Paint and Coating
Testing Manual 15th Edition of the Gardner-Sward Handbook
Joseph V. Koleske Editor
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
Paint and Coating Testing Manual Fifteenth Edition of the Gardner-Sward Handbook
Joseph V. Koleske, EDITOR ASTM Stock Number, MNL17-2ND
ASTM International 100 Barr Harbor Drive PO Box C700 West Conshohocken, PA 19428-2959 Printed in U.S.A.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
ii
Library of Congress Cataloging-in-Publication Data Paint and coating testing manual : 15th edition of the Gardner-Sward handbook / Joseph V. Koleske [editor]. p. cm. “ASTM Stock Number: MNL17-2nd” ISBN 978-0-8031-7017-9 1. Paint materials—Testing. 2. Paint materials—Analysis. I. Koleske, J. V., 1930TP936.5.P34 2011 667'.60284—dc23 2011034983 Copyright © 2012 ASTM International, West Conshohocken, PA. All rights reserved. This material may not be reproduced or copied, in whole or in part, in any printed, mechanical, electronic, film, or other distribution and storage media, without the written consent of the publisher. Photocopy Rights Authorization to photocopy items for internal, personal, or educational classroom use of specific clients is granted by ASTM International provided that the appropriate fee is paid to ASTM International, 100 Barr Harbor Drive, PO Box C700. West Conshohocken, PA 19428-2959, Tel: 610-832-9634; online: http://www.astm.org/copyright/ ASTM International is not responsible, as a body, for the statements and opinions advanced in the publication. ASTM does not endorse any products represented in this publication. Printed in Bridgeport, NJ January, 2012
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
iii
Contents Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .viii Part 1: Regulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Chapter 1—Regulation of Volatile Organic Compound Emissions from Paints and Coatings . . . . . . . . . . . . . . . . . . . . . 3 J. John Brezinski and Ronald K. Litton Part 2: Naturally Occurring Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Chapter 2—Bituminous Coatings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Ben J. Carlozzo Chapter 3—Cellulose Esters of Organic Acids. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Jos S. de Wit and Deep Bhattacharya Chapter 4—Drying Oils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Joseph V. Koleske Chapter 5— Driers and Metallic Soaps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 Marvin J. Schnall Part 3: Synthetic Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Chapter 6—Acrylic Polymers as Coatings Binders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 John M. Friel and Edwin Nungesser Chapter 7—Alkyds and Polyesters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 Dan Nelson Chapter 8—Amino Resins (Reaction Products of Melamine, Urea, etc., with Formaldehyde and Alcohols). . . . . . . . . 72 William Jacobs Chapter 9—Ceramic Coatings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 Richard A. Eppler Chapter 10—Epoxy Resins in Coatings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 Michael J. Watkins Chapter 11—Phenolics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 John D. Fisher Chapter 12—Polyamides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 Robert W. Kight Chapter 13—Polyurethane Coatings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 Han X. Xiao and Joseph V. Koleske Chapter 14—Silicone Coatings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 D. J. Petraitis Chapter 15—Vinyl Polymers for Coatings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 Joseph V. Koleske Chapter 16—Miscellaneous Materials and Coatings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 Joseph V. Koleske Part 4: Plasticizers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 Chapter 17—Plasticizers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 Peter Tan and Leonard G. Krauskopf Part 5: Solvents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 Chapter 18—Solvents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 Stephen A. Yuhas, Jr. and Rey G. Montemayor Part 6: Pigments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 Chapter 19—White Pigments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185 Juergen H. Braun Chapter 20—Black Pigments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204 Frank R. Spinelli Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
iv
CONTENTS
Chapter 21—Colored Organic Pigments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215 Paul Merchak Chapter 22—Inorganic Colored Pigments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234 Peter A. Lewis Chapter 23—Ceramic Pigments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239 Richard A. Eppler Chapter 24—Extender Pigments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242 Richard A. Eppler Chapter 25—Metallic Pigments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250 Russell L. Ferguson Chapter 26—Effect Pigments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256 Paul J. Nowak Chapter 27—Measurement of Gonioapparent Colors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272 Allan B. J. Rodrigues Chapter 28—Protective Coatings and Inorganic Anti-Corrosion Pigments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282 Lucien Veleva Chapter 29—Oil Absorption of Pigments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300 Charles W. Glancy Part 7: Additives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311 Chapter 30—Bactericides, Fungicides, and Algicides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313 Janet H. Woodward Chapter 31—Surfactants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320 Elvira Stesikova and Heinz Plaumann Chapter 32—Coalescing Aids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332 Kevin W. McCreight Chapter 33—Thickeners and Rheology Modifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341 Gregory D. Shay Part 8: Physical Characteristics of Liquid Paints and Coatings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373 Chapter 34—Density and Specific Gravity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375 Raymond D. Brockhaus and Ben J. Carlozzo Chapter 35—Characterizing Particle Size and Size Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389 George D. Mills Chapter 36—Rheology and Viscometry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415 Richard R. Eley Chapter 37—Surface Energetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 452 Gordon P. Bierwagen, Andrew Huovinen, and Bobbi Jo Merten Chapter 38—Solubility Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 470 Charles M. Hansen Part 9: Films for Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495 Chapter 39—Cure: The Process and Its Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 497 Thomas J. Miranda Chapter 40—Film Preparation for Coating Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505 Robert D. Athey, Jr. Chapter 41—Measurement of Film Thickness. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 514 John Fletcher and Joseph Walker Chapter 42—Drying Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 528 Thomas J. Sliva Part 10: Optical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 533 Chapter 43—Color and Light . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 535 Robert T. Marcus
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CONTENTS
v
Chapter 44—Gloss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 558 Gabriele Kigle-Böckler and Harry K. Hammond III Chapter 45—Hiding Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 569 Leonard Schaeffer Chapter 46—Mass Color and Tinting Strength of Pigments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591 Joseph V. Koleske Part 11: Physical and Mechanical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 599 Chapter 47—Adhesion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 601 Gordon L. Nelson Chapter 48—Abrasion Resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615 Daniel K. Slawson Chapter 49—Dynamic Mechanical and Tensile Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 624 Loren W. Hill Chapter 50—Flexibility and Toughness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 637 John Fletcher and Joseph Walker Chapter 51—Understanding Osmotic Activity in Paint Films and Determining Cause by Systematic Analysis of Blister Fluids and Blistered Coatings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 644 George Mills Chapter 52—Stress Phenomena in Organic Coatings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 655 Dan Y. Perera Chapter 53—Friction and Slip Resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 673 Joseph V. Koleske Part 12: Environmental Resistance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 685 Chapter 54—Prevention of Metal Corrosion with Protective Overlayers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 687 William H. Smyrl Chapter 55—Types of Metal Corrosion and Means of Corrosion Protective by Overlayers. . . . . . . . . . . . . . . . . . . . . . 697 Kenneth B. Tator and Cynthia L. O’Malley Chapter 56—Accelerated Weathering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 715 Valerie S. Sherbondy Chapter 57—Chemical Resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725 Latoska N. Price Chapter 58—Water-Resistance Testing of Coatings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 731 John Fletcher and Joseph Walker Part 13: Specific Product Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 737 Chapter 59—Aerospace and Aircraft Coatings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 739 Charles R. Hegedus, Stephen J. Spadafora, Anthony T. Eng, David F. Pulley, and Donald J. Hirst Chapter 60—Architectural Coatings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 751 Neal Rogers Chapter 61—Artists’ Paints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 765 Benjamin Gavett Chapter 62—Can Coatings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 770 Joseph V. Koleske Chapter 63—Testing of Industrial Maintenance Coatings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 778 Dwight G. Weldon Chapter 64—Pipeline Coatings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 787 Alfred Siegmund Chapter 65—Sealants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 792 Saul Spindel Chapter 66—Pavement Marking Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 799 James R. Swisher
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
vi
CONTENTS
Chapter 67—Water-Repellent Coatings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 807 Victoria Scarborough and Thomas J. Sliva Part 14: Analysis of Paint and Paint Defects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 811 Chapter 68—Analysis of Paint . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 813 Darlene Brezinski Chapter 69—The Analysis of Coatings Failures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 830 George D. Mills Part 15: Instrumental Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 849 Chapter 70—Atomic Absorption, Emission, and Inductively Coupled Plasma Spectroscopy . . . . . . . . . . . . . . . . . . . . . 851 Dwight G. Weldon Chapter 71—Chromatography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 856 Rolando C. Domingo and updated by Rey G. Montemayor Chapter 72—Electron Microscopy Overview with Coating Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 881 David R. Rothbard and John G. Sheehan Chapter 73—Infrared Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 895 Dwight G. Weldon Chapter 74—Methods for Polymer Molecular Weight Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 908 Thomas M. Schmitt Chapter 75—Ultraviolet/Visible Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 914 George D. Mills Chapter 76—X-Ray Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 920 A. Monroe Snider, Jr., Part 16: Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 941 Chapter 77—Paint and Coating Specifications and Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 943 Joseph V. Koleske Part 17: New Coating Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 949 Chapter 78—Radiation Curing of Coatings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 951 Joseph V. Koleske Chapter 79—Powder Coating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 957 Joseph V. Koleske Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 962 Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 965
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
vii
Preface For historical purposes, it is important to point out that at a January 1967 meeting of ASTM Committee D01 held in Washington, D.C., the American Society for Testing and Materials (ASTM International) accepted ownership of the Gardner-Sward Handbook from the Gardner Laboratory. It was through this laboratory that Dr. Henry A. Gardner published the previous twelve editions of the manual. Acceptance of this ownership gave ASTM an assumed responsibility for revising, editing, and publishing future editions of this well-known, respected manual. The undertaking was assigned to “Committee D01 on Paint and Related Coatings, Materials, and Applications.” This committee established a permanent subcommittee, “D01.19 on Gardner-Sward Handbook,” whose stated scope is delineated below. The 13th edition was published in 1972 as the Paint Testing Manual (STP 500) with Mr. G. G. Sward as editor and contributor. It was updated, expanded, and published in 1995 as the 14th edition, Paint and Coating Testing Manual (MNL 17) with Dr. Joseph V. Koleske as editor and contributor. The manual has served the industry well in the past by providing useful information that cannot be readily found elsewhere. It has been about fifteen years since the 14th edition was published. Interest in the manual has been strong through the years. This new edition of the Paint and Coating Testing Manual, the Fifteenth Edition of the Gardner-Sward Handbook (MNL 17), has been updated and expanded. The scope of the new edition is in keeping with the stated scope of Subcommittee D01.19: To provide technical, editorial, and general guidance for the preparation of the Fourteenth and subsequent editions of the Gardner-Sward Handbook. The handbook is intended for review of both new and experienced paint technologists and the past, present, and foreseeable trends in all kinds of testing within the scope of Committee D01. It supplements, but does not replace, the pertinent parts of the Society’s Book of Standards. It describes, briefly and critically all Test Methods believed to have significance in the world of paint technology, whether or not these tests have been adopted officially by the Society.
Once again, in this new edition, ASTM standard test methods, procedures, and other documents are described in minimal detail, with the various volumes of the ASTM Book of Standards remaining the primary source of such information. An effort was made to include references in the absence of ASTM documents concerning industrial, national, international, and other society test methods. The new edition contains either new chapters, or the previous topics/chapters in rewritten/revised form. In a few cases, the previous edition was merely updated, attesting to either the quality of the earlier writing, the lack of development in the area, or the apparent waning of interest in the topic. A variety of modern topics have been included. New chapters have been added as, for example, “Measurement of Gonioapparent Colors,” “Surfactants,” “Powder Coating,” and “Coalescing Aids.” As in the previous edition, individual authors, experts in their particular fields, were given a great deal of freedom in expressing information about their topics, but all chapters were subjected to peer review by two colleagues. Thus, style and content presentation may widely vary, but efforts were made to have understandable syntax and thus readers should find the information useful and “easy” to read and put to use. Manuals such at this one are prepared though a great deal of effort by the various authors and through the able assistance and behind-the-scenes concerted efforts of people such as Ms. Kathy Dernoga and Ms. Monica Siperko of ASTM International and Ms. Christine Urso, Ms. Barbara Carbonaro, Ms. Theresa Fucito, Ms. Patricia Mayhew, and Ms. Benita Hammer, of the American Institute of Physics, all of whom ensured that the manual was uniform in style and grammar and that manuscripts were submitted and processed in a timely fashion. The real unsung and unnamed contributors are the reviewers who gave encouragement to the various authors through constructive criticism, editorial information, and recommendations without deleteriously attempting to alter manuscripts from the author’s intent. To all of these people, a heart-felt “thank you.” Your talents have been utilized, you sacrificed much personal time, and you were patient with the numerous delays encountered on the road to making the manual a success. Joseph V. Koleske Editor
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
viii
Introduction PAST TO PRESENT
The previous edition of this manual, the 14th, described in detail the changes that took place in the coating industry from the early 1970s to the early- to mid-1990s. Published in 1995, the 14th edition classified powder coating, radiation-cured coatings, and higher-solids coatings as new, with a potentially reasonable growth curve. It noted that at the time, all liquid coatings were at higher solids content (lower volatile organic solvent content) than in the 1960s when Rule 66 came into being. Powder coating and radiation curing were sufficiently new enough that chapters related to testing them were not included in the manual. High-solids development still struggled with the difficulties of decreasing molecular weight for low viscosity purposes and achieving the low molecular weight with functionality on all molecules that quality coatings require. However, the solids level has increased in solvent-based coatings and achievements have been realized in decreasing volatile organic content (VOC). Since that time, powder coating has exploded. Today the technology is well established, has a significant share of the coatings market, is internationally accepted, and has a strong technical society that aids in future growth. To illustrate the widespread acceptance of powder coatings, one merely needs to look at advertisements. Outdoor metal furniture advertisements, for example, proudly include words that imply quality and durability—that is to say, “powdercoated finishes.” Of course, such furniture certainly is not the only commercial outlet for powder coatings. Applications include lighting fixtures, tubing and aerosol cans, automobile and bicycle wheels, rebars, store fixtures, agriculture and construction materials, and on and on. Initially, colors and color changeovers were considered to be a major obstacle to powder coating development, but today a broad variety of colors is available, including many metallic and special effect finishes with abrasion resistance, brilliance, and overall high quality. Powder coating provides quality, economy in manufacturing space, increased production, energy usage reduction, and other facets important to product development and sales in today’s marketplace. As with powder coating, radiation curing of coatings with either ultraviolet or electron beam radiation is no longer a new process. This technology also has been experiencing strong growth since the last edition of this manual. It is the technology in which, through an in situ means, a low viscosity liquid system is converted into a polymeric film or coating directly on a substrate that can be varied in nature—i.e., metal, wood, plastic, composite structures, etc. In effect, the originally liquid system is instantaneously converted into the final high molecular weight, cross-linked coating. Radiation curing of liquid systems is not limited to coatings, and it is growing in the printing ink and adhesive areas. It is considered to be “green” technology, is well established in the marketplace, has garnered a significant portion of the total coatings market, has a strong technical society dedicated to it, and is internationally accepted.
Radiation-curing technology has many facets that will ensure future growth. Harbourne1 has pointed out that over and above the usual advantages behind ultraviolet radiation curing technology—energy conservation, usage efficiency, and environmental conservation—its driving force is the fact that the UV process has enabled production and development of products that could not have been achieved with earlier existing technologies. Such products include flexible electronics for energy storage and circuit development, polymeric solar cells, printable electronics, medical devices, touch screens, optical films, and on and on. In the area of solar energy, highly efficient organic photovoltaic cells are being developed that are thinner and lighter in weight with significantly decreased production costs. Such cells are used in emergency power generation, lighting, and outdoor power generation. New smart materials with self-healing properties will provide overall cost savings through high value-added finishes on a variety of substrates. Solvent-based, high-solids coating systems continue to be developed. Such coatings have markedly decreased volatile organic content and provide high quality coatings and reduced environmental damage.
FUTURE
As described above, powder and radiation-cured coatings have been experiencing excellent growth over the past decade or so, with each technology growing on its own merits. More recently, a combination of the two technologies—UV–Curable Powder Coatings—has very good growth potential. New opportunities for the combination are due to the same benefits mentioned above—economic, environmental, process, energy savings, and increased productivity.2 The combination is meeting the less expensive, more rapid, and high quality challenges required by the demanding customers of today. The coatings are being used on medium-density fiberboard, plastics and other heat sensitive substrates, composites, and preassembled parts including completed items. Preassembled items often contain a number of different materials such as electronic components, gaskets, rubber seals, and the like--all of which are heat sensitive in nature. The ability to coat and cure such combinations with systems based on the combined technologies results in less thermal damage to the sensitive materials and thus greater efficiency and productivity along with cost savings. Nanotechnology is a field of emerging technology that may hold great promise in the future for the coatings, inks, and adhesives industry and certainly for a broad variety of other industries. Nanotechnology has broad implications
Harbourne, A. D. P., “The Evolution of UV Photopolymerization in Global Industrial Manufacturing Markets and the Promising Outlook for the Future of the Technology,” The 31st International Congress on Imaging Science, Beijing, China, pp. 013–015 (2010). 2 Schwarb, Ryan and Knoblauch, Michael, “New Opportunities for UV-Curable Powder Coatings, “ Coatings World, Volume 16, Number 5, pp 43-48 (May 2011). 1
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
INTRODUCTION
for new products and there are multi- and interdisciplinary efforts in progress. The technology deals with science on the nano, or one billionth-size, scale. Nanometer “particles” are 0.000000001 meter or 0.001 of a micrometer in size. Within the technology, an assembler or molecular manufacturing technique is used to position molecules through chemical reaction or interaction into new products or existing products with enhanced properties. Although the term “nanotechnology” was initially used to define efforts conducted on a molecular scale, currently the term has taken on a loose connotation for anything that is very small where small means something that is most usually smaller than a micrometer. Many examples of nanometer-designed products exist and a few of these are given below. Recently a plant was built of produce carbon nanotubes3. Such tubes in combination with aluminum result in new lightweight, high strength composite materials that have promise in the energy, electrical, and computer industries. In another area, a multilayered, polymeric nanocomposite has been devised and it is thought to have the potential to make a self-healing paint.4 In this technology, emulsion polymerization processes are used to develop a polymeric product that is covered with a silica-based layer of nanoparticles. Nanocomposite coatings for fabrics have also been described5. These coatings improve gas barrier properties as well as enhance mechanical characteristics. Another area that is receiving attention is additives for coating formulation. An additive that improves properties of water-based metal coatings has been described6. Although the additive is not chemically described, it is said to increase crosslink density and thereby various mechanical properties of cured films. An additive to accelerate the radiation-curing process is a small particle-sized version of nepheline syenite that is prepared by a micronizing process7. The micronized, ultra-fine form of this combination mineral—soda feldspar, potash feldspar, and nepheline--is said to enhance optical and physical performance in clear industrial and wood coatings. Properties such as gloss, Anon, “Bayer MaterialScience Builds Carbon Nanotubes Plant,” Paint and Coatings Industry, Volume 25, Number 11, p. 12 (Nov. 2009). 4 Colver, Patrick J., Colard, Catheline A. L., and Bon, Stefan A. F., “Multilayered Nanocomposite Polymer Colloids Using Emulsion Polymerization Stabilized Solid Particles,” J. American Chemical Society, Volume 150, No. 50, pp. 16850–16851 (2008). 5 Eberts, Kenneth, Ou, Runquing, and Shah, Kunal, “Nanocomposite Coatings for High-Performance Fabrics,” Paint and Coatings Industry, Volume 26, No. 4, pp. 32–36 (April 2010). 6 Herold, Marc, Burgard, Detlef, Steingrover, Klaus, and Pilotek, Steffen, “A Nanoparticle-based Additive for the Improvement of Water-Based Metal Coatings,“ Paint and Coatings Industry, Volume 16, Number 8, pp. 24–27 (Aug. 2010). 7 Van Remortel, Scott P. and Ratcliff, Robert E., “Ultrafine Nepheline Syenite as a Durable and Transparent Additive to Accelerate Radiation Cure,” Paint and Coating Industry, Volume 27, Number 3, pp. 27–34 (Mar. 2011).
ix
hardness, and scratch resistance are altered in a desirable manner. Cure rate via double bond conversion was enhanced in the presence of these very small mineral particles.
TESTING
As listed in Table 1, ASTM International has developed several documents that are useful in the area of nanotechnology. Although the documents are not necessarily directly related to coatings and paints, they provide useful background for investigators in this field and, as is apparent, useful guides for laboratory efforts in the areas of terminology, particle handling, effect of nanoparticles on red blood cells, particle mobility through a graduated index, and other areas. In the future, it is expected that this area will further develop within ASTM International. Joseph V. Koleske Editor
TABLE 1—ASTM Standard Documents Related to Nanotechnology ASTM Designation
Document Title
E2456-06
Terminology Relating to Nanotechnology
E2490-09
Standard Guide for Measurement of Particle Size Distribution of Nanomaterials in Suspension by Photon Correlation Spectroscopy (PCS)
E2524-08
Test Method for Analysis of Hemolytic Properties of Nanoparticles
E2525-08
Test Method for Evaluation of the Effect of Nanoparticulate Materials on the Formation of Mouse Granulocyte-Macrophage Colonies
E2526-08
Test Method for Evaluation of Cytotoxicity of Nanoparticulate Materials in Porcine Kidney Cells and Human Hepatocarcinoma Cells
E2530-06
Practice for Calibrating the Z-Magnification of an Atomic Force Microscope at Subnanometer Displacement Levels Using Si(III) Monatomic Steps
E2535-07
Guide for Handling Unbound Engineered Nanoscale Particles in Occupational Settings
E2578-07
Practice for Calculation of Mean Sizes/ Diameter and Standard Deviations of Particle Size Distributions
E2676-09
Practice for Tangible Property Mobility Index (MI)
3
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
Part 1: Regulations
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
1
MNL17-EB/Jan. 2012
Regulation of Volatile Organic Compound Emissions from Paints and Coatings J. John Brezinski1 and Ronald K. Litton2 INTRODUCTION
PRIOR TO THE 1960S, THE COATINGS INDUSTRY enjoyed a somewhat predictable regulatory and economic environment. The paint formulator selected solvents based on evaporation rate, solubility parameter, density, flammability, and, of course, cost. There was no apparent need to consider the relative photochemical reactivity of these materials, nor was there any appreciable incentive to reduce the solvent content of commercially acceptable coatings. It was recognized that objectionable odors were released from some paints and coatings. Further, air emissions resulting from the evaporation of solvents during high-temperature processing of oils and resins caused occasional complaints from persons living near the coatings plant. The prevailing view of this period was summarized by Francis Scofield in his article in the 13th edition of the Paint Testing Manual entitled “Atmospheric Pollutants” [1]. These “nuisance” types of pollution are a continuing problem but, in general, can be dealt with by dilution and dispersion of the objectionable materials to bring the concentration below a level that can be detected by the neighboring citizenry. Fortunately, most of the materials used by the paint industry are not toxic at concentrations significantly below the range at which they can be detected by the human nose, and sophisticated analytical procedures are rarely needed to deal with these “nuisance” problems. Since the 1960s, societal concern about health and the environment has increased appreciably. Actions taken by federal and state legislative bodies have resulted in a steady avalanche of new laws and associated regulations that affect virtually all of the chemical industry. Some of the federal laws administered by the U.S. Environmental Protection Agency (EPA) that impact the coatings industry are shown in Table 1. They are designed to control the emission of pollutants to air, to water, and to soil. In addition, among the new federal standards administered by the Occupational Safety and Health Administration are those that require manufacturers—including those making paints and coatings—to evaluate the hazards of products they make and to provide appropriate safety information to employees and users through the Material Safety Data Sheet (MSDS) and product labels:
1 2
Hazard Communication Standard (HCS), 1983 Occupational Exposure to Hazardous Chemicals in Laboratories, 1990 The discussion in this section will focus on the Clean Air Act and its amendments that, in the authors’ opinions, have had (and will continue to have) the greatest impact on coatings. However, it should be noted that regulatory activities in specific regions of the United States (for example, the state of California and the Ozone Transportation Commission— that includes 12 states in the U.S. Northeast plus the District of Columbia) have resulted in the development of VOC emissions rulings (see section on “Other Important U.S. Regulatory Activities”), which are more stringent than those enacted through EPA. In addition, other regions of the world are developing or have already enacted regulations addressing the emissions of volatile organic compounds. t t
THE CLEAN AIR ACT AND AMENDMENTS Photochemical Smog
A precipitating factor influencing the basis for selection of solvents for coatings in the 1960s and early 1970s was the recognition that solvents emitted to the atmosphere contributed to the growing “smog” problem, particularly in Southern California. The frequency of smog conditions in this area had increased steadily during the 1950s and 1960s as the number of automobiles, trucks, buses, and airplanes increased and as industrial development expanded with the accompanying growth of petroleum and chemical processing and power plant utilization.
Examples of Processes that Produce Hydrocarbons
Petroleum production, refining, transport Internal combustion engines Natural processes—forests and plants (isoprene and terpenes) t Surface coatings The smog problem was (and still is) very acute in the Los Angeles air basin, an area uniquely situated in a series of plains that originate in the high mountains to the east. The basin enjoys predominantly sunny days with cool, moist air flowing with a light westerly wind most of the year. These factors cause a nearly permanent temperature inversion layer, trapping air emissions that combine to produce persistent and eye-irritating smog in the basin. A summer t t t
Deceased, formerly of Hurricane, WV. Solvents Technical Service, Inc., 1015 Laurelwood Drive, Kingsport, TN 37660-8516.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed byASTM International This standard is for EDUCATIONAL USE ONLY. Copyright © 2012 by www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
3
4
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
TABLE 1—Federal Environmental Laws Administered by the U.S. Environmental Protection Agency. Law
Abbreviation
Clean Air Act, 1970
CAA
Amendments of 1977
CAAA-77
Amendments of 1990
CAAA-90
Clean Water Act of 1972
TABLE 2—Rule 66: Limits of Solvent Categories in Approved Mixtures.a 8%
20 %
Hydrocarbons,
Aromatic
Ethylbenzene, branched
alcohols,
hydrocarbons
ketones, toluene, or
Aldehydes,
(W/8 C atoms)
trichloroethane
esters, ethers or
CWA
ketones having
Amendments of 1977
an olefinic or
Safe Drinking Water Act, 1974
SDWA
Toxic Substances Control Act, 1975
TSCA
Resource Conservation and Recovery Act, 1980
RCRA
Comprehensive Environmental Response Compensation and Liability Act, 1980
CERCLA (Superfund)
Superfund Amendments and Reauthorization Act, 1986
SARA
Title III, Emergency Planning and Community SARA, Title III Right-toKnow, 1986
period in Los Angeles revealed that the use of organic solvents (for all purposes) accounted for about 18 % of the organic gases. About one half of the organic solvent emitted was attributed to the coatings industry, chiefly to the use in paint and coatings [1]. Based on the results of laboratory studies in “smog chambers,” in which a mixture of a solvent and nitrogen oxide was exposed for 6 h to light approximately the intensity of noon sunlight, the solvents could be classified as “low” or “high” in photochemical reactivity related to the amount of peroxides and ozone produced. These studies formed the basis for the well-known Rule 66, an air pollution control regulation passed by the Los Angeles Air Pollution Control District. Rule 66 identifies an “approved” solvent as one that contains less than 20 % by volume of specific chemicals and is further limited to certain combinations of these chemicals. Thus, approved solvents can contain no more than designated amounts of the combinations shown in Table 2. In effect, Rule 66 promoted the use of specific solvents such as aliphatic and naphthenic hydrocarbons, alcohols, esters, normal ketones, chlorinated hydrocarbons (except trichloroethylene), and nitroparaffins. Rule 66, superseded in 1976 by Rule 442, Usage of Solvents, by the California South Coast Air Quality Management District, was subsequently adopted by various other state jurisdictions.
VOC Definition
5%
The U.S. EPA was created in 1970 by Congress as part of a plan to consolidate several federal environmental activities. Studies directed by EPA laboratories in Research Triangle Park, NC, of the photochemical reactivity of materials in a laboratory smog chamber revealed that when organic materials and nitrogen oxide were irradiated for periods of up to 36 h, even those solvents considered acceptable under Rule 66 reacted to form peroxides and ozone. Only a few materials showed negligible photochemical reactivity,
Cycloolefinic Unsaturation a
Calculated as the percent by volume of the total solvent.
among which were: methane, ethane, methylene chloride, 1,1,1-trichloroethane, and fluorinated compounds. Since 1977, EPA has used the reactivity of ethane (based on a series of smog chamber experiments) as the benchmark for determining negligible reactivity. Compounds deemed less reactive than, or equally reactive to, ethane under the assumed conditions were classified as negligible. In contrast, compounds more reactive than ethane continued to be classified as reactive VOCs and were subject to appropriate control regulations. These studies, which were prompted in part by the passage of the Clean Air Act of 1970, led to the conclusion that most organic compounds emitted to the atmosphere contribute to the formation of ozone. On this basis, EPA adopted as a regulatory objective the limit of essentially all volatile organic compounds emitted to the atmosphere from all sources, including paint and coatings applications [2].
EPA Regulatory Definition of VOC
The regulatory definition of volatile organic compounds (VOC) was revised by EPA as of November, 2004. A part of this definition is as follows: 40 CFR Part 51 Section 51.100 Definitions3 Volatile organic compounds (VOC) means any compound of carbon, excluding carbon monoxide, carbon dioxide, carbonic acid, metallic carbides or carbonates, and ammonium carbonate, which participates in atmospheric photochemical reactions 1. This includes any such organic compound other than those which have been determined to have negligible photochemical reactivity. The original list is: methane; ethane; methylene chloride (dichloromethane); 1,1,1-trichloroethane (methyl chloroform); 1,1,2-trichloro-1,2,2-trifluoroethane (CFC-113); trichlorofluoromethane (CFC-11); dichlorodifluoromethane (CFC12); chlorodifluoromethane (CFC-22); trifluoromethane (HFC-23); 1,2-dichloro-1,1,2,2-tetrafluoroethane (CFC114); chloropentafluoroethane (CFC-115); 1,1,1-trifluoro 2,2-dichloroethane (HCFC-123); 1,1,1,2-tetrafluoroethane (HF-134a); 1,1-dichloro 1-fluoroethane Code of Federal Regulations (CFR) Part 51; Requirements for Preparation, Adoption and Submittal of Implementation Plans, Federal Register, Volume 2, 1 July 2002, pp. 131–136.
3
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 1
2.
3.
4.
5.
Q
REGULATION OF VOLATILE ORGANIC COMPOUND EMISSIONS
(HCFC-141b); 1-chloro 1,1-difluoroethane (HCFC142b); 2-chloro-1,1,1,2-tetrafluoroethane (HCFC-124); pentafluoroethane (HFC-125); 1,1,2,2-tetrafluoroethane (HFC-134); 1,1,1-trifluoroethane (HFC-143a); 1,1-difluoroethane (HFC-152a); parachlorobenzotrifluoride (PCBTF); cyclic, branched, or linear completely methylated siloxanes; acetone; perchloroethylene (tetrachloroethylene); 3,3-dichloro-1,1,1,2,2-pentafluoropropane (HCFC-225ca); 1,3-dichloro-1,1,2,2,3-pentafluoropropane (HCFC-225cb); 1,1,1,2,3,4,4,5,5,5-decafluoropentane (HFC 43-10mee); difluoromethane (HFC-32); ethylfluoride (HFC-161); 1,1,1,3,3,3-hexafluoropropane (HFC-236fa); 1,1,2,2,3-pentafluoropropane (HFC-245ca); 1,1,2,3,3-pentafluoropropane (HFC-245ea); 1,1,1,2,3-pentafluoropropane (HFC-245eb); 1,1,1,3,3-pentafluoropropane (HFC245fa); 1,1,1,2, 3,3-hexafluoropropane (HFC-236ea) 1,1,1,3,3-pentafluorobutane (HFC-365mfc); chlorofluoromethane (HCFC-31);1 chloro-1-fluoroethane (HCFC151a); 1,2-dichloro-1,1,2-trifluoroethane (HCFC123a); 1,1,1,2,2,3,3,4,4-nonafluoro-4-methoxy-butane (C4F9OCH3);2-(difluoromethoxymethyl)-1,1,1,2,3,3,3heptafluoropropane ((CF3)2CFCF2O CH3); 1-ethoxy1,1,2,2,3,3,4,4,4-nonafluorobutane (C4F9OC2H5);2(ethoxydifluoromethyl)-1,1,1,2,3,3,3-heptafluoropropane ((CF3)2CFCF2OC2H5); methyl acetate and perfluorocarbon compounds which fall into these classes: i. Cyclic, branched, or linear, completely fluorinated alkanes; ii. Cyclic, branched, or linear, completely fluorinated ethers with no unsaturations; iii. Cyclic, branched, or linear, completely fluorinated tertiary amines with no unsaturations; and iv. Sulfur-containing perfluorocarbons with no unsaturations and with sulfur bonds only to carbon and fluorine. For purposes of determining compliance with emissions limits—VOC will be measured by the test methods in the approved State implementation plan (SIP) or 40 CFR part 60, Appendix A, as applicable. Where such a method also measures compounds with negligible photochemical reactivity, these negligibility-reactive compounds may be excluded as VOC if the amount of such compounds is accurately quantified and such exclusion is approved by the enforcement authority. As a precondition to excluding these compounds as VOC or at any time thereafter, the enforcement authority may require an owner or operator to provide monitoring or testing methods and results demonstrating, to the satisfaction of the enforcement authority, the amount of negligibly-reactive compounds in the source’s emissions. For purposes of Federal enforcement for a specific source, EPA shall use the test methods specified in the applicable EPA-approved SIP in a permit issued pursuant to a program approved or promulgated under title V of the Act, or under 40 CFR part 51, subpart I or Appendix S, or under 40 CFR parts 52 or 60. EPA will not be bound by any State determination as to appropriate methods for testing or monitoring negligibly-reactive compounds if such determination is not reflected in any of the above provisions. The following compound(s) are VOC for purposes of all recordkeeping, emissions reporting, photochemical
5
dispersion modeling and inventory requirements which apply to VOC and shall be uniquely identified in emission reports, but are not VOC for the purpose of VOC emissions limitations or VOC content requirement: t-butyl acetate Note—The category of “VOC-exempt compounds” can be modified (i.e., add compounds to or delete them from the list) by EPA. Since the development of the initial list, numerous petitions requesting “VOC exemption” on specific compounds were submitted to EPA by various companies and trade associations. The petitions submitted requested that compound A be exempted from VOC control based on its low reactivity relative to ethane. As a result of those initiatives, several solvents were exempted by EPA. This list includes: Methyl Formate (2004) t-Butyl Acetate (2004) Propylene Carbonate (2009) Dimethyl Carbonate (2009) The majority of VOC exempt solvents in the initial list were not useful in formulating coatings with good solubility and application characteristics. The delisting of acetone, parachlorobenzotrifluoride, volatile methyl siloxanes (VMS), and methyl acetate in the mid-late ′90s (coupled with the recent delisting of the 4 aforementioned solvents) has provided coatings formulators with greater latitude in developing lower VOC coatings. States may also have their own list of VOC-exempt compounds. Although state lists are often modeled after EPA definition, users of coating products should confirm that a solvent deemed VOC exempt by the agency is classified similarly by the respective state.
Metrics for Defining “Negligible Photochemical Reactivity”
In an effort to define chemicals as having “negligible photochemical reactivity” (and, therefore, exempt from VOC regulations) EPA designated ethane as the benchmark for separating reactive from negligibly reactive compounds under the assumed conditions. Prior to 1994, EPA had only granted VOC exemptions based on the metric known as the kOH value. This value represents the molar rate constant for reactions between the given compound and the OH radical in the air. If the kOH value of a compound is less than ethane, the compound may be less reactive than ethane and may be declared to be “negligibly reactive.” In the mid-1990s, in response to a petition for VOC exemption, EPA used another type of comparison to ethane based on the ozone forming potential of other reactions of the compound in addition to the initial reaction with the OH radical. This method was based on the concept of maximum incremental reactivity (MIR). The MIR values are usually expressed either as grams of ozone formed per mole of VOC (molar basis) or as grams of ozone formed per gram of VOC (mass basis). Both metrics were used to exempt two compounds in the 1990s. Thus, since 1997, EPA has considered three different metrics to compare the reactivity of a specific compound to that of ethane: 1. the reaction rate constant with the hydroxyl radical (known as kOH) 2. MIR expressed as reactivity per gram basis 3. MIR expressed as reactivity per mole basis
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
6
PAINT AND COATING TESTING MANUAL
Q
Even though much debate and critique has ensued on the merits of a mass versus mole comparison, in “Interim Guidance on Control of Volatile Organic Compounds in Ozone State Implementation Plans” published on September 13, 2005 (70FR 54046), EPA stated: “A comparison to ethane on a mass basis strikes the right balance between a threshold that is low enough to capture compounds that significantly affect ozone concentrations and a threshold that is high enough to exempt compounds that may usefully substitute for more highly reactive compounds. . . . When reviewing compounds that have been suggested for VOC exempt status, EPA will continue to compare them to ethane using kOH expressed on a molar basis and MIR values expressed on a mass basis.” The MIR values (both by mole and mass) are developed measures of photochemical reactivity derived from a computer-based photochemical model. This concept, which demonstrates that VOCs have different reactivity, i.e., different ozone-forming potential, is now shaping revised ozone control strategies at both the federal and state levels (view section on “Alternative Concept for Controlling Ozone Formation”).
VOC and Ozone Formation
The understanding of “photochemical smog” and its contribution to the formation of ground level ozone has increased dramatically since the days of Rule 66. Basically, VOCs react with oxides of nitrogen (NOX) in the presence of heat and sunlight to form ground level. However, the interrelationship of VOCs and NOX in ozone formation is a complex series of reactions. One suggested pathway is depicted below [3]. NO2 + Sunlight → NO + O O + O2 → O3 O3 + NO → NO2 + O2 VOCs + Sunlight → Radicals NO + Radicals → NO2 Primary Sources of NOx: On-road mobile sources, electricity generating units, and nonroad mobile sources. Other articles have been published on factors affecting photochemical ozone formation and the potential of a given compound to contribute to ozone formation in the troposphere (lower atmosphere) [4,5]. In addition to the concentration of NO2 and available sunlight in the atmosphere, ozone formation is also affected by temperature, humidity, as well as the concentration and composition of other VOCs present in the atmosphere. Also, substantial biogenic (natural) VOCs released from trees and other green plants can impact ozone formation at ground level. Considerable interest has developed recently in the consideration of individual solvent photochemical reactivity in state, federal, and international programs related to air quality control (see section on “New Concept for Controlling Ozone Formation”).
Other VOC Definitions
The U.S. EPA defines a category of VOC-exempt compounds (as described in the section on “EPA Regulatory Definition of VOC”). However, for many consumer products (i.e., floor polishes, glass cleaners, automotive rubbing or
15TH EDITION
polishing compounds, etc.), EPA has determined that low volatility compounds may also be exempted from regulations as VOCs. These include solvents: Having a vapor pressure (VP) < 0.1 mm Hg at 20°C; or Consisting of > 12 carbon atoms, if the VP is unknown; or Having a MP > 20°C, and which does not sublime (i.e., does not change directly from solid into a gas without melting), if the VP is unknown. Currently, there is no explicit LVP (low vapor pressure) exemption for products used in paints and coatings. Thus, some LVP products are regulated as VOCs in coatings but not when used in consumer products. Lastly, a wide disparity in the definition of a VOC exists across countries. A product classified as a non-VOC in the Unites States does not automatically receive the same classification in Europe. For example, the criteria for the EU eco-labeling scheme is that a VOC is “any organic compound with, at normal conditions for pressure, a boiling point (or initial boiling point) lower than or equal to 250°C” [ref—Official Journal 39 (L4), 6 January (1996)]. Thus, coalescing aids for latex paints with a boiling point (BP) >250°C would not be a VOC under the EU ecolabeling definition. Other countries are becoming more proactive in developing rules and definitions addressing VOC issues.
The Ozone Standard
The Clean Air Act of 1970 targeted six criteria pollutants for control: carbon monoxide, lead, nitrogen dioxide, ozone, particulates, and sulfur dioxide. Criteria pollutants are those for which criteria were issued by EPA. These documents include national ambient air quality standards (NAAQS) for each criteria pollutant—levels that protect against adverse effects to health and to plants and materials [6]. These criteria documents and standards were to be reviewed every 5 years and if necessary, revised to assure that the standards provided adequate health protection. Criteria pollutants are measured using a network of monitors nationwide. Standards for ozone and nitrogen oxides are: Ozone: In July 1997, the U.S. EPA revised its NAAQS for ground-level ozone and particulate matter. The complete standards appeared in the July 18, 1997, edition of the Federal Register (pp. 38652–38896). The EPA began phasing out and replacing the previous 1-hour standard (last revised in 1979) with a new 8-hour standard. The new standard would be 0.08 ppm measured over 8 hours, replacing the old standard of 0.12 ppm measured over 1 h. In establishing the 8-hour standard, the EPA defined the new standard as a “concentration based” form, and it called for measuring the 3-year average of the annual 4th-highest daily maximum 8-hour ozone concentration. On June 20, 2007, the EPA administrator signed proposed revisions to the NAAQS for ozone. The end result of the proposal was the 8-hour “primary” standard for O3 set at 0.075 ppm in the March, 2008 final rule. On January 6, 2010, the EPA proposed to strengthen the NAAQS for ozone. This provision proposed that the 8-h “primary” standard for ozone be set at a lower level within the range of 0.060–0.070 ppm. This action would provide increased protection for children and other “at risk” populations against the adverse health effects related to ozone exposure. According to the publication “Ozone Air Quality
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 1
Q
REGULATION OF VOLATILE ORGANIC COMPOUND EMISSIONS
Standards: EPA’s Proposed January 2010 Revisions” (following public hearings and comments)—a final rule was scheduled to be issued by August 31, 2010. The Clean Air Act requires that the EPA set a standard based on the need to protect public health “with an adequate margin of safety.” For policy guidance, timelines, recent actions, etc., on the new ozone standard, see the following: www.epa.gov/air/ozonepollution/. Note—Primary standards set limits to protect public health, including the “health” of sensitive populations such as asthmatics, children, and the elderly. The implementation of a new ozone standard is of vital interest to the paint industry since the CAAA-90 specifically identifies paint and coatings for various controls for VOC emission reductions, one precursor for ozone formation. Under the new standard, there could be additional ozone nonattainment areas that will increase the need for compliant coating technology. Nitrogen Dioxide: The nitrogen dioxide concentration in the atmosphere cannot exceed 0.053 ppm as the annual arithmetic mean concentration. In 2010, the EPA supplemented the existing annual standard for NO2 by establishing new short-time standard based on the 3-year average of the 98th percentile of the yearly distribution of 1-hour daily maximum. EPA set the level of this new standard at 100 ppb. Final rule is listed at: http://www.regulations.gov. As of February, 2010, there are no areas in the United States that are designated as nonattainment of the NOx, NAAQS. Although no change has occurred in NAAQS for nitrogen dioxide, decreasing NOx emissions [various nitrogen compounds like nitrogen dioxide (NO2) and nitric oxide (NO)] are receiving more attention from the EPA in an effort to minimize ground level ozone, which is formed when NOx and VOCs react in the presence of heat and sunlight. This is an important shift in strategy for controlling ozone formation since it recognizes that in geographical regions in which the ratio of VOCs to NOx is high (“NOx -limited conditions”), additional reductions in VOC emissions will have a minimal impact on air quality.
CONTROL OF VOC EMISSIONS FROM COATINGS
The Clean Air Act addresses air pollution emanating from both existing sources and that from future new plant construction or significant modification of existing sources. States with areas that did not comply with the ozone standard were given primary responsibility to develop appropriate regulations for existing sources to meet the time schedule for compliance specified by Congress. The plan outlines the measures that the state will take in order to improve air quality. The Federal EPA was assigned oversight responsibility for the state programs that were described in “State Implementation Plans” (SIPs). The plan outlines the measures that the state will take in order to improve air quality. No SIP can mandate weaker pollution controls than those established by the EPA. The SIP is reviewed by the EPA and if deemed unacceptable, the EPA must prepare one for it. For example, failure by a state to submit an adequate SIP can result in restrictions on federal highway funds.
Control Technique Guidelines
In 1977, the Agency issued the first of a series of guidance documents for the states related to various industrial coat-
7
ing operations or end-use categories. These documents, called “Control Technique Guidelines (CTG) Series, Control of Volatile Organic Emissions from Stationary Sources,” include recommended VOC emission limits, based on the EPA’s assessment of Reasonably Available Control Technology (RACT): the limits are expressed as pounds of VOC per gallon of coating (minus water), as applied. The EPA has defined RACT as the lowest emission limit that a source can meet by the application of control technology that is reasonably available considering technological and economic feasibility. The Clean Air Act Amendments of 1977 directed that states had to revise their implementation plans for areas out of compliance with the national ozone standard. The revised SIPs were to include sufficient control of VOC emissions from stationary sources (buildings, structures, facilities, and installations), such controls to incorporate the RACT limits for coatings operations for which a CTG was published. Note—CTGs are not federal regulations. However, they do provide federally prescribed control measures to be incorporated as a part of approved SIPs. Under the CAAA-90, the EPA scheduled issuance of CTG documents for 29 categories of VOC sources, which covered a variety of surface coating operations. Section 183 of the amended act also requires that the EPA issue CTGs for an additional number of surface coatings processes by 1993. In addition, the EPA planned to integrate HAP (Hazardous Air Pollutant) rulemaking with VOC requirements under 183 (e). For the industry source categories, the approach to reduce VOC emissions has been for applicators to either adopt an alternate coating technology (high-solids, waterborne, powder, UV cure), or install engineering controls (carbon adsorption, incineration, etc.).
NEW SOURCE PERFORMANCE STANDARDS
The control of VOC emissions from new coatings plants and from significant modifications of existing plants was addressed by the EPA in a series of New Source Performance Standards (NSPS), the first of which issued in 1980. These mandatory standards, which apply uniformly to all parts of the country, define the emission sources more narrowly and impose a tighter level of emission control than that for related existing sources. Facilities that are constructed, modified, or reconstructed after the NPSP was proposed by the EPA are subject to NSPS. The VOC limits defined in the NSPS, expressed as kilograms of VOC per liter of applied solids, are based on the best demonstrated technology (BDT) for the specific coating operation. The NSPS requirements can be found in the Code of Federal Regulations at Title 40 (Protection of Environment), Part 60 (Standards of Performance for New Stationary Sources): http://ecfr.gpoaccess.gov/cgi/tZtext/ text-idx?sid=474f779beade290997e4611971d078f4&c=ecfr &tpl=/ecfrbrowse/Title40/40tab_02.tpl. The emission limits in both the CTG and NSPS documents, in the majority of cases, focus on restricting the VOC content per unit of coating or of coating solids applied in the operation, rather than placing a ceiling on individual plant emissions. The responsibility for establishing emission limits for particular plants, if appropriate, was left to the states [7].
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
8
PAINT AND COATING TESTING MANUAL
Q
TABLE 3—Clean Air Act Amendments—1990 major source identification based on VOC I emissions; limits for area classifications Ozone Nonattainment Area Classification
VOC Emission Limit, tons/year
Marginal or moderate
100
Serious
50
Severe
25
Extreme
10
DETERMINATION OF VOC CONTENT EPA Federal Reference Method 24
The procedures specified by the federal EPA for testing paint products for compliance with VOC limits are described in Federal Reference Method 24: Determination of Volatile Matter Content, Density, Volume Solids, and Weight Solids of Surface Coatings [8]. This standard employs several ASTM test standards, including those shown in Table 3. In addition, Reference Method 24A: Determination of Volatile Matter Content and Density of Publication Rotogravure Inks and Related Publication Rotogravure Coatings incorporates these ASTM standards. These two documents provide the framework for referencing these ASTM standards in determining VOC levels under regulation. Method D2369 is a key procedure of Federal Method 24. Since 1980, several important revisions have been made in this standard to make it compatible with revisions in Method 24, including the addition in 1990 of instructions for testing multicomponent coatings and the deletion of sections dealing with testing at shorter times. The revised version of Federal Reference Method 24 is also included in the ASTM Manual on Determination of Volatile Organic Compound (VOC) Content in Paints, Inks, and Related Coating Products, 2nd ed., 1993 [9]. Substantial revisions were also made in ASTM D3960, Practice for Determining Volatile Organic Compound (VOC) Content of Paints and Related Coatings, a standard developed in ASTM Subcommittee D01.21 to provide a guide for the calculation of VOC and to establish a base for the investigation in ASTM of the precision of VOC content determination. The definitions and symbols used in D3960 are those adopted by the EPA and included in the Agency document “Procedures for Certifying Quantity of Volatile Organic Compounds Emitted by Paint, Ink and Other Coatings” that was published in 1984 [10]. Studies and discussions in ASTM Subcommittee D01.21 that led to the modification and improvements of ASTM standards referenced in Federal Method 24 and in ASTM Practice D3960 were conducted with the cooperation of EPA personnel of the Office of Air Quality Standards Development at Research Triangle Park, NC. EPA Method 24 was designed to be used for measuring the VOC content of all coatings that are intended for either ambient or baking film conditions. However, that method was not applicable to ultraviolet (UV) radiationcured coatings. Therefore, the method was subsequently amended to incorporate ASTM Method D5403-93, which does contain those procedures. The test methods in
15TH EDITION
D5403 determine the weight percent volatile content of paint, coatings, and inks that are designed to be cured by exposure to ultraviolet light or to a beam of accelerated electrons. After radiation cure, the specimens are baked at 110±5°C for 60 min. The general expression for calculating VOC content [ASTM D5201-05a (2010) Standard Practice for Calculating Formulation Physical Constants of Paint (Physical Constants of Paint and Coatings)] is available from ASTM International at the following link: http://www.astm.org/ Standards/D5201.htm. The expression “VOC” includes all organic emissions from a coating, not just the solvent in a coating. These emissions can include volatile additives, by-products of the cure reaction, etc. “Formulation” VOC content may or may not be an acceptable means for compliance, depending on the specific wording of the applicable regulation. It would be acceptable if the same VOC content is obtained when tested using EPA Method 24. The EPA would have preferred to limit VOC emissions in the CTG on the basis of the unit volume of coating solids applied. However, the general expression for calculating VOC content in which VOC content is defined as mass per unit volume of coating less water and less exempt solvent was necessary as no acceptable consensus procedure was available for determining the volume percent nonvolatile content. In a presentation in Copenhagen in 1990, James C. Berry of the U.S. EPA stated: “Though certainly less than ideal, the major attraction is that the expression permits the determination of compliance from the analysis of a coating sample obtained during a plant inspection. In the simplest case, these units require only one volumetric and one gravimetric measurement” [7]. The measurement of low VOC content waterborne coatings (e.g., architectural) using EPA Reference Method 24 was found to be unreliable (confirmed by industry round-robin lab trials). Studies demonstrated that the error in VOC measurements in waterborne coatings was inversely proportional to the VOC content of the coating (i.e., the lower the VOC content, the poorer the precision to be expected). As industry moved toward lower VOC coatings for compliance purposes, it was prudent that a more reliable test method be found that more accurately quantified VOC content in waterborne coatings. After much stakeholder corroboration, the following test was developed: ASTM D6886-03 “Speciation of the Volatile Organic Compounds (VOCs) in Low VOC content Waterborne Air-Dry Coatings by Gas Chromatography,” ASTM International. The method was designed primarily for the analysis of waterborne coatings in which the material VOC content is NH) and methylol (>NCH2OH) groups on the amino resin, which can form strong hydrogen bonds with unshared electrons on nitrogen and oxygen. The high-solids amino resins have much higher levels of combined formaldehyde than the conventional solids resins. Typical values for combined formaldehyde are in the range 2.0–2.7 for urea resins and 3.5–6.0 for melamine resins. The etherifying alcohol is most often methanol, although resins made with both methanol and butanol or even butanol alone are also widely used. These resins are less polymeric, with DP NCH2OH groups, and (2) the solvent is not low boiling. Apparently, the increase in viscosity as solvent evaporates slows the diffusion rate and effectively prevents complete removal of solvent within the time frame of the test. There may also be a hydrogen-bonding effect between solvent and resin that contributes to the retention of solvent. Other test methods involve much higher temperatures, where resin condensation/degradation does occur. One standard method is the ASTM Test Methods for Volatile Content of Coatings (ASTM D2369), where a small resin sample (0.3–0.5 g) is diluted with xylene and placed in a 110°C oven for 1 h. There are a number of other, similar tests.
Viscosity Measurement
Amino resin viscosities are most commonly measured by the Gardner bubble viscometer method. This method is similar to the Test Method for Viscosity of Transparent Liquids by Bubble Time Method (ASTM D1545). A tube containing the resin under test is placed in a rack containing reference tubes of known viscosity. The tubes are equilibrated to 25°C in a constant temperature bath. The rack is quickly inverted, and the rate of rise of an air bubble in the sample tube is compared against similar bubbles in the reference tubes. The reference tubes are letter graded A–Z and Z1–Z6.
Solvent Tolerance
There are a number of different solvent tolerance tests. All involve titrating a weighed sample of the amino resin with a standard reagent (solvent). The object of the test is to measure how much of the reagent the amino resin can accept before the solution turns cloudy/milky. Results are typically reported in milliliters of reagent per gram of sample. Typical reagents used include xylene, iso-octane, and the isooctane/decahydronaphthalene/toluene mixture described in ASTM Test Method for Solvent Tolerance of Amine Resins (D1198), which was withdrawn (with no replacement) in March 2007 in accordance with section 10.5.3.1 of the Regulations Governing ASTM Technical Committees. While the immediate objective of the solvent tolerance test is to determine the amount of reagent that the amino
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 8
resin can accept before solution clouding occurs, the real purpose of the test is to gain insight into the structure and composition of the resin and hence have a better understanding of how it will perform in a given coating application. In general, amino resins of high molecular weight, or having high levels of polar functional groups, i.e., >NH and >NCH2OH, will have limited compatibility with the typical hydrocarbons used and hence will give low tolerance test results. Experience shows that a low tolerance value means a faster curing resin and vice versa, especially in the absence of acid catalyst. However, although the tolerance test represents a quick and easy way to measure potential cure response, it does not uniquely define the resin structure. Thus, a low tolerance reading can be caused by either high polarity or high molecular weight or both.
Size Exclusion, High-Performance Liquid Chromatography, and Mass Spectrometry
To obtain more detailed knowledge of resin structure, amino chemists now rely very heavily on gel permeation or size exclusion chromatography (SEC) and on highperformance liquid chromatography (HPLC) sometimes coupled with mass spectrometry (Mass Spec). The size exclusion chromatograph provides an excellent measure of number and weight-average molecular weight and molecular weight distribution (polydispersity), while HPLC, which fractionates the resin components primarily by functional groups, provides information on resin composition, especially if coupled with Mass Spec. Typically the more polar species are eluted first, followed by the less polar fractions. Thus, taken together, SEC and HPLC provide detailed information on molecular weight and functionality, which cannot be directly obtained or inferred from any of the various solvent tolerance tests. Size exclusion and liquid chromatograms for a representative commercial high-solids methylated melamine resin are shown in Figs. 4 and 5. More recently the combined technique of liquid chromatography and mass spectrometry has been used to obtain very detailed knowledge of amino resin structure. Advances in ionization techniques have resulted in mass spectra capable of discerning individual components in even the higher molecular weight oligomeric portions in amino resin compositions. This unique ability to discern numerous individual components and their concentration
Q
AMINO RESINS
75
Fig. 5—High-performance liquid chromatogram of a typical high-solids methylated melamine resin.
by LC-Mass Spec techniques is described in various publications by Chang [13].
Combining Ratios
Amino resins may also be characterized by measurement of the amounts of formaldehyde and alcohol that have reacted. For example, see hexa(methoxymethyl) melamine (HMMM) (Fig. 6), which has exactly 6 mol each of combined formaldehyde and methanol per mole of melamine. Unlike HMMM, most resins are, of course, mixtures of products, which are best described by an average composition. One of the most widely sold commercial high-solids methylated melamine resins has an average combining ratio melamine/formaldehyde/methanol of about 1/5.6/5.1. Because methanol reacts with an already-reacted formaldehyde molecule, a resin can never have combined methanol greater than the combined formaldehyde. The excess formaldehyde, 0.5 mol in the commercial example, represents formaldehyde that has not reacted with methanol and which must therefore be present as methylol (>NCH2OH), bridging groups (>NCH2NNCH2OCH2NNCH2OCH2OCH3>). Acetals are formed when an excess of formaldehyde is used in the synthesis. They are therefore present in many high-solids amino resins. Determination of combining ratios may be done most easily by either 1H or 13C NMR techniques [9,10]. Older methods involve complete hydrolysis of the resin to the starting materials, followed by wet-chemical analysis for nitrogen and formaldehyde and gas chromatographic determination of alcohol(methanol or butanol).
Free Formaldehyde
Amino resins always contain some unreacted formaldehyde, usually referred to in product specifications as “free” formaldehyde. Free formaldehyde may be analyzed quantita-
Fig. 4—Size exclusion chromatogram of a typical high-solids methylated melamine resin.
Fig. 6—Hexa(methoxymethyl)melamine.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
76
PAINT AND COATING TESTING MANUAL
Q
tively by a number of methods. One of the most commonly used is the sodium sulfite method [14]. Formaldehyde reacts rapidly and completely with aqueous sodium sulfite to form a bisulfite addition complex. Sodium hydroxide is liberated quantitatively on a mole-for-mole basis. CH 2 O + Na 2SO3 + H 2 O → NaOH + HOCH 2SO2 Na
(7)
The NaOH is either titrated directly with a standard HCl solution, or neutralized with a known excess of standard HCl, which is then back-titrated with NaOH. Care must be taken to ensure that reacted formaldehyde, particularly methylol groups, is not analyzed as free formaldehyde. This can occur because of the following reaction, which can be minimized by performing the titration as rapidly as possible at cool temperatures, e.g., room temperature or lower. > NCH 2 OH →> NH + CH 2 O
(8)
PHYSICAL PROPERTIES General
Amino resins are typically viscous liquids with an aminelike odor. Depending on composition, they may also smell of formaldehyde and/or solvent. They are readily soluble in alcohols, ketones, hydroxyl-functional glycol ethers, esters, etc., but have limited solubility in hydrocarbons. Some resins, especially methylol-rich resins with low levels of both combined formaldehyde and combined methanol, are water soluble. Many more are water reducible in the presence of other solvents, e.g., alcohols and glycol ethers. Because of their resinous nature, aminos have neither a well-defined freezing point nor boiling point. Uncured resins typically have glass transition temperatures around −40°C. When heated at temperatures above about 140°C, some aminos, especially urea resins, may undergo decomposition with release of formaldehyde and alcohol. This tendency to decompose causes difficulties in determining the solids content of resin solutions, as described in Analysis/ Analytical Methods. The problem is particularly acute with resins having high methylol functionality.
15TH EDITION
Surface Tension
The surface tension of amino resins is quite strongly related to the nature of the etherifying alcohol and is much less affected by the level of combined formaldehyde and alcohol. Surface tension measurements on high-solids, solvent-free resins using a DeNouy tensiometer have given values ranging from about 45 dynes/cm for methylated resins to about 28 dynes/cm for butylated resins [16]. Mixed methyl/butyl resins give intermediate values, depending on the methyl and butyl content. The reduction in surface tension when butanol is the etherifying alcohol may be one reason that high-solids butyl and methyl/butyl resins provide improved flow and leveling in high-solids formulations compared to their fully methylated counterparts.
REACTIONS OF AMINOS IN COATINGS Cure Reactions
Amino resins in coating formulations cure by reactions that are chemically and mechanistically similar to those that take place during synthesis of the resin. The principal reaction of cure is one of trans-etherification, wherein a hydroxyl group on the primary film-former (acrylic, polyester, or alkyd) reacts with an alkoxymethyl group on the amino resin > NCH 2 OR + HO − A →> NCH 2 O − A + ROH where: R = alkyl and A = primary film-former. Additionally, direct etherification may take place, the end result being the same > NCH 2 OH + HO − A →> NCH 2 O − A + H 2 O
(10)
where: A = primary film-former These two reactions both result in chemical bond formation between the amino and the primary film-former (co-condensation). Two other reactions may also take place, both of which involve reaction of the amino resin with itself (self-condensation). These are > NCH 2 OR + HN NCH 2 N < + ROH
Viscosity
The viscosity of an amino resin is a function of (1) polymer content (degree of polymerization) and (2) the nature of its functional groups. The latter may be a more important contributor to viscosity than the former. Amino resins are not generally very polymeric, especially in comparison with other coating resins, e.g., polyesters, alkyds, and acrylics. Typically, average degrees of polymerization are in the range of 1–5. High-molecular-weight “tails” increase viscosity significantly. Because of strong hydrogen bonding, resins carrying significant amounts of >NH and >NCH2OH functionality are quite viscous, even though they may not be highly polymerized. There is a marked drop in viscosity when amino resins are diluted with solvent, largely due to breaking of hydrogen bonds. Good solvents (e.g., alcohols) are more effective at reducing viscosity than poor ones [15]. Methanol is probably the best, although it is not widely used because of its low boiling point. Isopropanol is almost as effective, and because it is somewhat higher boiling, represents a good compromise.
(9)
(11)
where: R = H, alkyl > NCH 2 OH + HOCH 2 N NCH 2 OCH 2 N < + H 2 O
(12)
Besides the co-condensation and self-condensation reactions, hydrolysis and deformylation reactions may also occur:
> NCH 2 OR + H 2 O →> NCH 2 OH + ROH > NCH 2 OCH 2 OR + H 2 O →> NCH 2 OH + CH 2 O + ROH > NCH 2 OH + H 2 O →> NH + CH 2 O
(13) (14) (15)
The relative contributions to cure of the co-condensation and self-condensation reactions will depend on a variety of factors. These include: 1. The functionality of the amino resin, i.e., the relative proportions of >NCH2OR, >NCH2OH, and >NH groups present initially, as well as those generated during formulation and/or cure.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 8
2.
The functionality (hydroxyl, carboxyl, amido, carbamyl, etc.) and the equivalent weight (i.e. hydroxyl number, carboxyl number, etc.) of the primary film-former (co-reactant). 3. The amino/co-reactant ratio. 4. The level and type of catalyst (weak acid/strong acid). 5. Cure time and temperature. A co-reactant resin with a low hydroxyl number is best if formulated with a “polar” amino (i.e., one rich in >NH and/or NCH2OH) since these groups help build molecular weight during cure via self-condensation, particularly if little or no catalyst is present. Conversely, a high hydroxyl resin is best if matched with an alkoxymethyl-rich amino and cured with a strong acid catalyst. Where high cure temperatures are employed (e.g., can or coil coating operations), the choice of amino resin is less obvious, and, in practice, both polar and nonpolar aminos are used. Acid catalysts are usually used as an aid in curing amino-based formulations. These catalysts include very strong acids, such as p-toluenesulfonic acid (PTSA), dodecyl-benzenesulfonic acid (DDBSA), dinonylnaphthalenedisulfonic acid (DNNDSA), etc., and weaker acids, such as phenyl acid phosphate (PAP), butyl acid phosphate (BAP), etc. Amine blocking agents are sometimes used to help minimize resin advancement prior to cure. Some coatings, particularly those designed for high-bake temperatures, need no catalyst, relying instead on the combination of high temperature and perhaps carboxylic acid functionality on the primary film-former to bring about cure [17]. While all of the various reactions that take place during cure are accelerated by either acid or heat, it is fair to say that reactions of trans-etherification are most influenced by catalyst level and type, while reactions of self-condensation are most influenced by heat. The trans-etherification reaction takes place very rapidly under strong acid catalysis, even at low temperatures. This is especially true for aminos with a high level of alkoxymethyl substitution, i.e., a very low NH content, which tends to inhibit catalysis. Thus, most formulations involving resins with high alkoxymethyl ether content and designed for low-temperature cure (250°F or lower) will call for a sulfonic acid catalyst, either blocked or free. Although the individual reactions of cure are reasonably well understood and have been described in numerous papers [18–23], there is still much to be learned about the overall behavior of amino resins during cure, in particular the relative contributions of each of the various reactions. One of the difficulties is, of course, that the coating becomes intractable as cure progresses. Hence, a majority of studies involve analysis of the by-products of cure [18,22,24]. Other methods, such as dynamic mechanical analysis [25], nuclear magnetic resonance [26,27], FTIR [22], ESCA, etc., investigate the structure of the cured film. These techniques are useful not only for analyzing the freshly cured coating, but also as a means of following the coating through its lifetime, either natural or accelerated.
Degradation and Weathering
Amino-based crosslinked coatings exposed to the atmosphere are subject to both hydrolysis and UV-degradation at different rates dependant upon their structures. The mechanisms by which melamine resins hydrolyze have been described in detail by Berge and co-workers [28–30],
Q
AMINO RESINS
77
who was the first to distinguish between mono- and disubstituted nitrogen with respect to their behavior towards acid or base hydrolysis. Thus, in an alkaline medium, hydrolysis of an alkoxymethyl group on a singly substituted nitrogen is initiated by removal by the base of the proton attached to nitrogen: − NHCH 2 OR + B → − − NCH 2 OR + BH + −
−
(16)
− NCH 2 OR → −N = CH 2 + OR
(17)
−N = CH 2 + H 2 O → − NHCH 2 OH
(18)
−
+
OR + BH → ROH + B
(19)
This mechanism is clearly not applicable to di-substituted nitrogen (N(CH2OR)2), and these groups are in fact extremely resistant to alkaline hydrolysis. On the other hand, acid hydrolysis takes place readily for both mono- and di-substituted nitrogen. Berge proposed two mechanisms: (a) specific acid catalysis > NCH 2 − OR + H + →> NCH 2 OHR + +
+ 2
> NCH 2 OHR →> NCH + HOR + 2
> NCH + H 2 O →> NCH 2 OH + H
+
(20) (21) (22)
and (b) general acid catalysis − NHCH 2 OR + HA → −NHCH 2 OHR + + A − +
−
(23)
− NHCH 2 OHR + A → −N = CH 2 + ROH + HA
(24)
−N = CH 2 + H 2 O → − NHCH 2 OH
(25)
The work of Berge and co-workers with melamine resins is undoubtedly relevant to acid hydrolysis of paint films based upon hydroxyl functional primary film formers, which has been studied by a number of workers. English and co-workers [31,32] found that coatings prepared from highly alkylated melamines underwent extensive hydrolysis of residual methoxy groups during two years’ exposure in Florida, but there was no evidence of hydrolysis of bonds between melamine and the primary film-former. Bauer and Briggs [33,34] used IR to analyze acrylic-melamine coatings exposed to both UV and moisture and found evidence of hydrolysis of both residual methoxy groups and acrylic-melamine bonds, with the rate of hydrolysis being faster in the presence of UV light. The rate of hydrolysis was slowed considerably when a hindered amine light stabilizer was used. During the early 1990s, degradation of melaminecontaining automotive coatings had become particularly severe because of etching and spotting due to acid rain. The problem was compounded because the high-solids automotive coatings used very high levels of melamine resins (35 %–45 % of total binder weight) to help meet the lower amount of volatile organic (VOCs) requirements, giving rise to correspondingly high levels of acrylicmelamine bonds and residual alkoxymethyl groups in the cured film, all of which are susceptible to hydrolysis under acid conditions. The suppliers of high-solids coatings for automobiles were faced with a dilemma—either switch to more expensive alternative crosslinkers, such as isocyanates and aliphatic epoxies, which are more stable under
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
78
PAINT AND COATING TESTING MANUAL
Q
acid rain conditions, or modify the hydroxyl functional backbone to achieve increased resistance to spotting due to acid rain. An interesting aspect of the acid etch problem was that the damage was always more severe on relatively new coatings. If a newly painted automobile was protected from the acid environment for the first 6 to 12 weeks, damage thereafter was much less severe. One theory at the time was that the paint is undergoing some type of additional cure. Another possibility is that some type of protective surface oxidation occurs as it ages. In fact, Lamers et al. [35] eventually presented evidence that the melamine methylene ether carbon, involved in crosslinking, can be oxidized to a more acid etch-resistant urethane linkage in the presence of UV light and oxygen, lending support to the latter, protective surface oxidation theory. This protection mechanism (by oxidation to urethane) has since been all but proved by Wu et al. [36] with the introduction of acid etch-resistant automotive top coats based on tris-(alkoxycarbonylamino) triazine (TACT). The chemical structure of crosslinked hydroxyl functional acrylics and TACT is the same as those first prepared by Jacobs and DiLeone [37] using melamine tri-isocyanate as the crosslinker for hydroxyl functional automotive acrylic top coats. TACT forms crosslinks with hydroxyl functional acrylics by a trans-carbamylation reaction, affording acid etch-resistant melamine-urethane bonds. Since these early observations, suppliers of high-solids coatings have essentially solved the acid etch problem for automobiles in a very clever way by modifying their acrylic backbones (primary film-former) from hydroxyl functionality to primary and/or secondary carbamate functionality. This is usually done by a trans-carbamylation reaction on the poly-hydroxyl functional acrylic similar to TACT chemistry, but with mono-functional alkyl carbamates, such as methyl carbamate, to avoid premature cross-linking [38,39]. High levels of melamine resins can once again be used with the new carbamate functional acrylics, but this time resulting in melamine-primary carbamate cross-links, or melamine-secondary carbamate cross-links, both of which are very resistant to acid etch damage [40–42]. Automotive paint manufacturers are also actively pursuing water-borne systems, which use higher molecular weight, less hydroxyl functional co-reactant resins, and lower levels of melamine cross-linker. Mostly, these waterborne systems are used in the base coat, where in any case, the protective clear top coat minimizes acid attack. It is the clear top coat, with its high melamine content, that provides the excellent gloss and “distinctness of image” (DOI), characteristic of basecoat/clear coat technology. The melamine resin also minimizes the amount of solvent required because of its low viscosity at high-formulated solids, behaving in some ways as a reactive diluent and plasticizer.
End Uses of Amino Resins
Amino-based surface coatings protect and decorate the substrate to which they are applied. Their technology and use has developed over many years. As already mentioned, resins based on urea and melamine dominate the field. Urea resins are traditionally used in clear coatings for wood, e.g., furniture, kitchen cabinets, in paper, film, and foil applications, and in some appliance and general industrial coatings. They are also used to some extent in automotive
15TH EDITION
Fig. 7—Benzoguanamine.
primers. Urea resins cannot be used in automotive topcoats because of their sensitivity to hydrolysis. Melamine resins are much more widely used than urea resins. They give better chemical resistance, as well as resistance to weathering in exterior applications, despite some of the earlier difficulties described in connection with acid etch of automobiles before the use of carbamate functional acrylics. Besides automobiles, they are used in appliance formulations (both coil appliance and conventional postsprayed), general metal applications, container coatings (beer and beverage cans), etc. In choosing an amino resin for a particular application, consideration must be given not only to interior versus exterior use, but also to possible restrictions on cure conditions and compatibility of the amino resin with its co-reactant resin, both when formulated and as the paint film is formed during solvent flash-off and cure, etc. Compatibility of the amino is especially important in water-borne coatings, which are becoming more widely used. Another factor is the stability of the amino toward advancement (molecular weight buildup) during storage of the formulated paint. Benzoguanamine-based (Fig. 7) amino resins are used where film flexibility and hardness are required, as in some appliance applications (e.g., refrigerator doors made from coil stock, etc.). They also have good corrosion, humidity, and detergent resistance. Their use is limited by cost and poor exterior durability due to the pendant phenyl group on the benzoguanamine molecule. Glycoluril (Fig. 8) resins have been available since the late 1970s. In some pigmented formulations, they may require a higher cure temperature or a higher catalyst level than melamine-based resins, but show excellent corrosion and humidity resistance and release lower amounts of formaldehyde during cure [43]. Because of their increased resistance to hydrolysis, the glycolurils often find uses in container coatings where retort resistance is important and in some very interesting specialty applications, such as rheological thickeners for waterborne paints [44] and as curatives for optical electronics and photoresist fine line applications [45,46]. The tetra functional, fully methylated methylol glycolurils are high melting solid resins that find use in durable powder coatings [47], especially the wrinkle or “textured” finish applications [48].
Fig. 8—Glycoluril.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 8
ENVIRONMENTAL/TOXICITY
At least since the end of the 1970s, and perhaps starting even earlier, there has been an increased emphasis on the quality of the environment both in the workplace and beyond. In the coatings industry, this has meant strict controls on exposure of workers to hazardous ingredients in the coating formulation when applied, as well as on the nature and amounts of VOCs released to the environment when the formulation is cured. Amino resin suppliers have responded to these environmental challenges in a number of ways. Chief among these has been a progressive shift towards higher-solids, lower-molecular-weight aminos, which are now the resins of choice of coatings formulators. Many amino resins are supplied at 100 % non-volatiles, especially for the automotive industry. Where solvents are needed, those presenting the least hazard to worker and environment are selected. For their part, paint producers have increased the functionality of the co-reactant resin while lowering its molecular weight to minimize solvent use with the object of building molecular weight to the maximum possible extent during cure. This has meant using higher levels of amino resin, as much as 40 %–50 % of total binder weight in some cases. Perhaps the most intractable environmental problem with amino resins is the use of formaldehyde in their manufacture. Formaldehyde is recognized by the International Agency for Research on Cancer as a carcinogen. The American Conference of Governmental Industrial Hygienists lists formaldehyde as an “A2” substance, i.e., one suspected of carcinogenic potential for man, and the Occupational Safety and Health Administration (OSHA) has set workplace exposure limits of 0.75 ppm (8 h time weighted average) and 2 ppm (15 min short-term exposure limit). The formaldehyde content of amino resins is predominantly “combined,” i.e., chemically reacted, and represents about 30 %–50 % by weight of the resin. A small amount, ranging from about 0.1 % to about 3 % is present free, or un-reacted (see the section entitled Analysis/Analytical Methods). Amino resin suppliers have made considerable progress over the past several years in lowering the level of free formaldehyde in their products, which is important because of OSHA labeling requirements. In an ideal situation, all of the combined formaldehyde would remain in the coating after cure as part of the polymer network. In practice, however, some of the combined formaldehyde and all of the free formaldehyde is released during cure and may reach the environment, depending on the mechanics of the coating and curing operation. Since typically the free formaldehyde from reputable amino resin suppliers is very low, it is the partial release of combined formaldehyde during cure that is of concern. Efforts should be taken to optimize formulations for minimal release, not just for performance properties. Of course, scrubbing or incineration of any off gases, where possible, is one of the surest solutions.
References [1] Challener, C., J. Coat. Technol., Vol. 1, 2004, p. 46. online: http://goliath.ecnext.com/coms2/gi_0199-195058/Marketupdate-resins-Market-Update.html. [2] DeJong, J. I., and DeJonge, J., Recueil de Travail Chimie PayBas, Vol. 71, 1952, p. 643.
Q
AMINO RESINS
79
[3] Gordon, M., Halliwell, A., and Wilson, T., J. Appl. Polym. Sci., Vol. 10, 1966, p. 1153. [4] Gordon, M., et al., “The Chemistry of Polymerization Processes,” SCI Monograph No. 20, Society of Chemical Industry, London, 1966, p. 187ff. [5] Aldersley, J. W., et al., Polymer, Vol. 9, 1968, p. 345. [6] Okano, M., and Ogata, Y., J. Am. Chem. Soc., Vol. 74, 1952, p. 5728. [7] Braun, D., and Legradic, V., Angew. Makromol. Chem., Vol. 35, 1974, p. 101. [8] Tomita, B., J. Polym. Sci., Vol. 15, 1977, p. 2347. [9] Christensen, G., “Analysis of Functional Groups in Amino Resins,” Prog. Org. Coat., Vol. 8, 1980, p. 211. [10] Tomita, B., and Ono, H. J., J. Polym. Sci., Polym. Chem. Ed., Vol. 17, 1979, p. 3205. [11] Larkin, P. J., Makowski, M. P., and Colthup, N. B., Spectrochim. Acta, Part A, Vol. 55, No. 5, 1999, p. 1011. [12] Kambanis, S. M., and Rybicki, J., J. Coat. Technol., Vol. 52, No. 667, 1980, p. 61. [13] Chang, T. T., “Recent Developments in the Characterization of Melamine Resin Crosslinking Agents by Mass Spectrometry and Liquid Chromatography,” Polym. Prepr. (Am. Chem. Soc. Div. Polym. Chem.), Vol. 45, No. 2, 2004, p. 205; see also Chang, T. T., “Novel Approaches to Characterization of Melamine Coating Resins,” Prog. Org. Coat., Vol. 29, No. 1–4, 1996, p. 45. [14] Walker, J. F., Formaldehyde, 3rd ed., Robert E. Krieger Publishing Co., Huntington, NY, 1975, p. 486. [15] Hill, L.W., and Wicks, Z., Prog. Org. Coat., Vol. 10, 1982, p. 55. [16] Santer, J. O. (unpublished). [17] Yamamoto, T., Nakamichi, T., and Ohe, O., J. Coat. Technol., Vol. 60, No. 762, 1988, p. 51. [18] Blank, W., J. Coat. Technol., Vol. 51, No. 656, 1979, p. 61. [19] Blank, W., J. Coat. Technol., Vol. 54, No. 687, 1982, p. 26. [20] Santer, J. O., and Anderson, G. J., J. Coat. Technol., Vol. 52, No. 667, 1980, p. 33. [21] Santer, J. O., Prog. Org. Coat., Vol. 12, 1984, p. 309. [22] Lazzara, M.G., J. Coat. Technol., Vol. 56, No. 710, 1984, p. 19. [23] Nakamichi, T., Prog. Org. Coat., Vol. 14, 1986, p. 23. [24] McGuire, J. M., and Nahm, S. H., J. High Resolut. Chromatogr., Vol. 14, 1991, p. 241. [25] Hill, L. W., and Kozlowski, K., J. Coat. Technol., Vol. 59, No. 751, 1987, p. 63. [26] Bauer, D. R., Prog. Org. Coat., Vol. 14, 1986, p. 45. [27] Bauer, D. R., Prog. Org. Coat., Vol. 14, 1986, p. 193. [28] Berge, A., Kvaeven, B., and Ugelstad, J., Eur. Polym. J., Vol. 6, 1970, p. 981. [29] Berge, A., Adv. Org. Coat. Sci. Technol. Ser., Vol. 1, 1979, p. 23. [30] Berge, A., Gudmundsen, S., and Ugelstad, J., Eur. Polym. J., Vol. 5, 1969, p. 171. [31] English, A. D., Chase, D. B., and Spinelli, H. J., Macromolecules, Vol. 16, 1983, p. 1422. [32] English, A. D., and Spinelli, H. J., J. Coat. Technol., Vol. 56, No. 711, 1984, p. 43. [33] Bauer, D. R., J. Appl. Polym. Sci., Vol. 27, 1982, p. 3651. [34] Bauer, D. R., and Briggs, L. M., “Characterization of Highly Crosslinked Polymers,” American Chemical Society Symposium Series No. 243, American Chemical Society, Washington, DC, 1984. [35] Lamers, P. H., Johnston, B. K., and Tyger, W. H., Polym. Degrad. Stab., Vol. 55, 1997, p. 309; see also U.S. Patent No. 5, 106, 651 (1992). [36] Wu, K. J., Essenfeld, A., Lee, F. M., and Larken, P., Prog. Org. Coat., Vol. 43, No. 1, 2001, p. 167. [37] Jacobs, W., and DiLeone, R. R., U.S. Patent No. 4,939,213 (1990). [38] Singer, D. L., Swarup, S., and Mayo, M. A., PCT International Application No. WO 94/10213 A1 (1994). [39] Ohrbom, W. H., et al., European Patent Application No. EP 710676 A1 (1996). [40] Rehfuss, J. W., and St. Aubin, D. L., U.S. Patent No. 5,356,669 (1994). [41] Swarup, S., et al., PCT International Application No. WO 94/ 10211 A1 (1993).
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
80
PAINT AND COATING TESTING MANUAL
Q
[42] Higginbottom, H. P., Bowers, G. R., Ferrell, P. E., and Hill, L. W., J. Coat. Technol., Vol. 71, No. 849, 1999, p. 49. [43] Parekh, G. G., J. Coat. Technol., Vol. 51, No. 658, 1979, p. 101. [44] Glancy, C. W., and Steinmetz, A. L., U.S. Patent No. 5,914,373 (1999). [45] Pavelchek, E. K., and Trefonas, P., U.S. Patent No. 6,887,648 (2005).
15TH EDITION
[46] Barclay, G. G., and Puglino, N., U.S. Patent No. 7,211,365 (2007). [47] Jacobs, W., et al., “Durable Glossy, Matte and Wrinkle Finish Powder Coatings Crosslinked with Tetramethoxymethyl Glycoluril,” Prog. Org. Coat., Vol. 29, 1996, p. 127. [48] Cramer, M. L., and Osenbach, N. L., U.S. Patent No. 6,897,259 (2005).
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
9
MNL17-EB/Jan. 2012
Ceramic Coatings Richard A. Eppler1 CERAMIC COATINGS ARE AN ALTERNATIVE TO organic or polymer based coatings for selected surface coating applications. When protecting a surface, it is usually less expensive to use organic coatings rather than other materials such as ceramic coatings. However, organic paints have limitations and will not meet all service requirements. Though more expensive, for some applications ceramic coatings may be a more suitable form of protection. Vitreous (glassy) ceramic coatings are chosen for application over a substrate for one or more of several reasons [1]. These reasons for selecting a ceramic coating in preference to an organic paint include rendering the surface more chemically inert, impervious to liquids and gases, more readily cleanable, and more resistant to service temperature, abrasion, and scratching. The chemical durability of ceramic coatings in service substantially exceeds that of organic paints [2]. Vitreous coatings are formulated to be resistant to a variety of reagents, from acids to hot water to alkalies, as well as to essentially all organic media. The only important exception is hydrofluoric acid, which readily attacks all silicate glasses. This outstanding durability, combined with a very smooth surface, renders many ceramic coatings suitable for applications requiring the highest standards of cleanability, such as ware that comes in contact with food and drink. These coatings are also suitable for applications requiring true hermeticity, usually to protect sensitive electronic equipment. No organic resins are truly hermetic. Even the most thermally stable organic resins depolymerize at temperatures on the order of 300°C. Hence, organic paints are not suitable for applications requiring thermal stability above 300°C. For example, stove side panels are painted, but stove tops are porcelain enameled. A similar argument can be made for abrasion resistance. Organic resins are soft (Moh 2–3). By contrast, vitreous coatings are harder (Moh 5–6), and some plasma coatings are much harder. For example, alumina coatings, plasma sprayed, have Moh = 9. Vitreous coatings are thin layers of glass fused onto the surface of the substrate. When the substrate is a ceramic, the coating is called a glaze. When the substrate is a metal, the coating is called a porcelain enamel. When the substrate is a glass, the coating is called a glass enamel.
GLAZES
A ceramic glaze is a vitreous coating applied to a ceramic substrate, usually a whiteware. Glazes are applied to their substrates by one of several powder-processing techniques:
1
dipping, spraying, and waterfall or bell application. The raw materials are both crystalline oxides and frits. In these wet processes the raw materials are dispersed in an aqueous slip for application. After application, the coatings must be dried and fired at high temperatures (up to 1300°C, typically 1000–1100°C) to fuse them onto the substrate.
Applications for Glazes
Ceramic glazes find their way into a wide range of applications ranging from coffee mugs to automotive spark plugs. The major markets for ceramic coatings have different requirements, but one common theme is chemical durability and cleanability. The major products that normally use glazes are distributed as follows: 44.5 % 31.1 % 12.4 % 11.9 % 10 %
sanitary-ware, wall and floor tile, tableware, artware, electrical porcelain and electronics.
The total market for these products in the United States is estimated to be $12.4 billion for 2005 [3], of which the glaze component typically consumed 10 %–15 % of the total manufacturing cost. Hence, the value of the protective, functional, and decorative properties provided by the coating usually far outweighs the cost.
Leadless Glazes
Glazes are essentially mixtures of silica with other oxides added to permit the glaze to form at a readily achievable temperature. In a leadless glaze, the alkali and alkaline earth oxides, together with magnesia (MgO), zinc oxide (ZnO), and boron oxide (B2O3), are used to provide the fluxing action. Table 1 gives the formulas of a few typical ceramic glazes. Glaze 1 is a feldspathic glaze suitable for use on soft paste porcelains or hard stoneware [4]. This glaze is typical of that used on medieval Chinese porcelains. Glaze 2 is a sanitary-ware glaze [5]. It is derived from the soft paste porcelain glaze by the addition of ZnO. In the tile industry, the trend to ever faster firing rates (as low as 35 min cold-to-cold) has led to the formulation of glazes such as glazes 3 or 4 [6]. Here the melting rate is increased by both increasing the percent of fluxes, and increasing the alkaline earths and zinc oxide at the expense of the alkalis. Glaze 3 is opacified, while glaze 4 is a clear base for dark colors. To produce a glaze for tableware, the coefficient of
Consultant, Eppler Associates, 400 Cedar Lane, Cheshire, CT 06410.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed byASTM International This standard is for EDUCATIONAL USE ONLY. Copyright © 2012 by www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
81
82
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
TABLE 1—Typical ceramic glazes in weight percent Glaze
Li2O
Na2O
K2O
CaO
MgO
ZnO
SrO
BaO
PbO
B2O3
Al2O3
SiO2
ZrO2
1
0.00
2.24
3.24
9.71
4.44
0.00
0.00
0.00
0.00
0.00
14.44
69.90
0.00
2
0.00
2.05
3.12
11.15
0.00
5.39
0.00
0.00
0.00
0.00
18.58
59.71
0.00
3
0.00
3.24
1.56
8.81
0.07
3.50
2.29
0.00
0.00
3.29
9.54
56.51
11.20
4
0.00
2.78
2.82
11.68
0.08
3.16
0.11
0.00
0.00
4.55
11.89
62.37
0.57
5
0.00
1.81
2.71
9.16
0.62
10.94
30.7
2.50
0.00
5.47
7.37
55.79
0.57
6
0.20
4.24
0.43
2.18
0.00
1.86
4.73
12.23
0.00
17.78
8.13
48.22
0.00
7
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
88.14
0.00
0.00
11.86
0.00
8
0.00
2.15
2.24
16.52
0.00
1.07
2.10
0.00
0.00
3.26
7.95
55.37
9.35
thermal expansion must be reduced to match that of the ware. Glaze 5 is an example of a glaze for vitreous hotel china [7]. When the durability requirements of a food contact surface are not needed, processing can be improved by the use of a more fluid glaze, such as the spark plug glaze 6 [6].
concern are ASTM methods used to control release of lead and cadmium from glazed surfaces. These include: C738— Test Method for Lead and Cadmium Extracted from Glazed Ceramic Surfaces; and C895—Test Method for Lead and Cadmium Extracted from Glazed Ceramic Tile.
Lead-Containing Glazes
Porcelain enamel coatings are ceramic coatings designed for application to metals. Conventional porcelain enamel coatings are prepared in an aqueous system and applied to the substrate by spray, dip, or flow coating. The coating is dried before firing. Newer technology involves dry application of powdered porcelain enamel by electrostatic spray. The total market for porcelain-enameled products was reported to be $6.0 billion in 1999 [10]. About 86 % of the products are appliances, such as ranges, water heaters, home laundry, and dishwashers. About 6 % are cast-iron sanitary ware, and 8 % are architectural, cookware, and miscellaneous items. A porcelain enamel must be formulated such that it will bond to the metal substrate. For proper adherence of the enamel to the metal, it is necessary to develop a continuous electronic structure across the interface [11]. This structure is developed by saturating the enamel coating and the substrate metal with an oxide of the metal [12], which for iron and steel substrates is ferrous oxide. Certain transition metal oxides, such as cobalt oxide, nickel oxide, and cupric oxide, can be added to an enamel formulation to improve the adherence between the metal and the substrate. Ground coat enamels contain adherence oxides, while cover coat enamels do not.
Although historically important, the use of lead oxide in glazes is no longer acceptable, except in special applications [6]. The cost of meeting the regulalatory requirements for handling lead oxide are prohibitively expensive for most applications [6]. An exception is the coatings used on integrated circuit packages to seal them [8].
Satin and Matte Glazes
Satin and matte effects are due to dispersed oxide crystals of appropriate refractive index in the glaze [5]. Calcium aluminosilicate and zinc silicate crystals are commonly used. The crystals must be very small and evenly dispersed if the glaze is to have a smooth, velvet appearance. Glaze 8 in Table 1 is an example of a matte glaze.
Testing of Glazes
ASTM Committee C21 on Ceramic Whitewares and Related Products has developed several test methods to evaluate the physical properties of ceramic glazes. These are listed in Table 2 [9]. These tests form the basis for most quality control testing programs. There are several methods concerned with the fit of the glaze to the substrate. These include: C554—Test Method for Crazing Resistance of Fired Glazed Ceramic Whitewares by a Thermal Shock Method; C424—Test Method for Crazing Resistance of Fired Glazed Whitewares by Autoclave Treatment; C1300—Test Method for Linear Thermal Expansion of Glaze Frits and Ceramic Whiteware Materials by the Interferometric Method; and C372—Test Method for Linear Thermal Expansion of Porcelain Enamel and Glaze Frits and Fired Ceramic Whiteware products by the Dilatometer Method. Several other ASTM methods are concerned with chemical durability. These include: C650—Test Method for Resistance of Ceramic Tile to Chemical Substances; C1378—Test Method for Determination of Resistance to Staining; and C556—Test Method for Resistance of Overglaze Decorations to Attack by Detergents. Of particular
PORCELAIN ENAMELS
Ground Coat Enamels
A general-purpose ground coat enamel such as Enamel 1 in Table 3 is an alkali borosilicate containing small amounts of adherance oxides to promote the bonding process. Enamel 2 is a home laundry enamel that has been formulated for outstanding alkali resistance through the addition of large quantities of zirconia [13]. Hot water tank coatings such as Enamel 3 have very stringent thermal- and corrosionresistance requirements. Enamel 4 is a continuous clean coating. This is a porous coating that provides a means of volatilizing and removing food soils from the internal surfaces of ovens during normal operation [14].
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 9
Q
CERAMIC COATINGS
83
TABLE 2—Test methods for ceramic glazes [9] Number
Title
C1378
Test Method for Determination of Resistance to Staining
C1027
Test Method for Determining Visible Abrasion Resistance of Glazed Ceramic Tile
C650
Test Method for Resistance of Ceramic Tile to Chemical Substances
C609
Test Method for Measurement of Light Reflectance Value and Small Color Differences Between Pieces of Ceramic Tile
C554
Test Method for Crazing Resistance of Fired Glazed Ceramic Whitewares by a Thermal Shock Method
C424
Test Method for Crazing Resistance of Fired Glazed Whitewares by Autoclave Treatment
C556
Test Method for Resistance of Overglaze Decorations to Attack by Detergents (withdrawn)
C1300
Test Method for Linear Thermal Expansion of Glaze Frits and Ceramic Whiteware Materials by the Interferometric Method
C372
Test Method for Linear Thermal Expansion of Porcelain Enamel and Glaze Frits and Fired Ceramic Whiteware Products by the Dilatometer Method
C1028
Test Method for Determining the Static Coefficient of Friction of Ceramic Tile and Other Like Surfaces by the Horizontal Dynamometer Pull Meter Method
C584
Test Method for Specular Gloss of Glazed Ceramic Whitewares and Related Products
C738
Test Method for Lead and Cadmium Extracted from Glazed Ceramic Surfaces
C895
Test Method for Lead and Cadmium Extracted from Glazed Ceramic Tile
C1192
Standard Practice for Safe Spraying of Ceramic Glazes (withdrawn)
TABLE 3—Typical porcelain enamels in weight percent Oxide
Enamel 1
Enamel 2
Enamel 3
Enamel 4
Enamel 5
Enamel 6
Enamel 7
Li2O
0.88
0.81
1.33
0.52
0.89
1.10
1.76
Na2O
13.15
12.60
13.92
7.30
9.41
8.58
12.23
K2O
2.30
1.56
0.00
1.47
6.13
9.15
3.83
CaO
6.18
2.80
2.04
0.65
0.00
0.00
0.00
MgO
0.00
0.18
0.00
0.00
0.00
0.00
0.00
ZnO
0.00
0.26
1.27
0.00
0.00
1.04
0.00
BaO
7.27
0.73
0.56
0.00
0.00
0.00
0.00
CoO
0.47
0.36
0.47
0.03
0.00
0.00
0.00
NiO
0.29
0.31
0.00
0.03
0.00
0.00
0.00
CuO
0.20
0.00
0.00
13.99
0.00
0.00
0.00
B2O3
15.37
15.99
7.60
1.18
16.13
16.53
7.11
Al2O3
6.354
11.50
2.02
41.38
2.25
1.34
2.72
Cr2O3
0.00
0.00
0.00
1.24
0.00
0.00
0.00
Sb2O3
0.00
0.00
0.00
0.30
0.00
0.00
0.00
SiO2
44.01
41.55
56.05
24.20
40.97
46.74
59.07
ZrO2
0.00
6.36
11.66
7.24
0.00
0.00
7.86
TiO2
0.00
2.55
0.00
0.03
20.97
13.25
3.58
MnO2
0.20
0.66
1.81
0.03
0.00
0.00
0.00
P2O5
0.70
0.45
0.00
0.00
1.30
0.00
0.00
Nb2O5
0.00
0.00
0.00
0.00
0.06
0.00
0.00
WO3
0.00
0.00
0.00
0.00
0.05
0.00
0.00
MoO3
0.00
0.00
0.00
0.00
0.00
0.00
0.47
F
2.71
2.31
2.19
0.72
3.17
3.93
2.35
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
84
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
TABLE 4—Test methods for porcelain enamels [16] Number
Title
C148
Test Methods for Polariscopic Examination of Glass Containers
C282
Test Method for Acid Resistance of Porcelain Enamels (Citric Acid Spot Test)
C614
Test Method for Alkali Resistance of Porcelain Enamels
C756
Test Method for Cleanability of Surface Finishes
C538
Test Method for Color Retention of Red, Orange, and Yellow Porcelain Enamels
C839
Test Method for Compressive Stress of Porcelain Enamels by Loaded-Beam Method
C536
Test Method for Continuity of Coatings in Glassed Steel Equipment by Electrical Testing
C743
Test Method for Continuity of Porcelain Enamel Coatings
C374
Test Methods for Fusion Flow of Porcelain Enamel Frits (Flow-Button Methods)
C346
Test Method for 45-degree Specular Gloss of Ceramic Materials
C872
Test Method for Lead and Cadmium Release from Porcelain Enamel Surfaces
C539
Test Method for Linear Thermal Expansion of Porcelain Enamel and Glaze Frits and Ceramic Whiteware Materials by the Interferometric Method
C537
Test Method for Reliability of Glass Coatings on Glassed Steel Reaction Equipment by High Voltage
C283
Test Method for Resistance of Porcelain Enameled Utensils to Boiling Acid
C285
Test Method for Sieve Analysis of Wet-Milled and Dry-Milled Porcelain Enamel
C703
Test Methods for Spalling Resistance of Porcelain Enameled Aluminum
C385
Test Method for Thermal Shock Resistance of Porcelain Enameled Utensils
Cover Coat Enamels
Cover coat porcelain enamels are formulated to provide specific color and appearance characteristics, abrasion resistance, surface hardness, and resistance to corrosion, heat, and thermal shock. They can be clear, semi-opaque, or opaque. Opaque enamels such as Enamel 5 are used for white and pastel coatings [15]. They contain high concentrations of titania to provide the opacification. Semi-opaque enamels such as Enamel 6 are used for most medium-strength colors. Clear enamels such as Enamel 7 are used to produce strong bright colors. They are similar to ground coat formulations without the adherance oxides.
Testing of Porcelain Enamels
Test methods for porcelain enamel coatings are under the jurisdiction of ASTM Committee B-8 on Metallic and Inorganic Coatings. The methods are listed in Table 4. Again, they form the basis for most quality control test programs. Several of these test methods are concerned with the chemical durability of porcelain enamels. They include: C282—Test Method for Acid Resistance of Porcelain Enamels (Citric Acid Spot Test); C614—Test Method for Alkali Resistance of Porcelain Enamels; C756—Test Method for Cleanability of Surface Finishes; C538—Test Method for Color Retention of Red, Orange, and Yellow Porcelain Enamels; C872—Test Method for Lead and Cadmium Release from Porcelain Enamel Surfaces; and C283—Test Method for Resistance of Porcelain Enameled Utensils to Boiling Acid. A related issue is the possibility of defects pro-
viding a pathway from the surface to the substrate, usually called continuity of coating. Methods in this area include: C536—Test Method for Continuity of Coatings in Glassed Steel Equipment by Electrical Testing; C743—Test Method for Continuity of Porcelain Enamel Coatings; and C537— Test Method for Reliability of Glass Coatings on Glassed Steel Reaction Equipment by High Voltage.
GLASS ENAMELS
Glass enamels are vitreous coatings applied on glass. They provide a means of decoration, not an improvement in chemical durability or in cleanability. These coatings must be matured at temperatures below the deformation point of glass (1000–1200°F, or 538–649°C). Hence, they require large quantities of fluxing elements so that chemical durability is difficult to achieve. Glass enamels are produced in ready-to-use form (paste, thermoplastics, spray mediums, ultraviolet curable mediums) by a few select manufacturers. They represent a specialty product that is more akin to organic paints than to other ceramic coatings. The markets for this specialty product are categorized as tableware, glass containers, architectural, lighting, and automotive. As supplied to the user, glass enamels are mechanical mixtures of pigments, fluxes, and organic suspending media. The requirement for low maturing temperatures necessitates the use of very high lead oxide containing borosilicates for the flux. Leadless fluxes are now available, but have not yet achieved properties equal to the lead-containing fluxes. The organic suspending media are similar to materials used to make organic paints.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 9
Q
CERAMIC COATINGS
85
TABLE 5—Test methods for glass enamels [9] ASTM Method
Subject
C724
Test Method for Acid Resistance of Ceramic Decorations on Architectural-Type Glass
C735
Test Method for Acid Resistance of Ceramic Decorations on Returnable Beer and Beverage Glass Containers
C675
Test Methods for Alkali Resistance of Ceramic Decorations on Returnable Beverage Glass Containers
C676
Test Method for Detergent Resistance of Ceramic Decorations on Glass Tableware
C824
Standard Practice for Specimen Preparation for Determination of Linear Thermal Expansion of Vitreous Glass Enamels and Glass Enamel Frits by the Dilatometer Method
C927
Test Method for Lead and Cadmium Extracted from the Lip and Rim Area of Glass Tumblers Externally Decorated with Ceramic Glass Enamels
C978
Test Method for Photoelastic Determination of Residual Stress in a Transparent Glass Matrix Using a Polarizing Microscope and Optical Retardation Compensation Procedures
C777
Test Method for Sulfide Resistance of Ceramic Decorations on Glass
Testing of Glass Enamels
Test methods for glass enamels are under the jurisdiction of Subcommittee 14.10 on Glass Decoration of ASTM Committee C-14 on Glass and Glass Products. These methods [9] are listed in Table 5. Most of these methods are concerned with the chemical durability of glass decorations. They include: C724—Test Methods for Acid Resistance of Ceramic Decorations on Architectural-Type Glass; C735—Test Method for Acid Resistance of Ceramic Decorations on Returnable Beer and Beverage Glass Containers; C675—Test Methods for Alkali Resistance of Ceramic Decorations on Returnable Beverage Glass Containers; C676—Test Method for Detergent Resistance of Ceramic Decorations on Glass Tableware; and C927—Test Method for Lead and Cadmium Extracted from the Lip and Rim Area of Glass Tumblers Externally Decorated with Ceramic Glass Enamels.
REFRACTORY COATINGS
Flame spray techniques can be used to apply ceramic coatings in the molten state to heat-sensitive or massive substrates that cannot themselves be heated to high temperatures. Most ceramic coating materials used currently can be applied by flame spraying [17]. Silicates, silicides, carbides, oxides, and nitrides have all been deposited by this process. In these processes, the coating material is melted and projected as heated particles onto the substrate, where it instantaneously solidifies as a coating. Three methods of heating and propelling the particles in a plastic condition to the substrate surface include: (1) combustion flame spraying, (2) plasma arc flame spraying, and (3) detonation gun spraying. Combustion flame spraying is used for coating materials that melt readily. Plasma arc flame spraying is used for very refractory materials such as metal carbides. Detonation gun spraying is used for hard, wear-resistant materials such as tungsten carbide. Flame spray coatings generally lack smoothness and are usually porous. They are, therefore, limited to applications such as thermal barrier coatings, where porosity is a virtue, and wear-resistant coatings, where the materials cannot be applied readily by any other technique.
Testing of Refractory Coatings
There is only one test method for flame spray coatings in the ASTM standards: C633—Test Method for Adhesion or Cohesive Strength of Thermal Spray Coatings [16].
COATING APPLICATION
Ceramic coatings are applied to their substrates by one of several powder-processing techniques. In wet processes, the raw materials are dispersed in a slip. Slip preparation involves mixing the ingredients, particle-size reduction, dispersion in water, and the addition of minor amounts of additives to modify the rheological properties of the slip [18,19]. These processes are carried out together in a ball mill comprising a rotating cylinder partly filled with freely moving impact-resistant shapes. The application process for a ceramic coating must be straightforward and foolproof, reproducible, economical, and flexible [6]. Selection of the application technique is one of the most important decisions the coatings engineer makes. Criteria for this selection are type of ware, shape and size of ware, throughput required, energy and labor costs, and space available. All of these factors affect the quality and the cost of a coating process so that the best solution must be determined on an individual basis. Dipping is a simple, efficient, rapid technique requiring no capital equipment. The ware is immersed in the coating slip, moved around in a controlled way, removed from the slip, shaken to remove excess slip, and set down to drain and dry. Any bare spots are touched up with a finger wet with coating material. Its limitations are extreme sensitivity to operator skill and difficulty in automating volume production. Spraying is a process whereby a coating slip is broken down into a cloud of fine particles that are transferred to the substrate by either pneumatic, mechanical, or electrical forces. The method requires a gun, a container or feed mechanism, an impelling agency, and a properly designed hood or booth maintained under negative pressure [20], Spraying lends itself to high-volume automated systems [21]. The articles are continuously fed under a battery of angled spray guns. Coating reclaim is an essential part of automated systems. Slip can also be applied mechanically with a rotating atomizer. Slip is passed onto a set of closely spaced rotating
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
86
PAINT AND COATING TESTING MANUAL
Q
disks, which throw the coating into a fan of droplets. Costs are similar to spraying. The primary use of this technique is in producing textured coatings on tile. If the substrate is conductive (that is, a metal), the surface quality and uniformity of a ceramic coating can be improved by using the electrostatic spray coating technique [22,23]. In this system, the slip is broken into droplets either by air atomization or by centrifugal force from a sharp-edged rotating surface. The drops acquire a high negative charge and are dispersed as a fine mist. They are driven forward to the grounded substrate following the lines of force. Hence, coating material can reach the underside of the ware, and full-edge coverage is achieved. There are other techniques for specific applications. Tile require only one face to be glazed, but with a very smooth coating. This suggests the Bell or waterfall technique [6], where a continuous feed of tiles is carried under a curtain of fluid slip made by pouring a stream of slip over a bell shaped device. Painting and brushing are seldom used except for special effects and for applying glaze to inaccessible areas. For substrates that require precisely positioned areas of coating, the silk screen process can be used [20]. Finely powdered dry coating material is dispersed as a smooth paste. Using a squeegee, this paste is pressed through the open areas of a fine mesh screen stretched on a frame. For coating a total piece, costs are excessive. There are a few techniques of application that do not require the preparation of a slip. They include flame spraying, dry powder cast iron enameling, and electrostatic dry powder enameling. Flame spraying can be used to apply ceramic coatings in the molten state to heat-sensitive or massive substrates. Flame spray coatings generally lack smoothness and are usually porous. Equipment and material costs are generally high. In dry powder cast-iron enameling, a casting is heated in a furnace to red heat. It is then withdrawn from the furnace and, while still hot, dusted with dry powdered frit by means of a vibrating sieve placed over the surfaces to be coated. The powdered frit melts and adheres as it falls on the hot surface. This process is also extremely operator sensitive. Recently, it has been adapted for robot application, which serves to reduce variations over time. The most important dry application method, and the one most recently introduced, is dry powder electrostatic application of all-fritted coatings to conductive substrates. This technique involves charging individual coating particles at a high voltage and then spraying them towards the substrate surface. Charging of particles is accomplished by encapsulating the coating material with an organic silane. It is then suspended in clean compressed air in a fluidized bed container [24]. The fluidized powder is siphoned and propelled through powder feed tubes to special electrostatic powder guns for low-pressure application. The powder carries a potential of up to 100 kV, which causes it to seek out and attach itself to the grounded workpiece. Capital costs of this process are substantial, but operating costs are reduced through elimination of slurry preparation and drying of the ware.
15TH EDITION
[2] Eppler, R. A., “Corrosion of Glazes and Enamels,” Chap. 12, Corrosion of Glass, Ceramics, and Ceramic Superconductors, D. E. Clark and B. K. Zoitos, Eds., Noyes Publications, Park Ridge, NJ, 1992. [3] Grahl, C., “Tile and Sanitary Markets Benefit from Bathroom Trends; Dinnerware Industry Faces Significant Challenges,” Ceramic Industry, Vol. 156, No. 12, 2006, pp. 22–27. [4] Tichane, R., Ching-te-Chen; Views of a Porcelain City, N.Y. State Institute for Glaze Research, Painted Post, New York, 1983. [5] Singer, F., and German, W. L., “Ceramic Glazes,” Borax Consolidated, 1964. [6] Eppler, R. A., and Obstler, M., Understanding Glazes, American Ceramic Society, Westerville, OH, 2005. [7] O’Conor, E. F., Gill, L. D., and Eppler, R. A., “Recent Developments in Leadless Glazes,” Ceram. Eng. Set Proc., Vol. 5, Nos. 11–17, 1984, pp. 923–932. [8] Tummala, R. R., and Shaw, R. R., “Glasses in Microelectronics in the Information-Processing Industry,” Adv. Ceram., Vol. 18, 1986, pp. 87–102. [9] “Glass, Ceramic Whitewares,” Part 15.02, ASTM Annual Book of Standards, ASTM International, West Conshohocken, PA, 2008. [10] Sheppard, L. E., “The Porcelain Enamel Industry—New Developments and Challenges,” Ceram. Ind., Vol. 150, No. 10, 2000, pp. 30–35. [11] Pask, J. A., “Chemical Reaction and Adherance at Glass-Metal Interfaces,” Proceedings of the PEI Technical Forum, Vol. 22, 1971, pp. 1–16. [12] King, B. W., Tripp, H. P., and Duckworth, W. H., “Nature of Adherance of Porcelain Enamels to Metals,” J. Am. Ceram. Soc., Vol. 42, No. 11, 1959, pp. 504–525. [13] Eppler, R. A., Hyde, R. L., and Smalley, H. F., “Resistance of Porcelain Enamels to Attack by Aqueous Media: I—Tests for Enamel Resistance and Experimental Results Obtained,” Am. Ceram. Soc. Bull., Vol. 56, No. 12, 1977, pp. 1064–1067. [14] Monteith, P. G., Linhart, O. C., and Slaga, J. S., “Performance Tests for Properties of Low Temperature Thermal Cleaning Oven Coatings,” Proceedings of the PEI Technical Forum, Vol. 32, 1970, pp. 73–79. [15] Shannon, R. D., and Friedberg, A. L., “Titania-Opacified Porcelain Enamels,” Illinois University Engineering Experimental Station Bulletin, No. 456, 1960, pp. 1–49. [16] “Metallic and Inorganic Coatings,” Part 2.05, ASTM Annual Book of Standards, ASTM International, West Conshohocken, PA, 2008. [17] Taylor, T. A., Bergeron, C. G., and Eppler, R. A., “Ceramic Coating,” Metals Handbook, 9th ed., Vol. V, ASM International, Metals Park, OH, 1982, pp. 532–547. [18] Taylor, J. R., and Bull, A. C., Ceramics Glaze Technology, Pergamon Press, Oxford, England, 1986. [19] Reed, J. S., Introduction to the Principles of Ceramic Processing, John Wiley & Sons, New York, 1988. [20] Bloor, W. A., and Eardley, R. E., “Environmental Conditions in Sanitary Whiteware Shops, II. Glaze Spraying Shops,” Trans. J. British Ceramic Soc., Vol. 77, No. 2, 1978, pp. 65–69. [21] Whitmore, M., “Spraying of Earthenware Flatware,” Transactions, Journal of the British Ceramic Society, Vol. 73, No. 4, 1974, pp. 125–129. [22] Hebberlein, K., “Electrostatic Glazing of Tableware,” Ber. Dtsch. Keram. Ges., Vol. 53, No. 2, 1976, pp. 51–55. [23] Lambert, M., “Industrial Application of Electrostatic Enamelling to Parts in Sheet Steel and Cooking Equipment,” Vitreous Enameller, Vol. 24, No. 4, 1973, pp. 107–109. [24] ASM Committee on Porcelain Enameling, “Porcelain Enameling,” Metals Handbook, 9th ed., Vol. 5, ASM International, Metals Park, OH, 1982.
References [1] Eppler, R. A., “Glazes and Enamels,” Glass Science and Technology, Chap. 4, Vol. 1, Academic Press, New York, 1983, pp. 301–337. Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
10
MNL17-EB/Jan. 2012
Epoxy Resins in Coatings Michael J. Watkins1 PREFACE
THE AUTHOR COLLABORATED WITH RONALD Bauer and Edward Marx in preparing this chapter for the previous or 14th edition of this manual. This revised edition includes expanded descriptions of epoxy materials and curing agents currently used in coating formulations. Epoxy resins are versatile materials that are used in a wide variety of coatings. In the interests of brevity, this chapter focuses on the epoxy resins, curing agents, and coating types that are commercially most important and account for the largest usage of epoxy resins. As an example, epoxy coating technologies, which reduce volatile organic compound (VOC) emissions, have grown rapidly. Waterborne, high solids, and powder epoxy coatings account for about 80 % of all epoxy resin used in coatings. On the other hand, low solids epoxy coatings have declined significantly in importance and account for less than 15 % of epoxy resin usage. So this chapter naturally focuses on the low VOC technologies. Similarly, epoxy ester coatings were described in the previous edition, but have declined in importance, so they are not discussed here. Also, epoxy coatings are formulated with other materials described in this manual, and attempts have been made to reference those chapters where appropriate.
INTRODUCTION
Generically, epoxy resins can be characterized as a group of oligomeric materials that contain one or more epoxy (oxirane) groups per molecule. Almost all commercially significant epoxy functional materials are derived by reacting epichlorohydrin with various materials containing groups with active hydrogen (such as phenolic hydroxyl, aliphatic hydroxyl, carboxylic acid, or amine). The initial reaction yields a chlorohydrin, which is subsequently dehydrochlorinated to yield the glycidyl (epoxy) group. One exception to this generalization is the class of cycloaliphatic epoxy resins discussed in Chapter 16, Miscellaneous Materials and Coatings, in this manual. Another exception is represented by epoxy functional acrylic resins that are made by copolymerizing various acrylic monomers with glycidyl methacrylate or similar epoxy functional acrylic monomers. Epoxy resins based on bisphenol A (BPA) and epichlorohydrin are commercially the most important epoxy resins by a very wide margin. These resins have become technologically important materials that find extensive application in high-performance coatings, adhesives, and reinforced composites. Almost since their commercial introduction in about 1947 [1], epoxy resin systems have been used in
1
protective coatings. Historically, protective coatings were the largest single end use for epoxy resins. Although in recent years the non-coating applications of epoxy resin have been growing, coatings still represent about half of the annual epoxy resin usage. The principal components of any epoxy coating system are the epoxy resin and the curing agent or hardener. Epoxy resins are reactive intermediates that can be liquid or solid, and they are converted into the final coating by reaction with curing agents (hardeners). Curing agents function by reacting with specific groups in the epoxy resin molecule to give a three-dimensional, infusible polymer network. Although the resin and curing agent are common to all epoxy coatings, other materials are incorporated to achieve the desired rheological characteristics, cure speed, appearance, and film performance.
BPA EPOXY RESINS
BPA-based epoxy resins were developed independently by Pierre Castan in Switzerland and by Sylvan Greenlee in the United States during the 1930s and 1940s [1–3]. The generalized structure for these resins is given in Fig. 1. In commercial products, the n value ranges from 0 to about 60. Table 1 displays a range of typical epoxy resins that are commercially available, along with their properties and applications. As n increases, the epoxy equivalent weight increases, as does the number of hydroxyl groups. Thus epoxy resins with low n values are normally cured by reaction of the epoxy group, whereas those resins with higher n values are cured by reaction of the hydroxyl functionality. The highest molecular weight BPA epoxy resins, or phenoxy resins, are described in Chapter 16, Miscellaneous Materials and Coatings, in this manual. Resins having n values less than 1 are viscous liquids; they are used mainly in two-pack, ambient-temperature cure coatings, as well as in electrical castings, flooring, electrical laminates, and fiber-reinforced composites. Resins having n values in the range of about 1–2 are low melting solids that are used in solution in two-pack, ambient-temperature cure coatings. Resins having n values in the range of about 2–6 are solids, which do not sinter at room temperature. They are predominately used in powder coatings. All of these applications are cured through the epoxy groups. The higher n value resins, particularly those with n > 10, are normally used in solution and find their greatest application in heat-cured coatings. In these resins, the concentration of epoxy groups is low, and so they are cured with materials that react with the hydroxyl groups along the backbone.
Sr. Staff Research Chemist, Hexion Specialty Chemicals, Westhollow Technology Center, 3333 Highway 6 South, Houston, TX 77082.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed byASTM International This standard is for EDUCATIONAL USE ONLY. Copyright © 2012 by www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
87
88
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 1—Idealized structure of a bisphenol epoxy resin.
Coatings based on BPA epoxy resin offer a unique combination of performance characteristics and are used in a wide variety of technologically important end uses. These characteristics include exceptional adhesion and corrosion resistance, excellent chemical resistance, low shrinkage, high strength, good heat resistance, toughness, and excellent electrical properties. Because of the adhesion and corrosion resistance, epoxy resins have been used in corrosion-resistant primers and coatings since epoxy resins were first commercialized. In addition, the excellent chemical resistance has resulted in the formulation of epoxy coatings that provide outstanding protection against severe corrosive environments. They are used extensively in coatings for refineries, chemical plants, and marine equipment, such as ships and offshore platforms. Other important applications where epoxy resin coatings are used almost exclusively because of the corrosion protection they afford include automotive, aircraft, and appliance primers as well as coatings for both the inside and outside of pipelines. The chemical and corrosion resistance resulted in the development of BPA epoxy coatings used for the interior linings of beer and beverage cans, the interior linings of food cans, and for chemical-resistant linings of pails and drums. Guidelines for the safe use of coatings based on BPA epoxy resins for direct food contact have been established by the U.S. Food and Drug Administration [4]. These coatings are used not only to protect the metal of the container from corrosion, but also to protect the flavor of the contents, which can be affected by direct contact with metal.
The one significant weakness of coatings based on BPA epoxy resins is that they are not resistant to ultraviolet (UV) exposure. When exposed to direct sunlight, they tend to yellow and chalk. The aromatic structure of the BPA backbone absorbs the UV energy and the backbone degrades. Thus, BPA epoxy resins are typically used as primers, which are subsequently top coated with coatings that are resistant to UV degradation. This strategy takes advantage of the excellent corrosion resistance of the epoxy primer, while protecting it from its one weakness.
OTHER EPOXY MATERIALS USED IN COATINGS
Although BPA-based epoxy resins are far and away the largest volume commercial epoxy resins, other epoxy functional materials are also important. The diglycidyl ether of bisphenol F is a low viscosity liquid epoxy resin that it useful in high solids or 100 % solids coatings or floorings. For comparison, the viscosity of the standard commercial grade of BPA liquid resin is about 120–140 Poises at 25°C, while that for a typical BPF liquid resin is about 25–45 Poises. This advantage over BPA resins diminishes at higher molecular weights so BPF solids resins are not generally used. Epoxy phenol novolacs, epoxy alkylphenol novolacs (i.e., epoxy cresol novolacs), or epoxy BPA novolacs are multifunctional epoxy resins that are used to increase cross-link density and subsequently increase the chemical resistance, hardness, heat deflection temperature, etc. for epoxy coatings and other applications. These
TABLE 1—Typical properties of BPA-based epoxy resins Average Molecular Weight
Average EEWa
Approximate Average Value of n
Viscosity (P @ 25°C)
Softening Point (°C)b
350
182
0
80
. . .
Solventless and solvent-borne ambient cure coatings, electrical encapsulation, flooring, and filament winding
380
188
0.2
140
. . .
Solvent-borne, ambient cure coatings
600
310
1
Semi-solid
40
Solvent-borne, ambient cure coatings
900
475
2
Solid
70
Solvent-borne, ambient cure coatings
1400
900
4
Solid
100
Powder coatings
2900
1850
10
Solid
130
Heat cured, solvent-borne coatings for cans, drums, primers, etc.
3750
3050
13
Solid
150
Heat cured, solvent-borne coatings for cans, drums, primers, etc.
Applications
EEW—epoxide equivalent weight, grams of resin providing 1 mole of epoxide. Also referred to as WPE (weight per epoxide) and EMM (epoxy molar mass). All three terms are interchangeable. b Softening point by Duran’s mercury method (ASTM D1763). a
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 10
resins have epoxy functionality in the range of 2–8. (By comparison, BPA resins and BPF resins have typical functionality of about 2.) As an example, epoxy novolac resins are used in combination with BPA resins in powder coating formulations to enhance chemical and corrosion resistance. Another multifunctional resin that may be used is the glycidated tetraphenol of ethane. This has a nominal epoxy functionality of 4. As with the epoxy novolacs, it can be used along with BPA epoxy resins to improve chemical and corrosion resistance. Brominated epoxy resins can be prepared from epichlorohydrin and tetrabromobisphenol A, or from liquid BPA epoxy resin and tetrabromobisphenol A. These have a broad use in electrical laminating (i.e., circuit board) applications. The bromine content makes them suitable for flame retardant coatings. Epoxy resins such as the polyglycidyl ether of castor oil are multifunctional resins that are used with BPA epoxy resins to improve the flexibility and water resistance. A whole series of epoxy functional materials are used as reactive diluents and modifiers in epoxy coatings. Monofunctional epoxies are typically used as reactive diluents in high solids or 100 % solids, two-pack ambient cure coatings. One group of monofunctional epoxies are made by reacting epichlorohydrin with alcohols. One widely used example is the glycidyl ether of mixed alkyl C12−C14 alcohols. Another commercially important example is the glycidyl ether of mixed alkyl C8−C10 alcohols. The glycidyl ether of n-butanol gives the best viscosity reduction of any reactive diluent in BPA epoxy systems and is also important commercially. Another group of monofunctional epoxies are made by reacting epichlorohydrin with phenol or alkylphenols. The most widely used example is o-cresyl glycidyl ether. Other examples include phenyl glycidyl ether, p-tertbutylphenyl glycidyl ether, and nonylphenyl glycidyl ether. A third type of monofunctional epoxy is the glycidyl ester of neodecanoic acid. Monofunctional epoxies of this type are useful as reactive diluents to reduce viscosity and VOC. However, because they are monofunctional, they also decrease crosslink density of the cured coating, generally resulting in some loss of chemical resistance, corrosion resistance, hardness, etc. So the coatings formulator must carefully balance the need for low viscosity with other coating performance requirements. A group of multifunctional epoxies are available to serve as reactive diluents, but still maintain functionality and cured film properties. These are generally prepared by reacting epichlorohydrin with polyols. Examples include neopentyl glycol diglycidyl ether, butanediol diglycidly ether, cyclohexanedimethanol diglycidyl ether, trimethylolpropane triglycidyl ether, and trimethylolethane triglycidyl ether. Resorcinol diglycidyl ether is an effective difunctional reactive diluent. Unfortunately, it is a strong skin irritant and sensitizer and can cause severe allergic reactions. Therefore, relative to other reactive diluents, it is difficult to handle safely. Another specialty epoxy resin is made by reacting epichlorohydrin with hydrogenated BPA. Since this material is saturated, rather than aromatic, it has UV resistance superior to that for BPA epoxy resins.
Q
EPOXY RESINS IN COATINGS
CURING AGENTS
89
Epoxy resins are reactive intermediates composed of mixtures of oligomeric materials containing one or more epoxy groups per molecule. To convert epoxy resins into useful products, they must be cross-linked or “cured” into a three-dimensional polymer network. Cross-linking agents, or curing agents as they are generally called, function by reaction with or cause the reaction of epoxide or hydroxyl groups in the epoxy resin. The number of curing agents that have been developed over the years for epoxy resins is overwhelming. Selection of the curing agent is as important as that of the base resin; it is dependent on the performance requirements of the film and the constraints dictated by the specific method of application. The most widely used types of curing agents employed in epoxy resin coatings are the amine-functional materials for ambient-cure coatings, dicyandiamide or acid-functional polyesters for powder coatings, and amino resins (see Chapter 8, Amino Resins, in this manual) or phenolic resole resins (see Chapter 11, Phenolics, in this manual) for heatcured liquid coatings. The principal amine-functional curing agents used in two-pack, ambient-cure epoxy coatings are polyamides (see Chapter 12, Polyamides, in this manual). Amidoamines, aliphatic amines, and epoxy-amine adducts are also used. Specialty amine curing agents include phenalkamines, cycloaliphatic amines, and ketimines. These materials cure epoxy resins by reaction of the amine with the epoxy groups. Typical aliphatic amines used include diethylenetriamine (DETA), triethylenetetramine (TETA), and tetraethylenepentamine. Aliphatic amines such as DETA and TETA can be pre-reacted with low molecular weight epoxy resins to make epoxy-amine adducts that are very useful as epoxy curing agents. Amidoamines are made by reacting fatty acids with aliphatic amines (i.e., DETAor TETA). Polyamides are made by reacting dimerized fatty acids with aliphatic amines (i.e., DETA or TETA). More pounds of polyamide curing agents are consumed annually in the United States than any other type of epoxy resin curing agent. Polyamide cured epoxy coatings develop superior adhesion to moist and poorly prepared surfaces, and they provide a high degree of corrosion resistance. Like epoxy resins, polyamides are also mixtures of oligomers. Thus, a range of polyamides that vary in viscosity, amine equivalent weight, and reactivity is available. Polyamide cured coatings exhibit somewhat better retention of flexibility and impact resistance on aging than polyamine adducts. Although resistance to solvents and acids is not quite as good as with other types of amine curing agents, polyamides are adequate and cost-effective for most applications where amine cure epoxy coatings are used. Typical cycloaliphatic amines include isophorone diamine, bis(p-aminocyclohexyl)methane, and 1,2-diaminocyclohexane. Ketones add reversibly to primary amines with the loss of water to give ketimines. The ketimines obtained from the typical polyamine curing agents have rather low volatility compared to the precursor polyamine. Ketimine curing agents can be considered blocked polyamines, which in the presence of water hydrolyze to produce a ketone and a polyamine. These ketimines react at a practical rate of cure under ambient conditions. Atmospheric moisture, which is absorbed during and following application of the coating,
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
90
PAINT AND COATING TESTING MANUAL
Q
serves as the source of water required to activate the curing agent. Ketimine curing agents are similar in behavior to the aliphatic amine polyamines and amine adducts in rate of cure and performance of cured films, but they provide much longer pot lives. Epoxy resins are formulated with acid-functional polyesters to make hybrid powder coatings. Catalysts are often used to facilitate this reaction. The ratio of epoxy to polyester is typically in the range from 50:50 to 30:70. When the powder coating is baked, the acid groups on the polyester react with the epoxy groups to make a cured polymer matrix. The principal curing agent for epoxy powder coatings is dicyandiamide. At elevated temperature, this reacts with the epoxy groups to cure the coating. Catalysts such as imidazoles or epoxy-imidazole adducts are commonly used to speed up this reaction. Strong bases such as imidazoletype catalysts can also be used alone as curatives. They cause homopoly-merization of epoxy groups as a means of cure. Specialty curing agents for powder coatings include phenolic-functional polyethers, phenolic resoles (see Chapter 11, Phenolics), and anhydrides. Baked liquid coatings employ high molecular weight epoxy resins. These resins contain relatively low epoxide functionality and relatively high hydroxyl functionality. These are cured through the hydroxyl groups with melamine-formaldehyde resins and urea-formaldehyde resins (see Chapter 8, Amino Resins), as well as phenolformaldehyde resins (see Chapter 11, Phenolics). Strong acids, such as phosphoric acid or p-toluene sulfonic acid, are used as catalysts. Strong organic acids blocked with volatile amines can also be used to achieve enhanced package stability. The acid-amine salt is non-catalytic. However, at elevated temperature, the amine volatilizes, leaving the acid to catalyze the curing reaction.
EPOXY COATINGS
The four coating types that account for the highest usage of epoxy resins are powder coatings, high solids solvent-borne coatings, electrodeposition coatings, and waterborne coatings. These coating types account for over three quarters of epoxy resin usage in coatings. Low solids, solvent-borne epoxy coatings are clearly declining in volume. This is not surprising, as a major driver for change in the coatings industry is the reduction of VOCs in coatings. These four coating types are used to reduce VOC emissions.
EPOXY POWDER COATINGS
Powder coatings are produced by melt blending homogenous dispersions of solid resins, curing agents, pigments, fillers, and various additives. The dispersion is solidified by cooling, ground into a finely divided powder form, and classified by particle size for subsequent use. The resultant powder is normally electrostatically deposited onto grounded substrates and, through the application of heat, converted into very high performance thermoset films. The process of applying coating powders allows nearly 100 % powder utilization and evolves almost no VOCs. The 1970s volatiles regulations and energy concerns raised interest in powder coating technology. The real sustaining driving forces for growth, however, have been improvements in powder coating raw materials, formulations, manufacturing technology, and application equipment. The advantages for the use of powder coatings can best be summed up in the
15TH EDITION
“Four E’s,” used by the Powder Coating Institute: (1) Excellence of finish, (2) Economy in use, (3) Energy efficiency, and (4) Environmental acceptability. The Clean Air Act, as amended in 1990, has contributed to even greater interest in the use of powder coatings to meet more stringent volatile organic requirements. Powder coatings comprise one of the fastest growing areas of coatings technology. Current growth rate for epoxy powder coatings is approximately 6 % versus about 2 %–4 % for other coating types. The unique characteristics of solid epoxy resins account for their choice by formulators for use in powder coatings applications. BPA-based epoxides with equivalent weights greater than about 650 are non-sintering and friable. They have relatively low melt viscosity and high reactivity via the terminal oxirane functionality. The addition reaction with amines, phenolics, or carboxylic acid functional curatives allows a wide range of formulations. The primary limitations for BPA-based epoxy resins in powder coatings are yellowing and loss of gloss that occur when these coatings are exposed to exterior weathering conditions. Powder coatings are broadly divided into either “functional” or “decorative” uses. Functional coatings are normally applied at film thicknesses greater than about 3 mils and are expected to withstand some rather severe service. Examples of functional uses are coatings for exterior and interior pipe, rebar, and various electrical devices. Although decorative powder coatings are functional, these are normally used at a film thickness of 3 mils or less and are not expected to perform significantly better than baked films derived from “wet” coatings. Some examples of decorative uses are coatings for appliances, furniture, and under hood automotive parts.
HIGH SOLIDS, SOLVENT-BORNE COATINGS
Industrial maintenance and marine paints account for nearly all of these coatings. These are two-package, ambient cure coatings. A two-package coating is comprised of the epoxy component and the curing agent, which are packaged separately and often in volume ratios of 2 to 1 or 4 to 1 of epoxy component to curing agent. Two-package epoxy coatings are mixed just prior to application and are characterized by a limited working life or pot life after the resin and curing agent components are mixed. Commercial systems will have pot lives of a few hours to a couple of days, with typical working times of about 4–8 h. Historically, maintenance and marine coatings were formulated with solid epoxy resins (average n approximately 2) in solution, cured with high molecular weight polyamides. These gave excellent corrosion protection, but were high in VOC (about 4 lb/gal, or more). In order to achieve lower VOC and higher solids, formulations have shifted to using lower molecular weight BPA epoxy resins (n = 0.2) or BPF liquid epoxy resin with lower viscosity curing agents, such as amidoamines, aliphatic amines, phenalkamines, cycloaliphatic amines, and ketimines. By using lower viscosity resins and curing agents, higher solids systems can be achieved. Another strategy to reduce VOC has been to formulate coatings with some exempt solvents, such as acetone, methyl acetate, t-butyl acetate, and p-chlorobenzotrifluoride.
ELECTRODEPOSITION COATINGS
Epoxy resin electrodeposition coatings are waterborne coatings formulated from either anionic or cationic epoxy
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 10
resin polymers. The part to be coated is dipped into the electrodeposition bath, and an appropriate electrical charge is applied, causing the coating to deposit onto the part. The part is then removed from the bath, rinsed, and baked to cure the coating. In the United States, epoxy-based electrodeposition coatings account for over 90 % of all electrodeposition coatings. Epoxy-based cathodic electrodeposition (CED) automotive primers dominate this application, accounting for over 70 % of all electrodeposition coatings. Over 40 million pounds of epoxy resin are used in the United States in CED automotive primers, making this one of the largest single end uses for epoxy resins in coatings. Virtually every automobile made in the United States, Europe, and Japan is primed with an epoxy CED primer. CED primers are used because they afford exceptional corrosion protection and because they are deposited uniformly to all areas of the automobile, even in areas that would be inaccessible to other coating application methods such as spray. Because of their major importance, the remainder of this discussion will deal with CED automotive primers. The preparation of CED [5] coatings generally begins by reacting a BPA-based liquid epoxy with BPA to give an epoxy resin with an epoxy equivalent weight in the range of 500–1000. This epoxy resin may then be reacted with a flexibilizing diol. This diol can be an aliphatic diol or a polyether diol. The principal requirement is that the diol contain primary hydroxyl functionality. These primary hydroxyls are reacted with the epoxy groups in the presence of a suitable catalyst (e.g., a tertiary amine) to form ether linkages between the epoxy and the flexibilizing diol. At this point, the resin may have an epoxy equivalent weight in the range of 1000–1500. The remaining epoxy functionality is then reacted with amines. Generally, secondary amines are chosen to minimize further chain extension. One favored method to accomplish this is to use a diketimine of diethylenetriamine. During coating preparation, the ketimine groups decompose to give primary amines. These primary amines are fairly basic, resulting in stable dispersions at a relatively high bath pH (pH > 6). At this point, the CED resin preparation is complete. In practice, specialized CED resins are used to make the pigment grind pastes. These are developed to efficiently make stable pigment dispersions, which retain good stability in the CED coating bath. Curing agents used are generally blocked isocyanates. These are chosen to be stable and unreactive in the coatings bath, but to unblock and cure the coating at baking temperature. An example of such a curing agent would be the reaction product of 3 mol of toluene diisocyanate with 1 mol of trimethylolpropane. This is then reacted with 3 mol of a suitable blocking agent. Historically, 2-ethyl-1-hexanol has been used. However, much research has been done in recent years to find blocking agents that unblock at lower temperatures, permitting lower bake temperatures and energy savings. One example of many is provided in Ref. [6], where oximes are used as blocking agents. Catalysts such as tin or lead salts are generally used to facilitate unblocking and coating cure. Recent formulation efforts have eliminated lead catalysts. One example of many is provided in Ref. [7]. The coating is prepared by blending the resin with pigment paste, curing agent, catalysts, additives, and solvents. A low-molecular-weight organic acid, such as lactic or acetic acid, is then added to the mixture to
Q
EPOXY RESINS IN COATINGS
91
make ammonium salts with the amine groups in the resin. This mixture is then dispersed in water to make the CED coating. Solvents may be required in the preparation of the CED resin or other components. In order to reduce the VOC content of the finished coating, it is usually subjected to a vacuum stripping step, which can reduce VOC to less than 0.7 lb/gal. When the automobile is dipped into the CED bath, a negative charge is applied to it (making it the cathode) relative to counter electrodes in the bath. Electrolysis of water occurs, forming hydroxide ions in the immediate vicinity of the automobile surface. These hydroxide ions react with the ammonium ion groups in the resin near the surface, regenerating the neutral amine groups and causing the coating to be deposited onto the surface. In this way, a uniform film is applied to the entire conductive surface of the automobile. The automobile is then removed from the bath, rinsed, and baked.
WATERBORNE COATINGS
It was determined early on that epoxy coatings are safe and highly effective linings for food and beverage cans. Historically, these coatings were solvent borne. Recently, there has been concern about low molecular weight materials that may be extractable from can linings, such as BPA, BPA diglycidyl ether (BADGE), or derivatives of BADGE. Although research is ongoing, the current consensus is that epoxy-based can linings are safe and effective [8–10]. The need to reduce VOC has driven the development of waterborne food and beverage can linings [11]. Waterborne linings for beer and beverage cans are used exclusively and are an important use for epoxy resins. This is not surprising when one considers that over 100 billion beer and beverage cans are manufactured in the United States every year. These coatings are based on high-molecular-weight epoxy resins (average n of about 10–13), onto which are grafted acrylic terpolymers (i.e., styrene/methacrylic acid/ethyl acrylate). These epoxy/acrylic graft polymers are neutralized with base, such as dimethylethanolamine, to give a resin easily dispersible in water. The dispersed resin may be cured with an amino resin (see Chapter 8, Amino Resins) to give coatings with properties that make them suitable for beer and beverage containers. Food can linings are typically solvent-borne high-molecular-weight epoxy resins (average n of about 10–13), which are cured with amino resins (see Chapter 8, Amino Resins) or phenolic resins (see Chapter 11, Phenolics). Generally, food can linings require more chemical resistance than beer and beverage cans. Waterborne epoxy food can linings have been developed, which are based on similar technology to the beer and beverage can linings. These coatings are being used, and are growing in share of the food can lining market. Waterborne two-pack ambient-cure coatings comprise a small but growing end use for epoxy resin. BPA-based epoxy resins are rather hydrophobic and are not easily dispersed in water. Initial waterborne epoxy resins were modified with large amounts of surfactants to form stable dispersions. But the surfactant made subsequent coatings relatively hydrophilic, resulting in loss of corrosion resistance when applied to metallic substrates. These early waterborne epoxy coatings performed well when applied to non-metallic (e.g., cementitious) substrates. One solution to this problem was to make aqueous epoxy dispersions by using nitroparaffin cosolvents in place of surfactants [12–14]. The nitroparaffin
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
92
PAINT AND COATING TESTING MANUAL
Q
evaporates from the coating after application and does not affect the hydrophobicity or performance of the coating. This approach has been used to formulate waterborne aerospace coatings. Advances in surfactant technology and in dispersion technology have resulted in waterborne epoxy resins and curing agents, which give excellent performance on metallic substrates [15–17]. The need to lower VOC is encouraging the use of these systems in industrial maintenance, aerospace, and railroad coatings.
References [1] May, C., and Tanaka, Y., Epoxy Resins Chemistry and Technology, Marcel Dekker, Inc., New York, 1973. [2] Lee, H., and Neville, K., Handbook of Epoxy Resins, McGrawHill Book Co., New York, 1967. [3] Ellis, B., Chemistry and Technology of Epoxy Resins, Blackie Academic and Professional, London, 1993. [4] U.S. Code of Federal Regulations 21, Part 175.300. [5] Bauer, R. S., “Epoxy Resin Chemistry,” ACS Symposium Series 114, American Chemical Society, Washington, DC, 1979. [6] Garner, A. W., “Low Temperature Curing Cathodic Electrocoat,” U.S. Patent No. 6,517,695 (February 2003). [7] Kaufman, M. L., “Cationic Electrocoating Compositions, Method of Making, and Use,” U.S. Patent No. 5,820,987 (August 1996).
15TH EDITION
[8] The Society of the Plastics Industry, Safety of Epoxy Can Coatings, 2004. [9] European Commission Scientific Committee on Food, Statement of the Scientific Committee on Food on Bisphenol A diglycidylether (BADGE), 2002. [10] European Commission Scientific Committee on Food, Opinion of the Scientific Committee on Food on Bisphenol A, 2002. [11] Bauer, R. S., “Epoxy Resin Chemistry II,” ACS Symposium Series 221, American Chemical Society, Washington, DC, 1983. [12] Albers, R. A., “Water-Reducible Epoxy Coating Compositions Without Emulsifier,” U.S. Patent No. 4,352,898 (October 1982). [13] Albers, R. A., “Water Reducible Epoxy Coating Composition,” U.S. Patent No. 4,495,317 (January 1985). [14] Albers, R. A., “Water-Reducible Epoxy Coating Compositions,” U.S. Patent No. 4,501,832 (February 1985). [15] Galgoci, E. C., Komar, P. C., and Elmore, J. D., “High Performance Waterborne Coatings Based on Dispersions of a Solid Epoxy Resin and an Amine-Functional Curing Agent,” J. Coat. Technol., Vol. 71, No. 891, 1999, pp. 45–52. [16] Elmore, J. D., Kincaid, D. S., Komar, P. C., and Nielsen, J. E., “Waterborne Epoxy Protective Coatings for Metal,” J. Coat. Technol., Vol. 74, No. 931, 2002, pp. 63–72. [17] Watkins, M. J., Weinmann, D. J., and Elmore, J. D., “Formulating High-Performance Waterborne Epoxy Coatings,” Thermoset Resin Formulators Association 2006 Annual Meeting, September 11–12, 2006, Montreal, Quebec, Canada.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
11
MNL17-EB/Jan. 2012
Phenolics* John D. Fisher1 HISTORY
PHENOLIC RESINS, INITIALLY COMMERCIALIZED in 1909, were the first completely synthetic materials of the burgeoning plastics business. The expansion of several new technologies of the time, namely the electrical, communications, and automotive industries, all required and depended on new materials that had better electrical insulating properties, higher heat resistance, and improved resistance to chemicals, acids, oils, and moisture. The heat-reactive or “resole” resins, developed by Dr. Leo H. Baekeland [1], were formulated into blends that were convenient for mass production of compression molded parts that satisfied these requirements. Improved and new items, such as coil supports, commutators, distributor heads, telephone sets, vacuum tube bases, radio parts, and electrical switches, all blossomed onto the market within a few years.
FIRST PHENOLIC RESIN-BASED COATINGS
Concurrent with the above developments, the non-heatreactive phenolic resins or “novolak” resins were prepared as a hoped-for substitute for shellac. These resins were not as resilient as shellac and, when used alone, were not successful in coatings. However, combined with a formaldehyde donor such as hexamethylene tetramine, the novolaks could be compounded into another family of thermosetting molding materials, which found early use in phonograph records. While novolaks had to wait for success in coatings, the resole resins in alcohol solutions by 1911 were found to form excellent films when cross-linked by baking [2]. These coatings, still in wide use today, are hard and glass-like and have excellent resistance to chemicals, acids, water, and solvents. Early applications included protective coatings for brass beds as well as other hardware items. These solution resins also initiated the manufacture of laminates, which were used to make early radio circuit boards and, later, printed circuit boards.
PHENOLIC RESINS IN COATINGS
Coatings Based on Phenolic Resins
The early coatings based on phenolic resoles developed over time into a family of products used as protective coatings. Coatings based on similar technology are still found in a variety of applications. Chemically resistant protective coatings for stationary and mobile tanks,
drums, and pipes are produced based on phenolic resole solutions that are applied and then cured in place with the application of heat. Regulations limiting the release of volatile organic compounds led to the development of higher solids versions and waterborne versions in the 1970s and 1980s, but other than these changes the products are essentially the same as the products developed many years ago. Coatings of this nature possess excellent chemical resistance, good high temperature performance, high hardness, and hence good abrasion resistance and good specific adhesion to metal substrates. They suffer from low flexibility and deterioration upon prolonged exposure to UV light. Their use is limited to applications with rigid substrates and applications where UV light exposure will not be a concern. One innovation by scientists at the General Electric Company in the early 1950s led to the development of phenolic resins modified by reaction with allyl chloride [3–8]. Coatings produced from phenolic resins that use this technology have even greater chemical resistance than typical phenolic resins due to the replacement of the mildly acidic proton of the aromatic hydroxyl group with the allyl group. This yields a coating that is less susceptible to reaction with strong bases than a typical phenolic resin-based coating and hence has higher chemical resistance. However, this modification does not significantly help the flexibility of the coating (Fig. 1).
Coatings Based on Polymer Alloys with Phenolic Resins
As noted, coatings based wholly or primarily on phenolic resins possess some properties that make them desirable coatings, but they suffer from being extremely brittle. While they are suitable for applications where little or no flexibility is required, the lack of flexibility precludes them from use in many applications where this property is required or desired. Use of phenolic resin technology in a wider variety of applications had to wait for further developments. This came in two ways. The first was modification of the phenolic resins to achieve compatibility with traditional resins used in coating applications. The second was the development of other synthetic resin technology for use in the coatings market. One of the traditional coating technologies is the use of drying oils of various types as the principal vehicle to
*DEFINITION: a polymeric, resinous reaction product of a phenol with an aldehyde. Said products may be used alone or in formulations with other polymers to produce useful coatings. 1 Schenectady International, Inc., PO Box 202, Pattersonville, NY 12137.
93
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed byASTM International This standard is for EDUCATIONAL USE ONLY. Copyright © 2012 by www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
94
PAINT AND COATING TESTING MANUAL
Q
Fig. 1—Modification of resoles with allyl chloride.
carry and bind pigment to a variety of substrates. Over time it was found that “cooking” the oils with various other ingredients improved the performance of the coating made from the oil. This technology developed into what are now known as alkyd resins, which are essentially oil modified polyester resins. Modification of alkyd resins with phenolic resins held the promise of improving performance of the alkyd resin-based coatings, but the compatibility of early phenolic resins with the oil-based alkyd resins was not good and found limited application. Once technology advanced to the production of phenolic resins from alkylated phenolic monomers, the higher aliphatic content provided by the alkylated phenolic monomers greatly improved the compatibility with the oil-based alkyd resins. A variety of improvements in performance was observed for the phenolic resin modified alkyds, including greater abrasion resistance, better high temperature performance, improved chemical resistance, and improved adhesion to metallic substrates. Early phenolic resin modified alkyds were made from resole resins based on alkylated phenolic monomers that had to be cooked into the alkyd resins during manufacture of the alkyd resins. Later, higher molecular weight phenolic resins that could be “cold cut” into the alkyd resins were developed. Both technologies are still in use today [2]. The use of alkylated phenolic monomers to produce phenolic resins with improved compatibility with other coating ingredients led to a wide range of new products that took advantage of the improved heat resistance and hardness of the phenolic resins and the good film-forming properties of the drying oils and the alkyd resins derived from them. One of the areas that depended strongly on this developing technology was electrical insulation coatings where phenolic resins were alloyed with other coating resins to upgrade the performance of both primary and secondary electrical insulation coatings where products with much better performance allowed manufacturers to build electrical motors with improved performance and durability [9]. The development of other new synthetic polymers with potential applications in the coatings industry also created new opportunities for the use of phenolic resins in a wider range of coating applications. Particularly significant was the development of epoxy resins. When cured, epoxy resins exhibit a level of flexibility far higher than the phenolic resins. Also, the similarity in structure between phenolic resins and the predominant commercial epoxy resins makes them compatible across a wide range of blending ratios giving the coatings formulator the opportunity to balance the flexibility of the epoxy resin with the chemical resistance of the phenolic resin for each given application. Further, the aromatic hydroxyl group present throughout the phenolic resin structure is reactive with the oxirane group characteristic of the epoxy resins. When fully reacted the blend of the two polymers becomes one homogeneous film.
15TH EDITION
Use of phenolic/epoxy alloys in coatings has been employed in a variety of coating applications. One area where they have found wide acceptance is the formulation of coatings for metals, particularly in the packaging coatings industry. In this industry, the ability of phenolic resole/ epoxy alloys to provide very good chemical resistance allows packaging of food stuffs and beverages in metal cans coated with very thin film weight coating. These coatings take advantage both of the reaction between the epoxy and the phenolic resin as well as the homopolymerization of the phenolic resole, both of which occur when the coating is baked onto the substrate. Phenolic resole/epoxy alloys are also used in wash or pretreatment coatings. In these applications, a very dilute coating is applied during the wash process and in preparation for painting for the purpose of preventing flash rust formation and/or to improve adhesion of the permanent coating to the substrate. Phenolic novolak/epoxy alloys have also found use in coating applications. Powder coating formulations take advantage of the relatively low melt viscosity, the good friability, and film forming of the phenolic resin. In addition, the thermally induced reaction of the phenolic hydroxyl with the oxirane encourages the use of the novolak resins as cross-linkers for the epoxy resins to make chemically resistant powder coatings. This addition reaction has the added benefit that no volatiles need to be released, which is especially important in powder coatings where no solvent is present to facilitate the release of volatile components from the film without pinhole formation (Fig. 2). In addition to being used in alloys to produce epoxy and alkyd based coatings, phenolic resins have proved useful as modifiers at low levels for a variety of other synthetic resins for coating applications. Some examples are the use of phenolic resins as adhesion promoters for a variety of synthetic polymer-based coatings, These include SBR, NBR, and poly-(vinyl chloride) resins, to name a few. In addition the resole phenolic resins are often added as a cross-linking agent to formulations based on polyvinyl butylral or polyvinyl alcohol resins.
PHENOLIC RESINS AS PHOTO-IMAGABLE COATING
Phenolic novolak resins have also found application in the area of photo-imagable coatings. For this application, a coating of phenolic novolak mixed with a photo-sensitive base is applied to a substrate. The coated article is partially exposed to an image and the photo-sensitive base forms a salt with the phenolic novolak where the light activates the base. The phenolic resin salt has a distinct solubility difference from the phenolic novolak that has not reacted with base and so may be selectively removed, exposing the substrate below. The article is then able to be further processed and the exposed portion of the substrate may be selectively reacted while the unexposed substrate is protected by the remaining phenolic resin coating. This technology is used commercially on silicone wafers to make semiconductor chips and on aluminum plates to make planographic printing plates.
PHENOLIC RESIN CHEMISTRY
The reaction between phenols and aldehydes to produce resinous products was difficult to understand in the early years because many of the products were insoluble or
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 11
Q
PHENOLICS
95
Fig. 2—Reaction between epoxy resin and phenolic resin.
Fig. 3—Typical phenolic starting materials.
Fig. 4—Typical phenolic novolak structure.
Acid Catalysis infusible. With the advent of modern analytical tools, the chemistry of the reactions has been more fully defined by various workers. [10–12].
Raw Materials
The commercially important phenols used in coatings resins are shown in Fig. 3. While phenol is the most common, the substituted phenols are also used to vary the solubility, reactivity, and physical properties of resins. The cresols, alkylated phenols and bisphenol-A, are widely employed in various coating applications. Phenolic resins based on other phenolic monomers are used but have limited or specialty uses. Phenol has three ring positions that are active for reaction with aldehydes; the two and six carbon atoms (ortho) and the four position (para). Phenols with substitution in the above positions have lower functionalities and are frequently used to modify resin properties. The aldehyde co-reactant of choice for reaction with the phenols is formaldehyde, the most reactive of those commercially available. Formaldehyde is a gas but is conveniently handled as an aqueous solution (formalin), as an alcohol solution (formcel), or in a solid polymeric form known as paraform. Formaldehyde in aqueous solution exists as hydrated glycols or low-molecular-weight glycol ethers, which are easily broken down into formaldehyde under normal reaction conditions. Alternative aldehydes and other bridging agents are employed in the production of phenolic resin, but these alternatives comprise a minute minority of the commercially available phenolic resins.
Acid-catalyzed phenol-formaldehyde reaction proceeds through an unstable addition intermediate to form condensed, methylene-linked phenolic rings (Fig. 4). Acid catalyzed phenol-formaldehyde oligomers and polymers are generally referred to as “novolaks.” Novolaks are thermoplastic and require the addition of other materials to further polymerize. When phenol is used, highly branched novolaks are obtained. However, when substituted phenols are used, the functionality of the phenolic monomer is reduced to two and linear resins or cyclic calixarenes are formed [13]. Most novolak resins are produced as solids but are readily soluble in a variety of solvents.
Base Catalysis
The use of base to catalyze the reaction between phenol and formaldehyde produces initial reaction products such as the methylolated phenols, as shown in Fig. 5. Further reaction causes the methylol groups to condense with other ring positions to form a methylene link or to etherify with other alcohol groups to form dibenzyl ether links (Fig. 6). Base catalyzed phenol-formaldehyde oligomers and polymers are
Fig. 5—Methylolated phenols.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
96
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 6—Typical phenolic resole structure.
Fig. 7—Homopolymerization of phenolic resole resins.
generally referred to as “resoles.” Unlike novolaks, resoles are thermosetting and will further polymerize upon heating (Fig. 7). Additional reaction raises the molecular weight of resoles. As with novolaks, phenol is poly-functional towards formaldehyde and will form highly branched polymers. These highly branched polymers will ultimately reach a highly cross-linked “gelled” state with continued exposure to heat. Phenolic resins produced with substituted phenols will have lower functionality if the reactive ortho or parasites are blocked. Resins produced using substituted phenols will chain extend or form cyclic calixarenes upon continued exposure to heat. Resole resins are typically produced as solutions in solvents, but some are available as solids.
TESTING OF PHENOLIC RESIN PRODUCTS
Typical quality control tests for phenolic resin products may include the following: 1. Gel time [ASTM Test Method for Determining Stroke Cure Time of Thermosetting Phenol-Formaldehyde Resins (D4640-86)] (heat-reactive resins). 2. Viscosity (solution). 3. Color (Gardner). 4. Specific gravity. Other tests to characterize resins may be used: 1. Molecular weight distribution by gel permeation chromatography. 2. Structure analysis—NMR and IR.
3. 4. 5.
Thermal analysis—TGA, DSC, and TMA (curing curves). Free residual formaldehyde by various methods. Free residual phenols by gas chromatography.
References [1] Baekeland, L. H., “The Synthesis, Constitution, and Uses of Bakelite,” Ind. Eng. Chem., Vol. 1, No. 3, 1909, pp. 149–161. [2] Richardson, S. H., Paint and Varnish Production, August 1955. [3] Martin, R. W., “Trimethylol Phenol Compound and Derivatives Therof,” U.S. Patent No. 2,579,329 (1951). [4] Martin, R. W., U.S. Patent No. 2,579,330 (1951). [5] Martin, R. W., “Compositions Containing Methylol Phenyl Esters,” U.S. Patent No. 2,579,331 (1951). [6] Martin, R. W., U.S. Patent No. 2,598,406 (1951). [7] Martin, R. W., U.S. Patent No. 2,606,929 (1951). [8] Martin, R. W., “Alkylene Oxide-Methylol Phenol Reaction Products,” U.S. Patent No. 2,606,935 (1951). [9] Myer, J. F., “Coating Compositions of a Dibasic Polycarboxylic Acid/Tris(2-hydroxyethyl) Isocyanurate Polyester and a Phenol Formaldehyde Resin,” U.S. Patent No. 3,249,578 (1966). [10] Megson, N. J. L., Phenolic Resin Chemistry, Academic, New York, 1958. [11] Martin, R. W., The Chemistry of Phenolic Resins, Wiley, New York, 1956. [12] Gardziella, A., Phenolic Resins: Chemistry, Applications, Standardization, Safety and Ecology, Springer, New York, 2000. [13] Gutsche, C. D., Calixarenes, Royal Society of Chemistry, Cambridge, 1989.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
12
MNL17-EB/Jan. 2012
Polyamides Robert W. Kight1 POLYAMIDES
POLYAMIDE RESINS ARE POLYCONDENSATION products of dimerized fatty acids and polyamines. Reactive liquid polyamide resins are oligomers designed primarily for use in the manufacture of two-component epoxy/polyamide coatings and adhesives. The two-component coatings are generally labeled Part A and Part B, with the liquid polyamide resin usually (though not always) contained in Part B. The polyamide resin may function as the curing agent, core-actant, or hardener for epoxy resin. Polyamide resins should not be considered as catalysts although they may initiate the reaction; the polyamide resin reacts with the epoxy resin and becomes part of the polymer. The majority of polyamide resins used in coatings are viscous liquids that are usually supplied by the coatings manufacturer as a solution in organic solvents. The solution may be a clear amber liquid or may contain pigments in colored systems. Current trends in the coatings industry are toward higher solids, lower volatile organic compound (VOC)containing products. These products require lower viscosity reactants, such as liquid epoxy resins cured with amidoamines rather than polyamides. Amidoamines are condensation products of monobasic fatty acids and polyamines and are therefore lower in viscosity. These products are less compatible with epoxy resins so commercial examples are adducts of amidoamines. Adduction improves compatibility, though generally increases viscosity, so high boiling, compatible solvents, such as benzyl alcohol, are commonly used to reduce viscosity. Benzyl alcohol becomes trapped in the cured film due to its structural similarity to epoxy resins based on bisphenol A and does not migrate through the film to the surface. Many commercial high-solids curing agents are amidoamine adducts containing benzyl alcohol.
ACIDS
The dibasic fatty acids of commercial importance used to manufacture polyamide curing agents are prepared by dimerizing unsaturated C18 fatty acids from linseed, soya, or tall oils. Linseed and soya fatty acids are extracted from flax and soybeans, respectively. Crude tall oil is a byproduct from the Kraft process for papermaking and is a mixture of fatty acids and rosin acids, from which the fatty acids are separated by distillation. The tall oil fatty acids are a mixture of C18 isomers with a variable number of double bonds. Some of the isomers combine via Diels–Alder addition and other mechanisms to form C36 dibasic acids or dimer acids. The dimer acids produced may be acyclic, monocyclic, or polycyclic in structure, depending on the
1
location and number of double bonds in the feedstock. Many isomers are present in commercial dimer acids, most of which are difunctional carboxylic acids [1]. Examples of three possible isomer types follow: acyclic (Fig. 1), monocyclic (Fig. 2), and polycyclic (Fig. 3).
AMINES
The dimer acids are reacted with various polyamines to form polyamide resins and a variety of other useful products. The liquid polyamide resins commonly used in industry are polyethylene polyamines of various chain lengths that are linear, branched, or cyclic. The linear polyethylene polyamines are characterized as secondary amine groups separated by ethylene chains, terminated on either end by primary amine groups. Diethylenetriamine is an example of a simple linear polyethylene polyamine (Fig. 4). The cyclic and branched polyamine isomers contain tertiary amine groups in addition to the primary and secondary amine groups. Longer chain length polyethylene polyamines available commercially, such as triethylenetetramine and tetraethylenepentamine, are mixtures of linear, branched and cyclic polyamine isomers. Aminoethylpiperazine is an example of a cycloaliphatic polyamine (Fig. 5). The reaction between dimerized fatty acids and polyamines yield amide oligomers with amine group termination. These amide oligomers are used as coreactants with epoxy resins in high-performance coatings, as well as components of a variety of other useful commercial compositions including two-component adhesives.
EARLY HISTORY
Polyamide resins were commercialized in the late 1950s for use with epoxy resins in the manufacture of two-component adhesives and high-performance coatings. The early commercial epoxy/amine coatings contained aliphatic amines, primarily diethylenetriamine, which had several negative features, such as requiring critical mix ratios and toxicity. The introduction of liquid polyamide resins allowed the coatings manufacturer to produce high-performance coatings characterized by convenient mix ratios, such as 1:1 or 2:1, with very low toxicity. Epoxy/polyamide coatings find utility in a wide variety of applications, such as industrial maintenance coatings, machinery and equipment enamels, and marine applications. The presence of the long fatty chains gives coatings with much better flexibility as well as better wetting and adhesion than was obtained with the earlier aliphatic amine cured systems.
Sr. R&D Associate, Arizona Chemical Company, 1201 West Lathrop Ave., Savannah, GA 31415.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed byASTM International This standard is for EDUCATIONAL USE ONLY. Copyright © 2012 by www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
97
98
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 1—Acyclic dimer acid.
SYNTHESIS OF POLYAMIDES
Polyamide resins are polycondensation products of difunctional fatty acids and polyfunctional amines. In a typical commercial example, 1 mol of dimer acid is reacted with 2 mol of diethylenetriamine. During this condensation reaction, 2 mol of water are evolved. As the reaction proceeds, an interesting side reaction occurs: An additional mole or so of water evolves from a secondary reaction. One of the primary amine groups reacts with the dimer acid to form an amide linkage, which is a nitrogen bonded carbonyl. In addition, the ethylene chain next to the amide function and the secondary amine nitrogen are incorporated into a five-membered ring, known as an imidazoline ring. This condensation reaction, which also evolves water, eliminates an active hydrogen to yield a tertiary amine group. The degree of cyclization obtained is controlled to yield a product with specific useful properties, such as improved solubility and compatibility and longer pot life. Similar reactions occur at the other carboxylic acid group of the dimer [2]. If 50 % of the diethylenetriamine present in the polyamide is cyclized to imidazoline, a total of 3 mol of water of reaction is evolved. These products are shown in the following structures: the polyamide (Fig. 6) and the imidazoline (Fig. 7).
Commercial products range from about 35 % to more than 80 % imidazoline to allow the coating formulators latitude in customizing the properties of their products. Other properties of the polyamide resin that are important to the coatings formulator are the amine value, which is related to active hydrogen equivalent weight, and the viscosity level of the polyamide resin in organic solvents. Amine values range from about 100 to about 400, with active hydrogen equivalent weights of about 550-125, respectively. The active hydrogen equivalent weight is used to calculate the amount of polyamide resin required to react with a given amount of an epoxy resin of known epoxide equivalent weight. The ratio of these values, known as the stoichiometric mix ratio, is most often only a starting point for the formulator. By varying the mix ratio of the polyamide resin to the epoxy resin, certain properties of the cured coating are enhanced (and others are sacrificed) to obtain specific application properties.
CHEMICAL PROPERTIES
The total amine value of liquid polyamide resins is determined by potentiometric or colorimetric titration using dilute hydrochloric or perchloric acid to neutralize the amine base. Generally, 0.1N hydrochloric acid dissolved in an alco-
Fig. 2—Monocyclic dimer acid. Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 12
Q
POLYAMIDES
99
Fig. 3—Polycyclic dimer acid.
hol is used to titrate liquid polyamide resins having 100 amine value. The amine value is commonly listed in the specification properties of commercial liquid polyamide resins and is defined as the number of milligrams of potassium hydroxide equivalent to the basicity in 1 g of sample. The acid value, generally less than 5, which is also specified in commercial liquid polyamide resins, is defined as the number of milligrams of potassium hydroxide required to neutralize 1 g of sample. The acid value may be determined using ASTM Test Method for Acid Number of Naval Stores Products Including Tall Oil and Other Related Products (D465-05). The imidazoline content is not specified in commercial liquid polyamide resins except in special cases where the level is deemed critical. The level of imidazoline is usually controlled by the polyamide resin manufacturer to provide products with specific compatibility and/or solubility. Imidazoline level can best be measured by scanning the polyamide resin with an infrared spectrophotometer and comparing the absorption at 6.25 μm to the absorption at 6.05 μm. The imidazoline ring absorbs at 6.25 μm, and the nitrogen-bonded carbonyl, or amide, absorbs at 6.05 μm. The result is reported as either a ratio of imidazoline:amide (I/A) or as a percentage. In the example reaction described previously, the imidazoline ratio would be 1.0 and the percentage would be 50 %.
PHYSICAL PROPERTIES
storage requirements of the coatings manufacturer. For ease of handling, they may be supplied in various solvents. Most liquid polyamide resins suitable for coatings applications are quite viscous, and these polyamide resins are soluble in a variety of organic solvents including alcohols, glycol ethers, ketones, and aromatic hydrocarbons. Thus the coatings manufacturer has considerable latitude in selecting specific solvents for optimum applications properties. The percent nonvolatile content of polyamide resin solutions may be determined in accordance with ASTM Test Method for Nonvolatile Content of Resin Solutions (D1259-06). Commercial liquid polyamide resins are generally supplied in a single organic solvent at between 60 % and 80 % solids, which provides a handleable viscosity. The coatings formulator further dilutes the polyamide resin solution with more of the same solvent, or with a solvent blend, to form one component of the two-component system. The polyamide component may be clear or may contain pigments in colored coatings formulations. The color of the liquid polyamide resin or polyamide resin solution is determined in accordance with ASTM Test Method for Color of Transparent Liquids (Gardner Color Scale) (D1544-04). The viscosity of the liquid polyamide resin may be measured at elevated temperature in accordance with ASTM Test Method for Rheological Properties of NonNewtonian Materials by Rotational (Brookfield type) Viscometer (D2196-05). The viscosity of commercial liquid polyamide resins is typically specified at 40 or 75°C. The
Polyamide resins are supplied commercially in solution or as 100 % reactive liquids depending on the handling and
Fig. 4—Diethylenetriamine.
Fig. 5—Aminoethylpiperazine.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
100
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 7—Imidazoline structure. Fig. 6—Polyamide structure.
viscosity of polyamide resin solutions that may also be measured by this method is generally specified at 25 or 40°C. The viscosity of polyamide resin solutions may also be measured in accordance with ASTM Test Method for Viscosity of Transparent Liquids by Bubble Time Method (D1545-07). A modification of ASTM D1545 is usually used in which the polyamide resin solution is placed in a sample tube, the viscosity is compared to Gardner-Holdt letter standard tubes, and the observation is reported as the alphabetic letter of the tube most closely matching the sample. A plus (+) or a minus (–) is then used to indicate that the viscosity is greater or less than the designated letter.
REACTION OF POLYAMIDE RESINS IN COATINGS
Polyamide resins react with epoxy resins in several stages to form a complex insoluble cross-linked matrix. The initial reaction is between the terminal primary amine groups of the polyamide resin and the oxirane ring of the epoxy resin. The active hydrogen opens the ring and the oligomers join end to end. This initial reaction can be represented by the following simplified structure (Fig. 8). The reaction of the oxirane ring and active hydrogen, which also occurs at the secondary amine sites in the polyamide resin, is one mechanism for the cross-linking that occurs. A secondary reaction occurs between pendant hydroxyl groups in the epoxy resin molecule and other oxirane rings present, which is another mechanism for the cross-linking reaction [3]. Because of this later reaction, it
is desirable to mix the polyamide resin and epoxy resin in less than a stoichiometric ratio to provide coatings with maximum cross-link density. Such coatings will be characterized by excellent impact and chemical resistance but will tend to be less flexible. To provide more flexible coatings with greater elongation, the use of close to or greater than the stoichiometric ratio is recommended. Epoxy/polyamide coatings contain organic solvents, which when applied appear to dry because of solvent evaporation. This early dry time is not, however, an indication of cure. Cure results from a chemical reaction between the liquid polyamide resin and epoxy resins that generally requires about 8–10 h before the film will resist mechanical deformation. The rate can be accelerated by adding a catalyst, such as 2,4,6-tri(dimethylaminomethyl)phenol, to achieve a 4–6 h cure. Chemical resistance of the coating is not achieved before 3–4 days, and ultimate cure is achieved after about three weeks with ambient curing. Cure rates are faster at elevated temperature and become slower as application temperatures decrease. Below 50°F, liquid polyamide resins cure very slowly, or often will not react with epoxy resins. Epoxy/polyamide coatings may be applied by any conventional commercial applicator, including spray, brush, or roller. The coatings formulator may design the solvent system for a particular type of applicator. Epoxy/polyamide coatings may be applied to wood, concrete, or steel. For optimum adhesion to the substrate, the surface to be coated should be thoroughly cleaned and degreased. Epoxy/polyamide coatings are not normally applied to wood: When so
Fig. 8—Polyamide resin reaction with epoxy resin. Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 12
used, the wood surfaces should be first cleaned thoroughly and any old loosely adhering paint removed. Concrete surfaces may be chemically acid etched or mechanically brushed. Steel surfaces should be sandblasted, if possible, or at a minimum should be wire brushed and chemically cleaned [4,5].
ENVIRONMENTAL/TOXICITY CONSIDERATIONS
In response to environmental concerns over the emission of organic compounds into the atmosphere, many coatings manufacturers have begun to produce high solids coatings that contain much lower levels of VOCs. The VOC content is measured in accordance with ASTM Standard Practice for Determining VOC Content of Paints and Related Coatings (D3960-05). These products contain low molecular weight polyamide resins or polyamide adducts often dispersed in benzyl alcohol; or amidoamines, products formulated from monomer fatty acid rather than dimer acid [6]. A significant volume of higher VOC epoxy/polyamide coatings continues to be used. Though liquid polyamide resins are less toxic than aliphatic amines and amine adducts, direct contact exposure with the skin, eyes, and the respiratory system must be avoided. Polyamide resin solutions must also be handled with care to avoid exposure to ignition sources as they contain flammable or combustible solvents and the vapor level from polyamide resin solutions must be monitored in the workplace to avoid overexposure to the organic solvents present. Polyamide resin manufacturers supply material safety data sheets (MSDSs), which should be consulted for hazard information and guidance on the safe use of the products. The MSDS also contains information regarding procedures to follow if a spill occurs, as well as guidelines for hazardous waste disposal. Those polyamide resin solutions that
Q
POLYAMIDES
101
are classified as hazardous waste due to the presence of organic solvents must be incinerated. Liquid (100 %) polyamide resins are not generally classified as hazardous waste though their disposal may be regulated as an oil because of their liquid nature: These products must be either incinerated or absorbed by a suitable solid absorbent medium, such as a ground clay absorbent product, and placed in a sanitary landfill. Though liquid polyamide resins are reactive in the presence of epoxy resin, they are quite stable compounds when kept in a cool, dry environment, and they may remain unchanged for a year or more. Liquid polyamide resins may be stored in phosphatized steel drums or tanks constructed of stainless steel or aluminum. Carbon steel tanks should be avoided because of darkening of the product from iron contamination.
References [1] McMahon, D., and Crowell, E., “Characterization of Products from Clay Catalyzed Polymerization of Tall Oil Fatty Acids,” J. Am. Oil Chem. Soc., Vol. 51, No. 12, 1974, p. 522–527. [2] Lee, H., and Neville, K., “Amides and Miscellaneous Nitrogen Compounds as Epoxy-Resin Curing Agents,” Handbook of Epoxy Resins, McGraw-Hill, New York, Chap. 10, 1967, pp. 2–12. [3] “Epoxy Resins,” Encyclopedia of Polymer Science and Engineering, 2nd ed., John Wiley & Sons, New York, Vol. 6, 1988, pp. 348–354. [4] “Polyamides from Fatty Acids,” Encyclopedia of Polymer Science and Engineering, 2nd ed., John Wiley & Sons, New York, Vol. 11, 1988, pp. 476–489. [5] Allen, R., “Epoxy Resins in Coatings,” Federation Series of Coatings Technology, 1972, Unit 20. [6] Bozzi, E., “Epoxy Resins in High Solids Coatings,” The Epoxy Resin Formulators Training Manual, James Kaszyk, Ed., The Society of the Plastics Industry, Inc., New York, Chap. XIII, 1984, pp. 149–162.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
13
MNL17-EB/Jan. 2012
Polyurethane Coatings Han X. Xiao1 and Joseph V. Koleske2 INTRODUCTION
THE CHEMISTRY OF POLYISOCYANATES WAS FIRST described by Professor Otto Bayer in the laboratories of the I. G. Farbenindustry, today’s Farbenfabriken Bayer, in Leverkusen, Germany. Polyurethanes are mainly characterized by the linkage –NH–C(CO)–O–, though they may also contain other functional groups such as ester, ether, urea, and amide. The most important commercial route for the synthesis of such polymers is the addition polymerization that occurs when di- or higher functionality isocyanates and di- or higher functionality hydroxyl compounds, such as hydroxyl-terminated acrylics, polyesters, or polyethers, are combined and undergo rearrangement reactions. When di-functional reactants are used, linear or thermoplastic polyurethanes are produced. Cross-linked or thermoset polyurethanes are formed if the functionality of at least one of the reactants is greater than 2. The historical and commercial developments as well as the chemistry and applications of polyurethanes have been reviewed by a number of authors [1–23]. Polyurethanes have found extensive applications in the coating industry due to the fact that they exhibit excellent abrasion resistance, toughness, chemical and corrosion resistance, as well as a wide range of useful mechanical properties. Polyurethanes are widely used in coatings, adhesives, sealants, foams, elastomers, and RIM (reaction injection molding, composites, fibers, etc.). Excluding coatings, the 1988 United States consumption of polyurethanes was about 2,750 million pounds (1.25 million metric tons). In 1991, the national market for polyurethane coatings was about 209 million pounds (95,000 metric tons) [24, 25].3 Although the market for polyurethane coatings is large and growing, it is readily apparent that it represents only about 5 %–10 % of the total domestic polyurethane market. The growth in this industry is exemplified by the fact that in 2002 the amount used in the United States was about 5,500 million pounds (2.51 million metric tons), and about 6,390 million pounds (2.90 metric tons) in North America. Reasons for the use of polyurethane coatings include high performance characteristics such as flexibility, toughness, strength, abrasion, chemical, and stain resistance, good light stability when aliphatic isocyanates are used, and good low temperature properties. The latter factor is an important reason for use of polyurethane coatings on plastic substrates.
DEFINITIONS
ASTM [8] in its 2003 document defines urethane coatings as “coatings based upon vehicles containing a minimum of 10 % by weight (nonvolatile vehicle basis) of a polyisocyanate monomer reacted in such a manner as to yield polymers containing any ratio, proportion or combination of urethane linkages, active isocyanate groups, or polyisocyanate monomer. The reaction products may contain excess isocyanate groups available for further reaction at time of application or may contain essentially no free isocyanate as supplied.” ASTM has further classified such polyurethanes into six general types [8]: “Type I, one-package prereacted—urethane coatings characterized by the absence of any significant quantity of free isocyanate groups. They are usually the reaction product of a polyisocyanate and a polyhydric alcohol ester of vegetable oil acids and are hardened with the aid of metallic soap driers.” The curing cross-linking reaction functions by means of an oxidation of double bonds present in the system, that is, the same reaction that takes place with drying oils. For example, linseed oil and glycerol may be first reacted and then modified with a diisocyanate that reacts with a part or all of the available hydroxyl groups. If any residual isocyanate is present, it is removed by addition of a monofunctional alcohol. Catalysts such as dibutyltin oxide and dibutyltin dilaurate are used to promote urethane-linkage formation. Type I urethane coatings are often used as wood and floor finishes because they provide improved scuff, water, and stain resistance over those of conventional alkyds. “Type II, one-package moisture cured—urethane coatings characterized by the presence of free isocyanate groups and capable of conversion to useful films by the reaction of these isocyanate groups with ambient moisture.” The curing mechanism results in mainly urea linkages forming by water molecules reacting with free isocyanato groups. The final coating is a polyurethane/polyurea coating. The rate of cure depends on ambient humidity and the presence of certain tertiary aminecatalysts that accelerate the isocyanato-water reaction. For the most part, type II
University of Detroit Mercy, Polymer Institute, 8200 W. Outer Drive, Detroit, MI 48219. 1513 Brentwood Road, Charleston, WV 25314-2307. 3 The purpose of this chapter is not to give current market information, and the data used are used for illustration purposes. The numbers change and are usually quite far behind any current date. Interested readers might consider browsing various websites, particularly national government sites, for detailed information regarding production figures. 1 2
102
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 This standard is for EDUCATIONAL USE ONLY. www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
Downloaded/printed byASTM International Copyright © 2012 by
CHAPTER 13
urethane coatings are clear, but pigmented systems are becoming somewhat popular. They are often used as sealers for concrete and wood as well as floor and deck finishes. “Type III, one-package heat cured—urethane coatings that dry on cure by thermal release of blocking agents and regeneration of active isocyanate groups that subsequently react with substances containing active hydrogen groups.” Because of the mechanism of cure—release of blocking agents—these urethane coatings are stable at room temperature. The deblocking reaction takes place at elevated temperatures that usually are greater than 150–160°C, releasing active isocyanato groups that react with active hydrogen groups contained in the formulation. Type III coatings will also cure at lower temperatures and at short times if a catalyst is included in the formulation. The coatings are often used in coil coatings and electrical wire coatings. “Type IV, two-package catalyst—urethane coatings that comprise systems wherein one package contains a prepolymer or adduct having free isocyanate groups capable of forming useful films by combining with a relatively small quantity of catalyst, accelerator, or cross-linking agent such as a monomeric polyol or polyamine contained in a second package. This type has limited pot-life after the two components are mixed.” Two groups of catalysts are usually utilized with these coating systems. One group is reactive in nature and is comprised of molecules that contain hydroxyl groups such as alkanolamines. The other group are nonreactive catalysts, such as tertiary amines and metal salts of carboxylic acids. These coatings are not widely used. “Type V, two-package polyol—urethane coatings that comprise systems wherein one package contains a pre-polymer or adduct or other polyisocyanate capable of forming useful films by combining with a substantial quantity of a second package containing a resin having active hydrogen groups with or without the benefit of catalyst. This type has limited pot-life after the two components are mixed.” The compounds containing an active hydrogen group are usually low to medium molecular weight polyols with, for example, a polyester, polylactone, polyether, or polyacrylic backbone. These coatings, which are usually high-solids in nature, are used in high performance areas, such as automobile re-finish coatings, original automotive equipment clear coats over pigmented decorative coatings, aircraft, bus, and bus coatings, and industrial-structure maintenance coatings. “Type VI, one-package, nonreactive lacquer—urethane coatings characterized by the absence of any significant quantity of free isocyanate or other functional groups. Such coatings convert to solid films primarily by solvent evaporation.” Basically, these urethane coatings are solutions of high molecular weight polyurethanes (weight-average molecular weight of about 40,000–100,000) with thermoplastic properties. They are characterized by the absence of nil or essen-
Q
POLYURETHANE COATINGS
103
tially nil free isocyanato groups, and properties are developed by controlled solvent evaporation. In addition, they are low solids, about 10 %–15 %, in nature because of the high molecular weights involved and concomitant high viscosity of such molecules in solution. The films resulting from type VI urethane coating solutions have very high gloss and are used in the textile and furniture industries to achieve the “wet look” that was popular in the late 1970s. They are currently used in the cast transfer-process fabric coatings as well as other fabric coatings. There are other polyurethane coatings besides these six types included in the ASTM classifications. They may be described as follows. Two-package polyurea and poly(urethane-urea) coatings, which are composed of one package that contains amines along with fillers, pigments, and additives and a second package that contains monomeric multifunctional isocyanates and/or prepolymeric adducts of diisocyanates [26–33]. If the second package contains no pre-polymers, it will produce polyurea coatings after the two packages are blended and reaction takes place. If there are prepolymers in package 2, the resultant coatings will be poly(urethaneurea) in nature. In these systems, the isocyanates react very rapidly with amines to form urea groups, which have very good adhesion and mechanical properties. The success achieved when preparing this type of coating depends on the following alternatives: 1. The use of hindered aromatic amines to slow down the reaction that forms the urea groups. 2. The use of aliphatic isocyanate-based prepolymers or adducts followed by reacting with aromatic amines or long-chain aliphatic amines to slow down urea formation that takes place when or if aromatic isocyanates are used in prepolymer preparation. 3. The use of especially designed spray systems in which a small, powerful mixing chamber with high efficiency and very short residence time is employed. A solvent-free, 100 % solids polyurea can be formulated by proper selection of a low viscosity, multifunctional isocyanate such as polyarylpolyisocyanate, often termed PAPI, or the adduct/trimer of hexamethylenedisocyanate (HDI). In addition, to achieve such a goal, an amine with low viscosity such as polypropylene diamines is needed. Waterborne polyurethane coatings, which are mainly anionic and cationic in nature, though nonionic types are available [33–59]. Most commercial waterborne polyurethane coatings or waterborne polyurethane dispersions (WPUDs) are anionic in nature and a few are cationic in nature. The anionic WPUDs are made in the following manner: 1. Preparation of isocyanate-terminated and carboxylcontaining prepolymers from polyols, diisocyanates and dimethylolpropionic acid (DMPA). 2. Neutralization of the above prepolymers with an organic base such as triethylamine to form the pendant, internal hydrophilic salt groups. 3. Dispersion of the above neutralized isocyanateterminated prepolymers with water. 4. Chain extension of the above prepolymer dispersion to introduce urea groups.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
104
PAINT AND COATING TESTING MANUAL
Q
A large number of anionic WPUDs with extensive mechanical properties can be produced by changing the type of polyols, diisocyanates, neutralizing agents, and chain extenders as well as the acid group concentration from DMPA. If desired, the anionic WPUDs can be crosslinked with melamine compounds, azeridine, polyisocyanates, and blocked polyisocyanates. These could be made as one- or two-package systems. Cationic WPUDs are prepared in a similar manner as the above with the following changes: 1. A tertiary amine-containing diol such as N-methyldiethanolamine is used to replace DMPA. This introduces a tertiary amine group into the polymer backbone. 2. Inorganic or organic acidic compounds are used to neutralize the tertiary amines and form the internal hydrophilic salt group. 3. In the chain extension step, care should be taken to avoid destroying the hydrophilic internal salt groups. Various cationic WPUDs can be produced by changing the nature of the raw materials and the concentration of tertiary amine groups. Nonionic WPUDs are also used. They have a preparation procedure that is similar to that of the anionic/cationic WPUDs except that an internal hydrophilic nonionic group is introduced to replace the internal ionic groups. In comparison to the anionic/cationic WPUDs, nonionic WPUDs have excellent stability in any reasonable range of pH values and are commercially available. WPUDs are suitable for a variety of applications including automotive, furniture, textile, wood, leather, paper, nonwoven fabrics, and construction coatings. Polyurethane powder coatings [33,60,61], are finely divided, powdered polyurethanes prepared from polyols, difunctional isocyanates, urethane-modified polyesters, and hydroxyl-containing polyacrylics. They are usually cured with melamines or blocked isocyanates. When polyurethanes are prepared for this industry, factors such as ability to convert the polymer to a finely divided state, the final flexibility or impact strength, glass transition temperature, and sintering characteristics are factors that need to be considered. High glass transition temperatures, which might enhance powdering characteristics, may have a deleterious effect on impact strength. In contrast, a low glass transition temperature might result in good impact strength but have a negative effect on powdering factors and cause sintering difficulties. The key to successful powder coatings, polyurethane or other, is related to the ability to balance molecular weight and its concomitant effect on melt viscosity, glass transition temperature, and a crosslinking mechanism that is stable under storage conditions and that is not affected to any significant degree until flow and leveling takes place at the curing temperature. The major end uses for polyurethane powder coatings are the major appliance markets and automotive. Interpenetrating polymer network (IPN) coatings [62–77] are a relatively novel type of polymer alloys that consist of two or more cross-linked polymers, which in the general case of IPNs need not be polyurethanes. They are more or less intimate mixtures of two or more distinct cross-linked polymer networks held together by permanent entanglements and with only accidental covalent bonds between two networks—for example, they are polymeric “catenanes.” IPNs are produced either by swelling a
15TH EDITION
cross-linked polymer with monomer and cross-linking agent of a different polymer and curing these compounds in situ or by blending the linear polymers, prepolymers, or monomers in some “liquid” form—solution or bulk— together with cross-linking agents and simultaneously curing the component polymers. Combination of various chemical polymer types into IPNs results in different compositions that have controlled morphologies and synergistic behavior. Xiao and Frisch et al. [67–77] have prepared many different types of IPN coatings. Sperling et al. [64,65] have developed two-layer coating systems termed “silent paint,” which is capable of attenuating noise and vibration over a broad temperature range.
CHEMISTRY AND REACTIONS Basic Urethane Chemistry
The high reactivity of the isocyanate, especially with nucleo-philic compounds, has always been of intriguing interest for the organic chemist. However, multifunctional isocyanates only gained technical importance through polyaddition chemistry. The reactivity of the —N=C=O group is mainly determined by the pronounced positive charge of the carbon atom in the double bond sequence that consists of nitrogen, carbon, and oxygen. The positive charge on this carbon atom becomes obvious if one considers the resonance structure as shown below [33]:
With R an aromatic radical, the negative charge can be localized into it as described. This also explains that aromatic isocyanates have higher reactivity than aliphatic isocyanates. In the aromatic isocyanates, the electrondonating substituents lower reactivity of the isocyanato (NCO) group. Therefore, the major chemical reactions of the isocyanato group are as follows.
URETHANE
Hydroxyl-containing compounds can be reacted with NCO groups to form the major and important urethane structure, which is the main structure in polyurethane products.
ALLOPHANATE
The urethane that has formed is capable of further reaction with an NCO group to form an allophanate linkage and this results in crosslinking as described below.
UREA
If an amine or water is reacted with an isocyanate, urea groups are formed in the following manner.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 13
BIURET
If a urea group is reacted with an isocyanate, the biuret compound is formed.
Q
POLYURETHANE COATINGS
105
POLYCARBODIIMIDE
Polycarbodiimide molecules can form when multiple isocyanates react with themselves.
AMIDE
When carboxylic acids react with an isocyanate, amides are formed.
URETDIONE OR ISOCYANATE DIMER
Isocyanates can react with themselves to form isocyanate dimers or uretdione.
ISOCYANURATE OR ISOCYANATE TRIMER
Six ASTM Conventional Type Polyurethane Coatings
The major chemical reactions that take place with the six types of polyurethane coatings that are defined by ASTM [8] are as follows. Type I polyurethanes involve a transesterification that introduces the hydroxyl group and yields the di- or monoglycerides that subsequently will be reacted with diisocyanates at NCO/OH ratios equal or greater than 1.0/1.0. This forms urethane modified drying oils—the generalized urethane oil described below wherein R is an unsaturated, aliphatic chain of drying oil and R′ is the aromatic diisocyanate.
A six-member ring structure is formed when three isocyanate groups react to form an isocyanurate or isocyanate trimer.
CARBODIIMIDE
Carbodiimides are formed when two isocyanates groups react and carbon dioxide is eliminated.
URETONE-IMIME
Three isocyanate molecules can react and form the Uretoneimime molecule.
Type II polyurethanes involve isocyanate terminated prepolymers that are reacted with moisture (water) in the presence of catalyst to form amines and carbon dioxide. The resultant amines then react with other isocyanate groups to form ureas that also can further react with isocyanate groups to form biuret linkages (crosslinks) as described in the following reactions of prepolymer.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
106
PAINT AND COATING TESTING MANUAL
Q
Type III ASTM polyurethanes are prepared by blocking the isocyanate groups with a blocking agent that has an active hydrogen atom, BA-H in the following scheme, and then blending this blocked isocyanate with a polyol, fillers, pigments, and additives. The polyurethane system prepared is a stable one-package system that is cured by deblocking and freeing the isocyanato groups at elevated temperature. These freed groups then react with the hydroxyl groups and form the final polyurethane.
Type IV ASTM polyurethanes are produced with the same chemistry as was used for Type II polyurethanes. Type V ASTM polyurethanes with or without multifunctionality isocyanates (greater than two) are cured with polyols or urethane/urea linkage-containing polyols with or without multifunctionality (greater than two); however, one of the components with have a functionality greater than 2. The following is one example of the urethane-formation reaction.
15TH EDITION
an NCO/OH ratio at 1.0 or very close to this value. If this ratio exceeds 1.0 by any amount, there is the possibility of producing a polyurethane that will contain gel particles when dissolved.
Thermoplastic polyurethanes are used in type VI lacquers as well as in many industrial end uses that require solid polyurethanes. The solid polyurethanes have a similar chemistry except that short chain diol extenders, such as 1,4-butanediol or 1,6-hexanediol, are used and these can form blocks of hard segments with the polyol portion known as the soft segment. In effect, these useful polyurethanes are block copolymers of the (AN)n type. Thus, one of the blocks is a relatively long, number-average molecular weight of 300 to 3,000, polyether or polyester that forms the soft or flexible segment. The other block is formed by the reaction of a diisocyanate and the chain extender and is termed the hard segment. Again, the overall ratio of isocyanate to hydroxyl is maintained at 1.0 to allow thermal forming. The hard segments act as pseudo cross-links and the result is a tough, strong, elastomeric macromolecule. In a mole sense, these polyurethanes can be viewed as polyol/diisocyanate/short-chain extender formed in an equivalents ratio of 1/X/(X–1). The number X can vary from 1 or less to as much as 20 or more, though more typically in coatings, X has a value of one or less to about 3 or 4 [32,33,78]. Because of solubility characteristics, a ratio of about 1/2/1 is often used. A small excess of hydroxyl groups is used to keep final free isocyanate content and storage reactivity at a nil level. When the wide range of X values, the types of isocyanates, the types and molecular weight of polyols are considered, it is readily apparent that a myriad of polyurethanes can be prepared and that a broad range of mechanical and chemical properties can be achieved. The chemistry is basically that of isocyanates reacting with hydroxyl groups to form urethane linkages.
Waterborne Polyurethane Coatings Type VI ASTM polyurethanes are high molecular weight, thermoplastic polyurethanes that are formed with
Anionic waterborne polyurethane coatings or WPUDs [34– 59,78] are prepared by means of four main chemical reactions as described below. Preparation of NCO-terminated and pendant COOH-containing prepolymer:
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 13
Q
POLYURETHANE COATINGS
107
NEUTRALIZATION WITH TRIETHYLAMINE (TEA)
DISPERSION WITH WATER FOLLOWED BY CHAIN EXTENSION WITH DIAMINE
The chemistry involved in cationic WPUDs is similar to that used for anionic WPUDs except that dimethylolpropionic acid is replaced with N-methyldiethanolamine
followed by neutralization with an acid such as acetic acid. The prepolymer is then dispersed in water and chain extended with a diol.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
108
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Radiation-Curable Coatings
Powder Coatings
These reactions that lead to urethane acrylates are almost always carried out in an inert solvent. The reactions, as depicted above, have been idealized. In all commercial and laboratory preparations, there is a significant amount of reaction between the ingredients and as a result chain extension and accompanying molecular weight increases take place. This causes the final products to have a markedly higher than expected viscosity. Oligomeric compounds such as these are formulated with triacrylates such as trimethylolpropane triacrylate and various multifunctional acrylates to provide cross-linking, with monomeric acrylates, N-vinyl pyrrolidone and other low molecular weight compounds to provide viscosity reduction. In effect, 100 % solids systems that will rapidly cure when exposed to actinic radiation. In most formulations, the urethane acrylate is considered as the main ingredient that contributes to mechanical properties in the cured film. When the actinic radiation is ultraviolet in nature, a photoinitiator (for example, 2,2-diethoxyacetophenone or benzophenone in combination with an amine synergist) is added as a freeradical source. Electron beam curable formulations do not require a photo-initiator. Acrylated polyurethanes can be prepared as waterbased coatings [88]. The anionic and cationic oligomeric materials are prepared by introducing an internal salt in the backbone similar to the procedures described above for conventional waterborne polyurethane dispersions. Radiation-cured polyurethanes are often used on plastic substrates that will tolerate only low or moderate temperatures such as clear overprint lacquers on vinyl decals, electronic circuit boards, “no wax” vinyl flooring, tile, wood flooring, packaging, and a host of other end uses. Although radiation-cured, colored, and pigmented coatings and inks are widely used in the marketplace, it should be readily appreciated that radiation penetration is more difficult than in clear coatings.
high that the powder does not block during shipping and storage. The polyurethanes used as powder coating are usually of the ASTM Type III. The main end use for powdered polyurethanes is in the major appliance market—refrigerators, dryer drums, range cabinets, etc.
Radiation-curable polyurethane coatings, in particular the ultraviolet radiation-curable polyurethane-acrylate coatings, are prepared by combination of urethane and acrylate chemistry [79–90]. Acrylate-terminated polyurethanes are used in a number of ultraviolet radiation and electron beam curable formulations. The products are commonly termed “urethane-acrylates” or “acrylated urethanes.” They are prepared by first forming an isocyanate terminated prepolymer from a difunctional polyol, and then end-capping the prepolymer with a hydroxy acrylate such as 2-hydroxyethyl acrylate or 2-hydroxyethyl methacrylate.
Polyurethane powder coatings are usually urethanemodified polyesters and polyacrylates that cure at high temperatures [40,41]. High temperatures are required for the powdered polymer to flow and level to the extent needed for a particular end use. The key to successful powder coatings is related to a balance between molecular weight and related viscosity as well as a cross-linking mechanism that is stable under storage conditions and not effected to any significant degree until flow and leveling have taken place at the cure temperature. Another requirement is that the glass transition temperature be sufficiently
RAW MATERIALS Isocyanates
Two types of isocyanates are used to prepare polyurethanes for coating end uses—aliphatic and aromatic [33]. Polymers prepared from either type of isocyanate have excellent chemical and physical properties. Aromatic isocyanate-based products are used in places where weathering resistance, particularly sunlight or ultraviolet radiation resistance, is not important, since these isocyanates will cause discoloration, which almost always manifests itself as yellowing. Yellowing in itself causes a loss of an aesthetic property, but its cause and result do not deleteriously affect mechanical properties. Ultraviolet radiation attacks the labile hydrogen atoms on the aromatic ring structure. When non-yellowing polyurethanes are required, it is necessary to use an aliphatic isocyanate since their structure yields excellent sunlight and ultraviolet radiation resistance. However, it should be kept in mind that aliphatic isocyanates are less reactive and more costly than aromatic isocyanates, and while these factors should be considered, aliphatic isocyanates are very widely used for both interior and exterior applications. The two main aromatic isocyanates currently used are 4,4′-diphenylmethane diisocyanate (MDI) and toluene diisocyanate (TD), which is a mixture of 2,4- and 2,6-isomers. Moisture-cure urethanes and urethane alkyds usually employ TDI, though MDI has some use due to its low vapor pressure. The extensive use of TDI has been gradually, but markedly, reduced because of its toxicity. Xylylene diisocyanate (XDI) is a mixture of aromatic and aliphatic structures in which there are methylene groups between the aromatic
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 13
Q
POLYURETHANE COATINGS
109
ASTM Type III polyurethane coatings, including powder coatings, employ blocked isocyanates that provide roomtemperature-stable, one-package systems that are activated at elevated temperatures. At such temperatures, the molecules dissociate and the blocking agent leaves freeing the isocyanate functionality that then readily reacts with available active hydrogen-containing molecules. The reaction is rapid due to its nature and to the kinetic effect of elevated temperature. Blocking groups are proton donors such as ε-caprolactam, malonic and acetoacetic esters as well as other enolizable esters, ketoximes, phenol, etc. For example, trimethylolpropane (TMP) can be reacted with a diisocyanate and then blocked by reacting the free isocyanate group with a blocking agent, BA-H, such as phenol.
hardness and rigidity to the coating. The long chain polyols function as “soft segments” and impart flexibility. There are four major types of polyols—polyether, polyester, polyacrylic, and hydrocarbon. Polyether polyols are of three many types—poly(ethylene glycol), poly(oxypropylene) glycol, and poly(tetramethylene oxide) glycol with different functionalities and molecular weights. Copolymeric glycols of ethylene oxide and propylene oxide are also available and used. The polyester polyols are aliphatic (both diol and dimer acid used are aliphatic), aliphatic from ε-caprolactone via ring opening of the cycloaliphatic ring, or aromatic (either diol or dimer acids or both are aromatic). These polyols are also available with different functionalities and a variety of molecular weights. Acrylic polyols are various acrylates copolymerized with 2-hydroxyethylmethacrylate or 2-hydroxyethylacrylate. The equivalent weight based on hydroxyl functionality is dependent on the concentration of hydroxy-acrylate used, and in addition to a distribution of molecular weights, these polyols have a distribution of hydroxy-functional acrylate within the various molecules. Hydrocarbon polyols are homopolymers or copolymers of butadiene, isobutene, and isopentadiene with special initiators. Coatings based on these polyols have excellent water resistance and electrical insulation, but they have lower adhesion to polar surfaces because of their very low backbone polarity than the other classes of polyols. However, hydrocarbon-based polyols may have potential as intermediates for primers on thermoplastic polyolefins used in the automotive industries. As would be expected, primary hydroxyl groups react much more rapidly than secondary hydroxyl groups and tertiary hydroxyl groups react slower than either other type. For example, primary hydroxyl groups reacted about 3.5 times faster with phenyl isocyanate than secondary hydroxyl groups and about 200 times faster than tertiary hydroxyl groups [11]. It was also found that n-butanol reacted five times faster with the isocyanate group in the para- or 4-position than with the isocyanate next to the methyl group in 2,4-toluene diisocyanate. This demonstrated that neighboring groups can significantly affect isocyanate reactivity. Polyfunctional amines, which form urea linkages with isocyanates, are also used as chain extenders. Low molecular weight compounds, such as ethylenediamine, are used for this purpose when waterborne polyurethane dispersions are prepared. Amine terminated oligomers based on alky-lene oxides such as poly(oxypropylene)diamine are also available.
ACTIVE HYDROGEN-CONTAINING COMPOUNDS
Catalysts
ring and the isocyanate groups. It is used to some extent, but mainly in Japan. MDI has been modified into a liquid form that has a functionality between 2.0 and 2.5, and it has been used in some coating areas [75]. Recently, Dow Chemical Co. has produced a liquid isomer of MDI that comprises a mixture of 4,4′- and 3,4′-diphenylmethane diisocyanate, and it also has been used in coatings. Crude MDI, polyarylpolyisocyanate (PAPI) that has functionalities between 2.0 and 2.7, is dark brown in color and is low in cost. PAPI in combination with poly(oxypropylene) diamines (Jeffamine™) is used for two-package, solventfree polyurea coatings. Other aromatic diisocyanates, such as naphthalene diisocyanate, have also been used. Aliphatic isocyanates are more costly than aromatic isocyanates, and they are used for urethane coatings that require excellent sunlight resistance with no discoloration. These coatings are used on plastics, automobile signs, and similar outdoor end-use products. The main aliphatic isocyanates are hydrogenated MDI (4,4′-dicyclohexylmethane diisocyanate, H12MDI), HDI particularly in a biuret or trimer form for improved vapor pressure, tetramethylxylylene diisocyanate (TMXDI), isopropenyldimethyltoluene diisocyanate (TMI), mixtures of 2,2,4- and 2,4,4-trimethyl hexam-ethylene diisocyanate (TMHDI), 1,4-cyclohexane diisocyan-ate, and isophorone diisocyanate (3-isocyanatomethyl-3,5,5-trimethylcyclohexyl isocyanate (IPDI) [9]. The HDI trimer and adducts are major raw materials together with acrylic polyols for two-package, solvent-based polyurethane coatings used in automobile applications. The three significant isocyanates in the above listing are H12MDI, IPDI, and TMXDI.
BLOCKED ISOCYANATES
Isocyanate functionality readily reacts with active hydrogencontaining groups such as OH, NH2, NH, COOH, and SH [33]. Polyfunctional compounds such as glycols, triols, tetraols, polyester polyols, polylactone polyols, polyether polyols, acrylic polyols, and hydrocarbon polyols usually supply hydroxyl groups. Low molecular weight compounds or short-chain extenders, such as 1,4-butanediol, in combination with isocyanates, are termed “hard segments” that function as pseudo cross-links in the final polyurethane [13]. In high molecular weight polyurethanes, such as those used for ASTM Type VI products, these chain extenders represent only a few percent of the total polymer molecular weight yet play a significant role in final physical properties. Glycols and low molecular weight triols require relatively large amounts of isocyanates, usually the most costly ingredient, and impart
There are two main types of catalyst used to promote the urethane formation reactions. One type is tertiary aminecontaining compounds and the other is metal salts or metal oxides—in particular organic metal salts. These catalysts most often are used to promote the reaction between isocyanates and active-hydrogen-containing compounds. Only small amounts, on the order of 10 to 100 parts per million, are needed to cause marked increases in reaction rate. Popular catalysts that have been used are dibutyltin dilaurate, stannous octoate, diaza(2.2.2)bicyclooctane, dibutyltin diacetate, bismuth stearate, and zirconium octoate [91].
Additives
Many additives are used in polyurethane coating formulations, and the particular ones used depend on the final
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
110
PAINT AND COATING TESTING MANUAL
Q
TABLE 1—Examples of polyurethane end uses Home Furnishing
Optical Fibers
Drum Dryers
Printed Circuit Boards
Furniture
Sheet Molding Compound
“No Wax” Flooring and Tile
TPO Bumpers
Range Cabinets
Recreational Products
Refrigerators
Baseball Bats
Wood Flooring
Golf Balls
Industrial Maintenance
Golf Clubs
Bridge
Gym Floors
Industrial Buildings
Playground Equipment
Marine Coatings
Toys
Plant Equipment
Golf Balls
Roof Coatings Windows Miscellaneous
Textiles
References
Fabric Coatings
[1] Bayer, O., “Polyurethanes,” Mod. Plast., Vol. 24, 1947, p. 149. [2] Wright, P., and Cumming, A. P. C., Solid Polyurethane Elastomers, Elsevier Publishing Company, Amsterdam, 1969. [3] Bayer, O., Rinke, H., Siefken, W., Orthner, L., and Schild, H., “A Process for the Production of Polyurethanes and Polyureas,” German Patent No. 728,981 (November 13, 1942). [4] Bayer, O., “The Diisocyanate Polyaddition Process (Polyurethanes). Description of a New Principle for Building up HighMolecular Compounds,” Angew. Chem., Vol. A59, 1947, p. 257. [5] Schollenberger, C. S., Scott, H., and Moore, G. R., “A Virtually Crosslinked Elastomer,” Rubber World, Vol. 137, No. 4, 1958, p. 549. [6] Heiss, H. L., Saunders, J. H., Morris, M. R., and Davis, B. R., “Preparation of Polymers from Diisocyanates and Polyols,” Ind. Eng. Chem., Vol. 46, 1954, p. 1498. [7] Baldin, E. J., Cummin, A. S., and Bieneman, R. A., Off Dig., Vol. 30, 1958, p. 1070. [8] ASTM D16, “Standard Terminology for Paint, Related Coatings, Materials, and Applications,” Annual Book of ASTM Standards, Vol. 6.01, ASTM International, West Conshohocken, PA, 2003. [9] Pansing, H. E., “Chemistry and Theory of Polyurethane Coatings,” Off Dig., Vol. 30, No. 376, 1958, p. 37. [10] Toone, G. C., and Wooster, G. S., “Characterization of Polyurethane Foams from Soybean Oil,” Off Dig., Vol. 32, 1960, p. 230. [11] Bristol, F. A., Paint and Varnish Prod., Vol. 52, No. 11, 1962, p. 71. [12] Patton, T. C., Off Dig., Vol. 34, 1962, pp. 342–348. [13] Lowe, A., J. Oil Colour Chem. Assoc., Vol. 46, 1963, p. 820. [14] Tenhoor, R. E., Chem. Eng. News, Vol. 41, No. 5, 1963, p. 94. [15] Sempert, R. E., Official Digest, Vol. 36, No. 475, 1964, p. 16. [16] Gruber, G., Journal Oil Colour Chemists Association, Vol. 48, 1965, p. 1069. [17] Nylen, P., and Sunderland, E., Modern Surface Coatings, Interscience, London, 1965, p. 209. [18] Damusis, A., and Frisch, R. C., Treatise on Coatings, R. R. Myers and J. S. Long, Eds., Vol. 1, Marcel Dekker, New York, Chap. 12, 1967, p. 435. [19] Wells, E. R., Technology of Paints, Varnishes, and Lacquers, C. R. Martens, Ed., Reinhold Book Corp., New York, Chap. 12, 1968, p. 205. [20] Hampton, H. A., Hurd, R., and Shearing, H. J., “Recent Developments in Polyurethanes,” J. Oil Colour Chemists Assoc., Vol. 43, 1969, pp. 96–123. [21] Doyle, E. N., Development and Use of Polyurethane Products, McGraw-Hill, New York, 1971.
Leather
Luggage
Tarpaulins
Magnetic Tape
Upholstery Transportation
Medical Equipment
Aircraft and Aerospace
Safety Glass
Automotive, OEM
Shoes
Automotive, Refinish
Vinyl Decal Overprints
Bed Liners
Wire Coatings
Golf Carts
Plastic Substrates
Motorcycles
Fascia
Railroad Cars
Electronic Parts and Equipment
Vans
applications. Some of the important additives include are anti-oxidants, UV absorbers, wetting agents, anti-sagging agents, dispersants, defoamers, thixotropic agents, adhesion promoters, flatting agents, etc. [92].
MARKETS
are factors that offset cost factors. For example, polyurethanes are replacing poly(vinyl chloride) plastisols as undercoatings and sealants in the automotive and other transportation markets. The ability to use lower coating thickness at equivalent or improved performance makes the applied cost of polyurethane competitive with the vinyl plastisols [93]. The textiles represent a moderate growth area for thermoplastic polyurethane lacquers with their excellent combination of properties as the main driving force for use. These include good elasticity at low temperatures, abrasion resistance, solvent and water resistance, dry cleansability, machine washability, and an ability to be manufactured with a broad variety of tensile/elongation properties [94]. In addition, high performance can be achieved with very thin coatings that do not markedly increase fabric weight or change styling factors such as drape. To decrease volatile organic content, new low viscosity, aliphatic isocyanates [95] and polyurethane polyols [96] are being developed. Although it is not a complete listing, Table 1 is a summary of many end uses for polyurethane coatings.
Apparel
Aerospace
Mast and Spar Varnishes
15TH EDITION
The various types of polyurethanes are used in a number of market areas and end uses [33]. A number of these have been mentioned above. Two features of polyurethane coatings that often have been looked on as disadvantages are cost and special handing of the potentially hazardous isocyanates that are used in manufacturing or as curing agents. However, various industry segments have been able to develop safe handling and use methods that overcome one objection. The very high performance characteristics of polyurethanes, their ability to cure at low baking temperatures, and improved total coating solids, i.e., decreased volatile-organic-compound content, that can be obtained
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 13
[22] Frisch, K. C., Applied Polymer Science, J. K. Carver and R. W. Tess, Eds, American Chemical Society, ORPL, Chap. 54, 1975, p. 828. [23] Solomon, D. H., The Chemistry of Organic Film Formers, R. E. Krieger Publ Co., New York, Chap. 8, 1977, p. 211. [24] Smith, R. M., “Polyurethanes,” Supplement C, Report No. 10C, SRI International, Menlo Park, CA, May 1991. [25] Linak, E., Kalt, F., and Takei, N., “Urethane Surface Coatings,” Chemical Economic Handbook, SRI International, Menlo Park, CA, August 1992, p. 592.8000. [26] Harada, K., Mizoe, Y., Furukawa, J., and Yamashita, S., Makromol. Chem., Vol. 132, 1970, pp. 281–294. [27] Buist, J. M., and Gudgeon, H., Advances in Polyurethane Technology, Wiley-Interscience, New York, 1968. [28] O’Shaughnessy, F., and Hoeschle, G. K., Rubber Chem. Technol., Vol. 44, 1971, p. 52. [29] Brennan, J. P. (1970). [30] Farbenfabriken Bayer, Netherlands Patent No. 7,104,911 (1971). [31] Trapasso, L., “Polyurethane Elastomeric-Shaped Articles Containing Reactive Sites,” U.S. Patent No. 3,627,735 (1971). [32] Frisch, K. C., and Kordomenos, P. “Applied Polymer Science,” ACS Symposium Series 285, R. W. Tess and G. W. Poehiein, Eds, 1985, p. 985. [33] Oertel, G., Polyurethanes Handbook, Chemistry-Raw MaterialsProcessing-Applications-Properties, Hanser Publisher, Munich, 1985. [34] Dieterich, D., Keberle, W., and Wuest, R. J., J. Oil Colour Chem. Assoc., Vol. 53, 1970, p. 363. [35] Dieterich, D., Angew. Makromol. Chem., Vol. 98, 1981, p. 133. [36] Suskind, S. P., “Polyurethane Latex,” J. Appl. Polym. Sci., Vol. 9, No. 7, 1965, pp. 2451–2458. [37] Hill, F. B. Jr., “Polyalkylene Ether Glycol-Arylene DiIsocyanate Elastomer Sponge and Process for Preparing Same,” U.S. Patent No. 2,726,219 (1965). [38] Mallones, J. E., “Stable Polyurethane Latexes and Process,” U.S. Patent No. 2,968,575 (1961). [39] Rembaum, A., J. Macromol. Sci., Chem., Vol. A3, No. 1, 1969, pp. 87–99. [40] Rembaum, A., Baumgarten, W., and Eisenberg, A., J. Polym. Sci., Vol. B6, 1968, p. 159. [41] Somoano, R., Yen, S. P. S., and Rembaum, A., J. Polym. Sci., Vol. B8, 1970, p. 467. [42] Rembaum, A., Rile, H., and Somoano, R., J. Polym. Sci., Vol. B8, 1970, p. 457. [43] Dieterich, D., Bayer, O., and Peter, J., German Patent No. 1,184,946 (1962). [44] Dieterich, D., and Bayer, O., British Patent No. 1,078,202 (1965). [45] Keberle, W., and Dieterich, D., British Patent No. 1,076,688 (1966). [46] Keberle, W., Dieterich, D., and Bayer, O., German Patent No. 1,237,306 (1964). [47] Keberle, W., and Dieterich, D., British Patent No. 1,076,909 (1966). [48] Dieterich, D., Muller, E., and Bayer, O., German Patent No. 1,178,586 (1962). [49] Keberle, W., and Muller, E., British Patent No. 1,146,890 (1969). [50] Witt, H., and Dieterich, D., German Patent No. 1,282,962 (1966). [51] Scriven, R. I., and Chang, W. H., “Water-Reduced Urethane Coating Compositions,” U.S. Patent No. 4,046,729 (1977). [52] Scriven, R. I., and Chang, W. H., “Water-Reduced Urethane Coating Compositions,” U.S. Patent No. 4,066,591 (1978). [53] Scriven, R. I., and Chang, W. H., “Polylactone-Polyurethanes and Aqueous Dispersions Thereof,” U.S. Patent No. 4,098,743 (1978). [54] Mulligan, C., “Water Dilutable Polyurethanes,” U.S. Patent No. 3,412,054 (1968). [55] Liu, W., Yang, S., and Rende, T., Paint and Coating India, 1999, p. 58. [56] Martin, L., Dearth, R., Feng, S., Baumbeach, B., and Kerznar, A., Paint and Coating India, 2000, p. 44.
Q
POLYURETHANE COATINGS
111
[57] Greissel, M., Ind. Paint and Powder, Vol. 10, 2000, p. 22. [58] Feng, S. X., Lumney, P., and Wargo, R., “Effects of Additives on the Performance of Two-Component Waterborne Polyurethane Coatings,” J. Coat. Technol., Vol. 71, No. 897, 1999, p. 143. [59] Tauber, A., Scherzer, T., and Mehnert, R., “UV Curing of Aqueous Polyurethane Acrylate Dispersions. A Comparative Study by Real-Time FTIR Spectroscopy and Pilot Scale Curing,” J. Coat. Technol., Vol. 72, No. 911, 2000, p. 51. [60] Lu, S. P., Xiao, H. X., Frisch, K. C., Witt, F. W., and Ploeg, A. V. D., “Crosslinking Kinetics Studies on IPN Powder Coatings,” ACS Meeting, Washington, D.C., August 1992. [61] Thometzek, P., Freudenberg, U., and Grahl, M., “Maeschneiderte Polyurethan-Pulverlacke fur hochqualitative Beschichtungen,” Powder Coating, 2000, p. 54. [62] Millar, J. R., “Interpenetrating Polymer Networks. StyreneDivinylbenzene Copolymers with Two and Three Interpenetrating Networks, and Their Sulphonates,” J. Chem. Soc., 1960, p. 1311. [63] Frisch, H. L., Lempner, D., and Frisch, K. C., “Crosslinkable Isocyanate Compositions,” J. Polym. Sci., Part B: Polym. Lett., Vol. 7, 1969, p. 775. [64] Sperling, L. H., and Friedman, D. W., “Synthesis and Mechanical Behavior of Interpenetrating Polymer Networks: Poly(ethyl acrylate) and Polystyrene,” J. Polym. Sci., Part A-2, Polym. Phys., Vol. 7, No. 2, 1969, pp. 425–427. [65] Sperling, L. H., George, H. F., Huelek, Volker, and Thomas, D. A., “Viscoelastic Behavior of Interpenetrating Polymer Networks: Poly(ethyl acrylate)Poly(methyl methacrylate),” J. Appl. Polym. Sci., Vol. 14, 1970, pp. 2815–2824. [66] Klempner, D., Frisch, H. L., and Frisch, K. C., “Topologically Interpenetrating Polymeric Networks,” J. Elastoplastics, Vol. 3, 1971, p. 2. [67] Xiao, H. X., Frisch, K. C., and Frisch, H. L., “Interpenetrating Polymer Networks from Polyurethanes and Methacrylate Polymers. I. Effect of Molecular Weight of Polyols and NCO/ OH Ratio of Urethane Prepolymers on Properties and Morphology of IPNs,” J. Polym. Sci.: Ploym. Chem. Ed., Vol. 21, No. 8, 1983, pp. 2547–2557. [68] Xiao, H. X., Frisch, K. C., and Frisch, H. L., “Interpenetrating Polymer Networks From Polyurethanes and Methacrylate polymers. II. Interpenetrating Polymer Networks with Opposite Charge Groups,” J. Polym. Sci.: Ploym. Chem. Ed., Vol. 22, No. 5, 1984, pp. 1035–1042. [69] Cassidy, E. F., Xiao, H. X., Frisch, K. C., and Frisch, H. L., “Three-component Interpenetrating Polymer Networks (IPNs) from Polyurethanes, Epoxides, and Poly(methacrylates),” J. Polym. Sci.: Ploym. Chem. Ed., Vol. 22, No. 10, 1984, pp. 2667–2683. [70] Kordomenos, P. I., Frisch, K. C., Xiao, H. X., and Sabbah, N., “Coating Compositions Based on Acrylic-Polyurethane Interpenetrating Polymer Networks,” J. Coat. Technol., Vol. 57, No. 723, 1985, pp. 22–28. [71] Patsis, A., Xiao, H. X., Frisch, K. C., and Khahtib, S., “Ionomer/ Semi-IPN Coatings From Polyurethanes and Vinyl Chloride Copolymers,” J. Coat. Technol., Vol. 58, No. 743, 1986, pp. 41–47. [72] Tehranisa, M., Ryntz, R. A., Xiao, H. X., Kordomenos, P. I., and Frisch, K. C., “Urethane Acrylic Interpenetrating Polymer Networks (IPNs) for Coating Applications,” J. Coat. Technol., Vol. 59, No. 746, 1987, pp. 43–49. [73] Frisch, K. C., and Xiao, H. X., Polym. Mater. Sci. Eng., Vol. 57, 1987, p. 222. [74] Shah, J., Ryntz, R. A., Gunn, V. E., Xiao, H. X., Frisch, K. C., Feldpausch, A., and Kordomenos, P. I., J. Coat. Technol., Vol. 61, No. 772, 1989, p. 61. [75] Xiao, H. X., and Frisch, K. C., J. Coat. Technol., Vol. 61, No. 770, 1989, p. 51. [76] Shah, J., Rynz, R. A., Xiao, H. X., Gunn, V. E., and Frisch, K. C., J. Coat. Technol., Vol. 62, No. 785, 1990, p. 63. [77] Xiao, H. X., and Frisch, K. C., Advances in Interpenetrating Networks, D. Klemper and K. C. Frisch, Eds, Technomic Publishing Co., Lancaster, PA, 1991, Vol. 3, p. 223. [78] Dormish, J. F., Lau, C., Kinney, C., and Slack, W. E., Adhes. Age, 2000, pp. 33–36.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
112
PAINT AND COATING TESTING MANUAL
Q
[79] Salim, M. S., Polymer Paint Colour J., Vol. 177, No. 4203, 1987, p. 762. [80] Martin, B., Radiat. Curing, Vol. 13, No. 4, 1986, p. 8. [81] Hodakowski, L. E., and Carder, C. H., “Radiation Curable Acrylated Polyurethane,” U.S. Patent 4,131,602 (1978). [82] Johnson, O. B., and Labana, S. S., “Acrylic Rubber-Urethane-Acrylate Paint and Painting Process,” U.S. Patent No. 3,660,143 (1972). [83] Kehr, C. L., and Wazolek, W. R., “Radiation Curable Ink,” ACS Organic Coat. Plast. Prepr., Vol. 33, No. 1, 1973, p. 295. [84] Thomas, D. C., U.S. Patent No. 3,665,625 (1972). [85] Chang, W. H., Canadian Patent 3,655,625 (1972). [86] Smith, O. W., Weizel, J. E., and Trecker, D. J., “Polycaprolactone Production,” German Patent No. 2,103,870 (1971). [87] Tu, S. T., “Recent Advances in Radiation Curing,” The 1078 Modern Engineering Technology Seminar, Twain, China, July 1978. [88] Moss, M., Coatings World, Vol. 3, 1999, p. 33. [89] Peeters, S., and Loutz, J. M., Coatings World, 1998, p. 40. [90] Xiao, H. X., RadTech Report, September/October 1998, p. 27.
15TH EDITION
[91] Seefried, C. G., Jr., Koleske, J. V., and Critchfield, F. E., “Thermoplastic Urethane Elastomers. I. Effectof Soft-segment Variations,” J. Appl. Polym. Sci., Vol. 19, 1975, pp. 2493–2502. [92] Bailey, F. E., and Koleske, J. V., Alkylene Oxides and Their Polymers, Marcel Dekker, Inc., New York, 1991, p. 218. [93] Critchfield, F. E., Koleske, J. V., Magnus, G., and Dodd, J. L., “Effect of Short Chain Diols on Properties of Polycaprolactone Based Polyurethanes,” J. Elastoplastics, Vol. 4, 1972, p. 22. [94] Seefried, C. G., Jr., Koleske, J. V., and Critchfield, F. E., “Thermoplastic Urethane Elastomers. II. Effect of Variations in Hard-Segment Concentration,” J. Appl. Polym. Sci., Vol. 19, 1975, pp. 2503–2513. [95] Wojcik, R. T., Modern Paint and Coatings, Vol. 83, No. 7, 1993, p. 39. [96] Gardon, J. L., “Polyurethane Polyols: Ester-bond Free Resins for High Solids Coatings,” J. Coat. Technol., Vol. 65, No. 819, 1993, pp. 25–33.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
14
MNL17-EB/Jan. 2012
Silicone Coatings1 D. J. Petraitis2
UNIQUE PROPERTIES OF SILICONES THAT MAKE THEM USEFUL AS COATINGS
SILICONE-BASED COATINGS ARE AMONG THE MOST useful materials for a wide variety of applications. Because the basic bond energies of Si–C and Si–O bonds are high, the chemical processes usually associated with aging of coated surfaces are often much slower and in many situations virtually eliminated for silicone coatings. Also, because the Si–C and Si–O bonds are not present in the natural organic world, biocompatibility and resistance to degradation via biochemical and biophysical processes are significantly reduced. In a similar manner, some silicone polymeric coatings and fluorosilicone-based coatings, in particular, have excellent solvent resistance. Silicone coatings based on trifluoropropylmethyl polysiloxanes have resistance to swelling from such agents as gasoline, jet fuel, solvents, and various other reagents. Chemically, highly branched polymeric silicone coatings begin to approach the properties of silica surfaces as the organic pendant content is reduced. As the organic pendant groups are reduced, the SiO4/2 content increases and the chemical resistance increases. Such polymeric coatings can provide physical scratch resistance as well as chemical resistance. Elastomeric silicone coatings, however, do not provide good resistance to strong acids and/or bases. Strong acids or bases, in particular at elevated temperatures, can cause depolymerization of the siloxane backbone, resulting in failure or, in the case of silicone elastomeric coatings, dissolution of the coating. In a similar manner, silicone coatings are resistant to virtually all frequencies of the electromagnetic spectrum. For compliant coatings, silicones are unsurpassed in resistance to hard radiation, such as that from a Cobalt-60 source for doses in excess of 20 Mrd, as well as from the ultraviolet, visible, and infrared frequencies. When combined with their hydrophobicity, oxygen, and ozone resistance properties, silicones provide excellent weatherability characteristics, and when these properties are combined with the resistance to atomic oxygen encountered in low earth orbit conditions, silicone coatings provide protection for organic substrates in various spacecraft applications. Coating various medical devices is another applications area that utilizes the high quality chemical and biochemical performance characteristics associated with silicone coatings. Such coatings are used to encapsulate and seal permanent implants such as heart pacemakers. They have also
been used to coat temporary implants such as catheters and surgical drains. Thin elastomeric silicone coatings are used to provide soft tissue replacements by forming an envelope to encapsulate gels and/or normal saline solutions. Recent applications for biocompatible silicone coatings include drug delivery devices for both transdermal and long-term implantable, controlled-release drug delivery. A final characteristic that makes silicone coatings useful is their inherently low or nonflammability. Typically, silicone elastomeric coatings have been rated SE-l when tested via Underwriters’ Laboratories Flame Test (UL-94). This property makes silicone coatings ideal for conformal coating of various electrical circuits and devices. In the event of catastrophic thermal degradation, the silicone coatings can and do provide a SiO2 “ash” coating that may permit the emergency operation of the electrical device on a shortterm, temporary basis. Lynch et al. [1] have investigated silicone and other coatings as thermal barrier coatings. They found that the only system that met their requirements of protecting a thin steel plate during a direct flame impingement test and withstanding low temperature flexure tests was a fiberglass-polysilicone composite. Other investigators have studied the effect of silicone fabric coatings on mechanical properties when used in glass fabric/polyester composites [2] and on water absorption of such fabrics [3].
FORMS OF SILICONE COATINGS
Silicone coatings are available in various forms ranging from hard, rigid polymers, to compliant elastomeric products, to soft, almost gel-like, character materials. The rigid polymers are typically supplied in a solvent solution and are mixed with curing agents prior to application. Lead and zinc octoates are among the most common curing agents used. The cure process usually requires approximately 1 h at 250°C to attain complete cure. The cross-linking mechanism involves the condensation of silanol groups SiOH + HOSi → Si – O – Si + H2O Specific coating applications include jet engine components, furnace parts, incinerators, high-temperature appliances, and missile coatings. In addition, specific silicone polymers have been designed to mix with organic coatings and paints, providing improved performance under moderate heat environments than are realized by the organic materials alone.
This chapter is from the previous edition. The ASTM documents have been updated and/or added to, selected references have been added, and in certain instances the editor has added small amounts of information. 2 Nusil Technology, 1050 Cindy Lane, Carpinteria, CA 93013. 1
113
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed byASTM International This standard is for EDUCATIONAL USE ONLY. Copyright © 2012 by www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
114
PAINT AND COATING TESTING MANUAL
Q
By varying the R-group to Si ratio, the hardness of the final coatings can be changed. As the R:Si ratio is decreased, the cross-link density of the polymeric mass effectively increases. Similarly, variation of the R group itself can result in somewhat different flexibility and other physical properties. Properly designed and condensed resins can be formulated to provide hard, silica-like, abrasion-resistant coatings. Such coatings often involve the condensation of alkoxy groups with silanol groups as well as condensation between silanol groups alone. Technology to minimize shrinkage and maximize adhesion during the cure needs to be incorporated to prevent cracking and subsequent flaking of the coating from the substrate. Amino-functional alkoxy silanes are often incorporated into the formulation to simultaneously optimize cure rate and adhesion. Silicone elastomeric coatings incorporate the use of polymeric siloxanes with appropriate cross-linking agents to provide compliant, flexible coatings. Among the cure mechanisms that result in elastomers are the following: 1. SiOH + HSi → Si–O–Si + H2 (Sn catalyst) 2. SiOH + CH3COOSi → Si–O–Si + CH3COOH (Sn catalyst and presence of water) 3. SiOH + ROSi → Si–O–Si + ROH (Sn catalyst) 4. Si–CH=CH2 + HSi → SiCH2CH2Si (Platinum catalyst) 5. SiOH + R2NOSi → Si–O–Si + R2NOH (Sn catalyst and presence of water) These elastomeric coatings can range from extremely tough, high-strength elastomers to soft gel-like coatings. Typically, the elastomers will have properties within the following ranges: Durometer:
Type 00 = 10 Type A = 70
Tensile strength:
0.34 to 13.8 MPa
Elongation:
50 % to 1500 %
Tear strength:
0.88 to 43.8 kN/m
The properties and the cure systems that are chosen for these elastomeric coatings depend, to a large extent, on the end use and the method of application. For instance, the SiOH + HSi (reaction 1) mechanism is often used to provide release coatings for backing paper used with pressure-sensitive adhesives. The actual coating itself has poor strength but attains its properties by simply impregnating the substrate and imparting its nonadhesive properties. Such coatings are repellent to tacky substances and a high level of repellency is achieved when the coating contains a considerable proportion of diorganosilicone units. Depending on particular end uses, various degrees of repellency are needed and this can be achieved by replacing methyl groups with alkenyl groups in the cross-linked coating [4]. Systems with slight adherent qualities have been developed using radiation-cure systems [5–7]. The acetoxy cure system (reaction 2) is used where onepackage convenience is desired, where relatively slow cure is acceptable, and where the acetic acid given off during the cure is not a problem. The oxime (reaction 5) cure system provides many of the properties of the acetoxy cure system, but results in an oxime-leaving group instead of an acetic acid-leaving
15TH EDITION
group. Among the applications for the oxime cure systems are coatings for electronic components and protection for organic composites to prevent atomic oxygen degradation, and coating of quartz blankets to provide adequate emissivity and reflectivity characteristics for certain thermal protection surfaces on the space shuttle. The alkoxy 2-part (reaction 3) cure system, when combined with certain thermal enhancing fillers, such as iron oxide, glass microballoons, and various fibers, is often used to provide ablative and thermally insulating coatings. Various products incorporating the alkoxy two-part cure system are used to protect surfaces and components exposed to plume radiation from various rocket motors and jet engines. The addition cure system (reaction 4) has characteristics that permit rapid heat-accelerated cure, tough physical properties, virtually nil shrinkage; and, due to the platinum catalyst, the best overall flame resistance. Applications include solar cell protection, particularly for satellites, and bum-through protection for the liners of solid rocket motors. The only negative characteristic of the addition cure system is its susceptibility to inhibition. Because the system contains parts-per-million levels of platinum catalyst, it can be readily “poisoned.” Among the most common inhibitors are sulfur-containing organic rubbers and organo-tin compounds that are often used as plasticizers in plastics and also as catalysts for other silicone coatings. There are other silicone elastomeric cure systems, and one of the most significant applications is to coat fiberglass blankets for fire resistance. Spark protection welding blankets are a common application for peroxide-cured silicone coatings. Since peroxide-cured silicones require higher temperature cures, their usefulness is constrained by the substrate upper temperature limits. Also, selectivity of the specific peroxide is critical to prevent poor cures due to the oxygen inhibition; a characteristic of many peroxides. Another novel silicone elastomeric coating that has been developed is a combination cure involving the ultraviolet photoinitiation via free radical formation to provide cross-linking. This ultraviolet radiation mechanism is often combined with a standard cure mechanism to provide a combination cure. This system provides quick surface cure followed by a slower room temperature cure of unexposed, shadowed areas to ultimately provide a fully cured conformal coating important to the electronic applications area. Processes using the combination cure can be used to minimize the time and space required to hold the coated parts until cure is completed before downstream assemblies can take place. Other cure systems have been developed for silicone elastomers, but they find limited use as coating materials and were generally developed for specific applications such as building sealants or glazing compounds. ASTM C1564, Standard Guide for Use of Silicone Sealants for Protective Glazing Systems, deals with the use of the sealants for building construction. Protective glazing includes applications that are subject to natural disasters, such as earthquakes, hurricanes, and windstorm, as well as forms of forced entry, such as blasts, burglary, and ballistic attack. The most common form for silicone coatings is a dispersion of the silicone in an organic liquid. If the coating is based on a tough elastomeric silicone, the uncured elastomer base is most commonly described as a “dispersion” because it contains insoluble components such as
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 14
high surface area fumed silica for reinforcement and often other solid components such as titanium dioxide pigments for coloration or reflectivity properties. The carrier liquid for these dispersions may include chlorinated hydrocarbons, fluoro-chlorohydrocarbons, and both aromatic and aliphatic hydrocarbons. The dispersions also often include organic liquid blends to provide the proper combinations of flow, evaporation, and application ease. Among the most common liquids for silicone dispersions are 1,1,1-trichloroethane, VM&P naphthas, and xylene. Low-molecular-weight alcohols, such as ethanol and isopropanol, and ketones, such as acetone, are not suitable because silicones are generally incompatible with these lower-molecular-weight oxygen-containing solvents. Fluorosilicones require the use of such solvents as methyl ethyl ketone and methyl isobutyl ketone for adequate dispersion. Fluorosilicone-dimethyl copolymer-based silicones can be dispersed adequately in 1,1,1-trichloroethane for thin layer application. True solutions can be made if the silicone contains no insoluble components. For example, true solutions can be made for unfilled silicones or for silicones that are polymer reinforced. These coatings have limited use, however, because the final cured elastomeric coating lacks the overall toughness of filled materials. Recent developments have resulted in silicone coatings that have not involved the use of solvents. Because of environmental concerns, the use of solvent carriers for dispersions and solutions has become less desirable. In particular, fluorochlorocarbons and chlorinated hydrocarbons, despite their low toxicity and nonflammability, are being phased out because of international agreements. Similarly, hydrocarbon solvents are undesirable because of their flammability, toxicity, and environmental effects. Silicone-based conformal coatings have been developed without solvent carriers. However, thin layer applications are difficult unless the viscosity is low enough to permit proper coating. Unfortunately, the technology for high-strength, low-viscosity, 100 % solids silicone coatings does not exist. The current products, therefore, when cured, are very low strength and do not provide coatings that are resistant to handling. Research is ongoing to develop water-based dispersions, but to date, the demonstrated physical properties, although higher than the 100 % solids coatings, are significantly less than the solvent-based silicone coatings.
METHODS OF APPLICATION
The method used for application of silicone coatings to substrates depends on the article being coated and the specific type of silicone being used. Dipping, spraying, and brush painting are the most common types of application. The thinnest coatings result from spraying of two-solvent dispersions utilizing standard aerosol spray guns. Needless to say, experience involving aerosol spraying is critical for acceptable coatings. Among the variables to consider are viscosity, solvent, percent solids, pot life, and cure system choices. The most securely sealed surface layer is accomplished by dip coating. Again, variables including solvent, bath life, and cure systems must be optimized. Additionally, the evaporation of solvent during the dip processing needs to be compensated for by periodically or continuously adding make-up solvent to maintain optimal dip-bath viscosity. If a one-part, humidity-actuated cure system is used for
Q
SILICONE COATINGS
115
film formation, consideration must be given to provide a dry, inert blanket over the bath to prevent a partially cross-linked elastomeric skin from forming. Dry argon is often utilized to prevent moisture in the air from reacting with the silicone base coating. Another consideration for the dip coatings is the possibility of air bubble inclusion. Again, several variables need to be considered. Low viscosity, controlled immersion and withdrawal rates, and vibration of the bath and/or object to be coated can be used to minimize bubble entrapment. Similarly, two distinct liquids with different rates of evaporation are often used to ensure uniform coating with minimal drip regions and minimal bubble formation. Painting or brush coating substrates is yet another method to apply a uniform silicone coating. Painting, however, is usually not applicable for either large areas or mass production coatings. For brush application, virtually all of the variables discussed in the above dipping and spraying also apply. Regardless of the methods of application, the cure parameters demand significant considerations. Vacuum exposure may be used to remove air bubbles and to ensure flow under surface irregularities or impregnation of porous substrates. Vacuum treatment may also be used to enhance removal of the solvents, but care should be taken to prevent evaporation of the reactive volatile components that would prevent cure even after removal from the vacuum. Of course, most commonly, the vacuum removal of solvent is unwarranted and therefore solvent is merely evaporated at ambient pressures. The solvent evaporation can also be enhanced by air circulation and by the use of thermal energy. However, the application of heat should be limited or applied in a step-wise manner to prevent solvent entrapment below the surface resulting in bubble formation. Also, for one-part silicone coatings that are cured through moisture activation, it is ineffective to use heat acceleration because humidity is obviously reduced in a normal air circulating oven. If accelerated cure is required for one-part coatings, a steam autoclave may be used, but only after all of the carrier solvent is removed.
TESTING CONDITIONS
The test requirements for silicone coatings include MILI-46058C, Insulating Compound, Electrical (for Coating Printed Circuit Assemblies), for qualifying silicone coatings as insulating compounds for electrical coating applications of printed circuit board assemblies. MIL-I-46058C includes the following tests: Curing Time and Temperature Appearance Coating Thickness Fungus Resistance Insulation Resistance Dielectric Withstanding Voltage Leakage Current Testing Q Resonance Q Resonance after Immersion Thermal Shock Flexibility Thermal Humidity Aging Flammability Materials are used in spacecraft applications are evaluated by means of ASTM E595, Standard Test Method for
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
116
PAINT AND COATING TESTING MANUAL
Q
Total Mass Loss and Collected Volatile Condensable Materials from Outgassing in a Vacuum Environment. This test is used to determine the amount of collected volatile condensable material and total mass loss that leave the specimen from a material when exposed to 125°C for 24 h at a vacuum less than 7 × 10−3 Pa (5 × 10−5 torr) and condense on a collector set at 25°C. Basically, the maximum CVCM value for coatings intended for space applications is 0.1 % and the maximum TML is 1.0 %. The coatings intended for satellite applications require these high levels of purity to prevent the contamination of solar cells, optical surfaces, and other sensitive instrumentation. For most silicone materials, extended devolatilization is required for the polymeric components prior to compounding into the finished product. For silicone elastomeric coatings, the physical properties of the cured elastomer are critical parameters. The tensile strength, elongation, and modulus are defined in ASTM D412, Standard Test Methods for Vulcanized Rubber and Thermoplastic Elastomers-Tension. Kim et al. [8] have investigated the interaction of thickness, modulus, and shear rate on adhesion forces involved in silicone coatings. Durometer and tear strength measurements are defined in ASTM D2240, Standard Test Method for Rubber Property-Durometer Hardness, and ASTM D624, Standard Test Method for Tear Strength of a Conventional Vulcanized Rubber and Thermoplastic Elastomers, respectively. The viscosity, nonvolatile content, and specific gravity tests are defined in ASTM D1084, Standard Test Methods for Viscosity of Adhesives, ASTM D2288, Standard Test Method for Weight Loss of Plasticizers on Heating, and ASTM D792, Standard Test Methods for Density and Specific Gravity (Relative Density) of Plastics by Displacement, respectively. Other tests have been developed for silicone coatings to meet the requirements of specific applications. Included among these is the “blocking” test to determine the propensity of silicone coatings to cause “sticking” to contacted surfaces after application and cure. This test involves contact between the silicone-coated surface and the surface to be tested by subjecting the interface to an applied pressure for a fixed time followed by measurement of the force needed to separate the surfaces. A variety of tests have been developed to measure the adhesive force of the coating on the substrate. Again, a number of tests that are oriented toward the specific application have been developed and revised.
SPECIFIC APPLICATIONS FOR SILICONE COATINGS
Among the various applications for silicone coatings is the conformal coating of electronic circuit boards. Because of their previously described stability properties, silicones make ideal conformal coatings. Silicone coatings typically have stiffening points of 65°C, and can be formulated with stiffening points as low as 115°C. This makes them ideal for electronic device protection under extreme environmental conditions. Investigators have developed a thermoformable coating comprised of a thermoplastic film containing microencapsulated silicone that can be vacuum formed onto an object as a removable protective coating for packaging electronic components or equipment [9]. During vacuum forming or melting onto, say, a printed circuit board, there is a release of the silicone from the microcapsules and a film of silicone with a protective thermoplastic top layer is formed. If desired, as in repair, the protective films may be removed by peeling. Removal is complete since the silicone
15TH EDITION
will adhere more tightly to the thermoplastic film than the electronic equipment. Recently, there have been a number of studies that deal with ship-hull and other fouling [10–14] and how silicone coatings can play a role in alleviating the problem. The studies are concerned with the release properties of silicone coatings and how these can eliminate or reduce fouling by barnacles, barnacle larvae, and other organisms or materials that cause fouling. Silicone coatings are used almost exclusively to provide protection from atomic oxygen degradation in low earth orbit (approximately 100–500 miles high). Atomic oxygen degradation is sufficiently significant to cause rapid erosion and degrade organic substrates including epoxides, urethanes, and polyester-based thermoset coatings. Coating protection permits the use of composite materials in space applications where the advantages of high strength and low weight associated with composite materials would be unusable due to their atomic oxygen degradation. The high-temperature stability and excellent dielectric properties of silicone polymers make them ideal impregnate coatings for high-energy capacitors used in jet engine ignitions. The inherent stability of silicone coatings when combined with specific fillers including zinc oxide, titanium dioxide, and zinc orthotitanate are often used to provide the specific emissivity and reflectance required for thermal-control coatings. Similarly, silicone elastomeric coatings that provide ablation protection are produced by the addition of iron oxide, glass or ceramic microballoons, or graphite fibers to the polymeric matrix. Launch vehicles launch equipment, and thrust reversers are often coated with specially formulated silicone ablative coatings. The incorporation of phenyl siloxanes into the basic silicone polymeric species provides increased ablative properties, and various copolymers—including silicone-boranes and silylphenylenes—have been and are being evaluated to provide protection from impingement of high-energy lasers. As previously discussed, the biocompatibility of silicones makes them ideal for medical applications. Coating permanent implants as well as temporary implants with silicones provides improved safety and efficacy. Foley catheters coated with silicone elastomers result in less patient discomfort and reduced infection rates. For similar reasons, temporary pressure-sensitive silicone adhesive coatings are used to provide adhesion directly to the skin. Combinations of silicone coatings are being investigated for use in various drug delivery devices. Specifically layered coatings of silicones impregnated with drugs can be used for transdermal drug delivery. When combined with a silicone pressure-sensitive adhesive, a complete system of controlled drug delivery devices can be fabricated.
NEW REQUIREMENTS FOR SILICONE COATINGS
Research and development efforts continue to provide silicone coatings with even more stringent requirements and specifications. Electrical coatings with semiconducting properties for electronic applications and elastomeric coatings with volume resistivities in the 104–105 Ω cm range have been studied. Silicone coatings with variable electric properties are also being researched. Similarly, silicone coatings that provide specific biological properties are also being developed. Specifically, hydrophilic silicone coatings are being developed for reduced
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 14
thrombogenicity, and microporous coatings are under development for controlled tissue in-growth response.
References [1] Lynch, J. K., Nosker, T. J., Ondre, D., Mazar, M., and Nosker, P., “Development of a Composite Thermal Barrier Coating,” Society of Plastics Engineers Annual Technical Conference Proceedings, Cincinnati, OH, May 2007 (unpublished). [2] Pavlidou, S., Mai, S., Zorbas, T., and Papaspyrides, C. D., “Mechanical Properties of Glass Fabric/Polyester Composites: Effect of Silicone Coatings on the Fabrics,” J. Opt. Soc. Am. A, Vol. 91, No. 2, 2003, pp. 1300–1308. [3] Pavlidou, S., Krassa, K., and Papaspyrides, C. D., “Woven Glass Fabric/Polyester Composites: Effect of Interface Tailoring on Water Absorption,” J. Appl. Polym. Sci., Vol. 98, No. 2, 2005, pp. 843–851. [4] Achenbach, F., Fehn, A., Hechtl, W., and Kinne, M., “Regulating the Release Force of Silicone Coatings Which Repel Tacky Substrates,” U.S. Patent No. 6,046,294 (April 4, 2000). [5] Gordon, G. V., Moore, P. A., Popa, P. J., Tonge, J. S., and Vincent, G. A., “Radiation-Cured Silicone Release Coatings: ‘Sticking Lightly,’” Technical Conference Proceedings of RadTech 2000, Baltimore, MD, 9–12 April 2000 (unpublished), p. 994. [6] Kerr, S. R., III, “Electron Beam Curing of Epoxy-Silicone Release Coatings,” Adhesives Age, Vol. 41, No. 11, 1998, p. 27, 4p, 4 charts, 1 diagram, 2 graphs. [7] Riding, K. D., “Controlled Release Additives in UV Curable Epoxysilicone Chemistry,” Proceedings of RadTech ‘90— North America, Vol. 1, Chicago, IL, 25–29 March 1990, p. 377. [8] Kim, J., Chisholm, B. J., and Bahr, J., “Adhesion Study of Silicone Coating: The Interaction of Thickness, Modulus, and
[9] [10]
[11]
[12]
[13]
[14]
Q
SILICONE COATINGS
117
Shear Rate on Adhesion Forces,” Biofouling: J. Bioadhesion Biofilm Res., Vol. 23, No. 2, 2007, pp. 113–120. Cavallaro, A. and Akesson, J., “Silicone Dispensing with a Conformal Coating,” U.S. Patent No. 7,101,817 (September 2, 2006). Rittschof, D., Orihuela, B., Stafslien, S., Daniels, J., Christianson, D., Chishom, B., and Holm, E., “Barnacle Reattachment: A Tool for Studying Barnacle Adhesion,” Biofouling: J. Bioadhesion Biofilm Res., Vol. 24, No. 1, 2007, pp. 1–9. Stein, J., Truby, K., Darkangelo-Wood, C., Takemori, M., Vallance, M., Swain, G., Kavanagh, C., Kovach, B., Schultz, M., Wiebe, D., Holm, E., Montemarano, J., Wendt, D., Smith, C., and Meyer, A., “Structure-Property Relationships of Silicone Biofouling-Release Coatings: Effect of Silicone Network Architecture on Pseudobarnacle Attachment Strengths,” Biofouling: J. Bioadhesion Biofilm Res., Vol. 19, No. 2, 2003, pp. 87–94. Stein, J., Truby, K., Darkangelo-Wood, C., Stein, J., Gardner, M., Swain, G., Kavanagh, C., Kovach, B., Schultz, M., Wiebe, D., Holm, E., Montemarano, J. Wendt, D., Smith, C., and Meyer, A., “Silicone Foul Release Coatings: Effect of the Interaction of Oil and Coating Functionalities on the Magnitude of Macrofouling Attachment Strengths,” Biofouling: J. Bioadhesion Biofilm Res., Vol. 19, No. 1, 2003, pp. 71–82. Kavanagh, C. J., Swain, G. W., Kovach, B. S., Stein, J., Darkangelo-Wood, C., Truby, K., Holm, E., Montemarano, J., Meyer, A., and Wiebe, D., “The Effects of Silicone Fluid Additives and Silicone Elastomer Matrices on Barnacle Adhesion Strength,” Biofouling: J. Bioadhesion Biofilm Res., Vol. 19, No. 6, 2003, pp.381–390. Sun, Y., Guo, S., Walker, G. C., and Kavanagh, C. J., “Surface Elastic Modulus of Barnacle Adhesive and Release Characteristics from Silicone Surfaces,” Biofouling: J. Bioadhesion Biofilm Res., Vol. 20, No. 6, 2004, pp. 279–289.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
15
MNL17-EB/Jan. 2012
Vinyl Polymers for Coatings Joseph V. Koleske1 PREFACE
IN PREPARATION OF THIS CHAPTER, THE CONTENTS of the 14th edition of this manual were drawn upon. The author acknowledges the author of the chapter in the 14th edition, Richard J. Burns. The current edition will review, alter, and update the topics as addressed by the previous author and ensure that any ASTM documents cited are current.
VINYL POLYMERS FOR COATINGS Definition
The vinyl polymers used in solvent-based coatings, inks, and adhesives are low to- medium molecular weight copolymers of vinyl chloride, vinyl acetate, or other monomers to improve solubility. Functional monomers contribute specific properties; thus, carboxylic acid-containing monomers provide adhesion, while hydroxyl-containing monomers contribute to reactivity, compatibility with other polymeric species, or adhesion to specific surfaces. These modified vinyl chloride copolymers are most often used as thermoplastic, solvent-soluble lacquers, though by formulating with appropriate cross-linking agents and modifiers, air-dry or baking finishes can be produced having thermoset-like properties. Special techniques have been developed that enable the use of high molecular weight vinyl chloride homopolymers as dispersions in organic media called plastisols or organosols that require a heat fusion step to form films or coatings. Vinyl chloride homopolymers and copolymers are also compounded for use as powder coatings that can be applied by either electrostatic spray or fluidized bed techniques. Water-based vinyl chloride polymers and copolymers include high-molecular-weight polymer latexes that require heat to fuse, and also aqueous dispersions of lowmolecular-weight polymers that utilize coalescents to form films at room temperature.
General
Important characteristic features of vinyl polymers/coatings are: (1) relatively high glass transition temperature; (2) excellent resistance to water, alcohols, aliphatic hydrocarbons, vegetable oils, dilute acids, and alkali; and (3) inertness in contact with foods [U.S. Food and Drug Administration (FDA)-listed polymers or copolymers only]. Vinyl copolymer films can be degraded by exposure to high temperatures or by long-term exposure to ultraviolet radiation, with a resultant change in color from clear to amber, red, and, with sufficient exposure, black. Suitable thermal stabilizers are employed to permit the processing of vinyl 1
coatings at high temperature, and proper pigmentation helps to protect vinyl coatings from attack by ultraviolet radiation. Some stabilizer systems can provide limited protection to clear vinyl chloride copolymer films.
History
Between 1912 and about 1929, Ostromislensky pioneered investigations into the polymerization and properties of vinyl polymers, and he also made other valuable contributions to the development of poly(vinyl chloride) [1–4]. Early studies with this polymer showed that it was difficult to process and is thermally unstable. These factors hindered its early commercialization, and it was the development of thermal stabilizers as well as internal (comonomeric) and external plasticizers that opened the commercial door and has led to the wide usage of vinyl chloride polymers. Reid invented the copolymers of vinyl chloride and vinyl acetate [5,6]. In 1933, Davidson and McClure described various applications for vinyl polymers and copolymers including their use as swimming pool coatings [7]. Commercial production of vinyl chloride–vinyl acetate copolymers began in 1936. Carboxyl-modified copolymers were introduced in 1939 and hydroxyl-modified copolymers in 1945. The first commercial use of these vinyl copolymers was in 1936 as a coating used as the sanitary interior lining of beer cans. In general, the above described copolymers are of relatively low to moderate molecular weight. The viscosity requirements of spray and roll coating applications do not permit the use of very high molecular weight vinyl chloride polymers. In about 1943, organosol and plastisol coating technology that allowed use of such very high molecular weight polymers was developed [8,9].
Polymerization
Vinyl chloride monomer is a gas at standard conditions with a boiling point of −13.9°C. Polymerization is carried out in an autoclave under moderate to high pressure. The reaction is typically initiated by free radical-generating compounds such as peroxides. The polymerization is exothermic, and reaction temperature regulation is necessary to control the growth (molecular weight increases) of the polymer. The use of high pressure and low temperature generally favors the formation of very high molecular weight polymers. To control molecular weight, chain transfer agents are commonly employed. The number-average molecular weight (Mn) of commercially available solvent-soluble vinyl chloride homopolymers and copolymers ranges from a low of a few thousand to about 45,000. The Mn of vinyl chloride
1513 Brentwood Road, Charleston, WV 25314.
118
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 This standard is for EDUCATIONAL USE ONLY. www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
Downloaded/printed byASTM International Copyright © 2012 by
CHAPTER 15
Q
VINYL POLYMERS FOR COATINGS
119
polymers used for plastisol and organosol coatings ranges between about 60,000 to 110,000 [10].
suspension vinyl polymers are characterized as spherical in shape with a size between 100 and 300 μm.
Manufacture
Emulsion Polymerization
Vinyl polymers manufactured for the coatings market are made by several processes. Polymerization by the solution and suspension processes is used to make solvent-soluble polymers, while emulsion or dispersion polymerization is used to make much higher molecular weight polymers for plastisols and organosols. Some solvent-soluble grades are also made by the emulsion process. Post-polymerization processes are applied to some copolymers to achieve special properties. The following are brief descriptions of the manufacturing processes. Detailed information is available in many books dealing with vinyl and in particular vinyl chloride-containing polymers In addition, a relatively detailed, but still concise, description of the processes including flow diagrams is available in a government publication [11].
Like the suspension process, emulsion polymerization is also carried out in aqueous media, but in place of watersoluble polymers as stabilizers, surfactants are normally used to form and stabilize the very small monomer droplets associated with this process during polymerization. Another differentiation from the suspension process is that in the emulsion process the initiator is soluble in the media rather than in the monomer droplets. A special form of emulsion polymerization called “dispersion polymerization” uses an oil-soluble rather than watersoluble initiator and produces polymer particles that range from about 0.2 to 2 μm in size. These high molecular weight powdered products are used in plastisol and organosol coatings.
Bulk Process
Post-Polymerization Process
In the bulk polymerization of most monomers, the monomer acts as a solvent for the initiator and the forming polymer with the end product being a solid mass of polymer. However, in the case of vinyl chloride, polymerization takes place in solution only during the early stages. As the polymer particles grow and reach a sufficient size, they precipitate as a fine powder and polymerization continues in the solid phase of these particles. A slurry results and when the reaction is about 50 % completed, it becomes necessary to separate and remove the polymer particles from the slurry. If this is not done, an extremely rapid, dangerous reaction takes place because of poor heat transfer and autoacceleration [12].
Solution Process
Polymerization is carried out in a solvent in a batch or continuous process. The viscosity of the reaction medium increases as monomer is converted to polymer, and the extent of polymerization can be monitored and controlled by viscosity measurements. When the appropriate viscosity is attained, the autoclave polymer solution is stripped of unreacted vinyl chloride monomer, and the polymer is precipitated in a controlled manner by the addition of water, water/ alcohol mixtures, or other precipitant. The slurry is next centrifuged to remove most of the liquid, and the co-polymer is dried in fluid-bed dryers. The particle size of the dried polymers produced by this process ranges from about 75 to about 200 μm. Final particle shape is irregular in nature.
Suspension Polymerization
Suspension polymerization involves the mechanical dispersion of monomer in an aqueous medium. High-molecularweight, water-soluble colloidal polymers are used in small amounts to stabilize the droplets of suspended monomer(s) and to control particle size. The stabilizer used remains with the polymer during and after polymerization and final polymer recovery. The monomers and associated materials (initiator, stabilizer) exist as discrete, small droplets before and during polymerization and form the final polymer particles after polymerization. The initiator is soluble in the monomer mixture. Usually, when solutions of suspension vinyl polymers are prepared, mild heating is required to achieve maximum solution clarity at minimum viscosity. Particles of
Some vinyl-alcohol modified polymers are prepared in a two-step process. The first step consists of the preparation of a vinyl chloride-vinyl acetate copolymer by either a solution or suspension process. Next, the copolymer is dissolved in a suitable solvent and a reactant is added to partially hydrolyze the pendant acetoxy groups and yield a vinyl alcohol moiety.2 The modified polymer is then precipitated from solution and dried as described for the solution process. The vinyl chloride/vinyl alcohol/vinyl acetate or vinyl chloride/vinyl alcohol copolymer thus formed has only secondary hydroxyl groups. These polar groups account for the copolymers’ unique solubility/compatibility properties. These vinyl-alcohol-containing polymers differ from those prepared directly using other hydroxyl-containing monomers in their compatibility with alkyds and in the rate of reactivity with co-reactants such as polyfunctional isocyanates or amino-formaldehyde cross-linking agents.
Applications for Vinyl Chloride-Based Copolymer Coatings
The main solvent-soluble vinyl chloride copolymers available in industry are described in Table 1. These copolymers produced by the solution polymerization process are based on vinyl chloride/vinyl acetate copolymers, and they have the following generalized compositions. 1. Vinyl chloride/vinyl acetate copolymers. 2. Carboxyl-modified vinyl chloride/vinyl acetate copolymers. 3. Hydroxyl-modified copolymers of two types: a. Hydroxyalkyl acrylate modified vinyl chloride/ vinyl acetate copolymers via direct polymerization. b. Vinyl alcohol-modified copolymer derived from vinyl chloride-vinyl acetate copolymer in a postpolymerization reaction process. 4. Carboxyl/hydroxyl modified vinyl chloride/vinyl acetate copolymers 5. Sulfonate modified vinyl chloride/vinyl acetate copolymers. It is well known that vinyl alcohol does not exist as a monomer and thus its use in a direct polymerization step is not possible.
2
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
120
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
TABLE 1—Typical (i.e., not specifications) properties of vinyl chloride copolymers for coatings [13,14]. Specific Gravity of the copolymers vary from 1.34 to 1.39 [determined by ASTM D792, Standard Test Methods for Density and Specific Gravity (Relative Density) of Plastics by Displacement]. Molecular Weight Number Averageb
Glass Transition Temperature°C
Copolymer Composition, wt %
Reactive Functionality
Vinyl Chloride
Vinyl Acetate
Other Monomer
Type
wt %
Inherent Viscositya
Centipoise
Solids, wt %
Solvent, MEK/ toluene
90
10
. . .
None
. . .
0.74
250
15
67/33
44 000
79
86
14
. . .
None
. . .
0.50
200
20
50/50
27 000
72
86
14
. . .
None
. . .
0.40
175
25
33/67
22 000
72
80
13
1
Carboxyl
1.0
0.50
150
20
50/50
27 000
74
83
16
1
Carboxyl
1.0
0.38
250
25
25/75
19 000
72
81
17
2
Carboxyl
2.0
0.32
370
30
25/75
15 000
70
90
4
6
Hydroxyl
2.3
0.53
350
20
50/50
27 000
79
90
4
6
Hydroxyl
2.3
0.44
400
25
50/50
22 000
77
81
4
15
Hydroxyl
1.8
0.56
171
20
50/50
33 000
70
81
4
15
Hydroxyl
1.9
0.44
184
30
50/50
24 000
65
81
4
15
Hydroxyl
2.0
0.30
340
30
25/75
15 000
65
82
4
14
Hydroxyl/ carboxyl
2.0
0.56
170
20
50/50
35 000
72
85
13
2f
Sulfonate
1.0
0.33
500
20
50/50
17 000
72
c c c d d e e e c,e
Solution Viscosity
Determined by ASTM D1243, Standard Test Method for Dilute Solution Viscosity of Vinyl Chloride Polymers. Based on a polystyrene standard. c Maleic Acid. d Vinyl Alcohol. e Hydroxyalkyl acrylate. f Sulfonate-containing monomer. a
b
Soluble polymers similar to those described in Table 1 and prepared by either a solution or suspension process are available from various suppliers.
Food and Drug Administration Considerations
FDA regulations for various end uses list vinyl chloride copolymers by chemical identity. Designations for adhesives and coatings used on food contact surfaces are listed in Table 2. Particular copolymers are listed in these documents by chemical identity are for use on metal and paper substrates that are used as food contact surfaces of articles used in processing, manufacturing, packing, producing, heating, packaging, holding, or transporting food, or as components of closures with sealing gaskets for food containers. Vinyl chloride/acetate copolymers, hydroxyl-modified vinyl chloride-acetate copolymer, and several other vinyl chloride copolymers made with monomers having acid or ester functionality are commercially available. Regulations such as these are subject to change or expansion, so users should always search for up-to-date FDA information.
Vinyl Chloride Copolymers—Analysis
There are many references to chemical methods for identifying and characterizing vinyl chloride copolymers [15].
Infrared spectra of vinyl polymers are very useful for qualitative and quantitative purposes. Spectra of neat vinyl polymers can be found in a variety of sources such as atlases, encyclopedia of plastics, or specific papers dealing with the subject [16–18]. Also, several ASTM documents deal with the identification and characterization of vinyl polymers used in coatings materials. ASTM D2621, Standard Test Method for Infrared Identification of Vehicle Solids From Solvent-Reducible Paints, details the qualitative characterization of separated paint vehicle solids by infrared spectroscopy. An analysis spectrum for an ortho-phthalic alkyd, vinyl chloride-acetate modified vehicle is detailed. ASTM D2124, Standard Test Method for Analysis of Components in Poly(Vinyl Chloride) Compounds Using an Infrared Spectrophotometric Technique, presents methods through which vinyl systems can be separated into components including polymers, copolymers, plasticizers, stabilizers, and fillers. Each component can then be analyzed by infrared techniques. A particularly useful document was ASTM D4368, Standard Guide for Testing Poly(Vinyl Chloride) Resins. Unfortunately, this standard guide was withdrawn in 2005. The ASTM website indicates that a working document
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 15
Q
VINYL POLYMERS FOR COATINGS
121
TABLE 2—United States Food and Drug Administration (FDA) regulations [14] FDA Regulationa
Intended Use
21CFR 175.105
In adhesives used in articles intended to be used for holding, packaging, or transporting food
21CFR 175.300(b)(3)(XV)
In polymeric coatings used as continuous films for food contact surfaces intended for use in heating, holding, manufacturing, packaging, packing, processing, producing, or transporting food.
21CFR 175.320
As coating components for coatings to be applied as a continuous film on one or both sides of a base polyolefin film that is in compliance with 21CFR 177.1520.
21CFR 176.170(b)
Component of coatings on food contact surfaces of paper and paperboard used to package aqueous and fatty foods.
21CFR 176.180
Component of coatings on food contact surfaces of paper and paperboard that will be in contact with dry food.
21CFR 177.1210
Components of coatings used on closures with sealing gaskets used for food containers.
a
Regulations such as these are found in the Code of Federal Regulations (CFR) and users should be aware that they can be and are revised from time to time.
exists, so there may be a replacement in the future. ASTM D4368 described methods for testing homo- and copolymer vinyl chloride-containing polymers to determine important characteristics such as total chlorine content of the composition, dilute solution viscometry to assess polymer molecular weight, and high and low shear viscosity measurements to characterize vinyl dispersion polymers that are used for plastisols and organosols.
TABLE 3—Adhesion of vinyl copolymers to various substrates. Rating: 10 = pass scotch tape test, no loss of adhesion; 5 = some loss of adhesion, not recommended; 0 = no adhesion
Formulation of Solution Coatings
Substrate
No modification
Carboxyl modified
Hydroxyl modified
Steel
0
10
0
Galvanized
0
10
0
Paper (glassine)
10
10
10
Aluminum Foil
0
10
10
Polyethylene, treated
0
0
0
Polypropylene, treated
0
0
0
Polysulfone
7
10
10
Acrylic
10
10
10
PVC Plastic
10
10
10
ABS Plastic
10
10
10
Polycarbonate
10
10
10
Poly(phenylene oxide)
4
10
0
Poly(ethylene terephtalate)
0
0
0
Impact Polystyrene
0
0
0
Inked Surface
0
5
5
Typical vinyl coatings formulations consist of copolymer, solvent(s), plasticizer, pigments (required for exterior exposure), and optional ingredients such as stabilizers, modifying polymers(s), and cross-linking agent(s). The polymer, almost always a copolymer, is normally selected on the basis of its ability to provide adhesion to the substrate. While, in most cases, strong adhesion is desired, there are special coatings such as strippable or peelable coatings where adhesion is not wanted. Table 3 contains the relative adhesion of a few vinyl copolymers to various polymers and substrates. The copolymer selection may be made on the need for reactive functionality to produce crosslinked coatings that change the nature of the coating from thermoplastic to thermoset a factor that is characterized by improved solvent and/or stain resistance in the final coating.
Solubility
Variation in the ratio of vinyl chloride to vinyl acetate, the degree of polymerization, and the modifying third monomer results in a wide variety of vinyl copolymers with different solubility characteristics. Highest solubility is favored by low vinyl chloride content and low molecular weight. This relationship in terms of copolymer solution viscosity is apparent from the viscosity data given in Table 1. Vinyl solution-polymerized copolymers can be dissolved in ketones, esters, certain chlorinated solvents, and some nitroparaffins. As a class, ketones are the best solvents in terms of the ability to dissolve large amounts of solids with minimal solution viscosity. Hydrocarbons are chiefly used as diluents primarily to lower cost. Both aromatic and aliphatic hydrocarbons can be used as diluents. Aromatic hydrocarbons, particularly toluene and xylene as well as higher boiling fractions such as Aromatic 100 or 150, are preferred because they can be used at high levels, in the
Vinyl Chloride/Vinyl Acetate Copolymer
range of 50 %–65 % of the solvent blend depending on the copolymer composition, molecular weight, and desired solids. The Aromatic 100 and 150 are usually used only in baking finishes. Aliphatic hydrocarbons can be used in limited amounts, up to about 30 % of the solvent blend. Higher levels can lead to viscosity instability, and only low boiling aliphatic hydrocarbons, those with boiling points up to 117°C, are suitable. The use of higher boiling aliphatic hydrocarbons can cause
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
122
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
TABLE 4—Typical solvent mixtures for spray applied coatings, compositions in wt % Non Regulated Mixture
Rule 66 Compliant Mixture
Rule 66/EPA 33/50 Initiative Compliant Mixture
Methyl isobutyl ketone
50
Methyl ethyl ketone
14
Methyl ethyl ketone
7
Toluene or xylene
50
Butyl acetate
46
Acetone
3
100 %
Cyclohexanone
9
Methyl isoamyl ketone
15
Toluene
12
Butyl acetate
40
Xylene
7
Cyclohexanone
9
VM&P Naphtha
12
Toluene
6
100 %
VM&P Naphtha
20 100 %
precipitation of the copolymer during film formation and final drying. Alcohols are strong precipitants for vinyls and generally are not used in unmodified vinyl copolymer lacquers. However, in some cases vinyl copolymers, usually hydroxyl-modified versions, are readily formulated with other polymers, oligomers, or copolymers that are alcohol soluble. With these, up to 15 %–20 % alcohol may be used in the solvent blend. Careful attention must be paid in vinyl coating formulations that contain alcohols to ensure that problems do not develop during application and drying of the coatings. Glycol ethers and glycol ether esters are sometimes used in vinyl coatings to improve “flow out” of baked coatings. In response to regulations restricting the type and amount of solvents used in coatings, such as the early forerunner regulation known as the “Los Angeles Rule 66” and later versions,3 vinyl coatings were reformulated as compliant systems by reducing the amount of branched ketones and aromatic hydrocarbons and making up the difference with esters and aliphatic hydrocarbons [19]. Though it was necessary to use more oxygenated solvents, the performance requirements of compliant coatings remained the same. Table 4 has some typical solvent blends that had been used for spray applications and the reformulated compliant systems. These particular mixtures are for demonstration purposes to illustrate how regulations can markedly change formulations.
Solution Characteristics
From the time a vinyl polymer is dissolved, the viscosity of solutions increases with time until equilibrium is reached, after which the viscosity remains constant. This behavior is believed to be due to the formation of regions of microcrystallinity or other strong association between polymer molecules in solution. The extent of the viscosity increase is dependent on (a) polymer molecular weight, (b) solids content of the solution, and (c) the “strength”4 of the solvent blend. The viscosity increase may be small or so large that the solution sets to a gel. Properly formulated vinyl polymer solutions usually reach an equilibrium viscosity in about 3–5 weeks. Guidelines for the preparation of viscositySee chapters in this manual that deal with regulations and with solvents. 4 “Strength” in this case refers to the quality of the solvent and its ability to alter the size, configuration, and conformation of the dissolved molecules. 3
stable solutions for polymer of varying molecular weight are shown in Table 5. Vinyl chloride copolymer solutions also exhibit what is known as the “memory effect.” When a vinyl chloride copolymer solution is heated to about 60°C, the effect of microcrystallinity is eliminated. If the solution is then cooled to its original temperature, the viscosity will not immediately return to its original value because of the time lag needed for the effect of the microcrystallinity to redevelop. With time, the viscosity of the solution will return to the same value as it had before the heating process. The converse relationship holds when vinyl chloride copolymer solutions are cooled below the original storage temperature and then returned to the original condition. A graphical representation of the memory effect is given in Fig. 1.
Plasticizers
Internal plasticization of vinyl chloride polymers is achieved by copolymerization of vinyl chloride with monomers such as vinyl acetate. The comonomer reduces softening and processing temperatures and markedly improves solubility. However, such copolymerization often cannot achieve all of the desired physical flexibility and toughness characteristics required in many end uses. Plasticizers are often used with vinyl chloride copolymers to improve flexibility, formability, and impact resistance of the coating. Monomeric as well as polymeric plasticizers or compatible polymers with low glass transition temperature (Tg) may be used to plasticize a vinyl coating. Phthalate, phosphate, and glycol ester plasticizers are typically used. Plasticizers are selected to meet the requirements of the coating that may include low-temperature flexibility, resistance to extraction by solvents, resistance to migration, to humidity, etc. Blends of plasticizers may be required to meet specific requirements. FDA regulations as well as other health considerations must be taken into account when selecting these additives. Table 6 presents a listing of plasticizers that have been commonly used with vinyl copolymers.5 Care must be exercised in choosing the level of plasticizer since excessive amounts tend to make the film soft and prone to dirt retention and can lead to oozing from the surface under certain conditions. Ordinarily, a level of 25 phr (parts per hundred Also see the chapter in this manual that deals with plasticizers in general.
5
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 15
Q
VINYL POLYMERS FOR COATINGS
123
TABLE 5—Guidelines for viscosity stable solutions Compositiona
Third Monomer
Mnb
Maximum Solids
Typical Solvent Blend, wt %c
90/10/0
None
44 000
15
MEK/toluene, 67/33
86/14/0
None
27 000
25
MIBK/toluene, 50/50
86/14/0
None
22 000
28
MIBK/toluene, 50/50
86/13/1
Maleic acid
27 000
25
MIBK/toluene, 50/50
83/16/1
Maleic acid
19 000
28
MIBK/toluene, 50/50
81/17/2
Maleic acid
15 000
33
MIBK/toluene, 33/67
90/4/6
Vinyl alcohol
27 000
25
MIBK/toluene, 50/50
90/4/6
Vinyl alcohol
22 000
28
MIBK/toluene, 50/50
81/4/15
Hydroxyalkyl acrylate
33 000
25
MIBK/toluene, 50/50
81/4/15
Hydroxyalkyl acrylate
24 000
28
MIBK/toluene, 50/50
81/4/15
Hydroxyalkyl acrylate
15 000
33
MIBK/toluene, 33/67
Vinyl chloride/vinyl acetate/third monomer, weight percentages. Number average molecular weight, polystyrene reference standard. c MEK is methyl ethyl ketone and MIBK is methyl isobutyl ketone. a
b
parts polymer) plasticizer is considered about maximum for use with coating polymers.
Pigmentation
Vinyl copolymer coatings are pigmented (1) to achieve the desired colors and other aesthetic characteristics and (2) to prevent degradation of the vinyl copolymer caused by ultraviolet radiation when they are to be used outdoors [20]. Most organic and inorganic pigments can be used. However, basic pigments must be avoided with carboxylmodified copolymers since these pigments can and probably will react with the copolymer to form irreversible gel and may alter the desired color. Prime or color pigments that absorb ultraviolet radiation must be used at a level sufficient to protect the vinyl copolymer. Extender pigments or fillers do not absorb ultraviolet radiation and can only be used in combination with an ultraviolet-radiation adsorbing pigment, such as
titanium dioxide (TiO2). A minimum level of about 75 phr TiO2 is required to provide resistance to ultraviolet radiation. Other inorganic pigments can be used to replace TiO2 by substituting on an equal volume basis. Organic pigments that are manufactured with very small particle size are used at a lower concentration, and blends of inorganic
TABLE 6—Typical plasticizers that are compatible with solution vinyl chloridebased copolymers. Note—some of these plasticizers may be out of favor for health reasons, but all are compatible with solution vinyl chloride copolymers. Any users should check with suppliers regarding health considerations. Phthalates
Linear Dibasic Acid Esters
Butyl benzyl phthalate (BBP)
Di-n-butyl sebacate (DBS)
Di-2-ethylhexyl phthalate (DOP)
Di-2-ethylhexyl adipate (DOA)
Di-isooctyl phthalate (DIOP)
Di-isononyl adipate (DINA)
Di-isononyl phthalate (DINP)
Di-2-ethylhexyl azelate (DOZ)
Di-isodecyl phthalate (DIDP Citrates
Phosphates
Acetyl tri-n-butyl citrate
Tri-2-ethylhexyl phosphate (TOP) Isodecyl diphenyl phosphate
Epoxides
Fig. 1—Memory effect in vinyl chloride copolymer solutions.
Polymerics
Epoxidized soybean oil (ESO)
Adipic acid polyesters
2-Ethylhexyl epoxytallate
Azelaic acid polyesters
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
124
PAINT AND COATING TESTING MANUAL
Q
and organic pigments are often used to achieve the desired color. Excessively high pigment concentrations can lead to early chalking.
Organosols and Plastisols
A plastisol is a dispersion of discreet particles of high molecular weight vinyl chloride homopolymer in plasticizer. The combination almost always contains a low level of thermal stabilizer in sufficient quantity to prevent degradation during the baking or fusing process. Plastisols normally require a minimum amount of about 55–60 parts plasticizer per hundred parts of polymer to form a fluid mixture. The viscosity of the dispersion is dependent on particle packing effects, the volume of dispersed polymer relative to the volume of liquid plasticizer, the size and shape of the suspended particles, solvating or swelling effect of the plasticizer on the polymer particles, and the viscosity of the liquid plasticizer. The relatively high levels of plasticizer needed to produce a flowable liquid mix results in the formation of fused films that are too soft for use as coatings. Plastisol coatings are usually formulated from the addition of coarse particle-size poly(vinyl chloride) called extenders to the mix. These extenders are obtained from suspension or bulk (mass) polymerization and allow the use of less plasticizer and useful films with improved hardness result. Additionally, small amounts of thinner, usually aliphatic hydrocarbon are used (up to about 10 wt %) to reduce viscosity and provide better flow and leveling of the plastisol coating. Plastisol coatings do not adhere well to most substrates and most often require the use of a suitable primer before application. An organosol differs from a plastisol in that much lower levels of plasticizer are used. Combinations of weak solvents that are termed “dispersants” in combination with hydrocarbon solvents, termed “diluents,” are used to provide sufficient liquid to make a fluid dispersion. Because lower levels of plasticizer are used, films with much greater hardness can be obtained. Commercial organosols are usually modified with a solvent-soluble polymer to prevent mud cracking or film splitting during the bake used to fuse the film. The modifier polymer at times contains carboxyl functionality to make the coating self-adherent. In other instances, it may be a hydroxyl-containing copolymer to provide functionality for reaction with cross-linking agents, such as amino or phenol/formaldehyde oligomers and thus achieve a degree of thermoset properties. Though vinyl chloride copolymers are usually the preferred modifier for organosols, other polymers, such as acrylic polymers, are used. Careful consideration must be given to the selection of the solvent/diluent mixture for organosols, if one is to attain the highest solids coupled with good viscosity stability. Commercial organosols of 50 %–55 % non-volatiles by weight are typical. Plastisols and organosols require a high baking temperature of about 350°F (177°C) to fuse the films. At elevated temperatures, the plasticizer or plasticizer diluent mixture exerts a strong solvating or swelling effect on the dispersed poly(vinyl chloride) particles. At fusion, the polymer no longer exists in discreet particle forms, but rather as a continuous homogeneous film. Films of plastisols or organosols need to only reach fusion temperature, and they
15TH EDITION
do not need to be held at the fusion temperature for a long time period to form the final film. Undercuring or baking at temperatures lower than that required for fusion will yield films deficient in tensile strength, elongation, abrasion resistance, and other properties. Plastisols and organosols also require the use of thermal stabilizers to protect the vinyl polymer against degradation during the fusion/bake operation. Thermal stabilizers are usually combinations of metal salts of organic acids in combination with epoxidized oils or liquid epoxy compounds. Special attention must be given to the selection of thermal stabilizers for organosols modified with solvent-soluble polymers, especially when carboxyl-modified polymers are used. In such cases, metallic salts must be avoided as these will usually cause gellation; typically, mercapto tin or tin ester compounds are used in combination with an epoxy stabilizer. The type pigment and pigment concentration used in pigmented organosols follow the guidelines given for solution vinyl copolymers. It is, however, more difficult to prepare pigmented plastisols because there is generally little solvent used to control viscosity. Low oil absorption pigments must be used to avoid excessively high viscosity and difficult-to-work-with formulations.
Primers for Plastisols and Organosols
To develop good adhesion when used on metal substrates, plastisol coatings require a primer. An organosol coating may also require a primer if it is not modified with an adhesion-promoting modifier. Suitable primers can be formulated from carboxyl-modified vinyl copolymers and may require employing thermoset technology for best results. This is accomplished by using cross-linking agents such as amino-formaldehyde or phenolic compounds to provide resistance to excessive softening from highly plasticized plastisol or organosol coatings.
MAJOR MARKET AREAS FOR VINYL COPOLYMER COATINGS
Rigid Packaging LINERS FOR INTERIOR SURFACE COATINGS, CANS, CAN ENDS, CLOSURE/CAPS AND CROWNS
The first commercial use for vinyl chloride copolymer coatings was as the topcoat lacquer used on the inside of beer cans. As beverage cans evolved from three- to twopiece construction, the vinyl coating also changed from lacquer to hydroxyl-functional vinyl chloride copolymers in combination with amino-formaldehyde cross-linking agents. Thermoset coatings, such as these, were needed to meet the need for greater corrosion conditions encountered. Thermoset coatings of epoxy-modified vinyl chloride copolymers with carboxyl-modified vinyl copolymers were used to coat coil stock. The coated coil stock is then formed into the stay-on-tab can ends, an application that requires excellent mechanical properties to withstand the forming steps without cracking. Organosol coatings containing a solution vinyl copolymer component, usually carboxyltype for adhesion, have also been used on precoated stock for can ends. Vinyl organosols are further modified with amino-formaldehyde or phenolic compounds to upgrade chemical resistance and permit the use of such coatings for packaging food that will be autoclaved to sterilize the contents [21].
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 15
Vinyl lacquer and vinyl thermoset coatings are used as size coats for metals that are formed in caps and closures for jars or as crowns for beverage bottles. These systems serve as the primer coat for gasketing compounds made with plastisol or vinyl copolymer dry blends.
Flexible Packaging
Solvent-soluble carboxyl-modified vinyl chloride copolymers have good adhesion to most materials used in flexible or semi-rigid packaging including aluminum foil, paper, and plastic films, such as poly(ethylene terephthalate), polycarbonate, poly (vinyl chloride), and cellophane. This type copolymer is used for its adhesion characteristics, ease of heat sealing, and resistance to attack by the packaged product. The vinyl copolymer may be used alone or modified with plasticizers or other compounds and polymers to formulate heat-sealable coatings for applications requiring varying degrees of force needed to open the container. This could range from applications such as blister packaging where the bond needs to be strong enough to cause substrate failure when the package is opened, to items such as jellies or cream containers found in restaurants where a tight but readily peel-able bond is required. Vinyl coatings are also used to coat collapsible metal tubes for packaging materials such as pharmaceutical preparations, toothpastes, and the like where the need is for a very flexible coating that will not crack nor be attacked by the contents of the package even though high stresses from collapsing and rolling up the tube are encountered multiple times. Other applications include decorative coatings for aluminum foil/paper laminates used in cigarette packaging, food wrappers for fast food restaurant items, for butter, margarine, soups, and so on. Decorative foil for floral wrappings, decorative labels, and coatings for aluminum foil used on vapor barrier insulation in construction applications are also coated with vinyl copolymers.
Inks and Overprint Coatings
The major markets for vinyl copolymer-based inks are for products that have a vinyl surface such as floor and wall coverings, swimming pool liners, vinyl upholstery, and garment fabrics. The main reason for use on these substrates is related to excellent adhesion as well as the toughness, elongation, and stain resistance that are obtained with overprint lacquers—usually transparent wear layers. Other areas of importance include treated poly(ethylene terephthalate) and polyolefin films, aluminum, paper composites, and metalized substrates. Ink formulation is quite similar to that used when formulating coatings except solvent choices are somewhat narrowed and higher pigment or other colorant loadings are needed to achieve hiding in the thin films typical of inks. Vinyl inks are often reverse printed on a clear vinyl or other polymeric film, and the printed film is then laminated to substrates such as wood or metal to make articles having simulated wood finish. Vinyl inks are also printed by the gravure or screen process because these presses are compatible with the strong solvents needed for vinyls. Flexographic printing is not suitable for vinyls because the plates are susceptible to solvent attack. Inks for highly plasticized vinyl surfaces are usually formulated with ester solvents to avoid excessive softening of calendered films and puckering of the films.
Q
VINYL POLYMERS FOR COATINGS
125
Dry Film Printing (Hot-Stamp Transfer Process)
Dry film printing involves printing vinyl inks onto a carrier film such as poly(ethylene terephthalate), polyethylene, polypropylene, or other suitable surfaces to which the ink will not strongly adhere. The inks are applied and dried usually in web form. When ready for use, the printed carrier film is placed with the inked side on the surface to be decorated. A heated die presses the composite to make intimate contact with the surface so that when the die is removed, the ink is firmly bonded to the substrate and the carrier is cleanly peeled away.
Maintenance and Marine Finishes
Heavy duty marine finishes were developed in the mid-1940s. These systems were comprised of a poly(vinyl butyral) wash primer, vinyl chloride copolymer red lead anti-corrosive intermediate coatings (based on vinyl alcohol modified copolymers, which were needed to provide adhesion to wash primer), and vinyl copolymer/wood rosin/cuprous oxide anti-foul top coats. Such systems have become the subject of numerous specifications. Many United States government agencies as well as agencies of other governments have written specifications with such a coating system specified for use below the waterline of marine vessels. Because of their good water resistance, weathering qualities, flexibility, fast drying characteristics, ease of application, and repair, vinyl chloride-based copolymers quickly became established as maintenance finishes. This application area includes coatings for locks, dams, appurtenant structures for waterways, interior linings for potable water tanks, steel structures such as bridges, electrical towers, equipment in chemical plants, dams and locks, storage tanks, and the like. Many specifications have been written that require the use of vinyl copolymers as maintenance paints [22,23]. Air atomizing spray guns at low solids were used for application of vinyl maintenance and marine coatings in the early usage days. The low solids required several coats to attain coverage sufficient for good corrosion protection. High-build airless spray-applied vinyl coatings were developed in the 1970s to fill the need for coatings systems that could be applied in fewer coats at less expense [24].
Wood Finishes
Reactive heavy-duty vinyl finishes, coatings and sealants for wood have been developed. These consist of a hydroxylmodified vinyl chloride copolymer cross-linked with amino/ formaldehyde compounds. Alkyd copolymers were often added to improve film build. Such finishes became established as the standard for kitchen cabinets because of their retention of excellent adhesion and water resistance, particularly when the coated wood becomes wet from high humidity or water splashing. These finishes also have excellent resistance to a variety of household chemicals, solvents, and stains and have been used as fine furniture finishes [25].
Magnetic Recording Media
Vinyl chloride copolymers, especially hydroxyl-modified copolymers, have been used as binders for magnetic iron oxide tapes since the beginning of the development of tape recording. The vinyl copolymers are used because of their good adhesion, abrasion resistance, and pigment wetting properties. The early binder formulations used alkyd
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
126
PAINT AND COATING TESTING MANUAL
Q
copolymers as plasticizers, and later polyesters were used. The polyesters were followed by polyurethanes as the plasticizer as the technology of tapes advanced and placed more stringent requirements on the performance of magnetic tape for audio and video [26].
Powder Coatings
Vinyl powder coatings are formulated with vinyl chloride homopolymers and copolymers for application by fluidized bed, powder spray, or electrostatic powder spray. Powder coatings are prepared by dry compounding the polymers, plasticizer, pigments, and additives in ribbon blenders followed by attrition or dispersion to powder in mixers such as a Henschel mixer. Some powder coatings are prepared by a melt mix technique followed by cryogenic grinding. This latter technique produces powders of smaller particle size [27]. Powder coatings prepared by dry compounding are usually applied by fluidized bed or by spray techniques. For fluid bed powder application, the metal parts are heated so that the powder will adhere to the part, begin to coalesce, and start film formation. An oven bake after the powder application is needed to complete the film-forming process by fusion or melting. Cryogenically ground powder coatings are applied by electrostatic powder spray. With the electrostatic method, it is not necessary to preheat the parts, but an oven bake is necessary after application to fuse the powder to a coherent film. The finer particle size allows deposition of and thinner and smoother films than is attainable from the fluidized bed process. Poly(vinyl chloride) powder coatings are used to coat products such as metal pipe, fencing, and metal furniture.
Poly(Vinyl Chloride) Latex
Emulsion polymerized vinyl chloride homopolymers and copolymers are used in the latex form not so much to make finished coatings, but rather as substance coated onto a base or support and thereby providing the substrate for items such as wall coverings, backing for carpeting, and the like. In a sense, such use could be considered analogous to a waterborne version of an organosol coating. The vinyl chloride homopolymers need to be modified with a substantial loading of plasticizer, and some grades are sold as preplasticized latexes. These water-based materials require a high temperature bake to fuse the polymer plasticizer mix into a continuous film. By varying the type and amount of co-monomer used to make emulsion polymerized copolymer latexes, products with a decreased glass transition temperature and lower film-forming temperature are available. These allow lower temperature bakes for film formation.
Waterborne Vinyl Dispersions
Waterborne vinyl dispersions made from solution-polymerized vinyl copolymers became available in the 1980s. These colloidal aqueous vinyl dispersions are of medium molecular weight and have high glass transition temperatures of about 80°C. Coalescents are needed with these products to form a film. Some dispersions are available with a glycolether coalescent already present in the product, and a co-solvent free variety is also available. With the latter, the formulator can choose whichever coalescent, glycol-ether,
15TH EDITION
glycol-ether ester, plasticizer, or blend of coalescents that best meets performance requirements. Waterborne vinyl dispersions are used in many adhesive, ink, overprint lacquer, industrial coating, and heatsealable coating applications where solvent-based vinyl coatings had been used and still are used.
Trends in Vinyl Coatings
New and modified regulations of concern to vinyl chloride copolymers have been requiring improved volatile organic solvent content—that is, lower volatile organic content (VOC). Studies over the recent past have centered on high solids and waterborne systems. In the case of high solids formulations, reductions in copolymer molecular weight have allowed viscosity stable solution at twice and more the solids content. While higher solids can be achieved, the reductions in molecular weight have affected the performance of coatings made from such copolymers—decreases in chemical resistance and physical properties were noticed in films formed from low molecular weight lacquers. To overcome such decreases in properties caused by lower molecular weight copolymers, the copolymers designed for high solids vinyl copolymers are modified to contain hydroxyl functionality and to allow for reaction with added co-reactant materials to build molecular weight through the thermoset process of cross-linking. Here difficulties can be encountered in trying to achieve a minimum of one functional group on each oligomeric molecule. Thus, the high solids lacquer designed copolymers can be used alone for applications that do not have very demanding requirements, but the copolymers are best used when they contain functionality for cross-linking with amino-formaldehyde or isocyanate cross-linking agents or as modifiers for alkyds, polyester-isocyanate, or epoxy-amine coatings to improve initial drying or set-to-touch rate, or to improve recoatability [27]. Modification of the vinyl copolymers to improve specific characteristics such as the copolymer containing sulfonate groups for improved pigment dispersion provides new tools for formulators [28]. The waterborne vinyl dispersions previously described represent an alternative to high solids vinyls as a way to formulate low VOC coatings. The waterborne vinyls are compatible with a wide variety of other waterborne polymers that have low VOC content, such as acrylics, alkyds, urethanes, and amino-formaldehyde cross-linking agents. Investigators are looking into the interactions that take place between poly(vinyl chloride) and poly(vinyl acetate) in various solvents [29]. The results indicate that the solvent has a marked effect on the interactions that take place in solution and these in turn have an effect on films prepared from the mixtures. There may be ramifications from these studies with homopolymers as to copolymer interactions. Although this chapter deals with polymers and oligomers prepared from vinyl chloride, it should be pointed out that vinyl chloride (that is, the monomer) is a hazardous chemical and a known human carcinogen. Thus, precautions should be taken by anyone who comes in contact with the monomer. It is beyond the scope or intent of this chapter to deal with all of the details and ramifications of the health hazards associated with vinyl chloride monomer and readers are directed to information available from organizations such as the United States
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 15
Occupational Safety and Health Administration, from Europe’s Restriction on Hazardous Substances (RoHS), from Waste Electronic and Electrical Equipment, as well as other groups. Of course, manufacturers’ Material Safety Data Sheets, for both the monomer and polymers derived from it, should always be consulted when dealing with these materials. The internet contains an abundance of information about these and other organizations, vinyl chloride monomer, and polymeric species derived from vinyl chloride.
References [1] Ostromislensky, I. I., J. Russ. Phys.-Chem. Soc., Vol. 44, 1912, p. 204. [2] Ostromislensky, I., “Polymer of Vinyl Chloride and Process of Making the Same,” U.S. Patent No. 1,721,034 (1929). [3] Cowfer, J. A., and Gorensek, M. B., “Vinyl Chloride,” KirkOthmer Encyclopedia of Chemical Technology, John Wiley and Sons, Inc., New York, 2006. [4] Mulder, K., and Knowt, M., “PVC Plastic: A History of Systems Development and Entrenchment,” Technol. Soc., Vol. 23, 2001, pp. 265–286. [5] Reid, E. W., “Process for Producing Vinyl Resins,” U.S. Patent No. 2,064,565 (1936). [6] Reid, E.W., “Vinyl Resins,” U.S. Patent No. 1,935,577 (1933). [7] Davidson, J. G., and McClure, H. B., “Applications of Vinyl Resins,” Ind. Eng. Chem., Vol. 25, 1933, pp. 645–652. [8] Treatise on Coatings, Film Forming Compositions, R. Myers, and J. S. Long, Eds., Dekker, New York, Vol. 1, Part II, 1968. [9] Powell, G. M., Federation Series on Coatings Technology, Unit 19, Federation of Societies for Paint Technology, Philadelphia, 1972. [10] Brezinski, J. J., Koleske, J. V., and Potter, G. H., “Hydrodynamic Properties of Vinyl Chloride-Vinyl Acetate Copolymers in Dilute and Concentrated Solutions,” Proceedings of XI Congress FATIPEC, Florence, Italy, 1972. [11] Khan, Z. S., and Hughes, T. W., “Source Assessment, Polyvinyl Chloride,” Document EPA—600/2-78-0041, U.S. Environmental Protection Agency, May 1978. [12] Koleske, J. V., and Wartman, L. H., “Poly(vinyl chloride),” Polymer Monographs, Gordon and Breach Science Publishers, New York, Vol. 3, 1969, p. 112.
Q
VINYL POLYMERS FOR COATINGS
127
[13] Burns, R. J., “Vinyl Resin for Coatings,” Paint and Coating Testing Manual, 14th ed., ASTM, West Conshohocken, PA, Chap. 15, 1995. [14] Dow Chemical Company, UCARTM Solution Vinyl Resins: Flexible Solutions for Coatings, Dow Chemical Company, Midland, MI, 2006, p. 34. [15] Crompton, T. R., Analysis of Plastics, Pergamon Press, New York, 1984. [16] Infrared Spectra Atlas of Monomers and Polymers, Sadtler Research Labs, Philadelphia, 1980. [17] Burley, R. A., and Bennett, W J., “Spectroscopic Analysis of Poly(Vinyl Chloride) Compounds,” Appl. Spectrosc., Vol. 14, 1960, pp. 32–38. [18] An Infrared Spectroscopy Atlas for the Coatings Industry, 4th ed., Vols. I and II, D. R. Brezinski, Ed., Federation of Societies for Coating Technology, Philadelphia, 1991. [19] Burns, R. J., and McKenna, L. A., Paint and Varnish Production, Vol. 62, No. 2, 1972, p. 29. [20] Hardman, D. E., and Brezinski, J. J., “Pigmented Vinyl Copolymer Coatings: A Discussion of Factors Influencing Exterior Durability,” Off. Dig. Fed. Soc. Paint Tech., Vol. 36, 1964, pp. 963–984. [21] Good, R. H., ACS Symposium Series 365, American Chemical Society, Washington, DC, 1988, pp. 203–216. [22] “Corps of Engineers,” CW-099040, U.S. Department of the Army, August 1981. [23] Steel Structures Painting Council, Pittsburgh, PA, Paint No. SSPC-9. [24] Martell, R. J., and Yee, A., J. Protective Coatings Linings, Vol. 5, No. 9, 1988. [25] Mayer, W. P., “High Performance, High Solids Coatings Using Solution Vinyl Resins,” J. Oil and Colour Chem. Assoc., Vol. 73, No. 4, 1990, p.159. [26] Kreiselmaier, K. W., “Pigmentation of Magnetic Tapes,” Pigment Handbook, Vol. III: Applications and Markets, T. C. Patton, Ed., John Wiley & Sons, New York, 1973. [27] Ginsberg, T., “Vinyl-Modified Epoxy Coatings,” Modern Paint and Coatings, No. 11, 1988. [28] Dow Chemical Company, UCARTM Solution Vinyl Resins: Proven Performance—New Solutions, Dow Chemical Company, Midland, MI, 2006, p. 6. [29] Zhang, Y., Qian, J., Ke, Z., Zhu, X., Bi, H., and Nie, K., “Viscometric Study of Poly(vinyl chloride/poly(vinyl acetate) Blends in Various Solvents,” Eur. Polym. J., Vol. 38, 2002, pp. 333–337.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
16
MNL17-EB/Jan. 2012
Miscellaneous Materials and Coatings Joseph V. Koleske1
THIS CHAPTER IS CONCERNED WITH A VARIETY OF products that are not discussed elsewhere in the manual. Some topics are mentioned only briefly to indicate that the area has not been forgotten and that the topic is not within the scope of the manual.
POLYOLS
Polyols, or polyalcohols as they are sometimes known, are compounds containing one or more, but usually two or more, free hydroxyl groups. Most definitions, and particularly those over 10 years old, list typical polyols as compounds such as ethylene glycol, propylene glycol, neopentyl glycol, glycerol or glycerin, trimethylolpropane, pentaerythritol, and sorbitol that were used in the preparation of alkyds and polyesters. Today the word “polyols” is far more encompassing and more often than not refers to alkylene oxide [1] and ∈-caprolactone [2] adducts of the above-mentioned and other monohydric or polyhydric alcohols, low-molecular-weight polyesters prepared from the above mentioned as well as other polyhydroxyl compounds and dicarboxylic acids (particularly adipic acid) [3–5], polytetrahydrofurans prepared by a cationic ringopening polymerization of tetrahydrofuran [6,7], and lowmolecular-weight polycarbonates [8–10]. Ortho-phthalate based polyesters used in the manufacture of polyurethanes have hydrolytic stability advantages based on the aromatic
ingredients in high solids and cationic photocure systems, as well as in a number of other end uses including elastomeric fibers, dentistry, artifact preservation, and pharmaceutical preparations. The two main classes of polyols used in coatings are the polyether polyols, which are typified by the poly-(propylene oxide) polyols (PPO), and the polyester polyols, which include both poly(glycol adipates)(PEA) and poly-∈-caprolactone polyols (PCP). Both classes of polyols are available as difunctional and trifunctional hydroxyl compounds though the adipates are almost always difunctional in nature. Higher functional polyols are known and available, but their usage is less common than that of the di- and trifunctional products. In the above structural formulas, R and R′ may be the same or different and –O–R–O– and –O– R′–O– are the residues of the polyhydric alcohol initiators. Difunctional and trifunctional PPOs are usually initiated with 1,2-propylene glycol and glycerol, respectively. The adipate polyols are usually prepared with an excess of diol, so most end groups are hydroxylic rather than carboxylic in nature. Since these polyols are prepared by a condensation reaction, there is no need for an initiator. Caprolactone polyols are initiated with a variety of diols and triols such as diethylene glycol, ethylene glycol, 1,4-butanediol, trimethylol propane, glycerol, etc. The above structure for PPO indicates that the hydroxyl
HO—[CH(CH3)—CH2]a—O—R—O—[CH2—CH(CH3)b—OH Poly(propylene oxide)Polyol H—[O(CH2)4O—CO(CH2)4CO]u—O(CH2)4O—[CO(CH2)4CO—O(CH2)4O]v—H Poly(1,4-butanediol adipate), a Polyester Polyol H—[O(CH2)5CO]s—O—R’—O—[CO(CH2)5O]t—H Poly-∈-caprolactone Polyol substitution pattern within the polyol [11,12]. There are other compounds that meet the above definition, but they are not usually termed polyols. Compounds such as these are certain acrylic oligomers [13], vinyl chloride copolymers, hydroxyl-containing glycidyl ether compounds, vinyl alcohol copolymers, and so on. This chapter will not be concerned with these latter compounds since they are dealt with elsewhere in the manual. Polyols are important compounds used in the manufacture of alkyds and polyurethane coatings, of intermediates used in radiation curable formulations, as copolymerizable
1
groups are both secondary, which is the usual case. However, from time to time, a primary hydroxyl group will be found due to an unexpected opening of the propagating 1,2-epoxide. The subscripts a, b, u, v, s, and t in the above structural formulas can be the same or different, and they can take on a wide variety of values with the number average molecular weight ranging from about 150 to 3,000 for polyols usually used in coatings. Details about preparation of urethane coatings based on polycaprolactone polyols for rigid substrates [14] and flexible substrates [15] are available.
1513 Brentwood Road, Charleston, WV 25314-2307.
128
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 This standard is for EDUCATIONAL USE ONLY. www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
Downloaded/printed byASTM International Copyright © 2012 by
CHAPTER 16
Q
MISCELLANEOUS MATERIALS AND COATINGS
A variety of other specialty polyols also exist, such as poly(butylene oxide) and polybutadiene polyols, which are useful when very high levels of barrier hydrophobicity are needed [16]. Poly(tetramethylene oxide) polyols also have good hydrophobic character. New polyols are also being developed, including polyols based on lactose that have flame-retardant characteristics as well as polyols with different end capping, etc. [17]. Poly(phenylene ether) has been made into low-molecular-weight polyols that are effective in enhancing high temperature performance of products [18]. Hydroxyl terminated, low-molecular-weight polysulfides have been prepared by splitting a relatively high-molecular-weight polysulfide in aqueous latex form with a dithiodialkylene glycol and sodium sulfite [19]. Such polyols have been used as polysulfide-based polyurethane glass sealants. Although new polyols such as these are often designed for use in the manufacture of polyurethane foams and elastomers, they can be and are used in coating formulations. Recently, there has been a noticeable interest in highly branched and highly functional polyols that are hyperbranched or dendritic and polydisperse in nature [20–22]. Perfect dendrimers have a uniform distribution of branches and functional groups around a central core molecule and are monodisperse or substantially monodisperse in character [23]. At present, such molecules are more of academic than practical interest due to the multistep procedures required for their preparation. In contrast, hyperbranched dendritic polyols, which contain a nonuniform distribution of branches and functional groups around a central core, have a certain or definite degree of polydispersity that can have a Mw/Mn ratio of about 1.5 or more [24]. Such polyols are of commercial interest since they can be prepared by a one-step process. The polyols are most conveniently prepared from glycerol in molecular weights of about 1,000 to 35,000 g/mol. The molecules have a functional group on the end of every branch, and a molecule with a molecular weight of about 5,000 has about 68 hydroxyl end groups and a 1,500 molecular weight product will have 24 functional groups. With such high degrees of branching the compounds are amorphous in nature and soluble in a variety of organic solvents as well as with water in the case of polyglycerol. They have low solution viscosities because chain entanglements are nil, low vapor pressure, and high reactivity. The polyglycerol polyols have been further reacted by copolymerization with glycidyl ethers [25] and propylene oxide [26], esterification [27], and condensation with suitable acrylates to form radiation-curable products [28]. The hyper-branched polyols can be cross-linked with a variety of cross-linking agent and formed into useful products such as high solids, thermoset binders [29] and radiation-curable printing inks [30]. Such cured products have improved scratch and abrasion resistance, adhesion, hardness, and flexibility. End capping polyols can provide adducts with different properties. For example, poly(propylene oxide) polyols, which contain terminal secondary hydroxyl groups, can be end-capped with ethylene oxide to provide polyols with more reactive primary hydroxyl groups [1,7]. Ways to apply nuclear magnetic resonance to measure the ethylene oxide content of these and other propylene oxide/ethylene oxide copolymers are detailed in ASTM D4875, Test Methods of Polyurethane Raw Materials: Determination of the
129
Polymerized Ethylene Oxide Content of Polyether Polyols. Also described in the literature [1,7] are polyols modified to have amine, allyl, carboxyl, cyano, and vinyl ether end groups. Glycols that are solid and/or that have subliming characteristics, as 2,2′-dimethyl-3-hydroxypropyl 2,2′dimethyl-3-hydroxypropionate, can be modified with a few ethylene or propylene oxide groups to yield new polyols that are liquid, have low viscosity, and do not sublime with even a few molecules of ethylene oxide having nil or very little effect on moisture resistance [31]. Polyols can be end-capped with an anhydride to form adducts that have free carboxylic acid functionality or a mixture of it and hydroxyl functionality as has been done with the poly-∈caprolactone polyols [32] or the alkylene oxide capped glycols [33]. In other instances, poly(propylene oxide) polyols have had carboxyl groups grafted to their backbone with acrylic or methacrylic acid. These grafted polyols retain their original hydroxyl end groups and are used in coating formulations [34]. Polyols can be incorporated into alkyds, made into moisture-curing urethanes, can be cross-linked with aminoplasts, and can be cross-linked with cycloaliphatic epoxides when terminated with carboxylic acid end groups. In using the polyols, the hydroxyl number [35] is their most important physical characteristic to be measured and used. Five wet chemical methods and two nuclear magnetic resonance methods for determining the hydroxyl number are given in ASTM D4274, Test Methods for Testing Polyurethane Polyol Raw Materials: Determination of Hydroxyl Numbers of Polyols and in ASTM D4273, Test Methods for Testing Polyurethane Raw Materials: Determination of Primary Hydroxyl Contents of Polyether Polyols, respectively. An infrared method can be found in ASTM D6342. Standard Practice for Polyurethane Raw Materials: Determining Hydroxyl Number of Polyols by Near Infrared (NIR) Spectroscopy. The equivalent weight or combining weight of a polyol is determined from the hydroxyl number by the following relationship Equivalent Weight = 56 100/Hydroxyl Number when potassium hydroxide is used as the titrating agent. Of course, if functionality is known, polyol molecular weight can be calculated by multiplying the equivalent weight by the functionality. Manufacturers provide information about hydroxyl number and usually about methods for analytically determining it. Another important reactivity parameter is the acid number described in ASTM D4662, Test Methods for Polyurethane Raw Materials: Determination of Acid and Alkalinity Numbers of Polyols. Acidity and alkalinity in polyols can affect reactivity, shelf life, color, and hydrolytic stability of coatings prepared from polyols. Polyethers and poly-∈caprolactone polyols usually have very low acid numbers. However, due to the nature of the condensation reaction coupled with transesterification used to produce polyester polyols, these polyols have relatively high acid numbers. The alkalinity in polyols with low alkalinity content can be determined with ASTM D6437, Test Method for Polyurethanke Raw Materials: Alkalinity in Low-Alkalinity Polyols (Determination of CPR Values of Polyols). Color, which has obvious implications, can be determined with ASTM D4890, Test Methods for Polyurethane Raw Materials: Determination of
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
130
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
epoxides polymerize by nucleophilic attack on the epoxide ring to form an ether linkage and a hydroxyl group on the ring. The hydroxyl group that is formed on the ring is quite acidic in character and will readily open other cycloaliphatic epoxide groups. Recently, a variety of new cycloaliphatic epoxides have been prepared by a transesterification process involving methyl-3,4-epoxycyclohexanecarboxylate and a variety of hydroxyl-terminated compounds [38]. The hydroxyterminated compounds included 1,4-butanediol, 1-8-octanediol, 1,4-cyclohexanediol, and others. The new epoxides provided tougher films when cross-linked with ultraviolet radiation in the presence of cationic photoinitiators than conventional cycloaliphatic epoxides. In the coatings industry, cycloaliphatic epoxides are used as a major formulating ingredient in cationic, photocurable formulations [39]. Usually they are formulated with polyols, onium-salt photoinitiators, and other ingredients. The onium salts photolyze in the presence of ultraviolet radiation to form strong protic acids that cause rapid polymerization of the epoxides as well as their copolymerization with active hydrogen compounds such as polyols. The presence of alkalinity including even very weak bases can result in neutralization of the protic acids formed by photolysis. Since the protic acids function as initiators, their neutralization will cause a marked decrease in polymerization rate. It may even result in nil reactivity. Coatings such as these are used as conformal coatings [40–42] in the electronics industry because of their excellent electrical (MIL-I-46058C approved, QPL Type ER) flammability (UL QMJU2 at a 2-mil thickness) and water permeability properties, as exterior can and other packaging coatings, overprint varnishes, printing inks for paper and metal, etc. Cycloaliphatic epoxides have been reacted with the free carboxylic acid groups on anhydride adducts of polyols [43]. Such coatings are characterized by pot lives of less than 8 h, high solids, and low-temperature curing capabilities with very high gloss and depth of image, high hardness, excellent solvent resistance, adhesion, and toughness. In other instances, the epoxides have been reacted with polyols in the presence of triflic acid salts (as diethylammonium triflate, 3M Co.). In this case, shelf lives of more than 8 months have been obtained and the formulated systems have high solids coupled with low viscosity and low temperature-cure characteristics. Cured coatings have an excellent balance of properties similar to those described above.
Gardner and APHA Color of Polyols. Other factors such as ethylene oxide content, specific gravity, suspended matter, unsaturation content, and water content can be determined with various ASTM Test Methods, D4875, D4669, D4670, D4671, D4672, respectively.
CYCLOALIPHATIC EPOXIDES
Although the topic of epoxides in coatings is the subject of a separate chapter in this manual, that chapter deals with glycidyl or 1,2-epoxides that are not attached to a ring structure. Such epoxides are the largest volume products of all epoxides used, and the main products in this class are the diglycidyl ethers of bisphenol A. However, there is a special class of epoxides, termed “cycloaliphatic epoxides,” that are used in specialty coatings and in cationic radiation-cure coatings. These epoxides are characterized by a saturated ring structure that imparts a high degree of weatherability and excellent electrical properties such as dielectric constant, dissipation factor, dielectric breakdown voltage, etc., to coatings and other products made from them. The good weatherability of the cycloaliphatic epoxides is apparent from the fact that they have been used for decades to make the large electrical insulators used in substations [36]. These compounds react well with carboxylic acids, as evidenced by their time-honored use as acid scavengers, and this reactivity often forms the basis for their use in coating formulations. The main commercial cycloaliphatic epoxide is 3,4-epoxycyclohexylmethyl 3,4-epoxycyclohexane carboxylate that has the structure
Cycloaliphatic Epoxide This epoxide is well known by the familiar name designation ERL-4221. Table 1 contains the properties of this epoxide and other cycloaliphatic epoxides that are commonly known in the industry. Epoxide equivalent weight can be determined with ASTM D1652, Test Methods for Epoxy Content of Epoxy Resins. Manufacturers can also be helpful in supplying information about methods of analysis for specific products. Usually these epoxides are reacted with polyols that function as flexibilizing agents for the highly cross-linked polymeric network that results. These
TABLE 1—Commercial cycloaliphatic epoxides and their physical properties [36,37] Viscosity, cP at 25°C
Specific Gravity 25/25°C
Color, 1993 Gardner (max)
Epoxide Equivalent Weight
Boiling Point, °C (mm Hg)
Vapor Pressure at 20°C, mm Hg
Solidification or Glass Point°C
3,4-Epoxycyclohexylmethyl 3,4epoxycyclohexane carboxylate
350–450
1.175
1
131–143
354 (760)
γsl and 90° when γsv < γsl. Fig. 5 lists the possible scenarios for wetting. The work of spreading, WS, is a measure with which a liquid favors spontaneously wetting a solid. This value is the difference between the work of adhesion, WA, and work of cohesion, WC, for the solid-liquid interface [39]. WC = 2 sl
(11)
WA = sv + sl – lv
(12)
WS = WA – WC = sv – ( sl + lv )
(13)
A positive spreading coefficient indicates the liquid possesses sufficient excess energy in the liquid to spread
Fig. 5—Schematic of the possible scenarios for wetting. (Image duplicated from Fig. 4 of Ref. 37). Reprinted from Advances in Colloid and Interface Science, Vol. 133, Kumar, G. K., Prabhu, N., “Review of Non-Reactive and Reactive Wetting of Liquids on Surfaces,” pp. 61-89, 2007, with permission from Elsevier.
15TH EDITION
spontaneously under equilibrium conditions, whereas a finite contact angle is formed for WS less than zero [40,41]. The surface energy of a substrate is the determining factor in the degree of wetting that occurs. Polarity, which often describes affinity to water, has also been discussed as a basis for predicting extent of wetting and is dependent on the nature of the substrate [42]. Low surface energy, hydrophobic solids such as hydrocarbons or fluorocarbons are not easily wet and form high contact angles with most pure solvents or solutions; poly(tetrafluoroethylene), or Teflon®, is an extreme example of this. Superhydrophobicity and supero-leophobicity describe solid materials whose contact angle with water and organic liquids, respectively, is larger than 150° [43]. Also unique to these surfaces is that they have near-zero contact angle hysteresis, or the difference between the advancing, θa, and receding, θr, contact angles [44]. Materials of this nature have been the subject of extensive research within the last decade, especially in areas including stimulus-response chemistry, biologically inspired synthesis, and hierarchical structuring [45]. Substrates with these properties are of great interest and are considered for a variety of possible uses such as self-cleaning and self-healing applications. Additional publications concerning these topics and the chemical and topographical sources of their extreme contact angles are found in Refs. [42–54]. High surface energy materials include metals and metal oxides that form low contact angles and are easily wet. When coating these materials, caution should be taken during surface preparation; oils and low surface energy impurities readily adsorb and have the effect of lowered surface energy and reduced wetting. Wettability is essential in the application of coatings to polymeric or metallic substrates; the liquid must wet the surface completely to obtain good adhesion [40]. This is commonly achieved by cleaning the substrate well, or adding surfactants to the wetting liquid to depress its surface tension and reduce the contact angle to zero. Cationic surfactants are most commonly used because the surfaces to be wet generally have a negative charge [38]. Methods for measuring the extent of cleanliness, such as contact angle, have been discussed by Durkee and Kuhn [55]. Surface preparation is the single most important step in coating application to ensure substrate/coating adhesion and guarantee its protection. In addition to cleaning with detergents, physical and chemical surface pre-treatment options for altering the substrate surface energy are available; a section at the end of this chapter is dedicated to their discussion. Pre-treatments such as flame and corona are especially necessary for low surface energy polyolefin substrates such as polyethylene or polypropylene. Adjusting the surface roughness of a substrate also affects a coating’s ability to wet a substrate [40]. Substrate pre-treatments may often include sanding or other roughening techniques. Increased surface roughness promotes wetting when the contact angle is below 90° but decreases wetting for those above 90°. Roughened surfaces generally have the adverse effect of creating a hysteresis in the contact angle measurements, with the advancing drop being the larger than the receding [39]. In order to maintain the applicability of Young’s equation, Wenzel developed a relation to account for surface roughness. Here, cos θrough is the measured contact angle, cos θsmooth is the thermodynamic
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 37
Q
SURFACE ENERGETICS
457
contact angle for an atomically smooth surface, and r is the roughness ratio [37]: cosrough = r cos smooth
(14)
cos θsmooth is also the theoretical Young’s equilibrium contact angle, θY. Aside from wettability, contact angle measurements are also valuable for studying the weathering and degradation of organic coatings. Wenzel’s equation has been applied to polymer photodegradation; chemical bonds are broken at the surface, polymer segments are liberated, and the contact angle is decreased [56,57]. This roughening is not unlike surface pre-treatments. Hysteresis effects also arise from compositional variations or impurities at the surface [37]. Heterogeneity in a polymer surface is almost inevitable when considering the phase separation that occurs between crystalline and amorphous regions. The Cassie-Baxter equation can be applied to chemically heterogeneous surfaces if an estimate of the partial surface area of each material is known [58]: cos = f1 cos 1 + f2 cos 2
(15)
Here, θ is the contact angle and f is the fractional surface area for the chemically homogeneous materials of respective indices. The combination of these and other variables result in contact angle measurements that are slightly ambiguous and irreproducible within small standard deviations. However, with great care, techniques such as the low-rate dynamic automated axisymmetric drop shape analysis-profile (ADSA-P) have yielded contact angle accuracy greater than 0.3° [59]. It may be of interest to determine the surface tension required of a coating to wet a given substrate. Zisman and co-workers derived a method using contact angle measurements for determining the “critical surface tension,” or the energy needed for a liquid to spontaneously wet the solid surface of interest. This is the γlv value for a contact angle of zero. Experimentally it is determined by plotting the cos θ versus γlv for various liquids of known surface tension. A linear extrapolation of the data to zero contact angle, or cos θ = 1, gives the critical surface tension. Pure liquids eliminate the preferential absorption issues associated with surfactant solutions although both have been used in practice [60,61]. Caution is advised by van Oss [42] in regards to polar liquids, especially water and aqueous solutions, as they introduce interactions and error in the measurement of thermodynamic properties. Therefore, Zisman’s technique should be used in situations where van der Waals forces dominate [62]. The study of deriving surface tensions from contact angles has evolved into two approaches: surface tension components and equations of state [43]. Surface tension components include the previously mentioned method by Zisman as well as major contributions by van Oss and Good, Fowkes, Owens and Wendt, and Wu and the acid– base approach; discussion and supplemental references can be found within Refs. [58,62]. The equation of state approach obtains the relation between the three interfacial tensions theoretically. As mentioned before, Zisman’s technique revealed a linear trend for cos θ versus γlv; Neumann and co-workers went on to show that γlv cos θ versus γlv yielded a smooth curve representative of the solid surface
Fig. 6—Schematic of the parallel curve that suggests an equation of state, found by changing the substrate. (Image duplicated from Fig. 2 of Ref. 61). Reprinted with permission of John Wiley & Sons, Inc.
tension [61,63]. Changing the substrate produced a parallel, smooth curve as shown in Fig. 6; this indicated that an equation of state exists for surface tension, which is independent of inter-molecular interactions. The equation of state was developed using contact angle measurements on polymeric substrates and does not work well for polar substrates; other criticisms also persist as indicated in the publications by Morrison [64] and Janczuk et al. [65]. One main advantage of this approach is that it alleviates the deviations seen in the Fowkes equation for large liquid and solid surface tension differences.
DYNAMIC PROPERTIES OF LIQUID SURFACES
Fresh surfaces in liquid solutions have an initial surface tension unlike their equilibrium value. Considerations of this are necessary in manufacturing processes, such as coating application, where fresh surfaces are formed. The instantaneous γlv is characteristic of the bulk solution and approaches an equilibrium surface tension, γeq, as a function of time, γ(t). The time-dependent behavior associated with fresh liquid interfaces is referred to as dynamic surface tension (DST) in the literature; this is represented by Eq (16):
eq = lim (t ) for
t→∞
(16)
A minimized free energy conformation is approached as a result of adsorption of surface active agents, or surfactants, at the liquid/vapor interface. Aside from fresh surfaces, changes in the surface area of an existing interface or bulk surfactant concentration also tend toward a new γeq and are applicable to DST models [66]. For additional details on the state-of-art, refer to the reviews by Miller et al. [67] and Eastoe and Dalton [68] for air-water interfaces and Ravera et al. [66] for liquid-liquid interfaces. By the end of the 19th century, researchers recognized that instantaneous γlv differed from γeq; however, it was not until 1946 that a quantitative model was published [69]. Utilizing Fick’s equations, Ward and Todai [69] proposed the first analysis to support the theory that surfactant
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
458
PAINT AND COATING TESTING MANUAL
Q
Fig. 7—Schematic of surfactant flux and surface excess concentration with regard to liquid surface area. (Image duplicated from Fig. 1 of Ref. 68). Reprinted from Advances in Colloid and Interface Science, Vol. 85, Eastoe, J., Dalton, J.S., “Dynamic Surface Tension and Adsorption Mechanisms of Surfactants at the Air-Water Interface,” pp. 103–144, 2000 with permission from Elsevier.
molecules were driven to the interface by diffusion processes. The work also includes definition of the subsurface, the region a few molecular diameters below the interface, which regulates adsorption to the interface. Initially the surface excess is zero and each molecule in or reaching the subsurface subsequently diffuses to the interface, represented as the adsorbing flux, jads. Over time, the number of occupied surface sites increases and adsorption is no longer automatic. The desorbing flux, jdes, or back diffusion to the subsurface becomes significant as the system continues towards equilibrium. The surface excess concentration, Γ, over time is the result of balancing the flux of surfactants to and from the interface, respectively. d / dt = jads – jdes
(17)
Fig. 7 shows how the surfactant flux and F adjust to changes in the liquid surface area where Γeq is the equilibrium surface excess concentration. Significant contributions to the DST models were later made by Sutherland [70], Hansen [71], and Miller and co-workers [67,72]. References [28,29] discuss the importance of dynamic effects in relation to micellar solutions. Two models persist for DST; they are the diffusioncontrolled model and the mixed kinetic-diffusion model. In the diffusion-controlled model, diffusion to the subsurface is the rate-limiting step followed by a fast adsorption to the interface as theorized for short-time scenarios. On the contrary, the mixed kinetic-diffusion model assumes the transfer of surfactant molecules from the subsurface to the interface to be the rate-limiting step. Also, there is an energy barrier for adsorption which is possibly related to the number of available sites for adsorption or increased surface pressure [68]. Comparatively, the diffusion-controlled model contains one kinetic parameter, that is diffusion, D, whereas the mixed kinetic-diffusion model contains an additional parameter for the adsorption energy barrier denoted by β [73]. Once a surfactant molecule has adsorbed at the interface, additional reorientation and relaxation further minimizes the surface excess energy, or surface tension. Both models are used experimentally in association with a chosen adsorption isotherm to measure γeq and Γeq
15TH EDITION
for surfaces of interest. Experimental techniques for measuring DST are explained in the following section along with those for static surface tension. It must be understood that DST is interested in solution, particularly surfactant solutions, and the appropriate measurement should reflect this. For example, the Wilhelmy Plate method is appropriate for solutions whereas Du Nuoy ring is reserved for pure liquids. DST is especially important in coating application and all other processes that generate fresh surfaces [35,74]. Wetting in thin film spreading [75] and the dynamics effects of contact angles on moving substrates [76] are also pertinent. Schunk and Scriven [28] review the dynamic effects of surfactants in continuous application processes and give an extensive bibliography. More recent studies of dynamic wetting effects include wetting phenomena at structured surfaces, and surfaces of variable composition [77,78].
MEASUREMENT OF THE SURFACE TENSION OF LIQUIDS Introduction
There are many ways of measuring surface tension, but in essence, they are all related to two effects of capillarity. The first effect is the excess pressure due to surface tension at a curved interface. The Young–Laplace equation describes this as Δp = (1 / R1 + 1 / R2 )
(18)
where ∆p is the excess pressure due to the curved interface, and R1 and R2 are the two principal radii of curvature of the interface. In the case of a sphere of radius r, where the radii of curvature are both equal to r, this equation reduces to
p = 2 /r
(19)
The other capillary effect is that the surface tension of a liquid exerts a force upon a solid body immersed in it equal to the surface tension times the perimeter of that body times the contact angle the liquid makes with the solid. If one is using a balance, one can write
W = P cos
(20)
where P is the perimeter of the solid, θ is the contact angle of the solid-liquid interface, and ∆W is the extra force on the solid body due to surface tension. These two effects yield—together with various modifications of geometry, etc.—many methods to measure static and dynamic surface tension. The most widely used methods are described in the following sections. It should be noted that all surface tension measurements require the cleanliness of all apparatus and the purity of all materials. Organic impurities in aqueous systems will have drastic effects in reducing the surface tension values measured. The concentration levels necessary to alter surface tension measurements can be as little as 10–8 M. Trace amounts of impurities on solid apparatus surfaces can alter contact angles and, as will be shown, the measured surface tension values. All water used in surface tension measurements should be at least double distilled, and often the presence of a strong oxidizing agent in one step of the distillation ensures that trace surfactants are removed. The water should be used fresh, as surface active
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 37
Q
SURFACE ENERGETICS
459
impurities can be leached from glass and plastic containers. The same holds true for all solvents used in surface characterization studies.
Static Surface Tension Measurements
Certain surface tension measurement methods work well only in a static system. These methods must thus be used on systems where there is no formation of fresh surfaces, or with pure liquids.
Capillary Rise
Within a small capillary tube, there exists a pressure on a liquid relative to the pressure on a large vessel of the liquid. Using Eq (19) and allowing for the contact angle between the liquid and the capillary, we have at equilibrium between the force of gravity and the capillary pressure Δ gh =
2 cos r
(21)
where ∆p is the density difference between the liquid and air, g is the acceleration of gravity, h is the height of the capillary rise, and the rest of the terms are defined above. This equation allows one to measure the surface tension from a simple measurement of the height of the rise of the liquid in a capillary of known radius and contact angle with the test liquid. This method is shown schematically in Fig. 8. As a measurement technique, it has remained largely unchanged since its inception, with only the addition of computer-aided analysis of the instrument.
Drop Weight
The weight of a drop when it is formed slowly is the weight just to exceed the force of surface tension times the radius of the capillary tip from which it is formed, with a correction factor required for the formation of small satellite drops. Thus, one has W = 2 rf
(22)
where W is the weight of the drop and f is the correction factor, which is discussed in Adamson [1]. This method is simple to use and accurate if precautions are taken for cleanliness and very slow flow rates of the liquid in the formation of drops. The method is shown schematically in Fig. 9. The simplicity of this measurement, and easy availability of the
Fig. 8—Schematic of capillary rise method of measuring surface tension.
Fig. 9—Schematic of drop weight method of measuring surface tension. (Image duplicated from Fig. 1 of Ref. 79). “A Critical Review: Surface and Interfacial Tension Measurement by the Drop Weight Method,” Boon-Beng, L., Ravindra, P., Chan, E.-S., Chem. Eng. Comm, Vol. 195, 2008, pp. 889–924, reprinted by permission of the publication (Taylor & Francis Group, http://www.informaworld.com).
equipment, results in it being a popular technique. Reference [79] provides a review of this technique.
Du Nouy Ring
Applying Eq (13) to the case where the object is a ring being pulled from the surface of a liquid, one has W = Wring + 4 R
(23)
where W is the total weight sensed by a balance, Wring is the ring weight, and R is the ring radius as shown in Fig. 10. However, this formula requires a correction to be accurate and holds only for a zero contact angle between the ring and test liquid. Again, see Adamson [1] for further details. The force measured by the balance is that force just at detachment of the ring and so involves motion of the sensor; the method is thus not an appropriate technique to use in studying solutions. In general, this method is infrequently used, compared to the other options available.
Fig. 10—Schematic of Du Nouy ring method of measuring surface tension.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
460
PAINT AND COATING TESTING MANUAL
Q
Fig. 11—Schematic of Wilhelmy plate method of measuring surface tension.
Wilhelmy Plate
If one uses a thin plate instead of a ring and measures the force exerted on the plate just touching the surface of a liquid, one has W = Wplate + P
(24a)
where P is the perimeter of the plate. If the liquid does not perfectly wet the plate, the expression W = Wplate + P cos
(24b)
must be used as illustrated in Fig. 11. The perimeter of the plate may be determined in practice as an experimental constant from γ measurements with liquids of known surface tension. Again cleanliness is most crucial, and a roughened platinum plate cleaned in a Bunsen burner flame is often used to create a clean plate that is wet by most liquids. Because there is no motion of the plate in the measurement, this method can be used to measure dynamic as well as static values of γ. This technique may be used as well with a rod [80] or cone [81] instead of a plate and may also be used to measure the contact angle of a liquid of known surface tension against a specific plate, fiber, or cone-shaped substrate (see later in this chapter under “Contact Angle Measurements”). Also, the method does not require corrections as with the drop weight or the Du Nouy ring methods.
Sessile or Pendant Drop Shape Methods
If the shape of a sessile or pendant drop is measured photographically or by a digital camera, a solution of Eq (18) for the specific shape of the drop in the presence of a
Fig. 12—Schematic of sessile and pendant drop methods of measuring surface tension.
15TH EDITION
Fig. 13—Schematic of maximum bubble pressure method of measuring surface tension.
gravitational field can be used with the measurements of the drop profile to back calculate the surface tension of the drop. Fig. 12 gives a schematic of the method. This method works well at liquid–air or liquid–liquid interfaces and can also be used to make dynamic measurements if the camera response is continuous and rapid [82]. A variation of the drop-shape method of measurement is used wherein the drop is deformed by a centrifugal field or an electrical field. The spinning drop shape method allows very small interfacial tensions to be measured between two liquids [83].
Maximum Bubble Pressure Methods
A bubble created by forcing a gas through the end of a capillary in a liquid and the maximum pressure on the resultant bubble can be used to measure surface tension by Eq (19). This measurement procedure is illustrated in Fig. 13. However, this is another measurement that requires correction, in this case for the nonspherical shape of the bubble. With rapid formation of the bubbles and accurate differential pressure measurements, the technique may be used to make dynamic measurements [84]. If capillaries of two different radii are used simultaneously and the pressure difference between the bubbles formed and the two capillaries is measured, a rapid, accurate measurement of surface tension can be made and has been used as a process monitor in emulsion polymerization [85]. This is known as the differential bubble pressure method and is illustrated in Fig. 14. This technique has been improved by the use of computers with the ADSA/ ABSA technique [86,87].
Dynamic Surface Tension Methods
When confronted with situations where fresh surfaces are formed, flow is occurring, or there are polymer solutions with long equilibrium times, it becomes necessary
Fig. 14—Schematic of differential bubble pressure method of measuring surface tension.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 37
Q
SURFACE ENERGETICS
461
Fig. 15—Schematic of oscillating jet method of measuring surface tension.
to consider dynamic surface tension. As previously mentioned, bubble pressure, drop-shape methods, and the Wilhelmy plate method can be used for static as well as dynamic γ measurements. These methods are appropriate only for relatively slow changes in γ, and they will be discussed briefly below with literature references that discuss their use in characterizing γ(t) effects. In addition to the already mentioned methods, other techniques have been developed to measure γ for short-time scales and fresh, rapidly changing surfaces.
Oscillating Jet
The mechanical instability of a liquid jet leaving an elliptical orifice causes the stream to oscillate about a circular shape and eventually break up into droplets, as illustrated in Fig. 15. The oscillations are periodic, and measurement of their wavelength can be used to determine γ as given by
app
⎛ 37b2 ⎞ 4 2 ⎜1 + ⎟ 24r 2 ⎠ ⎝ = ⎛ 5 2 r 2 ⎞ 6 r2 ⎜1 + ⎟ 32 ⎠ ⎝
(25)
where ρ is the liquid density, ν is the velocity of the jet, λ is the wavelength of the oscillation, r is the sum of the minimum and maximum radii, and b is their difference [1]. This dynamic measurement of γ has been considered by Vijian and Ponter [88], who give references to the earlier work of Ray-leigh, Bohr, Sutherland, Hansen, and others.
Falling Curtain
In an analysis of curtain coating, Brown [89] observed that the angle θ in the falling curtain formed by the break around a small, non-wettable obstacle could be used to measure surface tension as sin =
2 Qu
(26a)
where Q is the mass flow rate per unit orifice slit length in the curtain and u is the velocity of the liquid. This can enable one to calculate the dynamic surface tension in the falling film at the point where the obstacle intersects the falling curtain if the velocity u is greater than the “bursting velocity” of the falling sheet. A schematic of this method is given in Fig. 16. This has been considered further by Van Havenburgh and Joos [90]. Antoniades et al. [91] reconsidered the work of Brown and suggested that Eq (26), when derived correctly, should be written as sin 2 =
2 Qu
(26b)
and give experimental evidence for this corrected equation.
Fig. 16—Schematic of falling film method of measuring surface tension.
Capillary Waves
The properties of capillary waves, small wavelength waves on the surface of a liquid with the dominant restoring force being surface tension, can be studied to provide a measurement of surface tension, surface elasticity, and other dynamic surface properties. For clean surfaces of an inviscid liquid, Kelvin [92] determined that on a liquid of density, ρ, waves of frequency, ω, and wave number k, the surface tension is given for small wavelengths.
=
2 k3
(27)
The situation for interfaces in real systems, viscous liquids with surfactants present, has been studied in detail [15,17], and the use of mechanically generated capillary waves for studying interfaces has been reviewed by Hansen and Ahmad [93]. The capillary waves generated by temperature fluctuations at a surface can be measured by laser-light scattering techniques, and these data can be used to generate very accurate values of surface and interfacial tensions [94]. Capillary wave studies can be used to study the time dependence of γ in some detail and are probably the most accurate and complete of the methods for measuring all dynamic surface properties.
The Falling Meniscus Method
The measurement of the height of a column of liquid in a tube with a small opening on top can be used to measure the dynamic surface tension of aqueous systems. The height measurement versus time may be analyzed to calculate the surface tension as a function of time. This is discussed in further detail, together with a full description of the experimental apparatus and analytical equations, by Defay and Hommeln [95].
Modified Static Surface Tension Measurements Maximum Bubble Pressure
If one can monitor the time response of pressure to the time-dependent surface tension, one can use the maximum bubble pressure technique described above to measure dynamic surface tension [84]. The equipment required for such measurements is the same as the static maximum bubble technique plus instrumentation for time-dependent measurement of the bubble pressure. Various authors have examined the theoretical and experimental aspects of these methods [96,97].
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
462
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Wilhelmy Plate
If a time-recording balance is used to monitor the force of surface tension pulling on the immersed plate, the Wilhelmy plate technique can be used for characterizing dynamic surface tension. As discussed above, the method is not useful for short times, but for slowly varying γ(t) values. The method is used considerably with Langmuir film balances and has been applied to various other problems [98].
Other Methods
If a sensing technique that can be time resolved is coupled to a specific static surface tension measurement, γ(t) data can be acquired. Reference [82] discusses this for dropshape methods, and Jho and Burke [99] present a modification of the drop weight technique for γ(t) characterization. Reference [29] discusses the general issue of characterizing surfactant effects in dynamic systems.
CONTACT ANGLE MEASUREMENTS Introduction
Contact angle measurement remains the most common method for determining γsv or wettability. Arguments suggest, however, that θY is difficult to obtain for a variety of reasons. Surface roughness results in contact angle hysteresis, as mentioned earlier, in which θa gives the closer fit to θY. A rigid, physically smooth, and chemically homogeneous surface should be used in order to get the best approximation. Other assumptions include: the use of pure liquids to eliminate preferential adsorption effects, physical and chemical reactions do not occur between phases, γlv is higher than γsv such that a drop forms on the surface, and that γsv is constant and independent of the liquid used. The review by Kwok and Neumann [61] includes further details on contact angle measurement techniques and their interpretations. In addition to the common contact angle measurements, a partial list of practiced solid surface tension measurement techniques not discussed here includes: solidification front techniques, sedimentation techniques, gradient theory, film flotation, Lifshitz theory of van der Waals forces, and theory of molecular interactions. All approaches share in common the assumed validity of Young’s equilibrium equation. References relevant to these solid surface tension determining techniques are given in the review by Tavana and Neumann [63]. Recent work has also been done to determine contact angles at high temperatures [100] and under flow conditions [101]. This chapter includes information on direct/indirect sessile drop methods, Wilhelmy plate, capillary rise, rate of penetration, ADSA-P, and atomic force microscopy (AFM).
Direct Measurement of Contact Angle by Sessile Drop
The common contact angle measurement is also known as the sessile drop measurement and refers to optical techniques for determining contact angles [102]. These include contact angle goniometry of the drop profile using a photograph or digital video image, see Fig. 17. The drop must be small enough that gravitational forces have insignificant effects on the drop shape; therefore, magnification is necessary either using a digital image or goniometer-fitted microscope. Once the contact angle is known, Eq (10) can be used to calculate the surface tension of the solid if desired. Often
Fig. 17—Schematic of contact angle goniometer.
experimentation with two different liquids is necessary to obtain two equations with two unknowns and solve for γsv algebraically. The contact angle measured by this method is referred to as the equilibrium angle, θe, which lies between θa and θr but nearer to the former and is not necessarily equal to θY. It may be desirable to know the values for θa and θr; this is accomplished by slowly adding or removing liquid from the sessile drop until the maximum and minimum contact angles are recorded, respectively [63]. Differences of 5° to 20° are common between the two angles. The method was first performed by Zisman and co-workers using a thin platinum wire to add liquid and a fine glass capillary to remove liquid [63]. Later, Good and co-workers [103] applied the technique using micrometer syringes. Two similar techniques for measuring the contact angle are the captive bubble method and the tilting plate method [104]. The captive bubble method places a drop of air on the solid surface in a liquid environment instead of vice versa. The same contact angle is measured as before, however, it is now nearer to θr. Although this method deviates from the simplicity of a liquid drop on a surface, it assures one that the saturated vapor pressure of the liquid and the solid-vapor interface are in equilibrium [63]. The tilting plate method suspends a plate or cylinder of the solid material in a bath of liquid with known γlv. The solid is then tilted until the capillary rise on the plate is zero, such that the liquid-vapor interface is flat at the contact line. A simple goniometer scale gives the contact angle to be the plate’s displacement from the position normal to the liquid. Placing a sessile drop on the tilted plate at this position gives a good approximation of θa and θr. Considerations must be taken when choosing the liquid used in this method as it requires appreciable quantities. Contact angle measurements are considered to be dynamic when there is a time dependency. For example, θa and θr are found through dynamic techniques in which the drop volume is steadily increased or decreased, respectively. The key factor is that the contact line is not static; the review by Blake [105] contains theoretical information on dynamic contact angles, velocity of the wetting line, and the models that describe them.
Measurement of Sessile Drop Dimensions
Using the same drop preparation as the direct measurement, the contact angle may be calculated through trigonometric equations by measuring drop diameter and height, volume, or mass (only if the liquid is of known density to obtain drop volume). Reference [62] further explains the methods and derives the necessary equations. The technique assumes that the drop is a section of a sphere of
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 37
Q
SURFACE ENERGETICS
463
radius, r, such that the surface tension is satisfied by the Laplace equation
WA is known, the Young equation can be combined with Eq (12) to obtain the Young-Dupre equation [37]:
Δp = 2 /r
cos = WA / lv – 1
(28)
where ∆p is the difference in pressure across the liquidvapor interface. Error is commonly associated surface heterogeneities, leading to variations in the diameter and contact angle along the drop perimeter.
Wilhelmy Plate
The equation for Wilhelmy plate method is adjusted to give the contact angle of a solid if the liquid surface tension is known. Eqs (29a) and (29b) both satisfy the solution of this method. cos = ΔW / P
(29a)
sin = 1 – ( gh2 /2 sl )
(29b)
As before, the surface tension of the solid is determined from the Young equation using this contact angle. This is the static Wilhelmy plate method; information on the dynamic method is given above and by Erbil [62].
Capillary Rise Method
The capillary rise method is also used to determine contact angles: cos = Δ rgh/2 sl
(30)
Automated Curve-Fitting Approaches
In an effort to enhance the reliability of contact angle measurements, curve-fitting programs have been developed to analyze the profile of a sessile drop by transferring the image to a computer for digitizatiion and analysis [58]. The first such approach ADSA-P by Neumann and co-workers [106,107]; adjustments have consistently been made to improve its applicability as a high-accuracy technique. Assuming that the profile is both Laplacian and axisymmetric, an integration of the Laplace equation gives the contact angle [104]. The method is used as a low-rate dynamic technique to measure the advancing and receding contact angles [61]. A similar approach, the automated polynomial fitting, was proposed using a polynomial fitting of the drop profile. This technique has similar accuracy to ADSA-P but is applicable to non-axisymmetric systems also [108].
Direct Force Measurement by AFM
Drelich et al. [109] have reviewed the use of AFM in determining the surface tension of a solid. Although this technique is very sensitive to experimental conditions and requires extensive setup and calibration, it is applicable to microscopic surfaces that are too small for regular sessile drop techniques [104]. Corrections must also be made for the elastic deformation of the materials. The theory for this technique is based on the work of adhesion between two materials, being the solid surface of interest and the instrument probe. The pull-off (adhesion) force, F, is measured directly in this experiment as a function of WA F = 2 RWA
(31)
where R is the diameter of the probing tip. The coefficient 2π assumes the particles in contact are perfectly rigid. Once
(32)
Similarly the surface tension can be solved using a constant, c, generally 1.5 or 2 depending on the model chosen; more details can be found in Ref. [109].
sv = F/2c R
(33)
Rate of Penetration into Powder
The rate of penetration relation is utilized when it is desirable to know the amount of time it will take for a liquid to wet a porous medium; examples include enhanced oil recovery operations, the movement of water through geological systems, or absorption in the clean-up of spilled liquids [110]. In this dynamic contact angle measurement, the liquid enters the powder at a given velocity, ν, and subsequently wets the surfaces by capillary forces as a function of γlv and θ as shown in the Washburn equation [111]: cos = 4 l/ lv r
(34)
Here, η is the viscosity of the penetrating liquid, l is the length of the capillary path in the powder, and r is the average pore size; the equation assumes inertial and gravitational effects to be negligible. Often the technique is used to find the γlv of the wetting liquid that minimizes the rate of penetration. Refer to Lavi and Marmur [110] or Subrahman-yam et al. [112] for more information on this method.
SOME SPECIFIC APPLICATIONS OF SURFACE ENERGETICS TO ORGANIC COATINGS Coatings Application and Defects
Flow phenomena, and their control, at surfaces are very important to coatings film application technology [12,113]. It is such an important consideration that there have been several books written on this topic since the first version of this chapter [12,114,115]. The creation of uniform thin films at high speeds involves many problems that are determined by a combination of the application geometry, the velocity of the substrate relative to the liquid coating, the physical properties of the liquid coating, especially the energetics of the surface of the liquid coating, the surface of the substrate, and the new liquid/solid and gas/liquid interfaces created in the application processes. Analysis of several application techniques is given below to illustrate the importance of the use of concepts discussed in the prior portion of this chapter.
Dip Coating
Dip coating is a coating method in which a sheet of material is constantly withdrawn vertically from a coating liquid bath at constant velocity, producing a constant film thickness coating adhering to the sheet, which is usually then dried or cured to produce the final coating on the sheet (see Fig. 18). This problem was originally analyzed by Landau and Levich [116], and its solution has been given in the most commonly used form by Probstein [117]. Only the solution of the problem will be given here, but it is representative of the problems identified by chemical engineers as coating flows. In these cases, the geometry of the coating device, the relative velocity, ν, of the object to be coated, and
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
464
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
where g is the acceleration of gravity. Since the flux of material is constant through the slot, we know that Q = v0 h0 = hv
(38)
and therefore h( x) =
the physical properties of the coating fluid (surface tension, γ; viscosity, η; density, ρ; and acceleration of gravity, g) fix the solution to the coating flow problem. In this case, the limiting film thickness, δf, at large distances from the coating bath surface, is given by ( )2/ 3
1/ 6 g
(35)
This solution is derived from the physical effects of surface tension, as described above, plus the Navier-Stokes equation fluid mechanics. These authors assume Newtonian viscosity behavior of the liquid, i.e., the viscosity, η, is independent of shear rate, time-independent surface tension, and an absence of surface tension gradient-driven flows (Marangoni flows), with the appropriate boundary conditions determined by the coating process. This solution can be used by the coating designer and user to establish a first order estimate of the film thickness of the liquid film by the properties of the coating and the velocity of sheet withdrawal. Further analyses of the dip coating processes are given in more recent monographs [3,12,118].
( x) =
As described above in the section on measurement of dynamic surface tension, analysis of the curtain coating process has yielded a measurement method for dynamic surface tension. In the analyses of this problem, Brown [89] and Ha-venburgh and Joos [96] both modeled curtain coating by a falling liquid film held in place by two vertical wires (see Fig. 19). The flux Q of falling liquid is
Qv > 2
v2 = v02 + 2gx
(37)
(41)
The application of surface energetics to the problem of charged liquid droplets was first done by Rayleigh [121]. In this study, the electrical repulsion forces between charges on a drop are equated to the surface tension forces holding a drop together to give the following equation as an estimate of the upper limit on the charge that can be held on a drop q = 16 r 2
(42)
where ε is the dielectric constant of the drop and r is the drop radius. This can be used to gain an estimate of the drop size in electrostatic spray of liquid paints. Further stability analysis of this problem indicates that if a drop at this charge limit breaks up, the total energy of the system will be minimized if it does so into n droplets, where n is given by n=
q2 4 r02
(43)
where r0 is the radius of the original drop [122]. This charge limit has been studied and verified experimentally by several workers [123,124] and also applied to problems of spray painting [125]. All of this work further illustrates the importance and extensive applications of surface energetics to coatings use. A computer simulation of electrostatic spray processes has been used to predict spray patterns on planar substrates [126].
Powder Coating
Fig. 19—Schematic of curtain coater.
(40)
Electrostatic Spray
(36)
where ν0 is the velocity of the falling film and h0 is the width of the falling film at time zero at the slot through which the liquid exits. The velocity of the falling film at a position x from the slot is given by
h( x) v02 2
These equations give the application engineer a reasonably complete description of the curtain coating process. Stokes and Evans [3] and Guthoff [119] give more recent descriptions with a discussion of multilayered curtain coating processes. Fig. 20 illustrates the system discussed in these references. Miamoto and Katagari [120] give an extensive discussion of curtain coating stability with respect to the falling film and and give the following criteria for stability
Curtain Coating
Q = v0 h0
(39)
This equation gives the user a description of the thickness of liquid film as a function of distance from the slot of the curtain coater. This can also be used to estimate the surface tension at a given position x by the following relation [96]:
Fig. 18—Schematic of dip coating process.
f = 0.946
Q v02 + 2 gx
An environmentally friendly process that often yields highperformance coating films is the powder coating process. This process is one of the fastest growing in terms of market for recently introduced application methods for organic coatings [127]. The process can be described by many of the variables that describe liquid electrostatic painting for both
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 37
Q
SURFACE ENERGETICS
465
Fig. 20—Schematic of multiple layer (A) curtain coater and (B) slide coater. (Image duplicated from Figs. 4.2 and 4.3 of Ref. 119). Reprinted with permission of John Wiley & Sons, Inc.
electrostatic spray powder coating and electrostatic fluidized bed coating [128,129].
Coating Defects
There are many ways in which surface energetics are involved in coating defects, and these are discussed in many references (for example, see Ref. [12]). There is indeed, a whole book devoted to this topic [12]. Surface defects occur by either imperfect coating applications or distortions of the film during the film formation process [130]. The control and prevention of defects in organic coating films requires a knowledge of surface energetics and film formation processes. Cratering occurs, for example, when a suface active impurity with a lower surface tension than the bulk coating spreads over the surface of a wet film displacing or thinning the coating film. The leveling process occurs because of the Laplace pressure at sites of curvature, such as brushmarks, in coatings. Bubbles stabilized by surfactant adsorption can be carried into films during application and cause subsequent imperfections. Any thinning
or non-uniformity in film is a potential site for corrosion attack on the underlying substrate, especially in purely barrier coatings [131].
Cleaning and Pretreatment of Substrates for Coating
A very important step in the total process of creating a high-performance organic coating/substrate system is the cleaning and pretreatment of the substrate prior to the coating application step. The processing time and costs of the cleaning step often exceed that of the coating and curing steps, indicating the importance often assigned to this part of the coating process by many users. The specific purpose of cleaning and pretreating substrate surfaces is to control and/or modify the surface energy of the substrate so that a coating may be successfully applied to the substrate and that the coating will subsequently adhere properly to the substrate and provide the performance desired. Cleaning, as a general rule, involves the removal of substances foreign to the substrate by a surfactant/detergent
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
466
PAINT AND COATING TESTING MANUAL
Q
solution, followed with rinsing by a solvent that leaves the substrate surface clean of contaminants. This may have to be preceded by a mechanical sandblasting, grinding, sanding, scrubbing, etc., to remove thick layers of mil scale, rust, scum, and other built-up material on what is the real substrate. Vapor or organic solvent degreasing/cleaning is also often used to yield a clean surface for coating. These mechanical cleaning steps, along with the cleaning and rinsing of the surface with low surface tension solutions that leave no residue, are best done just before the coating step lest recontamination occur while waiting for the coating step. The goal of all these cleaning procedures is to insure a uniform surface that has a uniform contact angle, usually and desirably zero, for wetting by the coating, and to insure the interface created is the coating liquid/substrate interface only. The mechanical sanding, polishing, etc., will also contribute to a lower contact angle by the surface roughness effect discussed above. Oil, dirt, rust, or other contaminants will give poor and incomplete wetting and poor adhesion of the coating at the contaminated sites. Complete wetting is further assisted by insuring the surface tension of the coating liquid is low so that all air is displaced by the coating and the contact angle between the liquid coating and the substrate is zero, or near to it. The surface tension of the liquid coating is best lowered by the polymer of the system or by surfactant additives to the liquid system. Attempts to control the surface tension by the solvent can cause problems in having the surface tension increase as the solvent evaporates. As mentioned above, control of surface defects in coating processes is often achieved by proper cleaning and handling of the objects/ substrates coated in the process [12]. Substrate pretreatment is usually performed just after the cleaning discussed above and is done to further insure complete wetting and adhesion of the coating, as well as, in the case of metallic substrates, to deposit a corrosion inhibitive layer [132]. In the case of plastic substrates, the pretreatment may be corona or flame modification to oxidize the surface layer and lower the contact angle to near zero. Metal pretreatments often involve the deposition or creation of a rough-surfaced crystalline layer of metallic phosphates, which give an easily wetted surface.
Standard Surface Phenomena Testing Methods
ASTM International is a source of testing methods and procedure documents (www.ASTM.org). Table 1 contains a listing of several ASTM documents that are related to surface technology and may be useful to those investigating in this area of science. In addition to this listing, other ASTM methods exist that may be pertinent and can be found at the above indicated web site.
SUMMARY
The concepts of surface energy, surface tension, and wetting and contact angle phenomena are of exceptional importance to the science and technology of organic coating; these are topics that are ongoing. Their understanding is vital for the proper formulation and application of coating. Many of the features of the final organic coating/ substrate system are controlled by proper understanding of the surfaces of the liquid coating and the substrate, as well as creation of a proper coating/substrate interface. Both the static and dynamic aspects of liquid surface properties
15TH EDITION
TABLE 1—ASTM standards used in testing surface phenomena (partial listing) Designation
ASTM Title
C813
Hydrophobic Contamination on Glass by Contact Angle Measurements
D971
Interfacial Tension of Oil Against Water by the Ring Method
D2578
Wetting Tension of Polyethylene and Polypropylene Films
D3825
Dynamic Surface Tension by the Fast-Bubble Technique
D5725
Surface Wettability and Absorbency of Sheeted Materials Using an Automated Contact Angle Tester
D5946
Corona-Treated Polymer Films Using Water Contact Angle Measurement
D7334
Surface Wettability of Coatings, Substrates and Pigments by Advancing Contact Angle Measurement
D7490
Measurement of the Surface Tension of Solid Coatings, Substrates and Pigments Using Contact Angle Measurements
D7541
Standard Practice for Estimating Critical Surface Tensions
should be considered and the dynamic aspects properly accounted for in coating production and application. These concepts have been reviewed, and references to further reading in this important area of organic coatings science have been given.
References [1] Adamson, A. W., Physical Chemistry of Surfaces, 5th ed., Wiley, New York, 1990, Chaps. 2–3. [2] Bakker, G., “Kapillarität und Oberflachenspannung,” Handbuch der Experimentalphysik, Vol. VI, Akadem, Verlag, Leipzig, 1928. [3] Stokes, R. J., and Evans, D. F., Fundamentals of Interfacial Engineering, Wiley-VCH, New York, 1997. [4] Butt, H. J., Graf, K., and Kappl, M., Physics and Chemistry of Interfaces, Wiley-VCH Verlag, Weinheim, 2003. [5] Butt, H.-J., Berger, R., Bonasccurso, E., Chen, Y., and Wang, J., “Impact of AFM on Interface and Colloid Science,” Adv. Colloid Interface Set, Vol. 133, 2007, pp. 91–104. [6] Wolf, D. E., Griffiths, R. B., and Tang, L., “Surface Stress and Surface Tension for Solid-Vapor Interfaces,” Surf. Sci., Vol. 162, 1985, p. 114. [7] Mueller, R., and Saul, A., “Elastic Effects on Surface Physics,” Surf. Sci. Rep., Vol. 54, 2004, pp. 157–258. [8] Walton, J. R R. B., Tildesley, D. J., and Rowlinson, J., Mol. Phys., Vol. 50, 1983, p. 1357. [9] Jones, R. A. L., Soft Condensed Matter, Oxford University Press, Oxford, UK, 2004, Chap. 9. [10] Liapatov, Y. S., Polymer Reinforcement, Chemtec Publishing, Toronto, 1995. [11] Braun, J. H., “Titanium Dioxide’s Contribution to the Durability of Paint Films,” Prog. Org. Coat., Vol. 15, 1987, pp. 249–260. [12] Bierwagen, G. P., “Surface Defects and Surface Flows in Coatings,” Prog. Org. Coat., Vol. 19, 1991, pp. 59–68. [13] Cohen, E. D., and Guthoff, E. B., Coating and Drying Defects, John Wiley & Sons, New York, 1995.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 37
[14] Burshtein, A. I., “Simple Liquid Surface Structure and Surface Tension,” Adv. Colloid Interface Sci., Vol. 11, 1979, pp. 315–374. [15] Navascues, G., “Liquid Surfaces: Theories of Surface Tension,” Rep. Prog. Phys., Vol. 42, 1979, pp. 1132–1186. [16] Davis, H. T., Statistical Mechanics of Phases, Interfaces and Thin Films, VCH, New York, 1996, Ch. 7–8. [17] Rowlinson, J. S., and Widom, B., Molecular Theory of Capillarity, Clarendon Press, Oxford, UK, 1982, Ch. 2–4. [18] Guggenheim, E. A., J. Chem. Phys., Vol. 13, 1945, p. 253. [19] Pitzer, K. S., and Brewer, L., Thermodynamics, 2nd ed., McGraw-Hill, New York, 1961, Ch. 29. [20] Adamson, A. W., and Gast, A. P., Physical Chemistry of Surfaces, 6th Ed., Wiley Interscience, New York, 1997, Chaps. 1–3. [21] Meyers, D., Surfactant Science and Technology, VCH Publishers, 1988, Ch. 3. [22] Rosen, M. J., Surfactants and Interfacial Phenomena, Wiley & Sons, New York, 1978. [23] Tanford, C., The Hydrophobic Effect, Wiley & Sons, New York, 1980. [24] Hamley, I. W., Introduction to Soft Matter, John Wiley & Sons, Chichester, UK, 2000, Chaps. 1–4. [25] Noskov, B. A., “Kinetics of Adsorption from Micellar Solutions,” Adv. Colloid Interface Sci., Vol. 95, 2002, pp. 237–293. [26] Rusanov, A. I., and Krotov, V. V., “Gibbs Elasticity of Liquid Films, Threads, and Foams,” Progress in Surface and Membrane Science, Vol. 13, J. Danielli, Ed., Academic Press, New York, 1979. [27] Tadros, T., “Polymeric Surfactants in Disperse Systems,” Adv. Colloid Interface Sci., Vol. 147–148, 2009, pp. 281–299. [28] Schunk, P. R., and Scriven, L. E., “Surfactant Effects in Coating Processes,” Liquid Film Coating, S. F. Kistler and P. M. Schweizer, Eds., Chapman & Hall, London, 1997, Chap. 11d. [29] Lucassen-Reynders, E. H., Lucassen, J., Garrett, P. R., Giles, D., and Hollway, F., “Dynamic Surface Measurements as a Tool to Obtain Equation-of-State Data for Soluble Monolayers,” Advances in Chemistry Series, No. 145, American Chemical Society, Washington, DC, 1975, pp. 275–285, Chap. 21. [30] Hansen, R. S., and Mann, J. A., “Propagation Characteristics of Capillary Ripples,” J. Appl. Phys., Vol. 35, 1964, pp. 152–158. [31] Hansen, R. S., Lucassen, J., Bendure, R. L., and Bierwagen, G. P., “Propagation Characteristics of Interfacial Ripples,” J. Colloid Interface Sci., Vol. 26, 1968, p. 198. [32] Lucassen, J., and Hansen, R. S., J. Colloid Interface Sci., Vol. 22, 1966, p. 32. [33] Lucassen, J., “Dynamic Properties of Free Liquid Films and Foams,” Physical Chemistry of Anionic Surfactants, Vol. 6 in Surfactant Science, E. H. Lucassen-Reynders and M. Schiff, Eds., MarcellDekker, New York, 1981, Chap. 6. [34] Benney, D. J., Gutoff, E. B., and Foley, J. A., “The Effect of Surface Elasticity on the Stability of Flow Down an Inclined Plane,” Pre-Print 3544, 1979, Session on Fundamental Research in Fluid Mechanics, 72nd Annual Meeting, American Institute of Chemical Engineering, San Francisco, Nov. 25–29. [35] Witten, T. A., and Pincus, P. A., Structured Fluids; Polymers, Colloids, Surfactants, Oxford, New York, 2004, pp. 151–172. [36] Young, T., Philos. Trans. R. Soc. London, Vol. 95, 1805, pp. 65–82. [37] Kumar, G. K., and Prabhu, N., “Review of Non-Reactive and Reactive Wetting of Liquids on Surfaces,” Adv. Colloid Interface Sci., Vol. 133, 2007, pp. 61–89. [38] Churaev, N. V., and Sobolev, V. D., “Wetting of Low-Energy Surfaces,” Adv. Colloid Interface Sci., Vol. 134–135, 2007, pp. 15–23. [39] Chibowski, E., “On Some Relations Between Advancing, Receding and Young’s Contact Angles,” Adv. Colloid Interface Sci., Vol. 133, 2007, pp. 51–59. [40] Packham, D. E., “Surface Energy, Surface Topography and Adhesion,” Int. J. Adhes. Adhes., Vol. 23, 2003, pp. 437–448. [41] de Gennes, P.-G., Brochard-Wyart, F., and Quéré, D., Capillarity and Wetting Phenomena: Drops, Bubbles, Pearls, Waves, Springer, New York, 2004, Chap. 1.
Q
SURFACE ENERGETICS
467
[42] van Oss, C. J., Giese, R. F., and Docoslis, A., “Hyperhydrophobicity of the Water-Air Interface,” J. Dispersion Sci. Technol., Vol. 26, 2005, pp. 585–590. [43] Tuteja, A., Choi, W., McKinley, G. H., Cohen, R. E., and Rubner, M. F., “Design Parameters for Superhydrophobicity and Superoleophobicity,” MRS Bull., Vol. 33, 2008, pp. 752–758. [44] Gao, L., Fadeev, A. Y., and McCarthy, T. J., “Superhydrophobicity and Contact-Line Issues,” MRS Bull., Vol. 33, 2008, pp. 747–751. [45] Youngblood, J. P., and Sottos, N. R., “Bioinspired Materials for Self-Cleaning and Self-Healing,” MRS Bull., Vol. 33, 2008, pp. 732–738. [46] Ferrari, M., “Surfactants Adsorption at Hydrophobic and Superhydrophobic Solid Surfaces,” Contact Angle, Wettability, and Adhesion, Vol. 5, K. L. Mittal, Ed., VSP, Leiden, The Netherlands, 2008, pp. 295–308. [47] Genzer, J., and Marmur, A., “Biological and Synthetic SelfCleaning Surfaces,” MRS Bull., Vol. 33, 2008, pp. 742–746. [48] Denny, M. W., “The Intrigue of the Interface,” Science, Vol. 320, 2008, p. 886. [49] Ma, M., and Hill, R. M., “Superhydrophobic Surfaces,” Curr. Opin. Colloid Interface Sci., Vol. 11, 2006, pp. 193–202. [50] Tuteja, A., Choi, W., Ma, M., Mabry, J. M., Mazelle, S. A., Rutledge, G. C., McKinley, G. H., and Cohen, R. E., “Designing Superoleophobic Surfaces,” Science, Vol. 318, 2007, pp. 1618–1622. [51] Marmur, A., “Super-Hydrophobicity Fundamentals: Implications to Biofouling Prevention,” Biofouling, Vol. 22, 2006, 107–115. [52] Gould, P., “Smart, Clean Surfaces,” Mater. Today, Vol. 6, 2003, pp. 44–48. [53] Feng, L., Li, S., Li, Y., Li, H., Zhang, L., Zhai, J., Song, Y., Liu, B., Jiang, L., and Zhu, D., “Super-Hydrophobic Surfaces: from Natural to Artificial,” Adv. Mater., Vol. 14, 2002, pp. 1857–1860. [54] Li, X. M., Reinhoudt, D. N., and Crego-Calama, M., “What Do We Need for a Superhydrophobic Surface? A Review on the Recent Progress in the Preparation of Superhydrophobic Surfaces,” Chem. Soc. Rev., Vol. 36, 2007, pp. 1350–1368. [55] Durkee, J. B., and Kuhn, A., “Wettability Measurements and Cleanliness Evaluation without Substantial Cost,” Contact Angle, Wettability, and Adhesion, Vol. 5, K. L. Mittal, Ed., VSP, Leiden, The Netherlands, 2008, pp. 115–138. [56] Croll, S. G., “Quantitative Evaluation of Photodegradation in Coatings,” Prog. Org. Coat., Vol. 15, 1987, pp. 223–248. [57] Hinderliter, B. R., and Croll, S. G., “Monte Carlo Approach to Estimating the Photodegradation of Polymer Coatings,” JCT Res., Vol. 2, 2005, pp. 483–491. [58] Karbowiak, T., Debeaufort, F., and Voilley, A., “Importance of Surface Tension Characterization for Food, Pharmaceutical and Packaging Products: a Review,” CRC Crit. Rev. Food Sci. Nutr., Vol. 46, 2006, pp. 391–407. [59] Li, D., Xie, M., and Neumann, A. W., “Vapour Adsorption and Contact Angles on Hydrophobic Solid Surfaces,” Colloid Polym. Sci., Vol. 271, 1993, pg. 573–580. [60] Liao, W.-C., and Zatz, J. L., “Surfactant Solutions as Test Liquids for Measurement of Critical Surface Tension,” J. Pharm. Sci., Vol. 68, 1979, pp. 486–488. [61] Kwok, D. Y., and Neumann, A. W., “Contact Angle Measurement and Contact Angle Interpretation,” Adv. Colloid Interface Sci., Vol. 81, 1999, pp. 167–249. [62] Erbil, H. Y., “Surface Tension of Polymers,” Handbook of Surface and Colloid Chemistry, K. S. Birdi, Ed., CRC Press, Boca Raton, FL, 1997, pp. 265–312, Chap. 9. [63] Tavana, H., and Neumann, A. W., “Recent Progress in the Determination of Solid Surface Tensions from Contact Angles,” Adv. Colloid Interface Sci., Vol. 132, 2007, pp. 1–32. [64] Morrison, I. D., “Does the Phase Rule for Capillary Systems Really Justify an Equation of State for Interfacial Tensions?,” Langmuir, Vol. 7, pp. 1833–1836. [65] Jan´czuk, B., Bruque, J. M., González-Martín, M. L., Moreno del Pozo, J., Zdziennicka, A., and Quintana-Gragera, F., “The Usefulness of the Equation of State for Interfacial Tensions Estimation in Some Liquid-Liquid and Solid-Liquid Systems,” J. Colloid Interface Sci., Vol. 18, 1996, pp. 108–117.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
468
PAINT AND COATING TESTING MANUAL
Q
[66] Ravera, F., Ferrari, M., and Liggieri, L., “Adsorption and Partitioning of Surfactants in Liquid-Liquid Systems,” Adv. Colloid Interface Sci., Vol. 88, 2000, pp. 129–177. [67] Miller, R., Joos, P., and Fainermann, V., “Dynamic Surface and Interfacial Tensions of Surfactant and Polymer Solutions,” Adv. Colloid Interface Sci., Vol. 49, 1994, pp. 249–302. [68] Eastoe, J., and Dalton, J. S., “Dynamic Surface Tension and Adsorption Mechanisms of Surfactants at the Air-Water Interface,” Adv. Colloid Interface Sci., Vol. 85, 2000, pp. 103–144. [69] Ward, A. F. H., and Tordai, L., “Time Dependence of Boundary Tensions of Solutions. I. The Role of Diffusion in Time Effects.,” J. Chem. Phys., Vol. 14, 1946, pp. 453–461. [70] Sutherland, K. L., “The Kinetics of Adsorption at Liquid Surfaces,” Aust. J. Sci. Res., Ser. A, Vol. A5, 1952, pp. 683–696. [71] Hansen, R. S., “The Theory of Diffusion Controlled Absorption Kinetics with Accompanying Evaporation,” J. Phys. Chem., Vol. 64, 1960, pp. 637–641. [72] Fainerman, V. B., Makievski, A. V., and Miller, R., “The Analysis of Dynamic Surface Tension of Sodium Alkyl Sulfate Solutions, Based on Asymptotic Equations of Adsorption Kinetic Theory,” Colloids Surf., A, Vol. 87, 1994, pp. 61–75. [73] Moorkanikkara, S. N., and Blankschtein, D., “New Methodology to Determine the Rate-Limiting Adsorption Kinetics Mechanism from Experimental Dynamic Surface Tension Data,” J. Colloid Interface Sci., Vol. 302, 2006, pp. 1–19. [74] Bierwagen, G. P., “Surface Dynamics of Defect Formation in Paint Films,” Prog. Org. Coat., Vol. 3, 1975, p. 101, and Bierwagen, Ref. [5]. [75] de Gennes, P.-G., Brouchard-Wyart, F., and Quere, D., Capillarity & Wetting Phenomena, Springer, New York, 1997. [76] de Gennes, P. G., “The Dynamics of Wetting,” Fundamentals of Adhesion, L.-H. Lee, Ed., Plenum Press, New York, 1991, Chap. 5. [77] Ishimi, K., Hikita, H., and Esmail, M., “Dynamic Contact Angles on Moving Plates,” Am. Inst. Chem. Eng. Symp. Ser. Vol.32, 1986, pp. 486–492. [78] Mittal, K. L., Ed., Contact Angle, Wettability and Adhesion, VSP, Leiden, The Netherlands, 2008, Parts 1 &2. [79] Boon-Beng, L., Ravindra, P., and Chan, E.-S., “A Critical Review: Surface and Interfacial Tension Measurement by the Drop Weight Method,” Chem. Eng. Commun., Vol. 195, 2008, pp. 889–924. [80] Lyons, C. J., Elbing, E., and Wilson, I. R., “The Rod-in-Free Surface Technique for Surface-Tension Measurement Using Small Rods,” J. Colloid Interface Sci., Vol. 101, 1984, pp. 292–294. [81] Ugarcic, Z., Vohra, D. K., Atteya, E., and Hartland, S., “Measurement of Surface Tension Using a Vertical Cone,” J. Chem. Soc., Faraday Trans. 1, Vol. 77, 1981, pp. 49–61. [82] Girault, H. H., Schiffrin, D. J., and Smith, B. D. V., “Drop Image Processing for Surface and Interfacial Tension Measurements,” J. Electroanal. Chem., Vol. 137, 1982, pp. 207–217. [83] Seeto, Y., and Scriven, L. E., “Precision Spinning Drop Interfacial Tensiometer,” Rev. Sci. Instrum., Vol. 53, 1982, pp. 1757–1761. [84] Bendure, R. L., “Dynamic Surface Tension Determination with the Maximum Bubble Pressure Method,” J. Colloid Interface Sci., Vol.35, 1971, pp. 238–248. [85] Shork, F. J., and Ray, W. H., “On-Line Measurement of Surface Tension and Density with Applications to Emulsion Polymerization,” J. Appl. Polym. Sci., Vol. 28, 1983, pp. 407–430. [86] Loubière, K., and Hébrard, G., “Influence of Liquid Surface Tension (Surfactants) on Bubble Formation at Rigid and Flexible Orifices,” Chem. Eng. Process., Vol. 43, 2004, pp. 1361–1369. [87] Loglio, G., Pandolfini, P., Tesei, U., and Noskov, B., “Measurements of Interfacial Properties with the Axisymmetric BubbleShape Analysis Technique: Effects of Vibrations,” Colloids Surf., A, Vol. 143, 1998, pp. 301–310. [88] Vijian, S., and Ponter, A. B., “Dynamic Surface Tension Studies Using an Oscillating Jet,” Indian Chem. Eng., Sect. A, Vol. 14, 1972, pp. 26–32. [89] Brown, D. R., “A Study of the Behaviour of a Thin Sheet of Moving Liquid,” J. Fluid Mech., Vol. 10, 1961, pp. 297–305.
15TH EDITION
[90] Van Havenburgh, J., and Joos, P., “The Dynamic Surface Tension in a Free Falling Film,” J. Colloid Interface Sci., Vol. 95, 1983, pp. 172–182. [91] Antoniades, M. G., Goodwin, R., and Lin, S. P., J. Colloid Interface Sci., Vol. 77, 1989, p. 583. [92] Kelvin, L. W. T., Philos. Mag., Vol. 42, 1871, p. 368. [93] Hansen, R. S., and Ahmad, J., Progress in Surface and Membrane Science, Vol. 4, Academic Press, New York, 1971. [94] Lofgren, H., Neuman, R. D., Scriven, L. E., and Davis, H. T., “Laser Light-Scattering Measurements of Interfacial Tension Using Optical Hetrodyne Mixing Spectroscopy,” J. Colloid Interface Sci., Vol. 98, 1984, pp. 175–183. [95] Defay, R., and Hommeln, J., “II. Measurement of Dynamic Surface Tensions of Aqueous Solutions by the Falling Meniscus Method,” J. Colloid Sci., Vol. 14, 1959, pp. 401–410. [96] Joos, P., and Rillaerts, E., “Theory on the Determination of the Dynamic Surface Tension with the Drop Volume and Maximum Bubble Pressure Methods,” J. Colloid Interface Sci., Vol. 79, 1981, pp. 96–100. [97] Lunkenheimer, K., Serrien, G., and Joos, P., “The Adsorption Kinetics of Octanol at the Air Solution Interface Measured with the Oscillating Bubble and Oscillating Jet Methods,” J. Colloid Interface Sci., Vol. 134, 1990, pp. 407–411. [98] Montgomery, D. D., and Anson, F. C., “Time-Resolved Measurement of Equilibrium Surface Tensions at the Electrified Mercury-Aqueous NaF Interphase by the Method of Wilhelmy,” Langmuir, Vol. 7, 1991, pp. 1000–1004. [99] Jho, C., and Burke, R., “Drop Weight Technique for the Measurement of Dynamic Surface Tension,” J. Colloid Interface Sci., Vol. 95, 1983, pp. 61–71. [100] Eustathopoulos, N., Sobczak, N., Passerone, A., and Nogi, K., “Measurement of Contact Angle and Work of Adhesion at High Temperature,” J. Mater. Sci., Vol. 40, 2005, pp. 2271–2280. [101] Gajewski, A., “A Method for Contact Angle Measurements Under Flow Conditions,” Int. J. Heat Mass Transfer, Vol. 48, 2005, pp. 4829–4834. [102] Marmur, A., “Equilibrium Contact Angles: Theory and Measurement,” Colloids Surf., A, Vol. 116, 1996, pp. 55–61. [103] Neumann, A. W., and Good, R. J., “Technique of Measuring Contact Angle,” Surface and Colloid Science, Vol. II Experimental Methods, R. J. Good and R. R. Stromberg, Eds., Plenum Press, New York, 1979, Chap. 2. [104] Chau, T. T., “A Review of Techniques for Measurement of Contact Angles and Their Applicability on Mineral Surfaces,” Minerals Eng., Vol. 22, 2008, pp. 213–219. [105] Blake, T. D., “The Physics of Moving Wetting Lines,” J. Colloid Interface Sci., Vol. 299, 2006, pp. 1–13. [106] Rotenberg, Y., Boruvka, L., and Neumann, A. W., “Determination of Surface Tension and Contact Angle from the Shapes of Axisymmetric Fluid Interfaces,” J. Colloid Interface Sci., Vol. 93, 1983, pp. 169–183. [107] Hoorfar, M., and Neumann, A. W., “Axisymmetric Drop Shape Analysis (ADSA) for the Determination of Surface Tension and Contact Angle,” J. Adhes., Vol. 80, 2004, pp. 727–743. [108] Bateni, A., Susnar, S. S., Amirfazli, A., and Neumann, A. W., “A High-Accuracy Polynomial Fitting Approach to Determine Contact Angles,” Colloids Surf., A, Vol. 219, 2003, pp. 215–231. [109] Drelich, J., Tormoen, G. W., and Beach, E. R., “Determination of Solid Surface Tension from Particle-Substrate PullOff Forces Measured with the Atomic Force Microscope,” J. Colloid Interface Sci., Vol. 280, 2004, pp. 484–497. [110] Lavi, B., and Marmur, A., “The Capillary Race: Optimal Surface Tensions for Fastest Penetration,” Colloids Surf., A, Vol. 282–283, 2006, pp. 263–271. [111] Washburn, E. W., “Dynamics of Capillary Flow,” Phys. Rev., Vol. 17, 1921, pp. 374–375. [112] Subrahmanyam, T. V., Monte, M. B. M., Middea, A., Valdiviezo, E., and Lins, F. F., “Contact Angles of Quartz by Capillary Penetration of Liquids and Captive Bubble Techniques,” Minerals Eng., Vol. 12, 1999, pp. 1347–1357. [113] Bierwagen, G. P., “Film Coating Technologies and Adhesion,” Electrochim. Acta, Vol. 37, 1992, pp. 1471–1478.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 37
[114] Kistler, S. F., and Schweizer, P. M., Eds., Liquid Film Coating, Chapman & Hall, London, 1997. [115] Cohen, E. D., and Guthoff, E. B., Eds., Modern Coating and Drying Technology, VCH, New York, 1992. [116] Landau, L. D., and Levich, V. G., “Dragging of a Liquid by a Moving Plate,” Acta Physicochim. URSS, Vol. 17, 1942, pp. 42–54. [117] Probstein, R. F., Physicochemical Hydronamics, Butterworths, London, 1989, pp. 280–289. [118] Schunk, P. R., Hurd, A. J., and Brinker, C. J., “Free-Meniscus Coating Processes,” Liquid Film Coating, S. F. Kistler and P. M. Schweizer, Eds., Chapman & Hall, London, 1997, pp. 673–708, Chap. 3. [119] Guthoff, E., “Premetered Coating,” Modern Coating and Drying Technology, E. D. Cohen and E. B. Guthoff, Eds., VCH, New York, 1992, pp. 117–168, Chap. 4. [120] Miamoto, K., and Katagari, Y., “Curtain Coating,” Liquid Film Coating, S. F. Kistler and P. M. Schweizer, Eds., Chapman & Hall, London, 1997, pp. 463–492, Chap. 11c. [121] Strutt, J. W., “On the Equilibrium of Liquid Conducting Masses Charged with Electricity,” Philos. Mag. A, Vol. 14, 1882, pp. 184–186. [122] Gopal, E. S. R., Emulsion Science, P. Sherman, Ed., John Wiley, New York, 1963, Chap. 1. [123] Richardson, C. B., Pigg, A. L., and Hightower, R., “On the Stability Limit of Charged Droplets,” Proc. R. Soc. London, Ser. A, Vol. 422, 1989, pp. 319–328.
Q
SURFACE ENERGETICS
469
[124] Schweitzer, J. W., and Hanson, D. N., “Stability Limit of Charged Droplets,” J. Colloid Interface Set, Vol. 35, 1971, pp. 417–423. [125] Hines, R. L., “Electrostatic Atomization and Spray Painting,” J. Appl. Phys., Vol. 37, 1966, pp. 2730–2736. [126] Colbert, S. A., and Cairncross, R. A., “A Computer Simulation for Prediction Electrostatic Spray Coating Patterns,” Powder Technol., Vol. 151, 2005, pp. 77–86. [127] Meng, X., Zhu, J., and Zhang, H., “Influences of Different Powders on the Characteristics of Particle Charging and Deposition in Powder Coating Processes,” J. Electrost., Vol. 67, 2009, pp. 663–671. [128] Barletta, M., and Tagliaferri, V., “Influence of Process Parameters in Electrostatic Fluidized Bed Coating,” Surf. Coat. Technol., Vol. 200, 2006, pp. 4619–4629. [129] Barletta, J., Gisario, A., Guarino, S., and Tagliagerri, V., “Fluidized Bed Coating of Metal Substrates by Using High Performance Thermoplastic Powders—Statistical Approach and Neural Modeling,” Eng. Applic. Artif Intell, Vol. 21, 2008, pp. 1130–1143. [130] Kheshgi, H. S., “The Fate of Thin Liquid Films After Coating,” Liquid Film Coating, S. F. Kistler and P. M. Schweizer, Eds., Chapman & Hall, London, 1997, pp. 183–205, Chap. 6. [131] Bierwagen, G., “The Physical Chemistry of Organic Coatings Revisited—Viewing Coatings as a Materials Scientist,” J. Coat. Technol., Vol. 5, 2008, pp. 133–155. [132] Landolt, D., Corrosion and Surface Chemistry of Metals, CRC Taylor & Francis, Boca Raton, FL, 2007.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
38
MNL17-EB/Jan. 2012
Solubility Parameters Charles M. Hansen1 NOMENCLATURE c D
= =
DM ∆Ed ∆Ep ∆Eh ∆EV ∆GM H
= = = = = = =
∆Hv = ∆HM = P = R Rs RO RED ∆SM T Tb Tc VM ∆T α ␦d ␦h ␦p ␦T i X
= = = = = = = = = = = = = = = = =
Dispersion cohesion energy from Figs. 2 and 3 Dispersion cohesion (solubility) parameter—in tables and computer printouts Dipole moment—Debyes Dispersion cohesion energy Polar cohesion energy Hydrogen bonding cohesion energy Energy of vaporization (=) cohesion energy Molar free energy of mixing Hydrogen bonding cohesion (solubility) parameter— in tables and computer printouts Molar heat of vaporization Molar heat of mixing Polar cohesion (solubility) parameter—in tables and computer printouts Gas constant (1.987 cal/mole K) Distance in Hansen space Radius of interaction sphere in Hansen space Relative energy difference, RS/R0 Molar entropy of mixing Absolute temperature (Normal) boiling point Critical temperature Molar volume Lydersen critical temperature group contribution Thermal expansion coefficient Dispersion cohesion (solubility) parameter Hydrogen bonding cohesion (solubility) parameter Polar cohesion (solubility) parameter Total cohesion (solubility) parameter Volume fraction of component “i” Polymer-liquid interaction parameter (FloryHuggins)
INTRODUCTION
SOLUBILITY PARAMETERS ARE USED IN THE COATings industry to select solvents. Liquids with similar solubility parameters will be miscible, and polymers will dissolve in solvents whose solubility parameters are not too different from their own. The basic principle is “like dissolves like.” Solubility parameters help put numbers into this simple qualitative idea. The solubility parameter approach has been used for many years to select solvents for coating materials. The lack of total success has stimulated research. The skill with which solvents can be optimally selected with respect to
1
cost, solvency, workplace environment, external environment, evaporation rate, flash point, etc., has improved over the years as a result of a series of improvements in the solubility parameter concept and widespread use of computer techniques. Most, if not all, commercial suppliers of solvents have computer programs to help with solvent selection. One can now easily predict how to dissolve a given polymer in a mixture of two solvents, neither of which can dissolve the polymer by itself. This contribution to the paint testing manual unfortunately cannot include discussion of all of the significant efforts leading to our present state of knowledge of the solubility parameter. An attempt is made to outline developments, provide some background for a basic understanding, and give examples of uses in practice. The key is to determine which affinities the important components in a system have for each other. For coatings, this means affinities of solvents, polymers, pigment surfaces, additives, and substrates. It is noteworthy that the concepts presented here have developed toward not just predicting solubility, which requires high affinity between solvent and solute, but to predict affinities between different polymers leading to compatibility, and affinities to surfaces to improve pigment dispersion and adhesion. Attempts are also being made to extend these developments, largely attributable to the coatings industry, to understand affinities and phenomena for a large number of other materials not specifically related to coatings. In these applications, the solubility parameter has become a tool, using well-defined liquids as energy probes, to measure the similarity, or lack of same, of key components. Materials with widely different chemical structures may be very close in affinities. Only those materials that interact differently with different solvents can be characterized in this manner. Many inorganic materials, such as fillers, do not interact differently with these energy probes since their energies are (presumed to be) very much higher. Changing their surface energies by various treatments can lead to a surface that can be characterized. It is also known that the surfaces of such materials are covered with water. The extent to which this is bound will influence practical performance, i.e., how easily can it be replaced? Solubility parameters are cohesion energy parameters since they derive from the energy required to convert a liquid to a gas. The energy of vaporization is a direct measure of the total (cohesive) energy holding the liquid’s molecules
Jens Bornoes Vej 16, 2970 Hoersholm, Denmark.
470
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 This standard is for EDUCATIONAL USE ONLY. www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
Downloaded/printed byASTM International Copyright © 2012 by
CHAPTER 38
together. All types of bonds holding the liquid together are broken by evaporation, which has led to the concepts described in more detail below. The term cohesion energy parameter is more appropriately used when referring to surface phenomena.
HILDEBRAND PARAMETERS
The term solubility parameter was first used by Hildebrand and Scott [1,2]. The solubility parameter is the square root of the cohesive energy density 1/ 2
⎛ ΔE ⎞ = ( c ⋅ e ⋅ d )1/ 2 = ⎜ v ⎟ (cal / cm 3 )1/ 2 or MPA A1/2 ⎝ VM ⎠
(1)
where VM is the molar volume and ∆EV is the (measurable) energy of vaporization [see Eq (14)]. The solubility parameter is an important quantity in predicting solubility relations, as can be seen from the following brief introduction. Thermodynamics requires that the free energy of mixing must be zero or negative for the solution process to occur spontaneously. The free energy change for the solution process is given by the relation ΔG M = ΔH M − TΔS M
(2)
where ∆GM is the free energy of mixing, ∆HM is the heat of mixing, T is the absolute temperature, and ∆SM is the entropy change in the mixing process. The heat of mixing, ∆HM, is given by Hildebrand and Scott as ΔH M ≈ ΔE M = 1 2VM (1 − 2 )2
(3)
where the ’s are volume fractions and VM is the average molar volume of the solvent. It is important to note that the solubility parameter, or rather the difference in solubility parameters for the solvent-solute combination, is important in determining the solubility. It is clear that a match in solubility parameters leads to a zero heat of mixing, and the entropy change should ensure solution. The maximum difference in solubility parameters that can be tolerated where solution still occurs is found by setting the free energy change equal to zero in Eq (2). It is, in fact, the entropy change that dictates how closely the solubility parameters must match each other. It can also be seen that solvents with smaller molecular volumes promote lower heats of mixing, which, in turn, means that smaller solvent molecules will be thermodynamically better than larger ones when their solubility parameters are equal. A practical aspect of this effect is that solvents with relatively low molecular volumes, such as methanol and acetone, can dissolve a polymer at larger solubility parameter differences than expected from comparisons with other solvents with larger molecular volumes. The converse is also true. Larger molecular species may not dissolve even though solubility parameter considerations might predict this. This can be a difficulty with plasticizers. A first impression ADD of the Hildebrand approach is that negative heats of mixing are not possible. Likewise, the approach is limited to regular solutions as defined by Hildebrand and Scott [2] and does not account for association between molecules, such as polar and hydrogen bonding interactions would require. The latter problem
Q
SOLUBILITY PARAMETERS
471
seems to have been largely solved with the use of multicomponent solubility parameters. Patterson and co-workers have cleared up many questions in polymer solution thermodynamics [3-8]. One of the more important matters was to clearly show that negative heats of mixing are allowed by the solubility parameter theory, thus eliminating one of the major mental stumbling blocks for further use of this concept. A more detailed description of the theory presented by Hildebrand and the succession of research reports which have attempted to improve on it can be found in Barton’s extensive handbook [9]. The slightly older excellent contribution of Gardon and Teas [10] is also a good source, particularly for coatings and adhesion phenomena. The approach of Burrell [11], who divided solvents into hydrogen bonding classes, has found numerous practical applications, and the approach of Blanks and Prausnitz [12], who divided the solubility parameter into two components, nonpolar and “polar,” are worthy of mention, however, in that these have found wide use and greatly influenced the author’s earlier activities, respectively. It can be seen from Eq (2) that the entropy change can be considered beneficial to mixing. When multiplied by the temperature, this will work in the direction of promoting a more negative free energy of mixing. Higher temperatures will also promote this more negative free energy change. The entropy changes associated with polymer solutions will be smaller than those associated with liquid-liquid miscibility, for example, since the “monomers” are already bound into the configuration dictated by the polymer they make up. They are no longer free in the sense of a liquid solvent and cannot mix freely to contribute to a larger entropy change. This is one reason polymer-polymer miscibility is difficult to achieve. The free energy criterion dictates that the polymer solubility parameters match extremely well since there is little help from the entropy contribution when progressively larger molecules are involved. However, polymer-polymer miscibility can be promoted by introduction of suitable co-polymers or co-monomers that interact specifically within the system.
HANSEN SOLUBILITY PARAMETERS
A widely used solubility parameter approach to predicting polymer solubility is that proposed by the author. The basis of these so-called Hansen solubility parameters is that the total energy of vaporization of a liquid consists of several individual parts [13–17]. Needless to say, without the work of Hildebrand and Scott [1,2] and others not specifically referenced here such as Scatchard, this postulate could never have been made. The total cohesive energy, ∆Et, can be measured by evaporating the liquid, i.e., breaking all the cohesive bonds. It should also be noted that these cohesive energies arise from interactions of a given solvent molecule with another of its own kind. The basis of the approach is, therefore, very simple, and it is surprising that so many different applications have been possible over the past 40 years. A number of applications are discussed below. A lucid discussion by Barton [18] enumerates typical situations where problems occur when using solubility parameters. These are most often where the environment causes the solvent molecules to interact with or within themselves differently than when they make up their own environment, i.e., as pure liquids.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
472
PAINT AND COATING TESTING MANUAL
Q
Materials having similar (Hansen) solubility parameters have high affinity for each other. The extent of the similarity in a given situation determines the extent of the interaction. The same cannot be said of the total or Hildebrand solubility parameter [1,2]. Ethanol and nitromethane, for example, have similar total solubility parameters (26.1 versus 25.1 MPa1/2, respectively), but their affinities are quite different. Ethanol is water soluble, while nitromethane is not. Indeed, mixtures of nitroparaffins and alcohols were demonstrated in many cases to provide synergistic mixtures of two nonsolvents that dissolved polymers [13]. There are three major types of interaction in common organic materials. The most general are the “nonpolar” interactions, which derive from atomic forces. These have also been called dispersion interactions in the literature. Since molecules are built up from atoms, all molecules will contain this type of attractive force. For the saturated aliphatic hydrocarbons, for example, these are essentially the only cohesive interactions, and the energy of vaporization is assumed to be the same as the dispersion cohesive energy, ∆Ed. Finding the dispersion cohesive energy as the cohesion energy of the homomorph, or hydrocarbon counterpart, is the starting point for the calculation of the three Hansen parameters for a given liquid. The permanent dipole-permanent dipole interactions cause a second type of cohesion energy, the polar cohesive energy, ∆Ep. These are inherently molecular interactions and are found in most molecules to one extent or another. The dipole moment is the primary parameter in calculating these interactions. A molecule can be primarily polar in character without being water soluble, so there is misuse of the term “polar” in the general literature. The polar solubility parameters referred to here are well-defined, experimentally verified, and can be estimated from molecular parameters as described below. As noted above, the polar solvents include those with relatively high total solubility parameters, which are not particularly water soluble such as nitroparaffins, propylene carbonate, tri-n-butyl phosphate, and the like. Induced dipoles have not specifically been treated by Hansen, but are recognized as a potentially important factor, particularly for solvents with zero dipole moments. The third major cohesive energy source is hydrogen bonding, ∆Eh. Hydrogen bonding is a molecular interaction and resembles the polar interactions in this respect. The basis of this type of cohesive energy is attraction among molecules because of the hydrogen bonds. In this perhaps oversimplified approach, the hydrogen bonding parameter has been used to more or less collect the energies from interactions not included in the other two parameters. Alcohols, glycols, carboxylic acids, and other hydrophilic materials have high hydrogen bonding parameters. Other researchers have divided this parameter into separate parts, for example, acid and base cohesion parameters, to allow both positive and negative heats of mixing. These approaches will not be dealt with here, but can be found described in Barton’s Handbook [9] and elsewhere [19–21]. The most extensive division of the cohesive energy has been done by Karger et al. [22], who developed a system with five parameters: dispersion, orientation, induction, proton donor, and proton acceptor. The Hansen hydrogenbonding parameter may be termed an electron interchange
15TH EDITION
parameter as well. As a single parameter, it has remarkably accounted well for the experience of the author and keeps the number of parameters to a level, which allows ready practical usage. It is clear that there are other sources of cohesion energy in various types of molecules arising, for example, from induced dipoles, metallic bonds, electrostatic interactions, or whatever type of separate energy can be defined. Hansen stopped with the three major types found in organic molecules. It was and is recognized that additional parameters could be assigned to separate energy types. The description of organometallic compounds could be an intriguing study, for example. This would presumably parallel similar characterizations of surface-active materials, where each separate part of the molecule requires separate characterization for completeness. The Hansen parameters have mainly been used in connection with solubility relations, mostly, but not exclusively, in the coatings industry, but their use is now spreading to other industries. Solubility and swelling have been used to confirm the solubility parameter assignments of many of the liquids. These have then been used to derive group contribution methods and suitable equations based on molecular properties to arrive at estimates of the three parameters for additional liquids. The goal of a prediction is to determine similarity or not of the cohesion energy density parameters. The strength of a particular type of hydrogen bond or other bond, for example, is important only to the extent that it influences the cohesive energy density. Hansen parameters do have direct application in other scientific disciplines of interest to the coatings industry, such as surface science, where they have been used to characterize the wettability of various surfaces and adsorption properties of pigment surfaces [10,14,16,23–25]. Many other applications of widely different character have been discussed by Barton [9] and Gardon [26]. Surface characterizations have not been given the attention deserved in terms of a unified similarity-of-energy approach. The author can certify that thinking in terms of similarity of energy, whether surface energy or cohesive energy, can lead to rapid decisions and plans of action in critical situations where data are lacking. In other words, the everyday industrial crisis situation can often be reduced in scope by appropriate systematic approaches based on similarity of energy. The basic equation that governs the assignment of Hansen parameters is that the total cohesion energy, ∆Et, must be the sum of the individual energies that make it up ΔEt = ΔEd + ΔEp + ΔEh
(4)
Dividing this by the molar volume gives the square of the total (or Hildebrand) solubility parameter as sum of the squares of the Hansen D, P, and H components. ΔEt ΔEd ΔEp ΔEh = + + VM VM VM VM
(5)
t = d2 + p2 + h2 = D2 + P 2 + H 2 (computer printouts ) (6)
METHODS AND PROBLEMS IN THE DETERMINATION OF PARTIAL SOLUBILITY PARAMETERS
The best method to calculate Hansen solubility parameters depends to a great extent on what data are available.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 38
Hansen originally adopted an essentially experimental procedure and established numbers for 90 liquids based on solubility data for 32 polymers [13]. This procedure involved calculation of the nonpolar parameter according to the procedure outlined by Blanks and Prausnitz [12]. This calculational procedure is still in use and is considered to be the most reliable and consistent for this parameter. It is outlined below. The division of the remaining cohesive energy between the polar and hydrogen bonding interactions was done by trial and error to fit experimental polymer solubility data. A key to parameter assignments in this initial trial and error approach was that mixtures of two nonsolvents could be systematically found to synergistically (but predictably) dissolve given polymers. This meant that these had parameters placing them on opposite sides of the solubility region, a spheroid, from each other. Having a large number of such predictably synergistic systems as a basis, reasonably accurate divisions into the three energy types were possible. Using the experimentally established, approximate, ␦p and ␦h parameters, Skaarup [15] found that the Böttcher equation could be used to calculate the polar parameter quite well, and this led to a revision of the earlier values to those now in wide use for these same liquids. These values were also consistent with the experimental solubility data for 32 polymers available at that time and Eq (6). Furthermore, Skaarup developed the equation for the solubility parameter “distance,” Rs, between two materials based on their respective partial solubility parameter components Rs = 4( d1 − d 2 )2 + ( p1 − p2 )2 + ( h1 − h2 )2
(7)
This equation was developed from plots of experimental data where the constant 4 was found convenient and correctly represented the solubility data as a sphere encompassing the good solvents. When the scale for the dispersion parameter is doubled compared with the other two parameters, essentially spherical rather than spheroidal regions of solubility are found. This greatly aids two-dimensional plotting and visualization. There are, of course, boundary regions where deviations can occur. These are most frequently found to involve the larger molecular species being less-effective solvents compared with the smaller counterparts that define the solubility sphere. Likewise, smaller molecular species, such as acetone and methanol, often appear as outliers in that they dissolve a polymer even though they have solubility parameters placing them at a distance greater than the solubility sphere radius, R0. This dependence on molar volume is inherent in the theory developed by Hildebrand and Scatchard as discussed above. Smaller molar volume favors lower heats of mixing, which in turn promotes solubility. Such smaller molecular volume species, which dissolve “better” than predicted by comparisons based on solubility parameters alone, should not necessarily be considered outliers. This statement is justified by Eq (3), where it can be seen that the molar volume and the square of the solubility parameter difference are weighted equally in estimating the heat of mixing in the Hildebrand theory. The molar volume is frequently used as a fourth parameter to describe molecular size effects. These are especially
Q
SOLUBILITY PARAMETERS
473
important in correlating diffusion phenomena with the solubility parameter, for example. The author has preferred to retain the three well-defined, partial solubility parameters with a separate, fourth, molar volume parameter, rather than to multiply the solubility parameters by the molar volume raised to some power to redefine them. In response to a reviewer request stating that molar volumes calculated by software that is generally available do not agree with those found by the author, it can be said that the approach of dividing the molecular weight by the density was the way the molar volume was found when the ideas for this chapter emerged over 40 years ago. The author has seen no reason to diverge from this practice. When densities are not known or cannot be found in standard reference works, recourse can be taken to comparison with related compounds or other suitable procedure for determining the density. There are various types of free volume, which can be a cause of some difference in estimations depending on the assumptions made, but a deeper discussion at this point appears to be beyond the scope of this chapter. The exact reason for the constant 4 in Eq (7) is discussed below. It is currently considered both as an experimental result related to the entropy changes in the systems described and as a theoretically well-defined constant. The author has also found in unpublished studies that values close to 5 could represent solubility data equally well for a few cases studied. The differences in solubility parameters between the solvent and solute in the polar and hydrogen bonding parameters are larger by a factor of two than is tolerated when nonpolar solvents dissolve the same polymer. This factor of 2 is squared to give 4 in Eq (7). The term “specific interactions” is often applied to the molecular polar and molecular hydrogen bonding (electron interchange) interactions implying that these are especially beneficial to achieving a solution since larger differences in the characteristic parameters are allowed with a successful result. Another way to view this is as follows. It is assumed that the (center-of-the-sphere) partial solubility parameters assigned by computer optimization techniques to polymers using Eq (7) are the theoretically correct ones. A solvent with parameters corresponding to the center is to be changed in quality. If the nonpolar parameter difference only is changed by one unit, the effect on Rs is four units. If the polar or hydrogen bonding parameter difference is changed by one unit, the effect on Rs is also one unit. The entropy changes associated with the polar and hydrogen interactions have reduced the total (free energy change) effect by a factor of 4 and are thus four times larger than those associated with the nonpolar interactions. The discussion above follows from the fact that the boundaries of the regions of solubility are characterized by a free energy change of zero for the solution process. The Flory-Huggins limiting chi parameter, χ, of about 0.5 is also characteristic for the boundary of the solubility region. Patterson [6,27], in particular, has been instrumental in showing the relations between the chi parameter and solubility parameters. Patterson’s work led to the developments reported in the next section. This is strictly valid only for the interactions described by this theory. So-called theta solvents will also be located in boundary regions on solubility parameter plots with these same restrictions. Much polymer research has focused on these boundary
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
474
PAINT AND COATING TESTING MANUAL
Q
regions only, for the above reasons and because relatively small changes in temperature, molecular weight, solvent quality, etc., give large easily measurable changes in other quantities. The approach of computer optimizing solubility data to spheres, which is currently in use, still seems most favorable, at least until an improvement is offered by an improved theory. Plotting experimental solubility data defines boundaries of solubility, which in fact are fixed by the free energy of mixing being experimentally equal to zero. Equation (7) is readily used on a computer (or on a hand calculator), and supplementary relations allow easier scanning of large sets of data. It is obvious that solubility, or high affinity, requires that Rs be less than R0. The ratio Rs/R0 has been called the RED number, reflecting the relative energy difference. A RED number of 0 is found for no energy difference. RED numbers less than 1.0 indicate high affinity, RED equal to or close to 1.0 is a boundary condition, and progressively higher RED numbers indicate progressively lower affinities. Scanning a sizeable computer output for RED numbers less than 1.0, for example, rapidly allows location of the most interesting liquids for a given application. The revised set of parameters for the 90 solvents was the basis for group contribution procedures developed by (most notably) Van Krevelen [28] and Beerbower [17,29], who also used Fedors, work [30]. These various developments have been summarized by Barton [9], although Beerbower’s latest values have only appeared in the NASA document [29]. Table 1 is an expanded table of Beerbower group contributions, which was distributed among those who were in contact with Beerbower in the late 1970s. Beerbower also developed a simple equation for the polar parameter [17], which involved only the dipole moment and the square root of the molar volume. This is also given below and has been found quite reliable by Koehnen and Smolders [31]. This equation has also been found reliable by the author as well, giving results generally consistent with Eq (6), which, again, is the basis of the whole approach. Koehnen and Smolders also give correlation coefficients for other calculation procedures to arrive at the individual Hansen parameters. A sizeable number of liquids have now been assigned Hansen parameters using the procedures described here. Many of these have not been published. Exxon Chemical Corporation [32,33] has indicated a computer program with data for over 500 solvents and plasticizers, 450 resins and polymers, and 500 pesticides. The author’s files contain the three parameters for about 1200 liquids, although several of them appear with two or three sets of possible values awaiting experimental confirmation. In some cases, this is due to questionable physical data, for example, for latent heats of vaporization or wide variations in reported dipole moments. Another reason for this is that some liquids are chameleonic [34] as defined by Hoy in that they adopt configurations depending on their environment. Hoy [34] cites the formation of cyclic structures for glycol ethers with (nominally) linear structure. The formation of hydrogen bonded tetramers of alcohols in a fluoropolymer has also been cited [35]. The term “compound formation” can be found in the older literature, particularly where mixtures with water were involved, and structured species were postulated to
15TH EDITION
explain phenomena based on specific interactions among the components of the mixtures. Barton has recently discussed some of these situations and points out that Hildebrand or Hansen parameters must be used with particular caution where the extent of donor-acceptor interactions, and in particular hydrogen bonding within a compound, is very different from that between compounds [18]. Amines and water, for example, are known to associate. Pure component data cannot be expected to predict the behavior in such cases. Still another reason for difficulties is the large variation of dipole moments reported for the same liquid. The dipole moment for some liquids depends on their environment, as discussed below. A given solvent can be listed with different values in files to keep these phenomena in mind. Large data sources greatly enhance searching for similar materials and locating new solvents for a polymer based on limited data, for example. Unfortunately, different authors have used different group contribution techniques, and there is a proliferation of different “Hansen” parameters for the same chemicals in the literature. This would seem to be an unfortunate situation, but may ultimately provide benefits. In particular, partial solubility parameter values found in Hoy’s extensive tables [9,36] are not compatible with the customary Hansen parameters reported here. Hoy has provided an excellent source of total solubility parameters. He independently arrived at the same type division of cohesion energies as Hansen, although the methods of calculation are quite different. Many solvent suppliers have also presented tables of solvent properties and/or use computer techniques to get these in their technical service. Partial solubility parameters not taken directly from earlier well-documented sources should be used with caution. In particular it can be noted that the Hoy dispersion parameter is consistently lower than that found by Hansen. Hoy subtracts estimated values of the polar and hydrogen bonding energies from the total energy to find the dispersion energy. This allows for more calculation error and underestimates the dispersion energy since the Hoy procedure does not appear to fully separate the polar and hydrogen bonding energies. The Van Krevelen dispersion parameters appear too low. The author has not attempted these calculations, being completely dedicated to the full procedure described here, but values estimated independently based on the Van Krevelen dispersion parameters are clearly low. A comparison with related compounds, or similarity principle, gives better results than those found from the Van Krevelen dispersion group contributions. In the following, calculation procedures and experience are presented according to the procedures found most reliable for the experimental and/or physical data available for a given liquid.
RECENT DEVELOPMENTS—THEORY
A new understanding of the Hansen solubility parameter approach has developed since the publishing of the 14th edition of the Gardner-Sward Handbook [37,38]. Chapter 2 in Ref. [38] and also Chapter 2 in Ref. [39] include a comparison of the HSP approach with different theories of polymer solution behavior. These include those developed or modified by Huggins, Flory, Hildebrand, Prigogine, and Patterson [6,27,38–42].
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 38
SOLUBILITY PARAMETERS
Q
475
TABLE 1—Group contributions to partial solubility parameters
Functional Group
Molar Volume, ∆V (cm3/mol)
London parameter, ∆V2D (cal/mol)
Electron transfer parameter, ∆V2H (cal/mol)
Polar parameter, ∆V2P
Total parameter, ∆V2 (cal/mol)
Aliphatic Aromatice Alkane
Cyclo
Aromatic Alkane
Cyclo
Aromatic
Aliphatic
Aromatic
Aliphatic Aromatic
CH–3
33.5
Same
1125
Same
Same
0
0
0
0
0
1,125
Same
CH2<
16.1
Same
1180
Same
Same
0
0
0
0
0
1,180
Same
—CH<
−1.0
Same
820
Same
Same
0
0
0
0
0
820
Same
>C<
−19.2
Same
350
Same
Same
0
0
0
0
0
350
Same
CH2 = olefin
28.5
Same
850±100
?
?
25±10
?
?
180±75
?
1,030
Same
—CH = olefin 13.5
Same
875±100
?
?
18±5
?
?
180±75
?
1,030
Same
>C = olefin
−5.5
Same
800±100
?
?
60±10
?
?
180±75
?
1,030
Same
Phenyl-
. . .
71.4
. . .
. . .
7530
. . .
. . .
50±25
50±50a
. . .
7,630
C-5 ring (saturated)
16
. . .
. . .
250
. . .
0
0
. . .
0
. . .
250
. . .
C-6 ring
16
Same
. . .
250
250
0
0
0
0
0
250
250
—F
18.0
22.0
0
0
0
1,000±150
?
700±100
0
0
1,000
800e
—F2 twid
40.0
48.0
0
0
0
700±250
?
500±250
0
0
1,700
1,360e
—F3 tripletd
66.0
78.0
0
0
0
?
?
?
0
0
1,650
1,315e
—Cl
24.0
28.0
400±100
?
1,300±100
1,250±100
1,450±100
800 ±100
100±20
Same
2,760
2,200e
—Cl2 twind
52.0
60.0
3,650±160
?
3,100±175a 800±150
?
400±150a
165±10a
180±10a
4,600
3,670e
—CL3 tripled
81.9
73.9
4,750±300a ?
?
300±100
?
?
350±250a
?
5,400
4,300e
—Br
30.0
34.0
1,050±300a 1,500±175
1,650±140
1,250±100
1700±150
800 ±100
500±100
500 ±100
3,700
2,960e
—Br2 twind
62.0
70.0
4,300±300a ?
3,500±300a 800±250a
?
400±150a
825±200a
800±250a
5,900
4,700e
—Br3 tripletd
97.2
109.2
5,800±400
?
?
350±150
?
?
—I
31.5
35.5
2,350±250
2 200±250
1,250±100
1,350±100
—I2 twind
66.6
74.6
5,500±300a ?
4,200±300a 800±250a
d — —I3 triplet
111.0
123.0
?
?
?
—O— ether
3.8
Same
0
0
0
500±150
600±150
>CO ketone
10.8
Same
. . .c
2,350±400
2,800±1352 25,000±7%)/V
1,000±300
—CHO
(23.2)
(31.4)
950±300
?
350±275
2,100±200
3,000±500
2,750±200
—COO−ester
18.0
Same
. . .
?
. . .
—COOH
28.5
Same
3,350±300
3,550±250
3,600±400
—OH
10.0
Same
1,770±450
1,370±500
—(OH)2 twin or adjacent
26.0
Same
0
—CN
24.0
—NO2
24.0
—NH2 amine >NH2 amine
d
a
a
a
7,650
6,100e
a
1,000±200
4,550
3,600e
1,650±250a
1,800±200a
8,000
6,400e
?
?
11,700
9,3 50e
450±150
450±25
1,200±100
800
(1,650±150)
950±300
800±250b
400±125a
4,150
Same
1,000±200
750±150
(4,050)
Same
(56,000±12%)/V ?
(338,000±12%)/V 1,250±150
475±100
4,300
Same
500±150
300±50
750±350
2,750±250
2,850±250a
6,600
Same
1,870±600
700±200
1,100±300
800±150
4,650±400
4,650±500
7,120
Same
?
?
1,500±100
?
?
9,000 ±600
9,300±600
10,440
Same
Same
1,600±850a ?
0
4,000±800a
?
3,750±300a
500±200b
400±12a
4,450
Same
32.0
3,000±600
?
2,550±125
3,600±600
?
1750±100
400±50b
350±50a
7,000
(4,400)
19.2
Same
1,050±300
1,050±450
150±150
600 ±200
600±350
800 ±200
1,350±200
2250±200
3,000
Same
4.5
Same
1,150±225
?
?
100±50
?
?
750±200
?
2,000
Same
—NH2 amide
(6.7)
Same
?
?
?
?
?
?
2,700±550a
?
(5,850)
Same
— —PO4
28.0
Same
. . .
?
?
(8100±10%)/V
?
?
3,000±500
?
(7,000)
Same
a
a
a
?
c
c
a
a
2,000±250
a
a
1,500±300
a
?
575+100
1,000±200
a
?
400±150a
?
?
a
a
b
Based on very limited data. Limits shown are roughly 95 % confidence; in many cases, values are for information only and not to be used for computation. b Includes unpublished infrared data. c Use formula in ΔV2p column in calculation, with V for total compound. d Twin and triplet values apply to halogens on the same C atom, except that ΔV2p also includes those on adjacent C atoms. e These values apply to halogens attached directly to the ring and also to halogen attached to aliphatic double-bonded C atoms. f From R. F. Fedors25. a
For those familiar with the widely used Flory interaction coefficient, χ12 it can be shown [38,39] that there is a relation between this and the RED number given by Eq (8). χ12 corresponds to χc (RED)2
(8)
χc is usually taken as being near 1/2 for the interaction of typical solvents with polymers of very large molecular weight. Unfortunately, χ12 does not explicitly take account of hydrogen bonding. Equation (8) is therefore only indicative of a relation had the Flory approach included polar and hydrogen bonding interactions explicitly.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
476
PAINT AND COATING TESTING MANUAL
Q
It is instructive to arrange Eq (8) in the form given in Eq (9): (RED)2 = (Rs/Ro)2 corresponds to χ12/χc
(9)
(Rs)2 corresponds to χ12, both being interaction parameters. These are respectively divided by the corresponding limiting or critical values that allow solubility to find (RED)2. These limiting values are (Ro)2 in the Hansen system and χc in the Huggins-Flory approach. Thus Rs greater than Ro indicates nonsolubility just as values of χ12 greater than χc have the same function. Values of Rs and χ12 that are less than zero are not allowed (even though one finds many measured values of the latter in the literature which are negative). The coefficient 4 in Eq (7) has been found correct for all practical purposes in over 1,000 correlations using Hansen solubility parameters. The coefficient 4 in front of the difference in the nonpolar interactions shows that the specific interactions described by the differences in the polar and hydrogen bonding parameters are only 0.25 times as significant as the differences in the nonpolar term. This factor (0.25) is predicted theoretically by the Prigogine corresponding states theory of polymer solutions for the importance of specific interactions relative to the atomic/ nonpolar interactions [38–42]. This same factor can also be traced back to still earlier approaches (Lorentz-Berthelot mixtures) studying affinities among gases, for example [42]. The Hansen approach and the (first term in the) Prigogine approach are in agreement with each other. The so-called geometric mean to estimate the interaction between two unlike species is used in both of these, and the coefficient 4 experimentally confirms that this type of mean is also valid for hydrogen bonding interactions. The coefficient 4 must be used to differentiate the behavior of the atomic (nonpolar, dispersion, or London) type forces from that of the molecular dipolar and molecular hydrogen bonding forces. The Hansen approach cannot be considered empirical. The Huggins-Flory and subsequent “new Flory” approaches seem inadequate for other than strictly nonpolar systems since hydrogen bonding is not included. The pioneering Hildebrand solubility parameter approach did not treat specific interactions either, although it led the author to develop the Hansen approach that does. The Prigogine approach is too difficult to use and lacks specific consideration of hydrogen bonding as well. This leaves the Hansen approach as the only proven, reliable, and generally useful means to systematically study the common types of interactions involved in systems with hydrogen bonding and with permanent dipoles. It is not in conflict with the other theories. It extends them in a general way to applications with polar and hydrogen bonding. Panayiotou has recently used a statistical thermodynamica approach to calculate the HSP that is reported in Chapter 3 in Ref. [39]. Expanding on earlier work, he first calculates the hydrogen bonding solubility parameter. After this the other HSP and the total solubility parameter are calculated. There is also a new large set of group contributions to help those not fully competent in statistical thermodynamics. The values for all the parameters for about 50 liquids calculated by Panayiotou are in amazingly good agreement with those reported by Hansen over 40 years ago [13–16]. There is also exceptionally good
15TH EDITION
agreement for a number of the common polymers. These results, coupled with the agreement, or perhaps better, lack of disagreement, with the other theories of polymer solutions, appear to confirm that the three-parameter approach based on a division of the cohesion energy is fundamentally sound.
CALCULATION OF THE DISPERSION SOLUBILITY PARAMETER, δd
The ␦d parameter is calculated according to the procedures outlined by Blanks and Prausnitz [12]. Figs. 1 and 2, or 3, can be used to find this parameter depending on whether the molecule of interest is aliphatic, cycloaliphatic, or aromatic. These figures have been inspired by Barton [9], who converted earlier data to SI units. All three of these figures have been straight line extrapolated into a higher range of molar volumes than that reported by Barton. Energies found with these extrapolations have also provided consistent results. The solubility parameters in SI units, MPa1/2, are 2.0455 times larger than those in the older centimeter, gram, second system, (cal/cc)1/2, which still finds extensive use in the United States, for example. The figure for the aliphatic liquids gives the dispersion cohesive energy, ∆Ed, whereas the other two figures directly report the dispersion cohesive energy density, c. The latter is much simpler to use since one needs only take the square root of the value found from the figure to find the respective partial solubility parameter. Barton also presented a similar figure for the aliphatic solvents, but it is inconsistent with the energy figure and in error. Its use is not recommended. When substituted cycloaliphatics or substituted aromatics are considered, simultaneous consideration of the two separate parts of the molecules is required. The dispersion energies are evaluated for each of the types of molecules involved, and a weighted average for the molecule of interest based on numbers of significant atoms is taken. For example, hexyl benzene would be the arithmetic average of the dispersion energies for an aliphatic and an aromatic liquid, each with the given molar volume of hexyl benzene. Liquids such as chlorobenzene, toluene, and ring compounds with alkyl substitutions with only two or three carbon atoms have been considered as cyclic compounds only. Such weighting has been found necessary to satisfy Eq (6). The author has directly used these figures to find the cohesion energy for nonpolar molecules. This introduces a small but unaccountable error since, for example, ∆ED , for an aromatic compound such as toluene, is not equal to the total cohesion energy ∆Et. In practice (so far) this has not led to any apparent problems. The error becomes less significant as the polar and hydrogen bonding effects increase. The critical temperature, Tc, is required to use the dispersion energy figures. If the critical temperature cannot be found, it must be estimated. A table of the Lydersen group contributions, ∆T, [43] as given by Hoy [36] for calculation of the critical temperature, is included here as Table 2. In some cases, the desired groups may not be in the table, which means some educated guessing is required. The end result does not appear too sensitive to these situations. The normal boiling temperature, Tb, is also required in this calculation. This is not always available, either, and must be estimated by similarity, group contribution, or
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 38
Q
SOLUBILITY PARAMETERS
477
Fig. 1—Energy of vaporization for straight chain hydrocarbons as a function of molar volume and reduced temperature.
other technique. The Lydersen group contribution method involves the use of Eqs (10) and (11), Tb / Tc = 0.567 + ∑ Δ T − ( ∑ Δ T )
Tr = T / Tc
2
(10) (11)
where T has been taken as 298.15 K. The dispersion parameter is an atomic force parameter. The size of the atom is important. It has been found that corrections are required for atoms significantly larger than carbon, such as chlorine, sulfur, bromine, etc., but not for
Fig. 2—Cohesive energy density for cycloalkanes as a function of molar volume and reduced temperature.
oxygen or nitrogen, which have a similar size. The carbon atom in hydrocarbons is the basis of the dispersion parameter in its present form. These corrections are applied by first finding the dispersion cohesive energy from the appropriate figure. This requires multiplication by the molar volume for the cyclic compounds using data from the figures here since these figures give the cohesive energy densities. The dispersion cohesive energy is then increased by adding on the correction factor. This correction factor for chlorine, bromine, and sulfur has been taken as 1,650 J/mol for each of these atoms in the molecule. Dividing by the molar volume and then taking the square root gives the (large atom corrected) dispersion solubility parameter. The need for these corrections has been confirmed many times, both for interpretation of experimental data
Fig. 3—Cohesive energy density for aromatic hydrocarbons as a function of molar volume and reduced temperature.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
478
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
TABLE 2—Lydersen group constants Group
Aliphatic, ∆T
Cyclic, ∆T
∆TP
Aliphatic, ∆P
Cyclic, ∆P
CH3
0.020
. . .
0.0226
0.227
. . .
CH2
0.020
0.013
0.0200
0.227
0.184
>CH—
0.012
0.012
0.0131
0.210
0.192
>C<
0.000
−0.007
0.210
0.210
0.154
—CH2
0.018
. . .
0.0192
0.198
. . .
—CH—
0.018
0.011
0.0184
0.198
0.154
—C<
0.000
0.011
0.0129
0.198
0.154
—CH aromatic
. . .
. . .
0.0178
. . .
. . .
—CH aromatic
. . .
. . .
0.0149
. . .
. . .
—O—
0.021
0.014
0.0175
0.16
0.12
>O epoxide
. . .
. . .
0.0267
. . .
. . .
—COO—
0.047
. . .
0.0497
0.47
. . .
>C— —O
0.040
0.033
0.0400
0.29
0.02
—CHO
0.048
. . .
0.0445
0.33
. . .
—CO2O
. . .
. . .
0.0863
. . .
. . .
—OH→
. . .
. . .
0.0343
0.06
. . .
—H→
. . .
. . .
−0.0077
. . .
. . .
—OH primary
0.082
. . .
0.0493
. . .
. . .
—OH sec.
. . .
. . .
0.0440
. . .
. . .
—OH tert.
. . .
. . .
0.0593
. . .
. . .
—OH phenolic
0.035
. . .
0.0060
−0.02
—NH2
0.031
. . .
0.0345
0.095
. . .
—NH—
0.031
0.024
0.0274
0.135
0.09
>N—
0.014
0.007
0.0093
0.17
0.13
—C — — —N
0.060
. . .
0.0539
0.36
. . .
—NCO
. . .
. . .
0.0539
. . .
. . .
HCON<
. . .
. . .
0.0546
. . .
. . .
—CONH—
. . .
. . .
0.0843
. . .
. . .
—CON<
. . .
. . .
0.0729
. . .
. . .
—CONH2
. . .
. . .
0.0897
. . .
. . .
—OCONH—
. . .
. . .
0.0938
. . .
. . .
—S—
0.015
0.008
0.0318
0.27
0.24
—SH
0.015
. . .
. . .
. . .
. . .
—CI 1°
0.017
. . .
0.0311
0.320
. . .
—CI 2°
. . .
. . .
0.0317
. . .
. . .
Cl2 twin
. . .
. . .
0.0521
. . .
. . .
CI aromatic
. . .
. . .
0.0245
. . .
. . .
—Br
0.010
. . .
0.0392
0.50
. . .
—Br aromatic
. . .
. . .
0.0313
. . .
. . .
—F
0.018
. . .
0.006
0.224
. . .
—I
0.012
. . .
. . .
0.83
. . .
Conjugation
. . .
. . .
0.0035
. . .
. . .
c is double bond
. . .
. . .
−0.0010
. . .
. . .
trans double bond
. . .
. . .
−0.0020
. . .
. . .
4 Member ring
. . .
. . .
0.0118
. . .
. . .
5 Member ring
. . .
. . .
0.003
. . .
. . .
6 Member ring
. . .
. . .
−0.0035
. . .
. . .
7 Member ring
. . .
. . .
0.0069
. . .
. . .
Ortho
. . .
. . .
0.0015
. . .
. . .
Meta
. . .
. . .
0.0010
. . .
. . .
Para
. . .
. . .
0.0060
. . .
. . .
Bicycloheptyl
. . .
. . .
0.0034
. . .
. . .
Tricyclodecane
. . .
. . .
0.0095
. . .
. . .
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 38
and to allow Eq (6) to balance. Research is definitely needed in this area. The impact of these corrections is, of course, larger for the smaller molecular species. The taking of square roots of the larger numbers involved with the larger molecular species reduces the errors involved in these cases since the corrections themselves are relatively small. It can be seen from the dispersion parameters of the cyclic compounds that the ring also has an effect similar to increasing the effective size of the interacting species. The dispersion energies are larger for cycloaliphatic compounds than for their aliphatic counterparts, and they are higher for aromatic compounds than for the corresponding cycloaliphatics. Similar effects also appear with the ester group. This group appears to act as if it were, in effect, an entity that is larger than the corresponding compound containing only carbon (i.e., its homomorph), and it has a higher dispersion solubility parameter without any special need for corrections. The careful evaluation of the dispersion cohesive energy may not have major impact on the value of the dispersion solubility parameter itself because of the taking of square roots. Larger problems arise because of Eq (4). Energy assigned to the dispersion portion cannot be reused when finding the other partial parameters using Eq (4) [or Eq (6)]. This is one reason why group contributions are recommended for the polar and hydrogen bonding components in some cases below. The author has not used older tables of group contributions for the dispersion parameter. These were found to be not as reliable as the longer procedure discussed above when data are available. Frequent use has been made of simple comparisons of similarity to molecules whose dispersion parameters have already been established. The group contributions for the dispersion solubility parameter presented by Panayiotou in Chapter 3 in Ref. [39] appear to be a good third alternative, however.
CALCULATION OF THE POLAR SOLUBILITY PARAMETER, δp
The earliest assignments of a polar solubility parameter were given by Blanks and Prausnitz [12]. These parameters were in fact the combined polar and hydrogen bonding parameters as used by Hansen and cannot be considered polar in the current context. The first Hansen polar parameters [13] were reassigned new values by Skaarup according to the Böttcher equation [15]. This equation requires the molar volume, the dipole moment (DM), the refractive index, and the dielectric constant. These are not available for many compounds, and the calculation is somewhat more difficult than using the much simpler equation developed by Beerbower [17]
p =
37.4 (DM ) (VM )1/ 2
(12)
The constant 37.4 gives this parameter in SI units. Equation (12) has been consistently used by the author over the past few years, particularly in view of its reported reliability [31]. This reported reliability appears to be correct. The molar volume must be known or estimated in one way or another. This leaves only the dipole moment to be found or estimated. Standard reference works have tables of dipole moments, with the most extensive listing still being McClellan [44]. Other data sources also have this
Q
SOLUBILITY PARAMETERS
479
parameter as well as other relevant parameters and data such as latent heats and critical temperatures. The so-called DIPPR database has been found useful for many compounds of reasonably common usage, but many interesting compounds are not included in DIPPR. This abbreviation is for Design Institute for Physical Property Research,2 Project 801 of the American Institute of Chemical Engineers at the Pennsylvania State University [45]. When no dipole moment is available, similarity with other compounds, group contributions, or experimental data can be used to estimate the polar solubility parameter. It must be noted that the fact of zero dipole moment in symmetrical molecules is not basis enough to assign a zero polar solubility parameter. An outstanding example of variations of this kind can be found with carbon disulfide. The reported dipole moments are mostly 0 for gas phase measurements, supplemented by 0.08 in hexane, 0.4 in carbon tetrachloride, 0.49 in chlorobenzene, and 1.21 in nitrobenzene. There is a clear increase with increasing solubility parameter of the media. The latter and highest value has been found experimentally most fitting for correlating permeation through a fluoropolymer film used for chemical protective clothing [46]. Many fluoropolymers have considerable polarity. The lower dipole moments seem to fit in other instances. Diethylether has also presented problems as an outlier in terms of dissolving or not, or rapid permeation or not. Here the reported dipole moments [44] vary from 0.74 to 2.0 with a preferred value of 1.17, and with 1.79 in chloroform. Choosing a given value seems rather arbitrary. The chameleonic cyclic forms of the linear glycol ethers would also seem to provide for a basis of altered dipole moments in various media [34]. When Eq (12) cannot be used, the polar solubility parameter has been found using the Beerbower table of group contributions, by similarity to related compounds, and/or by subtraction of the dispersion and hydrogen bonding cohesive energies from the total cohesive energy. The question in each case is: “Which data are available and judged most reliable?” New group contributions have also been developed from related compounds where their dipole moments are available. These new polar group contributions then become supplementary to the Beerbower table. For large molecules, especially those with long hydrocarbon chains, the accurate calculation of the relatively small polar (and hydrogen bonding) contributions present special difficulties. The latent heats are not generally available with sufficient accuracy to allow subtraction of two large numbers from each other to find a very small one. In such cases the similarity and group contribution methods are thought best. Unfortunately, latent heats found in a widely used handbook [47] are not clearly reported as to the reference temperature. There is an indication that these are 25°C data, but checking indicated many of the data were identical with boiling point data reported elsewhere in the literature. A more recent edition of this handbook has a completely different and less voluminous section for the latent heat of evaporation [48]. Again, even moderate variations in reported heats of vaporization can cause severe Design Institute for Physical Property Research, Department of Chemical Engineering, 167 Fenske Laboratory, The Pennsylvania State University, University Park, PA 16802.
2
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
480
PAINT AND COATING TESTING MANUAL
Q
problems in calculating the polar (or hydrogen bonding) parameter when Eqs (4) or (6) are strictly adhered to.
CALCULATION OF THE HYDROGEN BONDING SOLUBILITY PARAMETER, δh
In the earliest work, the hydrogen bonding parameter was almost always found from the subtraction of the polar and dispersion energies of vaporization from the total energy of vaporization. This is still widely used where the required data are available and reliable. At this stage, however, the group contribution techniques are considered reasonably reliable for most of the required calculations and, in fact, more reliable than estimating several of the other parameters to ultimately arrive at the subtraction step just mentioned. Therefore, in the absence of reliable latent heat and dipole moment data, group contributions are judged to be the best alternative. Similarity to related compounds can also be used, of course, and the result of such a procedure should be essentially the same as for using group contributions. The work of Panayiotou in Chapter 3 of Ref. [39], discussed above, also provides an alternative method to arrive at the hydrogen bonding parameter.
SUPPLEMENTARY CALCULATIONS AND PROCEDURES
The procedures listed above are those most frequently used by the author in calculating the three partial solubility parameters for liquids where some data are available. There are a number of other calculations and procedures that are also helpful. Latent heat data at 25°C have consistently been found from latent heats at another temperature using the relation given by Fishtine [49]. ΔHv (T1 ) ⎡ Tc − T1 ⎤ =⎢ ⎥ ΔHv (T2 ) ⎣ Tc − T2 ⎦
0.38
⎡ 1 − Tr1 ⎤ =⎢ ⎥ ⎣1 − Tr 2 ⎦
0.38
⎡ 1 − Tr1 ⎤ =⎢ ⎥ ⎣1 − Tr 2 ⎦
0.38
(13)
This is done even if the melting point of the compounding being considered is higher than 25°C. The result is consistent with all the other parameters, and to date no problems with particularly faulty predictions have been noted in this respect, i.e., it appears as if the predictions are not significantly in error. When the latent heat is given in cal/mole, the above equation is used to estimate the latent heat at 25°C. RT equal to 592 cal/mole is then subtracted from this according to Eq (14) to find the total cohesion energy, ∆EV , in cgs units at this temperature: ΔEv = ΔEt = ΔHv − RT
(14)
Only limited attempts have been made to calculate solubility parameters at a higher temperature. Solubility parameter correlations of phenomena at higher temperatures have generally been found satisfactory when the established 25°C parameters have been used. Recalculation to higher temperatures is possible, but has not been found necessary. In this direct but approximate approach, it is assumed that the parameters all demonstrate the same temperature dependence, which, of course, is not the case. It might be noted in this connection that the hydrogen bonding parameter, in particular, is the most sensitive to temperature. As the temperature is increased, more and more hydrogen bonds are progressively broken, and this parameter will decrease more rapidly than the others.
15TH EDITION
The gas phase dipole moment is not temperature dependent, although the volume of the fluid does change with temperature, which will change its cohesive energy density. Beerbower has suggested relations to predict the changes of the partial solubility parameters with temperature [17]. The coefficient of thermal expansion, α, appears in all of these relations. These are d( d ) = −1.25 d dT
(15)
d( p) = −0.5 p dT
(16)
d( p) = −H (1.22 × 10 −3 + 0.5 ) dT
(17)
A computer program has been developed by the author to assign the three Hansen parameters for solvents based on experimental data alone. This has been used in several cases where the parameters for the given liquids were desired with a high degree of accuracy. The procedure is to enter solvent quality, good or bad, into the program for a reasonably large number of polymers where the solubility parameters and appropriate radius of interaction for the polymers are known. The program then locates that set of ␦d, ␦p, and ␦h parameters for the solvent that best satisfies the requirements of a location within the spheres of the appropriate polymers where solvent quality is good and outside of the appropriate spheres where it is bad. An additional aid in estimating the Hansen parameters for many compounds is that these parameters can be found by interpolation or extrapolation, especially for homologous series. The first member may not necessarily be a straight line extrapolation, but comparisons with related compounds should always be made where possible to confirm assignments. Plotting the parameters reported in Table 3 for homologous series among the esters, nitroparaffins, ketones, alcohols, and glycol ethers will aid in finding the parameters for related compounds. Table 3 contains Hansen solubility parameters for a large number of liquids and plasticizers. These are given in SI units. Over 800 HSP values can be found in Refs. [38,39].
SOLUBILITY PARAMETERS FOR POLYMERS
The solubility parameters for numerous polymers and film formers are given in Table 4. Suppliers and trademarks are given in Table 5. These data are based on solubility determinations unless otherwise noted. There are four parameters, the three describing the nonpolar, polar, and hydrogen bonding interactions as for the liquids, and the fourth, R0, a radius of interaction for the type of interaction described. Most of these are taken from a report [50] from the Scandinavian Paint and Printing Ink Research Institute. (This institute unfortunately no longer exists.) Additional values have been contributed according to the notes in the table to indicate the types of data, which have been correlated with these techniques. Barton [51] has also provided solubility parameters for many polymers in a handbook. Additional values are also found in Refs. [38,39]. Experimental determination of polymer solubility parameters involves trying to dissolve the polymer at a
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 38
Q
SOLUBILITY PARAMETERS
481
TABLE 3—Hansen solubility parameters for selected liquids (solvents in alphabetical order) No.
Solvent
Dispersion
Polar
Hydrogen Bonding
Molar Volume
1
Acetaldehyde
14.7
12.5
7.9
56.6
2
Acetic acid
14.5
8.0
13.5
57.1
3
Acetic anhydride
16.0
11.7
10.2
94.5
4
Acetonea
15.5
10.4
7.0
74.0
5
Acetonitrile
15.3
18.0
6.1
52.6
6
Acetophenonea
19.6
8.6
3.7
117.4
7
Acrylonitrile
16.0
12.8
6.8
67.1
8
Allyl alcohol
16.2
10.8
16.8
68.4
9
Amyl acetate
15.8
3.3
6.1
148.0
10
Aniline
19.4
5.1
10.2
91.5
11
Anisole
17.8
4.1
6.7
119.1
12
Benzaldehyde
19.4
7.4
5.3
101.5
13
Benzenea
18.4
0.0
2.0
89.4
14
1.3-Benzenediol
18.0
8.4
21.0
87.5
15
Benzoic acid
18.2
6.9
9.8
113.1
16
Benzonitrile
17.4
9.0
3.3
102.6
17
Benzyl alcohol
18.4
6.3
13.7
103.6
18
Benzyl butyl phthalate
19.0
11.2
3.1
306.0
19
Benzyl chloride
18.8
7.1
2.6
115.0
20
Biphenyl
19.7
1.0
2.0
155.1
21
Bromobenzene
20.5
5.5
4.1
105.3
22
Bromochloromethane
17.3
5.7
3.5
65.0
23
Bromoform
21.4
4.1
6.1
87.5
24
1-Bromonaphtalene
20.3
3.1
4.1
140.0
25
Bromotrifluoromethane
9.6
2.4
0.0
97.0
26
Butane
14.1
0.0
0.0
101.4
27
1.3-Butanediol
16.6
10.0
21.5
89.9
28
l-Butanola
16.0
5.7
15.8
91.5
29
2-Butanol
15.8
5.7
14.5
92.0
30
Butyl acetatea
15.8
3.7
6.3
132.5
31
Sec-butyl acetate
15.0
3.7
7.6
133.6
32
Butyl acrylate
15.6
6.2
4.9
143.8
33
Butylamine
16.2
4.5
8.0
99.0
34
Butyl lactate
15.8
6.5
10.2
149.0
35
Butyraldehyde
15.6
10.1
6.2
90.5
36
Butyric acid
14.9
4.1
10.6
110.0
37
Gamma butyrolactonea
19.0
16.6
7.4
76.8
38
Butyronitrile
15.3
12.4
5.1
87.3
39
Carbon disulfide
20.5
0.0
0.6
60.0
40
Carbon tetrachloridea
17.8
0.0
0.6
97.1
41
Chlorobenzenea
19.0
4.3
2.0
102.1
42
1-Chlorobutane
16.2
5.5
2.0
104.5
43
Chlorodifluoromethane
12.3
6.3
5.7
72.9
44
Chloroforma
17.8
3.1
5.7
80.7
45
3-Chloro-l-propanol
17.5
5.7
14.7
84.2
46
m-cresol
18.0
5.1
12.9
104.7
47
Cyclohexane
16.8
0.0
0.2
108.7
48
Cyclohexanola
17.4
4.1
13.5
106.0
49
Cyclohexanone
17.8
8.4
5.1
104.0
50
Cyclohexylamine
17.2
3.1
6.5
113.8
51 Cyclohexylchloride Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT17.3 2014
5.5
2.0
118.6
Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
482
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
TABLE 3—(Continued) No.
Solvent
Dispersion
Polar
Hydrogen Bonding
Molar Volume
52
Cis-decahydronaphthalene
18.8
0.0
0.0
156.9
53
Trans-decahydronaphthalene
18.0
0.0
0.0
156.9
54
Decane
15.7
0.0
0.0
195.9
55
1-Decanol
16.0
4.7
10.0
191.8
56
Diacetone alcohola
15.8
8.2
10.8
124.2
57
Dibenzyl ether
19.6
3.4
5.2
197.4
58
Dibutyl phthalate
17.8
8.6
4.1
266.0
59
Dibutyl sebacate
13.9
4.5
4.1
339.0
60
o-dichlorobenzenea
19.2
6.3
3.3
112.8
61
2.2-Dichlorodiethyl ether
18.8
9.0
5.7
117.2
62
Dichlorodifluoromethane
12.3
2.0
0.0
92.3
63
1.1-Dichloroethane
16.5
7.8
3.0
84.2
64
1.1-Dichloroethylene
16.4
52
24
79.9
65
Di-(2-chloro-isopropyl) ether
19.0
8.2
5.1
146.0
66
Dichloromonofluoromethane
15.8
3.1
5.7
75.4
67
1.2-Dichlorotetrafluoroethane
12.6
1.8
0.0
117.6
68
Di-iso-butyl carbinol
14.9
3.1
10.8
177.8
69
Diethanolamine
17.2
10.8
21.2
95.9
70
Diethylamine
14.9
2.3
6.1
103.2
71
2-(Diethylamino) ethanol
14.9
5.8
12.0
133.2
72
Para-diethy l benzene
18.0
0.0
0.6
156.9
73
Diethyl carbonate
15.1
6.3
3.5
121.0
74
Diethylene glycol
16.6
12.0
20.7
94.9
75
Diethylene glycol butyl ether acetate
16.0
4.1
8.2
208.2
76
Diethylene glycol hexyl ether
16.0
6.0
10.0
204.3
77
Diethylene glycol monobutyl ether
16.0
7.0
10.6
170.6
78
Diethylene glycol monoethyl ether
16.1
9.2
12.2
130.9
79
Diethylene glycol monomethyl ether
16.2
7.8
12.6
118.0
80
Diethylenetriamine
16.7
13.3
14.3
108.0
81
Diethyl ethera
14.5
2.9
5.1
104.8
82
Diethyl ketone
15.8
7.6
4.7
106,4
83
Diethyl phthalate
17.6
9.6
4.5
198.0
84
Diethyl sulfate
15.7
14.7
7.1
131.5
85
Diethyl sulfide
16.8
3.1
2.0
107.4
86
Di(isobutyl) ketone
16.0
3.7
4.1
177.1
87
Di(2-methoxyethyl) ether
15.7
6.1
6.5
142.0
88
N,N-dimethylacetamide
16.8
11.5
10.2
92.5
89
Dimethylformamidea
17.4
13.7
11.3
77.0
90
1,1-Di methyl hydrazine
15.3
5.9
11.0
76.0
91
Dimethyl phthalate
18.6
10.8
4.9
163.0
92
Dimethyl sulfone
19.0
19.4
12.3
75.0
93
Dimethyl sulfoxidea
18.4
16.4
10.2
71.3
94
Dioctyl phthalate
16.6
7.0
3.1
377.0
95
1,4-Dioxanea
19.0
1.8
7.4
85.7
96
Dipropylamine
15.3
1.4
4.1
136.9
97
Dipropylene glycola
16.5
10.6
17.7
130.9
98
Dipropyiene glycol methyl ether
15.5
5.7
11.2
157.4,
99
Dodecane
16.0
0.0
0.0
228.6
100
Eicosane
16.5
0.0
0.0
359.8
101
Epichlorohydrin
18.9
7.6
6.6
78.4
102 Ethanethiol Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT15.7 2014
6.5
7.1
74.3
a
Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 38
Q
SOLUBILITY PARAMETERS
483
TABLE 3—(Continued) No.
Solvent
Dispersion
Polar
Hydrogen Bonding
Molar Volume
103
Ethanola
15.8
8.8
19.4
58.5
104
Ethanolaminea
17.0
15.5
21.2
59.8
105
Ethyl acetate
a
15.8
5.3
7.2
98.5
106
Ethyl acrylate
15.5
7.1
5.5
108.8
107
Ediyl amyl ketone
16.2
4.5
4.1
156.0
108
Ethylbenzene
17.8
0.6
1.4
123.1
109
Ethyl bromide
16.5
8.4
2.3
74.6
110
2-Ethyl-l-butanol
15.8
4.3
13.5
123.2
111
Ethyl butyl ketone
16.2
5.0
4.1
139.0
112
Ethyl chloride
15.7
6.1
2.9
70.0
113
Ethyl chloroformate
15.5
10.0
6.7
95.6
114
Ethyl cinnamate
18.4
8.2
4.1
166.8
115
Ethylene carbonate
19.4
21.7
5.1
66.0
116
Ethylene cyanohydrin
17.2
18.8
17.6
68.3
117
Ethylenediamine
16.6
8.8
17.0
67.3
118
Ethylene dibromide
19.2
3.5
8.6
87.0
119
Ethylene dichloridea
19.0
7.4
4.1
79.4
120
Ethylene glycola
17.0
11.0
26.0
55.8
121
Ethylene glycol butyl ether acetate
15.3
4.5
8.8
171.2
122
Ethylene glycol monobutyl ethera
16.0
5.1
12.3
131.6
123
Ethylene glycol monoethyl ethera
16.2
9.2
14.3
97.8
124
Ethylene glycol monoethyl ether acetate
15.9
4.7
10.6
136.1
125
Ethylene glycol monomethyl ethera
16.2
9.2
16.4
79.1
126
Ethylene glycol monomethyl ether acetate
15.9
5.5
11.6
121.6
127
Ethyl formate
15.5
8.4
8.4
80.2
128
2-Ethyl hexanol
15.9
3.3
11.8
156.6
129
Ethyl lactate
16.0
7.6
12.5
115.0
130
Formamidea
17.2
26.2
19.0
39.8
131
Formic acid
14.3
11.9
16.6
37.8
132
Furan
17.8
1.8
5.3
72.5
133
Furfural
18.6
14.9
5.1
83.2
134
Furfuryl alcohol
17.4
7.6
15.1
86.5
135
Glycerol
17.4
12.1
29.3
73.3
136
Heptane
15.3
0.0
0.0
147.4
137
Hexadecane
16.3
0.0
0.0
294.1
138
Hexamethylphosphoramide
18.5
8.6
11.3
175.7
139
Hexanea
14.9
0.0
0.0
131.6
140
Hexylene glycol
15.7
8.4
17.8
123.0
141
Isoamyl acetate
15.3
3.1
7.0
148.8
142
Isobutyl acetate
15.1
3.7
6.3
133.5
143
Isobutyl alcohol
15.1
5.7
15.9
92.8
144
Isobutyl isobutyrate
15.1
2.9
5.9
163.0
145
Isooctyl alcohol
14.4
7.3
12.9
156.6
146
Isopentane
13.7
0.0
0.0
117.4
147
Isophoronea
16.6
8.2
7.4
150.5
148
Isopropyl palmitate
14.3
3.9
3.7
330.0
149
Mesitylene
18.0
0.0
0.6
139.8
150
Mesityl oxide
16.4
6.1
6.1
115.6
151
Methacrylonitrile
15.8
15.1
5.4
83.9
152
Methanol
15.1
12.3
22.3
40.7
8.2
13.3
109.5
a
153 o-Methoxyphenol Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT18.0 2014
Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
Continued
484
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
TABLE 3—(Continued) No.
Solvent
Dispersion
Polar
Hydrogen Bonding
Molar Volume
154
Methyl acetate
15.5
7.2
7.6
79.7
155
Methyl acrylate
15.3
6.7
9.4
90.3
156
Methylal
15.0
1.8
8.6
169.4
157
Methyl amyl acetate
15.2
3.1
6.8
167.4
158
Methyl butyl ketone
15.3
6.1
4.1
123.6
159
Methyl chloride
15.3
6.1
3.9
55.4
160
Methylcyclohexane
16.0
0.0
1.0
128.3
161
Methylene dichloridea
18.2
6.3
6.1
63.9
162
Methylene diiodide
17.8
3.9
5.5
80.5
163
Methyl ethyl ketone
16.0
9.0
5.1
90.1
164
Methyl isoamyl ketone
16.0
5.7
4.1
142.8
165
Methyl isobutyl carbinol
15.4
3.3
12.3
127.2
166
Methyl isobutyl ketonea
15.3
6.1
4.1
125.8
167
Methyl methacrylate
15.8
6.5
54
106.1
168
1-Methylnaphthalene
20.6
0.8
4.7
138.8
169
Methyl oleate
14.5
3.9
3.7
340.0
170
2-Methyl-1-propanol
15.1
5.7
15.9
92.8
171
Methyl-2-pyrrolidonea
18.0
12.3
7.2
96.5
172
Methyl salicylate
18.1
8.0
13.9
129.6
173
Morpholine
18.8
4.9
9.2
87.1
174
Naphtha.high-flash
17.9
0.7
1.8
181.8
175
Naphthalene
19.2
2.0
5.9
111.5
176
Nitrobenzene
20.0
8.6
4.1
102.7
177
Nitroethanea
16.0
15.5
4.5
71.5
178
Nitromethanea
15.8
18.8
5.1
54.3
179
1-Nitropropane
16.6
12.3
5.5
88.4
a
180
2-Nitropropane
16.2
12.1
4.1
86.9
181
Nonane
15.7
0.0
0.0
179.7
182
Nonyl phenol
16.5
4.1
9.2
231.0
183
Nonyl phenoxy ethanol
16.7
10.2
8.4
275.0
184
Octane
15.5
0.0
0.0
163.5
185
Octanoic acid
15.1
3.3
8.2
159.0
186
1-Octanol
16.0
5.0
11.9
157.7
187
2-Octanol
16.1
4.9
11.0
159.1
188
Oleic acid
16.0
28
6.2
317.0
189
Oleyl alcohol
16.0
2.6
8.0
316.0
190
Pentane
14.5
0.0
0.0
116.2
191
2.4-Pentanedione
17.1
9.0
4.1
103.1
192
1-Pentanol
15.9
5.9
13.9
108.6
193
Perfluoro(dimethylcyclohexane)
12.4
0.0
0.0
217.4
194
Perfluoroheptane
12.0
0.0
0.0
227.3
195
Perfluoromethylcyclohexane
12.4
0.0
0.0
196.0
196
Phenol
18.0
5.9
14.9
87.5
197
Bis-(m-phenoxyphenyl) ether
19.6
3.1
5.1
373.0
198
1-Propanol
16.0
6.8
17.4
75.2
199
2-Propanol
15.8
6.1
16.4
76.8
200
Propionitrile
15.3
14.3
5.5
70.9
201
Propylamine
16.9
4.9
8.6
83.0
202
Propyl chloride
16.0
7.8
2.0
88.1
203
Propylene carbonatea
20.0
18.0
4.1
85.0
9.4
23.3
73.6
a
a 204 Propylene glycol Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT16.8 2014
Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 38
Q
SOLUBILITY PARAMETERS
485
TABLE 3—(Continued) No.
Solvent
Dispersion
Polar
Hydrogen Bonding
Molar Volume
205
Propylene glycol monobutyl ether
15.3
4.5
9.2
132.0
206
Propylene glycol monoethyl ether
15.7
6.5
10.5
115,6
207
Propylene glycol monoisobutyl ether
15.1
4.7
9.8
132.2
208
Propylene glycol monomethyl ether
15.6
6.3
11.6
93.8
209
Propylene glycol monophenyl ether
17.4
5.3
11.5
143.2
210
Propylene glycol monopropyl ether
15.8
7.0
9.2
130.3
211
Pyridine
19.0
8.8
5.9
80.9
212
2-Pyrolidone
19.4
17.4
11.3
76.4
213
Quinoline
19.4
5.6
5.7
118.0
214
Stearic acid
16.3
3.3
5.5
326.0
215
Styrene
18.6
1.0
4.1
115.6
216
Succinic anhydride
18.6
19.2
16.6
66.8
217
1.1.2.2-Tetrabromoethane
22.6
5.1
8.2
116.8
218
1.1.2.2-Tetrachloroethane
18.8
5.1
9.4
105.2
219
Tetrachloroethylene
18.3
5.7
0.0
101.2
220
Tetraethylorthosilicate
13.9
4.3
0.6
224.0
221
Tetrahydrofurana
16.8
5.7
8.0
81.7
222
Tetrahydronaphthalene
19.6
2.0
2.9
136.0
223
Tetramethylurea
16.7
8.2
11.0
120.4
224
Toluenea
18.0
1.4
2.0
106.8
225
Tributyl phosphate
16.3
6.3
4.3
345.0
226
Trichlorobiphenyl
19.2
5.3
4.1
187.0
227
1.1.1-Trichloroethane
16.8
4.3
2.0
99.3
228
Trichloroethylenea
18.0
3.1
5.3
90.2
229
Trichlorofluoromethane
15.3
2.0
0.0
92.8
230
1.1.2-Trichlorotrifluoroethane
14.7
1.6
0.0
119.2
231
Tricresyl phosphate
19.0
12.3
4.5
316.0
232
Tridecyl alcohol
16.2
3.1
9.0
242.0
233
Triethanolamine
17.3
22.4
23.3
133.2
234
Triethylamine
17.8
0.4
1.0
138.6
235
Triethyleneglycol
16.0
12.5
18.6
114.0
236
Triethylene glycol monooleyl ether
16.0
3.1
8.4
418.5
237
Triethylphosphate
16.7
11.4
9.2
171.0
238
Trifluoroacetic acid
15.6
9.9
11.6
74.2
239
Trimethylbenzene
18.0
1.0
1.0
137.3
240
2.2.2,4-Trimethylpentane
14.1
0.0
0.0
166.1
241
2.2.4-Trimethyl-1.3-pentanediol M.I. butyral
15.1
6.1
9.8
227.4
242
Trimethyl phosphate
16.7
15.9
10.2
115.8
243
Water
15.5
16.0
42.3
18.0
244
Xylene
17.6
1.0
3.1
123.3
245
o-xylene
17.8
1.0
3.1
121.2
a
Indicates use in author’s standard set of test solvents.
given concentration, usually 10 % by weight, in a selection of solvents intended to maximize information regarding all types of interaction. Whenever possible, the author uses a set of parameters indicated with an “a” in Table 3. The “yes” or “no” solubility data can be plotted by hand or processed by computer to yield a “spherical” characterization as described above. Teas [52] has developed a triangular plotting technique, which helps visualization of three param-
eters on a plain sheet of paper. Examples are found in Refs. [9,10] as well. Swelling, weight gain, solvent resistance, and surface attack have also been used as a primary data to characterize polymers. The weighted-averaging of the respective Hansen solubility parameters for the test liquids should be done with caution. Many polymers have polar and hydrogen bonding parameters higher than any liquids, for which
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
486
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
TABLE 4—Solubility parameters for polymers D
P
H
TABLE 4—(Continued)
R
D
16.6
Cell it BP-300
12.0
6.7
10.2
18.2
Hycar 1052a
18.2
R
8.6
4.1
9.4
2.2
3.3
6.4
1.4
−0.8
9.4
Polybutadiene
Cellulose acetate Cellidora Aa
H
Acrylonitrile-butadiene elastomer
Cellulose acetobutyrate a
P
12.4
10.8
7.4
Buna Hüls B-10a
17.1 Polyisoprene
Cethyl cellulose Ethocel HE 10 indb
17.9
4.3
3.9
5.9
Ethocel Std 20 indb
20.1
6.9
5.9
9.9
Epoxy
Cariflex IR 305a
16.2 Polyisobutylene
Lutonal IC/1203
14.2
2.5
4.6
12.4
Lutonal I60
16.9
2.5
4.0
7.2
a
Araldite DY 025
14.0
7.4
9.4
13.7
Polyvinylbutyl ether
17.4
4.3
8.4
7.4
Epikote 828
21.3
14.2
6.1
17.7
Lignin powdera
19.7
14.3
14.7
11.4
Epikote 1001
18.1
11.4
9.0
9.1
Modaflow Multiflow
16.1
3.7
7.9
8.9
Epikote 1004
17.4
10.5
9.0
7.9
7.4
8.2
3.4
Epikote 1007
21.0
11.1
13.4
11.7
Epikote 1009
18.9
9.6
10.7
7.8
Phenoxy PKHH
23.4
7.2
14.8
14.9
Epoxy curing agents Versamid 100
23.8
5.3
16.2
16.1
Versamid 115
20.3
6.6
14.1
9.6
Versamid 125
24.9
3.1
18.7
20.3
Versamid 140
26.9
2.4
18.5
24.0
Polyvinylchloride Vipla KR
17.2
a
Chlorparaffin Cereclor 70
20.0
8.3
6.8
9.8
Chlorparaffin 40
17.0
7.6
7.9
11.9
Chlorinated rubber Pergut S 5
17.4
9.5
3.8
10.0
Allopren R 10
17.4
4.3
3.9
6.1
5.3
10.4
Chlorinated polypropylene Parlon P 10
19.8
a
Polyurethane
6.2
Chlorosulfonated polyethylene
Desmophen 651
17.7
10.6
11.6
9.5
Desmophen 800
19.1
12.2
9.9
8.0
Desmophen 850
21.1
14.6
12.0
16.2
Desmophen 1100
16.0
13.1
9.2
11.4
Desmophen 1150
20.6
7.8
11.6
13.1
Desmophen 1200
19.4
7.4
6.0
9.8
Desmophen 1700
17.9
9.6
5.9
8.2
Hypalon 20
18.1
3.4
4.9
3.6
Hypalon 30
18.2
4.7
2.0
5.0
0.0
0.0
9.4
14.4
8.6
11.2
21.2
0.9
8.3
15.4
b b
Cyclized rubber Alpex
19.9
b
Nitrocellulose 1/2-sec-Nitrocellulose H 23a
15.1
Rosin derivatives
Desmolac 4200
18.7
9.6
9.9
8.2
Cellolyn 102a
Macrynal SM 510N
19.9
8.1
6.0
9.8
Pentalyn 255a
17.2
9.2
14.0
10.4
Pentalyn 830
19.6
5.7
10.7
11.4
19.2
4.6
7.6
10.4
a
Phenolic resins Super Beckacite 1001a
22.7
6.4
8.2
19.4
Phenodur 373 Ua
19.3
11.4
14.3
12.4
Ester Gum 8L
a
Polyamide Versamid 930
17.0
−1.9
14.9
7.4
Versamid 961
18.9
9.6
11.1
6.2
20.4
0.4
14.0
12.9
a
Hydrocarbon resins Piccolyte S-100a
16.14
0.4
2.8
8.4
Piccopale 110a
17.2
1.2
3.5
6.4
Piccoumarone 450 La
19.0
5.4
5.6
9.4
Styrene-butadiene elastomer (SBR) Polysar 5630
a
17.2
3.3
2.6
6.4
Versamid 965
Isocyanate Desmodur L
17.5
11.3
5.9
8.5
Demodur Na
17.6
10.0
3.7
9.3
Suprasec F-5100a
19.7
12.9
12.8
11.4
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 38
TABLE 4—(Continued) D
P
SOLUBILITY PARAMETERS
Q
487
TABLE 4—(Continued) H
R
D
Polyvinylbutyral
P
H
R
Amino resins
Mowithal B 30 H
18.6
12.9
10.3
8.3
BE 370
20.7
6.1
12.7
14.8
Mowithal B 60 H
20.2
11.2
13.2
11.2
Beetle 681
22.2
−0.4
10.1
18.4
Butvar B 76
18.2
4.3
12.7
10.4
Cymel 300a
19.9
8.3
10.4
14.4
Cymel 325
25.5
15.2
9.5
22.2
Dynomin MM 9
18.8
14.0
12.3
10.5
Dynomin UM 15
19.9
15.8
13.4
11.7
Soamin M 60
15.9
8.1
6.5
10.6
Synresen A 560
22.1
5.0
11.3
15.5
Plastopal H
20.3
8.1
14.6
12.4
Uformite MX-61
22.7
2.8
5.4
16.2
a
Polyacrylate Lucite 2042a
17.2
9.4
3.9
10.4
Lucite 2044
16.2
6.8
5.7
9.1
Plexigum MB 319
18.6
10.8
4.1
11.5
Plexigum M 527
18.4
9.4
6.5
10.7
PMMAa
18.2
10.3
7.7
8.4
a
Polyvinylacetate Mowilith 50a
20.5
11.0
9.4
13.4
Acrylate resins
Polystyrene Polystyren LGa
20.8
5.6
4.2
12.4
Vinylchloride/copolymers
Uracron 15
19.2
7.7
5.7
10.6
Paraloid P 400
19.2
9.6
9.3
12.2
Laroflex MP 45
18.4
8.4
5.8
9.0
Paraloid P 410
19.6
9.1
6.8
12.2
Vilit MB 30
20.0
8.3
6.7
9.4
9.8
10.0
12.4
20.0
8.3
6.7
9.4
Paraloid experimental resin QR 954
18.4
Vilit MC 31 Vilit MC 39
18.4
7.6
6.7
6.8
Vinylite VAGD
17.1
10.4
6.5
7.5
Baysilon UD 125
19.4
9.9
10.1
6.9
Vinylite VAGH
16.5
10.9
6.4
7.7
Wacker 190 F
16.6
1.9
8.0
8.0
Vinylite VMCA
17.7
11.1
6.9
8.7
Vinylite VMCC
17.6
11.1
6.8
8.8
Vinylite VMCH
17.6
11.1
6.4
8.6
Vinylite VYHH
17.4
10.2
5.9
7.8
Vinylite VYLF
18.1
10.3
4.2
8.3
Alkyds and polyesters
Silicone resins
Additional Special Data DEN 438 (Dow epoxy novolak)
20.3
15.4
5.3
15.1
DEN 444 (Dow epoxy novolak)
19.5
11.6
9.3
10.0
Zink silicate (CR)-Chemical resistance
23.5
17.5
16.8
15.6
Alftalat AC 366
18.6
10.0
5.0
10.4
2-Comp epoxy (CR)—Chemical resistance
18.4
9.4
10.1
7.0
Alftalat AM 756
23.0
2.2
4.2
16.9
Polyvinylidine fluoride
17.0
12.1
10.2
4.1
Alftalat AN 896
22.9
15.2
7.6
18.1
Coal tar pitch
18.7
7.5
8.9
5.8
Alftalat AN 950
22.6
13.8
8.1
17.1
3.4
10.6
5.1
20.5
9.3
9.1
12.4
PA6 polyamide chemical resistance
17.0
Alftalat AT 316 Alftalat AT 576
19.2
5.3
6.3
11.9
PA66 polyamide solubility
17.4
9.8
14.6
5.1
Alkydal F 261 HS
23.6
1.0
7.6
19.0
4.4
10.6
5.1
20.6
4.6
5.5
12.6
PA11 polyamide chemical resistance
17.0
Alkydal F 41 Duroftal T 354
17.3
4.2
7.9
9.3
Cellophane swelling
16.1
18.5
14.5
9.3
Dynapol L 812
22.6
13.1
5.8
16.8
20.5
10.5
12.3
7.3
Dynapol L 850
20.0
6.2
7.0
9.5
EVOH—solubility (ethylenevinyl alcohol)
Plexal C-34
18.1
9.0
4.8
10.4
Note: D = dispersion; P = permanent dipoles; H = hydrogen bonding; R = interation radius.
Soalkyd 1935-EGAX
18.0
11.6
8.5
9.0
a
Vesturit BL 908
18.8
12.0
6.0
11.5
Vesturit BL 915
17.7
13.0
7.6
11.5
a
Taken from Hansen, C. M., “Solubility in the Coatings Industry,” Fårg och Lack, Vol. 17, No. 4, 1971, pp. 69–77. b Calculated from solubility data in Polymer Handbook, 2nd Ed., John Wiley & Sons Inc., New York, 1975.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
488
PAINT AND COATING TESTING MANUAL
TABLE 5—List of suppliers and trademarks for paint binders and polymers Suppliers
Trademarks
Bayer (D)
Cellit, Desmophen, Desmolac, Pergut, Cellidora, Desmodur, Baysilon Alkydal
Hercules (USA)
Piccolyte, Cellolyn, Pentalyn, Ester Gum, Parlon
Ciba-Geigy
Araldite
Shell (D)
Epikote, Cariflex
Union Carbide (USA)
Vinylite, Phenoxy
Hoechst (D)
Macrynal, Phenodur, Alpex, Mowithal, Alftalat, Mowilith
Reichhold (CH)
Super Beckasite, Uformite
Polymer Corp. (CAN)
Polysar
Goodrich (USA)
Hycar
Huls (D)
Vilit, Vesturit, Buna Huls
BASF (D)
Lutonal, Laroflex, Plastopal, Polystyren
Monsanto (USA)
Modaflow, Multiflow, Butvar
Montecatini Edison (I)
Vipla
ICI (GB)
Cereclor, Allopren, Suprasec
Du Pont (USA)
Lucite
Hagedorn (D)
1/2-sec. nitrocellulose H 23
Röhm (D)
Plexigum
Rohm and Haas (USA)
Paraloid
Dynamit Nobel (D)
Dynapol
SOAB (S)
Soamin
BIP Chemicals (GB)
Beetle
Dyno Cyanamid (N)
Dynomin
DSM Resins (S)
Uracron
Reichhold Chemie (CH)
Super Beckasite, Uformite
Wacker (D)
Wacker
Dow Chemical (CH)
Ethocel
Cray Valley Prod. (GB)
Versamid
W. Biesterfeld (D)
Chlorparaffin
Synres (NL)
Synresen
American Cyanamide (USA)
Cymel
Polyplex (DK)
Plexal
Pennsylvania Industrial Chemical Corp. (USA)
Piccopal, Piccoumarone
Q
15TH EDITION
reason any average will be too low. This is discussed in more detail in Refs. [38,39].
APPLICATIONS
There are many applications documented in the literature where solubility parameters have aided in selection of solvents, understanding and controlling processes, and, in general, offered guidance where affinities among materials are of prime importance. In all of the examples below the technical aspects only are discussed. It is clear that worker safety and environmental impact must also be considered in any solvent usage. Discussion of these aspects is beyond the scope of this report. To find the optimum solvent for a polymer using solubility parameters, it is most desirable to have the solubility parameters for the polymer. Matching the parameters of an already existing solvent or combination of solvents can be done, but does not necessarily optimize the new situation. The optimum depends on what is desired of the system. A solvent with highest possible affinity for the polymer is both expensive and probably not necessary. Most coatings applications involve solvents safely within the solubility limit with a maximum of cheaper hydrocarbon solvent. Some safety is advised because temperature changes, potential variations in production, etc., can lead to a situation where solvent quality changes in an adverse manner. Balance of solvent quality on evaporation of mixed solvents is also necessary. Here again computer approaches are possible. An oxygenated solvent frequently added to hydrocarbon solvent and that has been cost effective in increasing the very important hydrogen bonding solubility parameters has been n-butanol. The mixture of equal parts xylene and n-butanol can be used in conjunction with many polymers, but a third solvent, such as a ketone or ester, is often included in small amounts to increase the polar parameter/ solvency of the mixture. Glycol ethers can also be added to hydrocarbon solvents with advantage, and the polar and hydrogen bonding parameters are higher than had n-butanol been added to the same concentration. There are many possibilities, and a solubility parameter approach is particularly valuable in quickly limiting the number of candidates. Coalescing solvents in water-reducible coatings are often those with somewhat higher hydrogen bonding parameters than the polymer, which also means they are water soluble or have considerable water solubility. The distribution between the water phase and the dispersed polymer phase depends on the relative affinities for water and the polymer. Solvents that are not particularly water soluble will preferentially be found in the polymer phase. Such coalescing solvents may be preferred for applications to porous substrates, making certain they are where they are needed. Otherwise a water-soluble coalescing solvent would tend to follow the aqueous phase penetrating the substrate and not be available to do its job in the film itself. When water evaporates, the solvent must dissolve to some extent in the polymer to promote coalescence. This can be determined and adjusted by either increasing or decreasing the affinity for the polymer. Amines are frequently added in water-reducible coatings to neutralize acid groups in polymers, thus providing a water-solubilizing amine salt. Amine in excess of that required
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 38
for total neutralization acts like a solvent. Such amine salts have been characterized separately to demonstrate that they have higher solubility parameters than either (acetic) acid or organic bases [53]. These salts are hydrophilic and have very little affinity for coatings polymers, which means they are to be found in a stabilizing role in interfaces in the aqueous phase, still being attached to the polymer. Electrostatic repulsion contributes to stability as well. Surface active agents, whether nonionic or ionic, are also to be found where the affinities of the respective parts of their molecules dictate their placement. The hydrophilic end with a high hydrogen bonding parameter will be in the aqueous phase, and the hydrophobic end will seek out an environment where energy differences are lowest. Increases in temperature lead to lower hydrogen bonding parameters, especially. For this reason, solvents with high hydrogen bonding parameters, such as glycols, glycol ethers, and alcohols, become better solvents for most polymers at higher temperatures. This can markedly affect hot-room stability in water-reducible coatings, for example, since more of the solvent will partition to the polymer phase, which swells, becomes more fluid, and has altered affinities for stabilizing surface active agents, for example. They may dissolve too readily in the dispersed polymer. Carefully controlled, these temperature effects are an advantage in water-reducible, oven-cured coatings, leading to higher film integrity. A simple approach to many practical problems is to make a two-dimensional plot of polar versus hydrogen bonding parameters with a circle (or estimated circle) for the polymer in question. One can plot the points for potential solvents and quickly arrive at a starting composition for an experiment. This can subsequently be adjusted if necessary. It should also be kept in mind that cyclic solvents generally have higher nonpolar parameters than aliphatic solvents. A long list of applications, in coatings and elsewhere, was provided as early as 1975 in a review article by Barton [54]. Many applications are given in Table 6 and in the present discussion to give an idea of what can be studied systematically using this concept. More recent references and varied applications can be found in Refs. [9,10,38,39]. Beerbower [55] has given many of the more theoretical applications for solubility parameters including correlations of the Rehbinder effect of crushing strength of aluminum oxide under various liquids, the work of adhesion for liquids on mercury, the Joffe effect of the consequences of immersion in various liquids on the fracture strength of soda-lime glass, and correlations of friction on polyethylene treated with fuming sulfuric acid H2SO4 + SO3 [55]. Direct and practical applications of solubility parameters in coatings have included their use as an aid in the selection of solvents and solvent blends for many years. Most solvent suppliers and frequent solvent users have computer programs for this purpose, although as noted above, such programs are not an absolute necessity. Reformulation to meet environmental requirements is especially important in this respect since one can quickly evaluate which of the alternatives is most likely to meet the given requirements. This includes reducing amounts of volatile organic solvent, worker safety, or other environmental concern. Optimum formulation in such cases often involves mixtures. The main thing to remember in solvent selection using solubility parameters is that the resultant values for mix-
Q
SOLUBILITY PARAMETERS
489
TABLE 6—Examples of the use of the solubility parameter Activity coefficients Aerosol formulation Biological materials and compatibility Chromatography Coal solvent extraction Compressed gases Cosmetics Cryogenic solvents Dispersion Dyes Emulsions Gas-Liquid solubility Grease removal Membrane permeability and swelling Paint film appearance Pharmaceutical Pigments Plasticizers, polymers, resins Plasticization Polymer and plasticizer compatibility Printing ink Reaction rate of radical polymerization Resistance of plastics to solvents Rubber blends Solid surface characterization—organic and inorganic Solid surface modification Solvent extraction Solvent formulation, environmental aspects Surface tension Urea-water solutions Vaporization of plasticizers Viscosity of polymer systems Water-based polymer systems, coalescents
tures can be estimated from volume fraction averages for each solubility parameter component. Solvent quality can be adjusted by the RED number concept or graphically as described above. A computer search for nearest neighbors for a given single solvent has been used many times to locate alternates. A similar application is to predict which other solvents will probably be aggressive to a chemically resistant coating where very limited data have indicated a single solvent or two are somewhat aggressive. A nearest neighbor search involves calculation of the quantity Rs for a whole
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
490
PAINT AND COATING TESTING MANUAL
Q
database, for example, and then arranging the printout in RED number order with the potentially most aggressive at the top of the list. Solvents with RED less than 1.0 are “good” and easily recognized. In many cases a marginal solvent is desired, in which case RED numbers just under 1.0 will be sought. Marginal solvent quality will ensure that polymer adsorbed onto pigment surfaces has little reason to dissolve away from that surface where it is desired as a stabilizing factor in the product. The solvent in this case should have a RED number for the pigment surface greater than 1.0 to aid in the planned affinity approach to pigment dispersion stability. A sketch of the optimum relations is given in Fig. 4, where the marginal solvent is Number 1. Solvent 2 would probably be too expensive and, in addition, will probably dissolve the polymer too well. In special applications this extended polymer chain configuration is desirable, but a solid anchor to the pigment surface is required. A good anchor has high affinity for the pigment surface and marginal or no affinity for the solvent. Solvent 3 would adsorb onto the pigment surface preferentially, and pigment dispersion stability will be poor. Pigments have been characterized by long-time suspension studies where some solvents will suspend the fines for days, months, or years, while others of similar viscosity yield rapid settling. The suspending solvents are the good ones and can be used to define a sphere for surface wetting/adsorption. Details of such characterizations are given in Refs. [38,39]. In other cases, better matches between solvent and polymer solubility parameters are required. This is true when two polymers are mixed and one of them precipitates. This is most likely the polymer with larger molecular weight, and it must be dissolved better. Lower RED numbers with respect to this polymer are desired, while still maintaining affinity for the other polymer. Miscible blends of two polymers have been found using a solvent mixture composed exclusively of nonsolvents. This is demonstrated schematically in Fig. 5, where it can be seen that different percentage blends of solvents 1 and 2 will have different relative affinities for the polymers. No other alternative theory of polymer solution thermodynamics can duplicate this predictive ability. Polymer miscibility is enhanced by larger overlapping solubility regions for the polymers as sketched in Fig. 6. Polymers A and B should be compatible, while C will not be. Such a systematic analysis allows modification of a given polymer to provide more overlap and enhanced compatibility. The advantages of a copolymer containing
Fig. 4—Solubility parameter relations for optimum pigment dispersion stability.
15TH EDITION
Fig. 5—Solubility relations for polymer mixtures can be quickly evaluated to ensure solution stability. Even mixtures of nonsolvents can be systematically used to regulate solution behavior.
the monomers of A or B and C should also be evident. Such a copolymer will essentially couple the system together. Van Dyk et al. [56] have correlated the inherent viscosity of polymer solutions with the solubility parameter. This is interesting in that the solubility parameter is a thermodynamic consideration, while the viscosity is a kinetic phenomenon. Solvents with higher affinities give greater polymer chain extension in solution, and the inherent viscosity—the solution viscosity divided by the solvent viscosity at polymer concentrations approaching zero—is an expression reflecting polymer chain extension in solution. Higher intrinsic viscosities were found for solvents with solubility parameters nearest the polymer solubility parameters. As stated above, this approach may not always be advisable since many polymers have polar and hydrogen bonding parameters higher than the test liquids that have been used. The use of supercritical gases as solvents has become more common in recent years. Space limitations prevent going into the details of these developments. It should be noted, however, that when a gas is compressed, its cohesive energy density increases. This means that nonpolar gases with their low nonpolar solubility parameters can begin to dissolve given organic materials, which otherwise have solubility parameters that are too high. Increasing
Fig. 6—Schematic representation showing expected miscibility of polymers A and B with each other but not with polymer C.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 38
the nonpolar solubility parameter of the gas by increasing the pressure causes a closer match with the corresponding parameter for potential solutes. Similar behavior is found for polar gases such as carbon dioxide. The prevailing pressure and temperature conditions determine its cohesive energy density and changes in pressure or temperature change solubility relations for this reason. Whereas nonpolar gases are most suitably used for relatively nonpolar solutes, carbon dioxide—and in principle other polar gases—are most suitably used in connection with more polar solutes. The solubility parameters for carbon dioxide have been reported [57] based on the room temperature solubility of the gas in different liquids (␦d, ␦p, ␦h equal to 15.3, 6.9, 4.1). These parameters resemble those of a higher ketone, but are subject to revision when larger data sets are processed. The ␦d, ␦p, ␦h parameters for carbon dioxide have been revised to 15.7, 6.3, and 5.7 based on a correlation with a perfect data fit and R0 equal to only 3.3, using a much larger data set collected by Williams and reported in Chapter 10 of Ref. [39]. The HSP are also reported in this reference at higher temperatures and pressures. The ␦p parameter experimentally confirms the presence of sizeable values for compounds with zero dipole moments, and the ␦h parameter emphasizes the need to use terminology electron exchange for this parameter, traditionally called the hydrogen bonding parameter, since there is no hydrogen atom. Chapter 3 in Ref. [39] has a plot of the HSP for water at temperatures above the normal boiling point. Pigment wetting/suspension characteristics have been presented earlier [10,14,16]. Mixtures of nonsuspending solvents could also be found which, when admixed, provided predictably higher affinity and suspension of pigment particles for prolonged periods of time. An exceptionally clear demonstration of pigment adsorption properties is given in Ref. [10] where a triangular plot of the three partial parameters shows the clearly different adsorption properties of untreated zinc oxide powder and organic phosphate surface-treated zinc oxide powder. This triangular approach to plotting was developed by Teas [52]. Other surface characterizations have also been given for surfaces such as coatings and metal substrates [23,24]. These characterizations have been of the type sketched in Fig. 7. Such cohesive energy plots can lead to systematic modifications of systems to improve adhesion. It might be added parenthetically that equal information can be obtained from plots of the cosine of the contact angle versus the solubility parameter as for plots of these same data versus liquid surface tension [23]. The energy information obtained in these types of studies, the critical surface tension, corresponds to the condition of RED equal to 1.0, with a solubility parameter approach, that is, marginal affinity. The list of applications has recently been expanded to include correlations of the breakthrough times for common types of chemical protective clothing such as butyl rubber, nitrile rubber, plasticized polyvinyl chloride, neoprene, Viton rubber, polyvinyl alcohol, etc. [58]. Fig. 8 shows the importance of combined use of the RED number and molecular volume in correlating 3 h breakthrough times for a fluoropolymer chemical protective product [59]. Monomers with terminal double bonds diffuse more rapidly than comparisons with non-double-bonded molecules of similar size and solubility parameter would have predicted. The smaller cross section at the end of the molecule reminds
Q
SOLUBILITY PARAMETERS
491
Fig. 7—Schematic diagrams showing surface energy/contact angle characterizations using the Hansen solubility parameters (cohesion energy parameters).
one of a nail, and the preferred direction and relative rapidity of transport become easily understandable. In addition it should be noted that even biological materials, some of which have interest for coatings applications and use, have been assigned Hansen solubility parameters [60]. These include keratin (which relates to skin permeation), fat, sucrose, blood serum and zein (which are proteins), urea, lignin (wood penetration), and chlorophyll (which closely resembles lignin) [57,60]. The characterization of other biological materials is also possible, of course, with the assignment of HSP to DNA being one of the most recent results [39]. Even inorganic salts have been characterized by solubility parameters [61]. The practice of dissolving polymers in solutions of organic liquids and inorganic salts can thus be explained by the solubility parameter. A solubility parameter correlation of the chemical resistance of an inorganic zinc silicate coating is reported in Table 4. Finally, it might be noted that Hildebrand presented a chapter on the solubility parameters of metals [1]. Unfortunately, we do not often coat metal, but rather metal oxides, for which no solubility parameter work has been reported. Examples in addition to the above are discussed in a larger context in Refs. [38,39]. The material in the following section has not been discussed in Ref. [38], but rather in Refs. [39,62]. Inclusion here is justified by the presence of solvents and additives in coatings, which may produce these catastrophic effects in rigid polymeric systems.
HANSEN SOLUBILITY PARAMETERS AND ENVIRONMENTAL STRESS CRACKING (ESC)
The Hansen solubility parameter approach allows systematic analysis of environmental stress cracking situations [62]. One can locate those liquids that are most likely to cause the problem based on limited experimental data for a reasonable number of different solvents. The state of the stress in a polymer is clearly important. Tensile stress is required for the phenomena to occur. Higher stress levels lead more easily to cracking. Test solvents to evaluate the stress level present in given polymers can be systematically located. However, it is clearly not just solubility relations that control the phenomena. The size and shape of the environmental challenge chemical are also important. These relate to the rate of uptake (diffusion coefficient). These factors collectively control how rapidly a challenge
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
492
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 8—Effects of molecular size (molar volume) and affinity (RED number) on the breakthrough time of Challenge 5100 [59].
chemical can absorb into the surface of a polymeric material, this being a requirement for the cracking to occur. The absorbed chemical changes the local stress-bearing properties, and cracking can occur in unfavorable situations. The times required may vary from seconds to years, depending on the situation. Stress relaxation caused by the plasticizing effect of absorbed liquid can reduce the risk of ESC, but can cause other problems, such as excessive softening of the polymer. ESC phenomena in polymers are not fully understood, but progress toward this goal is being made as discussed in the following. In general, those solvents causing cracking of a given polymer will be taken up in lesser amounts at equilibrium absorption than those amounts which will fully dissolve the polymer. In terms of the Hansen approach, this means that there will be the usual spherical characterization for polymer solubility. Essentially concentric to this, but with a larger radius, is another spherical characterization encompassing both the true solvents at RED numbers less than about 0.8, and the cracking solvents at RED numbers typically between about 0.8 and 1.0. Fig. 9 shows these relations for a COC type (Cyclo-olefinic Copolymer) polymer. The basis for this correlation was to consider as aggressive (RED < 1) both those solvents that truly dissolve the polymer as well as those that cause cracking. This type of plot can be made for other polymers. The above discussion is for exposure to a pure liquid. It is also known that aqueous solutions can leave smaller amounts of aggressive chemicals when the water has
evaporated. In this case almost any solvent, either good or swelling only, will be a potential problem in this respect. The small amount acts like a concentrated solvent, but only at the surface. Solution can not occur, other than at the
Fig. 9—Hansen solubility parameters correlate both solubility (inner sphere) and solubility plus environmental stress cracking (outer sphere). The liquids that crack the polymer without application of external stress have Hansen solubility parameters placing them in the clear shell in the figure. From Ref. [62]. Reprinted with permission from Industrial and Engineering Chemistry Research, Vol. 40, No. 1, 2001, pp. 21–25. Copyright 2001, American Chemical Society.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 38
surface, because of the limited amount of solvent, but the stress situation at the surface is altered unfavorably, thus allowing cracks to form. There are also examples of ESC where there was no measurable weight gain of the liquids causing the failure. Crack initiation by the rotation of segments of polymers at the surface exposed to these liquids is thought to be the reason. The rotation is driven by the HSP of the liquid resembling the HSP of a buried or partially buried polymer segment, thus reducing the free energy of the system by rotation to surroundings of similar (cohesion) energy. Higher energy segments of polymers are most often buried just beneath the surface, if this is possible. The polymer segments in contact with air are most often those of lower energy, again, to keep the free energy at a minimum [39].
CONCLUSIONS
The background and many uses of the solubility parameter concept have been described in detail. Tables of solubility parameters for many liquids and polymers have been presented. Systematic use of solubility, swelling, or permeation data for polymers allows their characterization by these same cohesive energy parameters. Surfaces can also be characterized using these same parameters. This provides data for optimizing solvent selection, improving compatibility, and enhancing pigment dispersion and adhesion. When all of the materials involved in a given product and application can be characterized with the same energy parameters, the possibility exists to predict interactions among them, even in complicated situations. Methods for estimating the three Hansen solubility parameters have been given with as much detail as has been possible to ensure their more uniform use in the future. It has not been possible to deal with all the primarily theoretical problems with the solubility parameter concept. Some of these are dealt with in the literature cited, most notably in Refs. [6,9,10,18,26,38,39]. A simple approach is described to understand affinities in such varied materials as gases, liquids, polymers, biological materials, surfaces, organic and inorganic coatings, inorganic salts, and metals. An appeal made to the scientific community earlier [37] to expand research on this seemingly universal approach appears to have been responded to by Panayiotou in Chapter 3 of Ref. [39]. Statistical thermodynamics calculations of the ␦d, ␦p, and ␦h parameters, starting with the hydrogen bonding parameter, and extending to the other parameters, have shown amazing agreement with the values reported by the author 40 years ago [13–16]. Information related to the Hansen solubility parameters is continually being updated on the website www.hansensolubility. com.
References [1] Hildebrand, J., and Scott, R. L., The Solubility of Nonelectrolytes, 3rd ed., Reinhold, New York, 1950. [2] Hildebrand, J., and Scott, R. L., Regular Solutions, PrenticeHall Inc., Englewood Cliffs, NJ, 1962. [3] Patterson, D., and Delmas, G., “New Aspects of Polymer Solution Thermodynamics,” Off. Dig. Fed. Soc. Paint Technol., Vol. 34, No. 450, 1962, pp. 677–692. [4] Delmas, D., Patterson, D., and Somcynsky, T., “Thermodynamics of Polyisobutylene-n-Alkane Systems,” J. Polym. Sci., Vol. 57, 1962, pp. 79–98.
Q
SOLUBILITY PARAMETERS
493
[5] Bhattacharyya, S. N., Patterson, D., and Somcynsky, T., “The Principle of Corresponding States and the Excess Functions of n-Alkane Mixtures,” Physica (Amsterdam), Vol. 30, 1964, pp. 1276–1292. [6] Patterson, D., “Role of Free Volume Changes in Polymer Solution Thermodynamics,” J. Polym. Set, Part C: Polym. Symp., Vol. 16, 1968, pp. 3379–3389. [7] Patterson, D. D., “Introduction to Thermodynamics of Polymer Solubility,” J. Paint Technol., Vol. 41, No. 536, 1969, pp. 489–493. [8] Biros, J., Zeman, L., and Patterson, D., “Prediction of the C Parameter by the Solubility Parameter and Corresponding States Theories,” Macromolecules, Vol. 4, No. 1, 1971, pp. 30–35. [9] Barton, A. F. M., Handbook of Solubility Parameters and Other Cohesion Parameters, CRC Press Inc., Boca Raton, FL, 1983. [10] Gardon, J. L., and Teas, J. P., “Solubility Parameters,” Treatise on Coatings, Vol. 2, Characterization of Coatings: Physical Techniques, Part II, R. R. Myers and J. S. Long, Eds., Marcel Dekker, New York, 1976, Chap. 8. [11] Burrell, H., “Solubility Parameters for Film Formers,” Off. Dig. Fed. Soc. Paint Technol., Vol. 27, No. 369, 1972, pp. 726–758; Burrell, H., “A Solvent Formulating Chart,” Off. Dig. Fed. Soc. Paint Technol., Vol. 29, No. 394, 1957, pp. 1159–1173; Burrell, H., “The Use of the Solubility Parameter Concept in the United States,” VI Federation d’Associations de Techniciens des Industries des Peintures, Vernis, Emaux et Encres d’Imprimerie de l’Europe Continentale, Congress Book, 1962, pp. 21–30. [12] Blanks, R. F., and Prausnitz, J. M., “Thermodynamics of Polymer Solubility in Polar and Nonpolar Systems,” Ind. Eng. Chem. Fundam., Vol. 3, No. 1, 1964, pp. 1–8. [13] Hansen, C. M., “The Three Dimensional Solubility Parameter—Key to Paint Component Affinities I,” J. Paint Technol., Vol. 39, No. 505, 1967, pp. 104–117. [14] Hansen, C. M., “The Three Dimensional Solubility Parameter—Key to Paint Component Affinities II,” J. Paint Technol., Vol. 39, No. 511, 1967, pp. 505–510. [15] Hansen, C. M., and Skaarup, K., “The Three Dimensional Solubility Parameter—Key to Paint Component Affinities III,” J. Paint Technol., Vol. 39, No. 511, 1967, pp. 511–514. [16] Hansen, C. M., 1967, “The Three Dimensional Solubility Parameter and Solvent Diffusion Coefficient,” Ph.D. thesis Danish Technical Press, Copenhagen. [17] Hansen, C. M., and Beerbower, A., “Solubility Parameters,” Kirk-Othmer Encyclopedia of Chemical Technology, Supplement Volume, 2nd ed., A. Standen, Ed., Interscience, New York, 1971, pp. 889–910. [18] Barton, A. F. M., “Applications of Solubility Parameters and Other Cohesion Energy Parameters in Polymer Science and Technology,” Pure Appl. Chem., Vol. 57, No. 7, 1985, pp. 905–912. [19] Sørensen, P., “Application of the Acid/Base Concept Describing the Interaction between Pigments, Binders, and Solvents,” J. Paint Technol., Vol. 47, No. 602, 1975, pp. 31–39. [20] Van Dyk, J. W., Fourth Chemical Congress of America, New York, 25–30 Aug. 1991. [21] Anonymous (Note: This was in fact Van Dyk, J. W., but this does not appear on the bulletin), “Using Dimethyl Sulfoxide (DMSO) in Industrial Formulations,” Report No. 102, Gaylord Chemical Corp., Slidell, LA, 1992. [22] Karger, B. L., Snyder, L. R., and Eon, C., “Expanded Solubility Parameter Treatment for Classification and Use of Chromatographic Solvents and Adsorbents,” Anal Chem., Vol. 50, No. 14, 1978, pp. 2126–2136. [23] Hansen, C. M., and Wallström, E., “On the Use of Cohesion Parameters to Characterize Surfaces,” J. Adhes., Vol. 15, 1983, pp. 275–286. [24] Hansen, C. M., “Characterization of Surfaces by Spreading Liquids,” J. Paint Technol., Vol. 42, No. 550, 1970, pp. 660– 664; Hansen, C. M., “Surface Dewetting and Coatings Performance,” J. Paint Technol., Vol. 44, No. 570, 1972, pp. 57–60. [25] Hansen, C. M., and Pierce, P. E., “Surface Effects in Coatings Processes,” Ind. Eng. Chem. Prod. Res. Dev., Vol. 13, No. 4, 1974, pp. 218–225.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
494
PAINT AND COATING TESTING MANUAL
Q
[26] Gardon, J. L., “Critical Review of Concepts Common to Cohesive Energy Density, Surface Tension, Tensile Strength, Heat of Mixing, Interfacial Tension and Butt Joint Strength,” J. Colloid Interface Set, Vol. 59, No. 3, 1977, pp. 582–596. [27] Flory, P. J., Principles of Polymer Chemistry, Cornell University Press, New York, 1953. [28] van Krevelen, D. W., and Hoftyzer, P. J., Properties of Polymers: Their Estimation and Correlation with Chemical Structure, 2nd ed., Elsevier, Amsterdam, 1976. [29] Beerbower, A., “Environmental Capability of Liquids,” Interdisciplinary Approach to Liquid Lubricant Technology, NASA Publication SP-318, 1973, pp. 365–431. [30] Fedors, R. F., “A Method for Estimating Both the Solubility Parameters and Molar Volumes of Liquids,” Polym. Eng. Sci., Vol. 14, No. 2, 1974, pp. 147–154. [31] Koenhen, D. N., and Smolders, C. A., “The Determination of Solubility Parameters of Solvents and Polymers by Means of Correlation with Other Physical Quantities,” J. Appl. Polym. Sci., Vol. 19, 1975, pp. 1163–1179. [32] Anonymous, “Co-Act—A Dynamic Program for Solvent Selection,” Brochure, Exxon Chemical International, Inc., 1989. [33] Dante, M. F., Bittar, A. D., and Caillault, J. J., “Program Calculates Solvent Properties and Solubility Parameters,” Mod. Paint and Coat., Vol. 79, No. 9, 1989, pp. 46–51. [34] Hoy, K. L., “New Values of the Solubility Parameters from Vapor Pressure Data,” J. Am. Oil Chem. Soc., Vol. 42, No. 541, 1970, pp. 76–118. [35] Myers, M. M., and Abu-Isa, I. A., “Elastomer Solvent Interactions III-Effects of Methanol Mixtures on Fluorocarbon Elastomers,” J. Appl. Polym. Sci., Vol. 32, 1986, pp. 3515–3539. [36] Hoy, K. L., Tables of Solubility Parameters, Union Carbide Corp., Research and Development Dept., South Charleston, WV, 1985. [37] Hansen, C. M., “Solubility Parameters” in Paint Testing Manual, Manual 17, J. V. Koleske, Ed., American Society for Testing and Materials, Philadelphia, 1995, pp. 383–404. [38] Hansen, C. M., Hansen Solubility Parameters—A User’s Handbook, CRC Press, Boca Raton, FL, 1999/2000. [39] Hansen, C. M., Hansen Solubility Parameters—A User’s Handbook, 2nd, Ed., CRC Press, Boca Raton, FL, 2007. [40] Hansen, C. M., “Polymeropløselighed—Prigogines Teori om Korresponderende Tilstande og Hansen Opløselighedsparameterteori Bekræfter Hinanden (Polymer Solubility— Prigogine’s Corresponding States Theory and Hansen Solubility Parameter Theory Confirm Each Other),” Dansk Kemi, Vol. 78, No. 9, 1997, pp. 4–6. [41] Prigogine, I. (with the collaboration of A. Bellemans, and A. Mathot), The Molecular Theory of Solutions, North-Holland, Amsterdam, 1957, Chaps. 16 and 17. [42] Rowlinson, J. S., Liquids and Liquid Mixtures, Butterworths Scientific Publications, London, 1959, pp. 254–257 and 313– 318. [43] Reid, R. C., and Sherwood, T. K., Properties of Gases and Liquids, McGraw-Hill, New York, 1958 (Lydersen Method, see also Ref [31]). [44] McLellan, A. L., Tables of Experimental Dipole Moments, W. H. Freeman, San Francisco, 1963. [45] Tables of Physical and Thermodynamic Properties of Pure Compounds, American Institute of Chemical Engineers Design Institute for Physical Property Research, Project 801, Data Compilation, R. P. Danner and T. E. Daubert, Project Supervi-
15TH EDITION
[46]
[47] [48] [49] [50]
[51] [52] [53] [54] [55]
[56] [57] [58]
[59]
[60] [61] [62]
sors, DIPPR Data Compilation Project, Department of Chemical Engineering, Pennsylvania State University, University Park, PA. Hansen, C. M., “Selection of Chemicals for Permeation Testing Based on New Solubility Parameter Models for Challenge 5100 and Challenge 5200,” Contract No. DTCG50-89-P-0333, U.S. Coast Guard, June 1989, Danish Isotope Centre, Copenhagen. CRC Handbook of Chemistry and Physics, 65th ed., R. C. Weast, Ed., CRC Press Inc., Boca Raton, FL, 1988-1989, pp. C-672–C-683. Majer, V., “Enthalpy of Vaporization of Organic Compounds,” Handbook of Chemistry and Physics, 72nd ed., D. R. Lide, Ed., CRC Press Inc., Boca Raton, 1991–1992, pp. 6-100–6-107. Fishtine, S. H., “Reliable Latent Heats of Vaporization,” Ind. Eng. Chem., Vol. 55, No. 4, 1963, pp. 20–28; IECHAD, Vol. 55, No. 5, pp. 55–60; IECHAD, Vol. 55, No. 6, pp. 47–56. Saarnak, A., Hansen, C. M., and Wallström, E., “Solubility Parameters, Characterization of Paints and Polymers,” Report from Scandinavian Paint and Printing Ink Research Institute, January 1990. Barton, A. F. M., Handbook of Polymer-Liquid Interaction Parameters and Solubility Parameters, CRC Press, Inc., Boca Raton, FL, 1990. Teas, J. P., “Graphic Analysis of Resin Solubilities,” J. Paint Technol., Vol. 40, No. 516, 1968, pp. 19–25. Hansen, C. M., “Some Aspects of Acid/Base Interactions,” (Einige Aspekte der Süure/Base-Wechselwirkung) (in German) Farbe und Lack, Vol. 83, No. 7, 1977, pp. 595–598. Barton, A. F. M., “Solubility Parameters,” Chem. Rev. (Washington, D.C.), Vol. 75, No. 6, 1975, pp. 731–753. Beerbower, A., “Boundary Lubrication—Scientific and Technical Applications Forecast, AD747336,” Office of the Chief of Research and Development, Department of the Army, Washington, DC, 1972. Van Dyk, J. W., Frisch, H. L., and Wu, D. T., “Solubility, Solvency, Solubility Parameters,” Ind. Eng. Chem. Prod. Res. Dev., Vol. 24, No. 3, 1985, pp. 473–478. Hansen, C. M., “25 Years with Solubility Parameters (25 Årmed Opløselighedsparametrene) (in Danish),” Dansk Kemi, Vol. 73, No. 8, 1992, pp. 18–22. Hansen, C. M., and Hansen, K. M., “Solubility Parameter Prediction of the Barrier Properties of Chemical Protective Clothing,” Performance of Protective Clothing: Second Symposium, ASTM STP 989, S. Z. Mansdorf, R. Sager, and A. P. Nielsen, Eds., American Society for Testing and Materials, Philadelphia, 1988, pp. 197–208. Hansen, C. M., Billing, C. B., and Bentz, A. P., “Selection and Use of Molecular Parameters to Predict Permeation Through Fluoropolymer-Based Protective Clothing Materials,” The Performance of Protective Clothing; Fourth Volume, ASTM STP 1133, J. P. McBriarty and N. W. Henry, Eds., American Society for Testing and Materials, Philadelphia, 1992, pp. 894–907. Hansen, C. M., “The Affinities of Organic Solvents in Biological Systems,” J. Am. Ind. Hyg. Assoc., Vol. 49, No. 6, 1988, pp. 301–308. Hansen, C. M., “The Universality of the Solubility Parameter,” Ind. Eng. Chem. Res., Vol. 8, No. 1, 1969, pp. 2–11. Hansen, C. M., and Just, L., “Prediction of Environmental Stress Cracking in Plastics with Hansen Solubility Parameters,” Ind. Eng. Chem. Res., Vol. 40, No. 1, 2001, pp. 21–25.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
Part 9: Films for Testing
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
39
MNL17-EB/Jan. 2012
Cure: The Process and Its Measurement Thomas J. Miranda1 INTRODUCTION
AMONG THE MOST CRITICAL AND OFTEN MISunderstood properties of a coating are cure and the measurement of cure. If ASTM standard technology [1] is examined, one will find no listing for the word “cure.” Cure is something that can have different meanings for different people. In the coating and paint field of study, cure is usually thought of as the chemical reaction between functional groups that results in cross-linking of a system by one or more of a variety of schemes including polymerization of monomers, rearrangement, condensation or elimination, reaction with adventitious moisture or oxygen, as well as others. However, the word can be taken to mean much more and should include the drying of a lacquer as when thermoplastic coatings are applied, the solidification and/ or crystallization of polyolefins or other hot-melt applied coatings, the drying of latexes or emulsions that can form films through the high-pressure compacting forces of surface tension, as well as of other ways of forming a coherent, useful thin film with good mechanical, electrical, or other properties. This chapter will mainly be concerned with cross-linked systems and cure will be taken to mean reactions that result in a thermoset or network-formed coating [2], unless otherwise indicated. The desired and often optimum chemical and physical properties of a coating are dependent on proper curing conditions—time, temperature, and humidity. For example, if a thermosetable, acrylic polymer-based lacquer is applied to a substrate and solvent is allowed to evaporate under ambient conditions, the resulting film might have cracks, low gloss, and be hard and brittle. Yet, the same system could result in a coherent, useful film if the solvent were evaporated under controlled temperatures for specific times. The film might still be brittle and have poor chemical resistance, but such factors may not be important in certain instances. However, strength properties will be markedly improved as a cross-linking agent is added, and the system is allowed to react under proper temperature conditions for an appropriate length of time. The resulting coating will be glossy, hard, tough, and chemical resistant. The time and temperature relationship and the interaction of these parameters has been reported by Neag and Prime [3] in the case of powder coating cure.
Concept of Cure
Polymer is a word meaning many (poly) units (mers) wherein the units are monomers or single-unit mers. Thus, polymers are high molecular weight molecules that are comprised of monomers that may be the same or different, though if the monomers comprising the polymer are different, the polymer is referred to as a “copolymer.” Polymers are formed by a variety of polymerization mechanisms including addition (such as the polyacrylates, polyvinyls, etc.), cationic (as polyepoxides), condensation (as polyesters), coordination (as polyolefins), ring opening (as Nylon 6 and polylactones), and rearrangement (as polyurethanes). These compounds known as polymers range from high to low molecular weight in nature. High molecular weight polymers can form coating films with high gloss and good mechanical properties; however, to achieve such useful properties, the applied solids must be low due to the high viscosities involved when such large-size molecules are used. In the coatings arena, usually low molecular weight polymers or oligomers comprising a relatively small number of chain units that either have one or more functional groups or a co-unit that has functional or polyfunctional groups on as many of the oligomer molecules as possible. If used alone, the oligomers would have poor mechanical and chemical properties and would either be sticky masses (glass transition temperature, Tg, below room temperature) or “rock hard” chunks (Tg above room temperature). To transform these oligomers into useful materials with good mechanical and other properties, it is necessary to build or increase molecular weight. This transformation is accomplished by the addition of a cross-linking agent to the coating formulation wherein said compound will react with the functional groups on the oligomers and then cure the liquid system into a film comprised of extremely high molecular weight molecules.2 It should be clear from the above that thermoplastic coatings are soluble in particular compounds we term “solvents” and, in addition, if heated will liquefy with the viscosity dependent on the molecular weight. In contrast, thermoset coatings are insoluble but will often swell in particular compounds and if heated may soften but will thermally degrade rather than liquefy. Of course, additives [4] that will markedly improve the path to achieving a coating with designed gloss, good
Senior Consultant, Consolidated Research Inc., 16731 Brick Road, Granger, IN 46530. The concept of molecular weight becomes lost when one discussed cross-linked systems. Dr. Herman Mark described the “rubber” in an automobile tire after vulcanizing or cross-linking as one gigantic molecule. Although the definition may be a bit ambitious, since defects and the laws of probability exist, it is a good model to keep in mind when thinking about cross-linked chemicals. Thus, a chemical coating can be thought of as a huge molecule with an extremely high molecular weight that is incapable of being dissolved.
1 2
497 Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed byASTM International This standard is for EDUCATIONAL USE ONLY. Copyright © 2012 by www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
498
PAINT AND COATING TESTING MANUAL
Q
mechanical and chemical properties, as well as rapid cure are very important to the cure process. Accelerators, catalysts, or other additives insure that cure or cross-linking will be effected in an acceptable time. For example, in radiation cure [5], monomeric and/or oligomeric molecules are combined into a very fluid formulation and ultraviolet radiationactivated cationic or free-radical initiator additives known as photoinitiators cause the molecules to react in an extremely rapid manner at coating line speeds of even several hundred feet per minute. When an electron beam is used to effect cure of such systems, photoinitiators are not required, though flow-control additives would be used. With radiation activation, cure is achieved in the “snap of a finger.” When less-energetic reactions than those used in radiation curing are involved, the cure process will be relatively slow, say on the order of minutes at an elevated temperature. Examples of such a cure system might be the acidcatalyzed, aminoplast cross-linking of hydroxyl-functional oligomers to form thermoset films or the controlled evaporation of solvent or other diluent from lacquers or latexes to form thermoplastic films. In the case of powder coatings, the coating material is in a dry form and thermal energy is used to melt the powder particles and time at a temperature is required to allow the particles to flow into a continuous coating. If a cross-linking agent is present in the powder coating, the cure system must be orchestrated in such a manner that liquefaction and flow take place before significant network formation occurs. Very slow reactions occur in thermoset systems that cure by reaction of adventitious oxygen with unsaturation in the oligomeric molecules. Many details regarding coating film formation [6] and the preparation of films for coating tests [7] can be found elsewhere.
15TH EDITION
decrease the glass transition temperature. As described, the curing results in a thermoplastic coating. If desired, these systems can be made into thermoset films by incorporating a functional group into the latex molecules and employing a cross-linking agent.
Oxidative Cross-linking
Oxidative cross-linking takes place in systems that contain reactive, unsaturated double bonds as are found in unsaturated fatty acids, drying oils such as linseed, soy, tung, and tall oil (see Chapters 4 and 5 for a discussion of drying oils and driers), as well as certain alkyds and oleoresinous varnishes. This vehicle is formulated with a suitable solvent, a drier (often termed a “siccative”) such as cobalt, manganese, or zirconium salts in the form of naphthenates, octanoates, or the like, and other ingredients. Curing takes place through adventitious oxygen interacting with the molecular unsaturation. The coating film is applied by brushing, spraying, or other means and solvent is removed by evaporation. The resultant liquid or tacky film then undergoes an induction period in which natural antioxidants inherent in the starting materials are overcome as oxygen is absorbed from the atmosphere. This is followed by hydroperoxide formation on carbon atoms alpha to double bonds. This is followed by peroxide decomposition and formation of free radicals that initiate cross-link formation at the unsaturated sites with the result formation of a cross-linked thermoset film [8,9]. A simplified reaction scheme in which R and R′ are alkyl groups is shown below [10].
CURING MECHANISMS Lacquers
Lacquers consist of thermoplastic polymers that are dissolved in a solvent. When the coating is applied, the solvent evaporates under controlled conditions and a thermoplastic film is formed. This is the type coating used in certain high quality furniture finishes. A number of years ago, many automobiles were coated in this manner, but today such coatings are thermoset in nature.
The starting materials are converted into polymer according to the following scheme:
Emulsions and Latexes
Another means of forming a thermoplastic film is film formation from an aqueous-based emulsion or latex. In this instance, a polymer such as a vinyl acetate-acrylic copolymer latex is formulated with colorants and additives, cast, and the water carrier is evaporated with a coherent film resulting. In such systems, the water is not a solvent but a dispersing medium. Microscopically, the polymer appears as tiny ball bearings suspended in water with surface active agents present to keep the polymer particles dispersed and free from settling. As the water evaporates, hydrodynamic pressure causes the small spheres to crowd closer and closer until capillary action forces the particles to fuse and form a continuous film. If the polymer is too hard in nature, a poorly adherent, brittle film is formed. Often, this is the reason paint manufacturers recommend that water-based latex paints not be applied at low temperatures. To facilitate fusion and continuity, small amounts of a coalescing solvent such as diethylene glycol monobutyl ether are added to
Termination by coupling increases the molecular size—say, doubling if the joined chains are the same in size—and eventually a network or cross-linked system is formed. If properly formulated, the final product has the desired chemical and physical characteristics.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 39
Q
Reactive Cross-linking
In this type of cure system, a reactive intermediate is added to a functionalized polymer or oligomer and a further chemical reaction is initiated by heating alone or in the presence of a catalyst or by some other means. The functionality on the polymer molecules often is hydroxy in nature. The reactive intermediate is often an amino compound such as hexa(methoxymethyl)melamine.
CURE: THE PROCESS AND ITS MEASUREMENT
499
UREA COMPOUNDS
A scheme similar to that described above for HMMA and other melamine compounds is also applicable to ureaformaldehyde (UF) compounds. To form the UF compounds, urea is reacted with formaldehyde and the product is end capped with alkyl alcohols—usually methyl, isopropyl, or butyl alcohols. A full reacted urea intermediate in which methanol has been used as the capping agent has the following structure:
This multifunctional compound is capable of reacting with hydroxyl groups to produce cross-links with by-products such as alcohol, water, and formaldehyde. Note that in the above examples, neither the melamines, ureas, or most polymers would form a good coating if used alone. The polymers act as flexibilizers and/or tougheners for the hard, brittle cross-linking agents.
MELAMINE
Melamine-formaldehyde condensates are often used to cure a reactive polymer. Melamine is 2,4,6-triamino1,3,5-triazine, a compound prepared by condensing three molecules of dicyandiamide. Melamine, in turn, can be condensed with up to six molecules of formaldehyde and 6 molecules of methanol to form hexa(methoxymethyl) melamine (HMMA) (see Chapter 8) or an oligomeric phenolic compound (see Chapter 11). Alcohols other than methanol can be used as the hydroxylcapping groups. Examples of commonly used alcohols are isobutanol and butanol. In practice, these compounds can be reacted with a functionalized polymer such as an acrylic or an alkyd that contain either acid or hydroxyl groups, which will react with the melamine compound to liberate an alcohol byproduct with the multifunctional nature of the melamine compound leading to a cross-linked coating film. This cure system is briefly indicated by the following reaction scheme:
wherein A is the residue from HMMA and P represents the hydroxyl functionalized polymer that is being cross-linked. Note that after the above reaction, the residue A still has five functional groups that can react with other hydroxyl groups on the same polymer molecule or on other polymer molecules to set up the network. In addition to the by-product methanol, it is possible for formaldehyde to be a by-product. Narayan and coworkers [11] have addressed various aspects of the cure and properties of polyesters cured with HMMA. Numerous other articles can be found in the literature.
EPOXIDES
Epoxides represent a ring-opening-mechanism form of curing coatings. Aromatic epoxides are produced by the reaction of epichlorohydrin and bisphenol A, with the latter compound a condensation product of acetone and phenol. Aliphatic epoxides are prepared by the peroxidation of unsaturated linear or cyclic olefins. A typical epoxide used in coatings, is the diglycidylether of bisphenol A, which is described below and wherein Z is a phenylene group.
The epoxide groups can react with active hydrogen atoms such as those found on amine, hydroxyl, and carboxylic acid groups. Aromatic epoxides react very rapidly with amines and hydroxyl groups and cycloaliphatic epoxides react very rapidly with cations derived from acid functionality or hydroxyl groups. Opening an epoxide ring results in linkage to the donating compound an formation of a secondary hydroxyl group that may undergo further reaction. Thus, the above described epoxide may be considered as tetrafunctional when it is in the presence of active hydrogen atoms and as a result will rapidly lead to a highly cross-linked, cured coating. Under the proper conditions, the formed hydroxyl group can react with another epoxide group or with a melamine-formaldehyde compound to produce coatings with improved properties such as hardness, toughness, chemical resistance, and/or solvent resistance in the cured film.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
500
PAINT AND COATING TESTING MANUAL
Q
URETHANES
Urethane linkages represent a rearrangement mechanism for the formation of cross-links and cure in polymeric systems. Urethanes are the reaction product of a hydroxyl group and an isocyanate group that undergo a rearrangement with the formation of a urethane linkage as described in the following.
This reaction is used in the manufacture of a broad range of products from insulation and seating foams, reaction injection moldings, in high-performance coatings, and other end uses. The isocyanate used in coatings and many other end uses is usually a di- or tri-functional compound that will lead to cross-linked systems. Monofunctional isocyanates are used to end-cap hydroxyl compound in the preparation of intermediates or to impart particular, desirable characteristics for certain end uses. Included are isocyanates such as iso-phorone diisocyanate, 4,4′-diphenylmethane diisocyanate, 4,4′-dicyclohexylmethane diisocyanate, trimers of hexam-ethylene diisocyanate, and others. The higher molecular weight isocyanates are preferred when practical, since they have a lower vapor pressure than their lower molecular weight counterparts and thus provide for improved working conditions. As would be expected, saturated epoxides have better weathering characteristics than aromatic epoxide coatings. Isocyanates are used to cross-link a variety of polymers containing hydroxyl or amine functionality. In the case of amine functionality, the reactions are very rapid and result in the formation of urea cross-links. The reactions with hydroxyl functionality are usually catalyzed and proceed at easily controllable rates. Among the coating systems cured isocyanates or that contain urethane linkages are a variety of acrylic polymers and copolymers; polyesters; poly ether, polyester, and polylactone polyols; polyols capped with isocyanates and cured with adventitious moisture, urethane acrylates cured with ultraviolet or electron beam radiation, as well as others.
PHENOLICS
Phenolic compounds used in coatings are the condensation reaction products of phenol or substituted phenols and formaldehyde. Two main types of phenolics are made with the type depending on the catalyst used and the reaction conditions. When an acid is used as the catalyst, a thermoplastic, soluble novolac phenolic is obtained. Under basic conditions, the product obtained is a resole phenolic that is thermoset in character and is cross-linked in the final stages of the reaction. A phenol-formaldehyde condensate (P-F) may have the following structure:
15TH EDITION
Hydroxyl-containing oligomers such as the alkyds can be cross-linked through reaction of the hydroxy-methyl function of the P-F with an hydroxyl group on the alkyd with the elimination of water.
CURE MEASUREMENT
Complete or essentially complete cure development is a very important factor in meeting coating specifications and/ or in determining the ultimate performance characteristics of any particular coating. Therefore, it is very important for the developer and the user of coatings to know how the curing process takes place and whether or not the final coating is adequately cured. As cross-links develop and molecular weight increases in a coating, performance characteristics such as adhesion, solvent resistance, toughness, hardness, gloss, and the like approach an optimum value, which is near a set, desired, or specification value. Coatings that are undercured will tend to be soft (lack hardness), have poor solvent resistance, and probably have poor adhesion. Overcured coatings may be hard and brittle, take on a dark, undesirable color, have low gloss, or not meet other criteria. Cure measurement can be determined by simple, qualitative testing or through analytical techniques that require sophisticated instrumentation. Often simple testing can be used to give an investigator a quick benchmark and tell if they are on the correct developmental track. Then sophisticated techniques can be used to completely understand the coating system and the coating’s ultimate properties. Finally, investigators attempt to correlate simple and sophisticated tests so performance features may be confidently determined during production, which is often taking place at high line speeds, with relatively simple testing. Also important is being aware that coating properties may change after leaving a thermal curing oven, a radiation source, or other curing device and properties after a length of aging time should be considered. Methods for measuring degree of cross-linking and the distance between cross-links can be found in Chapter 46. Rapid test methods are desirable because of time constraints and cost, but sometimes they do not tell the true story. Yet, they do provide much useful information. It should be kept in mind that one is attempting to ascertain a molecular process, cross-linking (unless thermoplastics are involved), and its relationship to coating performance. Most thermoplastic coatings cure or “dry” by simple evaporation of solvent. This is true for nitrocellulose, cellulose acetate, and solvent soluble acrylic lacquers wherein no cross-linking takes place. To measure cure of such systems, the time for either a print-free or tack-free condition is determined. As the evaporation process takes place, the nature of the coating goes from tacky to set-to-touch to tack free. On further drying, the coating becomes print free wherein a thumb print will not be visible on the coating surface. At times the effect of a twisted thumb print under moderate hand pressure will be observed. Another testing method involves dropping cotton linters or cotton balls onto the surface and noting the time after which they do not adhere to the coating. Thermoplastic or thermoset coatings on metal substrates can be quickly tested with a hand-held, fairly new nickel (U.S.A. five-cent piece) by applying a quick, rapid, gouging scratch with the edge of the coin to the coating and observing the results. One quickly finds out if the coating is tough and adherent, since
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 39
Q
the hard, sharp coin edge will have a marked effect on the coating.
Solvent Rubs
A convenient, often-used method for determining crosslinking involves solvent rubs. A cloth soaked with a solvent such as methyl ethyl ketone (MEK) or acetone is rubbed vigorously back and forth on the film, and then the film is examined after a specific number of rubs. If the film is removed, softened, or loses gloss, it is considered inadequately cross-linked or it is considered nonresistant to solvent. Often, the number of back-and-forth double rubs that the film does pass is reported. Thermoplastic coating are not expected to pass the test. The tests involved include ASTM D5402, Practice for Assessing the Solvent Resistance of Organic Coatings Using Solvent Rubs, and ASTM D4752, Test Method for Measuring MEK Resistance of Ethyl Silicate (Inorganic) Zinc-Rich Primers by Solvent Rubs. Related tests include ASTM D5109, Test Methods for Copper-Clad Thermosetting Laminates for Printed Wiring Boards; ASTM D1676, Test Methods for Film-Insulated Magnet Wire; and ASTM D2671, Test Methods for HeatShrinkable Tubing for Electrical Use. One difficulty associated with this test is that the individual testing the coating is not standardized. One person may exert more pressure than another and this can have a bearing on the final result. However, even with this deficiency, the test does provide useful information when the number of double rubs passed is reported.
CURE: THE PROCESS AND ITS MEASUREMENT
501
sharpening the lead point, variances of lead hardness from different pencil lots, and different manufacturers. Yet, as mentioned above, the test is widely used and does provide useful information.
KNOOP AND PFUND HARDNESS
The Knoop hardness test, described in ASTM D1474 Test Methods for Indentation Hardness of Organic Coatings, is a useful test in which a spherical, pyramidal, or conical indentor is pressed into the film to cause indentation with the depth of penetration a measure of hardness. In the test, a 25 g load is applied for 18 s and the length of the indention line is measured with a microscope. The Knoop hardness number (KHN) is calculated from the expression [12]
Hardness
where L is the applied load in kilograms, Ap is the projected indentation area in square millimeters, l is the length in millimeters of the long diagonal of the indentation mark, and Cp is an instrument constant. Because of the thinness of most coating films, this method is more useful for plastics, since it required almost a 75 % indentation into the film and can be influenced by substrate considerations. A more generalized formula than the above can be found in the referenced ASTM Test Method. The Pfund Hardness Number (PFN) is obtained in an analogous manner by using a hemispherical quartz or sapphire indenter that is pressed into the film. PFN is calculated from the expression
PENCIL HARDNESS
where L is the applied load in kilograms, A is the area of the projected indentation in square millimeters, and d is the diameter of the projected indentation in millimeters.
Hardness and scratch resistance provide another method for obtaining information about the degree of cross-linking and cure. There are a variety of tests that provide information about this characteristic. The topic is discussed in detail in another chapter in this manual. A very simple, but complex, test for measuring the hardness of coatings is pencil hardness, which is carried out according to ASTM D3363, Test Method for Film Hardness by Pencil Hardness. The test originated with manufacturers of pencils who tried to develop a consistent means for checking the quality of pencils. Along the development way, someone suggested scratching a paint film, and it was noticed that the different hardness of pencils were able to penetrate the coating down to the substrate or to scratch the coating to different degrees. The test was later used by coating technologists and currently is widely used in the industry to determine coating hardness. In this technique, a number of pencils of known hardness, usually from a specified manufacturer, are used. The pencil is trimmed so that 5 mm of “lead” is exposed. The lead diameter is 1.8 mm and is sharpened by rubbing it perpendicular to the surface of No. 400 carbide abrasive paper. The sharpened pencil is held at a 45° angle to the coating and pushed along the surface with hand pressure to attempt peeling away the coating. The pencil that fails to scratch or cut through the coating is considered the “Pencil Hardness Value.” The method is simple, the equipment is low in cost, and results are quickly obtained. However, values obtained can be operator dependent as well as dependent on the method of
SWARD HARDNESS
Sward hardness depends on the change in surface properties and the viscoelastic nature of a film. This method, which was discontinued in 1990, is used for automotive finishes and consists of a rocker containing two spirit-bubble indicators. The method is described in ASTM D2134, Test Method for Determining the Hardness of Organic Coatings with a Sward-Type Rocker and in ASTM D4366, Test Methods for Hardness of Organic Coatings by Pendulum Damping Tests. The pendulum is rocked and the number of swings are recorded and a reading is taken at the point where the swing distance is equal to half the original value. As cure increases, the Sward number increases. The value of a Sward test is that a number that can be compared to that of other coatings is obtained. This test must be conducted under very clean conditions as lint and surface imperfection can interfere with the results. Temperature control is another important factor. The difficulty with this test is that it only measures cross-link density or cure to a point. Then, it cannot detect further cure or over cure, since a plot of Sward numbers as a function of time or cure tends to reach a limiting value. A comparison [13] of hardness values obtained by different test methods is shown in Table 1.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
502
PAINT AND COATING TESTING MANUAL
Q
TABLE 1—Hardness test comparisons [13] Pencil (Brands) Sample
KHN
Sward
A
B
C
D
E
1
3.09
24
5B
6B
5B
6B
4B
2
4.33
28
4B
6B
6B
6B
4B
3
2.77
24
5B
6B
5B
4B
3B
4
2.61
22
3B
4B
5B
4B
3B
5
5.81
38
2B
2B
2B
2B
HB
10
25.7
54
H
H
H
H
2H
12
39.1
40
3H
2H
2H
4H
3H
14
—
40
8H
9H
7H
7H
9H
a
Note how the values differ in this investigation—especially between the different brands of pencils and the limiting values of Sward Hardness.
a
IMPACT RESISTANCE
Tests that measure the impact resistance of a coating can be used to correlate with cure. An undercured coating may exhibit a low impact value, but as cross-link density increases, impact values improve. Of course, overcured coating may be brittle and have a low impact value. ASTM D2794, Test Method for Resistance of Organic Coatings to the Effects of Rapid Deformation (Impact), is a widely used test. It consists of placing a flat coated panel under a weighted spherical ball assembly and then dropping the weighted ball onto the panel from different heights. The cylinder in which the ball assembly is mounted is calibrated such that an operator can directly read the impact resistance in inch-pounds. The measurements are carried out by dropping the ball directly on the coated surface or on the reverse side with the results being reported as direct or reverse impact, respectively. The impact causes a dimple to form in the test panel, and it is examined visually or with a ten-power lens to determine the extent of cracking or other failure. This is a simple test that is widely recognized in the coatings industry, and it gives useful information about the performance characteristics of the coating. Related tests include ASTM D5420, Test Method for Impact Resistance of Flat, Rigid Plastic Specimen by Means of a Striker Impacted by a Falling Weight (Gardner Impact); ASTM D950, Test Method for Impact Strength of Adhesive Bonds; and ASTM D2137, Test Methods for Brittleness Point of Flexible Polymers and Coated Fabrics. In addition to these methods, certain industries, such as the automotive industry, often develop their own in-house test methods and these will vary from manufacturer to manufacturer.
Thermal Analysis
Thermal analysis is an important analytical tool for determining the response of materials to changes in temperature and is useful in determining cure of coatings. The technique can be used to monitor the glass transition temperature, Tg, of a coating and this property can be related to cross-link density. For example, if a number of paint films that are cured at various times and/or temperatures are evaluated, thermal analysis can be used to determine the change in
15TH EDITION
Tg as a function of a bake schedule or, for that matter, of formulation composition. From such studies, optimum cure cycles can be determined to ensure quality finishes are produced. Thermal analysis devices have several modes for determining Tg and these include Differential Scanning Calorimetry (DSC), Thermomechanical Analysis (TMA), and Dynamic Mechanical Analysis (DMA). The glass transition temperature is a second-order transition, and it represents a temperature where segmental motion occurs in a polymer chain. This motion also causes a volume change in the polymer. Comparative changes of cross-linking measured by Sward hardness, evaporative rate analysis, and DSC indicated that the changes can be followed to high levels of cross-linking, but Sward measurements reached a limiting value while Tg continued to increase [14]. The importance of using the correct window of time and temperature when curing powder coatings has been described in the literature [15]. When DSC is used, a plot of endotherm/exotherm as a function of temperature is obtained at a set rate of temperature increase. As Tg is reached, there is an abrupt change in the endotherm that appears on the plot. Similarly, Tg can be measured using a plot of change in expansion as a function of temperature with the TMA method. In this method, a specimen is placed under a quartz probe and then heated. As the film expands, a plot is obtained in which a change in slope occurs. The intersection of the tangents of the expansion cure yields the Tg value. In the penetration mode, a depression occurs at Tg. Further details about thermal analysis can be found in the literature [16–19].
DYNAMIC MECHANICAL ANALYSIS
Another method for obtaining Tg is to use dynamic mechanical analysis in which a coating film is fixed to the ends of a tuning fork and the fork is driven over a frequency range and at different temperatures to provide a plot of damping as a function of temperature. Tg, being a second order transition, is a function of both temperature and of frequency. Information from both variables can be used to obtain information about the film [20].
TORSION PENDULUM
The torsion pendulum provides another way to obtain Tg as well as other dynamic properties such as the real and the imaginary shear modulus. A coating film is either imbibed into an inert substrate or used in a neat form (often difficult for this technique) and suspended between jaws and a counterbalanced disk that is used to apply a twisting torque to the specimen. The applied shear force then freely decays. A difficulty is that many coating films do not have the strength to support even the light stress of the counterbalanced disk, especially when the glass transition region is approached. This is why the coating is usually imbibed into an inert substrate such as fiberglass. If the sample is held at a constant temperature for a period of time, changes in cross-link density are reflected in modulus and damping changes. Details of the method and ways to construct a torsion pendulum can be found in the literature [21,22].
Impedance Measurements
Meyers [23] investigated the drying behavior of latex systems using ultrasonic impedance measurements. In his
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 39
Q
studies, a latex coating was cast onto a quartz crystal and ultrasonic energy was beamed at an eleven degree angle at the underside of the coating, reflected to a detector that measured the attenuation of the initial beam as absorbed by the drying coatings. As water evaporated, there were changes in impedance that could be correlated to curing in the sense of drying.
EVAPORATIVE RATE ANALYSIS
A novel method for studying cure is Evaporative Rate Analysis (ERA), which was developed by Anderson and coworkers [24]. This methodology was applied to determine the cleanliness of the surface of spacecraft. It was reasoned that if a radioactive C14 liquid were placed on a clean surface, the evaporation would be retarded due to cleanliness. In fact, the opposite was true. Nevertheless, this technique was developed to measure the degree of cleanliness. If a small amount of diethyl succinate-C14 is deposited on a surface and a controlled sweep of nitrogen is applied over the surface, the rate of evaporation can be monitored using a Geiger counter. The principle is that on a clean surface, there is no interaction between the solvent and a clean substrate and as a result normal evaporation takes place. If, however, the surface contains a contaminant, then the radioactive liquid will interact with the surface contaminant and retard the rate of evaporation of the radioactive component. Plots that relate cleanliness to evaporation rate can be prepared. The investigators applied this technique to the cure of organic coatings. They demonstrated that in an undercured surface, solvent retention increased and this led to longer residence times for the radioactive substance on the under-cured surface. By preparing coated panels having different cure times or temperatures, and then measuring the ERA of each panel, a plot of retention as a function of bake schedule was obtained. The method was applied to the cure of a variety of coatings by many coating technologists [14,24]. Other instruments are useful for determining the cross-linking of coatings including the Vibrating Needle Curometer that is used for liquid systems such as adhesive and sealants. The instrument drives a vibrating needle at a fixed frequency into the sealant or adhesive, and the damping amplitude is measured as a function of gel and cure time [25]. Products that can be monitored for cure with this technique include coatings, adhesive, foams, and sealants.
Recent Information
Readers are referred to other chapters in this manual for related cure-measuring techniques. The information provided in this chapter is also applicable to radiation-cured and powder coatings, since the performance of these coatings depends on proper extent of cross-linking consistent with the design of the polymers used [5,15]. In addition there have been recent advances in analytical techniques that provide coating chemists with better tools to understand the process of cross-linking [26]. Analytical techniques for measuring key polymer properties are reviewed in polymer chemistry texts [27– 29]. A recent application of space technology in which a technique known as Near Infrared Thermal Processing is being used to cure a wide range of coatings, printing inks, and adhesives [30]. The technique involves high
CURE: THE PROCESS AND ITS MEASUREMENT
503
energy sources that radiate near the infrared spectrum (more than 90 % of the energy emitted is below 2 μm) and allow markedly higher rates of cure than infrared or hot air systems. The systems have high process efficiencies and reliability. The importance of cure and its characterization is apparent from the recent April 2006 two-part, virtual-learning course offered about this topic by the Federation of Societies for Coating Technology and presented by Dr. T. Provder.
References [1] ASTM D16-10, “Standard Terminology for Paint, Related Coatings, Materials, and Applications,” Annual Book of ASTM Standards, Vol. 06.01, ASTM International, West Conshohocken, PA, 2010. [2] Gotro, J., and Prime, R. B., “Thermosets,” Encyclopedia of Polymer Science and Technology, 3rd ed., J. Kroschwitz, ed., John Wiley & Sons, NY, 2004. [3] Neag, C. M., and Prime, R. B., “The Application of Time-Temperature Superposition Techniques to Powder Coating Cure,” J. Coat. Technol., Vol. 63, No. 797, 1990, p. 37. [4] Koleske, J. V., Springate, R., and Brezinski, D., “2006 Additives Guide,” Paint and Coating Industry, Vol. 22, No. 5, 2006, p. 40. [5] Koleske, J. V., “Radiation Curing of Coatings,” ASTM Manual 45, ASTM International, West Conshohocken, PA, 2002. [6] Wicks, Z. W., Jr., Film Formation, Federation Series on Coating Technology, Federation of Societies for Coatings Technology, Philadelphia, PA, June 1986, p. 19. [7] Athey, R. D., Jr., “Film Preparation for Coating Tests,” Paint and Coating Testing Manual, Chapter 37, ASTM Manual Series MNL. 17, 14th ed., J. V. Koleske, ed., ASTM, West Conshohocken, PA, 1995. [8] Hurley, R., “Metal Soaps: Drier Stabilizers and Related Compounds,” Handbook of Coating Additives, Marcel Dekker, Inc., New York, 1987, Chap. 13, p. 485. [9] Godbole, V. A., “Use of Metallic Driers in Organic Coatings,” Paint India, April 1986, p. 28. [10] Swaraj, P., Surface Coatings: Science and Technology, 2nd ed., John Wiley & Sons, New York, 1996, pp. 452–453. [11] Narayan, R., Chattopadhyay, D. K., Sreedhar, B., and Raju, K. V. S. N., “Cure, Viscoelastic and Mechanical Properties of Hydroxylated Polyester Melamine High Solids Coatings,” J. Mater. Sci., Vol. 37, No. 22, 2002, pp. 4911–4918. [12] Swaraj, P., Surface Coatings: Science and Technology, 2nd ed., John Wiley & Sons, New York, 1996, p. 485. [13] Sato, K., Prog. Org. Coat., Vol. 8, No. 1, 1980, p. 12. [14] Miranda, T. J., J. Paint Technol., Vol. 43, No. 553, 1971, p. 51. [15] Anon., “EDC Industrial Quality Control Applications Series,” Guide to Product Cure Optimization, EDC, Inc., www.edc.com, 1998. [16] Seymour, R. B., and Carraher, C. E., Jr., Polymer Chemistry, An Introduction, 5th ed., Marcel Dekker, New York, 2000, p. 117. [17] Stevens, M. P., Polymer Chemistry, An Introduction, 3rd ed., Oxford University press, 1999, p. 169. [18] Miranda, T. J., Mechanical Behavior of Materials, Volume III, The Society of Materials Science, Japan, 1972, p. 392. [19] Lambourne, R., Ed., Paint and Surface Coatings, Ellis Horwood Ltd., 1987, p. 607. [20] Nielsen, L. E., Mechanical Properties of Polymers, Reinhold Publish Corp., New York, 1962, p. 274. [21] Allcock, H. R., and Lampe, F. W., Contemporary Polymer Chemistry, 2nd ed., Prentice Hall, Englewood, NJ, 1990, p. 427. [22] Myers, R. R., J. Polym. Set, Part C: Polym. Symp., Vol. 35, No. 3, 1971. [23] Anderson, J. L., Root, D. E., and Green, G., J. Paint Technol., Vol. 40, No. 320, 1968. [24] Rossi, A. G., and Paolini, A., J. Paint Technol., Vol. 40, No. 328, 1968.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
504
PAINT AND COATING TESTING MANUAL
Q
[25] Scott, K. W., Adhesive Age, Vol. 34, No. 11, 1991, p. 22. [26] Taylor, J. W., and Winnik, J., JCT Res., Vol. 1, No. 3, 2004, p. 163. [27] Fried, J. R., Polymer Science and Technology, 2nd ed., Prentice Hall Professional Technical Reference, Upper Saddle River, NJ, 2003.
15TH EDITION
[28] Allcock, H. R., Lampe, F. W., and Mark, J. E., Contemporary Polymer Chemistry, 3rd ed., Pearson Education, Inc., Upper Saddle River, NJ, 2003. [29] Carraher, C. E., Jr., Polymer Chemistry, An Introduction, 6th ed., Marcel Dekker, 2003. [30] Bar, K., “One, two…cured!,” http://www.beckers-bic.com, Oct. 2006.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
40
MNL17-EB/Jan. 2012
Film Preparation for Coating Tests Robert D. Athey, Jr.1
THE PERFORMANCE OF A COATING FILM IN A TEST is likely to be dependent on the physical form of the film. For instance, film thickness is an important factor in physical and appearance measurements (until the coating gets too thick), so there must be some control of the film thickness. Appearance is also related to how smooth the film surface is, and care to make the film appropriately will ensure that the appearance measurements are germane to the end use. The substrate used as the carrier for the test film, even temporarily, may affect the property measurements, as well. The primary concern in making films for tests is that the film prepared be homogeneous and consistent with previous or future films for the same test. The jargon of the trade calls the art of making films “casting,” but many film formation methods are used to form the film, and each method has its own set of advantages and drawbacks.
TEST REQUIREMENTS OF FILMS
Test requirements come in two classes, appearance (aesthetics) and physical properties. Choice of application method for the film may affect the appearance in some cases. For instance, the “multicolor” paints must be applied by dip or spray techniques for laboratory testing or the appearance does not give the desired mottle of nearly circular spots. Roller, brush, or even drawdown bar will cause these spots to become streaks. Choice of application method for appearance includes drying technique, as some films require special drying conditions to attain their desired special appearance. Examples include leafing aluminum flake “metallic”-look paints, hammer tones, and wrinkle finishes. Choice of application method driven by physical property testing is also necessary. Think carefully of the three kinds of stresses (tensile, compression, shear) and recognize that all are blended in a hardness or adhesion evaluation by indenter, pencil, mandrel bend, or dart impact test on a coating. Recognizing these stress combinations can make one aware of film preparation needs. In cases where adhesion is stronger than cohesion of the film, one should be able to distinguish between adhesive failure and cohesive failure. Film preparation should not conflict with these objectives. The physical properties to be tested in specific ASTM methods have a “significance and use” section in the method to ascribe the relation of the test value to some “in use” performance criterion. The preparation of the test film should thus correspond to the standard field application of the material, as well. 1
Another portion of film influence on the test is film thickness. The presentation by the Technical Committee from Toronto in the 1990 National FSCT Meeting and Paint Show [1] dealt with hardness measurements in films of varying thickness. However, the control of film thickness was not as straightforward as might have been thought. Hardness values turned out to be dependent on film thickness along with other equally important variables [2]. Fluid rheology, concentration, and other factors govern the amount of coating left after the application process. Knowing the process variables associated with a filmcasting technique are essential to getting what is needed for the ultimate test and results therefrom. Certain other properties of the film (opacity, permeability, erosion rate by some attacking mode, etc.) will depend on the thickness of the film. So, understanding thickness control in the casting process is crucial to obtaining meaningful results in terms of meeting specifications in reproducibility. In instances where fluid or gas permeation resistance or corrosion barrier properties are to be tested, a “pinhole,” “holiday,” or “mud-crack” in the film is a fatal flaw. The film preparation technique must eliminate (as much as possible) any such fatal flaw, or the test method should prescribe what is to be done (for example, replications) in cases where an unseen fatal flaw is detected by the test.
FILM CASTING TECHNIQUES Free Films
Free films, that is, films not applied permanently to a substrate, are used for a wide variety of tests and accordingly vary in thickness and size. The most common castings of free films are used for permeation or strength and elongation testing. They can also be used for cold flex tests or moisture/ solvent absorption studies related to permeation. Free film thickness can be measured as described in ASTM D1005, “Test Method for Measurement of Dry-Film Thickness of Organic Coatings Using Micrometers.” Probably the most common free film castings are laid down on a “nonstick” surface, such as the silicone-coated papers (available through Leneta and other suppliers), Teflon™, or polyethylene sheets. Release substrates are described in ASTM D823, Standard Practices for Producing Films of Uniform Thickness of Paint, Varnish, and Related Products on Test Panels. Lack of release-substrate wetting is the most common problem encountered when trying to obtain a continuous film. Water-borne coatings are especially difficult to formulate for good wetting of these substrates. One trick to use with such formulations
Athey Technologies, P.O. Drawer 7, El Cerrito, CA 94530-0007.
505 Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed byASTM International This standard is for EDUCATIONAL USE ONLY. Copyright © 2012 by www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
506
PAINT AND COATING TESTING MANUAL
Q
Fig. 1—Vacuum plate pulling a vacuum through the side nozzle holds paper or paperboard substrates tight for drawdown film casting. (Courtesy of Paul N. Gardner Co.)
is to prepare the sample at higher than normal solids content, which results in increased viscosity and resistance to beading is improved. An alternative is to apply several coats until the dried “beaded-up” portions merge to form a continuous film. A vacuum plate makes it easier to hold down the paper or polymer film for film casting and is an aid to good results. Fig. 1 is a diagram of the vacuum plate formed from aluminum. Older versions had eyes at the bottom surface through which screws could affix the plate to a tabletop. If this system is to be vacuum supplied from a sink faucet aspirator, a liquid trap (vacuum side-fitted Erlenmeyer flask) should be placed in line between the vacuum plate and the aspirator. This will ensure that any liquid coming from the aspirator will not get to the vacuum plate or coating substrate. Grenko [3] suggested several coated paper techniques for obtaining free films, such as decal paper coated on the face side or outdated matte or semimatte photopaper. The decal paper could be floated on water and the paper peeled off. Grenkodid not describe the removal of photopapers from the film. He also cited Sager [4] for a technique using an embroidery hoop for holding nonmoisture proof cellophane. After the coating film had been applied, one simply poured water on the inverted cellophane and the coating film dropped off. These techniques are likely to be most appropriate for solvent-borne or 100 % reactive formulations. Water-borne coatings may be inappropriately leached or weakened by the water used to loosen and isolate the films. Grenko [3] also mentioned the use of metal substrates for free film isolation. Brightwell [5] used glass onto which a thin silver film had been vacuum deposited. Bayor and Kempf [6] used thin aluminum foil that was later dissolved from most of the back by using hydrochloric acid doped with a minute amount of platinum chloride. The latter dissolved off only enough aluminum foil to leave a foil “frame” at the edges of the film to support it and to facilitate handling. This technique with aluminum foil dissolution in an acidic, aqueous medium presumes no sensitivity of the coating components to acid or water. For thicker films (1 cm or more), a Pyrex™ TV or Teflon™ coated baking pan can be used with some success. For instance, about 300 mL of a latex formulation in a 9 by 12 in. (23 to 30 cm) Pyrex™ baking dish will yield a sufficient quantity of thick film for testing. Occasionally, there is a “mud cracking” problem, but it can be reduced periodically
15TH EDITION
by releasing the edges of the film from the vertical sides of the dish with a spatula to allow uninhibited shrinkage. This way usually results in pieces of increased size, and 1 in.2 (6.25 cm2) is all that is required for the moisture absorption or solvent-swell test. Using this technique, it is possible to obtain large enough pieces of test coating for moisture vapor transmission tests at an application thickness that is common for roof coatings. The “mud cracking” problem is serious for any coating. The cause is the lack of “wet gel strength” to resist the tensile stresses within the shrinking membrane. Strictly speaking, unless the coating cross-links, it is not a gel, but simply a high-viscosity fluid or “pseudogel.” Assuring that the upper surface does not “skin over” by controlling the humidity or solvent-vapor environment over the drying surface may alleviate the problem. The literature describes mercury pool casting [7,8] or tin foil casting [9] with mercury amalgamation used to remove the tin without stressing the coating film. ASTM D4708, “Standard Practice for Preparation of Uniform Free Films of Organic Coatings,” also describes the tin foil amalgamation technique. These techniques are no longer in common use, as the mercury vapor is deemed a health hazard and caution is advised if the technique is attempted. Rarely, but occasionally, for instance, a free film of a polymer or binder material is needed for testing water vapor permeation or absorption. Some of these tend to be tacky, and powder applied to the surface will make them easier to handle. Although pearl cornstarch is preferred over the powdered talc for this purpose, baby powders and similar talcum powders are often used. Rotational casting is a means of obtaining nonporous, non-mud cracked-free films. The standard jar-mill roller (see Fig. 2) may be used with a variety of open containers (quart or gallon jars, earless paint cans) with a 4-in. (10cm) circle cut out of the bottom, to centrifugally cast the film with good thickness control. An alternative device, the Caframo REAX2 rotating mixer [10] may be used to hold the paint can with its main axis along the rotation axis of the clamp (see Fig. 3). The Los Angeles Society for Coating Technology Technical Committee successfully used this rotational casting technique in a study to obtain tensile and moisture vapor permeation film samples from acrylic latex compounded with silane-modified talcs [11]. It is derived from an older
Fig. 2—Lab roller mill—Although this device is most often used for pebble milling, an empty “earless” paint can on the rollers may be used to cast films. (Courtesy of Indco Inc.)
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 40
Fig. 3—Heidolph “REAX 2”—This device was designed as a rotational mixer for fluids, but a gallon paint can will fit in it for rotational film casting, as well. (Courtesy of Heidolph USA.)
practice at the Thiokol Rubber Co. that was used to obtain films of castable elastomers in a rolling pipe [12]. Equipment for this technique is easy to set up and use. Introduce 100 mL of coating into a 1-gal can and let it roll overnight. This will result in a film approximately 6 by 18 in. (15 by 46 cm). That is a sufficient quantity for 6-in. (15-cm) tensile bars and 2-in. (5-cm) permeation circles. Close control of film thickness is achieved by knowing the area of the cylinder and concentration. From these, the weight of coating needed can be calculated. Films of tacky coatings or pressure-sensitive adhesive formulations can be made by this technique. After the film is formed, the can is placed in a refrigerator or freezer to facilitate film removal. To obtain a centrifugally cast powder coating film, a sample of the powder is introduced into the can. Then, a hot air gun is held at the outside of the rolling can, and the powder coating will melt to coat the inside. Care is needed to eliminate scorching, and practice is needed to determine how close the flame should be to the can to get a good film. Other techniques to obtain a film include melting on a Teflon™ sheet on a hot plate or in between Teflon sheets in a heated press. Spraying one to a silicone-coated paper is occasionally used to make free films. There is no problem with this technique when a skilled practitioner does it. However, films made with this technique may not give as good permeation test results as from one made by a film from a fluid-flow technique. A note of caution should be added about handling free films made for physical testing. Any undue stresses may chip, stress, or crack the film. This is especially important for permeation or physical property tests. In the case of the latter, some materials are “notch sensitive,” that is, the results of a tensile or other physical test may be markedly reduced by the presence of a notch, scratch, or crack in the film [13]. Dry Coatings on Substrates Test coatings on a substrate are generally applied for a wide variety of appearance and physical property tests. Color and opacity are commonly evaluated on coated paper charts, though unsealed paper charts with a porous surface are available, as well. Coating on substrates to simulate actual use is common so the paint laboratory has many kinds
Q
FILM PREPARATION FOR COATING TESTS
507
of steel, aluminum, wood, or other substrates in stock as needed by the customer or the paint company. Plastic panels are rarer, but they can be made available when needed. Several companies now offer such products. Plastic panels may need special surface preparation techniques to assure wetting and adhesion (solvent washes, chemical etches with alkali or acid, low-temperature plasma, or corona discharge or chemical oxidation, etc.). As painting plastics become more and more important for automotive, computer, sign, and other industries, the manufacturers are making panels available to potential paint suppliers. Panels of metal are a special case for the preparation of test films for paint. It may be necessary to do a precoating surface preparation on the metal, and consultation with appropriate specifying agencies may be necessary. The Society for Protective Coatings (SSPC) has many grades of hot-rolled steel preparation techniques that vary from wire brushing through sand blasting to expose “white metal” [14]. Other preparations such as phosphatizing, galvanizing, anodizing, etc., are unique to specific industries or industry segments and will require lab preparation of panels or freshly treated panels from the customer or a commercial source. An annual handbook contains descriptions of some of these techniques for laboratory usage [15]. Preordered steel panels that are subsequently stored in special corrosion-inhibitor-treated packaging may skew exposure tests if they are not rigorously cleaned of corrosion inhibitor just prior to the coating steps. Cleaning procedures may require rinsing with hydrocarbon solvent or acetone and hot air drying just before paints are applied. Exposure panels made of metal must also have special edge protection to ensure that only the coated face has the exposure test applied. The casting technique on a paper or paperboard substrate is a skill that is learned through apprenticeship. First, the panel needs to be clean and dust free. For horizontal castings, which is a way to obtain good flat films, it is necessary to ensure that the surface is level during casting and drying. Tape the chart or panel to the level surface so it does not curl as the film loses solvent or dispersing media and starts to shrink. Unsealed paper charts will curl in a convex manner since any water in the coating swells the paper fibers, and the substrate may try to curl the sheet in a concave manner when the film tries to shrink across the top surface from which evaporation is taking place. Vacuum plates are available, as noted earlier, to hold down the paper, foil, or cardboard chart during the film-casting process. The paperboard chart should be stored in the same humidity environment that will be used in drying/curing the coating as the wood-pulp fibers are hygroscopic. Concrete, plaster, and asphalt/gravel test blocks or panels may be prepared in the laboratory or they may be purchased. However, whenever possible such blocks should be obtained from the customer since paint-laboratory results may not compare well with the customer’s results if details such as porosity, surface salinity, surface pH, soluble calcium or sodium ion content, etc., are not equivalent. These materials may be hygroscopic and will need equilibration in the drying/curing environment prior to coating. In addition, some concrete castings, that is, tilt-up wall constructions, often have surface residues of form coatings that had been used to aid release from the forms. Surface
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
508
PAINT AND COATING TESTING MANUAL
Q
characterization for such release aids or rigorous cleaning may be needed. There is an ASTM standard practice for preparation of mortar panels for testing paints: ASTM D1734, “Standard Practice for Making Cementitious Panels for Testing Coatings.” Grenko [3] has described press coating of bitumens on aluminum or felts, with shims placed in the press to attain appropriate thickness. His work was based on earlier studies by Greenfield [16].
Wet Films for Testing
There are a variety of reasons one may want to have a close look at a wet coating film. Simply watching the drying process may make any problems with moisture condensation on the film, pigment float, etc., apparent. There are test films made to assess wet hiding or sag resistance (ASTM D4400, “Standard Test Method for Sag Resistance of Paints Using a Multinotch Applicator”). Good film casting practice, for example, a level surface for casting, consistent technique, etc., is a requirement for reproducibility of results. Equipment required for special tests will be included in the following section.
EQUIPMENT FOR FILM PREPARATION Drawdown Bars
The drawdown bars are the simplest of the film casting devices to use. One simply makes a puddle of coating on the desired substrate and moves the drawdown bar over the puddle to make the desired wet film thickness. Application of the film with a drawdown bar nominally gives a good continuous film, but the spacing between the bar and the substrate may have little to do with the film thickness obtained. Speed of pulling the drawdown bar over the fluid will affect the film thickness since fluid rheology influences film thickness. High molecular-weight polymers may cause the edges of the film to draw in and increase the wet-film thickness. The formulation may not wet the drawdown bar, and, if this happens, it adds another factor to the control of film thickness. Wetz and coworkers [17] showed almost 100 % deposition of the wet film and found a relationship between wet-film and dry-film thickness. Tdf = 0.057 + 0.588GS(l /f )
15TH EDITION
or Tdf = 0.057 + 0.955Twf (l /f ) where: Tdf = dry film thickness, mils, Twf = wet film thickness, mils, ρl = density of liquid, ρf = density of dry film, G = clearance of drawdown barmils, and S = weight fraction of solids in liquid. Grenko [3] makes the point that these relationships are only true for drawdown blades having substantial flat surface between the leading edge of the bar and the trailing edge. The Byk-Gardner information [18] suggests the following expectations for cast films of varying thickness: 15 to 100 μm
50 %
100 to 300 μm
60 %
300 to 500 μm
80 %
over 500 μm
90 %
The hand-held drawdown bars are covered in an ASTM specification, Method E in ASTM D832, “Standard Practice for Rubber Conditioning for Low Temperature Testing.” A wide variety of drawdown bars are commercially available, and your local machine shop can make special orders of any design, if needed. Table 1 lists a few of the available types of drawdown bars, with commercial sources of some suppliers. Some rectangular design devices have differing gap depths on the sides so one may choose the film thickness needed for casting (see Figs. 4 and 5). Stainless steel or aluminum are the preferred materials of construction, as corrosion can damage the region of the drawbar controlling thickness of applied film. Good laboratory practice dictates immediate cleaning of the paint contact surfaces after every usage to minimize the threat of corrosion or other damage. A caution on marking the draw down bars, as some manufacturers label them not with the gap spacing, but a number half the gap spacing since that is the expected wet film thickness to be obtained.
TABLE 1—Drawdown bars for film casting Type
Design Details
Suppliera
Reference
“Bird” knife
Permanent fixed gap
Byk-Gardner
3
Two-path applicator
Bar has two cuts of different depth machined into top and bottom
P. N. Gardner
20
Eight-path applicator
Stainless-steel square tube with eight different depth cuts in each of top and bottom edges
P. N. Gardner
20
Gardner “Microm” applicator (also Hercules-Gardner adjustable [3])
Micrometers on each side lower or raise blade of “U” shaped device
P. N. Gardner
20
“Universal” blade applicator
Blade forms “U” shape with sides having slots and thumb screws to raise or lower
P. N. Gardner
20
Dow latex film applicator
“U”-shaped blade with wider cut for larger gap so second cast film
Byk-Gardner
3
a
There may be many other suppliers, but only one is cited herein as a space-savings technique.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 40
Fig. 4—Applicator frame/step gap applicator—Two film thickness choices are available with this device. (Courtesy of Byk-Gardner Inc.)
Some drawdown bars have film thickness adjustment choices by micrometers or other techniques (see Table 1 and Fig. 6). When these are used, the settings should be checked with feeler gages to make certain user wear has not made them inaccurate. These also require dismantling and cleaning after every use. If oil is used to inhibit corrosion and make the mechanics move smoothly, make sure they are rigorously cleaned of oil prior to usage. Several drawdown bars have gradations of cuts into the wet-film thickness controlling surfaces. For instance, the Erichsen suppliers offer the Kruse Multi Clearance applicator, having six to ten adjacent film strips of 10 to 200 μm in thickness for assessment of color yield, opacity, etc. (see Fig. 7) [19]. A similar device, the Leneta TG19 Logicator, is intended for hiding power and spreading rate measurements. It is available from Gardner [20] and has eight “gates” ranging from 2.65 to 10.4 mil in depth. There are motorized film application devices specified in Method C of ASTM D823, “Standard Practices for Producing Films of Uniform Thickness of Paint, Varnish, and Related Products on Test Panels.” The Model 509/1 Film Applicator from Erichsen [19] may be fitted with any sort of drawdown bar or blade (see Fig. 8). Speed of the motion of the applicator may be preset, so the variation among several samples may be minimized. Two devices Grenko [3] included with drawdown bars are more like bulk coating devices, as they do not require putting down a puddle of coating before applying the blade. Indeed, these devices contain the coating and apply it to a stack of sheets, one sheet at a time as the sheets are pulled out from underneath the device. One such device is the Parks Film-O-Graph cylinder. One puts the cylinder on a stack of Leneta charts or metal panels, fills the cylinder with the coating, and pulls each sheet from underneath the cylinder. A portion of the rim edge of the cylinder is milled
Fig. 5—Multiple clearance applicator—Eight film thickness choices are available on one device. (Courtesy of Byk-Gardner Inc.)
Q
FILM PREPARATION FOR COATING TESTS
509
Fig. 6—Fllm casting knife—Micrometers adjust the blade clearance. (Courtesy of Byk-Gardner Inc.)
out to act as the gate to allow coating at a desired thickness. One may even use it for very viscous coatings (even putty) by forcing the fluid through the tube with a loose-fitting plug from the top and wiping the bottom of the tube over the substrate to be coated. The second of the devices discussed was the Parks Rapid coater. Paper sheets were stacked in the bottom of a box with their ends out of a side slot at the bottom. The box was then filled with coating fluid, and each sheet was withdrawn individually. Grenko [3] noted that this was not a precision device, but it was adequate for some purposes. Grenko [3] also described the flat Parks Film-O-Graph, which used a flat plate with spring clips to hold shims along the sides of the sheet of substrate. One poured the coating onto the substrate between the shims and used a bar such as a ruler or other straight edge to doctor off the excess and make the coating as thick as the shims. Bending of the scraper bar by the very slightest amount would make very thin films almost impossible, but for thick films (for example, roof coatings), this technique worked well. Some applicators are meant only for the wet film tests. Two especially useful applicators for these are the sag test devices and the wet hiding test devices. The latter, most frequently the Pfund Cryptometers (available from Erichsen [19] and Byk-Gardner [20]), have black or white glass (ceramic) beds over which a transparent wedged cover levels the coating film (see Fig. 9). Although these devices are quite hard, care in cleaning will assure they do not get scratched. They must be stored carefully, wrapped, and covered. The sag test film casters in effect drawdown several narrow films in gradations of film thickness. Many are simply slotted U-shaped drawdown bars with variations in slot depth from 1 to 6 mil, 3 to 12 mil, and 14 to 60 mil.
Fig. 7—Multiple gap drawdown bar—Six or eight gaps are machined along the same edge for casting side-by-side films for comparison. (Courtesy of Erichsen GMBH & Co.)
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
510
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 10—New York paint club leveling test blade—The slotted blades are for assessment of sag and leveling. (Courtesy of Byk-Gardner Inc.)
Wire-Wound Rods
Fig. 8—Electrically driven drawdown device—Four speeds may be chosen to control drawdown rate. (Courtesy of Erichsen GMBH & Co.)
However, the New York Paint Club Technical Committee designed a special “leveling test blade” that had double slots for each depth with a wide space between each set of double slots [21]. Their slots are 1, 2, 4, 8 and 16 mil in depth (see Fig. 10). How a paint film drips or sags depends on the rheological characteristics, and the thickness of the coating that first sags may be related to the gravity-applied shear stress by the outer elements furthest away from the substrate on that film. Other rheological factors (high shear viscosity, ca. 10 000 s−1, for instance) are also important in sag and leveling. Sag test devices complying with ASTM D3730, “Standard Guide for Testing High-Performance Interior Architectural Wall Coatings,” and several federal specifications as well as the New York leveling test blades are available from Gardner [20] and Byk-Gardner [18].
Fig. 9—Pfund crytometer—The white style shown is for black/ dark paints, while a similar black style is used for white/light paints. (Courtesy of Byk-Gardner Inc.)
For very thin films, the wire-wound rods are quite useful for quickly and easily applying films. The wire-wound rods are simple rods of 12- or 16-in. (30- to 40-cm) length, of 1/4, 3/8, or 1/2 in. (0.63, 0.45, or 1.27 cm) diameter, with a spiral wind of wire tight about 75 % to 80 % of the rod. Table 2 shows a selection of the wet film thickness that is obtained from the rods with different wire diameter. The film laid down with a wire-wound rod is almost exactly a tenth the thickness of the wire winding. Strictly speaking, as the film has the rod depart, there are ridges in the film in the direction of travel of the rod. However, these collapse to make a quite smooth film unless there is rheological inhibition in the formulation or the coating is very fast drying. This technique is quite effective to simulate paper or can coating end products. There are industrial production systems that employ wire-wound rods to meter coatings onto roll substrates, as well.
TABLE 2—Selected coating thickness obtained from various wire-wound rods Wire
Wet Film Thickness
Size
Diameter (in.)
Mils
2.5
0.0025
0.25
6.4
3
0.003
0.3
7.6
3.5
0.0035
0.35
8.9
4
0.004
0.4
10.2
4.5
0.0045
0.45
11.4
5
0.005
0.5
12.7
5.5
0.0055
0.55
14.0
6
0.006
0.6
15.2
10
0.01
1.0
25.4
20
0.02
2.0
50.8
30
0.03
3.0
76.2
40
0.04
4.0
101.6
50
0.05
5.0
127.0
60
0.06
6.0
152.4
70
0.07
7.0
177.8
80
0.08
8.0
203.2
90
0.09
9.0
228.6
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
Micrometers
CHAPTER 40
Q
FILM PREPARATION FOR COATING TESTS
511
Fig. 11—Spiral wire drawdown applicator—The handle shown grips the end of the wire-wound rod for application with one hand. (Courtesy of Erichsen GMBH & Co.)
A similar device called the “spiral-film applicator” is available from Erichsen [19]. It is a wire-wound rod with a perpendicular handle (see Fig. 11). It is available to apply a film thickness of from 10 to 200 μm in widths of 80, 150, or 220 mm. Byk-Gardner offers a handle that attaches to the ends of the wire-wound rod for lab film preparation as well [18]. The Accu-Lab™ Lab Drawdown Machine has a rod holder and substrate holder to assure precision in laboratory manual film preparation [22]. There are also two-wire rods in which a second, smaller wire is wound in the grooves made by the first winding wire. These yield higher application rates than do the single-wound wire rods. There is potentially an advantage to the two-wire rods since they leave a different pattern in the coating. The ridges left are larger and may level faster and more uniformly.
Fig. 12—Single setting lab spray applicator. (Courtesy of Erichsen GMBH & Co.)
and others offer automated spray applicators for laboratory sample preparation (see Figs. 12 and 13). The spray technique is particularly important, but precautions are necessary to assure evenness of coating on the substrate. The spray pattern should extend beyond
Spray Outs
Spray application, including by weight on a panel hanging on a balance in a spray booth, can be done with good precision. Good spray-out application on panels can be done by those skilled in the art. To control the weight applied in a spray booth, a triple beam balance may be attached to a beam 5 or 6 ft (1.5 or 1.8 m) above the floor, and a hangerwire panel holder is hung from it as the spray target. A cardboard shield over the balance will keep off overspray off the positioned balance. Take off a few tenths of a gram to give a pointer indicator warning when the spray has almost reached the desired weight. Spray quickly to minimize evaporation. There are automated spray devices. Grenko [3] related the automated spray device design from Bell Labs [23] to Method A of ASTM D823. The Bell Labs design has the gun travel over stationary panels, while an alternative design has the panels on a moving belt under a stationary spray gun, similar to a design reported from Battelle [24]. In either case, the amount of coating applied to the panel is controlled by the speed of the moving belt and by nonvolatile content of the sprayed fluid. Grenko [3] notes the sample panels on the moving belt may be held to the belt with magnets or suction cups. Erichsen [19], Gardner [20],
Fig. 13—Programmable lab spray applicator. (Courtesy of Erichsen GMBH & Co.)
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
512
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
the edges of the substrate when the nozzle is aimed at the center of the target. The spray pattern—be it flat, fan, or circular—has less paint at the edges than at the center. Moving the spray nozzle across the target or the target under the center of the nozzle makes the deposition more likely to be consistent in thickness.
Dip Coating
Dip coating is particularly important in some special cases. For instance, the edge protection of corrodible panels in preparation for the application of a test coating is a common practice. Printed circuit board coatings, including conformal coatings and baseball bats, are other areas that employ this technology. A red lead vinyl 1/2-in. (1.27-cm) dip was successfully used for edge protection of exposure panels, since attack corroded only the desired exposure surfaces [25]. The point is that dipping assures film thickness without holidays and other techniques do not provide such assurance. This is particularly true for odd-shaped test specimens such as fragments of a pressurized gas cylinder and items as mentioned above. Grenko [3] briefly reviewed two-dip coaters. Bruins [26] first designed a dip coater using a tire pump and needle valve to control the rate of panel withdrawal from the coating, while Payne designed an electric motor-driven device adopted for Method B of ASTM D823. A commercial laboratory device is available from Gardner [20], with variable withdrawal speeds of 2 to 20 in. (~5to ~50 cm) per minute. It can dip a panel of up to 2 lb (0.9 kg) and 1 ft2 (0.3 m2) in area—12 by 12 in. (30.5 by 30.5 cm). Again, care must taken when dip coating. Film thickness control is a battle between wetting surface forces and the shear forces of drainage through the thickness of the film. Drip edges on the bottom edge may be avoided to some degree by having the panel holder inverted to hold the bottom edge upward for a portion (or intermittent portions) of the drying period.
Spin Coating
Grenko [3] described the work of Walker and Thompson [27] attaching a panel to a turntable and rotating for 1 min at 300 rpm to obtain a 25-μm varnish film thickness. He also described the Sward-Gardner [28] relation of film thickness (F) to viscosity (V in poise) and nonvolatiles, % N, as F = 0.4N + V 4 + 3 where the spin rate was 290 rpm for 60 s. Their work had a precision of 5% to 10%, and corrections for time or rpm variations were offered. Parker and Siddle [29] suggested modifying the method by adjusting viscosity to equivalence for all fluids to be compared and using volume solids rather than weight percent nonvolatiles. Plots of film thickness versus volume percent nonvolatile were straight lines for nonhixotropic fluids, but curvature existed for thixotropic fluids. A commercial lab spin coating device is currently available from Erichsen [19]. There are two versions, one with 600-rpm speed set, and another that is adjustable from 50 to 2000 rpm (see Fig. 14).
Fig. 14—Lab spin coating device. (Courtesy of Erichsen GMBH & Co.)
OTHER TIPS ON PRACTICE OF THE ART
Dust is always a problem, especially in formulation laboratories that have pigment dusting in the lab and plant. Cover cast films immediately to keep the dust off. The easiest cover is the top of a box or similar device. Simply cut out 1/2 in. (or ~1 cm) strip from two, three, or four sides for free flow of air. Of course, a similar cover can be made from thin plywood. Keep the cover on top of the lab refrigerator or bookshelf so it is always at hand. Make sure the film is drying under appropriate temperature and humidity conditions. Something in the formulation may respond in an adverse manner to condensing moisture that can form droplets on the surface as evaporating solvent rapidly cools the system. In moisture-cure and reactive two-package urethanes, the condensing moisture may react with the isocyanates to modify degree of cure, which can reduce strength, solvent resistance, or other properties. In systems that do not react with the condensing water, one may still get pits, pinholes, or haziness from the condensing water droplets on the surface. Paint applicators are learning to pay attention to humidity variation and its effects on the end product surface.
CONCLUSION
How the film is prepared for testing can have a dramatic effect on the test results. The thought put into film preparation prior to preparation and the care used in film formation can be crucial factors in obtaining meaningful and reproducible results.
References [1] Toronto Society for Coatings Technology Technical Committee presentation, 1990 National Paint Show, Voss/APJ Competition presentation, available from the Federation of Societies for Coatings Technology, 492 Norristown Rd., Blue Bell, PA 19422–2350. [2] Athey, R. D., Jr., “Coating Tests-Hardness of the Film,” European Coatings Journal, Vol. 92, No. 10, December 1992, p. 461.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 40
[3] Grenko, C., “Preparation of Films for Test,” ASTM STP 500, Paint Testing Manual, G. G. Sward, Ed., ASTM International, West Conshohocken, PA. [4] Sager, T. P., “The Preparation of Thin Films,” Ind. Eng. Chem. Anal. Ed., Vol. 9, 1937, p. 156. [5] Brightwell, E. P., “An Optical Method for Measuring Film Thickness of Paint Films,” Official Digest, Federation of Paint and Varnish Clubs, Vol. 28, 1956, pp. 413–416. [6] Bayor, E. H., and Kempf, L., “Preparing of Vehicle Films Free of Supporting Foundation,” Ind. Eng. Chem. Anal. Ed., Vol. 9, 1937, p. 49; Bayor, E. H., and Kempf, L., “Preparing Fragile Paint and Varnish Films,” Ind. Eng. Chem. Anal. Ed., Vol. 10, 1938, p. 280. [7] Clarke, G. L., and Tschentke, H. L., “Physicochemical Studies on the Mechanism of Drying of Linseed Oil: 1. Changes in Density of Films,” Ind. Eng. Chem., Vol. 21, 1929, pp. 621–627. [8] Gloor, W. E., “Effect of Heat and Light on Nitrocellulose Films,” Ind. Eng. Chem., Vol. 23, 1931, pp. 980–982. [9] Long, J. S., Egge, W. S., and Wetterau, P. C., “Action of Heat and Blowing on Linseed and Perilla Oils and Glycerides Derived from Them,” Ind. Eng. Chem., Vol. 19, 1927, pp. 903–906. [10] Caframo Lab Products, P.O. Box 70, Wiarton, Ontario, Canada NOH 2TD (519-534–1080). [11] Athey, R. D., Jr., et al., “Latex Coating Formulation Evaluation of Organosilane Treated Talcs: A Statistically Designed Study Part II. Experiment Design and Test Results,” J. Water Borne Coat., Vol. 8, No. 2, May 1985, p. 10. [12] DePugh, C. C., private communication. [13] Takano, M., and Nielsen, L. E., “The Notch Sensitivity of Sensitive Materials,” J. Appl. Polym. Sci., Vol. 21, 1976, pp. 2193–2197. [14] SSPC, 2010, “Systems and Specifications,” SSPC Painting Manual, Vol. 2, The Society for Protective Coatings, Pittsburgh, PA.
Q
FILM PREPARATION FOR COATING TESTS
513
[15] Metal Finishing Guidebook and Directory Issue, M. Murphy, Ed., Elsevier, Hackensack, NJ, 2010. [16] Greenfield, S. H., “A Method of Preparing Uniform Films of Bituminous Materials,” ASTM Bulletin No. 193, ASTM International, West Conshohocken, PA, October 1953, p. 30. [17] Wetz, J. M., Golding, B., and Case, L. C., “Film Thickness Relationships of Organic Coatings,” Official Digest, Vol. 31, Federation of Paint and Varnish Production Clubs, 1959, p. 419. [18] Instruments Catalog, Section 8, Byk-Gardner Inc., Silver Spring, MD, 2011. [19] Erichsen, T. J. Bell, Akron, OH, 2010. [20] P. N. Gardner Co. Inc., Pompano Beach, FL, 2010. [21] Industry Tech, Oldsmar, FL, 2011. [22] New York Paint Club Technical Committee, Official Digest, Vol. 32, No. 430, 1960, pp. 1435–1438. [23] Arlt, H. G., “Paint Films of Controlled Thickness,” Bell Lab. Rec., Vol. XIV, 1936, p. 216. [24] Mueller, E. R., “A Simple Semi-automatic Laboratory Spraying Device,” Prod. Finish., Vol. 15, No. 2, 1950, pp. 36–38. [25] Golden Gate Society for Coatings Technology, “Corrosion Inhibitive Performance of Some Commercial Water-Reducible Non-Toxic Primers,” J. Coat. Technol., Vol. 53, No. 682, 1981, p. 29. [26] Bruins, P. F., “Production of Uniform Test Films of Shellac and Other Finishes,” Ind. Eng. Chem. Anal. Ed., Vol. 9, 1937, pp. 376–378. [27] Walker, P. H., and Thompson, J. G., “Some Physical Properties of Paints,” Proceedings, ASTM International, West Conshohocken, PA, Vol. 22, Part II, 1922, p. 465. [28] Sward, G. G., and Gardner, H. A., “Uniform Varnish Films for Exposure Tests,” Ind. Eng. Chem., Vol. 19, 1927, pp. 972–974. [29] Parker, R. V., and Siddle, F. J., “The Hardness of Paint, Varnish and Lacquer Films,” Journal Oil and Colour Chemists Association, Vol. 21, 1938, p. 363.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
41
MNL17-EB/Jan. 2012
Measurement of Film Thickness John Fletcher1 and Joseph Walker2 PREFACE
IN PREPARATION OF THIS CHAPTER, THE CONTENTS of the 14th edition were drawn upon. The current edition will review and update the topics, introduce new technology that has been developed, and include up-to-date references.
INTRODUCTION
One of the major requirements of paint and coating testing is the measurement and control of film thickness. Ensuring that a thickness specification is achieved is important in laboratory tests, manual paint application, automatic spraying, and other coating application methods, be they solvent based or powder coatings. A number of test methods are available, and the choice is dependent on: (1) the location—laboratory or site; (2) the material coated—metal (ferrous or nonferrous), wood, plaster, brick, and plastic; (3) the condition of the coating—wet or dry, cured or uncured; and (4) the condition of the surface—rough or smooth, flat or shaped, thick or thin, etc.
WET FILM THICKNESS
The measurement of wet film thickness provides the first opportunity in the coating application process to check the coating thickness. It also offers an assessment of the spreading rate of paints. It is very important that wet film measurements are made as soon as the coating is applied to avoid error due to solvent loss during the curing process. Reference to the technical data for volume solids in the coating is required to establish the wet and dry ratio so that wet film thickness values can be converted to dry film equivalents. Wet film gages are manufactured from a range of materials, aluminum, stainless steel, etc., and in most cases, wet film thickness gages can be cleaned with solvents and reused. The main exception to this are the notched gages (wet film combs) molded from plastic. Although the materials used, such as loaded ABS, are resistant to many solvents, it is recommended that plastic wet film combs be used only once. Wet film measurement can be carried out on most substrate materials, both metal and nonmetal. However, the substrate must be rigid and not prone to distortion when the measurement device is applied. ASTM D1212, Standard Test Methods for Measurement of Wet Film Thickness of Organic Coatings, describes the Inmont wet film gage (formally known as Interchemical 1 2
gage and commonly known as the Wet Film Wheel) and the Pfund wet film gage. These two gages are detailed in the following sections, as Inmont Wet Film Gage (ASTM D1212 method A) and Pfund Wet Film Gage (ASTM D1212 method B). ASTM D4414, Practice for Measurement of Wet Film Thickness by Notched Gages, describes a third wet film measurement method using notch gages, which are also known as wet film combs. This method is described in the section called Notch Gage.
Inmont Wet Film Gage (Wet Film Wheel)
The Inmont wet film gage consists of two concentric outer disks with an inner eccentric disk and a smaller diameter positioned between them as shown in Fig. 1. The outer disks are scaled, with the clearance between the inner disk and the outer disk from zero to a maximum value as shown. The gage is used by placing the disks with the maximum clearance on the specimen of wet coating and rolled toward the minimum clearance on the gage in either direction. The paint will coat the inner disk until the clearance is greater than the wet film thickness. The point at which the coating stops on the inner wheel is referenced to the scale on the outer wheel and the value noted. Starting from maximum value avoids the possibility of pushing paint ahead of the inner disk, creating an error condition when the gage indicates a value higher than the true wet film thickness. A number of range options are available for the wet film wheel and wheels with notched outer rims are particularly useful for measuring wet films in coil coating, as the notches help to ensure even rotation on the fast-moving substrate.
Pfund Wet Film Gage
As shown in Fig. 2, this gage consists of a convex lens, L, whose radius of curvature is 250 mm, at the lower end of tube T1 that slides in the outer tube T2. Compression springs, S, keep the lens out of contact with the paint film until pressure is applied on tube T1. In making a measurement, the gage is placed on the painted surface and the lens is pushed slowly through the film until stopped by the substrate. The pressure is released, and the diameter of the spot of paint transferred to the lens is measured. A 1-to-1 ratio for the thickness added by the paint displaced by the lens to the actual thickness has been assumed and is accounted for in the equation t = D2/16R
Technical Support Manager Elcometer Limited, UK. VP Sales & Marketing, Elcometer Inc., Rochester Hills, MI.
514
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 This standard is for EDUCATIONAL USE ONLY. www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
Downloaded/printed byASTM International Copyright © 2012 by
(1)
CHAPTER 41
Q
MEASUREMENT OF FILM THICKNESS
515
Fig. 1—Inmont gage (Interchemical) wet film wheel schematic diagram (left) and photograph (right).
where D = diameter, mm of spot and R = radius of curvature mm of the lens. Table 1 gives film thickness and the corresponding spreading rate in square feet per gallon for spots from 3 to 38 mm in diameter. It has been observed that a substantial proportion of paints do not obey the 1-to-1 relationship. The actual thickness, obtained by independent methods, may be several times, or only a fraction of, the thickness calculated by the equation. A small amount of thinner added to a paint may increase the diameter of the spot on the lens and give a corresponding increase in the calculated thickness. This phenomenon has been ascribed to the effects of surface tension. Hence, for best results, a correction factor should be established for each type of paint based on the known thickness of a freshly prepared film measured by the Inmont gage. Reproducibility is within about 2 % for films 2 mils (50 μm) thick, decreasing to about 10 % for films 5 mils (125 μm) thick, and then becoming better as thickness increases.
Notch Gage (Wet Film Comb)
Notch gages (wet film combs) are formed on the edge of a strip of material so that each notch has a different clearance from the reference shoulders to that of its neighbors (Fig. 3). Many different materials are used to make these gages, such as stainless steel, aluminum, and plastic, in a variety of shapes: square, rectangle, triangle, hexagon, etc. The Notch gage is a simple low-cost device, which is useful when approximate values of wet film thickness are satisfactory, as the notches have discrete values and are not continuous across the range of thickness. The gage is dipped vertically into the film until the reference shoulders are resting firmly on the substrate. The thickness of the film is between the highest coated notch or tooth face and the next highest, uncoated, value. Several different methods of manufacture of these gages exist from spark-eroded stainless steel precision combs, through punched aluminum sheet, to plastic flow molded combs. The stainless steel combs can be certified with measurements of the tooth displacements, which are
Fig. 2—Pfund gage schematic (left) and photograph (right). Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
516
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
TABLE 1—Spreading rate by Pfund film gage Diameter of the spot in mm
Thickness of coat (mm)
Painted Coat (m2/litre)
Painted Coat (sq.ft./US Gallon)
Painted Coat (sq.ft./imp. Gallon)
3
0.00225
445
18.088
21.800
4
0.00400
250
10.175
12.250
5
0.00625
160
6.512
7.840
6
0.00900
111
4.522
5.440
7
0.01225
81.6
3.321
4.000
8
0.01600
62.5
2.543
3.340
9
0.02025
49.5
2.009
2.425
10
0.02500
40
1.628
1.960
11
0.03025
33
1.350
1.640
12
0.03600
28
1.130
1.372
13
0.04225
24
963
1.176
14
0.04900
20
830
980
15
0.05625
18
723
880
16
0.06400
15.6
636
765
17
0.07225
13.8
563
676
18
0.08100
12.3
502
603
19
0.09025
11
450
548
20
0.10000
10
407
490
21
0.11025
9.07
369
443
22
0.12100
8.26
336
405
23
0.13225
7.56
307
370
24
0.14400
6.95
282
340
25
0.15625
6.40
260
313
26
0.16900
5.92
241
290
27
0.18225
5.49
223
268
28
0.19600
5.10
207
250
29
0.21025
4.76
192
233
30
0.22500
4.44
180
218
31
0.24025
4.16
169
204
32
0.25600
3.90
158
192
33
0.27225
3.67
149
180
34
0.28900
3.46
141
169
35
0.30625
3.26
133
160
36
0.32400
3.08
125
152
37
0.34725
2.82
117
138
38
0.36100
2.77
113
129
traceable to national standards. As the stainless steel is hard wearing, this certificate can be valid over a period of up to one year. On the other hand, plastic combs, although manufactured from solvent-resistant ABS plastic, should only be used once as the solvent in the coating may soften
the plastic. Plastic combs can be tagged and kept as a permanent record of wet film measurement. Aluminum combs are prone to wear, and the condition of an aluminum wet film comb should be carefully checked before use.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 41
Q
MEASUREMENT OF FILM THICKNESS
517
Fig. 3—Wet film comb schematic (left) and photograph (right).
Wet film gages are available covering many ranges, from 0.5 mil (2.5 μm) to 160 mils (4 mm/4000 μm). The teeth are normally square ended, but for thicker coatings pointed teeth are sometimes used. Notch gages are supplied by various coating equipment suppliers, e.g., Elcometer Inc., Paul N Gardner, etc. and are fully described in ASTM D4414, Practice for Measurement of Wet Film Thickness of Organic Coatings by Notched Gages. ASTM D4414 describes methods; Method A uses a square or rectangular notch gage made from a thin rigid material, and Method B uses a circular notch gage made from a thin rigid material.
Needle Micrometer
This method was used to study the relationship between the clearance of a doctor blade and the thickness of the wet film left by the blade [1]. A needle is attached vertically to the objective holder of a microscope. The barrel is lowered until the needle just touches the film that has been spread on a plain metal panel. The contact is observed through a horizontal microscope. When contact is made, the needle and its image reflected by the film just meet. The needle is then lowered into the film until it just touches the metal panel. This contact is noted by the deflection of a galvanometer in series with the panel, a dry cell, and the needle. The thickness of the film is calculated from the number of turns made by the focusing screw of the vertical microscope between the two points of contact. It should be noted that this technique is only applicable to measurements made in a laboratory, as it is impractical for work on site. The equipment is no longer available but may still be in use in certain laboratories.
DRY FILM THICKNESS (DESTRUCTIVE METHODS)
As there are many circumstances under which coatings and paints are used with many different materials as the substrate, no single method of dry film thickness measurement is universal. Some methods are destructive, and these are most often used when nondestructive methods
cannot be applied. The following sections describe the use of the Micrometers and Dial Gages, the Gardner Needle Thickness Gage, the Gardner Carboloy Drill Thickness Gage, and the Gardner Gage Stand and the Gardner Micro-Depth Gage.
Micrometers and Dial Gages
When a chip or flake of coating is freed from the surface of the coated object, its thickness can be measured directly using a micrometer. Alternatively, the total thickness of the substrate and coating can be measured; the substrate can then be re-measured after removing the coating with a scraper or solvent. The coating thickness is then the difference between the two measurements. This “before and after” coating measurement is also the basis of the method for manufacturing coated thickness standards. ASTM D1005, Test Method for Measurement of DryFilm Thickness of Organic Coatings using Micrometers, describes four methods as follows: Procedure A—Stationary micrometer for measuring coatings applied to plane rigid surfaces. Procedure B—Stationary micrometer for measuring free films. Procedure C—Hand-held micrometer for measuring coatings applied to plane rigid surfaces. Procedure D—Hand-held micrometer for measuring free films. Procedures A and C require the thickness of the test panel and film to be measured and then the film is removed to allow just the thickness of the test panel to be measured. The difference between the two thickness values is then calculated and reported as the film thickness. It is also possible to measure the uncoated test panel in selected and identified locations before the coating is applied and the combined test panel and dry film measured together at the locations used for the initial measurements. Procedures B and C require the preparation of free films as outlined in ASTM D2370 and detailed in ASTM D823.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
518
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 4—Dial micrometer.
Procedure D, using a hand-held micrometer mounted on a support with a clamp to hold the specimen (see Fig. 4), is recommended only for thicknesses over 0.5 mil and is accurate to ±0.1 mil. A hand-held dial coating thickness gage (Fig. 5), may be used to measure coated samples. In this case the dial records the difference in position between the foot that sits on the surface of the coating and the stylus with its ball end, which passes through a hole prepared in the coating to the surface of the substrate. This is particularly useful for site work. However, for best accuracy and precision, the mounted dial gage specified in ASTM Method D1005 is preferred.
Gardner Needle Thickness Gage
This instrument is designed to measure the thickness of electrically nonconducting films on metal (conducting) substrates. It is small enough to be used in the field where the substrate can be made part of the electric circuit. The needle makes only a minute puncture in the film. In many instances, particularly in a go-no-go determination, the damage is so slight that the method may be considered nondestructive for many end users. The aluminum housing case contains the needle screws for forcing the needle through the film and a lamp to signal when the needle contacts the substrate. For use in the field and for occasional use in the laboratory, the electric circuit comprises the needle, the substrate, a dry cell, the lamp, and a cord that connects with the substrate. In the laboratory, if many measurements are to be made, it is advisable to use a step-down transformer and to connect to a 110 V source. The thickness is read on a dial attached to the screw. One turn of the dial raises or lowers the needle by 2 mils. The dial is graduated in steps of 0.05 mil. Range is 0 to 15 mils.
Fig. 5—Dial type coating thickness gage.
The zero setting of the needle is obtained by retracting the needle within the housing, placing the gage on a plane metal block, and lowering the needle until the lamp signals contact. The specimen replaces the block and the needle is lowered until the lamp again signals contact. The difference between the two readings is the thickness of the film.
Gardner Carboloy Drill Thickness Gage
In many instances films are so hard that they successfully resist penetration by the needle penetrometer described above. This difficulty has been overcome by substituting a Carboloy drill for the needle. The drill is a needle terminating in a pyramid having three faces. The drill is secured in a chuck that can be rotated and advanced independently. The rotation is controlled by finger action on a knob at the upper end of the chuck shaft. In all other respects, the operation is the same as that of the needle gage. This method is not now widely used.
Gardner Gage Stand
Although the Gardner needle gage and the Carboloy drill gage may be operated by manually holding the gage against the specimen, using the Gardner gage stand is less tiring and more accurate, especially when many measurements are to be made. The stand provides constant known pressure [up to 10 lb (5kg)] on the specimen and ensures that the needle or drill is always perpendicular to the specimen.
Gardner Micro-Depth Gage
Although in outward appearance this gage resembles the gages described in the last two subsections, only the
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 41
Q
MEASUREMENT OF FILM THICKNESS
519
establishing of the zero setting is the same. Measurement is not restricted to nonmetallic films on metal—any type of film on any type of substrate may be measured, and the film is always damaged. In this gage, a chisel replaces the needle of the gage. The zero setting having been established, the chisel is advanced by an amount estimated to be less than the thickness of the film. The gage is placed on the specimen and drawn toward the operator through a distance of a few millimeters. If the scratch made by the chisel does not penetrate the film, the chisel is advanced by a small increment, and another scratch is made. The procedure is repeated until the substrate is reached and exposed. Inspection is best made with the aid of a low-power magnifier. The range is 0 to 40 mils (1000 μm). Repeatability depends on the magnitudes of the increments and the compressibility of the film and substrate. For example, specimens with poor adhesion may be torn off, exposing the base, even if the chisel does not penetrate the film.
recorded. The procedure is repeated on line BC. The difference multiplied by the calibration factor equals the thickness of the film.
Microscope for Film Thickness CLASSICAL METHOD
ASTM D4138, Standard Test Methods for Measurement of Dry Film Thickness of Protective Coating Systems by Destructive Means, describes three methods as follows: Test method A: Using grove cutting instruments Test method B: Using grinding instruments Test method C: Using drill bit instruments The Tooke Inspection gage [4], also known as the Paint Inspection Gage (Fig. 6), provides for estimating the thickness of a film from the geometry of a V-groove cut in the film by a special tool. With the aid of a ×50 illuminated magnifier equipped with a reticle in the eyepiece, the operator measures the lateral distance from the top edge of the cut and the projection of the intersection of the cut and the substrate. To make a measurement, a “bench mark” of ink applied to the surface of the film serves to make the top edge of the cut readily visible. A short cut is then made at a right angle to the benchmark with the selected cutting tip. Film thickness is then obtained by counting the scale divisions as described previously.
To use a microscope to assess film thickness, a section is prepared and the width of the coating is measured using a graticule in the eyepiece of the microscope. For an approximate assessment, a flake of the coating can be used, but for best results from this method the specimen should be prepared as follows. The specimen is mounted in a block of wax. The face of the mount is cut or ground to a smooth surface. The prepared specimen is then inspected under the microscope.
BRIGHTWELL METHOD
This method does not require removal of a chip and elaborate mounting and preparation [2]. A tiny furrow is made in the film or a small chip is removed. A prism or ribbon of light is projected on the selected area at an angle of 45°. The distortion of the beam is examined with a microscope equipped with a micrometer eyepiece. Apparatus for this is available in the Schmaltz optical surface analyzer (Carl Zeiss). The apparatus is calibrated by measuring known depths milled in a smooth metal block. The ribbon of light is focused on line A, and the filar micrometer reading is
STOPPED METHOD
A cut is made in the film with a sharp knife. The microscope is focused, in turn, on the upper and lower edges of the cut. The thickness of the film is computed from the vertical adjustments of the microscope [3]. If the value is not known, it may be found as follows: Put a piece of plate glass on the stage. Lower the tube until it just touches the plate. Record the reading of the fine adjustment. Now raise the tube as far as possible and again record the fine adjustment. Now raise the tube as far as possible and again record the fine adjustment. The distance of the tube from the plate divided by the number of turns of the adjusting screw gives the value for each turn.
TOOKE INSPECTION GAGE (PAINT INSPECTION GAGE)
SÄBERG DRILL
ASTM D4138 Test method C is similar to the method described in the previous subsection; however, a circular drill is used to penetrate the film. The hole can then be
Fig. 6—Paint Inspection Gage (Tooke gage) with section through coating and view through the microscope. Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
520
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 7—Säberg Drill.
inspected using the ×50 magnifier with graticule, and the width of the cut from the outer edge to the print where the drill penetrates to the substrate is a measure of the coating thickness. For the Säberg Drill illustrated in Fig. 7, the calculation of coating thickness is as follows: 1. For measurement in micrometers, multiply graduations by 20.0. 2. For measurement in mils, multiply graduations by 0.79.
DRY FILM THICKNESS (NONDESTRUCTIVE METHODS)
There are two main groups of gages for the nondestructive measurement of dry films on metal substrates and these are the permanent magnet gages, which can be used for coatings on magnetic metal substrates and electronic gages, which can be used for coatings on ferro-magnetic substrates, such as steel, nonferrous metals and where ultrasonic measurement methods are used on some nonmetal substrates. The following sections describe the Permanent Magnet Thickness Gages, the Electronic Thickness Gages with some of their features and limitations, and Ultrasonic Gages.
Permanent Magnet Thickness Gages
Permanent magnet coating thickness gages can be used to determine the thickness of films applied to magnetic substrates such as steel, iron, magnetic stainless steel, etc., providing that the coating is nonmagnetic. Materials such as nickel and cobalt, which are naturally magnetic, have to be treated with care, while paints containing magnetic particles, such as some ferrous micaceous iron oxide, can cause errors when using magnetic gages, both mechanical and electronic. Simple magnetic coating thickness gages, also known as pull-off gages, use the principle that the attractive force between a permanent magnet and the magnetic metal substrate is inversely proportional to the distance between them. The principal limitations are (1) the film must be sufficiently hard to prevent indentation, and (2) the film must
Fig. 8—Magne-Gage. (Courtesy of Magne-Gage Sales & Service Co., Inc.)
not be tacky, causing the magnet to be held by the surface of the coating. Electronic magnetic coating thickness gages are also available, but these will be described in a separate section entitled Electromagnetic Thickness Gages; this section includes the electronic gage based on the magneto-resistor (Hall-effect) probe.
MAGNE-GAGE
This instrument (Fig. 8) consists of a small permanent bar magnet, 2 mm in diameter, suspended from a horizontal lever arm [5]. The arm is actuated through a spiral spring by turning a dial. The tip of the magnet is brought into contact with the paint film (on iron or steel), and the dial is then turned until the magnet is detached. The attractive force between the magnet and the film support is indicated on the dial, and the thickness of the nonmagnetic paint film is obtained from a calibration curve relating thickness to dial reading. The Magne-gage can be used to measure coatings on convex and concave surfaces as well as on flat ones provided the radius of the curvature is not too small. Unless special calibrations are made, cylinders should not be less than 1/2 in. (1.27 cm), in diameter spheres not less than 3/4 in. (1.9 cm), and flat pieces should be at least 3/4 in. (1.9 cm) square. Magnets for thicknesses in the following ranges are available: 0.0 to 0.002, 0.002 to 0.007, and 0.007 to 0.025 in. Within ASTM D7091 (Standard Practice for Nondestructive Measurement of Dry Film Thickness of Nonmagnetic Coatings Applied to Ferrous Metals and Nonmagnetic, Non-conductive Coatings Applied to Non-Ferrous Metals) the Magne-gage is described as a Type 1 or magnetic pull-off gage. It should be noted that the Magne-gage is not now in common use but the apparatus is still available in the USA.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 41
Fig. 9—Magnetic coating pull-off thickness gage.
HAND-HELD MAGNETIC PULL-OFF GAGES
Electronic gages are superseding the mechanical gages, and some—such as the Tinsley gage and the chemigage—are no longer manufactured. However, many thousands of these Type 1 gages are still in use, so it is appropriate to describe them. The simplest form of these gages contains a magnet suspended from a coil spring housed in a pen-style body manufactured from aluminum or plastic. A scale is drawn on the body and a marker is used to indicate the extension of the spring on the scale. The reading on the scale when the magnet lifts off the surface corresponds to the thickness of the coating. The scale on these instruments is nonlinear, leading to low reading resolution at the maximum range, usually 20 mils (500 μm), and great care must be taken to ensure that they are used vertically to avoid the influence of gravity on the spring-magnet combination. Some of these types of gages have design features to overcome the gravity effect when the gage is used horizontally. The accuracy of this type of gage is typically ±15 % of the readings and therefore these gages do not satisfy the accuracy requirement of ASTM D7091. A common form of the hand-held magnetic pull-off gage is shown in Fig. 9. A balanced beam with a magnet fitted to one end and counterbalanced by a brass weight at the other is attached at the pivot to a helical spring. The other end of the spring is attached to a ring holding the scale. Rotation of the ring raises or lowers the magnet. This type of gage is described in ASTM D7091 as a Type 1 gage.
Q
MEASUREMENT OF FILM THICKNESS
521
The gage is placed on the surface to be tested in any orientation as the balance weight ensures that gravity effects are neutralized. The ring or scale wheel is pushed forward (anticlockwise rotation) to bring the magnet in contact with the coating and to set the scale to a maximum reading. The ring or scale wheel is then rotated clockwise until the magnet breaks free, the thickness of the coating being indicated by a pointer. This type of gage is calibrated against plated standards available from NIST, and accuracy against these standards of ±5 % of reading is achieved. It should be noted that the steel used to make NIST standards is not always representative of engineering steels, and therefore a practical operating accuracy of ±10 % is obtained in the field unless coated standards using steel that is typical of the substrate on which the coating is applied can be used to verify the accuracy of the gage for the specific application.
THE MAGNETIC RELUCTANCE GAGE
The coating thickness gage based on the magnetic permeability principle (Fig. 10) is one of the first magnetic thickness gages to be commercially available, having been patented in 1948. The magnetic flux from a permanent magnet acts on an armature suspended in between two magnetic arms of the unit, forming north and south poles. The turning moment of the magnetic flux is countered by a helical spring, and the magnitude of the magnetic flux changes with the distance between the tips of the magnetic arms (ball feet) and the substrate beneath the coating, i.e., the coating thickness. A pointer attached to the armature indicates the thickness of the coating. Gages covering thickness ranges from 0 to 3 mil (0 to 80 μm) to 0 to 0.75 in. (0 to 18 mm) are still available. The accuracy of this type of gage is typically ±10 % of the reading and these gages do not, therefore, meet the requirements of ASTM D7091. However, this type of gage is useful where electronic gages cannot be used, for example in places where flammable vapors are present or underwater.
Electronic Coating Thickness Gages
Electronic coating thickness gages use the electromagnetic induction principle or the Hall effect to measure nonmagnetic coatings on magnetic substrates or the eddy current principle to measure nonconductive coatings on non-ferrous metals. Gages can also make use of the Hall effect to measure coatings on magnetic substrates.
Fig. 10—Magnetic reluctance coating thickness gage. Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
522
PAINT AND COATING TESTING MANUAL
Q
ELECTROMAGNETIC INDUCTION PROBES
The electromagnetic induction method for measuring film thickness is based on the effect of a magnetic metal substrate on the balance of the alternating magnetic field in the probe tip, generated in the probe by a signal applied to the central coil that has two neighboring search coils. When the probe is away from the influence of the substrate (free-air condition), the net voltage output from the two search coils tends to zero. As the substrate is brought toward the tip, the field is increasingly out of balance between the two search coils until, with the uncoated substrate in contact with the tip, the net voltage output from the coils tends toward V max. The voltage output from the probe can be amplified and calibrated and then used to display a thickness value. Three types of electronic coating thickness gages have been developed on the basis of this probe technology using analogue, digital, and microprocessor electronics. ASTM D7091 describes these gages as Type 2 gages.
EDDY CURRENT PROBES
The eddy current method for measuring film thickness is used to measure coatings on nonferrous metals. It is based on the effect that a high-frequency alternating field (∼3 000 000 Hz or 3 MHz) has an electrically conductive surface causing highly localized current (eddy currents) to flow. These currents generate their own magnetic fields, which influence the impedance of the coil, generating the highfrequency field. The magnitude of these changes is proportional to the distance from the probe coil to the substrate, that is, to the thickness of the coating. This type of gage is also described as a Type 2 gage in ASTM D7091. Calibration adjustment by adjustment to zero and to a known thickness on a piece of metal of the same type, shape, and thickness as the samples to be measured is vital to ensure accuracy. Modern eddy current instruments are available using microprocessor designs and some of these designs offer the facility for a dual-purpose ferrous (F) electromagnetic induction and nonferrous (N) eddy current principle combined in a single probe. Using the dual FNF probe, the gage can be made to measure automatically on either type of substrate, with the gage displaying the substrate type, F or N, when the measurement is complete and the reading displayed. This type of gage can be used to measure paint galvanizing on a steel substrate. The F probe is used to measure the total thickness of the zinc and the paint on the steel and, if the zinc layer is 2 mil (50 μm) or greater, the N probe can be calibrated to measure the paint thickness only. By subtracting the paint thickness from the total thickness, the thickness of the zinc (galvanizing) can be estimated.
HALL EFFECT PROBES
This type of gage makes use of a probe that combines a permanent magnet with a Hall-effect semiconductor sensor a signal, which varies with the intensity of the magnetic field. The strength of the magnetic field is influenced by the distance of the magnetic substrate from the tip of the probe, i.e., the thickness of the coating. The scale of these gages is nonlinear and they use either an analogue meter movement or digital electronics to indicate the thickness. In operation it is necessary to set zero on the
15TH EDITION
uncoated metal and calibrate to a thickness of known value to obtain the best accuracy, as with other electronic gages. These gages meet the requirements of ASTM D7091 and are also described as Type 2 gages.
ANALOGUE COATING THICKNESS GAGES
The simplest form of electronic coating thickness gage uses a nonlinear scale on a meter movement to display the voltage output from the probe against a thickness scale. These gages are now relatively expensive to manufacture and have almost completely been replaced by digital gages, although some of these gages may still be in use in the field. In common with most electronic coating thickness gages, calibration adjustment to the substrate to be tested is required. Setting the needle to zero with the probe on the uncoated magnetic metal substrate and then setting the upper scale point, using the calibration control, with the probe on a known thickness of coating, will adjust the gage for use on surfaces typical of the uncoated metal substrate used. Two forms of thickness standard are in common use: (1) the pre-coated type, which can be the plated type, as illustrated by the standards available from NIST, or the hard-wearing epoxy coating applied to metal substrates of known thickness type, and (2) the measured and unmeasured plastic shims available commercially from gage manufacturers. In many applications the plastic shims are preferred as the thickness standards for calibration adjustment because they can be used as a sample of the uncoated substrate to be measured, reducing the errors of calibration adjustment due to surface finish, curvature, substrate composition, and thickness of the substrate. More details of these sources of error are given later in this chapter under the heading “EFFECTS OF SURFACE FINISH, CURVATURE AND SUBSTRATE COMPOSITION ON ELECTROMAGNETIC AND EDDY CURRENT MEASUREMENTS.”
DIGITAL COATING THICKNESS GAGES
Advances in electronic components and instrument design techniques over the last 30 to 40 years have made it possible to significantly reduce the size of coating thickness gages while retaining and enhancing the features users find necessary in their applications. The use of digital electronics means that the voltage output from the probe can be converted to a numerical value early in the processing of the signal, thus reducing the effects of temperature changes and component drift on the accuracy of the result. It is also possible to use a linear scaling, making it possible to have a fixed resolution over the full range of the instrument scale although resolutions are often enhanced in the 0 to 125 or 0 to 250 μm (0 to 5 mil or 0 to 10 mil) ranges. It should be noted that this type of product has been superseded by microprocessor-based designs, although some digital gages are still being used. The same principles of calibration adjustment apply to digital electromagnetic coating thickness gages as they apply to the analogue types, and accuracy capabilities of ±5 % of readings are readily achieved using digital gages in the field.
MICROPROCESSOR ELECTROMAGNETIC THICKNESS GAGES
The application of microprocessor electronics to the design of portable coating thickness gages has made it possible to improve the accuracy and reproducibility of these
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 41
Q
MEASUREMENT OF FILM THICKNESS
523
Fig. 13—Coating Thickness Gage with Hall Effect probe. (Courtesy of DeFelsko Corporation) Fig. 11—Microprocessor electromagnetic thickness gage.
instruments as well as allowing developments in range, special calibration techniques for rough surfaces, memory of readings, statistical calculation, printouts and data transfer. In a microprocessor design, the characteristic of the probe voltage output against coating thickness value is stored in the memory of the instrument for many values over the range of the probe. The actual voltage output from the probe is digitized and then compared with the stored values. A thickness value is then calculated from these data and displayed. This is typically achieved in 0.3 s. Using this technique, accuracies of ±3 % of reading are possible. A simple form of this gage is shown as Fig. 10. Note that this gage has no user adjustments and this simple operation is made possible by a factory calibration to the specific requirements of the automotive industry for which it was designed. As a microprocessor instrument is in effect a dedicated computer, many calculations can be performed on the data, and features such as correction for temperature changes, storage of calibration conditions and corrections to these calibrations, averages, and other statistical values can be included in the instrument’s firmware. Fig. 11 illustrates one of these microprocessor-based electromagnetic thickness gages that have memory-operated features such as a large reading memory, statistics, printout and data transfer.
Fig. 12 illustrates the current trend in microprocessor design with a dual microprocessor coating thickness gage. This unit has a microprocessor to control the gage, display the readings, perform the calculations, and provide data transfer with a second microprocessor, contained in the separate probe electronics, so that the probe can be individually characterized to thickness standards for optimum accuracy, ±1 % or ±1 μm, whichever is the greater, across the full thickness range of the probe. Fig. 13 illustrates a coating thickness gage with the Hall effect type probe. The use of microprocessor electronics also makes it possible for a single gage to make use of a wide variety of probes with different ranges, different formats, and different access capabilities. In general, the greater the thickness range of a probe the larger the physical size of the probe.
EFFECTS OF SURFACE FINISH, CURVATURE, AND SUBSTRATE COMPOSITION ON ELECTROMAGNETIC, EDDY CURRENT, AND HALL EFFECT MEASUREMENTS
The accuracy of coating thickness measurements carried out using the methods described above depends on the technique used in adjusting the gages for the conditions of the measurement. The major influences on the adjustment of these gages are surface finish, curvature, thickness, and material composition of the substrate. In the case of magnetic substrates the substrate material composition will influence the magnetic properties and hence the calibration of the gage. In the case of nonferrous metals the effect is that different materials will have different electrical conductivity, which will influence the generation of eddy currents and the resulting calibration adjustment.
Surface Finish Fig. 12—Dual microprocessor electromagnetic coating thickness gage.
A variety of surface finishes are to be found on metal to which a protective or decorative coating is to be applied. In some cases the coatings used require an anchor pattern
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
524
PAINT AND COATING TESTING MANUAL
Q
of profile depth, which forms a part of the specification. Comparators such as the Keane-Tator Surface Profile Comparator or the International Standards Organization ISO 8503 are used to determine the surface finish after shot or grit blasting. Also Testex Tape and the Elcometer Surface Profile Gage can be used to measure peak-to-valley heights of profiles. These instruments are shown in Fig. 14. Surface finish also influences the calibration adjustment of a coating thickness gage, as the mass of the metal directly beneath the probe is reduced by the effects of shot and grit blasting and the probe tip sits on the highest peaks of the profile. This has the effect of increasing the value of thickness indicated using a gage calibrated on a smooth surface by as much as 1.5 mils (35 μm) at 4 mils (100 μm) coating thickness for the highest values of profile. It is possible to use a rough surface calibration technique to eliminate this error and make the instrument read the correct value of coating thickness over the peaks by using the statistical power of the microprocessor type gages to calibrate on the profile. This is achieved by using a thin foil 1.0 mil (25 μm) over the profile to set the lower calibration point and a thicker foil 5.0 mils (125 μm) or 10.0 mils (250 μm) to set the upper value over the profile. The instrument will then indicate the thickness over the peaks for the coating between the values of foil chosen. This method is most accurate and reproducible when 15 to 20 readings are taken on each calibration foil to
15TH EDITION
establish a mean value and the mean value is then reset to the correct value of the foil. Trials have shown that the mean of 15 to 20 readings taken over an area of coating give a mean value within a few percent of the actual value over peaks determined by sectioning. The method does not, however, take into account situations where access to the substrate is not possible for calibration purposes. In this case, the method described in SSPC PA2 where a correction value is applied to readings taken using a smooth surface calibration in the instrument is more appropriate. This method is also described in EN ISO 19840. For either case it is important to agree with the method before measurements start to avoid discrepancies in reporting.
Curvature
The shape and metal wall thickness can also influence the accuracy of the calibration. The degree to which a particular gage is affected depends on the design of the probe. Most modern instruments exceed the limits identified in SSPC PA2. The effect of shape is most evident when taking readings on an uncoated sample. With an instrument calibrated on a smooth piece of metal 0.125 in. (3.175 mm) thick, changes of more than 0.2 mil (5 μm) in the reading at zero will be seen on curves with a radius below 0.12 in. (3 mm) convex or 0.96 in. (25 mm) concave on a typical
Fig. 14—Surface profile gages. Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 41
electromagnetic induction probe. Values will vary between manufacturers and from different probe types. This error can also be eliminated by calibration on a shape closely representing the specimen to be tested. However, it should be noted that once below the values of curvature indicated in the manufacturer’s literature, changes in curvature have a significant effect on calibration, i.e., the calibration on a shape will not be applicable to another shape.
Substrate Thickness
The minimum thickness of substrate required to provide stable readings will vary with the type and design of probe. Typically both F and N probes can be used on substrates down to 6 mil (300 μm). Special gage adjustment techniques can be used to measure on substrates thinner than this value but it should be noted that flexing of these thin substrates also changes the measurement condition and support from a solid nonmetallic base may be required. Measurement close to the edge of a substrate can also lead to errors. A good rule of thumb for electronic gages is that the area to be measured should be at least the same as the total area of the probe face. The performance of the gage and probe close to an edge can be determined by noting the change in the zero value on an uncoated sample as the probe is brought closer to the edge of the sample. The zero reading will start to increase as the edge effect comes into play.
Substrate Composition
In the case of electromagnetic induction probes, most are insensitive to the differences between the majority of steel specifications in general engineering use. However, when high-carbon steels are coated, the carbon content sufficiently alters the magnetic properties of the steel to cause the normal calibration curve applied within the instruments to be in error with respect to linearity. Thus, an instrument calibrated at zero and say 5 mils (125 μm) may have an error at 2 mils (50 μm) of more than 0.2 mil (50 μm), or 10 %. A similar effect can be seen with some cast irons. This error can be overcome by calibrating as for the rough surface described in the section earlier in this chapter entitled “Surface Finish.” For best accuracy, choose a foil just below the expected coating thickness value for the lower calibration points and a value well above the expected coating thickness value for the upper calibration point. When coatings applied to nonferrous metals are being measured using eddy current techniques, the composition of the substrate and its effect on electrical conductivity are the important factors with respect to calibration. Materials such as aluminum and copper have very similar characteristics and similar calibration values. However, zinc, brass, and other nonferrous metals and alloys have different characteristics, and calibration on an uncoated sample is essential. Differences of up to 2 mils (50 μm) can be seen between “zero” with an aluminum calibration and zero on a brass component. D7091 explains calibration, adjustment, and verification for dry film thickness gages. Calibration of a coating thickness gage is normally carried out by the manufacturer of the gage or an authorized laboratory using a documented procedure and thickness standards that are traceable to a
Q
MEASUREMENT OF FILM THICKNESS
525
National Metrology Institute. The frequency of such calibration is determined by the quality assurance procedure of the user, particularly when operating to a quality management system that is independently audited such as ISO 9001. The re-calibration period for gages within such a system will be specified and is typically 12 months. The adjustment of a dry film thickness gage is a procedure to minimize the errors caused by the conditions under which the measurements are to be made. These include the topics covered in the paragraphs above, surface finish, curvature, substrate thickness, and substrate composition but may also include issues such as the temperature of the substrate or the prevailing climatic conditions. The verification of a coating thickness gage should be carried out by the user before the gage is used to measure the work in hand by checking the accuracy of measurements on known thicknesses standards, preferably on a sample of the substrate that is typical for the work that is being coated. The gage must be verified for the range of thickness for which it is to be used and if the gage readings are outside the combined accuracy of the coating thickness standards and the manufacturer’s accuracy specification for the gage, then the gage should be sent for re-calibration and repair should that be required to restore the correct calibration.
STATISTICS IN FILM THICKNESS MEASUREMENT
As many random variations can be expected in a coating process, it is appropriate to classify the thickness of the coating using a statistical analysis. In fact, many national specifications utilize a statistical approach in recognition of these variations, e.g., SSPC PA2. The sources of these variations are many, and only a few examples can be cited here—operator error in taking the measurement, recording error, variation due to surface or curvature or composition, local variation in substrate due to local heat treatment or due to forming or working the metal, inclusions in the metal or in the coating, etc. The influence of these factors can be greatly reduced by taking a statistically significant number of readings for each area of the coating to be tested. This group of readings can then be summarized using mean and either standard deviation or range to show the average and the spread of readings about the average. A statistically significant number of readings would be 20 to 50; however, if the process is under statistical control as defined, five readings in each group or subgroup are sufficient. Many of the microprocessor-based coating thickness instruments are capable of calculating and recording mean values (X), standard deviation (σ), and highest and lowest values (range) within a batch of readings. It is important to establish the method of evaluating the information before embarking on an evaluation of a coating system so that the correct disciplines are applied to collecting the data and evaluating it for further decisions.
DATA MANAGEMENT
Some gages have memory and data transfer features such as RS 232 serial data transfer by cable or wireless communication via Bluetooth. These features used together allow readings for a project to be collected in a batch or batches in the gage and then this data can be transferred to a
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
526
PAINT AND COATING TESTING MANUAL
Q
computer running Data Management software for analysis, reporting, and archiving and fulfilling the requirements of Paperless Quality Assurance systems. One of the major benefits of this approach is the removal of any potential transcription errors that can arise when manual records of coating thickness readings are kept in an inspector’s notebook and are then transcribed into a report or a spreadsheet for further analysis. It is possible to transfer data directly from the memory of the coating thickness gage into a spreadsheet if a suitable function is included either in the gage or in the supporting software. The spreadsheet can then be used in the normal way to provide statistical values and charts. There are Data Management systems that also allow supporting data such as paint datasheets, inspection procedures, photographs and other project-related information to be associated with the specific gage data from an inspection. This then allows detailed reports to be prepared for clients in a semiautomatic fashion.
FILM THICKNESS MEASUREMENT USING ULTRASONICS
Thickness measurements of certain dry films can be made using ultrasonic energy, measuring the time of flight of a pulse through the coating and back from the coating/ substrate interface. ASTM D6132, Nondestructive Measurement of Dry Film Thickness of Applied Organic Coatings Using an Ultrasonic Gage, describes this method. The pulse echo layer thickness measurement technique using an ultrasonic transducer with a scan display is capable of measuring individual layer thickness in a multiple layer system but calibration is sensitive to the differences in these layers and a change of pigment may be sufficient to change the calibration conditions. A rigid substrate is required to obtain a good reflection of the ultrasonic energy and the coating must be able to transmit the sound energy without attenuation or dispersion. Coatings that include bubbles of air or that have certain types of fillers or pigments cannot be measured using this technique as they absorb the ultrasonic energy. Some substrates exhibit characteristics that tend to disperse the ultrasonic energy, as they are rough or are not sufficiently different to the material of the coating to adequately reflect the energy, e.g., certain coating on plastic substrates.
15TH EDITION
Fig. 15— Notch Gage for uncured powder thickness measurement—schematic diagram.
tion can be checked by measuring the powder before the cure using the ultrasonic gage and after the cure using a dry film coating thickness gage as described earlier in this chapter. This procedure is awaiting a round robin test to be carried out.
X-RAY FLUORESCENCE (XRF)
Recent developments in performance and reductions in cost have pushed X-ray fluorescence to center stage, particularly for metal coatings on metal substrate applications and smaller parts, such as fasteners. Beta-ray backscatter (BBS) techniques had been widely used to measure plated coatings; however, limitations in performance—e.g., a minimum of 20 % difference in atomic number between the coating and the substrate is required—mean that while gold over nickel, copper, or Kovar can be measured, nickel
UNCURED POWDER COATING THICKNESS TO PREDICT FINAL FILM THICKNESS
ASTM D7378, Standard Practice for Measurement of Thickness of Applied Coating Powders to Predict Cured Thickness, describes three procedures. Procedure A uses a rigid metal notched (comb) gage (see Fig. 15). Procedure B uses a special magnetic or eddy current probe, depending on the substrate material. The probe is designed to minimize the damage to the uncured coating and measures the actual coating thickness of the uncured powder film. For both Procedure A and Procedure B a correction value has to be applied to the uncured powder thickness readings to predict the cured coating thickness. Procedure C describes noncontact ultrasonic powder gages (see Fig. 16). These gages use a characterization of the reflected sound energy from the uncured coating to predict the final thickness of the coating after cure. Calibra-
Fig. 16—Noncontact ultrasonic coating thickness gage for uncured powder coatings.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 41
TABLE 2—Thickness ranges for some common coating materials using X-ray fluorescence [7] Coating
Substrate
Approximate thickness range (μm)
Aluminium
Copper
0–100.0
Copper
Aluminium
0–30.0
Copper
Iron
0–30.0
Copper
Plastic
0–30.0
Gold
Ceramics
0–8.0
Gold
Copper or nickel
0–8.0
Lead
Copper or nickel
0–15.0
Nickel
Aluminium
0–20.0
Nickel
Ceramics
0–20.0
Nickel
Copper
0–20.0
Nickel
Iron
0–20.0
Palladium
Nickel
0–40.0
Palladium-nickel
Nickel
0–20.0
Platinum
Titanium
0–8.0
Rhodium
Copper or nickel
0–50.0
Silver
Copper or nickel
0–50.0
Tin
Aluminium
0–60.0
Tin
Copper or nickel
0–60.0
Tin-lead
Copper or nickel
0–25.0
Zinc
Iron
0–40.0
over copper or Kovar cannot be measured using BBS techniques. Other disadvantages exist, such as limits in the aperture/component geometry, and measurement times have led to the further development of XRF techniques and technology.
Principle of XRF Measurement
If sufficient light energy collides with an electron, it is possible for the electron to be driven out of its atomic orbit, a process known as the photoelectric effect [6]. An atom with an electron removed from its orbit is unstable, so to restore equilibrium, an electron from a higher shell must drop
Q
MEASUREMENT OF FILM THICKNESS
527
into the vacant orbit. This transition causes an emission of energy in the form of a light wave or photon. When the inner shell electrons are ejected from an atom, the emitted photon has high energy, and they fall into the region of the electromagnetic spectrum called X rays. X rays have characteristic energy levels determined by the element, which is emitting and can therefore be used to identify the elements in a sample. In XRF instruments an X ray source or tube is used to produce photon emissions as they have an energy distribution capable of fluorescing all elements commonly used in plating. The X ray beam can be accurately illuminated to provide a small focal spot and high-intensity energy suitable for noncontact measurement of complex layers on small components. The characteristic X rays emitted by the target materials are detected using a gas-filled “proportional counter” in which the passage of the X ray ionizes the gas and produces a pulse of electrical charge proportional to the energy of the X ray. The XRF instruments’ electronics convert the charge pulse into a digital signal that can be interpreted as thickness or analyzed for composition and produce the measurement information by comparison with standards of known thickness. XRF instruments have developed with optical alignment systems and motor-driven sample stages to position the sample and computerized analytical equipment to store calibration data to calculate and present data to the user in a suitable format. Table 2 shows some of the applications, which can be successfully measured using XRF.
References [1] New Jersey Zinc Co., “Leaves from a Paint Research Note Book,” No. 1, 1937, p. 33. [2] Brightwell, E. P., “An Optical Method for Measuring Film Thickness of Paint Films,” Official Digest, Federation of Paint and Varnish Production Clubs, ODFPA, Vol. 28, 1956, p. 412. [3] Stoppel, E. A., “Measurement of Thickness of Varnish Films,” Proceedings, ASTEA, Vol. 23, Part 1, 1923, p. 286. [4] Tooke, R., Jr., “A Paint Inspection gage,” Official Digest, Federation of Societies for Paint Technology, ODFPA, Vol. 35, 1963, pp. 691–698. [5] Brenner, A., “Magnetic Method for Measuring the Thickness of Non-magnetic Coatings on Iron and Steel,” J. Res. Natl. Bur. Stand., JRNBA, Vol. 2, 1938, p. 357. [6] Stebel, M. D., and Silvermann, W. M., “XRF Programmable Plating Thickness Measurement Instrumentation,” Proceedings of the International Coil Winding Association, The International Coil Winding Association, Peterborough, UK, November 1984. [7] Ray, G. P., “Thickness Testing of Electroplating and Related Coatings,” Institute of Metal Finishing, The Institute of Metal Finishing, Birmingham, UK, 1993.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
42
MNL17-EB/Jan. 2012
Drying Time Thomas J. Sliva1
THE PROCESS OF DRYING INVOLVES SEVERAL physical and/or chemical changes, such as solvent evaporation, oxidation, and polymerization, all of which are time dependent. The various stages of drying that occur in organic films may be subjective, difficult to measure reproducibly, and are influenced by many factors such as film thickness, substrate, temperature, humidity, light, and air circulation. Therefore, it is essential that most of these variables must be minimized in order to make drying time determinations more quantitative.
PREPARATIONS OF SPECIMENS Substrate Preparation
It is essential that the substrate to be used and the applied wet film thickness be agreed upon in advance, preferably to conform to the intended use of the coating. Flat glass panels are typically the substrate of choice. Ground and polished glass plates are more suitable for low-viscosity coatings that may tend to crawl. All panels must be thoroughly cleaned, dried, and placed in a horizontal position on a level surface.
Application
The test coating should be filtered to remove any dirt or contamination. Test films are typically prepared, in duplicate, using a drawdown bar or doctor blade adjusted to obtain a uniform film thickness. Films should be drawn down at a uniform rate of application to avoid drag on the coating. It is recommended that all test films should be prepared and tested by one operator properly skilled in the method to be used and that a control (known) coating be run alongside the test coating. All testing should be done within an area, any point of which is not less than 1/2 in. (∼ 15 mm) from the edge of the test film. Table 1 can be used as a general guide for film application when nothing more specific is agreed upon between the purchaser and the seller. The dry film thicknesses shown in Table 1 are suggested. Other methods of application, such as spraying, dipping, or flood coat, may be used provided the film thickness obtained is consistent with that recommended under actual usage. Other substrates, such as metal, may be used provided they are smooth and flat.
ENVIRONMENT
When determining drying time, a controlled environment is essential. Variations in temperature, relative humidity, circulation of air, and light will have an effect on the drying time of a coating. The typical standard environment used for determining the drying time of air dry coatings 1
is a temperature of 73.4° ± 3.6°F (23° ± 2°C) and a relative humidity of 50 % ± 5 % under diffuse daylight (about 25 fc). Relative humidity should be strictly controlled for moisture-cure and two-package urethane coatings since their cure is greatly affected by the existing relative humidity. The effect of variation in temperature was discussed by Algeo and Jones [1], who observed a difference of 4 h for a particular paint dried at 73 and 77°F (22.7 and 25°C), both at 50 % relative humidity. All testing should be conducted in a well-ventilated room free from direct drafts and dust. Airflow is important in determining drying time. For films that dry by oxidation, the rate of drying is a function of the concentration of oxygen at the interface. Since oxygen can reach the surface only by diffusion, the rate of drying is a function of the thickness of the stationary air layer. For films that dry by solvent evaporation, the continuous removal of solventladen air hastens drying [2].
TEST METHODS
ASTM D1640, Test Methods for Drying, Curing, or Film Formation of Organic Coatings at Room Temperature
Method D1640 is the most commonly used method to determine the various stages and rates of film formation in the drying of organic coatings normally used under conditions of ambient room temperature. The method describes eight stages of the drying process: 1. Set-To-Touch Time The test film is lightly touched with the tip of a clean finger, and the finger tip is immediately placed against a piece of clean, clear glass to determine when the film does not adhere to the finger or transfer to the glass. 2. Dust-Free Time This test is generally performed to determine when dust or cotton fibers lightly dropped on the test film can be removed by blowing over the test film. Individual absorbent cotton fibers are dropped from a height of 1 in. (25 mm). The film is considered to be dust free when the cotton fibers can be lightly blown off the test film. 3. Tack-Free Time The test film is considered to be tack free when no stickiness is observed under moderate pressure. This can be measured by either of two methods: a. Paper Test Method A special paper (K-4 Power Cable Paper) [3] is placed on the test film under a weight of 2 psi (13.8 kPa). After 5 s, the weight is removed and the test film inverted. If
Deceased, formerly of DL Laboratories, 116 East 16th St., New York, NY 10003.
528 Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 This standard is for EDUCATIONAL USE ONLY. www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
Downloaded/printed byASTM International Copyright © 2012 by
CHAPTER 42
TABLE 1—Recommended film thickness of material to be tested Material
Dry Film Thickness, mils
Oil paints
1.8±0.2
Enamels
1.0±0.1
Drying oils
1.0±0.1
Water-based paints
1.0±0.1
Varnishes
0.85±0.1
Lacquers, polymer solutions
0.5±0.1
4.
the paper drops off within 10 s, the film is considered tack-free. A variation of the above method is used to test the tack-free time of insulating varnishes. The varnish is considered tack free when the paper is placed on the test film under a weight of 1 lb (450 g) for 1 min and tested as above. b. Tack Tester This is a mechanical device that consists of a strip of metal 1 in. (25 mm) wide, 3 in. (75 mm) long, and 0.016 to 0.018 in. (0.41 to 0.46 mm) in thickness. It is bent to form a base 1 in. (25 mm) square and a vertical length 1 in. by 2 in. (25 mm by 50 mm) angled at 135°. The bottom of the base of the tester is covered with aluminum foil [4] (Fig. 1). A 300 g weight is placed on the center of the base and allowed to set for 5 s. The test film is tack free when the tester tips over immediately after the weight is removed. Occasionally, tack-free time may be longer than dry-hard or dry-through time due to the inclusion of external plasticizers in the coating. Dry-To-Touch Time The test film is considered dry-to-touch when no mark is left when the film is touched by a finger. The following variations are used: a. Drying Oils—The film is considered dry-to-touch when it does not rub up appreciably when a finger is rubbed lightly across the surface.
Q
DRYING TIME
b. Lacquers (and Sealers)—The film is considered dryto-touch when no pronounced marks are left by a finger touching the film. Sealers are generally tested on wood or other porous substrates. 5. Dry-Hard Time The test film is considered dry-hard after maximum downward thumb pressure (without twisting) applied to the test film leaves no mark when the contacted area is lightly polished with a soft cloth. 6. Dry-Through (Dry-To-Handle) Time The test panel is placed in a horizontal position at such a height that when a thumb is placed on the film, the arm of the operator is in a vertical line from the wrist to the shoulder. The operator bears down on the film with the thumb, exerting moderate pressure and at the same time twisting the thumb through an angle of 90°. The test film is considered dry through when the film is not distorted by bearing down with moderate thumb pressure and twisting 90°. 7. Dry-To-Recoat Time The test film meets this requirement when a second coat can be applied without causing any film irregularities, e.g., lifting, wrinkling. 8. Print-Free Time The test film meets this requirement when imprinting fabric under a pressure of 1/2 or 1 lb/in.2 (3.5 or 7.0 kPa) shows the coating to be print free. This procedure is similar to ASTM D2091, Test Method for Print Resistance of Lacquers. An indication of the accuracy of these methods is the precision statement developed in ASTM D1640 in which duplicate determinations within a laboratory should agree within ± 10 % [5].
Federal Test Method Standard 141C, Method 4061.2: Drying Time
This method is similar to ASTM D1640. It includes essentially the above stages of drying with the exception of dry-to-touch. However, it includes a test for free-from-aftertack. This test is applicable to coatings where tackiness persists beyond, or reappears at, the through-dry stage. It is similar to the Paper Test Method, discussed earlier in this chapter, except that a 2.8 kg (6.2 lb) weight is used.
ISO Standard 9117, Paints and Varnishes— Determination of Through-Dry State and Through-Dry Time—Method of Test
Fig. 1—Zapon Tack Tester. The base of the tester is padded and wrapped with aluminum foil. The weight, at right, is set on the base for a definite interval. After the weight is removed, the time required for the tester to tip over is the measure of tack.
529
This standard describes a method for determining under standard conditions whether a single coat or a multi-coat system of paint or related material has, after a specified drying period, reached the through-dry state, i.e., a pass/fail test. The test procedure may also be used to determine the time taken to achieve that state. 1. Through-Dry State This state defines the condition of a film in which it is dry throughout its thickness as opposed to that condition in which the surface of the film is dry but the bulk of the coating is still mobile. A single coating or a multi-coat system of paint or varnish is considered to be through-dry when a specified gauze attached to a plunger is placed on the test film under specified pressure (1,500 g) for 10 s, after which time the plunger head is turned through an angle of 90° over a period of
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
530
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
4.
5.
Fig. 2—Through-dry tester.
2.
2 s and removed (Fig. 2). If no damage or markings are noted on the test panel, the film is said to have achieved “through-dry state.” Through-Dry Time This is the period of time between application of a coating to a prepared test panel and the time to achieve the “through-dry state” as outlined above.
British Standard B.S. 3900, Methods of Test for Paints
Parts C-1 thru C-4 of British Standard B.S. 3900 describe drying tests for determining the wet edge time, surface drying, hard-drying, and freedom from residual tack tests. Part C-8 describes a test for determining print-free state or time. 1. Wet Edge Time (BS3900, C1) This procedure is used for determining whether the edge of a film of paint remains “alive” after a specified period of drying. Following a touch-up coat over the film after the specified drying period, the area is evaluated for lack of film continuity, absence of leveling, or variation in color or sheen. 2. Surface Drying (BS3900, C2) This procedure is used to determine the time after which a coating is applied and when approximately 0.5 g of the ballotini (small transparent solid glass spheres) can be poured onto the surface of the film from a height of between 50 and 150 mm and lightly brushed away without damaging the surface. 3. Hard-Drying Time (BS3900, C3) A rubber-faced plunger is covered with cotton twill, rough side outwards, and then loaded to a total weight of 1.8 kg (4lb). The rotating plunger drops into the panel and makes a three-quarter turn while in contact. The paint film is dry-hard when no damage is observed (Fig. 3).
Fig. 3—Hard-drying time apparatus: Assembly.
Freedom from Residual Tack (BS3900, C4) After a specified drying period, a paper-backed gold leaf is placed on the test panel and covered with a microscope slide and an 800 g weight. After 10 s, the weight and slide are removed and the panel is held vertically and lightly tapped to detach the gold leaf. The surface of the paint film is examined for adhesion of gold leaf. Print-Free (BS3900, C8) The state of a coating or varnish when gauze of a specified grade, under specified force and after a specified time, does not leave an imprint on the surface of a coating.
DIN 53 150, Determination of Drying Time of Paints
Drying time is determined in this method by the adherence or non-adherence of sand or paper to the film under various loadings. Stage 1 is determined with sand (0.16 to 0.315 mm) or glass beads (ballotini). The sand is allowed to remain on the film for 10 s. The remaining stages are determined using disks of typewriter paper (22 mm in diameter and weighing about 60 g/m2) and various loads ranging from 5 to 5,000 g/cm2. Interposed between the load and the test disk is a soft rubber cushion. The load remains on the disks for 60 s. The criteria for the seven stages are as follows: 1. Sand easily removed with a soft brush. 2. Disk under load of 5 g/cm2 does not adhere. 3. Disk under load of 50 g/cm2 does not adhere. 4. Disk under load of 500 g/cm2 does not adhere, but film is temporarily marred. 5. Disk under load of 500 g/cm2 does not adhere, but film is not marred. 6. Disk under load of 5,000 g/cm2 does not adhere, but film is temporarily marred. 7. Disk under load of 5,000 g/cm2 does not adhere, but film is not marred.
MECHANICAL DEVICES
In an attempt to improve the accuracy and reproducibility of the drying time test procedure, various mechanical devices have been developed. The following sections outline these devices and the procedures used in determining drying characteristics.
ASTM D5895, Test Methods for Evaluation Drying or Curing During the Film Formation of Organic Coatings Using Mechanical Recorders 1.
Circular Drying-Time Recorder The device consists of a synchronous motor in a metal case resting on a rubber-tipped tripod and rotating
Fig. 4—Straight line recorder. (Courtesy of Byk-Gardner.)
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 42
Q
DRYING TIME
531
Circular drying time devices have been developed for use when determining the drying time of bake finishes that cure at elevated temperatures [up to 500°F (260°C)]. The compactness of the instrument allows the user to place it in an oven at a specified temperature. 2. Straight Line Drying Time Recorder This device consists of multiple needles being drawn over multiple (up to six) parallel coated glass strips [7]. Its speed can be varied to cover drying periods of 6, 12, and 24 h (Fig. 4). It defines the following stages in the drying process: 1. The first stage is a pear-shaped depression corresponding to the time it takes for the solvent to evaporate. 2. The second stage is the cutting of a continuous track corresponding to a sol-gel transition. 3. The third stage is an interrupted track corresponding to the surface dry time. 4. In the fourth stage, the needle no longer penetrates the film, indicating the final drying time. Five-gram brass weights may be added to apply greater pressure on the needles and thus record through drying. The instrument has also been found useful in evaluating gel time of many two-component surface coatings.
NO PICK-UP TIME TRAFFIC PAINT ROLLER Fig. 5—No pick-up time traffic paint roller. (Courtesy of BykGardner.)
a vertical shaft. A pivotal arm assembly is attached to the shaft, operating a vertical stylus with a Teflon sphere that does not stick to the drying film [6]. Under a 12 g load, the Teflon stylus scribes an arc in the drying film. Programmable units that are designed to control motor speeds and cover drying times of 1 min to 96 h—i.e., one 360° rotation of the stylus. A transparent, plastic template with time increments can be placed over the dried coating at the end of the test. The appropriate scale number of circular degrees required for the film to cure hard dry is then used to determine the dry time. During the early stages of drying, the coating tends to flow back into the wake of the stylus. When the tendency of the flow has ceased, the film may be considered set. As the drying process continues, a skin will form. Visually, this part of the film formation is seen when the stylus begins to tear the surface of the film. The film may be considered surface dry or dust free when the skin is no longer ruptured by the stylus. It is considered through-dry when the stylus rides above the film.
This device is described in ASTM Standard D711, Test Method for No-Pick-Up Time of Traffic Paint. The apparatus consists of a steel cylinder weighing 11 lb, 14 oz (5385 g) with two O-rings [6]. It is rolled along a drying film of traffic paint which has been applied on a glass plate. The paint is dry when no paint adheres to the O-rings (Fig. 5).
References [1] Algeo, W. L., and Jones, P. A., “Factors Influencing the Accurate Measurement of Drying Rates of Protective Coatings,” J. Paint Technol., Vol. 41, 1969, p. 235. [2] Technical Subcommittee 37 of the New York Paint and Varnish Production Club, “Investigation of Methods for Measuring Drying Time,” Official Digest, Federation of Paint and Varnish Production Clubs, ODFPA, Vol. 20, 1948, pp. 836–843. [3] Crocker Technical Papers, Inc., Grade R 20–34, Fitchburg, MA. [4] Bonner, R. E., and Brewster, M. D., “Tack Testing Device,” U. S. Patent 2,406,989, Sept. 3, 1946. [5] Prane, J. W., “A Latin Square Drying Time Study,” Paint Industry Magazine, 1961, pp. 39–47. [6] Byk-Gardner, Inc, Gardner Laboratory, Silver Spring, MD; Paul N. Gardner Co., Inc., Pompano Beach, FL 33060. [7] T. Bell Inc., Akron, OH, as well as manufacturers listed in Ref 6.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
Part 10: Optical Properties
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
43
MNL17-EB/Jan. 2012
Color and Light Robert T. Marcus1 PREFACE
IN PREPARATION OF THIS CHAPTER, THE CONTENTS of the fourteenth edition were drawn upon. The author acknowledges the authors of the fourteenth edition, Fred W. Billmeyer Jr. and Harry K. Hammond III. The current edition will review and update the topics as addressed by the previous authors, introduce new technology that has been developed, and include up-to-date references.
TERMINOLOGY
To understand this chapter and to make the best use of it, the reader should be familiar with the terminology of appearance. The precise definition of terms is becoming increasingly important in today’s world community. The paint terminology standard, Definitions of Terms Relating to Paint, Varnish, Lacquer, and Related Products (D16), is the primary source of terms and definitions relating to paint, but it contains very few appearance terms. The reader should refer to Terminology of Appearance (E284) for terms and definitions relating to color and other appearance attributes. All significant terms used in this section are defined in Terminology E284. An important international source of appearance terms is the International Lighting Vocabulary [7], published jointly by the International Commission on Illumination (CIE) and the International Electrotechnical Commission (IEC). However, this is structured from the viewpoint of illuminating engineering. It is less readily available and a much more costly document than Terminology E284. Because color is a significant factor in the appearance of an object, it is an important characteristic of any paint. Appearance, of which color is a part, is one quality of a product that every customer can judge for himself. No matter how good the physical properties of a paint, if its color does not meet the expectation of the customer, the finished product will be rated as unsatisfactory. Color, often thought to be a property of the paint itself, depends on three objective aspects: (1) the spectral composition of the light in which the paint is viewed, (2) the spectral reflectance of the paint, and (3) the spectral response of the eye of the observer. The subjective interpretation of the response to these aspects by the brain is also an essential part of color. Describing the color of a paint or other material requires consideration of all of these and not merely the spectral character of the material. The three objective aspects of color are considered in sections entitled “Light Sources,” “Reflection and Transmission,” and “The Eye.” The sciences involved include chemistry, physics, physiol1
ogy, and psychology. These are broad subjects, and only enough discussion is included to provide a background for understanding the development of test methods. Readers desiring to pursue these subjects in detail should consult an appropriate text [1–6].
LIGHT SOURCES
Light is electromagnetic radiation weighted by the response of the normal human eye. It occupies a small portion of the electromagnetic spectrum between ultraviolet and infrared radiation. Its wavelength range is approximately 380 to 780 nm (Fig. 1).
Natural and Artificial Daylight
Despite modern dependence on interior illumination, daylight is still an important light source since most objects are at some time viewed in it. The spectral composition of daylight, however, is quite variable, depending upon the hour of the day, the season of the year, and the amount of cloud cover. One way of dealing with this variability is to use standard light sources and their spectral power distributions when making visual or instrumental color measurements and calculations (see CIE Standard Sources and Illuminants).
Incandescent Sources
Other light sources must replace daylight when appropriate. For use in homes, incandescent lamplight is generally preferred because it imparts a soft, mellow effect similar to that of candlelight.
Fluorescent Sources
In stores and offices, fluorescent lamps can provide high levels of illumination with low power consumption and heat generation. The most commonly used fluorescent lamp, known as cool white, has a spectral distribution consisting of a relatively smooth curve throughout the visible spectrum. This arises from the fluorescent emission from a phosphor coated on the inside of the lamp tube. The fluorescence is excited by ultraviolet radiation from mercury vapor inside the tube. This lamp is, however, deficient in power in the red end of the visible spectrum. Modifications of it, known as deluxe and super-deluxe versions, have been designed to overcome this deficiency. Fluorescent lamps have also been designed with phosphors emitting only in three rather narrow regions of the spectrum. When these three bands are selected to peak near 450, 530, and 610 nm, light is provided that is especially pleasing to the eye and is energy efficient.
Senior Color Scientist, Sun Chemical Corporation, 1701 Westinghouse Blvd., Charlotte, NC 28273.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed byASTM International This standard is for EDUCATIONAL USE ONLY. Copyright © 2012 by www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
535
536
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 1—Electromagnetic spectrum showing the relatively small portion that the visible spectrum occupies.
The color-balance and color-rendering properties among the various types of commercial fluorescent lamps can be vastly different one from another.
Other Sources
Other light sources have been developed for special purposes. They include arc lamps (sodium, mercury, neon, xenon), metal halide lamps, and high-intensity discharge lamps. None of these lamps has been adopted as a standard for use in color measurement.
Color-Matching Booths
Because of the variation in spectral composition of different natural and artificial sources, it is essential that visual color evaluation and matching be done under standardized illumination, such as that provided by a color-matching booth. This device allows the colorist to compare the colors of specimens under controlled and standardized illumination. Carefully manufactured and maintained light booths permit a colorist to make a visual match with confidence that the illumination duplicates that used at another time or place. However, the spectral power distribution of daylight illumination in color-matching booths is not the same as that of natural daylight.
REFLECTION AND TRANSMISSION
Opaque, Transparent, and Translucent Films
When light strikes an object, some of it may be reflected, some may be absorbed, and if the object is not opaque, some may be transmitted. The reflected light may be concentrated in a glossy, mirror-like, specular reflection, scattered uniformly in all directions (diffuse reflection), or distributed between these two extremes. A highly polished metal can reflect as much as 99 % of the incident light in the specular direction. A white powder, such as barium sulfate, scatters light uniformly in all directions, and it, too, can reflect as much as 99 % of the incident light. Specular reflection is related to the visual perception of gloss; diffuse
reflection is related to the visual perception of lightness and, when it is wavelength dependent, to that of color. Transmission can also be diffuse or regular, depending on whether or not light is scattered in passing through a material. Specimens that both transmit and reflect light are called translucent. A spectrophotometer is used to provide information on the spectrally selective character of a material. Fig. 2 shows typical spectral reflectance curves of some paints. A trained colorist can obtain valuable information from such curves, but spectral data alone are unsatisfactory as a means for color identification. Among the ASTM standards on reflectance and transmittance measurement [8], the most useful are shown in Table 12.
Retroreflection
Retroreflection is defined in Terminology E284 as “reflection in which the reflected rays are preferentially returned in directions close to the opposite of the direction of the incident rays.” It is important in paints and coatings used for signs viewed at night, pavement and pedestrian markings, and other safety devices. The measurement of retroreflection requires special instrumentation and special test methods for the determination of daytime and nighttime colors of retroreflecting materials. Table 2 lists the ASTM standards dealing with this subject. The last two items in the table are standards in preparation and may never become official ASTM standards. 2 As noted in Terminology E284, in the Discussion under reflectance, “The term reflectance is often used in a general sense or as an abbreviation for reflectance factor....” This simplifying convention is used in this chapter, as it is in many textbooks. The reader should refer to Terminology E284 for the definitions of reflectance, transmittance, and radiance, and the corresponding factors. Note that commercial instruments measure reflectance factor, not reflectance.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 43
Q
COLOR AND LIGHT
537
Fig. 2—Spectrophotometric curves typical of those measured on paint films. Note the sharp drop in the curve for rutile titanium dioxide (white) as the violet end of the spectrum is approached. The drop continues in the ultraviolet, where this pigment absorbs light strongly.
Gonioappearance
Gonioapparent coatings change one or more attributes of color with a change in the illuminating or viewing angle. Coatings containing metallic flakes exhibit a change in lightness, while coatings containing mica-based interference pigments or other effect pigments change their hue and chroma. Often metallic flakes are combined with other effect pigments so that the coating will change hue, lightness, and chroma with a change in the illuminating and viewing angles. Multiangle spectrophotometers, sometimes called goniospectrophotometers, are used to measure gonioapparent coatings. These are bidirectional instruments that have the capability of measuring with a variety of illuminating and/or viewing angles. More than one illuminating angle is required to characterize interference pigments, whereas only one illuminating angle can be used to satisfactorily characterize materials containing only metallic flakes. Practice for Multiangle Color Measurement of Metal Flake Pigmented Materials (E2194) describes how to measure gonioapparent materials containing only metallic flake pigments; Practice for Multiangle Color Measurement of Interference Pigments (E2539) describes how to measure gonioapparent materials containing thin film interference pigments; and Practice for Specifying the Geometry of Multiangle Spectrophotometers (E2175) provides a way of specifying geometric properties that are peculiar to multiangle spectrophotometry.
Infrared radiation, with wavelengths longer than 780 nm, is associated with heat transfer. It is widely used for the identification and analysis of chemical compounds. The near-infrared region, with wavelengths from 780 to about 1100 nm, is important for camouflage detection. Most paint pigments do not absorb radiation in this region, but some inorganic pigments reflect visible light and absorb radiation in the near-infrared.
TABLE 1—ASTM standards on reflectance and transmittance measurements Designation
Title
E1164
Practice for Obtaining Spectrometric Data for Object-Color Evaluation
E1331
Test Method for Reflectance Factor and Color by Spectrophotometry Using Hemispherical Geometry
E1347
Test Method for Color and Color-Difference Measurement by Tristimulus (Filter) Colorimetry
E1348
Test Method for Transmittance and Color by Spectrophotometry Using Hemispherical Geometry
E1349
Test Method for Reflectance Factor and Color by Spectrophotometry Using Bidirectional (45:0 or 0:45) Geometry
E805
Practice for Identification of Instrumental Methods of Color or Color-Difference Measurement of Materials
E1345
Practice for Reducing the Effect of Variability of Color Measurement by Use of Multiple Measurements
E2214
Practice for Specifying and Verifying the Performance of Color-Measuring Instruments
Ultraviolet and Infrared Spectral Regions
Ultraviolet and infrared radiation can have important effects on paint. Ultraviolet radiation, with wavelengths shorter than 380 nm, is the principal stimulus of fluorescence of certain pigments, is an aid to identification and analytical determination of certain ingredients of paint, and may promote decomposition of pigments or binders. Colorless pigments absorbing in the ultraviolet region can impart protection against such decomposition. Rutile titanium dioxide absorbs in the ultraviolet, as its spectral curve shows (see Fig. 2).
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
538
PAINT AND COATING TESTING MANUAL
Q
TABLE 2—ASTM standards on Retroreflectance Designation
Title
D4061
Test Method for Retroreflectance of Horizontal Coatings
E808
Practice for Describing Retroreflection
E809
Practice for Measuring Photometric Characteristics of Retroreflectors
E810
Test Method for Coefficient of Retroreflection of Retroreflective Sheeting Utilizing the Coplanar Geometry
E811
Practice for Measuring Colorimetric Characteristics of Retroreflectors Under Nighttime Conditions
E1501
Specification for Nighttime Photometric Performance of Retroreflective Pedestrian Markings for Visibility Enhancement
E1696
Test Method for Field Measurement of Raised Retroreflective Pavement Markers Using a Portable Retroreflectometer
E1709
Test Method for Measurement of Retroreflective Signs Using a Portable Retroreflectometer at a 0.2 Degree Observation Angle
E1710
Test Method for Measurement of Retroreflective Pavement Marking Materials with CENPrescribed Geometry Using a Portable Retroreflectometer
E1809
Test Method for Measurement of High-Visibility Retroreflective-Clothing Marking Material Using A Portable Retroreflectometer
E2176
Test Method for Measuring the Coefficient of Retroreflected Luminance of Pavement Markings in a Standard Condition of Continuous Wetting (RL-Rain)
E2177
Test Method for Measuring the Coefficient of Retroreflected Luminance (RL) of Pavement Markings in a Standard Condition of Wetness
E2302
Test Method for Measurement of the Luminance Coefficient Under Diffuse Illumination of Pavement Marking Materials Using A Portable Reflectometer
E2366
Test Method for Measurement of Daytime Chromaticity of Pavement Marking Materials Using a Portable Reflection Colorimeter
E2367
Test Method for Measurement of Nighttime Chromaticity of Pavement Marking Materials Using a Portable Retroreflection Colorimeter
E2540
Test Method for Measurement of Retroreflective Signs Using a Portable Retroreflectometer at a 0.5 Degree Observation Angle
F923
Guide to Properties of High Visibility Materials Used to Improve Individual Safety (Withdrawn 2006)
WK3833
New Test Method for Determination of the Coefficient of Retroreflection of Pavement Markings Using a 30 Meter Geometry Mobile Retroreflectometer
WK19806
New Test Method for Measuring the Coefficient of Retroreflected Luminance of Pavement Markings in a Standard Condition of Continuous Wetting (RL-Rain)
15TH EDITION
Fluorescence
Some materials have the property of fluorescing when irradiated by ultraviolet or visible radiation. They typically emit radiation at longer wavelengths in the visible range or even in the near-infrared. The effect of fluorescence is to increase the apparent reflectance since the eye responds to the sum of the fluoresced and the reflected energy. This sum may even exceed the amount of light reflected by an ideal white material at the wavelengths of maximum fluorescent emission. Fluorescent pigments had a reputation for poor lightfastness in outdoor applications, but improved technology is changing this. Some fluorescent pigments can now survive at significant levels for several years. Most modern colorimeters and spectrophotometers are designed to properly evaluate the colors of fluorescent materials. However, many of these instruments do not have light sources adequately simulating the ultraviolet content of natural daylight. In such a case, the instrument will not produce the same amount of fluorescence as would daylight. While many instruments will measure the color of fluorescent materials, generally it is necessary to use an instrument that both illuminates the specimen monochromatically and also views monochromatically to separate the fluorescence component from the reflection or the total radiance of a material. These instruments are called either spectrofluorimeters or bispectrometers. Five ASTM standards apply to the measurement of fluorescence. Practice for Color Measurement of Fluorescent Specimens Using the One-Monochromator Method (E991) specifies the instrument geometry required for the measurement and shows how to assess the performance of daylight-simulating instrument light sources. Practice for Detecting Fluorescence in Object-Color Specimens by Spectrophotometry (E1247) provides instrumental methods to supplement simple visual examination of the specimen under ultraviolet light to detect the presence of fluorescence. Practice for Computing the Colors of Fluorescent Objects from Bispectral Photometric Data (E2152) and Practice for Obtaining Bispectral Photometric Data for Evaluation of Fluorescent Color (E2153) specify how to collect and interpret fluorescence data using bispectral, fluorescence-measuring instruments. Test Method for Daytime Colorimetric Properties of Fluorescent Retroreflective Sheeting and Marking Materials for High Visibility Traffic Control and Personal Safety Applications Using 45°:Normal Geometry (E2301) specifies how to collect and interpret fluorescent-retro-reflective materials by bispectral colorimetry using a 45:0 or 0:45 optical measuring system. Recently, fluorescence has been used for the detection of defects in industrial coatings including fluorescing primer coatings, as well as non-fluorescent top coatings. A Specification for Light Source Products for Inspection of Fluorescent Coatings (E2501) was developed to provide the requirements for light source products intended for the excitation of fluorescent materials used for the detection of defects in industrial coatings. Test Method for Luminance Ratio of a Fluorescent Specimen using a Narrow Band Source (E2630) was developed as a companion to Specification E2501 to support the development and specification of industrial coatings that are used in a system for the detection of coating defects when inspected with a Specification E2501 light source. Specification E2501 establishes
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 43
Q
COLOR AND LIGHT
539
a quantitative measure of the optical property of a coating that correlates to its ability to enhance defect contrast under the specified inspection light source.
Phosphorescence
Phosphorescence is similar to fluorescence. However, while fluorescing materials will stop emitting radiation once the irradiation stops, phosphorescing materials continue to emit. This makes phosphorescing materials ideal for safety markings. Two ASTM standards and one specification apply to phosphorescing materials—Guide for Recommended Uses of Photoluminescent Phosphorescent Safety Markings (E2030), Test Method for Photopic Luminance of Photoluminescent (Phosphorescent) Markings (E2073), and Specification for Photoluminescent (Phosphorescent) Safety Markings (E2072).
THE EYE
The Visual System
The human eye functions in a manner similar to a camera. It has a lens to focus images of objects and an iris to control the amount of light that enters. A complex lightsensitive layer, called the retina, plays a role analogous to that of the film in a camera. Neither the structure of the retina nor its function are fully understood. It contains two different types of light receptors that send information along neural pathways to the visual cortex of the brain. They are called rods and cones because of their shapes. The rods are responsible for black-and-white vision at low light levels, and they are not considered further in this section. At usual daylight levels, the rods are overwhelmed and do not contribute to vision. The cones are responsible for color vision. There are three types of cones, each with a different spectral sensitivity. The exact spectral response of each type of cone is not known, although it is assumed that each cone response function is related to the absorption curve of its pigment. The absorption curves are broad and overlapping (Fig. 3). They peak in the short, middle, and long wavelength regions of the visible spectrum; thus the designations blue, green, and red (sensitive) cones are sometimes used [9-12]. Detailed models of color vision have been proposed [13,14], but they are presently based on incomplete information. What happens to the neural signals from the retina on the way to and in the brain is not well understood, but for most work related to color and appearance, it does not need to be.
Perception
Perception is defined as the translation of retinal images by the observer into meaningful information about the environment. The perception of objects and their colors thus represents the overall response of the visual system, including both the eye and the brain. Vision is called a psychophysical phenomenon—physical in the way light reaches the eye, psychological in how the brain interprets the neural signals. The psychological factor determines, for example, whether a given color combination is interpreted as pleasing or displeasing. The mechanism of seeing is physical; the interpretation of what is seen is psychological. Objective color measurement is, however, confined to physical aspects. For example, the perceived color of a specimen may be changed by changing the color of the area
Fig. 3—Spectral curves showing the relative sensitivities of the three types of cones in the eye, peaking in the short (S), medium (M), and long (L) wavelength regions.
surrounding it. This phenomenon, called simultaneous contrast, cannot yet be evaluated instrumentally. Another example of a perceptual phenomenon is chromatic adaptation, defined as the changes in the visual system’s sensitivities due to changes in the spectral quality of the illuminating and viewing conditions. These changes tend to compensate, for example, for the effect of the change in illumination from distinctly bluish daylight to distinctly yellowish incandescent lamplight. The perceived colors of familiar objects tend to appear the same (known as color constancy) when the observer goes between environments illuminated by the two kinds of light, while the actual colors have all been shifted because of the change in spectral composition of the incident light.
The Variables of Perceived Color
The fact that the eye perceives color because it has three types of cones with differing spectral sensitivities implies, and experience confirms, that perceived color should have three variables. Several sets of these variables are of interest because of their wide use.
Object Colors: L, C, and H Cylindrical Systems
Of great interest to the paint colorist are the variables applying to the perception of object colors. Hue is always one of the variables. Hue is defined as the attribute of color described by common names such as red, yellow, green, blue, etc. The hues are commonly arranged in a circle in the order of their appearance in the spectrum, with the circle closed by the purples, mixtures of the red and blue at the ends of the visible spectrum (Fig. 4). A second important variable of object colors is lightness, the attribute by which an object is judged to reflect more or less light. It is often represented graphically by a line through the center of the hue circle and perpendicular to its plane (also shown in Fig. 4). The upper and lower
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
540
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 4—Arrangement of the hue, lightness, and chroma axes in the usual cylindrical representation of color space.
ends of this line, often called the neutral or achromatic axis, are white and black, respectively. The third variable in this set has several different names, referring to variations among what we can perceive: Chroma, saturation, and colorfulness are examples. The common factor among these names is a sense of the amount, in contrast to the kind, of hue in the color. In this section we use chroma as the name for this third variable and show it in Fig. 4. This quantity is exemplified by the distance between the point representing the color and the neutral axis. The Munsell system (see Munsell System) is an example of a cylindrical system.
Colored Lights
When we deal with colored lights instead of objects, two changes need to be made in the above system: Lightness is replaced by brightness and chroma by saturation. Brightness in this sense is defined as the attribute by which an area appears to emit more or less light.
Object Colors: L, a, and b Opponent Systems
A widely used alternative to the hue-lightness-chroma system described above is an opponent-color system that mimics the behavior of the neural signals transmitted from the retina to the brain. The lightness axis, often labeled L, is retained, but the hue circle is replaced by two opponenttype axes at right angles and perpendicular to the lightness axis (Fig. 5). Commonly they are a redness-greenness and a yellowness-blueness axis labeled a and b, respectively, as in the figure. Scales of this type are displayed in many color-measuring systems; examples are given later in Color Order Systems.
Color Constancy and Metamerism Color Constancy
As previously noted under Perception, color constancy is the general tendency of colors to remain constant in appearance when the color of the illumination is changed. Note that this term refers to what happens to the color of a single specimen when the illumination is changed.
Metamerism
Of greater concern to the colorist is what happens to the relationship of two colors when the illumination is changed. This phenomenon is of industrial importance and largely under the control of the colorist. Suppose that two
Fig. 5—Arrangement of the lightness, redness-greenness, and yellowness-blueness axes in the usual opponent-color representation of color space.
colors, which match in daylight, are formulated by using different sets of pigments. The two colors may not match under another type of illumination, such as incandescent lamplight, since the two specimens may exhibit different types and degrees of color constancy. This phenomenon is known as illuminant metamerism, and the colors are said to be metameric. Illuminant metamerism is defined as the property of two specimens having different spectral characteristics and having the same color when viewed by a normal observer under a given illuminant, but different colors when viewed under a different illuminant when other conditions remain the same. The metameric samples whose spectral curves are shown in Fig. 6 were created by using different pigments in each specimen. Observer metamerism is the phenomenon in which two specimens having different spectral characteristics match to some observers but not to others under the same illuminating and viewing conditions. Only when colors have identical spectral curves can they be expected to match under all types of light and to all observers. To avoid metamerism, the same pigment formulation should be used when remaking the color. Correcting a batch by adding pigments that were not in the original formulation can result in metamerism. Whenever pigments used for the match have different spectral characteristics from those used in the target, the resultant color match should be tested for the metamerism by several observers and under several different types of illumination. Daylight, incandescent lamplight, and fluorescent lamplight are commonly used sources for evaluating metamerism. A colormatching booth should be used for making the evaluation. If the match is not satisfactory, spectrophotometric analysis of the two formulations should be carried out to determine their spectral differences, and the formulation should be adjusted to minimize metamerism. Practice for Visual Evaluation of Metamerism (D4086) specifies procedures
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 43
Fig. 6—Spectrophotometric curves of two highly metameric paint films. To most observers, the films match visually and are a dull green color in daylight, but sample A shifts color to a strong reddish brown in incandescent lamp light while sample B exhibits color constancy for this change of illumination.
for identifying the presence of metamerism and evaluating it semiquantitatively. Means of minimizing metamerism in both visual and instrumentally aided color matching are described later in “Color Matching.”
COLORIMETRY AND THE CIE SYSTEM
Colorimetry is defined as the science of color measurement. Its modern development began in 1931, when, in the interest of standardization and to focus attention on the properties of material objects such as paint films, international standards and recommendations were established by the International Commission on Illumination (Commission Internationale de l’Éclairage, CIE). These recommendations [15] define standard lights and observers and a methodology for combining their properties with those of the objects to describe color and related appearance parameters.
CIE Standard Sources and Illuminants
Here, it is necessary to note two conventions of CIE terminology [7] reflected in Terminology E284. A source is defined as a real emitter of light, whereas an illuminant is defined as a spectral power distribution.
Incandescent Source and Illuminant
In 1931 the CIE defined a tungsten-filament incandescent lamp of 2856 K color temperature (see “Other Features of the CIE System”) as Source A. Later, when measurement of spectral power distributions became easier, the fundamental definition was changed to the lamp’s spectral power distribution, known as Illuminant A. The spectral power distribution of Illuminant A is given in Ref. [15], in a CIE/ ISO standard [16], and in abbreviated form in Practice for Computing the Colors of Objects by Using the CIE System (E308).
Daylight Source and Illuminants
The CIE also recommended standard Source C in 1931, consisting of liquid filters used in combination with Source A, representing north-sky daylight. Later, the fundamental definition was changed to that of Illuminant C [15]. Source and Illuminant C do not duplicate the ultraviolet content of natural daylight and thus do not provide
Q
COLOR AND LIGHT
541
correct daylight color rendition of fluorescent materials. In 1968 the CIE adopted the D series of illuminants, which duplicate the spectral power distributions of various phases of natural daylight. They are designated by their correlated color temperatures (see “Other Features of the CIE System”). The most important of these daylight illuminants [15,16] is D65, with a correlated color temperature of 6500 K. The spectral power distributions of Illuminant C and several of the D series are tabulated in Practice E308. Unfortunately, very few real sources, whether for visual or instrumental use, simulated any of the D illuminants satisfactorily. The CIE has recommended procedures for assessing the quality of daylight simulators [17]. The relative spectral power distributions of CIE Standard Illuminants A, C, and D65 are shown in Fig. 7.
Fluorescent Illuminants
In 1986, the CIE defined [15], but did not recommend as standard illuminants, a series of 12 spectral power distributions representative of various types of fluorescent lamps, including cool white lamps simulating daylight well, and three-band lamps. These data should be used when calculations involving fluorescent lamps are required.
CIE Standard Observers 1931 CIE Standard Observer
In order to evaluate colors consistently, a standard observer was defined by the CIE in 1931 [15,18] by evaluating the spectral responses of a small group of well-trained individuals. The spectral responses of the CIE 1931 Standard Observer were determined by means of experiments, like those described later in Additive Mixing of Lights, in which the observer determined the amounts of three primary colors (red, green, and blue) required to match the colors of all wavelengths of the visible spectrum. These sets of three values are called tristimulus values. For convenience, the data were transformed mathematically to correspond to the use of imaginary primary lights designated X, Y, and Z. The Standard Observer is defined by the amounts, x (λ ), y (λ ) and z (h ), of these primaries required to match the spectrum colors; these are plotted in Fig. 8. The symbol (λ) indicates that the quantity depends on the wavelength, λ. The quantities x (λ ), y (λ ) and z (h ), are known as the color-matching functions of the 1931 CIE Standard Observer. The transformation from real primaries to X, Y, and Z was made so that the color-matching function y(h ) is equal to the spectral luminous efficiency function V(λ), that is, the effectiveness of radiation to stimulate the perception of light. This choice means that the tristimulus value Y of a given color, called its luminance, contains all the information about the lightness of the color.
1964 Supplementary Standard Observer
The data for the CIE 1931 Standard Observer were obtained with a visual colorimeter in which the field of view subtended an angle of only 2° at the eye of the observer. This was selected to correspond to the size of the fovea, which is that part of the retina containing only cones used in color vision. Later, the CIE studied color vision in a 10° field, with the central 2° portion disregarded. This corresponds to sample sizes more like those used in commerce, but for which the spectral sensitivity of the eye is somewhat differ-
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
542
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 7—Relative spectral power distributions of CIE standard illuminants A (incandescent lamp light), C (north-sky daylight), and D65 (actual daylight).
ent from that for the 2° field. With the newer data, the CIE established the 1964 Supplementary Standard Observer [15,18]; see also Practice E308. Where confusion might result, quantities referring to the 1964 Supplementary Standard Observer are given the subscript 10; for example, its color-matching functions are x10 (h ), y10 and z10 (h ).
Calculation of Tristimulus Values
The tristimulus values X, Y, and Z of a color can, in principle, be obtained by direct matching, as were the tristimulus values of the spectrum colors defining the standard observers. But this is impractical, and one of two other methods is always used. One of these involves the design and use
Fig. 8—The color-matching functions of the CIE 1931 standard observer; they are the tristimulus values of the colors of the spectrum. Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 43
Q
COLOR AND LIGHT
543
of tristimulus (filter) colorimeters and is discussed later in “Tristimulus (Filter) Colorimeters.” The other requires knowledge of the spectral reflectance curve of the specimen, obtained by spectrophotometry, and the following procedure. At any wavelength, the contribution to a tristimulus value is given by the product of the relative spectral power of the illuminant, S(λ), the reflectance of the specimen, R(λ), and one of the color-matching functions of the observer, for example, x (h ). These products are summed over the visible wavelengths, then normalized by multiplication by a normalization factor k; for example X = k ∑ S (h ) R (h ) x (h ) (and similar equations for Y and Z), where ∑ indicates the summation over the visible wavelength region. The quantity k is chosen to make Y for perfect white equal to 100 k=
100 S ∑ (h ) y (h )
The CIE had defined perfect white as the perfect reflecting diffuser, the ideal reflecting surface that neither absorbs nor transmits light, but reflects all of it. Most textbooks, Practice E308, and Ref. [15] provide examples of the use of these equations. Because the fundamental process is integration, for which summation is an approximation, the process is usually referred to as tristimulus integration. If a large number of sets of reflectance data are to be integrated for the same illuminant and observer, it is convenient to combine quantities such as S (h ) x (h ) by multiplying them together and normalizing the products so that k = 1. The resulting tristimulus weighting factors can then be stored and used by multiplying them by whatever function R(λ) is desired. The CIE has not tabulated or recommended specific sets of tristimulus weighting functions, but a substantial number can be found in Practice E308. They may be used for wavelength intervals of 10 or 20 nm. For closer intervals, such as 5 or 1 nm, the tables in Practice E308, Ref. [15], or Refs. [16,18] should be used. For the wavelength intervals of 10 or 20 nm, the raw spectral data acquired by the instrument are affected by the spectral bandpass error. This error is introduced because the spectral bandpass must be of significant bandwidth to allow sufficient energy to reach the detector. However, the spectral data being reported has a theoretical zero bandwidth. ASTM developed a Practice for Rectification of Spectrophotometric Bandpass Differences (E2729). This Practice outlines the methods that can be used to deconvolve, at least in part, the spectral bandpass error of raw spectral data acquired by abridged spectrophotometry.
Chromaticity Coordinates and Diagram
An important use of the CIE tristimulus values X, Y, and Z is the calculation of coordinates describing the chromaticity of a color, that is, its hue and chroma, ignoring its luminance or lightness. The CIE chromaticity coordinates x, y, and z are computed from the tristimulus values X, Y, and Z by dividing each of these by the sum X+Y+Z. Thus x=
X X +Y + Z
Fig. 9—The CIE 1931 x, y chromaticity diagram showing the various features described in the text.
y=
Y X +Y + Z
z=
Z X +Y + Z
Since x + y + z = 1, only two chromaticity coordinates need be given; usually they are x and y. These chromaticity coordinates can be plotted to yield the 1931 CIE x, y chromaticity diagram, shown in Fig. 9, or the 1964 CIEx10,y10 diagram for the Supplementary Standard Observer; the two are quite similar. Features of the chromaticity diagrams include: (1) the locations of the spectrum colors around the horseshoeshaped spectrum locus, from 400 nm (violet) at the lower left to 700 nm (red) at the right; (2) the straight line along which purples lie joining these two ends of the spectrum locus; and (3) the location of whites (illuminant points) near the middle of the diagram. As an alternative to the three tristimulus values, colors can be specified by the chromaticity coordinates x and y together with the luminance Y (or their equivalents in the 1964 system). These can be arranged in a three-dimensional color space.
Other Features of the CIE System Dominant Wavelength, Complementary Wavelength, and Purity
Dominant wavelength is defined as the wavelength along the spectrum locus at the end of a line drawn from the white point (usually Illuminant C) through the sample point to the spectrum locus. For purple colors where the line would end at the purple locus, it is extended back through the white point to the spectrum locus, and the wavelength at that point is designated as the complementary wavelength of the sample. The fractional distance from the white point to the sample point relative to the distance to the spectrum or purple locus is called the (excitation) purity of the sample. Dominant wavelength correlates well with the hue of the sample, but purity does not correlate well with any perceived quantity and is rarely used today.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
544
PAINT AND COATING TESTING MANUAL
Q
Blackbody Locus, Color Temperature, and Correlated Color Temperature
When a metal, such as a lamp filament, is heated, it first radiates heat in the infrared region, then light with a chromaticity at the red corner of the diagram. As it gets hotter, the chromaticity shifts through the oranges and yellows. The line along which these chromaticities lie for perfect absorbers and emitters of radiation is called the blackbody locus and is shown in Fig. 9. If the metal did not melt, its color would continue along this locus through white to light blue at infinitely high temperature. The temperature of a perfect blackbody can be correlated with its chromaticity and is called the color temperature of the body. It is measured in kelvins, K; the color temperature of Illuminant A is, for example, 2856 K. Many light sources, including the phases of daylight and fluorescent lamps, have chromaticities that are close to, but not on, the blackbody locus. In that case, the color temperature closest to the chromaticity of the source is used and called its correlated color temperature. An example is Standard Illuminant D65, with a correlated color temperature of 6500 K.
Uniform Color Spaces
From almost the beginning of the CIE system, it was recognized that distances in the CIE space did not correlate well with visual estimates of the magnitudes of color differences. Many proposals have been made for deriving quantities that are more uniformly visually spaced. In 1976, the CIE recommended two more nearly uniform color spaces that, although not perfect, have been widely used. Here we describe the one that is most widely used in the paint and related industries and mention briefly the second, more useful, when colored lights are considered. The equations for these two spaces, known by official acronyms CIELAB and CIELUV, are found in Refs. [1,6,15] and in the Practice E308.
CIE 1976 L* ,a*, b* (CIELAB) Space
CIELAB is an opponent-type color space, with a lightness axis L*, a redness (positive values)-greenness (negative) axis a*, and a yellowness (positive)-blueness (negative) axis b*, all mutually perpendicular, as illustrated in Fig. 5. The transformations from Y to L*, from X and Y to a*, and from Z and Y to b* are all nonlinear, using cube-root functions. The equations for these transformations are
L* = 116(Y/Yn)1/3 − 16 a* = 500[(X/Xn)1/3 − (Y/Yn)1/3] b* = 200[(Y/Yn)1/3 − (Z/Zn)1/3] X Y Z , , ≥ 0.008856 X n Yn Z n where Xn, Yn, and Zn are the tristimulus values of the illuminant or white point. See Practice for Calculation of Color Tolerances and Color Differences from Instrumentally Measured Color Coordinates (D2244) for alternate equations when X/Xn,Y/Yn, ·Z/Zn< 0.008856. Because CIELAB does not have tristimulus values or chromaticity coordinates as defined for the 1931 and 1964
15TH EDITION
CIE systems in Chromaticity Coordinates and Diagram, it does not have a chromaticity diagram. An alternative set of CIELAB coordinates results in an LCH cylindrical color space. L remains the same but a* and b* are combined to give a chroma C* and a hue angle h (measured in degrees) C * = ( a*2 + b*2 );
h = tan –1 ( b* / a* )
These correlate well with visual judgments of lightness, chroma, and hue, respectively. Perhaps the widest use of CIELAB is in the calculation of color differences (see “Color Difference Calculations”).
CIE 1976 L*,u*,v* (CIELUV) Space
The CIELUV space has a chromaticity diagram with coordinates u and v, which are linear transformations of x and y, respectively. The linearity is important for the additive mixing of colored lights (see Additive Mixing of Lights). For the three-dimensional CIELUV space, these are combined with the (nonlinear) L* transformation to give opponenttype axes u* and v*, whose meanings are the same as those of CIELAB a* and b*, respectively. An alternative set of hue angle and chroma coordinates, like those in CIELAB, and a color-difference equation, much less widely used than the CIELAB equation, are also part of this system.
COLOR ORDER SYSTEMS
A color order system is a rational method or plan of ordering and specifying all producible object or display colors, or all within a limited domain, by means of a set of physical standards selected and displayed so as to represent adequately the whole set of such colors under consideration. In this section the major color order systems, which have physical exemplifications or atlases illustrating underlying systems, are briefly discussed. References [19, 20] provide useful general coverage.
Munsell System
Dating from the early 1900s, the Munsell system is accepted by most users as the standard for equal visual spacing. It is described in the Practice for Specifying Color by the Munsell System (D1535). Its color solid is like that of Fig. 4 with the sole exception that lightness is called value in the Munsell system. Munsell Hue is designated by position around the hue circle in a notation combining letters designating five major hues (red, yellow, green, blue, purple) and their pairs (R, YR, Y, GY, G, BG, B, PB, P, RP) with numbers from 1 to 10. Munsell Value, to which CIE lightness L* is a good approximation, runs from zero for black to 10 for white. Munsell Chroma, which expresses the degree of departure of the color from the gray of the same lightness, starts at zero and is open-ended. To describe a color in the Munsell system, the hue, value, and chroma are noted in a prescribed sequence, as for example, 8R 4/10. This designation indicates that the hue is red (toward yellow-red, i.e., an orange color), the value is 4, and the chroma is 10. In 1943, the underlying Munsell system was defined in colorimetric terms [21], and since 1968 the chips in the glossy editions of the Munsell Book of Color have been produced to match these specifications. Fig. 10 shows contours of equal Munsell Hue and Chroma at value 5 on the 1931 CIEx,y chromaticity diagram.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 43
Q
COLOR AND LIGHT
545
Fig. 10—Lines of constant Munsell Hue and closed curves of constant Munsell Chroma, at Munsell Value 5 plotted on the CIE 1931 chromaticity diagram.
The Munsell Book of Color [22], which is available in the United States, exemplifies the Munsell system.
ISCC-NBS System
In the 1950s the Inter-Society Color Council (ISCC) and the National Bureau of Standards (NBS) developed the ISCC-NBS Method of Designating Colors [23] based on the Munsell system but grouping similar colors to produce a smaller number (267) of categories. These are designated by descriptive color names, for example, dark reddish orange for the group containing 8R 4 /10.
Universal Color Language
The Munsell system and the ISCC-NBS system are both parts of a Universal Color Language [23], a six-level system for describing color to any desired degree of accuracy. Level 1 consists of the use of 13 hue and neutral names. At this level, the color with Munsell notation 8R 4/10 would be described as orange. In Level 2, which has 29 categories, intermediate hue names are added. Here the color would be described as reddish orange. Level 3 is the ISCC-NBS system, and Level 4 is the Munsell system as
used in the Munsell Book of Color; designations at these levels are given above. Level 5 uses interpolated Munsell notations based on visual comparison of the color with Munsell Book chips. With practice it is possible to interpolate accurately to 1/10 value step, 1/4 chroma step, and from 1 hue step at chroma /2 to 1/4 hue step at chroma near /10 and above. Thus, at this level the color might be designated very accurately as 8.25 R 4.1/9.75. The final stage, Level 6, of the Universal Color Language is based on the results of color measurement, expressed as CIE chromaticity coordinates x,y, and Luminance Y. Now the color might be specified with greatest accuracy as x = 0.527,y = 0.343, Y= 12.5.
DIN System
The DIN Color System is the official German Standard Color System [24]. Its coordinates are hue (for which dominant wavelength is used), saturation (like chroma except representing difference from black instead of from gray of the same lightness), and relative darkness. From time to time, atlases of DIN samples have been available in glossy and matte finish and in transparent films. In the former
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
546
PAINT AND COATING TESTING MANUAL
Q
two, chips are arranged on pages of constant hue, similar to the Munsell arrangement.
NCS
Developed in the late 1960s and 1970s, the Natural Color System (NCS) [25] is the national standard color order system in Sweden and several other European countries. It is based on an entirely different principle from that of the other systems discussed, namely the resemblances of colors to six imaginary elementary colors, unique red, yellow, green, and blue, and black and white. The four hues are placed 90° apart on the hue circle. This does not lead to visually equal hue spacing. Two opponent-type axes result, perpendicular to the black-white lightness axis. The third variable of the system is chromaticness. In terms of resemblances, hue is defined as the resemblance of a color to the two nearest chromatic elementary colors. For example, orange might be said to have 80 % resemblance to red and 20 % to yellow; its hue notation would be Y80R. Chromaticness is the resemblance of the color to the (imaginary) color of the same hue having the maximum possible chromatic content. Lightness is not an explicit variable in the system; the third resemblance can be to either white or black. Scales of resemblances to white, black, and the aforementioned maximally chromatic color are chosen so that these three add to 100 %, so only one of the resemblances to white or black need be specified. For the notation, resemblance to black has been selected, so that the three variables of the system are blackness, chromaticness, and hue. The complete NCS notation for the orange sample considered above in the Munsell system is approximately 3070 Y80R. In the NCS atlas, pages of constant hue notation contain chips arranged on a grid in the form of an equilateral triangle. The corner points of the NCS triangles are imaginary elementary colors.
OSA-UCS System
The Optical Society of American Uniform Color Scales (OSA-USC) system [26,27] was developed by a committee of the OSA between 1947 and 1976. Although the committee concluded that no perfectly equally visually spaced system can exist because color space itself is not Euclidean, it attempted to achieve the best possible compromise and is generally thought to have succeeded. The axes of the color solid are lightness, L (with both positive and negative values around zero at middle gray) and two opponent axes, j (from the French jaune for yellow), with positive values toward yellow and negative values toward purple-blue, and g (for greenness), with blue-greens at the positive end and pinks at the negative end. The OSA-UCS system in described in Practice for Specifying Color by Using the Optical Society of America Uniform Color Scales System (E1360). The 558 samples of the OSA atlas are arranged in a rhombohedral lattice in which each sample, except those at the edges, has 12 nearest neighbors.
COLORCURVE SYSTEM [28]
This new color order system was introduced in the United States in 1989. It is based on a color space similar to CIE 1964 tristimulus space [29]. Aim points are laid out according to additive mixing (see “Additive Mixing of Lights”) of
15TH EDITION
CIE 1964 tristimulus values X10, Y10, Z10 for Illuminant D65, starting from eight specified CIELAB points. The CIELAB lightness axis L (designated only as L) is retained, but coordinates along the CIELAB-type opponent axes are replaced by a simple numbering system, giving the sample location in terms of the number of lattice steps between it and the neutral axis in the directions of two adjacent major axes directed toward reds, yellows, greens, and blues. A typical Colorcurve notation would be L40 R1Y3 representing at lightness level 40 the lattice point one step away from neutral toward the reds and three steps toward the yellows. This is an orange, only slightly different from the sample discussed above. Another feature of the Colorcurve system is that spectral reflectances are furnished for the aim points. These can be additively mixed like tristimulus values to obtain the spectral curve corresponding to any Colorcurve notation; this can be the basis for computer colorant formulation (see “Instrumental and Computer-Aided Color Matching”) to provide a match with minimum metamerism to the surrounding Colorcurve samples. Practice for Specifying and Matching Color Using the Colorcurve System (E1541) describes the system in more detail, including a tabulation of spectral reflectances, tristimulus values, and CIELAB coordinates for the sample aim points.3 The Colorcurve atlas [30] contains about 2,200 painted samples.
Color Collections
For each of the color order systems described above, atlas samples consist of painted chips individually matched to the specified aim points with a mean accuracy of about one CIELAB unit. A number of color collections have been produced by coating paint on paper, printing, molding plastic chips, and dying textiles. A colorist may be required to match these collections. Painted samples are probably the easiest to match because of the similarity of pigments and resins. Plastic samples can present problems because they are often translucent. However, many of the pigments used in plastics are similar to paints and may be fairly easy to match. Printed samples are often very difficult to match. Most are at incomplete hiding and have high chromas because they are printed over white (possibly fluorescent) paper. It may be impossible to achieve the high chromas in a conventional paint system at complete (or nearly so) hiding. The color reproducibility achieved by printing is usually significantly lower than can be achieved with a paint system. Textile samples are made by dying fabric. It may not be possible to match the spectral curve of a dyed fabric with the pigments used in a paint system.
Single-Number Color Scales
Occasions arise in which the colors of samples vary along a single direction in color space. In such a case the color can be described adequately by a single scale value. Three cases are of interest: whiteness, yellowness, and series ranging from colorless to highly colored as the concentration of a colored component increases. E1541 was withdrawn in 2007 after production of the Colorcurve atlas and related materials ceased. This section is included for historical purposes.
3
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 43
Q
COLOR AND LIGHT
547
TABLE 3—Coefficients for the Equations for CIE Whiteness Index and Tint CIE Standard Illuminant and Observer Value
C, 1931
D50, 1931
D65, 1931
C, 1964
D50, 1964
D64, 1964
x
0.3101
0.3457
0.3127
0.3104
0.3477
0.3138
yn
0.3161
0.3585
0.3290
0.3191
0.3595
0.3310
WI,x
800
800
800
800
800
800
WI,y
1700
1700
1700
1700
1700
1700
T,x
1000
1000
1000
900
900
900
T,y
650
650
650
650
650
650
Whiteness and Tint Indices
Whiteness scales start at the point corresponding to an ideal white, which may be the perfect reflecting diffuser (see “Calculation of Tristimulus Values”) or some other industry standard, often assigned a whiteness of 100. Practice for Calculating Yellowness and Whiteness Indices from Instrumentally Measured Color Coordinates (E313) extends the equations recommended by the CIE [15] to Illuminants C and D50. The equation for the whiteness index is WI = Y + (WI,x)(xn – x) + (WI,y)(yn – y) where: Y,x,y = the luminance factor and the chromaticity coordinates of the specimen, xn and yn = the chromaticity coordinates for the CIE standard illuminant and source used, and WI,x and WI,y = numerical coefficients. The whiteness index is valid only for specimens with 40 < WI < (5Y–280). E313 also includes a tint index T= (T,x)(xn – x) – (T,y)(yn – y) where the symbols have meanings analogous to those for the whiteness index. The tint index is valid only for specimens with –3 < T < +3. Values for all the coefficients, except those measured for the specimen, are given in Table 3.
Yellowness Indices
Yellowness indices show the departure of a specimen from achromatic toward yellow. The yellowness index described in Practice E313 is calculated using the equation YI =
100(C X X − CZ Z ) Y
Values for all the coefficients, except those measured for the specimen, are given in Table 4. This equation supercedes the one given in the ASTM Test Method for Yellowness Index of Plastics (D1925), which has been withdrawn.
Scales for Liquids
When only a limited range of color is involved, for example, in the testing of the color of oil, clear varnish, lacquer, or solvents used in the paint industry, simple methods are used consisting of comparison of the specimen with standard colored solutions or glasses ranging from colorless to highly colored. A standardized series of these colors is used to provide a specialized color scale. The color is often a measure of concentration of an ingredient. One difficulty in the use of these special color scales is that the color of the specimen may not match that of the standard; this can make rating on a single-number scale difficult. Nevertheless, their simplicity, low cost, and adaptability to special situations has resulted in wide use of single-number scales for certain applications. Scales useful to the paint industry are shown in Table 5. Test Method D6045 is an instrumental version of Test Methods D1500 and D156. Practice E313 can be used to determine the whiteness and yellowness indices of transparent liquids. A number of single-number color scales were compared by Johnston [31].
INDUSTRIAL COLOR MEASUREMENT
Instruments Using the Eye as Detector
The eye is the ultimate arbiter in color evaluation because, of course, it is the means for making the final judgment on the acceptability of a color. The first color-measuring instruments used the eye as a detector. This practice is still followed when use of a simple color comparator suffices, as in many of the test methods for single-number color scales described in the preceding paragraph. But there are many factors, such as fatigue, poor color memory, and subjectivity, as well as the perceptual phenomena described earlier under
TABLE 4—Coefficients of the Equation for Yellowness Index CIE Standard Illuminant and Observer Value
C, 1931
D65, 1931
C, 1964
D64, 1964
CX
1.2769
1.2985
1.2871
1.3013
C
1.0592
1.1335
1.0781
1.1498
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
548
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
TABLE 5—ASTM standards on scales for liquids Designation
Title
D156
Test Method for Saybolt Color of Petroleum Products (Saybolt Chromometer Method)
D365
Guide for Soluble Nitrocellulose Base Solutions
D1209
Test Method for Color of Clear Liquids (Platinum-Cobalt Scale)
D1500
Test Method for ASTM Color of Petroleum Products (ASTM Color Scale)
D1544
Test Method for Color of Transparent Liquids (Gardner Color Scale)
D1686
Test Method for Color of Solid Aromatic Hydrocarbons and Related Materials in the Molten State (Platinum Cobalt Scale)
D5386
Test Method for Color of Liquids Using Tristimulus Colorimetry
D6045
Test Method for Color of Petroleum Products by the Automatic Tristimulus Method
D6166
Test Method for Color of Naval Stores and Related Products (Instrumental Determination of Gardner Color)
“Perception,” that make the eye at least suspect for close color evaluation work. Therefore, the use of the eye as a detector has been almost entirely replaced by the use of two types of photoelectric instruments developed since the 1940s for color measurement.
Spectrophotometers
Spectrophotometers provide the spectral reflectances (or transmittances) required for the calculation of CIE tristimulus values and derived color coordinates, which were described earlier under “Colorimetry and the CIE System” and in Practice E308. When radiation strikes a sample, it may be reflected, absorbed, or transmitted. Each of these can be measured by most spectrophotometers. All spectrophotometers consist of a light source, a monochromator, arrangements for illuminating and viewing the sample, a photodetector, and an output device. In modern instruments, the latter consists of a computer that processes the signals from the detector and provides for the calculation of tristimulus values and a wide variety of other color-related quantities. Each of these components is described briefly.
The Illuminator
The illuminator is that part of a spectrophotometer that provides the illuminating beam on the specimen, including the source, occasionally a monochromator or spectral filters, a diffuser such as an integrating sphere, if used, and associated optics. In most cases the exact nature of the light source in a spectrophotometer is of little importance as long as it has adequate power and stability at all wavelengths in the visible spectrum. Incandescent lamps or xenon flash tubes are widely used. When fluorescent samples are measured (see “Fluorescence”), the source must illuminate the sample directly and may be filtered to simulate a standard
Fig. 11—A representation of the CIE hemispherical geometry. Light illuminates the integrating sphere to diffusely illuminate the sample. The receiver is placed slightly off of the normal so that the specular reflection can be either included or excluded during the measurement.
source, such as CIE daylight D65. Most color measuring spectrophotometers are designed with these features. Practice E991 provides more information on the measurement of fluorescent samples.
Illuminating and Viewing Geometry
In most spectrophotometers, the geometry of sample illumination and viewing follows CIE recommendations [15]. Two standard geometries are widely used. In hemispherical geometry, light from the source usually illuminates the white interior of a hollow, approximately spherical cavity called an integrating sphere, and the diffused light from the sphere illuminates the sample from all angles in the hemisphere bounded by its surface. The sample is viewed at an angle near the normal or perpendicular to its surface. If the sample is glossy, the specular or mirror reflection from its surface will result in a portion of the wall of the integrating sphere also being viewed. The user is given the option of making this part of the sphere white, thus including the specular component, or black, excluding it (Fig. 11). In bidirectional geometry, illuminating and viewing are at angles along the normal to the sample surface (designated 0°) and 45° from the normal, or the reverse (Fig. 12). The use of bidirectional geometry with the specimen illuminated by white light is recommended when fluorescent samples are measured. Hemispherical and bidirectional geometries are described in the Guide for Selection of Geometric Conditions for Measurement of Reflection and Transmission Properties of Materials (E179) and in Practice E1164, as well as by the CIE [15]. In typical 45/0 or 0/45 geometry, several illuminating or viewing beams, distributed around the azimuth at 45° to the normal, may be used. When a specimen exhibits directionality, that is, its reflectance changes when it is rotated in its own plane, the use of an instrument with multiple beams provides data that average over the directionality, giving a single number characteristic of the average properties of the specimen. If it is desired to measure the directionality, an instrument with one illuminating (or viewing) beam, or two 180° apart in azimuth, should be used and measurements made at several different specimen orientations.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 43
Q
COLOR AND LIGHT
549
photodiode. The photo-detector may consist of an array of diodes, each positioned permanently to receive light of a given wavelength, thus eliminating the need for a spectrum scanning device. The electrical signal from the detector is usually amplified, digitized, and entered in an interfaced computer. In addition to Practice E1164, the following ASTM test methods cover the operation of modern spectrophotometers: Test Method E1331, Test Method E1348, and Test Method E1349.
Spectroradiometers
Fig. 12—A representation of the CIE 45/0 bidirectional viewing. Light illuminates the sample from an angle of 45° and the receiver views the light reflected from the sample along its normal. The positions of the illuminator and receiver can be reversed to obtain the 0/45 illuminating/viewing geometry. For non-directional specimens, the two measurements will be equivalent.
For the measurement of specimens exhibiting gonioappearance, in which the reflectance changes when the illuminating or viewing angles are changed, the use of special instruments capable of measuring at different combinations of these angles is required. Several such multiangle instruments, known as goniospectrophotometers, are commercially available (Fig. 13). See the earlier section on “Gonioappearance.”
The Receiver
The receiver converts the light reflected or transmitted by the specimen into an electrical signal and includes all the optics necessary to focus the light onto a photodetector. The optics in the receiver direct the light to the spectral analyzer in which a narrow (typically 10 to 20 nm) band of wavelengths is selected from the full spectrum of the incident light. The spectral analyzers in modern instruments usually use holographic diffraction gratings or interference filters to isolate the narrow wavelength range. Spectral light is received by the photodetector, usually a silicon
Instruments closely related to spectrophotometers but made to measure light incident from external sources are known as spectroradiometers. They could be used in the paint industry when, for example, the specimen must be sensed remotely after being illuminated externally to the instrument. Methods for making such measurements are described in Test Method for Obtaining Colorimetric Data From a Visual Display Unit by Spectroradiometry (E1336), Practice for Obtaining Spectroradiometric Data from Radiant Sources for Colorimetry (E1341), and Practice for Obtaining Colorimetric Data from a Visual Display Unit Using Colorimeters (E1455).
Obtaining Tristimulus Values from Spectral Data
The first step in utilizing spectral data is the calculation of CIE tristimulus values, as described earlier under Calculation of Tristimulus Values. This step is performed automatically as part of the measurement sequence in all modern spectrophotometers designed for color measurement. It is usual that the spectral bandpass of the monochromator and the measurement interval are selected to be the same. For highest accuracy, the correct set of tristimulus weighting factors, corresponding to the instrument bandpass, selected from among those added to the Practice E308 in 1994 (Table 6), must be used in calculating tristimulus values. In addition, the user must select the standard illuminant and the standard observer used.
Spectrocolorimeters
Some spectrophotometers are designed so that the measured spectral reflectances or transmittances cannot be accessed for examination; only the resulting tristimulus values and other color coordinates can be printed out. Such instruments are designated spectrocolorimeters.
Tristimulus (Filter) Colorimeters
Fig. 13—A representation of a multiangle instrument (goniospectrophotometer) for measuring materials containing metallic flake pigments. One illuminating angle and three viewing angles are sufficient for these materials. To characterize interference pigments, at least two illuminating angles are required.
Among the earliest photoelectric color-measuring instruments were those in which the source-filter-photodetector combinations duplicate the tristimulus functions of the Standard Observer and a CIE Standard Illuminant, usually Illuminant C [32]. How well their filters are designed and matched to the spectral characteristics of the source and of the detector determines how accurately the instrument performs. Today a high degree of accuracy is attained in the resulting tristimulus values. Because of their ease of operation, good precision, and relatively low cost, tristimulus (filter) colorimeters had found wide application for industrial control. They are used primarily as color-difference meters to evaluate the difference in color between a production specimen and a standard of similar spectral character. This last limitation is important for most colorimeters. Because of their design,
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
550
PAINT AND COATING TESTING MANUAL
Q
TABLE 6—Standards pertinent to calibration and maintenance of color-measuring instruments Designation
Title
D3964
Practice for Selection of Coating Specimens and Their Preparation for Appearance Measurements
D5531
Guide for Preparation, Maintenance, and Distribution of Physical Product Standards for Color and Geometric Appearance of Coatings
E1164
Practice for Obtaining Spectrometric Data for Object-Color Evaluation
E1331
Test Method for Reflectance Factor and Color by Spectrophotometry Using Hemispherical Geometry
E1345
Practice for Reducing the Effect of Variability of Color Measurement by Use of Multiple Measurements
E1348
Test Method for Transmittance and Color by Spectrophotometry Using Hemispherical Geometry
E1349
Test Method for Reflectance Factor and Color by Spectrophotometry Using Bidirectional (45°:0° or 0°:45°) Geometry
E2214
Practice for Specifying and Verifying the Performance of Color-Measuring Instruments
most colorimeters provide colorimetric data for only one combination of illuminant and observer and therefore cannot detect metamerism. When specimens are metameric, a colorimeter can give incorrect data on the differences among them. Filter colorimeters should not be used to evaluate pairs of specimens that may be metameric. This limitation should be clearly understood. As a result of improved technology and lower costs, most tristimulus colorimeters have been replaced by spectrophotometers. The use of colorimeters is described in Test Method E1347.
Selection and Calibration of Instruments
The potential user of color-measuring instrumentation must carefully consider the selection of the proper instrument(s), depending on his samples and measurement needs. Guide E179 and Practice E805 address this topic. They should be consulted, if possible, before the purchase of a new instrument. The user of any color-measuring instrument must give regular attention to its calibration. Appropriate calibration standards, available from instrument manufacturers and, in some cases, national standardizing laboratories, should be obtained, maintained in good condition, and used at frequent intervals in an established routine [33]. A collaborative reference program [34] is available by means of which a user’s instrument performance can be compared to that obtained for the same samples by the color community as a whole and by national standardizing laboratories. With proper calibration and maintenance, modern color-measuring instruments are capable of a repeatability and reproducibility far greater than the repeatability with which the samples themselves can normally be prepared for
15TH EDITION
industrial measurement. Thus, special care must be taken to establish a highly reproducible method of paint sample preparation and to establish by careful measurement and periodic verification the uncertainties associated with this step in the measurement process. ASTM standards pertinent to this topic are shown in Table 6. The following two subsections apply mainly to spectrophotometers.
Precision
The modern age of computer-interfaced color measuring spectrophotometers began in the mid-1970s. By then it had been clearly demonstrated that such instruments could surpass the eye in the precision of color measurement. In 1981 the short-term repeatability of instrumental color measurement was reported to be about 0.1 CIELAB unit of color difference [35] (see “Color Difference Calculations”). By 2002 instrument manufacturers were specifying the short-term repeatability for top-of-the-line benchtop instruments to be in the neighborhood of 0.01 CIELAB units. In 1979 the average reproducibility within a group of similar instruments was reported to be about 0.2 CIELAB units [36]. By 2002 instrument manufacturers were specifying the inter-instrument agreement for top-ofthe-line benchtop instruments as less than 0.10. It should be noted that the very low repeatability and reproducibility numbers can only be achieved with stable standards at a tightly controlled temperature. Note that 0.5 CIELAB unit is approximately the smallest difference that can be detected visually. Repeatability and reproducibility values are published in the Precision and Bias section of Practice E1164 and Practice for Calculation of Color Tolerances and Color Differences from Instrumentally Measured Color Coordinates (D2244).
Bias
This quantity, the difference between a measured result for a standard sample and the result when it is measured in a reference laboratory, is difficult to quantify for color measurement where there are no absolute values. It is usual to accept the results obtained in a national standardizing laboratory as the reference. On this basis, when instruments are selected to minimize differences in measuring geometry and are properly calibrated and used, the average bias for a group of standard samples can be as low as 0.5 CIELAB unit [37]. In addition, the differences between the two sets of measurements can be utilized effectively to model and correct the systematic spectrophotometric errors leading to the bias [38].
Commercial Instruments
Specific information about instruments published in a manual such as this would almost certainly become obsolete by the time the manuscript is printed. For this reason, detailed information on suppliers and their instrumentation is not included. Such information, including types of instruments manufactured, their design features and performance characteristics, can best be found by first consulting buyers’ guides [39,40] that are revised annually. Another useful source of information is the lists of exhibitors at paint shows, such as the descriptions published annually in the JCT Coating Tech. An additional source of information on commercial instruments is the internet.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 43
Most color instrument suppliers maintain internet sites with current information on their products and services. Registered users can also obtain product updates and support. In contacting suppliers of instrumentation, the user might request that they send literature, measure samples sent them, or have a representative visit the user to demonstrate the instrument. Another procedure is to negotiate a rental-purchase contract. While some suppliers are reluctant to negotiate such contracts, the procedure provides users the opportunity to evaluate the instrument with their own operators at their own facilities. 1. Portable instruments. In earlier times, production samples had to be taken to a quality-control laboratory for evaluation. This procedure is still desirable in some instances to permit control of such variables as ambient light, temperature and humidity, and atmospheric contamination. However, time can be saved by making measurements on the production floor using portable instruments. Many manufacturers now supply portable spectrophotometers and colorimeters that provide small size and light weight. Some portable instruments perform nearly as well as benchtop instruments. In many, the precision is good, but limitations related to measurement geometry may increase bias. If very small illuminated and viewed areas are used, there may be an advantage in that smaller portions of samples can be measured, but averaging of multiple measurements (see Practice E1345) may be required to obtain representative values for nonuniform samples. 2. Multiangle Evaluation. Several manufacturers now supply instruments with multiangle geometry for the measurement of gonioapparent samples, such as those containing metal flake or pearlescent pigments. For research in this field, a goniospectrophotometer is required. For quality control, however, use of one angle of illumination with several angles of view, or several angles of illumination with one angle of view, may suffice. 3. Location and Type of Use. For continuous processes, online instrumentation is finding greater use. In addition, color measurement and computer-aided color matching (see Instrumental and Computer-Aided Color Matching) are used extensively at point-of-purchase locations such as retail paint stores.
COLOR DIFFERENCE EVALUATION FOR COLOR CONTROL
The color control of industrial paint products is generally aided by color measurement, as described in the last section. The instruments provide color coordinates of sample and standard from which accurate measures of color difference may be derived. These data can also be used to establish color tolerances. Of course, color differences may also be judged visually, and several ASTM standards address this topic: Practice for Visual Appraisal of Colors and Color Differences of Diffusely-Illuminated Opaque Materials (D1729), Test Method for Evaluation of Visual Color Difference With a Gray Scale (D2616), and Guide for Selection, Evaluation, and Training of Observers (E1499). This guide also describes materials and methods for the testing and evaluation of the observer’s color vision.
Q
COLOR AND LIGHT
551
Color-Difference Calculations
In the 1931 and 1964 CIE systems, equal distances in different parts of color space do not represent equally perceived color differences. Although many proposals for more uniform color spaces have been made, an ideal uniform space has not been and may not ever be devised. Sets of color coordinates can be derived from tristimulus values that correlate better with perceived color differences than do the original CIE tristimulus values. Modifications of CIE tristimulus values have been proposed for more uniform color spaces. Those proposed by Hunter [32], MacAdam [41-43], and the CIE [15] had been widely use in the paint industry. The color-difference equation based on the CIE 1976 L*a*b* (CIELAB) space discussed earlier in Uniform Color Spaces and colordifference equations based on that equation perform significantly better than the Hunter and MacAdam equations when compared to experienced visual color judgments. As a result, the Hunter and MacAdam equations can no longer be recommended. The total CIELAB color difference, ΔEab* , is given by * ΔEab = [( ΔL* )2 + ( Δa* )2 + ( Δb* )2 ]1/ 2
where ΔL* = L*TRIAL − L*STANDARD Δa* = a*TRIAL − a*STANDARD Δb* = b*TRIAL − b*STANDARD * It is convenient to express ΔEab in terms of differences in the Munsell-type coordinates hue, chroma, and lightness, but the hue angle h defined earlier in Uniform Color Spaces does not have the same dimensions and scaling as chroma C* and lightness L*. Instead of using h, it is necessary to define CIELAB hue difference ∆H* to combine with ∆C* and ∆L* 2 2 * * 2 ΔH ab = ⎡( ΔEab ) − ( ΔCab* ) − ( ΔL* ) ⎤⎥⎦ ⎢⎣
1/ 2
Then 2 2 * * 2 ΔEab = ⎡( ΔH ab ) − ( ΔCab* ) − ( ΔL* ) ⎤⎥⎦ ⎣⎢
1/ 2
* This method of computing ΔH ab loses the information about the sign of the hue difference (positive or negative) and can be unstable for pairs of colors near the neutral axis. Alterna* tive equations for determining ΔH ab and its sign have been proposed [44] that corrects both problems * * * * * * * ⎤ ΔH ab = s ⎡⎣ 2 ( Cab , B Cab , S − aB aS − bB bS ) ⎦
1/ 2
where s = 1 if aS* bB* > a*B bS* ,
else S = −1
in which the subscript B indicates the trial and the subscript S indicates the standard. The unit of CIELAB color difference is two to three times the just perceptible color difference.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
552
PAINT AND COATING TESTING MANUAL
Q
In the use of color differences, it is important to examine the components of the color difference, such as ∆L*, ∆a* and ∆b*, or ∆L*, ∆C*ab, and ∆H*ab, as well as ∆E*ab itself. The magnitudes and signs of these components can provide valuable information on remedial action needed to bring a production batch on shade. For near-neutral and pastel colors, there are advantages to using the ∆L*, ∆*a, and ∆*b forms of the color-difference equation. The interpretation of measured color differences becomes more complex when the specimens are gonioapparent and measurements are made at several combinations of illuminating and viewing angles. The quantitative meaning of color differences as the angles change has not yet been thoroughly studied. Since the adoption of the CIELAB equations, a number of new color-difference equations have been developed. One of these, known as the CMC (l:c) equation [45,46], is a modification of CIELAB that has improved uniformity of visual perception of its color differences. The CMC (l:c) equation has found widespread use in the coatings industry. The CMC (l:c) equation modifies the ∆L*, ∆C*, and ∆H* components of the CIELAB color difference 2
⎛ ΔL* ⎞ ⎛ ΔC * ΔECMC ( l : c ) = cf ∗ ⎜⎜ ⎟⎟ + ⎜⎜ ⎝ l ∗ S L ⎠ ⎝ c ∗ SC SL =
0.040975 ∗ L* 1 + (0.01765 ∗ L* ) S L = 0.511 SC =
for
for
2
⎞ ⎛ ΔH * ⎞ ⎟⎟ + ⎜⎜ ⎟⎟ ⎠ ⎝ SH ⎠
L’ = L* a’ = (1 + G) · a* b’ = b*
L* ≥ 16
( a)2 + ( b)2
C =
If b ≠ 0, then h = 180 – (180/π) arctan(a/b) – 90 sign(b) If b = 0, then h = 90 sign((a)2) – 90 sign(a) where sign is a function that returns the sign of the argument and arctan is the inverse tangent function returning angles in units of radians ⎛ (C *)7 G = 0.5* ⎜ 1 − ⎜ (C *)7 + 257 ⎝
∆L = LB – LS ∆C = CB – CS ΔH = s ∗ 2 ⎡⎣( C! ∗ CS ) − ( aB ∗ aS ) − ( bB ∗ bS ) ⎤⎦ where s = 1 if aS bB > aB*bS,
S H = SC {(T ∗ f ) + 1 − f } where 1/ 2
⎧ ( C * )4 ⎫ ⎪ ⎪ f =⎨ ⎬ * 4 ⎪⎩ ( C ) + 1900 ⎪⎭
if 1644 ° < h < 345°
else T = 0.36 + abs|0.4*cos(h + 35°)| in which “abs” indicates the absolute, i.e., positive value, of the term inside the brackets. The parameters l and c are to compensate for systematic bias or parameteric effects such as texture and sample separation. The most common values for l:c are 2:1. The parameter cf is a commercial factor [47] used to adjust the total volume of the tolerance region so that accept/reject decisions can be made on the basis of a unit value of the tolerance. In 2001 the CIE recommended an improved color-difference equation, CIEDE2000, based on the CIELAB color space [48]. This equation has been shown to outperform the CMC (l:c) equation [49]. CIEDE2000 is calculated from
else s = −1 2
2
L* < 16
⎞ ⎟ ⎟ ⎠
where C * is the arithmetic mean of the CIELAB C* values for the pair of specimens (standard and batch).
2
0.0638 ∗ C * + 0.638 1 + (0.0131 ∗ C * )
T = 0.56 + abs 0.2 cos(h + 168 ° )
15TH EDITION
⎛ ΔL′ ⎞ ⎛ ΔC′ ⎞ ⎛ ΔH ′ ⎞ ΔE002 = ⎜ ⎟ +⎜ ⎟ ⎟ +⎜ ⎝ K L ⋅ SL ⎠ ⎝ K C ⋅ SC ⎠ ⎝ K H ⋅ SH ⎠ ⎛ ⎞ ΔC′ ⋅ ΔH ′ + Rr ⋅ ⎜ ⎟ K ⋅ S ⋅ K ⋅ S ⎝ C C H H⎠
2
2 ΔE00 = ΔE00
The factors KL, KC and KH are correction terms for variation in perceived color-difference due to the viewing conditions. To obtain color-differences similar to CMC (2:1), set KL = 2 and KC = KH = 1. SL = 1+
0.015 ∗ ( L − 50 ) 20 + ( L − 50 )
2
2
where L ' is the arithmetic mean of the CIELAB L’ values for the pair of specimens (standard and batch). SC = 1 + (0.045 ∗ C) where C ' is the arithmetic mean of the CIELAB C’ values for the pair of specimens (standard and batch). S H = 1 + (0.015 ∗ C ∗ T ) RT = − RC ∗ sin( 2 ∗ Δe ) RC = 2 ∗
(C)7 (C)7 + 257
⎧⎪ ⎡ h − 275° ⎤ 2 ⎫⎪ Δθ = 30 ∗ exp − ⎨ ⎢ ⎥ ⎬ ⎪⎩ ⎣ 25 ⎦ ⎪⎭
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 43
T = 1− [0.17 * cos(h − 30°)] + [0.24 * cos(2h)] + [0.32*cos(3h + 6°)] − [0.20*cos(4h − 63°)] where h is the arithmetic mean of the CIELAB L’ values for the pair of specimens (standard and batch). All angles are in degrees. Take care calculating the mean hue-angle if the color-difference pair has samples in different quadrants. For example, a color-difference pair has hue angles of 30° and 300°. The mean hue-angle for this example is 345°. To determine the mean correctly, calculate the absolute difference of the hue angles. If the absolute difference is larger than 180° then add 360 to the smaller hue angle and divide that sum by 2.
Color Tolerance
A valuable means of recording the experience gained in the production of a given color is through the use of color tolerances, expressed as numerical limits to ∆E* and its components or as color-tolerance charts. Because no color space is entirely uniform, industrial color tolerances are often based on acceptability, not perceptibility, considerations between buyer and seller. General color differences should be used only as a guide until enough experience has been accumulated for each color to allow specification of a firm tolerance limit.
Tolerance Charts
Color-tolerance charts are usually enlarged sections of a nearly visually equally spaced diagram, such as the CIELAB a*,b* diagram. A color-tolerance chart is set up with the coordinates of the standard at the center, and the differences ∆a* and ∆b* of production batches are entered as they are made (Fig. 14(a)). The process is described in Practice for Establishing Color and Gloss Tolerance (D3134). As experience is gained, it should be possible to draw a tolerance-limit curve enclosing most, if not all, of the acceptable batches and excluding most, if not all, of the unacceptable batches. Because no known diagram is perfectly visually uniform or if there is a preferred direction for the batches relative to a standard, then the tolerance limit figure may not be a circle or an ellipse and may not be centered on the location of the standard. Lightness differences between the standard and batch must also be taken into account. This can be done by use of a separate tolerance for ∆L*, but better practice is to use a second chart (Fig. 14(b)), in which ∆L* is plotted against either ∆a* or ∆b*, depending on the data. The batch readings should fall within the tolerance figure on both charts. It must be emphasized that the final criterion for acceptance of a color is always its visual appearance, to which instrumental measurements, color differences, and tolerance charts provide only clues. Instrument malfunctions, miscalibrations, miscalculations, and operator errors can be revealed only by use of confirming visual observations.
Indices of Metamerism
Another important use of color-difference equations is to quantify indices of metamerism as an aid in reducing this defect between standard and batch. The CIE [15,50] has recommended that indices of metamerism be calculated
Q
COLOR AND LIGHT
553
as the color differences between two samples that match under a standard illuminant or to a standard observer when measured for a test illuminant or a test observer. For illuminant metamerism, the standard illuminant is usually daylight. The second observer is selected by the user and is usually incandescent light. Often a third illuminant is specified, which is usually a fluorescent lamp. For observer metamerism, the standard observer is either the CIE 1931 or 1964 observer. To determine the index of metamerism, a corresponding second standard deviate observer is provided [50] based on the range of normal color vision found in the human population.
COLOR MIXING
Additive Mixing of Lights
When a white card is illuminated, its apparent color is the color of the incident light; a red light, for example, makes it appear red. Three colored lights (red, green, and blue), if carefully chosen, would, when placed to overlap on the same area of the card, produce white again (Fig. 15). In this area, the reflected lights from the three sources add together. These three lights are called the additive primaries. By adjusting the intensities of each of the three lights, a wide range of colors can be produced. The procedure just described can provide an analytical tool for color measurement. If an adjacent spot on the white card is illuminated by light of an unknown color, it could be matched visually, in principle, with the three-light combination by adjusting their intensities. This is the basis for color matching by addition of lights. The additive principle is used to produce the colors in television and computer monitors [51]. The analytical device described above is also a simple version of the visual colorimeter used to generate the tristimulus values of the spectrum colors that form the basis for the CIE Standard Observers (see CIE Standard Observers). At each wavelength, the total light reflected is the sum of the power reflected from each of the three primary lights. Since tristimulus values are obtained by adding such sums across the spectrum, they too are produced by additive mixing. It is not necessary to know the spectral nature of the additive primaries to predict the resulting colors.
Subtractive Mixing in Transparent Films
Now consider the light falling on a white card after passing through a film that contains three transparent colorants, yellow, cyan (blue-green), and magenta (red-purple). This might be the situation with a free-standing transparent plastic film, for example. If the three colorants are ideally chosen, all the light would be absorbed when all three colorants are present, and the result would be black (Fig. 16). These three colorants are called the subtractive primaries. If varying quantities of them were combined in subtractive mixing, a wide variety of colors could be produced. However, these mixture colors cannot be predicted from the colors of the primaries alone, as in the additive mixing described above. One must also know their spectral character and compute, wavelength by wavelength, how much light is removed by each primary through absorption, using the well-known Beer’s law relationship. From what is not absorbed, tristimulus values can be calculated in the usual way.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
554
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 14—Color tolerance charts plotted in the CIELAB system: (a) chromaticity tolerance figure on a Δa*, Δb* diagram; (b) lightness tolerance figure on a ΔL*, Δa* diagram. In this case, a plot of ΔL* versus Δb* would have served as well. (Based on ASTM D3134.)
Pigment Mixing
Pigments are added to paints to produce a wide variety of colors. The pigmentation is chosen to control the three components of color—hue, lightness, and chroma. Four pigments are needed to control the components. Generally two pigments are used to control the hue, one
(a black or umber) to control the chroma and usually a white to control the lightness. The white pigment is also used as an opacifier. Mixing of colors using pigments is more complex than simple subtractive mixing in transparent films because of the scattering of light caused by the pigments. Again, it is necessary to know the spectral character of all the colorants present. The Beer’s law calculation of the transparent case is replaced by the Kubelka-Munk relation: For opaque films, as is often the case in the paint industry, there are simple equations relating the reflectance R(λ) of the film at each wavelength to the ratio of two constants describing what happens to the light in the film K (λ ) / S (λ ) = [1 − R(λ )]2 / 2 R(λ )
{
}
R(λ ) = 1 + K (λ ) / S (λ ) − [ K (λ ) / S (λ )]2 + [2 K (λ ) / S (λ )]
1/ 2
in which K(λ) is the Kubelka-Munk absorption coefficient, and S(λ) is the Kubelka-Munk scattering coefficient. Here, R(λ) must be expressed as a decimal instead of the usual percentage. The quantities K(λ) and S(λ) refer to the mixture of pigments in the paint film. These are calculated by adding the separate contributions of each pigment, using its K(λ) and S(λ) weighted by its concentration C in the mixture
Fig. 15—Representation of the additive mixing of colored lights showing the additive primaries red, green, and blue forming the mixture colors yellow, cyan, and magenta where the primaries overlap in pairs and white where all three overlap.
K (h ) MIXTURE [C1 ∗ K1 (h )] + [C2 ∗ K 2 (h )] + ... = S (h ) MIXTURE [C1 ∗ S1 (h )] + [C2 ∗ S 2 (h )] + ... in which there are as many terms as there are pigments used. These equations form the basis for computer color
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 43
Q
COLOR AND LIGHT
555
how to use it, coupled with knowledge of the behavior of the paint system and pigments available, to predict the colors that result from mixing them under carefully controlled conditions. Careful record keeping to document the results is essential. Visual methods of recognizing and controlling metamerism have been described [53,54]. These require no more than knowledge of the behavior in mixtures of the pigments used and a good color matching booth (see “Color Matching Booths”).
Instrumental and Computer-Aided Color Matching
Fig. 16—Representation of the subtractive color mixing of transparent colorants showing the subtractive primaries yellow, cyan, and magenta forming the mixture colors green, blue, and red where the primaries are mixed in pairs and black where all three are mixed together.
matching described in the next section. The equations for dealing with translucent films are more complex [52].
COLOR MATCHING
One of the major objectives of industrial coloring is to match the color a customer wants. Whether this is done visually or with the aid of instruments and computers is often a matter of work load and economics. Therefore, we address both visual and instrumental matching. In either case, a major objective of color matching in paint systems is the formulation of a nonmetameric match to a given paint sample. The difference in spectral character of the samples of a metameric pair is the property determining their metamerism, and achieving a nonmetameric match places several requirements on the formulation. First, the same pigments must be used. This requires identification of the pigments in the sample to be matched. Second, the pigments must be used in the same resin system since the spectral properties of pigments can depend upon the choice of binder. Finally, the same degree of dispersion must be achieved since the absorption and scattering properties of pigments that determine their spectral character change with degree of dispersion. This usually means that the same method of dispersion must be used.
Visual Color Matching
The selection of visual color matchers should be made with great care, and reference should be made to Guide E1499 for details of how to make the selection, test the color vision of the candidates, and train them in making visual judgments. Beyond this there is no substitute for the experience of the visual color matcher. He or she must learn the skill by practice. This should include becoming familiar with one of the better color order systems and atlases and learning
The importance of eliminating metamerism dictates that the instrument of choice is clearly a spectrophotometer. Display of the spectral curves of samples, and the computeraided color-matching operations that lead to spectral curve shapes minimizing metamerism, cannot be achieved by the use of other types of instruments. The spectrophotometer can also be used in a simple but powerful method of organic pigment identification [55] based on extraction of the pigments into solution followed by spectrophotometry. This can be an important aid to selecting the same pigmentation for the match as was used in the standard. Many modern color-measuring spectrophotometers can be obtained with computer software for color matching. Most of these systems work very well, but it must be emphasized that the investment in building up the database that is essential for their use, and in taking all the steps necessary to bring the coloring process under precise control, is not a small one. The details of how computer color matching works are beyond the scope of this chapter. In summary, most systems call for measurement of the sample to put its spectral data into the computer. These data are usually matched at each of 16 or 31 wavelengths across the spectrum by KubelkaMunk-type calculations [56] using stored values of K() and S() for the useful pigments in the product line. The results are the pigment concentrations in the formulation. Several formulations arising from mixing different pigments are usually produced. Additional calculations of metamerism indices and pigment costs allow the selection of the most suitable results. Computer color matching has been discussed in a book [57], and some textbooks [1,2,5,58,59] provide useful summaries. The underlying Kubelka-Munk theory, applied to paint systems, has been the subject of a number of articles [56,60-65] directed to paint colorists. More complex theory appears to be needed only in special cases, such as matching automotive paints containing metal flake or pearlescent pigments. The development of such theories is still in its early stages.
References
[1] Berns, R. S., Billmeyer and Saltzman’s Principles of Color Technology, 3rd ed., Wiley, New York, 2000. [2] Hunter, R. S., and Harold, R. W., The Measurement of Appearance, 2nd ed., Wiley, New York, 1987. [3] Wyszecki, G., and Stiles, W. S., Color Science: Concepts and Methods, Quantitative Data and Formulae, 2nd ed., Wiley, New York, 1982. [4] Hunt, R. W. G., Measuring Colour, 3rd ed., Fountain Press, England, 1998. [5] McDonald, R., Ed., Colour Physics for Industry, 2nd ed., Society of Dyers and Colourists, Bradford, England, 1997.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
556
PAINT AND COATING TESTING MANUAL
Q
[6] McLaren, K., The Colour Science of Dyes and Pigments, 2nd ed., Adam Hilger, Bristol, England, 1986. [7] Commission Internationale de l’Éclairage, Publication CIE No. 17, International Lighting Vocabulary, 4th ed., 1987. Central Bureau of the CIE, Vienna. Available from U. S. National Committee of the CIE, c/o Thomas M. Lemons, TLA-Lighting Consultants, Inc., 7 Pond St., Salem, MA 01970, http:// www. cie-usnc.org. [8] ASTM Standards on Color and Appearance Measurement, 7th ed., ASTM, West Conshohocken, PA, 2004. [9] Stockman, A., Sharpe, L. T., and Fach, C. C., “The Spectral Sensitivity of the Human Short-Wavelength Cones,” Vision Res., Vol. 39, 1999, pp. 2901–2927. [10] Stockman, A., and Sharpe, L. T., “Spectral Sensitivities of Middle- and Long-Wavelength Sensitive Cones Derived from Measurements in Observers of Known Genotype,” Vision Res., Vol. 40, 2000, pp. 1711–1737. [11] Stockman, A., and Sharpe, L. T., “Tritanopic Color Matches and the Middle- and Long-Wavelength Sensitive Cone Spectral Sensitivities,” Vision Res., Vol. 40, 2000, pp. 1739–1750. [12] Stockman, A., Sharpe, L. T., Merbs, S., and Nathans, J., “Spectral Sensitivities of Human Cone Visual Pigments Determined In Vivo and In Vitro,” in Vertebrate Phototransduction and the Visual Cycle, Part B, K. Palczewski, Ed., 2000. [13] Boynton, R. M., Human Color Vision, Holt, Rinehart and Winston, New York, 1979. [14] Hurvich, L. M., Color Vision, Sinauer, Sunderland, MA, 1981. [15] Commission Internationale de l’Éclairage, Publication CIE No. 15:2004, Colorimetry, Central Bureau of the CIE, Vienna, 1986. Available from U. S. National Committee of the CIE, c/o Thomas M. Lemons, TLA-Lighting Consultants, Inc., 7 Pond St., Salem, MA 01970, http://www.cie-usnc.org. [16] CIE Standard on Colorimetric Illuminants, Publication CIE No. S001 (ISO IS 10526), Central Bureau of the CIE, Vienna, 1986. Available from U. S. National Committee of the CIE, c/o Thomas M. Lemons, TLA-Lighting Consultants, Inc., 7 Pond St., Salem, MA 01970, http://www.cie-usnc.org. [17] A Method for Assessing the Quality of Daylight Simulators for Colorimetry, Publication CIE No. 51, Central Bureau of the CIE, Vienna, 1981. Available from U. S. National Committee of the CIE, c/o Thomas M. Lemons, TLA-Lighting Consultants, Inc., 7 Pond St., Salem, MA 01970, http://www. cie-usnc.org. [18] CIE Standard on Colorimetric Observers, Publication CIE No. S002 (ISO IS 10527), Central Bureau of the CIE, Vienna, 1986. Available from U. S. National Committee of the CIE, c/o Thomas M. Lemons, TLA-Lighting Consultants, Inc., 7 Pond St., Salem, MA 01970, http://www.cie-usnc.org. [19] Billmeyer, F. W., Jr., “A Survey of Color Order Systems,” Color Res. Appl.,Vol. 12, 1987, pp. 173–186. [20] Agoston, G. A., Color Theory and Its Application in Art and Design, 2nd ed., Springer, New York, 1987, Chaps. 8–10. [21] Newhall, S. M., Nickerson, D., and Judd, D. B., “Final Report of the O. S. A. Subcommittee on the Spacing of the Munsell Colors,” J. Opt. Soc. Am., Vol. 33, 1943, pp. 385–418. [22] Available from X-Rite, 4300 44th St. SE, Grand Rapids, MI 49512, http://www.xrite.com. [23] Kelly, K. L., and Judd, D. B., “Color: Universal Language and Dictionary of Names,” NBS Special Publication 440, U. S. Government Printing Office, Washington, 1976. [24] Richter, M., and Witt, K, “The Story of the DIN System,” Color Res. Appl., Vol. 11, 1986, pp. 138–145. [25] Hård, A., and Sivik, L., “NCS — Natural Color System: a Swedish Standard for Color Notation,” Color Res. Appl., Vol. 6, 1981, pp. 129–138. [26] MacAdam, D. L., “Uniform Color Scales,” J. Opt. Soc. Am., Vol. 64, 1974, pp. 1691–1702. [27] Nickerson, D., “Uniform Color Scales Samples: A Unique Set,” Color Res. Appl., Vol. 6, 1981, pp. 7–33. [28] Colorcurve is a registered U.S. trademark used by ASTM under the authorization of Colorcurve Systems, Inc. Aspects of Colorcurve technology are covered by U.S. Patent 5,012,482. [29] Stanziola, R., “The Colorcurve System®,” Color Res. Appl. Vol. 17, 1992, pp. 263–272.
15TH EDITION
[30] Formerly available from Colorcurve Systems, Inc., 200 Sixth St., Fort Wayne, IN 46808. After production of Colorcurve materials ceased, the Practice for Specifying and Matching Color Using the Colorcurve System (E1541) was withdrawn. [31] Johnston, R. M., “Colorimetry of Transparent Materials,” J. Paint Technol., Vol. 43, No. 553, 1971, pp. 42–50. [32] Hunter, R. S., “Photoelectric Tristimulus Colorimetry with Three Filters,” J. Opt. Soc. Am, Vol. 32, 1942, pp. 509–538. [33] “Guide to Material Standards and Their Use in Color Measurement,” ISCC Technical Report 2003-1, Inter-Society Color Council, 11491 Sunset Hills Road, Reston, VA 20190, 2003; see also Carter, E. C., and Billmeyer, F. W., Jr., “Material Standards and Their Use in Color Measurement,” Color Res. Appl., Vol. 4, 1979, pp. 96–100. [34] Color & Appearance Interlaboratory Testing, Collaborative Testing Services, Inc., Herndon, VA 22070. [35] Billmeyer, F. W., Jr., and Alessi, P. J., “Assessment of Color-measuring Instruments,” Color Res. Appl., Vol. 6, 1981, pp. 195–202. [36] Stanziola, R., Momiroff, B., and Hemmendinger, H., “The Spectro Sensor—A New Generation Spectrophotometer,” Color Res. Appl., Vol. 4, 1979, pp. 157–163. [37] Billmeyer, F. W., Jr., and Hemmendinger, H., “Instrumentation for Color Measurement and its Performance,” Golden Jubilee of Colour in the CIE, Society of Dyers and Colourists, Bradford, England, 1981, pp. 98–112. [38] Berns, R. S., and Petersen, K. H., “Empirical Modeling of Systematic Spectrophotometric Errors,” Color Res. Appl., Vol. 13, 1988, pp. 243–256. [39] Optical Industry and Systems Purchasing Directory, Optical Publishing Co., Pittsfield, MA, annually. [40] Modern Paint and Coatings Paint Red Book, Communications Channels, Inc., Atlanta, annually. [41] MacAdam, D. L., “Visual Sensitivities to Color Differences in Daylight,” J. Opt. Soc. Am, Vol. 32, 1942, pp. 247–274. [42] Chickering, K. D., “Optimization of the MacAdam-Modified 1965 Friele Color-Difference Formula,” J. Opt. Soc. Am, Vol. 57, 1967, pp. 537–541. [43] Chickering, K. D., “FMC Color-Difference Formulas: Clarification Concerning Usage,” J. Opt. Soc. Am, Vol. 61, 1971, pp. 118–122. [44] Sève, R., “New Formula for the Computation of CIE 1976 Hue Difference,” Color Res. Appl., Vol. 16, 1991, pp. 217–218. [45] McDonald, R., “Industrial Pass/Fail Colour Matching,” J. Soc. Dyers Colourists, Vol. 96, 1980; Part I, pp. 372–376; Part II, pp. 418–433; Part III, pp. 486–495. [46] News: “CMC Colour-Difference Formula,” Color Res. Appl., Vol. 9, 1984, p. 250. [47] AATCC Test Method 173-1992, “CMC: Calculation of Small Color Differences for Acceptability,” AATCC Technical Manual, AATCC, Research Triangle Park, NC, 1993. [48] Commission Internationale de l’Èclairage, Technical Report 142-2001, Improvement to Industrial Colour Difference Equation, Central Bureau of the CIE, Vienna, 2001. Available from U. S. National Committee of the CIE, c/o Thomas M. Lemons, TLA-Lighting Consultants, Inc., 7 Pond St., Salem, MA 01970, http://www.cie-usnc.org. [49] Luo, M. R., Cui, G., and Rigg, B., “The Development of the CIE 2000 Colour-Difference Formula: CIEDE2000,” Color Res. Appl., Vol. 5, 2001, pp. 340–350. [50] Special Metamerism Index: Change in Observer, Publication CIE No. 80, Central Bureau of the CIE, Vienna, 1989. [51] Hunt, R. W. G., The Reproduction of Colour in Photography, Printing and Television, 4th ed., Fountain Press, Tolworth, England, 1988, distributed by Van Nostrand Reinhold, New York. [52] Judd, D. B., Color in Business, Science and Industry, 1st ed., Wiley, New York, 1952. [53] Longley, W. V., “A Visual Approach to Controlling Metamerism,” Color Res. Appl., Vol. 1, 1976, pp. 43–49. [54] Winey, R. K., “Computer Color Matching with the Aid of Visual Techniques,” Color Res. Appl., Vol. 3, 1987, pp. 165–167. [55] Kumar, R., Billmeyer, F. W., Jr., and Saltzman, M., “Identification of Organic Pigments in Paints,” J. Coat. Technol., Vol.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 43
[56] [57] [58] [59] [60]
57, No. 720, 1985, pp. 49–54; see also Billmeyer, F. W., Jr., Saltzman, M., and Kumar, R., “Identification of Organic Pigments by Solution Spectrophotometry,” J. Coat. Technol., Vol. 7, 1982, pp. 327–337. Kubelka, P., “New Contributions to the Optics of Intensely Light Scattering Materials, Part I,” J. Opt. Soc. Am., Vol. 38, 1948, pp. 448–457. Kuehni, R. G., Computer Colorant Formulation, D. C. Heath, Lexington, MA, 1975. Judd, D. B., and Wyszecki, G., Color in Business, Science, and Industry, 3rd ed., Wiley, New York, 1975. Allen, E., “Colorant Formulation and Shading,” Chap. 7 in Color Measurement, F. Grum, and C. J. Bartleson, Eds., Academic, New York, 1980, pp. 290–336. Allen, E., “Basic Equations Used in Computer Color Matching, II. Tristimulus Match, Two-constant Theory,” J. Opt. Soc. Am., Vol. 64, 1974, pp. 991–993.
Q
COLOR AND LIGHT
557
[61] Gall, L., “Farbrezepturvorausberechnung mit Half von Timesharing,” Farbe und Lack, Vol. 77, Nr. 7,1971, pp. 647–655. [62] Billmeyer, F. W., Jr., and Abrams, R. L., “Predicting Reflectance and Color of Paint Films by Kubelka-Munk Analysis,” J. Paint Technol. Vol. 45, No. 579, 1973; Part I, pp. 23–30; Part II, pp. 31–38. [63] Mudgett, P. S., and Richards, L. W, “Kubelka-Munk Scattering and Absorption Coefficients for Use with Glossy, Opaque Objects,” J. Paint Technol., Vol. 45, No. 586, 1973, pp. 43–53. [64] Phillips, D. G., and Billmeyer, F. W., Jr., “Predicting Reflectance and Color of Paint Films by Kubelka-Munk Analysis. IV Kubelka-Munk Scattering Coefficient,” J. Coat. Technol., Vol. 48, No. 616, 1976, pp. 30–36. [65] Rich, D. C., “Computer-Aided Design and Manufacturing of the Color of Decorative and Protective Coatings,” J. Coat. Technol., Vol. 67, No. 840, 1995, pp. 53–60.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
44
MNL17-EB/Jan. 2012
Gloss
Gabriele Kigle-Böckler1 and Harry K. Hammond III2 THE APPEARANCE OF THE SURFACE OF AN OBJECT or material can be described by its color and gloss characteristics. Like color, gloss can be subdivided into several aspects depending on viewing conditions. In 1937, Hunter identified five aspects of gloss [1] and the functions of reflectance by which they could be evaluated. Later studies by Lex [2] expanded Hunter’s gloss terms and divided them into two groups. One group is based on visual observation with the eye focused on the surface of the material (Fig. 1). For the other group, the eye is focused on the image of the object reflected by the material (Fig. 2). However, investigations have shown that no single objective measurement of gloss will provide perfect correlation with the integrated subjective appraisal of glossiness that the eye so quickly renders. For this reason, the gloss evaluation requirement of an object or material should first be examined and the most useful gloss measurement aspects then selected.
ASPECTS OF GLOSS AND THEIR DEFINITION
The simple term “gloss” is defined in ASTM E284, Terminology of Appearance [3] as “angular selectivity of reflectance, involving surface-reflected light, responsible for the degree to which reflected highlights or images of objects may be seen as superimposed on a surface.” To indicate specific types of “angular selectivity,” such as those involving specular gloss (see below), sheen, or haze, and to illustrate the difference between an evaluation where the focus is on the surface and one where the focus is on the reflected image, the complexity of the phenomenon “gloss” is illustrated in Fig. 3. By focusing on the reflected image of an object, an observer obtains information on the image forming capabilities of the surface. A reflected light source may appear brilliant or diffuse depending on the specular gloss of the surface. The outline of a reflected object may appear distinct or blurred depending on the distinctness-of-image. A halo surrounding the image of the reflected object is an indication of haze. Focusing on the surface of an object provides information on the size, depth, and shape of surface structures contributing to such things as waviness or directionality of brush marks.
Specular Gloss
Specular gloss is defined in ASTM E284 [3] as the “ratio of flux reflected in specular direction to incident flux for a specified angle of incidence and source and receptor
1 2
angular apertures” (Fig. 4). This aspect of gloss has been measured most frequently because it is the one for which an instrument is most easily constructed. In practice the divergence angles of source and receptor are precisely specified in ASTM Test Method for Specular Gloss [4], D523, as are the directional angles of incidence and reflection. Tolerances are specified for all angles. For simplicity, glossmeter geometries are identified by reference to the incidence angles, most frequently 20°, 60°, and 85°. However, the associated source and receptor aperture angles and their tolerances play a vital role in determining the reproducibility of instrument readings. The measurement result is dependent on the amount of light reflected in the specular direction and the refractive index of the surface. ASTM D523 Test Method for Specular Gloss [4] lists the gloss values in dependence of the refractive index. A polished glass with a refractive index of n = 1.550 has a maximum 20° gloss value of 95.4. If the refractive index changes to n = 1.510, the 20° gloss value will decrease to 84.7. Therefore, gloss readings of materials with different refractive indices, such as 1 and 2 K clear coat systems, will not correlate with the visual perception (Fig. 5). As long as the refractive index of the material does not change, a glossmeter will give objective measurement results.
Sheen
Sheen is defined in ASTM E284 [3] as “the specular gloss at a large angle of incidence for an otherwise matte specimen.” The usual angle for sheen measurement is 85°; from the perpendicular to the specimen. This is about the maximum angle that can be used without encountering difficulty in positioning the optics to illuminate and view the specimen at near-grazing angles.
Haze
Haze of opaque products is called reflection haze, while haze of transparent products is encountered as near-forward scattering in transmission that is designated transmission haze. ASTM E284 [3] defines haze in reflection as “percent of reflected light scattered by a specimen having a glossy surface so that its direction deviates more than a specified angle from the direction of specular reflection.”
Distinctness-of-Image
This aspect of gloss is also referred to as “image clarity.” ASTM E284 [3] defines distinctness-of-image gloss as “the
Marketing Manager, BYK-Gardner GmbH, Lausitzerstrasse 8, 82538 Geretsried, Germany. Consulting scientist, BYK-Gardner USA, 9104 Guilford Road, Columbia, MD 21046.
558 Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 This standard is for EDUCATIONAL USE ONLY. www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
Downloaded/printed byASTM International Copyright © 2012 by
CHAPTER 44
Fig. 1—Observer focuses on the image of reflected object.
aspect of gloss characterized by the sharpness of images of objects produced by reflection at a surface.” During visual observation, the sharpness of the light-dark edge of a reflected object can be observed.
Waviness
One obvious type of waviness is designated “orange peel.” ASTM E284 [3] defines orange peel as “the appearance of irregularity of a surface resembling the skin of an orange.” A surface may be described as exhibiting orange peel when it has many small imperfections that are perceived as a pattern of both highlighted and non-highlighted areas. This pattern is interpreted by an observer as a three-dimensional structure of peaks and troughs.
Directionality
ASTM E284 [3] defines directionality, perceived as “the degree to which the appearance of a surface changes as the surface is rotated in its own plane, under fixed conditions of illumination and viewing.” A surface exhibits directionality when specular gloss measurements are a function of the direction for which measurements are made. When paint is applied by brushing in one direction, the brush marks can cause the surface to have a directional characteristic. Metallic materials frequently exhibit directionality when the surface is polished in one direction.
Q
GLOSS
559
Fig. 2—Observer focuses on the illuminated surface of object.
VISUAL GLOSS EVALUATION
Early investigations of gloss were carried out by observing differences in the characteristics of images reflected in the surfaces of specimens. In 1932, the Detroit Paint and Varnish Production Club [5] reported on investigations carried out with their distinctness-of-image gloss comparator. In 1936, Hunter [6] reported an investigation where reflected images of a target pattern were observed. Only in comparatively recent times has there been a major effort to investigate visual scaling of gloss and to endeavor to develop correlation between visual and instrumental measurements. O’Donnell did a doctoral thesis on visual gloss scaling at Rensselear Polytechnic Institute, Troy, New York. Results were first presented, in part, at an ASTM Symposium in 1984 [7] and more fully in a journal article in 1987 [8]. In 1997, the DFO (Deutsche Forschungsgesellschaft für Oberflachenbehandlung) started a research project to investigate which of the physically measurable parameters influence the visual assessment of gloss. The conclusion of this study was that the visual perception is influenced by scattered light reducing the contrast and microstructures distorting an object’s outlines, which results in a lower sharpness. Specular gloss measurement is dependent on the refractive index of the material and the curvature of the sample. Therefore, the visual assessment of gloss requires
Fig. 3—Block diagram depicting relationships of various appearance characteristics. Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
560
PAINT AND COATING TESTING MANUAL
Q
Fig. 4—Reflected light flux distribution from a semigloss surface is depicted by a broken curve. Three lines are shown proceeding from the specimen surface. The center one depicts the specular direction, the image of a ray reflected from a mirror-like surface. The two other rays represent the range of angles passing through the aperture (AP) to the detector (shown as a rectangle).
a multi-dimensional description [9]. At about the same time the ASTM committee E12 on Color and Appearance started a new subcommittee, E12.14, on Multidimensional Characterization of Appearance. The scope of this subcommittee is to develop standards for visual and instrumental assessment of appearance that address the multiple dimensions of appearance. In the first step the committee has conducted several round robin experiments to judge distinctness of image both visually and instrumentally [10].
Development of a Documentary Standard
In 1990, ASTM Committee D-1 on Paint and Related Coatings and Materials published ASTM D4449, Method for Visual Evaluation of Gloss Differences Between Surfaces of Similar Appearance [11], for making visual evaluations of gloss between surfaces of similar appearance. It uses two types of light sources. One source consists of a tubular fluorescent desk lamp modified by placing a matte-black reflecting material behind the tubes and a coarse wire-mesh screen in front. The directions of illumination and view can be adjusted to be 20°, 60°, or 85° as desired for comparing specimens having high gloss, intermediate gloss, or sheen. The other light source is a clear-bulb incandescent lamp. Light from the selected source illuminates the specimens at the chosen angle. The sharpness of reflected images permits a subjective comparison of the relative gloss of similar surfaces.
Use of Landolt Rings to Visually Analyze Distinctness of Image
Landolt rings have been used by ophthamologists to evaluate visual acuity for nearly a hundred years [12]. The test consists of locating the gaps in a graduated series of sizes of incomplete rings whose radial thickness and gap are equal to one fifth the diameter of the ring. For gloss evaluation of trans-illuminated rings, reflections are viewed on a Mylar polyester film. Rings have different sizes and different gap orientations. An image-gloss scale is associated
15TH EDITION
Fig. 5—Specular gloss is dependent on the refractive index of the material: The 1 K and 2 K clear coat system look the same. Due to the difference in refractive index the 2 K clear coat system has a lower specular gloss value than the 1 K clear coat system.
with the different sizes of rings. An image-gloss scale ranging from 10 to 100 in steps of 10 was established for 11 sizes of rings from the largest to the smallest. The development of the scale is not documented, but it ostensibility took place in the General Motors Automotive Division about January 1977. Visual observers select the smallest size of ring for which they can call the gap orientation correctly. The visual judgment is influenced by the loss of contrast and sharpness of the outlines. The numerical size of the rings is used as an inverse index of distinctness-of-image gloss (Fig. 6). ASTM has not published a method for visually evaluating distinctness-ofimage gloss by using Landolt rings, but equipment for this purpose is available.3
Visual Evaluation of Orange Peel
The automotive industry established a physical standard for orange peel consisting of ten high-gloss panels with various degrees of orange peel structure.4 The panels are visually ranked from 1 to 10 with Panel Number 1 depicting very pronounced orange peel and Panel Number 10 illustrating no orange peel. The visual observer can use these panels as a supportive tool to evaluate degree of orange peel. In order to understand this visual ranking it is important to realize that our visual impression is influenced by the structure size and the observing distance (Figs. 7–9)
INSTRUMENTAL MEASUREMENT TECHNIQUES Specular Gloss Measurement
The design of many gloss meters is based on the precise measurement of the specular component of reflected light [13]. A light source, simulating CIE illuminant C, is placed at the focal point of a collimating lens. The axis of the collimated beam is set to the desired angle of illumination. A receptor lens with an aperture in the focal plane followed by an illumination detector completes the basic optical design (Figs. 10 and 11). The size of the receptor aperture and the size of the source image in that aperture are the elements that complete the optics and that 3 Apparatus for evaluating distinctness of image using Landolt rings, available from Paul N. Gardner Co., Pompano Beach, FL. 4 Set of orange-peel panels can be obtained from Advanced Coatings Technology, 273 Industrial Dr., Hillsdale, MI 49242.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 44
Q
GLOSS
561
Fig. 7—Appearance changes with structure size.
Fig. 6—“Landolt Rings.” Broken rings of various sizes are used to provide a scale for visual evaluation of the distinctness of surface-reflected images.3
determine the high, low, and intermediate gloss scale readings. The specular angle largely determines the magnitude of the reflected light. However, the tolerances assigned to the source and receptor apertures are what determine the accuracy and reproducibility of measurements made with instruments having the same angles of illumination and view. Periodic calibration or verification of instrument performance requires the use of calibrated gloss standards. For permanence they should be made of glass or ceramic material. ASTM D523 recommends the use of a primary standard of polished black glass of known refractive index for which the Fresnel (specular) reflectance [14] has been computed for the angle of incidence of the geometry for which the instrument is designed. Since about 1990, national standardizing laboratories have preferred a primary reference standard made of a wedge of clear quartz. Polished black glass working standards can be calibrated by direct comparison of their reflectances with that of a quartz wedge. A single measurement geometry, such as 60°, may not provide instrument readings of gloss that correlate well with visual observations when comparing different gloss levels. This is why ASTM D523 [4] provides for measurement at three different angles of incidence, namely 20°, 60°, and 85° (Fig. 12). Each of the three geometries uses the same source aperture, but a different receptor aperture. The choice of geometry depends on whether one is: (1) making a general evaluation of gloss, (2) comparing high-gloss finishes, or (3) evaluating low-gloss specimens for sheen. ASTM D523 [4] states that the 60° geometry is used for intercomparing most specimens and for determining when
the 20° or 85° geometry may be more applicable. The 20° geometry is advantageous for comparing specimens having 60° gloss values higher than 70. The 85° geometry is used for comparing specimens for sheen or near-grazing shininess. It is most frequently applied when specimens have 60° gloss values lower than 10. The ASTM documentary gloss standard originally published in 1939 contained only the 60° geometry [15]. The desirability of using an auxiliary geometry of 85° for sheen evaluation was recognized shortly afterward. However, the use of another geometry with smaller angles of incidence and view, such as 20°, and a smaller receptor aperture to provide improved differentiation of high-gloss finishes was not published until 1947 [16]. The three geometries, 20°, 60°, and 85°, were originally published as separate ASTM standards. In 1953, ASTM D523 was revised to incorporate all three geometries, and it still does. Meanwhile, the Paint Committee of the International Organization for Standardization, ISO TC-35, was investigating gloss measurements with various commercial instruments prior to drafting an international standard. A paper documenting what the committee had been doing was published in 1976 by the Chairman of the Gloss Task Group, Dr. Ulrich Zorll [17]. In 1978, the ISO Paint Committee published ISO 2813, essentially an international version of ASTM D523 [18]. In keeping with the usual ISO procedure, the standard was made available in English, French, and German. Instrument manufacturers report that measurement precision, reproducibility, and data handling capabilities of gloss meters have been improved markedly in recent years. New instruments have been designed that are smaller, more portable, and more convenient to use. Data storage and analysis are frequently included as well as the capability of electronically transferring data to a personal computer [19,20].
Goniophotometry
“Gonio” means angle, and a photometer measures radiance; so a goniophotometer is an instrument for measuring the angular distribution of reflected or transmitted light intensity. This type of instrument is often used in the research laboratory to investigate the distribution of light flux (Fig. 13). A goniophotometer with appropriate apertures can also be used to provide gloss data for a wide variety of angles and apertures. When goniophotometric measurements are desired, reference should be made to ASTM E167, Practice for Goniophotometry of Objects and Materials [21]. Analysis of goniophotometric curves was treated by Nimeroff [22].
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
562
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 8—Contrast sensitivity of our eyes is highest at three periods per cycle. Therefore, our visual assessment is dependent on the observing distance.
Measurement of Reflection Haze
Haze is a gloss parameter that causes a high-gloss surface to appear milky and lighter, thereby losing contrast. When visually evaluating the reflected image of a hazy object, one observes halos around distinct reflection outlines caused by scattered light. Haze can result from various material or process parameters such as degree of dispersion, flocculation, incompatibility of raw materials (pigment-additive-resin), or from poor application procedures. Haze is most often associated with high-gloss surfaces when small amounts of reflected light are scattered in a
region 1° to 4° from the direction of specular reflection. Therefore, it is useful to place apertures several degrees wide on each side of the specular receptor aperture (Fig. 14). ASTM E430, Test Methods for Measurement of Gloss of High Gloss Surfaces by Goniophotometry [23], describes two methods for evaluating reflection haze, one at 20° and another at 30°. The instrument using a 30° angle is no longer commercially available.5 Since about 1992, a 20° laboratory gloss meter6 was equipped with auxiliary apertures for haze evaluation, thus permitting measurement of 20° specular gloss and haze with the same instrument (Fig. 15).
Fig. 9—At a 40 cm observing distance structures with a wave length size of 3–5 mm can be best seen. While at a 3 m observing distance structures with a wave length size of 10–15 mm can be best seen: Typical orange peel.
Fig. 10—Schematic diagram of a gloss meter. Source is on the left, detector on the right. Lenses are used to provide beam control. Source aperture is designated AP1, receptor aperture AP2.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 44
Q
GLOSS
563
Fig. 12—Diagram depicting the positions of source and receptor for the three geometries of ASTM D523. Fig. 11—Photograph of a modern miniature glossmeter5 (courtesy of BYK-Gardner).
Measurement of Distinctness-of-Image
A variety of different technologies is commercially available to measure Distinctness of Image (DOI) [24–28]. The two most-often used principles are: (1) evaluation of the steepness of the reflection indicatrix or (2) measurement of contrast in dependence of the structure elements. ASTM E430 [23] describes the design of an instrument based on the evaluation of the reflection indicatrix (Fig. 16). The instrument illuminates the specimen at a 30° angle and measures the light reflected at 0.3° from the specular angle with an aperture of 0.3° width. This instrument7 is no longer commercially available. Based on the DFO research project the visual perception of DOI is influenced by loss of contrast and microstructures distorting outlines. Therefore, in 1999 a new measurement technology8 [29] was developed to measure light scattering caused by structure sizes smaller than 0.1 mm. This new measurement parameter was named “Dullness.” A light emitting diode light source and a CCD detector are used to detect the reflected image of the source aperture (Fig. 17). In Fig. 18 two different surfaces are compared. Surface A is a polished black glass standard without dullness. In this case the camera will detect a perfectly sharp image— no light scattering outside the aperture. Surface B shows some light scattering, which results in a lighter image outside the aperture. The dullness measurement determines the amount of light scattering within and outside the aperture in a defined range. The dullness value is a ratio of these two values. Therefore, this measurement parameter is independent of the refractive index. In addition, an adaptive filter is used to separate between the inner and outer image, which minimizes the influence of curvature. Apparatus designated “micro-TRI-gloss,” available from BYKGardner USA, Columbia, MD. 6 Apparatus designated “haze-gloss,” available from BYK-Gardner USA, Columbia, MD. 7 Apparatus designated “Dorigon,” Hunter Associates Laboratory, Inc. VA. 8 Apparatus designated “wave-scan” available from BYK-Gardner USA, Columbia, MD. 5
Microstructures larger than 0.1 mm also influence our impression of DOI. They are in the wavelength range of 0.1–1 mm and can be objectively detected with the same measurement principle as used for orange peel measurement.
Measurement of Waviness (Orange Peel)
The phenomenon of waviness is most observable on a glossy surface, a critical appearance phenomenon in the automotive industry. Waviness has been evaluated by visual means and by use of a profilometer. The correlation between pro-filometry measurements and visual perception is satisfactory for surfaces with similar optical properties. The operation of a profilometer, however, is very time consuming and limited to laboratory use. When the eye of an observer is focused on a painted surface, various types of waviness can be identified that involve size, structure, and shape. Variations in process or material parameters can cause differences in surface structure. For example, poor flow or leveling properties of a coating will usually cause a long wave structure often called orange peel. Changes in substrate roughness, on the other hand, will exhibit a short wave structure of higher frequency. Because waviness is often caused on the production line, it is important to control it there. After considerable research, an instrument was introduced in 1992 to provide an objective evaluation of waviness with structure sizes between 0.3 and 12 mm [30]. In 1999 the measurement principle of this instrument was further developed to increase the resolution and expand the measurement range from 0.1 to 30 mm (Fig. 19). A diode laser source is used to illuminate the specimen at 60°. The reflected light intensity is evaluated at the specular angle. During the measurement the instrument is moved along the surface for a distance of about 10 cm. The intensity of the reflected light is a maximum when coming from a concave structure element. The detector receives less light from a convex structure element. The human eye cannot resolve the actual heights of the structural elements of a painted surface, but the contrast between light and dark areas provides an impression of depth. The contrast of a surface structure can be expressed by use of the statistical parameter “variance.” The final measurement results
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
564
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 13—Schematic diagram of a goniophotometer.
Fig. 14—Schematic diagram of an instrument for reflection haze measurement.
Fig. 15—Photograph of a haze-and-gloss-measuring instrument.6
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 44
Q
GLOSS
565
Fig. 19—Photograph of an orange-peel-measuring instrument.8
Fig. 16—Schematic diagram of an instrument for DOI measurement.7
Fig. 20—Structure spectrum helps to understand appearance of surface finishes
Fig. 17—Schematic diagram of an instrument for dullness measurement.
Fig. 18—Surface A without dullness; Surface B with dullness.
Fig. 21—Original surface with high amount of long-, shortand microstructures: Equivalent to the color gray.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
566
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 22—Structure spectrums of original surface and first improvement step: reduction of longer waves.
are divided into five wave length ranges (Wa–We) using electronic filtering procedures. As a final result a structure spectrum is obtained for structure sizes between 0.1 and 30 mm (Fig. 20). In order to get a complete picture of the surface appearance, this new instrument also includes the measurement principle for dullness. The combination of two measurement principles in one instrument simulates the visual perception at different distances and permits categorizing structure sizes with their causes.
Changes in Structure Spectrum Compared to Changes in Color Measurement
Fig. 23—Surface appearance is dominated by short waves after longer waves were reduced: Surface appears “dull.”
For better understanding of how various structure sizes influence the visual perception, changes in the structure spectrum can be compared to color measurement. The visible spectrum of color measurement is between 400 and 700 nm (blue → red). For simplicity three colors can be allocated to three structure sizes: Long waves correspond to the color red, short waves correspond to the color green, and micro-structures (DOI) correspond to the color blue. The ideal surface would be perfectly smooth, i.e., no waves. In colorimetric terms we would be dealing with
Fig. 24—Structure spectrum with reduced short waves by, e.g., sanding primer surfacer. Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 44
Fig. 25—Very brilliant surface with a low amount of short waves makes long waves very apparent.
the color black (no waves = no color). In the following example a coating with severe waviness is to be improved whereas the main target is to reduce the longer waves. The original surface has a high amount of long, short, and microstructures corresponding to the color gray (Fig. 21). The structure spectrum of the original surface is shown in Fig. 22. A reduction of the longer waves can be achieved by optimizing the basecoat/clearcoat application (e.g., higher clear coat film thickness). As a result the new look will be dominated by the shorter waves, which would be equivalent to a blue/green color (Fig. 23). In practice dominance in shorter waves can be caused by, e.g., a poor-quality primer. This high amount of shorter waves can be reduced by, e.g., sanding the primer. In our example the shortened microstructures were reduced below the amount of longer waves (Fig. 24). In color measurement the color red would be dominating the look. And the same is true in appearance measurement: The very low levels of shorter waves will make the surface very brilliant with a high DOI and consequently the small amounts of long waves will be very obvious (Fig. 25). It is important to realize that the impression of gloss is a multi-dimensional phenomenon. By changing the ratio of different structure sizes, the visual appearance can be dramatically changed. The structure spectrum is like the fingerprint of a surface and can be compared to the spectral curve of colorimetry. Currently the correlation of the different wavelength ranges to the visual perception is under investigation. The goal is to obtain a measurement system for appearance similar to the L*, a*, b* system of color measurement [31].
References [1] Hunter, R. S., “Methods of Determining Gloss,” J. Res. Natl. Bur. Stand., Vol. 18, No. 77, 1937, p. 281. [2] Lex, K., “Die erweiterte Glanzmessung und die Messung von Oberflaechenstrukturen,” Pruftechnik bei Lackherstellung und Lackverarbeitung, Vincentz Verlag, Hannover, Germany, 1992, pp. 70–74. [3] ASTM Standard E284, 1994, “Standard Terminology of Appearance,” Annual Book of ASTM Standards, Vol. 06.01, ASTM International, West Conshohocken, PA. [4] ASTM Standard D523, 1994, “Standard Test Method for Specular Gloss,” Annual Book of ASTM Standards, Vol. 06.01, ASTM International, West Conshohocken, PA. [5] The Detroit Club, “Accurate Gloss Measurement by a Practical Means,” Scientific Section Circular, No. 423, National Paint, Varnish, and Lacquer Association (NAPVA), 1932.
Q
GLOSS
567
[6] Hunter, R. S., “Gloss Investigations Using Reflected Images of a Target Pattern,” J. Res. Natl. Bur. Stand., Vol. 16, 1936, pp. 359–366. [7] O’Donnell, F. X. D. and Billmeyer, F. W., Jr., “Psychometric Scaling of Gloss,” Review and Evaluation of Appearance: Measurements and Techniques, ASTM STP 914, ASTM International, West Conshohocken, PA, 1986, pp. 14–32. [8] Billmeyer, F. W., Jr., and O’Donnell, F. X. D., “Visual Gloss Scaling and Multidimensional Scaling Analysis of Painted Specimens,” Color Res. Appl., Vol. 12, No. 6, 1987, pp. 315–326. [9] Schneider, M., and Schuhmacher, M., “Untersuchung zur Entstehung des visuellen Glanzein-druckes aus den Eigenschaften der Lackoberfläche [Correlation between visual assessment and phyisically measurable gloss parameters],” DFO research report, March 1999. [10] Tannenbaum P. M., “New Routes to surface Appearance assessment—the ASTM E12.14 Approach,” 4th Wave-Scan User Meeting by BYK-Gardner GmbH, Geretsried, Germany, September 1998. [11] ASTM Standard D4449, 1994, “Standard Test Method for Visual Evaluation of Gloss Differences Between Surfaces of Similar Appearance,” Annual Book of ASTM Standards, Vol. 06.01, ASTM International, West Conshohocken, PA. [12] Landolt, E., “Nouveaux Opto-types pour la determination de l’acuite visuelle,” Archives d’Ophthalmologie, Vol. 19, 1899, pp. 465–471. [13] Horst Schene, Untersuchungen über den optischphsiologischen Eindruch der Oberflächenstruktur von Lackfilmen, Springer Verlag, Berlin, 1990. [14] Fresnel, A., “Calcul des Tientes que Polarisation Developpe dan Lames Cristallesees,” Annal Chemie et Physic, Vol. 17, 1821, p. 312. [15] Hunter, R. S., and Judd, D. B., “Development of a Method of Classifying Paints According to Gloss,” ASTM Bull, No. 97, 1939, p. 11. [16] Horning, S. C., and Morse, M. P., “The Measurement of Gloss of Paint Panels,” Official Digest, Federation of Paint and Varnish Production Clubs, ODFPA, Vol. 19, 1947, pp. 153–160. [17] Zorll, U., “Progress towards International Agreement on Gloss Measurement of Paint Films,” Journal of the Oil and Color Chemists Association, Vol. 59, 1976, pp. 439–442. [18] ISO 2813, Paints and Varnishes—Measurement of Specular Gloss of Non-Metallic Paint Films at 20, 60 and 85, International Organization for Standardization. [19] New Product, “The Micro-Gloss Family,” Color Res. Appl., Vol. 15, No. 4, August 1990, p. 242. [20] Paint Red Book, Communication Channels, Inc., 6255 Barfield Road, Atlanta, GA 30328. [21] ASTM Standard E167, 1994, “Standard Practice for Goniophotometry of Objects and Materials,” Annual Book of ASTM Standards, Vol. 06.01, ASTM International, West Conshohocken, PA. [22] Nimeroff, I., “Analysis of Goniophotometric Reflection Curves,” J. Res. Natl. Bur. Stand., Vol. 48, No. 5, pp. 441–448. [23] ASTM Standard E430, Standard Test Methods for Measurement of Gloss of High-Gloss Surfaces by Abridged Goniophotometry, Annual Book of ASTM Standards, Vol. 06.01, ASTM International, West Conshohocken, PA, 1994. [24] Czepluch, W., “Visuelle und messtechnische Oberflaechencharakterisierung durch Glanz,” Industrie-Lack, Vol. 58, No. 4, 1990, pp. 149–153. [25] International Standard ISO 10 215, Anodized aluminum and its alloys—Visual method of image clarity of anodic oxidation coatings. [26] Loof, H., “Goniophotometry with the Zeiss GP 2,” Journal of Paint Technology, Vol. 38, No. 501, 1966, pp. 632–639. [27] Matsuta, M., Kito, K., and Kubota, T., “New Portable Orange Peel Meter for Paint Coatings,” Williamsburg Conference, Feb. 8-11, 1987, pp. 25–28.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
568
PAINT AND COATING TESTING MANUAL
Q
[28] Ladstaedter, E., and Gessner, N., “Die quantitative Erfassung von Reflexionsvermoegen, Verlaufsqualitaet und Glanzschleier mit dem Gonioreflektometer GR-COMP,” Farbe Lack, 1985, Nr. 11, 1979, pp. 920–924. [29] Lex, K., and Hentschel, G., “Neues Verfahren zur Glanz- und Verlaufsstrukturbewertung,” Berichtsband-Nr. 41: Jubiläumstagung 50 Jahre DFO, 14./15. September 1999, Düsseldorf, pp. 73–80.
15TH EDITION
[30] New Product, “Wave-Scan for the Measurement of Surface Structure,” Color Res. Appl., Vol. 18, No. 1, 1993, p. 69. [31] Hentschel G., “Weiterentwickelte wave-scan Technologie: Neue Möglichkeiten der Glanz—und Verlaufsmessung,” Automotive Circle International Conference: Die Ganzheitliche Qualitätssicherung bei der Karrosserielackierung, 19./20. February 2001, Bad Nauheim, pp. 67–113.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
45
MNL17-EB/Jan. 2012
Hiding Power Leonard Schaeffer
CONCEPTS, RELATIONSHIPS, TERMINOLOGY Opacity
WHEN LIGHT ENTERS A PAINT FILM, SOME OR ALL of it is absorbed or reflected by the film before reaching the substrate, thereby hiding the substrate to a lesser or greater degree. The light that reaches the substrate is partly absorbed by it and partly reflected back in conformance with the substrate’s visual pattern. Reflection from the substrate eventually emerges from the film carrying the substrate reflectivity information perceived as visibility or lack thereof, and referred to as hiding. Opacity may be qualitatively defined as the property of a paint film that enables it to prevent the passage of light and thereby to hide the substrate on which it has been applied. Note that opacity is a film property, whereas hiding power is a property of the whole paint. Hiding is a more general term used frequently to refer to either opacity or hiding power (HP).
Light Absorption
If most of the light is absorbed by the film before reaching the substrate, the film is dark in color and hides the substrate well, in which case hiding has been produced by light absorption.
Light Scattering
If most of the light entering the film is reflected back and reemerges without having reached the substrate, the film is white or light in color and hides the substrate well. The reflection mechanism of the film involves multiple internal refractions and reflections that scatter the light to produce a net reversal in its direction. Hiding in this case is produced by light scattering.
Incomplete Hiding
In all cases, the light-absorbing and light-scattering properties of the film act together to produce its opacity. If the film is low in both light-absorption and light-scattering ability, much of the light reaches the substrate. Such a film, therefore, hides poorly and is characterized as being low in opacity.
Test Substrates
An opacity test substrate generally has an ordered pattern of contrasting colors or shades, usually black and white, although black and gray and gray and white are also used. Juxtaposition of contrasting areas permits both visual
1 2
observation and photometric measurement of film opacity. Sealed paperboard charts are the most commonly used substrates of this description. For photometric measurements only, individual black glass and white glass panels are sometimes employed to take advantage of their excellent planarity. Clear plastic can also be used as an opacity test substrate by placing it over black-and-white backgrounds. Black and white painted metal panels are commercially available for use with powder coatings. Standard blackand-white opacity test substrates are defined in paint technology as having CIE-Y reflectances of 0.01 (1 %) maximum and 0.80 (80 %), respectively.2 White test areas are seldom exactly 80 %, but equations are available for correction of photometric values to that and any other standard (see “Kubelka–Munk Equations for Correcting Reflectance and CRn Measurements to a Standard White Substrate Reflectance”).
CR
The extent to which a paint film obscures or hides the contrasting features of the test substrate on which it is uniformly applied is the measure of its opacity. This is expressed photometrically as the ratio of the luminous (CIE-Y) reflectance over the darker, to that over the lighter, area of the substrate, which is referred to as the CR (CR). The Y-reflectance is employed because this CIE parameter is designed to match the sensitivity of the human eye. The CR and the reflectance are expressed as a percentage or as a decimal fraction, the latter to be assumed unless otherwise indicated. A CR value of unity indicates that too little light has reached the substrate for the substrate reflectance characteristics to have a measurable effect on the emergent light flux; thus, there is complete absence of contrast, or complete hiding. Lesser CR values define intermediate levels of contrast, or incomplete hiding. The CR of a given paint film varies with substrate reflectances, and therefore has significance only with respect to a known substrate and primarily to a standard black-and-white substrate as defined in Test Subtrates. In practice, the white area of a commercially available black-and-white substrate normally deviates somewhat from the ideal reflectance of 80 %, whereas the black area is normally 1 % or less, which has no measurable effect on test results, and is therefore treated mathematically as having zero reflectance. Conventional symbols used in this connection are as follows: W = the reflectance of the white area of the test substrate,
The Leneta Company, 15 Whitney Road, Mahwah, NJ 07430. (Author deceased 2002.) CIE = Commission Internationale d’Eclairage. Reflectances are measured with specular (mirror) reflection excluded.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed byASTM International This standard is for EDUCATIONAL USE ONLY. Copyright © 2012 by www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
569
570
PAINT AND COATING TESTING MANUAL
Q
Rw = the reflectance of the paint film over a white area of reflectance W, R0 = the reflectance of the paint film over the black area, Cw = R0/Rw, the CR of the applied paint film, C0.80 = R0/R0.80, the CR when W = 0.80, and C = C0.80, the abbreviation sometimes used in equations. In careful HP measurements, if the white substrate reflectance deviates more than 0.01 from the standard value of 0.80, one of the following correction equations3 is employed. Correction equations3 for normalizing CR values to a standard white substrate reflectance are given in “Calculation of Hiding Power from Tinting Data.” C0.80 = f ( R0 , Rw , W ) =
WR0 (1 – 0.80 R0 ) R0 (W – 0.80) + 0.80 Rw (1 – WR0 )
C0.80 = f (Cw , R0 , W ) =
15TH EDITION
SR and Film Thickness Relationships Let
SR of the coating (equivalent to SR),4 wet film thickness (equivalent to WFT),4 coating density (prior to loss of volatiles), DFT (exclusive of air)5 (equivalent to DFT), dry film density (displacement density),5 Nonvolatile fraction by weight (equivalent to nonvolatile fraction by weight (NVW)), and Nv = nonvolatile fraction by volume (equivalent to nonvolatile fraction by volume (NVV)). H T D t d N
U.S. UNITS
(1)
WCw (1 – 0.80 R0 ) (2) Cw (W – 0.80) + 0.80(1 – WR0 )
Visual Observations of Contrast
= = = = = =
H (ft 2 /gal) × T (mil) = 1604.2
(3)
H (ft 2 /gal) × t(mil) = 1604.2ND ÷ d
(4)
H (ft 2 /lb) × T (mil) = 1604.2 ÷ D(lb/gal)
(5)
H (ft 2 /lb) × t(mil) = 1604.2N ÷ d((lb/gal)
(6)
Although intermediate levels of contrast cannot be directly quantified by visual means, the eye is qualitatively very sensitive to contrast variations. It can identify equalities or nearly complete absence of contrast with considerable precision, which is the basis for the original, as well as several current, HP methods to be described. Indeed, such visual observations are the basic criteria of what constitutes hiding and HP, to which all instrumental hiding measurements trace their validity.
METRIC UNITS6
Film Thickness
U.S.—METRIC UNIT CONVERSIONS
This is usually expressed in thousandths of an inch (mils) or in micrometres (μm). A liquid paint usually contains a substantial quantity of volatiles, so that its dry film thickness (DFT) is substantially less than the original wet film thickness (WFT). The WFT of architectural paints applied in the field are typically in the neighborhood of 3 to 4 mils (75 to 100 μm). With other coating types, it might be as low as 1 mil (25 μm) or as high as 60 mils (1500 μm). With volatile-free liquid coatings, the WFT and DFT are the same except for a possible small increase in density during curing. With powder coatings, for which film formation and curing are concurrent, the term WFT is inapplicable and DFT redundant, so that it is appropriate to refer simply to film thickness.
H (m2 /L ) × T (m) = 1000
(7)
H (m2 /L ) × T (m) = 1000
(8)
H (m2 /kg) × T (m) = 1000 ÷ D(kg/L)
(9)
H(m2 /kg) × t(m) = 1000N ÷ d(kg/L)
(10)
H (ft 2 /gal) = H (m2 /L) × 40.746
(11)
H (ft 2 /lb) = H (m2 /kg) × 4.8882
(12)
T (mil) = T ( m) ÷ 25.4
(13)
D(lb/gal) = D( kg/gal) × 8.3454
(14)
DRY VERSUS WET FILM RELATIONSHIPS7
SR
When paint is applied, whether for test purposes or in actual usage, the area covered per unit quantity of paint is called the spreading rate (SR) for that particular application. When the quantity of coating is expressed volumetrically, as is usual with liquid paints, the SR is usually expressed in square feet per gallon (ft2/gal) or square metres per litre (m2/L). When the quantity is expressed gravimetrically, the SR is usually expressed in square feet per pound (ft2/lb) or square metres per kilogram (m2/kg). SR is inversely related to the film thickness; thus, for a given paint, the lower the SR, the higher the film thickness and film opacity. 3
Derived from Eq (41) by equating to W = 0.80.
ND = Nv d
(15)
t = Nv T
(16)
T = td ÷ ND
(17)
DEFINITION OF HP
Qualitatively, HP is the property of a paint that is manifested as opacity in its films. Quantitatively, it is the SR at Note that WFT and SR, when the latter is expressed by volume, are inverse ways of stating the same information. 5 Refers to films containing no air or hypothetically compressed to exclude air. 6 The following metric notations are identities: kg/L = g/mL = g/ cm3 = g/cc. 7 These are applicable to both common and metric units since the units all cancel. 4
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 45
which the film opacity is just sufficient to give complete hiding over the specified standard black-and-white substrate (see “Test Substrates”). The “complete hiding” point is determined visually in some test procedures and photometrically in others.
Visual HP End-Point
In the visual determination of HP, the operator increases the film thickness gradually and records the amount of paint applied at the supposedly exact point of complete hiding. In practice, instead of perceiving such a point, a range of uncertainty is reached beyond which, when hiding seems unquestionably complete, it also seems that the true end-point has been exceeded. To resolve this dilemma and to obtain repeatable results, the operator chooses an end-point at which it seems that only a negligible increase in film thickness is required to completely obscure the contrasting features of the substrate. This so-called complete hiding end-point is therefore more accurately described as just short of complete hiding.8
Photometric HP End-Point
Uncertainty as to the end-point also exists when measuring HP photometrically. A curve of film thickness versus CR approaches CR = 1 asymptotically, so in theory there is no point at which the contrast is completely obscured. Thus in practice the CR end-point for HP measurements must be less than unity. A CR value of 0.98 is generally accepted in paint technology as representing the point of photometric complete hiding because it is in fact very close to being visually complete, and a higher CR endpoint could not be identified with as much precision. The concept of 98 % as the CR for complete hiding was originally based on the Weber–Fechner law, from which it can be deduced that differences of 2 % in reflectance (with moderate illumination) are imperceptible to the human eye [1]. Actually, this level of contrast, though slight, is definitely visible.
THE ROLE OF PIGMENTS IN HP Binders and Pigments
A typical paint binder, by itself, forms a transparent and virtually colorless film that neither absorbs nor scatters light to any appreciable degree and therefore makes no contribution to the HP of the coating of which it is a part. This task resides entirely in the pigment constituent of the paint. Pigments are fine-particle-size, insoluble, and usually crystalline solids that when dispersed in paint vehicles contribute to the various properties of the mixture, among which are the optical properties of color and HP. Pigments that absorb light strongly over the entire visible spectrum are black; those that are optically selective, absorbing strongly in parts of the visible spectrum and poorly in other parts, are colored, viz. blue, red, yellow, etc., corresponding to the spectral region of nonabsorption. Those that absorb poorly over the entire visible spectrum are white.
It has been proposed to use ∆E of a uniform color space as the HP endpoint instead of CR. This is a valid and feasible concept, although to the writer’s knowledge it has not yet been employed in a published test method.
Q
HIDING POWER
571
White Pigments
When dispersed in a paint binder, some white pigments scatter light strongly and thereby contribute to hiding, while others scatter very poorly and make little, if any, contribution. On that basis, white pigments are classified as hiding pigments or as extenders. White hiding pigments in a paint formulation are sometimes called “prime pigments” as distinguished from the nonhiding “extender” types. The latter are also referred to as “inerts” in view of their apparent passivity with regard to both light absorption and scattering. The difference in scattering behavior between hiding and extender pigments is a function of their refractive indices.
Refractive Index
Most pigments are crystalline in nature. If a single crystal of white pigment were grown sufficiently large, it would be perceived as shiny and transparent like glass, and objects observed through it would look bent and distorted as when observed through a glass prism. This is due to the change in direction, referred to as refraction, that occurs when light passes between media in which it has different velocities, as illustrated in Fig. 1. The relationship between the angles in this figure is expressed by Snell’s law of refraction n = sin i/sin r
(18)
in which i and r are the angles of incidence and refraction, respectively, and n is a constant referred to as the refractive index, which is the ratio of the velocity of light in the incident to that in the refraction medium. If the large pigment crystal postulated previously is pulverized and dispersed in a paint film, each small particle will refract incident light in the same way as described for the large one. Light will also be partially reflected at the surface, and both refractions and reflections will occur within the pigment particle itself. This activity, endlessly repeated with a multitude of pigment particles as illustrated in Fig. 2 (Ref. [2], p. 1), results in the scattering of the original incident light with concomitant film opacity and paint HP. The greater the difference between the refractive indices of the pigment and the surrounding medium, the greater the amount of light scattering that will occur. Refractive indices are reported in Tables 1 and 2 with respect to a vacuum as the medium of incidence. Values with respect to air are practically the
8
Fig. 1—Bending of a light ray by refraction toward the normal as it enters a medium of lower light velocity (higher refractive index).
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
572
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
TABLE 2—Refractive index and relative HP of some extender pigments Pigment
Refractive Index
Relative HP F %a
Barium sulfate
1.64
0.20
Calcium sulfate
1.59
0.08
Calcium carbonate
1.57
0.05
Magnesium silicate
1.57
0.05
Aluminum silicate
1.55
0.03
Silica
1.55
0.03
a
Calculated from Eq (20).
in which it is dispersed. Since the refractive index of a paint binder is, in general, very close to 1.5, Eq (19) can be rewritten as Fig. 2—Light-scattering behavior of a pigmented film.
same. Since white pigments are crystalline in nature, they usually possess different refractive indices along the different crystal axes. Their indices also vary somewhat with the wavelength of the light, generally being higher at the blue (short wavelength) end of the spectrum than at the red (long wavelength) end. Tables 1 and 2 give average values [1,3]. A rough indication of the relative HP of a white pigment can be calculated from its refractive index using the Fresnell equation of reflectivity (Ref. [2], p. 1) ( n – nm )2 × 100 ( n + nm )2
(19)
where F is the Fresnell reflectivity, n is the refractive index of the pigment, and nm is the refractive index of the medium
TABLE 1—Refractive index and relative HP of some white hiding pigments Pigment
Refractive Index
Relative HP F %a
Titanium dioxide (rutile)
2.76
8.8
Titanium dioxide (anatase)
2.55
6.7
Zirconium oxide
2.40
5.3
Zinc sulfide
2.37
5.0
Antimony oxide
2.19
3.5
Zinc oxide
2.02
2.2
White lead carbonate
2.01
2.1
White lead sulfate
1.93
1.6
Lithopone
1.84
1.0
a
Calculated from Eq (20).
( n – 1.5)2 × 100 ( n + 1.5)2
(20)
Tables 1 and 2 illustrate the use of this equation and the general principle that the higher the refractive index of a pigment the greater its HP. The relative HP values shown therein indicate the magnitude of variation related to index of refraction. Other factors can also affect HP substantially, as discussed in “Factors Affecting White HP.”
Extender Pigments
White Hiding Pigments
F % =
F % =
Pigments in this category have low refractive indices in the neighborhood of 1.5. In the form of a powder, with the surrounding medium being air with a refractive index of 1.0, the difference in the two indices produces substantial light scattering, so that extender pigments look white. But dispersed in paint binders, which like themselves typically have a value of about 1.5, they scatter light very poorly and are virtually transparent. This is indicated by the low HP values listed for them in Table 2 as compared with the white hiding pigments in Table 1. Although extender pigments are also referred to as inerts, the latter term is somewhat of a misnomer. They have an indirect but strong influence on light scattering and HP through phenomena referred to as “crowding” and “dry hiding.” They also have important effects on other physical properties of paints such as consistency and gloss.
Colored Pigments
If a pigment absorbs some wavelengths of light more strongly than others, it reflects back a higher proportion of the weakly absorbed wavelengths and is perceived as having the color of the latter (e.g., red, blue, yellow, etc.). Light absorption of this nature is referred to as selective. Colored pigments can vary greatly in HP depending on their light absorption and light-scattering abilities. With regard to light scattering, as with white pigments this is a function of the refractive index or, more specifically, the difference in refractive index between the pigment and surrounding medium. Refractive indices of colored pigments vary widely with wavelength, ranging from 1.3 to 2.7. These variations cause such phenomena as bronzing, dichroism, color
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 45
change with film thickness, and differences in undertone upon dilution with white pigments (Ref. [4], p. 22).
EARLY VISUAL HP METHODS Brushouts
The earliest methods for determining HP employed the practical procedure of brushing the paint uniformly onto combination black-and-white test substrates, increasing the amount of paint in small increments until reaching the point of essentially complete hiding at which the amount of visual contrast was considered negligible. The quantity of paint was determined by weighing the container and brush in grams before and after painting. The corresponding SR is the HP by definition and was calculated from the equation SR(ft 2 /gal) =
test area(ft 2 ) × paint density(lb/gal) × 454 weight of paint (g) (21)
For single-pigment paints, the value calculated from Eq-(21) can be converted to pigment HP using the equation SR pigment (ft 2 /lb) =
SR pigment (ft 2 /gal) pigment concentration(lb/gal)
(22)
Variants of these equations provide for the use of metric instead of U.S., units.
Early Test Substrates
Originally in the study of HP, test surfaces were prepared in individual laboratories by painting black stripes on whitepainted panels. In response to the need for standardized test surfaces, studies were made on oil cloth and linoleum having printed checkerboard-type designs [5]. The Gardner Contrast HP Board was a two-square-foot area glass checkerboard with black and white squares painted on the underside of a thin piece of glass (Fig. 3). The first formalized ASTM method used a linoleum checkerboard surface in the brushout test procedure described in “Brushouts.” The Gardner glass board was used in the same way. Since the “complete hiding” end point in those early methods was
Fig. 3—Gardner contrast HP board.
Q
HIDING POWER
573
determined when the paint was freshly applied and still wet, the resultant HP value pertained only to the wet HP of the paint, not to the dry. This was not a problem in the earliest days of HP measurement, when typical paints contained relatively little volatile constituent and the opacity of the film therefore did not change markedly while drying. But, with the advent of modern paint formulations containing substantial amounts of volatiles, the composition and with it the opacity of the dry paint film could be substantially different than that of the initially applied film. The need to measure dry HP therefore became of paramount importance. As a practical problem in this connection, expensive linoleum and glass test surfaces had to be cleaned for reuse after each test, which made it very difficult to use them for the study of dry HP. This problem was partly overcome with the introduction of paper test charts circa 1931 that were printed in various designs such as checkerboard, concentric diamond-shaped bands, spirals, crescents, etc., and with various degrees of contrast such as black-white, black-gray, gray-white, and a graded series of stripes from black to white. After printing, a coat of nitrocellulose lacquer or other suitable clear sealer was applied. Many of those chart types became and continue to be commercially available.
Contrast Design and Visual Sensitivity
Kraemer and Schupp [6] evaluated contrast surfaces in a variety of designs prepared on glossy photographic paper. These included the customary checkerboard design, a design of narrow 15-mm-wide bands, another with much broader bands, and one with dark half circles on a light background. The results seemed to favor a narrow band design subsequently employed by the Krebs Pigment Co. in preparing the diamond stripe gray-and-white contrast charts illustrated in Fig. 4. The test area of that chart was 1 ft2 (0.0929 m2). The use of a gray-and-white contrast
Fig. 4—Krebs diamond-stripe HP chart.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
574
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
TABLE 3—HP (m2/kg) of some colored pigments measured with a Pfund cryptometera (Nelson and Norris)
Fig. 5—Diagram of early model of Pfund cryptometer.
combination was based on the idea that this is more representative than black and white of the degree of contrast encountered by paints in actual use.
Relative Dry HP—Krebs Method
Although the introduction of paper test charts as replacements for linoleum or glass made dry HP measurements easier, they were still not easy enough. The problem was that it required the preparation of a considerable number of paint-outs at various SRs to obtain one that after drying could be identified with confidence as representing the “complete hiding” end-point. One solution to this problem was to determine comparative or relative dry HP. In the Krebs Pigment Co. method, their square-foot gray-and-white diamond stripe chart was used for that purpose in the following manner: A partial hiding ladder of six to eight brushout standards is made by applying a standard paint at SRs ranging from 400 to 800 ft2/gal (10 to 20 m2/L) and allowing the brushouts to dry. The SRs are precontrolled approximately by syringing specified volumes of paint onto each chart and then determined accurately by weight measurements and calculation as described in “Brushouts.” A single test paint panel is likewise prepared at an intermediate SR. After drying, that panel is compared with the standard panels to determine the two that bracket it in contrast. Then, by visual interpolation, a fairly precise estimate is made of the SR of the standard paint required to match the contrast of the test paint panel. The relative dry HP of the test paint is the percent ratio of its SR to that of the standard paint at equal visual contrast, thus SR Test Paint % Relative Hiding Power = × 100 SR Stdd. Paint
Pfund Cryptometers ALL BLACK
(23)
Introduced in 1919, this was one of the first laboratory instruments made for determining HP [7]. Referring to Fig. 5, A is a plate of black glass whose upper surface is optically flat; B is a transverse groove 10 mm wide and about 2 mm deep. Beginning at the left edge of the groove is a millimeter scale etched in the upper surface of Plate A. C is a plate of clear glass whose lower surface is optically flat. D is a steel shim cemented to C so that a wedge of paint may be formed between the plates. This wedge abruptly becomes infinitely thick at B, and so long as hiding is not complete, the line of demarcation is visible. Sliding the wedge to the
Black Glass
White Glass
Lampblack
. . .
105
Carbon black
. . .
41
Chromic oxide
29
. . .
Prussian blue
72
. . .
Chinese blue
106
. . .
Blue toner
51
. . .
Light green
129
. . .
Medium green
187
. . .
Deep green
88
. . .
Light green
101
. . .
Medium green
167
. . .
Deep green
181
. . .
Light green
62
. . .
Medium green
98
. . .
Light green
154
. . .
Medium green
202
. . .
Deep green
150
. . .
Green toner
91
. . .
Green toner dark
130
. . .
Chrome yellow
23
27
Hansa yellow
31
34
Lt. chrome orange
44
56
Med. chrome orange
17
20
Dk. chrome orange
29
. . .
Lithol toner
66
. . .
Lithol toner
75
. . .
Maroon toner
65
. . .
Madder lake
36
. . .
Toluidine toner
137
. . .
Light para toner
224
. . .
Deep para toner
160
. . .
Light para toner
41
. . .
Deep para toner
35
. . .
a
Multiply by 4.9 to obtain HP in ft2/lb.
left eventually causes the line to disappear. The WFT at the point of complete hiding is determined from the scale reading at the toe of the wedge and the thickness of the shim at the heel, from which the HP in ft2/gal or m2/L can be calculated using Eq (3) or Eq (7). Dark-colored paints cannot be measured using this instrument because of the lack of contrast with the black glass background. Nelson and Norris of the New Jersey Zinc Co. used this cryptometer to determine HP of colored pigments with results as shown in Table 3. The pastes were prepared by rubbing the colors in No. 0000 lithograph varnish. The rubbing was regulated to represent the maximum development usually obtained in practice. In addition to the regular black glass instrument,
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 45
Q
HIDING POWER
575
they used one specially made with white glass for several measurements.
BLACK AND WHITE
Pfund [8] introduced the black-and-white cryptometer in 1930. It eliminated the well, making cleaning easier, and worked for use with paints of any color because of the black-and-white instead of all-black background. Referring to 6, the black glass B and the white glass W are fused along line LM. Longitudinal grooves catch the overflow of paint. The wedge is moved to the right to make the line disappear, then to the left to make it reappear. The position of the plate is reversed and the observations repeated. From the mean of all readings, the hiding is calculated as with the all-black cryptometer (see “All Black”). Comparison of results for white paints shows that the two cryptometer types (allblack and black-and-white) yield the same values within experimental error (Ref. [4], p. 22).
ROTARY TYPE
The rotary cryptometer was a short-lived device designed to overcome the jerky movement of the top plate of the regular cryptometer [9]. The wedge of the cryptometers of Fig. 5 and 6 was replaced with a circular glass plate mounted in a metal frame (Fig. 7). The thickness of the film was read on a scale located on the bottom plate. While the movement of the plate was much smoother with this instrument, it was found that bubbles often obscured the end-point.
ASSESSMENT OF CRYPTOMETERS (REF. [4], P. 25)
The cryptometer is a simple instrument requiring only small quantities of paint, and determinations are quickly made. However, reading the end-point is difficult, and the mean of a number of determinations is therefore advisable. Most users can repeat their own results, but agreement among different users is not satisfactory although it is improved by the use of a standard paint [10]. Another major disadvantage of cryptometers is that they measure only wet HP. One study [11] reported that cryptometers were satisfactory with low-opacity but not high-opacity paints. Consideration of its advantages and disadvantages suggests that the cryptometer is better suited for control work than for specification requirements. The cryptometers shown in Fig. 5 and 6 continue to be commercially available.
Fig. 6—Pfund black-and-white cryptometer.
Fig. 7—Rotary cryptometer.
Hallet Hidimeter
Along with the Pfund cryptometer, the Hallet hidimeter [12] was one of the very early devices for evaluating HP. The objective of a regular microscope is replaced by a long tube fitted with a plain ground glass objective; the eyepiece is replaced with a small hole. The principle of the device is the light-diffusing property of ground glass. If a contrast substrate is viewed through a plate of ground glass, the contrast boundaries become more blurred as the distance between plate and substrate increases. If a liquid paint sample is sandwiched between them, it blurs the boundary further, and the distance required to make the boundary disappear decreases. Since that distance is the thickness of the intervening paint film, it is a measure of the HP of the paint. This measurement is essentially comparative because it cannot be translated into regular HP units.
EARLY PHOTOMETRIC HP METHODS Pfund Precision Cryptometer
In this device (Fig. 8) a photoelectric cell is used to measure the reflectance of paint contained in a wedge-shaped layer [13]. The base plate consists of black-and-white areas B and W, whose boundary is parallel to the length of the plate instead of perpendicular as with the visual cryptometer. The photoelectric device is shifted until a position is found where the reflectance of the paint over the black area is 98 % of that over the white area. The film thickness and HP
Fig. 8—Pfund precision cryptometer. Uses a photoelectric cell instead of the eye to measure reflectance.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
576
PAINT AND COATING TESTING MANUAL
Q
calculations are the same as with the visual cryptometers described in “Pfund Cryptometers.” This cryptometer eliminates the uncertainties of the visual type, as there is no sliding of the top plate over the base plate and no need to estimate visually the appearance and disappearance of an indistinct line. However, it retains the disadvantage of permitting only wet hiding measurements and has therefore been superseded by other photometric devices and methods that permit the measurement of dry HP.
Hanstock Method
Hanstock [14] studied the relationship of light transmission through free paint films to opacity and HP on a black-andwhite substrate. For his transmission work, he employed a flicker photometer and found that paint films were perfectly diffusing and that films having the same degree of light transmission had approximately equal opacity. He further showed the correspondence between refractive index, the Fresnel relationship, and HP. The problem with the transmission concept is that modern paints have so much opacity it is difficult to accurately measure the transmission of films of commercial thickness. Moreover, HP is concerned in practice with paint in intimate contact with opaque surfaces and not as a free film. Consequently, measurement of light transmission through paint films is done today only for very specialized research.
An empirical relationship between SR and CR was found by Fell and reported by Sawyer [15] in the following form (24)
where m and b are experimental constants. Since the graph for this equation is a straight line, it is a simple matter after determining m and b from measurements at two CR levels to find the SR required for any desired CR. This procedure was adapted by Marchese and Zimmerman to determine the HP of paints at a CR of 0.98, and the method was used for many years by the Titanium Pigment Co. (Ref. [4], p. 24). Experience has shown that reasonably satisfactory results can be obtained if the equation is used for interpolation between points close to the desired CR. But, as pointed out by Switzer [16], extrapolation of results can lead to serious errors. He further pointed out that the Fell equation method allows only a single estimate of HP from at least two test applications, thus requiring a considerable effort to obtain an estimate of intralaboratory precision.
New York Paint Club (NYPC) Method
in this method was to determine film thicknesses. To minimize that effort they modified the method by casting films on black and white glass plates and determined thicknesses with an interchemical WFT gage in accordance with ASTM Test Methods for Measurement of Wet Film Thickness of Organic Coatings (D1212). Any error in film thickness, of course, carries over to the HP value. According to Mitton, the revised method sacrifices accuracy and precision for speed [17]. In addition, graphical averaging makes it burdensome to estimate the precision with which the HP has been determined for the same reason pointed out in “Use of the Fell Equation” in connection with the Fell equation method.
Van Eyken–Anderson Method
The method proposed by Van Eyken and Anderson [18] uses CRs and film thicknesses in the same way as the NYPC method described in New York Paint Club (NYPC) Method Section, except that films of different thicknesses are applied in a single operation by using a doctor blade having seven clearances. A die is used to prepare uniform area punch-outs of the paper charts to determine SR by the basic weight-area-density-NVW calculation (see Eq (26)). The defects of this method are that the small areas used for reflectance and weighing make the achievement of good precision difficult (Ref. [4], p. 31), and there is no provision for correcting CR if the white substrate reflectance differs from 0.80.
Federal Test Method for Dry Opacity
Use of the Fell Equation
log(CR × 10) = m × SR + b
15TH EDITION
This method employed doctor blades to apply films at several thicknesses on black-and-white cardboard HP charts. After the films had dried, reflectance, weight, and area measurements were made from which CRs and corresponding WFT were calculated. CR values (rather than log 10 CR as in the Fell equation) were plotted against reciprocal film thickness and the HP calculated from the WFT at 0.98 CR. If the white area of the chart deviated appreciably from the standard reflectance of 0.80, the CR was corrected using Eq (1) or Eq (2). The Club reported that most of the effort
This is Method 4121 of U.S. Federal Test Method Standard No. 141. It is a pass–fail test calling for a minimum dry film CR at a specified WFT. Black-and-white HP charts are used as the test substrate. For routine testing, the paint may be applied either by brush or doctor blade. For referee tests, application is by doctor blade only. The density and the nonvolatile content of the paint are also required. Several drawdowns are made to bracket the specified WFT. The weight of dry paint film is determined for a measured area on each drawdown and the WFT is then calculated from the equation WFT(mils) =
61M(g) A(in.)2 × N × D(g/mL)
(25)
where M = the dry film weight, N = the fractional nonvolatile content of paint by weight, A = the film area, and D = the density of the paint. The CR of each chart is measured and plotted against the corresponding WFT. From a smooth curve drawn through the points, the CR at the specified WFT is obtained. If this is equal to or greater than the specified CR, then the requirement for dry opacity has been met.
GENERAL HP METHODOLOGY Film Application
The objective is to determine the SR at a specified level of dry film opacity, which is usually full hiding as perceived visually or corresponding to the CR: C = R0/R0.80 = 0.98. The basic experimental procedure is to apply a uniform film on
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 45
a suitable test substrate, to observe its opacity either visually or photometrically, and to determine its SR. Since it is not possible to apply a film with precision at a predetermined dry opacity, several such applications need to be made over a range of SRs and their results plotted graphically or otherwise interpolated to the desired HP end-point. This laborious procedure is exemplified in the visual methods discussed in “Brushouts” and “Relative Dry HP–Krebs Method” and in the CR (CR) methods discussed in “Van Eyken–Anderson Method” and “Federal Test Method for Dry Opacity.” The Fell equation and NYPC methods (“Use of the Fell Equation” and “New York Paint Club (NYPC) Method”) attempt to reduce the workload to only two SR determinations by plotting SR against CR or log CR and finding the hiding end-point graphically on the basis of perceived empirical straight-line relationships. Kubelka-Munk (K-M) theory (see “K-M Two-Constant Theory”) shows how the end-point can be calculated with just one SR determination.
SR (or Film Thickness) Determination
In both visual and photometric HP methods, the procedures for observing film opacity are well-defined and can be performed with dispatch. The experimental task most demanding on the operator’s time and ingenuity is to determine the SR or film thickness of the applied coating with good precision. Although gages are available to measure WFT and DFT directly and quickly, the most accurate procedure, by far, is to determine the weight of applied paint film on a measured test area and then to calculate the SR or film thickness as described in “Brushouts” and “Federal Test Method for Dry Opacity.” The equations in both of those methods contain mixed metric and common units. When the units are all metric, the equations are simpler. Letting M = dry film weight and A = the film area and using the symbolism in SR and Film Thickness Relationships: A(cm)2 · N · D(kg/L) H (m2 /L) = 10M(g) T ( m) =
10 4 M(g) A(cm)2 · N · D(kg/L)
(26)
(27)
See “Metric Units” for equations interrelating SR, WFT, and DFT. See “U.S.—Metric Unit Conversion” for conversions between metric and U.S. units. If the volatiles have a relatively low evaporation rate as with most architectural coatings, the film might be weighed rapidly before appreciable loss of volatiles, in which case Eqs (26) and (27) would still apply but with M as the wet film weight and N as unity. The disadvantage of this procedure is that it demands very skillful and speedy manipulation to minimize loss of volatiles before weighing. For that same reason, it is not applicable at all to coatings containing fast-evaporating solvents. With powder coatings, for which the SR is normally expressed on a weight basis, Eq (26) becomes H (m2 /kg) =
A(cm)2 · N 10M(g)
(28)
Assuming negligible volatile content, the value of N in this equation can be taken as unity.
Q
HIDING POWER
577
Photometric Measurements
The CIE-Y reflectance is measured because this function defines the human eye’s quantitative response to the luminous character of light across the visual spectrum. This is valid for chromatic, as well as achromatic colors, as reported by Tough [19], who found good correlation in a large series of colored paints between visual HP measurements and CR values based on CIE-Y measurements with a spectrophotometer. The end-point of 0.98 CR is effective with colors, although it appears that other end-points, for example CIELAB color difference: ∆E = 1.5, would make some difference in the relative HP of various colored paints [20]. However, the simplicity of the 0.98 end-point and its history of validity and general agreement among various workers make it the best choice regardless of color (Ref. [4], p. 31). CIE-Y measurements can be made with the green filter of a tristimulus colorimeter or with a spectrophotometer. When properly standardized, results with the two instrument types should be the same. As a precaution, there should be coordination between correspondent laboratories with regard to instrumentation. In all cases, reflectance measurements must be made excluding surface reflection, which is implicit for instruments designed with 0°/45° geometry and optional with most other instrument types.
CURRENTLY USED TEST SUBSTRATES
The substrate is generally the major factor affecting the specific experimental details of a test procedure. It is selected or specified on the basis of its adaptability to the type of coating being tested and for its perceived advantages in the required or preferred test procedure.
Paperboard Charts
Substrates of this type are described in “Relative Dry HP— Krebs Method.” Their employment with baking finishes is limited because of distortion and discoloration at high temperatures, but they are used widely with air-dried coatings for general HP observations. Black-and-white charts can be used for precision photometric HP measurements by taking appropriate steps to allow for weight variations in the substrate due to humidity and inherent random variations in the area weight of paper. These steps include the use of unpainted control charts and the averaging of multiple test results. Charts with combinations of gray and black, gray and white, and gradations of gray on a white background are used in visual HP tests to obtain what are considered to be more practical HP measurements.
Clear Plastic Film
Polyester is the preferred chemical type. Because of heat distortion, its use is generally confined to air-dried coatings. After the film has dried, a square of convenient size is cut and the area measured. Values of R0 and Rw are read by placing the painted plastic film alternately on a black and a white background with the underside moistened with a suitable liquid (e.g., mineral spirits or dibutyl phthalate) to remove the air interface and establish good optical contact. The dry film weight is determined as the difference in the weight of the painted and unpainted substrate by stripping off the paint film with a strong solvent.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
578
PAINT AND COATING TESTING MANUAL
Q
Glass Panels
Individual black and white glass panels are used to take advantage of the superior levelness of glass for casting of uniform films and because the hard, smooth surface permits rapid WFT measurements with an ASTM-type of WFT gage (New York Paint Club (NYPC) Method). The same information is obtained less rapidly, but with much better precision by scraping off and weighing a defined area of dry film and calculating as described in SR (or Film Thickness) Determination. In some tests, CRs are calculated on the questionable assumption that separate film applications on black and white glass panels are identical in film thickness.
Painted Metal Panels
Panels of this type are generally used with coatings that are applied by spraying and cured by baking. The weight of the applied film is determined by weighing the panel before the coating is applied and again after drying. The SR or film thickness is then calculated as described in SR (or Film Thickness) Determination. If desired, the DFT can be determined without weighing, though with considerably less precision, by direct measurement with a magnetic or an eddy current thickness gage. Black-and-white panels are used for CR measurements or for visual observation of opacity. Mitton has described the use of all-black panels for measuring the HP of baking enamels [21].
K-M TWO-CONSTANT THEORY Introduction
The light that enters a paint film is subjected to scattering and absorption as described in “The Role of Pigments in HP,” and whatever is not absorbed by the film or substrate eventually re-emerges as reflected light. In 1931 Kubelka and Munk [22] published equations defining the optical behavior of a translucent material in terms of two constants referred to as coefficients of scattering and absorption. Steele [23] in 1935 showed how these equations were adaptable to the measuring of paper opacity, and Judd et al. [24] in 1937 did the same in connection with coatings. Kubelka [25] in 1948 rearranged the original equations into new and simplified forms from which Switzer [26] in 1952 developed equations designed specifically for the study of HP by expressing the film thickness (or SR) as a function of the CR. Using these equations and their derivatives, the CR of a coating can be calculated for any SR (or vice versa) from measurements made at only one and its photometric HP thereby determined by a single test application. This is in contrast with the more laborious procedure of obtaining CR values at two or more SRs for interpolation or extrapolation to the HP end-point. The calculations appear formidable, but are readily accomplished with a suitably programmed computer. Graphs and tables are also available for this purpose, although not as convenient and accurate as a computer. The experimental steps are straightforward, and, as with most HP methods, the most difficult and time-consuming operation is to determine the experimental SR (or film thickness) with sufficient precision. How that is accomplished is the essential difference between various K-M-based methods.
Equation Symbols
The symbols used here are based on ASTM Test Method for HP of Paints by Reflectometry (D2805-2003) as follows:
15TH EDITION
G = the substrate luminous reflectance (CIE-Y) For a white substrate G = W. For a standard white substrate G = W = 0.80. For a black substrate G = B. For a standard black substrate G = B ≤ 1 ≈ 0. RG = the luminous reflectance of a film applied over a substrate of reflectance G. R = reflectivity—a property of the paint—the limiting reflectance of the paint film as it is increased in thickness. Also defined as the reflectance at complete hiding as evidenced by R0 = Rw over a black-and-white substrate of uniform film thickness. Cw = the CR of a film applied at uniform thickness over a black-and-white substrate; thus, Cw = R0 /Rw, C0.80 = the CR over a standard black-and-white substrate, thus C0.80 = R0/R0.80. C = abbreviation for C0.80; the two are used interchangeably, thus C = C0.80 = R0/R0.80. T = the film thickness in any stated unit, e.g., μm, mils. H = the SR in any stated unit, e.g., m2/L, ft2/gal, m2/kg, ft2/lb, cm2/g. S = the scattering coefficient, a measure of the ability of the paint to scatter light, expressed in units reciprocal to T or the same as H. K = the absorption coefficient—a measure of the ability of the paint to absorb light, expressed in the same unit(s) as S. e = 2.718 28… the exponential base for natural logarithms. P = scattering power—a measure of the ability of a film to scatter light. A unitless film constant defined mathematically by the relationships: P = ST or P = S/H.
Subscripts
x = an experimentally determined value, e.g., Tx, Hx, Px. c = a value calculated for a specified CR C, e.g., Pc, Hc. 0.98 = a value calculated for C = 0.98, e.g., H0.98, T0.98. H = indicates a value pertaining to a SR e.g., CH , PH , SH. T = indicates a value pertaining to a film thickness, e.g., CT , ST. a and b = simplifying functions of R, defined by a = 1/ 2 (1/ R∞ + R∞ )
(29)
b = 1/ 2 (1/ R∞ − R∞ )
(30)
From these definitions are derived the additional relationships b = (a2 − 1)1/ 2 2
(31) 1/ 2
R∞ = a − b = a − (a − 1)
(32)
Note that R∞, a, and b are three forms of the same constant, so that the determination of any one of them is
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 45
equivalent to determining all three. Sometimes they are used together in the same equation. Additional simplifying functions which can be expressed in exponential-logarithmic form, or using hyperbolic cotangents are ⎛ e2 bp + 1⎞ U = f ( P , R∞ ) = b ⎜ 2 bp ⎟ = b coth bP ⎝ e − 1⎠
(33)
⎛ ⎞ 1 ⎛U + b ⎞ 1 −1 U ln ⎜ ⎟ = coth ⎜ ⎟ 2b ⎝ U − b ⎠ b ⎝b⎠
(34)
in which ln = designation for natural logarithms, viz., logex = ln x, coth = designation for hyperbolic co-tangents, defined by coth x =
e2 x + 1 e2 x − 1
and coth−1 = designation for inverse hyperbolic cotangents, defined by 1 x +1 coth –1 x = ln 2 x −1 Values of natural logarithms and hyperbolic functions are available in published tables and in calculators. Since the tangent function is frequently provided without the cotangent, the relationships between the two are stated here as follows coth x = 1/tanh x,
coth−1 x = tanh−1 1/x
Original K-M Equations
The original equations are as follows: For nonopaque films RG = f (ST , R∞ , G ) =
G / R∞ − 1+ (1 − GR∞ ) e G − R∞ + (1/ R∞ − G ) e
(1/ R∞ − R∞ ) ST
(1/ R∞ − R∞ ) ST
(35)
(36)
The product ST in Eq (35) is a unitless film constant referred to by Kubelka [23] and Judd [22] as the scattering power of the film and symbolized here by the letter P. Thus, given that P = ST and employing simplifying forms of R∞ and the function U of Eq (33), Eq (35) can be rewritten in the much abbreviated form 1 − G( a − U ) a+U −G
(37)
Functional forms are shown in this discussion along with the corresponding explicit forms for a clearer perception of the variables. Sometimes the functional form will be used by itself for both brevity and clarity. Eq (35) shows the reflectance of a paint film in terms of two basic optical characteristics of the paint: the scattering coefficient S, and reflectivity R∞, and two values that are characteristic of the particular application: the reflectance G of the substrate and the thickness T of the film.
(38)
Since film thickness T and SR H are reciprocally interdependent (see Eqs (6) and (7)), it follows that P =STT = SH/H, with the scattering coefficient (ST or SH) being expressed in a unit reciprocal to that of T (e.g., mil−1, μm−1) or in the same SR units as H (e.g., ft2/gal, m2/L, ft2/lb, m2/kg, cm2/g). A clarifying concept in which SR units are mandatory is to consider scattering as an entity quantifiable in area units, with the scattering coefficient as the amount of scattering per unit quantity of coating or coating ingredient, and scattering power as the amount of scattering per unit area of film. SR units have the further advantage over film thickness and reciprocal film thickness of being directly relatable to gravimetric, as well as volumetric quantities. Thus, for understandability, convenience, and standardization, it is preferable to use SR units for scattering coefficients and HP and more specifically the metric SR units m2/L and m2/kg. These are translatable into film thicknesses and U.S. units using the conversion equations in “U.S.—Metric Unit Conversions.” Equations for the numerical conversion of scattering coefficients expressed in various units to standardized metric SR units are given in Table 4.
General K-M HP Method
The experimental procedure, in brief, is to determine the reflectivity R∞ of the paint and R0 and Hx of a nonopaque paint film, from which the scattering coefficient S of the paint is calculated. From R∞ and S is then calculated the SR Hc at any specified CR C or vice versa, or more specifically the SR H0.98 when C = 0.98, which by definition is the HP of the paint. The K-M equations used in these calculations are derived from Eq (38) (the simplified form of Eq (35)) and can be programmed for quick computer solutions. A paint film is applied uniformly over a black-and-white substrate at normal SR (or film thickness) and dried in the manner usual for the particular coating. After drying, the
whose converse and more useful form is K / S = f ( R∞ ) = (1 − R∞ )2 / 2 R∞
579
Scattering Coefficient and Scattering Power
DETERMINATION OF REFLECTIVITY R∞
For opaque films R∞ = f ( K / S) = 1+ K / S − ( K 2 / S2 + 2 K / S)1/ 2
HIDING POWER
RG = f (U , R∞ , G ) = f ( P , R∞ , G ) =
and the converse of Eq (31) P = f (U , R∞ ) =
Q
TABLE 4—Unit conversion equations for scattering coefficients. S(m2/L) = S(ft2/gal) ÷ 40.746 S(m2/L) = S(mil−1) × 39.37 S(m2/L) = S(μm−1) × 1000 S(m2/L) = S(mm−1) × 1 S(m2/kg) = S(ft2 /lb) ÷ 4.888 S(m2/kg) = S(cm2/g) ÷ 10 S(m2/L) = S(m2/kg) × D(kg/L)
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
580
PAINT AND COATING TESTING MANUAL
Q
reflectance values R0, Rw, and W are measured. If the CR Cw=R0/Rw is less than 0.96, the application is repeated as a second coat or at a somewhat higher film thickness. A porous film should not be recoated, nor an impractically high film thickness applied in a single coat, due to the possible effect on R∞. If the original or recoated film hides completely, then R0 = Rw = R∞. If not, calculate R + W − Rw ⎞ 1⎛ ⎟⎟ a = f ( R0 , Rw , W ) = ⎜⎜ Rw + 0 WR0 2⎝ ⎠
(39)
(
2
)
1/ 2
The preceding two equations may be programmed sequentially to give R = f ( R0 , Rw , W )
2
H (m /kg) =
Having determined R∞ of the paint and R0 and Hx of the testfilm, H0.98 is obtained by the following sequence of calculations: (a) The scattering power, Px of the test film is calculated from Px = f ( R0 , R) =
(40)
or
(44)
1/ 2
2 ⎡⎛ ⎤ 1− C ⎞ 1 ⎥ 1− C U0 = f (C , R∞ ) = ⎢⎜ a + + ⎟ − ⎢⎣⎝ 1.60C ⎠ C ⎥⎦ 1.60C
(45)
and Eq (34) P = f (U , R∞ ) which are solved sequentially to give Pc = f (C , R∞ )
(45)
(d) The HP H0.98 is then calculated from Hc = S/Pc where C = 0.98.
(41)
CR AT A SPECIFIED SR
Having obtained a film within the specified CR range, R0 is recorded and Hx is determined by a suitable method. Various techniques for determining the SR are available, but the most precise is a weight-area-density-NVW method as discussed in “SR (or Film Thickness) Determination,” using applicable Eqs (26) and (27). The dry film weight M in those equations is usually obtained as the difference in the weight of the test area before and after application of the paint. Sometimes, as with black glass, it is obtained by removing a known area of film and weighing it directly. Such weighings can be performed on an analytical balance with great accuracy. The density D and nonvolatile N of the paint must, of course, also be determined if not already known. With considerably less precision, the thickness of the dry film can be measured using a caliper or electronic gage on a metal panel, from which the SR can be calculated using one of the following relationships 1000ND t(m) · d
⎛1 − aR ⎞ 1 ⎛ 1 – R R ⎞ 1 0 0 ∞ ⎟⎟ = ⎟⎟ coth –1 ⎜⎜ ln ⎜⎜ b ⎝ bR0 ⎠ 2b ⎝1 – R0 / R∞ ⎠
(b) The scattering coefficient S of the paint is calculated from S = PxHx. (c) The scattering power Pc of a paint film at the CR C = 0.98 is calculated from
This requires the application of a uniform film at a SR (or film thickness) such that the CR Cw is within the range of 0.96 to 0.985. These limits are established because too low a CR requires excessive extrapolation to the C = 0.98 endpoint, and higher CR values become increasingly insensitive to SR (or film thickness) variations. If the initial application is outside that range, the application is repeated at a higher or lower film thickness, as required. The film may be applied on a black-and-white or an all-black substrate. If black-and-white, then the test application can be the same one used for determining R∞ in Determination of Reflectivity R∞. If an all-black test surface is employed, the indicated CR range is still required, but since it cannot be measured directly, it is calculated from
H (m2 /L) =
(43)
where N = the nonvolatile fraction by weight NVW of the test paint, D = the density of a liquid paint, d = the density of the dry or cured film, and t = the thickness of the dry or cured film.
DETERMINATION OF R0 AND HX
R0 (1 − 0.80 R0 ) C0.80 = f ( R0 , R∞ ) = R0 + 0.80(1 − 2 a R0 )
1000N t(m) · d(kg/L)
CALCULATION OF H0.98
and from Eq (32) R∞ = a − a − 1
15TH EDITION
(42)
Although this is not HP as such, it is frequently used as an alternative HP criterion. After Step (b) of Calculation of H0.98, calculate the scattering power P at the specified SR H from: PH = S/H, then calculate the CR CH from Eq (33) UH = f ( PH , R∞ )
(46)
and CH = f (UH , R∞ ) =
a + U − 0.80 ( a + U )[1 − 0.80( a − U )]
(47)
which together give CH = f ( PH , R∞ )
Judd Graph (Information Included for Historical Purposes)
(48)
Prior to the availability of modern computers, K-M equations were much too complex for a practicable HP test method. Judd [24], therefore, laboriously worked out a general solution to Eq (35) in the form of a graph repro-
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 45
Fig. 9—Judd graph derived from Kubelka–Munk Eq (35).
duced in Fig. 9. The graph relates the four variables R0, C0.80, R∞, and P so that from any two of them the other two can be determined. It consists of two families of curves for constant values of R∞ and P, plotted on the coordinates R0 and C0.80. The P curves were referred to in the original Judd graph as curves of SX (or ST). The portion pertinent to white paints (R∞ ≥ 0.75) has been enlarged and is shown in Fig. 10. Experimentally, R0, RW, and Hx are determined as in Determination of R0 and Determination of Reflectivity R∞ for a film applied uniformly on a black-and-white test substrate. If W deviates from 0.80 by more than 0.01, C0.80 is calculated using correction Eqs (1) and (2). Px and R∞ are determined at the graph point corresponding to C0.80 and R0, and the scattering coefficient of the paint calculated from S = PxHx. The value of P0.98 is located at the intersection of the R∞ curve with the vertical line for C = 0.98. The HP is then calculated from H0.98 = S/P0.98. If desired, the SR can be determined for CRs other than 0.98 in the same way. Conversely, CH may be determined for any specified value of H by first calculating: PH = S/H, then finding the desired value of CH at the intersection of the curves for the determined PH and R∞. The Judd graph is also useful for depicting the basic optical properties of paints. It shows that paints with high S values are lighter over black backgrounds than paints of the same reflectivity with low S values. Also, if their S values and film thicknesses are the same, paints with high reflectivity are poorer in hiding than paints of low reflectivity. The latter fact may be demonstrated as follows: Suppose a portion of paint for which R∞ = 0.85 is tinted with a black colorant to an R∞ value of 0.78 and the untinted and tinted paints are applied at the same thickness such that P = 5.0. From Fig. 9, or more accurately from Fig. 10, it can be determined that the colorant addition has increased the
Q
HIDING POWER
581
Fig. 10—Judd graph derived from Kubelka–Munk Eq (35)— a portion of Fig. 9 enlarged.
CR to 0.965 from its original value of 0.945, representing a considerable increase in visual film opacity. To determine what this amounts to in terms of photometric HP, the P values of the untinted and tinted paints at the intersection of their R∞ curves with the vertical line C = 0.98 are found to be 7.5 and 6.0, respectively. Since S is unaffected by tinting, the SR change at C = 0.98 is calculated as Htinted P 7.5 = untinted = =1.2 25 Huntinted Ptinted 6.0 representing an increase of 25 % in HP by tinting to a lower R∞ value. This hiding increase was obtained at negligible monetary cost but at a sacrifice in paint quality in regard to brightness of appearance. For that reason, in evaluating a series of paints experimentally, a fair comparison requires that all R∞ values be adjusted by tinting to that of its lowest reflectivity member. Examination of the Judd graph shows that, after adjustment to the same R∞ value, films of the different paints applied at the CR C = 0.98 all have the same P0.98 value and, since H0.98= S/P0.98, their HPs will be directly proportional to their scattering coefficients. Thus the scattering coefficient alone can be an adequate HP comparator, without actually tinting the individual paints.
Mitton Graph and Table [27] (Information Included for Historical Purposes)
As with the Judd graph, these provide precalculated solutions to K-M equations, but with much greater precision. They were designed for the experimental procedure described in “General K-M HP Method,” in which R0 and Hx are determined for a film applied on an all-black test surface, and R∞ is determined in a separate test application. The test surface of choice is black float glass because the
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
582
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 11—Mitton graph of Kubelka–Munk Eq (44).
extremely level nature of the surface permits the application of very uniform films with a doctor blade. Mitton also describes the use of all-black metal panels to test spraying/ baking-type finishes [21]. The graph is derived from Eq (44): P = f(R0, R∞) and is plotted as a family of curves at constant R∞ on coordinates of scattering power P and reflectance R0. The ordinate is indicated as ST (Factor B), which is the same as P, and the abscissa as RB, which is usually, and in this case necessarily, the same as R0. It consists of a small-scale index graph (Fig. 11) divided into 31 sections, each then expanded to a much larger scale on a separate sheet. Fig. 12 shows one of the expanded sections. Associated with the Mitton graph is a table of factor A values derived from Eq (46): Pc = f(C, R∞), in which factor Ac = 1604.2/Pc. Values of factor Ac are given in this table for C = 0.98, 0.95, and 0.93, for all values of R∞ from 0.08 to 0.98 (8 to 98 %). The C value of most interest for HP calculations is 0.98, representing full photometric hiding as defined in Section Photometric HP End-Point. If desired, Pc is easily calculated from factor Ac. The graph and table are typically used as follows: After determining R∞ and R0 experimentally, the index graph is consulted in order to select the appropriate expanded graph on
which the scattering power Px of the experimental paint film is to be found. Factor A0.98 is determined from the table for the measured value of R∞. At this point either the film thickness Tx, or SR Hx of the test film associated with R0 is determined. If, as Mitton intended, Tx is determined in mils, then the scattering coefficient S is calculated in reciprocal mils from S = Px /Tx, and the HP is calculated from the equation
(
) (
)
H0.98 ft 2 / gal = S mil −1 · A0.98 The preceding simple relationship holds when S is expressed in reciprocal mils and HP is expressed in ft2/gal. If the SR in m2/L is determined instead of the film thickness, then after determining Px and R∞, the scattering coefficient is calculated as: S (m2/L) =Px · Hx(m2/L) and the HP calculated from H0.98 (m2 /L) =
S(m2 /L) · A0.98 S(m2 /L) = 1604.2 P0.98
At a later date, Mitton commented that graphical and tabular aids for K-M calculations had become unnecessary
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 45
Q
HIDING POWER
583
Fig. 12—Mitton graph—expansion of Sector 5 in Fig. 11.
with the advent of inexpensive programmable calculators [28]. Nevertheless, the Mitton graph and table continue to be used in a number of important test methods, and both the Judd and Mitton graphs are useful for instructional purposes. Currently, software is available with color-measuring instruments, and it can be used to carry out these calculations. For this reason, the graphical techniques are seldom used today.
Typical K-M HP Results
Tables 5 and 6 are based on the testing of various commercial paints and pigments. They are intended to illustrate magnitudes of HP and scattering coefficient values encountered in K-M HP measurements. The scattering coefficient values are intended to supplement and clarify, by specific examples, the relationships shown in Table 4. With regard to pigments are (Table 6), it is of course dispersions that are actually measured and the values for the pigments then calculated from their concentrations in the dispersions. For example Hpigment (m2 /kg) S pigment (m2 /kg g) 1 = = 2 2 Hcoating (m /L) S coating (m /L) Pigment Conc.(kg/L)
TABLE 5—Powder coatings—Representative HP and scattering coefficient data. White
Light Gray
Orange
R∞
0.8234
0.6860
0.4389
P0.98 (unitless)
7.011
4.655
2.449
H0.98 (m2/kg)
18.09
20.22
0.26
Density (kg/L)
1.60
1.66
1.41
T0.98 (μm)
34.55
29.79
69.12
S (m /kg)
126.8
94.13
25.13
H0.98 (ft2/lb)
88.3
98.7
50.1
Density (lb/gal)
13.35
13.85
11.77
T0.98 (mils)
1.36
1.17
2.72
S (ft /lb)
619.2
459.5
122.7
Metric Units
2
U.S. Units
2
The values in Table 6 should be considered as comparative because the scattering coefficient of a pigment can vary widely depending on the conditions of measurement (Ref. [4],
NOTE: Derived from test results obtained by ASTM Subcommittee D01.51 on Powder Coatings.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
584
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
TABLE 6—White pigments—HP and scattering coefficient valuesa Lead Carbonate
Zinc Oxide
Zinc Sulfide
Anatase TiO2
Rutile TiO2
R∞ (estimated)
0.91
. . .
. . .
. . .
. . .
P0.98 (unitless)
9.5
. . .
. . .
. . .
. . .
H0.98 (m2/kg)
3.1
4.1
11.9
23.5
30
S (m /L)
29
39
113
223
285
H0.98 (ft2/gal)
15
20
58
115
147
S (ft /lb)
140
190
550
1090
1390
Metric Units
2
U.S. Units
2
a
Based on reported hiding power values at a pigment volume concentration (PVC) of 28 % [30,31].
p. 34), being affected by pigment volume concentration (PVC), effectiveness of dispersion, the presence of other pigments in the same dispersion, and the nature of the vehicle. Even within a specific chemical class it can vary considerably depending upon the particular method of manufacture employed. Nevertheless, it is frequently useful to determine pigment HP values for a comparison of their efficiency under specified conditions.
Theoretical Problems and Practical Considerations
The validity and usefulness of the K-M equations in HP calculations are predicted on the constancy of the scattering coefficient S over a suitably wide film thickness range. Judd [24] studied this question in connection with water-borne paints and vitreous white enamels and concluded that at practical film thicknesses S is constant within experimental error. The author has experimentally obtained essentially constant S values within a WFT range of 50 to 100 μm (2 to 4 mils), equivalent to a SR of 10 to 20 m2/L (400–800 ft2, /gal), with white alkyd gloss, latex gloss, and latex flat paints. Moreover, the effect of any variation of S with film thickness that might occur is minimized in experimental practice by casting films with CRs fairly close to the 0.98 CR hiding endpoint, as called for in Determination of R0 and Hx. This is not difficult to do. Refractive indices and resultant scattering coefficients vary with the wavelength of light. Thus, the effective scattering coefficient of a paint is actually an average for all of the encountered wavelengths. With achromatic paint films, the wavelength composition of the light flux remains constant and, therefore, so does the scattering coefficient upon which constancy the validity of K-M equations is predicated. Chromatic paint films, however, absorb light selectively and therefore change the composition of broadband illuminants with a resultant change in the effective scattering coefficient as the light passes through the film. This would, in theory, appear to disqualify chromatic paints from K-M HP calculations. In practice, however, the equations are used successfully for that purpose (Ref. [29]; Ref [4], p. 31), which is undoubtedly related to the previously noted fact that the experimental measurements are made fairly close to the HP end-point (C = 0.98), so that the K-M extrapolation and thus any associated error is relatively small.
As discussed by Mitton (Ref. [4], p. 27), the use of K-M theory for HP calculations has been questioned because it is phenomenological rather than based on fundamental theoretical considerations, and the measurements and equations omit needed corrections for surface reflection that are theoretically substantial. However, in experimental practice the errors are generally small despite the theoretical defects. Simpson took note of this in his comment that when uncorrected values of S and K are inserted back into the uncorrected K-M equations, “it would appear that an approximately correct answer is obtained” (Ref. [2], p. 111).
Calculation of HP from Tinting Data
Initially, the S and R∞ values of a standard white paint are determined in accordance with the procedure described in “General K-M HP Method.” The K value of the paint can then be calculated from Eq (37): K/S = (1 – R∞)2/2R∞. From Scattering Coefficient and Scattering Power, S and K can be considered as concentrations of “scattering” and “absorption” per unit weight or volume. The K value of a black tinter is determined by adding a measured ratio to the standard paint sufficient to reduce the reflectivity to about 0.40. The K value of the tinted paint is its initial K value plus the tinter contribution, thus K2 = K1 + XKt
(49a)
XKt = K2 – K1
(49b)
from which
and dividing through (b) by the common value of S XKt /S = K2 /S − K1 /S
(49c)
in which Kt = the absorption coefficient of the tinter, X = the ratio of tinter to paint, K1 = the initial K value of the paint, K2 = the K value of the paint after tinting, and S = the scattering coefficinet of the paint. The ratios K2/S and K1/S are calculated from measured values of R∞ for the tinted and untinted paints using Eq (37): K/S = (1 – R∞)2/2R). If the standard paint is an untinted white with a reflectivity no lower than 0.93, then its absorp-
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 45
HIDING POWER
Q
585
after tinting and the four corresponding K/S values calculated from Eq (30). Then, as in Eq. (49c) of “Calculation of HP from Tinting Data,” for paint A: XKt /SA = KA2/SA – KA1/SA
(50a)
XKt / SB = K B2 / SB − K B1 / SB
(50b)
for paint B
Next, dividing Eq (49b) into Eq (49a), X and Kt cancel to give SB / SA = ( K A 2 / SA − K A1 / SA ) ÷ ( K B 2 / SB − K B1 / SB ) (50c) If the comparison paints are both untinted high reflectance whites then, as pointed out in “Calculation of HP from Tinting Data,” the untinted K-values can be considered negligible and Eq (49c) becomes
Fig. 13—Scattering coefficients determined by tinting and by HP tests.
tion contribution K1 is considered negligible compared with that of the tinter, in which case K1/S is dropped from Eq (49c) to give XKt /S = K2 /S
(49d)
The absorption coefficient Kt of the tinter can be calculated from Eqs (49c) or (49d) since all other terms in these equations are known [30,31]. Having determined Kt, the S-value of a test paint can be determined using the same tinting procedure and equations as before, but this time calculating unknown S from known Kt, instead of vice versa. With the values S and R∞ of the test paint having thus been determined, its HP H0.98 can be calculated as in Calculation of H0.98 (c) and (d) without the tedious requirement of measuring the SR. Experimental evidence for the validity of this procedure is given by Mitton and Jacobsen [32], who, equating the tinting strength of a white pigment with its scattering coefficient, measured S (cm2/g) for a number of white pigments by direct HP measurement and by the tinting procedure. As shown in Fig. 13, the correlation between the two methods is very close. If this simplified method is to work, the K value of the black tinter must be the same in any paint being tested. Also, the tinter must not change the degree of dispersion of the white pigment so as to cause a change in its S value. These conditions are not always met, so that it is safest to apply the method only under favorable circumstances, when interaction of tinter and paint are known to be negligible.
Determination of Relative HP of Untinted White Paints from Tinting Data
For this purpose there is no need to determine the K value of the tinter as in “Calculation of HP from Tinting Data.” An equal ratio of black tinter is added to comparison paints A and B, sufficient to reduce their R∞ values to about 0.40. The R∞ values of paints A and B are measured before and
SB / SA = K A 2 / SA ÷ K B 2 / SB
(50d)
As stated at the end of Judd Graph, at the same reflectivity R∞, the HP values of paints A and B will be in the same ratio as their scattering coefficients.
K-M EQUATIONS FOR CORRECTING REFLECTANCE AND CR MEASUREMENTS TO A STANDARD WHITE SUBSTRATE REFLECTANCE
The reflectances of applied paint films, and hence their CRs, vary with the substrate reflectance. In practice, black substrates are effectively zero and constant, but white substrates can vary appreciably. Normalization equations that correct for this variation, derived from Eq (39) of “Original K-M Equations,” are as follows:
RG = f ( Rw , R0 , W , G ) = Rw −
CG = f ( Rw , R0 , W , G ) =
(W − G )( Rw − R ) (51a) W (1 − GR0
WR0 (1 – GR0 ) R0 (W – G ) + GRw (1 – WR0 )
CG = f (Cw , R0 , W , G ) =
(51b)
WCw (1 − GR0 ) (51c) Cw (W − G ) + G(1 − WR0 )
TABLE 7—Scattering coefficient of a 20 % PVC TiO2-alkyd paint film versus crystal size of pigment. S, μm−1
Mean Crystal Size, μm
0.76
0.24
0.73
0.20
0.64
0.16
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
586
PAINT AND COATING TESTING MANUAL
Q
where R0 = reflectance of the coating over the black area W = measured white substrate reflectance Rw = measured reflectance of the coating over the white substrate Cw = measured CR G = standard white substrate reflectance Nominal:0.80 RG = normalize reflectance of the coating over the white substrate CG = normalized CR
FACTORS AFFECTING WHITE HP
As shown in Table 1 and Table 6, rutile titanium dioxide is by far the most effective of the white hiding pigments in producing light scattering and HP, which is true on a cost as well as a weight basis. This fact has effectively eliminated the use of other white hiding pigments except for special properties or considerations. The important variables that determine the scattering and hiding efficiency of a titanium dioxide pigment in a paint are: (1) its mean crystal and particle size, (2) the state of pigment dispersion, (3) its concentration in the paint film, and (4) film porosity.
Crystal and Particle Size
By decreasing the particle size of the pigment, the number of particles and surfaces for light reflection and refraction increase, and the light-scattering ability of a given quantity of pigment will, therefore, tend to be enhanced. However, if the particle size is too small in relation to the wavelength of light, the wave front passes around rather than through it, so there is no light scattering, and the dispersion is transparent. Obviously, there is some intermediate optimum size related to the wavelength of light at which maximum scattering efficiency is obtained. The wavelength of the visible spectrum ranges from approximately 0.4 to 0.7 μm, peaking in luminosity at 0.55 μm. The mean crystal size for maximum opacity ranges from approximately 0.20 to 0.30 μm depending on both the PVC and the fraction of the pigment consisting of single crystals. Commercial grades of titanium dioxide developed for high-gloss finishes exhibit a singlecrystal content of about 20 % and have a mean crystal size between 0.22 and 0.24 μm. The adverse effect of a lesser crystal size in such formulations is shown in Table 7 [33].
15TH EDITION
or floccules due to weak forces of cohesion. Floccules are easily broken down but can spontaneously and quickly recur in the wet paint or drying paint film. Despite their weak bonding, floccules have the optical effect of increasing the mean particle size, thereby decreasing the scattering efficiency of the pigment. An auxiliary phenomenon related to increased particle size is the preferential scattering of longer wavelengths. Balfour and Hird took advantage of this phenomenon to quantify pigment flocculation by measuring backscattered infrared radiation (wavelength 25 μm) from a dried paint film to obtain what they refer to as a “flocculation gradient” [34,35].
Pigment Concentration
In “Refractive Index,” it was pointed out that a very large single crystal of a white hiding pigment is actually transparent. Without undertaking a theoretical analysis, it is to be expected that as the concentration of pigment increases and its particles become more crowded, they approach the optical condition of a very large particle with resultant loss of scattering efficiency and HP. The crowding effect was studied by Stieg [36–38], whose results were used by Mitton (Ref. [4], pp. 34–35) to draw curves of HP H0.98 versus PVC for pure rutile and anatase titanium dioxide in alkyd enamels. These are shown in Fig. 14, in which HP is expressed in ft2/lb of nonvolatile matter. If the paint is formulated at 50 % nonvolatile by volume, the HP results would be half that shown in Fig. 14, but the shape of the curves would be unchanged. Note the maximums in the curves at 25 to 30 % PVC, above which HP actually begins to decrease with increasing concentration of pigment. When calculated in terms of ft2/lb of pigment, the results appear as shown in Fig. 15, clearly indicating the drastic decrease in TiO2 efficiency due to crowding. Stieg [36] found empirically that the relationship between TiO2 HP and PVC, as shown in Fig. 15, could be expressed by the equations rutile:H0.98 (ft 2 /lb) = 370 − 410(PVC)1/3
(52)
anatasse:H0.98 (ft 2 /lb) = 290 − 330(PVC)1/3
(53)
The PVC values in these equations are decimal fractions. Expressed in metric units, the equations become:
Pigment Dispersion
The process of obtaining a satisfactory dispersion involves the wetting of the pigment by the dispersion medium to displace air, breakdown of larger particles by milling, and stabilization after the dispersion has been obtained. With alkyd media, standard grades of titanium dioxide disperse easily and develop full hiding with very little milling. Thus, the main reason for milling alkyd dispersions is to reduce or eliminate oversize particles that affect the appearance of the film. With latex paints, milling can have an important effect on opacity depending on the grade of pigment employed [33], but the appearance factor is also an important consideration, particularly with semigloss and gloss finishes. A major factor affecting the efficiency of TiO2 in the completed formulation is the phenomenon referred to as flocculation, which is the formation of large particle groups
Fig. 14—HP H0.98 (ft2 /gal) of solids at various PVC levels.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 45
HIDING POWER
Q
587
dioxide for the esthetic purpose of producing a flat finish and to contribute HP by means of porosity. Stieg and Ensminger [38] showed that with paints over the CPVC that contain both TiO2 and extender, HP is in a straight-line relationship with the Porosity Index (P.I.), with the slope of the line depending on the percentage of prime pigment and the nature of the extender. The P.I. is calculated from the equation P.I. = 1
CPVC(1 – PVC) PVC(1 – CPVC)
(56)
The low-cost HP obtained from porosity is unfortunately accompanied by a deterioration in the quality of the film as manifested by poor scrub, soil, and stain resistance. This is due to an insufficiency of binder, resulting in an air phase continuum that gives ready capillary access to staining materials.
MICROVOIDS FOR WHITE HP Fig. 15—HP H0.98 (ft2/lb) of pigment at various PVC levels.
rutile: H0.98 (m 2 /kg) = 75.7 – 839(PVC)1/3
(54)
anata ase: H0.98 (m 2 /kg) = 59.3 – 67.5(PVC)1/3
(55)
The question has been studied [36,39] of whether extenders added to a gloss or semigloss paint film might tend to increase the spacing of the TiO2 pigment and thereby its scattering efficiency. The physical picture that emerges is of large particle-size extenders acting as massive intrusions having no effect on the original TiO2 spacing, and of fine particle-size extenders dispersing uniformly so as to increase TiO2 spacing, but no differently in this respect than an equal volume of binder. Consequently, when binder is replaced by an equal volume of large particle-size extender, TiO2 efficiency decreases, whereas with small particle-size extenders, TiO2 efficiency has been found to remain essentially the same and in no case improved.
Film Porosity
The preceding relationships pertain to pigment concentrations at which there is sufficient binder to wet the pigment completely and form a continuous phase, which means below the critical pigment volume concentration (CPVC). Above the CPVC, the dried film becomes porous, containing entrapped air that increases pigment-scattering efficiency by effectively lowering the refractive index of the surrounding medium. The air itself, as particulate matter in contact with the higher refractive index binder, contributes to light scattering. Thus, if the curves of Fig. 14 were extended to a sufficiently high PVC, the HP of the film would begin to rise again due to the opacification effect of film porosity. Obviously this is an extremely impractical use of expensive titanium dioxide with no relation to actual formulation practice. However, porosity does in practice make a major contribution to HP in the important interior flat wall paint sector. In paints of that type, inexpensive inert white pigments are included in the formulation, along with titanium
Through the use of encapsulated preformed microvoids, it has been found possible to obtain some of the HP benefit of entrapped air while avoiding or minimizing the deleterious effect of film porosity. The microvoids are supplied as a water dispersion of hollow beads having a plastic outer shell and water-filled core. Incorporated into a latex paint, the water in the core evaporates during the drying of the film and is replaced by air that functions as light-scattering particulates shielded from staining penetrants by the surrounding plastic shell. Because the microvoids alone are not able to produce the desired level of opacity in a film of normal thickness, the inclusion of titanium dioxide pigments in the paint formulation along with microvoids is essential. One widely used microvoid bead product is referred to as “opaque polymer” and employs a shell of thermoplastic polystyrene. Another type is a vesiculated bead in which titanium dioxide and water-filled “vesicles” are associated in a cross-linked polyester/styrene matrix. By using such products to partially replace titanium dioxide pigment, raw-material cost savings have been demonstrated with no loss in film integrity or HP [35,40].
FORMAL HP METHODS
ASTM Methods ASTM D344: TEST METHOD FOR RELATIVE HIDING POWER OF PAINTS BY THE VISUAL EVALUATION OF BRUSHOUTS
This is essentially the same as the Krebs Method described in “Relative Dry HP—Krebs Method,” differing only in requiring black-and-white instead of gray-and-white charts and in permitting checkerboard or other suitable contrast designs as well as the diamond-stripe pattern. Modern charts are 0.1 m2 in area (1.076 ft2) instead of 1 ft2 as specified originally. Provision is made for reporting results in m2/L as well as ft2/gal.
ASTM D2805: TEST METHOD FOR HIDING POWER OF PAINTS BY REFLECTOMETRY
This was adopted in 1969 and is actually a combination of two earlier methods, ASTM D1738 and ASTM D2614, that differed only in technique. It conforms with the general K-M method described in “General K-M HP Method” but is
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
588
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
The equation assumes that, for any one drawdown, the ratio of clearance to WFT for the several notches does not deviate appreciably. On that basis, WFT variations due to application technique or paint rheology would not affect the final test result. This is not a precision test, but provides significant information of a practical nature with minimal effort.
ASTM D5150: TEST METHOD FOR HIDING POWER OF ARCHITECTURAL PAINTS APPLIED BY ROLLER
Fig. 16—Multi-notch applicator for ASTM D5007.
designed specifically for air-dried coatings. Originally it provided for the use of either black glass or charts for determining Hx and R0. In later versions, black glass is mandatory. R∞ is determined by a separate application as described in “Determination of R0 and Hx.” The need for two test applications does not represent a significant extra effort since only the application on black glass requires the time-consuming SR determination. The latter is accomplished by placing a template of predetermined area on the dry film, scraping off and discarding the film outside the confines of the template, then carefully scraping off the remaining film in the defined test area and weighing it on an analytical balance. The SR is then calculated from the density and nonvolatile content of the paint using Eq (26). Having obtained the experimental values R0, Hx, and R∞, the scattering coefficients and HP H0.98 of the paint are calculated from these values using the K-M sequence shown in Calculation of H0.98, thus: Px = f(R0,R∞), S = PxHx, Pc = f(C;R∞), Hc = Pc/S for C = 0.98. Earlier versions of ASTM D2805 and its predecessor standards included or referenced the Mitton tables and graphs described in Mitton Graph and Table [27], for solving the K-M equations. The method can be adjusted by appropriate experimental modifications to the measurement of baked enamels on black-painted metal panels as discussed in “Painted Metal Panels,” or to other types of coatings and test substrates.
ASTM D5007: TEST METHOD FOR WET-TO-DRY HIDING CHANGE
This test method is concerned with determination of the change in hading power of an architectural coating during drying by visual evaluation of the wet and dry coating films. It is a rapid visual test designed to measure percent change in HP during drying. The paint is drawn down on a black-and-white test chart using a special multinotch applicator (Fig. 16) having eight notches with clearances in geometric progression ranging from 67 to 264 μm (2.65 to 10.4 mils). The clearance corresponding to an agreed visual endpoint (Visual HP End-Point) is estimated immediately after application and again after drying. The ratio of the two clearances multiplied by 100 gives the percentage change in HP: CLEARANCE WET ENDPOINT CLEARANCE DRY ENDPOINT =
=
This is a visual comparison method designed for use with interior wall finishes and intended to provide practical information from tests performed on a convenient laboratory scale. The test substrate is a large, sealed paper test chart (Fig. 17), with a series of stripes numbered 1 through 6 on a white background. The stripes range in shade from very light gray to black and were selected so that the color ∗ difference ΔEab between each successive stripe and the white surround is in a geometric progression from 2 to 64 CIELAB units. The dimensions of the test area are 24 by 36 in. = 6 ft2 (610 by 914 mm = 5575 cm2), sufficiently large to simulate practical application of paints with a roller. The paint is applied at a specified, controlled SR, and the HP is reported as the stripe number of the darkest stripe perceived as being completely obscured. The concept of this test is that in practical applications the levelness of the paint film and, hence, its effective opacity are affected by the rheological properties of the paint. Thus, in practice paints tend to have lower HP than indicated by more customary test methods in which films are applied with maximum uniformity using a blade-type applicator. Relative practical HP among paints can be influenced for that same reason.
ASTM D6441: TEST METHOD FOR MEASURING THE HIDING POWER OF POWDER COATINGS
This standard conforms with the power coating industry practice of reporting HP in terms of film thickness rather than SR. A “wedge” shape film providing a range of film thicknesses is applied by electrostatic spraying on black and white painted metal panels. After curing, film thicknesses are measured with an electronic film thickness gage and reluctances measured with a small aperture (e.g. 4 mm)
WFTWET ENDPOINT WFTDRY ENDPOINT
SPREADING RATE DRY ENDPOINT SPREADING RATE WET ENDPOINT
(57)
Fig. 17—Large gray scale chart (6 ft 2, 5575 cm 2) for roller application tests per ASTM D5150.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 45
spectrophotometer. Method A reports the CR at a specified film thickness, and Method B the film thickness at a specified CR.
Federal Test Method Std. 141 METHOD 4121, CR AT A SPECIFIED SR
This pass–fail test was previously described in some detail in “Federal Test Method for Dry Opacity.” Paint films are applied on black-and-white charts by brush or drawdown, and SRs are determined by a typical weight-area-density procedure, discussed in “SR (or Film Thickness) Determination.” CR values are plotted at several SRs to obtain graphically the CR at a specified SR. For the test paint to pass, the CR at the specified SR must have a specified minimum value.
ISO (International Standardization Organization) Methods ISO 2814, CR (CR) AT A NOMINAL SR (SR) OF 20 m2/L ON BLACK-AND-WHITE CHARTS OR POLYESTER FILM
A paint film is applied with a 100-μm clearance applicator to give a nominal WFT of 50 μm, corresponding to a SR of 20 m2/L. Black and white substrate reflectance are measured and the CR calculated without a determination of actual SR. Films cast on clear polyester film are measured, as described in Clear Plastic Film, by placing the film alternately on black and white glass. Because different paints and application techniques with the same applicator give films differing significantly in thickness, the method is satisfactory only as a rough guide for paints of the same type and color evaluated by one operator.
ISO 6504-1, KUBELKA-MUNK METHOD FOR WHITE AND LIGHT-COLORED PAINTS
This is in accordance with the general K-M HP method described in “General K-M HP Method” section. It calls for an all-black substrate, which can be glass or polyester film over black glass. The Mitton graph and table described in Mitton Graph and Table [27] are included, which makes it very similar to early versions of ASTM D2805.
ISO 6504-3, DETERMINATION OF CR (OPACITY) OF LIGHT-COLORED PAINTS AT A FIXED SR
This method is analogous to ISO 2814, but the SR is determined precisely by a weight-area-density procedure at several film thicknesses. Substrates are clear polyester film in Method A and black and white charts in Method B. The CRs and SRs are plotted graphically and the CR at 210 m2/L is reported. The CIE-Y reflectance of the white substrate is specified to be 80 ± 2 %.
British Standards Institute, BSI 3900
Part D4. Comparison of CR of Paints of the Same Type and Color—This method is technically identical with ISO 2814 (11.c). Part D6. CR at 20 m2/L Using Polyester Film—This method is technically identical with ISO 3906-1980 (11.c). Part D7. True HP (SR at C = 0.98) by the Kubelka-Munk Method—This method is technically identical with ISO 6504-1 (11.c) and in accordance with early versions of ASTM D2805 and the general K-M HP method described in “General K-M-HP Method” section.
Q
HIDING POWER
589
Canadian General Standards Board, (CGSB) 1-GP-71 METHOD 14.1, VISUAL HIDING AT A SPECIFIED SR
The test substrates are black-and-white or black-and-gray checkerboard charts with an area of 0.1 m2. The appropriate chart is specified according to a list of CGSB color numbers, with black-and-gray being used for lighter colors. The paint is applied by brush or drawdown. In brush application the SR is controlled accurately by weighing container and brush before and after application, with a specified volume being delivered to the chart surface by syringe. With drawdowns, presumably identical applications are made on glass and charts and the WFT determined on glass by means of an Interchemical (ASTM D1212) WFT gage. For the test paint to pass, the dry paint film is required to completely obscure the contrast pattern of the chart.
METHOD 14.2, SR DETERMINED AT FULL VISUAL HIDING (FOR QUICK-DRYING COATINGS)
Successive thin coats are applied by spraying onto blackand-gray or black-and-white charts until visual hiding of the dry film is complete. The SR is calculated from the difference in weight of the coated and uncoated chart. This can be expressed in m2/kg of dry film or m2/L of the original liquid coating.
METHOD 14.7, CR ON BLACK AND WHITE GLASS PANELS AT A GIVEN SR OR DRY FILM THICKNESS
This is modeled after the NYPC method described in “New York Paint Club (NYPC) Method.” WFT is determined with an interchemical gage or DFT with a micrometer. The target film thickness is bracketed experimentally to obtain two points on a CR versus reciprocal film thickness graph and the CR at the target thickness determined by interpolation. The experimental CR values are corrected for W = 0.80 before plotting the graph.
French Standards Association (AFNOR) NF-T30-075, SR AT A CR (CR) OF 0.98
Paint films are cast on clear polyester at several thicknesses and CR values are determined after drying by measuring reflectance over a black-and-white substrate. Dry films just below and above 0.98 in CR are measured by weight or micrometer to obtain experimental SRs in m2/kg or m2/L and results interpolated to obtain the SR at exactly CR = 0.98. The introductory text points out that this method measures true HP in preference to ISO methods that simply compare CR values at 20 m2/L. It also refers to the experimental film thickness not being limited to 50 μm as in ISO methods. No provision is made in this method to correct for deviations of the white substrate from W = 0.80.
NF-T30-076, SR AT COMPLETE VISUAL HIDING
This is referred to as a “simplified” method. Several films are cast on polyester to obtain one that shows full hiding when placed over a black-and-white background. The DFT is measured by difference with a micrometer and the HP calculated in m2/L. Potential users should consider whether this method, though simple in concept, might be excessively burdensome in execution.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
590
PAINT AND COATING TESTING MANUAL
Q
German Standards Institute (DIN) DIN 53162, HP OF AIR DRYING NONCHROMATIC PAINTS
This is a K-M method that is essentially the same as ISO 6504-1, but includes auxiliary test procedures for measuring paint density and nonvolatile content. The Mitton nomograph and table Mitton Graph and Table [27] are employed.
DIN 53164, RELATIVE SCATTERING POWER OF WHITE (TiO2) PIGMENTS
This method measures the K-M scattering coefficient S of a TiO2 pigment and reports its value as a percentage of the scattering coefficient of a reference pigment measured in the same way. The determination of S is based on the solution of Eq 44: P = f(R0, Rx) using the Mitton monograph. The method calls for the test pigment to be dispersed in an alkyd or a plasticized polyvinyl chloride vehicle. R∞ is measured from a thick, full hiding film of the dispersion and R0 from a nonopaque film applied on a black plastic substrate. The SR Hx of the pigment is determined in a unique way, by igniting a known area of film on plastic and weighing the residue. This method is basically the same as DIN 53162 and other K-M methods (General K-M HP Method), with the difference that only relative values are reported. There is no attempt to report actual scattering coefficients or to calculate HP in physical units, although this could easily be done on the basis of the accumulated data.
References [1] Gardner, H. A., and Sward, G. G., Physical and Chemical Examination of Paints, Varnishes, Lacquers and Colors, 9th ed., May 1939, p. 10. [2] Simpson, L. A., “Measuring Opacity, Part I,” Paint, Pigments and Coatings J., Vol. 179, February 1989. [3] Mitton, P. B., Vejnoska, L. W., and Frederick, M., “HP of White Pigments: Theory and Measurement—I,” Official Digest, Federation of Paint and Varnish Production Clubs, Vol. 33, 1961. [4] Mitton, P. B., Paint Testing Manual, ASTM STP 500, Chap. 1.3: HP, “Physical and Chemical Examination of Paints, Varnishes, Lacquers and Colors,” 13th ed., ASTM International, West Conshohocken, PA, 1972. [5] Gardner, H. A., Sward, G. G., and Levy, S. A., “HP and Tinting Strength of Pigments and Paints,” Scientific Section Circular, National Paint, Varnish, and Lacquer Association, No. 362, 1930. [6] Kraemer, E. O., and Schupp, O. E., “Determination of HP of White Paints,” unpublished paper presented at the Washington, DC meeting of the American Chemical Society, March 1933. [7] Pfund, A. H., “HP of White Pigments and Paints,” J. Franklin Inst., Vol. 188, 1919, p. 675. [8] Pfund, A. H., “HP Measurements in Theory and Application,” Proceedings, Vol. 30, Part II, ASTM International, West Conshohocken, PA, 1930, p. 878. [9] Sward, G. G., and Levy, S. A., “An Instrument for HP Determinations,” Scientific Section Circular, National Paint, Varnish, and Lacquer Association, No. 433, 1933. [10] Brodgen, D., “The Precision of the Pfund Black and White Cryptometer,” Official Digest, Federation of Paint and Varnish Production Clubs, Vol. 33, 1961, p. 1297. [11] Saxena, K. G., and Chowdhry, K. K., “Determination of Opacity of Wet Films of Ready-Mixed Paints and Enamels,” Paintindia, Vol. 12, No. 1, 1962, p. 103. [12] Hallet, R. L., “An Instrument for Measuring the HP of Paints,” Proceedings, Vol. 20, Part II, ASTM International, West Conshohocken, PA, 1920, p. 426.
15TH EDITION
[13] Pfund, A. H., “HP Measurements in Theory and Practice,” Proceedings, Vol. 30, Part II, ASTM International, West Conshohocken, PA, 1930, p. 882. Pfund, A. H., “The Photoelectric Cryptometer,” Proceedings, Vol. 31, Part II, ASTM International, West Conshohocken, PA, 1931, p. 876. [14] Hanstock, R. F., “The Opacity of Paints,” J. Oil Colour Chemists’ Assoc., Vol. 20, 1937, p. 5. [15] Sawyer, R. H., “HP and Opacity,” Symposium on Color, ASTM STP 50, ASTM International, West Conshohocken, PA, 1941, p. 22. [16] Switzer, M. H., “Critical Analysis of the Fell HP Relationship,” Am. Paint J., Vol. 40, No. 13, 1955, p. 72. [17] Mitton, P. B., “A Mathematical Analysis of the Precision in Determining HP,” Am. Paint J., Vol. 30, 1958, p. 156. [18] Van Eyken, W. W., and Anderson, F. T., Jr., “An Improved Method of HP Determination,” Am. Paint J., Vol. 43, No. 31, 1959, p. 78. [19] Tough, D., “The Use of CR in the Measurement of HP,” J. Oil Colour Chemists’ Assoc., Vol. 39, 1956, p. 169. [20] Gall, L., “On the HP of Colored Pigments in Paints and Printing Inks,” Farbe and Lack, Vol. 72, 1966, p. 1073. [21] Mitton, P. B., “Measuring HP of Baked Coatings on Metal,” Met. Finish., Vol. 72(G), 1974, p. 44. [22] Kubelka, P., and Munk, F., “Ein Beitrage zur Optik der Farbenstriche,” Z. Tech. Phys. (Leipzig), Vol. 12, 1931, p. 593. [23] Steele, F. A., “The Optical Characteristics of Paper,” Pap. Trade J., Vol. 100, No. 12, 1935, p. 37. [24] Judd, D. B., Harrison, W. N., Hickson, E. F., Eickhoff, A. J., Shaw, M. B., and Paffenbarger, G. C., “Optical Specification of Light-Scattering Materials,” Journal of Research, National Bureau of Standards, Vol. 19, p. 287. [25] Kubelka, P., “New Contributions to the Optics of Intensely Light-Scattering Materials—Part,” J. Opt. Soci. Am., Vol. 38, 1948, p. 448. [26] Switzer, M. H., “Equation for Calculating HP Index and SR of Paints,” ASTM Bulletin, No. 181, ASTM International, West Conshohocken, PA, 1952, p. 75. [27] Mitton, P. B., “Easy, Quantitative HP Measurements,” J. Paint Technol., Vol. 42, 1970, p. 159. [28] Mitton, P. B., to Weaver, J. C., personal communication, 1977. [29] Mitton, P. B., Madi, A. J., and Rode, J. W., “Development of a Test Method for HP,” J. Paint Technol., Vol. 39, 1967, p. 536. [30] Hallett, R. L., “HP and Tinting Strength of White Pigments,” Proceedings, Vol. 30, Part II, ASTM International, West Conshohocken, PA, 1930, p. 895. “HP of Pigments,” Proceedings, Vol. 26, Part II, ASTM International, West Conshohocken, PA, 1926, p. 538. [31] Titanium Pigment Company, “The Handbook,” 1956. [32] Mitton, P. B., and Jacobsen, E. E., “Reflectometry Method for Measuring Tinting Strength of White Pigments,” Official Digest, Federation of Paint and Varnish Production Clubs, Vol. 34, 1962, p. 704. [33] Simpson, L. A., “Measuring Opacity, Part II,” Paint, Pigment Coatings J., Vol. 179, March 1984. [34] Balfour, J. G., and Hird, M. S., J. Oil Color Chemists Assoc., Vol. 58, 1975, p. 331. [35] Simpson, L. A., “Measuring Opacity, Part III,” Paint, Pigment and Coatings Journal, Vol. 179, April 1989. [36] Stieg, F. B., “A New Look at the HP of Titanium Pigments,” Official Digest, Federation of Paint and Varnish Production Clubs, Vol. 29, 1957, p. 439. [37] Stieg, F. B., “The Effect of Extenders on the HP of Titanium Pigments,” Official Digest, Federation of Paint and Varnish Production Clubs, Vol. 31, 1959, p. 52. [38] Stieg, F. B., and Ensminger, R. I., “The Production and Control of High Dry Hiding,” Official Digest, Federation of Paint and Varnish Production Clubs, Vol. 33, 1961, p. 792. [39] Stieg, F. B., “The ABCs of White HP,” J. Coat. Technol., Vol. 49, 1977. [40] Fasano, D. M., Hook, J. W., Hill, W. H., and Equi, R. S., “Formulating High PVC Paints with Opaque Polymer Additives,” Resin Review, Vol. 37–2, 1987.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
46
MNL17-EB/Jan. 2012
Mass Color and Tinting Strength of Pigments Joseph V. Koleske1 PREFACE
IN PREPARATION OF THIS CHAPTER, THE CONtents of the Fourteenth Edition of this manual were drawn upon. The author acknowledges the author of the chapter in the Fourteenth Edition, Julio I. Aviles. The current edition will review and update the topics as addressed by the previous author, introduce new technology, and include up-to-date references.
INTRODUCTION
Color is a common but complex subject that can be described by perspective and in mathematical terms. As early as 1666, when he was 23, Isaac Newton developed his famous and useful “Newton color circle” that provided an understanding about additive color mixing and complementary colors. His experiments with prisms demonstrated that ordinary sunlight is made up of all wavelengths in the visible spectrum and thus all colors. Thomas Young,2 an English physicist, in the early 1800s suggested that color perception is three-fold in nature, and he speculated that there were three kinds of nerve fibers or receptors in the eye’s retina—something that was experimentally proven in 1959. These eye receptors are for long, middle, and short wavelengths that correspond to the primary colors of red, green, and blue (the RGB primaries) that had been useful in matching many visual colors by additive mixing. Later Hermann von Helmholtz put the qualitative analysis on a quantitative basis wherein three parameters are used to describe a color sensation—the well-known tristimulus values. While James Maxwell was examining usage of the three primary colors in the 1860s, he found that no additive combination would cover the entire range of hues perceivable to the eye. It was found that the set of primary colors was not unique and that more widely wavelength-separated spectral primaries would produce a more complete range of perceived hues. In addition, with some subtractive combinations, the entire range of perceived colors could be included. Maxwell also demonstrated that the hue and saturation or chromaticity of a colored surface is in effect insensitive to brightness. His studies are considered the basis of modern colorimetry. In the 1920s and early 1930s, experimental efforts by W. David Wright and John Guild pointed out that all colors
within a particular range could be matched with the additive RGB primaries but not all spectral colors could be matched. The lack of matching was particularly true in the green portion of the range. However, if an amount of red wavelength light were added to the color being matched, all colors were possible. These quantitative studies were expressed in the modern terms of tristimulus values for the RGB primaries, but negative values had to be imposed for the red values so that all colors could be matched. In 1931 the International Commission on Illuminations (CIE)3 defined a color system in which all tristimulus values were positive and all visible colors could be represented by two color coordinates, x and y, that specify the color point on a chromaticity diagram. The third or Y parameter used in the CIE system for defining colors is a luminance parameter. This CIE system has some difficulties associated with it, and a new system was introduced in 1976. While the 1976 CIE standard removes the difficulties, it has not gained acceptance, and currently the 1931 CIE standard is the basis for almost all quantitative color measurements. Basic tristimulus color measurement is usually carried out with either a colorimeter or a spectrophotometer. Tristimulus colorimeters are a combination of an illumination source, an array of filters, and a photoelectric output instrument. Measurements are comparative with the device standardized by using ceramic or glass standards that have colors similar to those of the materials to be measured. If precise measurements are needed, spectrophotometers that measure reflectance at each wavelength are used to determine the tristimulus values. This brief passage into the history of color merely scratches the surface of color technology. Color and its management in paint and coatings are important from both an aesthetic and a practical or economic standpoint. It plays a role in our everyday lives and it affects our emotions and moods as we react to the colors around us. From an economic standpoint, it was well pointed out by Rich [1], who stated, “Getting the customer’s color right the first time, whether it involves paint, coatings, or ink, is often the first and most crucial test of quality.” The tinting strength of pigments is often used as a guide for estimating relative hiding power of a pigment or paint to completely obliterate the background to which it has been applied. It may seem as if one is stating the
1513 Brentwood Road, Charleston, WV 25314. Young is also the investigator who devised the concept of a modulus of elasticity that is known in mechanical properties as Young’s modulus. 3 The abbreviation CIE is taken from the French title for the commission, Commission Internationale de l’Eclairage. The address for the commission, which is an independent organization that was conceived in 1913, is International Commission on Illuminations, CIE Central Bureau, Kegelgasse 27, A-1030 Wien, Austria. 1 2
591 Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed byASTM International This standard is for EDUCATIONAL USE ONLY. Copyright © 2012 by www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
592
PAINT AND COATING TESTING MANUAL
Q
obvious, but when light falls on a painted substrate, some of the light is reflected from the surface and the remainder enters the paint or coating film. The light that enters the film either emerges from the various faces, including the top face, or is absorbed. If all or most of the light is absorbed, the film appears dark in color or black, the coating effectively hides the substrate, and the pigment involved is said to have good tinting strength. The factors that have an effect on tinting strength and mass color include pigment refractive index, particle size, concentration, degree of dispersion, and inherent light-absorbing characteristics. Applied coating thickness also has an effect on this characteristic. It has been suggested to the author that most industrial colorists consider tinting strength as a “quality control measurement comparing a batch to a standard of the identical materials.” The ASTM definitions of tinting strength are given in a following section dealing with definitions. For both mass color and tinting strength, it is the final appearance or reflectivity (R∞-value) of the opaque film that is of importance. In turn, reflectivity depends on pigment light-scattering and light-absorption characteristics and particularly with variation of these properties with wavelength(s) of the impinging light. In the case of mass color, the R∞-value depends primarily on the relative quantity of each characteristic and secondarily on the absorption property of the vehicle. For tinting strength, reflectivity depends on the amount of each characteristic as related to the amount found in a standard paint or pigment. Historical information about these parameters and early studies related to their development can be found in the literature, and they have been summarized by Mitton [2] and in a follow-up publication by Aviles [3]. A fairly recent publication by Völz [4] deals with color testing and includes chapters that deal with determination of tinting strength and related hiding power. An easy-to-read handbook concerned with color can be found on the Internet [5].
DEFINITIONS
Mass color is the color, when viewed by reflected light, of a pigment-vehicle mixture of such thickness as to obscure completely the background. Sometimes mass color is referred to as over-tone or mass-tone [6]. In a different ASTM publication [7], mass color is not listed, but masstone (note, no hyphen) is defined as “in paint technology, a pigment-vehicle mixture containing a single colorant only,” but the discussion points out that “at times colorants are developed that contain more than one pigment, but are tested and used as if they contained only a single pigment. This definition is meant to include such colorants.” Mass color is produced by the reflected light, R∞, of an opaque coating, and it depends on the pigment concentration, degree of dispersion, coating thickness, and the light absorption, K, and scattering, S, of pigments and binders. Mass color is applicable to both chromatic and achromatic pigments. Tinting strength is a “measure of the effectiveness with which a unit quantity of a colorant alters the color of a material. For scattering and absorbing colorants, both scattering and absorption tinting strength must be specified” [7]. ASTM D16 [6] defines tinting strength as “the power of coloring a standard paint or pigment.” ASTM E284 defines tinting strength as a “measure of the effectiveness
15TH EDITION
with which unit quantity of a colorant alters the color of a material.” Other definitions that are of interest include the following [7]. The listing is by no means complete and it is given to help guide a reader/user to the terms used in the technology. In certain cases, designation symbols are listed to again provide guidance. t Absorption is the transformation of radiant energy to a different form of energy by interaction with matter. t Absorption tinting strength is the relative change in the absorption properties of a standard white material when a specified amount of an absorbing colorant, black or chromatic, is added to it. t Achromatic (1) for primary color sources, the computed chromaticity of the equal-energy spectrum; (2) for surface colors, the color of a whitish light, serving as the illuminant, to which adaptation has taken place in the visual system of the observer; (3) the perception of having no hue, that is, as white, gray, or black. t Brightness (1) is an aspect of visual perception whereby an area appears to emit more or less light or (2) brightness of an object is the combination of lightness and saturation. Different definitions of brightness are used in the textile industry, the paper industry, and by dyers. t Chroma is an attribute of color used to indicate the degree of departure of the color from a neutral color of the same lightness. Munsell chroma is an attribute of color used in the Munsell color-order system to indicate the degree of departure of a color from a gray to the same Munsell value, in steps that are visually approximately equal in magnitude. t Chromatic describes the perception of something having a hue; not white, gray, or black. t Chromaticity Diagram is a plane diagram in which points specified by chromaticity coordinates represent the chromaticities of lights (color stimuli). t CIE is an abbreviation for the French title of the International Commission on Illumination, Commissione Internationale de l’Eclairage. t CIE 1931 standard colorimetric system is a system for determining the tristimulus values of any spectral power distribution using the set of reference color stimuli X, Y, Z and the three CIE color-matching functions x(λ), y(λ), and z(λ) adopted by the CIE in 1931. t Colour Index International is a listing of colors by name and number by the Society of Dyers and Colourists and American Association of Textile Chemists and Colorists [8]. t Hue is the attribute of color perception by means of which a color is judged to be red, orange, yellow, green, blue, purple, or intermediate between adjacent pairs of these, considered in a closed ring (red and purple being an adjacent pair). Neutral colors are judged to have no “hue.” Munsell hue is an attribute of color used in the Munsell color-order system to indicate the hue of a specimen viewed in daylight. t Lightness is (1) the attribute of color perception by which a non-self-luminous body is judged to reflect more or less light; (2) the attribute by which a perceived color is judged to be equivalent to one of a series of grays ranging from black to white. t Luminance is the luminous flux in a beam, emanating from a surface, or falling on a surface, in a given
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 46
t
t
t
t
t
t
t
Q
MASS COLOR AND TINTING STRENGTH OF PIGMENTS
direction, per unit of projected area of the surface as viewed from that direction, per unit solid angle. Munsell Color System was developed just before the end of the nineteenth century by A. Munsell, an American artist. The system is depicted as a globe with a band of colors running equatorially and an axis of gray values wherein white is the North Pole and black is the South Pole. At each gray value, there is a change in color that gradually ranges from full saturation to neutral gray, and in this manner a myriad of colors could be described by their hue, chroma, and value. Saturation is the attribute of a visual sensation that permits a judgment to be made of the proportion of pure chromatic color in the total sensation. Scattering is the process by which light or other electromagnetic radiant flux passing through matter is redirected over a range of angles. Scattering tinting strength is the relative change in the scattering properties of a standard black material (with no scattering colorant present) when a specified amount of a white or chromatic scattering colorant is added to it. Shade is (1) a color produced by a dye or pigment mixture including black dye or pigment; (2) an expression of color difference from a reference dyeing such that another dye must be added to produce a match; and (3) a color slightly different from a reference color. Shade is related to “tint.” Tint is a color produced by the mixture of white pigment or paint with a chromatic pigment or paint. When used as a verb, “tint” means to adjust the color of a test specimen to be a closer color match to the standard. Also see “Shade” above. Tristimulus values are the amount of three specified stimuli required to match a color. In the CIE system, these three stimuli are assigned the symbols X, Y, and Z.
TINTING STRENGTH
As mentioned above, tinting strength is a measure of the effectiveness with which a unit quantity of a colorant changes the color of a material. One may think of it as a pigment’s “coloring power.” For colorants that both absorb and scatter light, scattering and absorption tinting strengths must be specified. Scattering tinting strength is the relative change in scattering properties of a standard black material that has no scattering colorant present when a specified amount of white or chromatic scattering colorant is added to it. Absorption tinting strength is the relative change in absorption properties of a standard white material when a specified amount of an absorbing colorant, black or chromatic, is added to it. Since pigment concentration is important to coating strength and cost, tinting strength is an important economic factor when selecting one paint over another. There is no particular value of tinting strength that can be stated as desirable unless an end use is stated. In certain cases, a high value is desirable and in other cases a low value is required to achieve a desired color/strength effect. Pigments each have a different ability to vary the color of a mixture. When economy is a factor in paint formulation, stronger tinting strength is desirable. Some colors, such as phthalocyanine blue, are very strong in their tinting strength whereas other colorants can be relatively weak. Tinting strength increases
593
as pigment particle size decreases with synthetic organic pigments generally having greater tinting strength than mineral pigments. In a general sense, tinting strength is determined by dilution of a test paint and a reference paint with a standard “mixing white paint” in the case of chromatic paints or a standard “tinting color” in the case of white paints. These diluted pastes are then drawn down on a suitable substrate, and then either instrumentally measuring tristimulus values or visually comparing the specimens. As would be expected, the visual comparison technique has lower precision than the instrumental method. Details for preparation of standard white paints are described [9] or a commercial titanium dioxide-white artists’ paint may be used as the standard. It should be understood that the mixing white paint must be made with the same vehicle chemical type— acrylic, alkyd, or oil—as the paint to be tested. When color tints are considered, differences in gloss and haze can be mistaken for a lighter tint or lower tinting strength than really exists. In grays, these factors may be interpreted as higher white pigment strength than exists. Instruments cannot compensate for specular gloss or haze differences between a sample and a standard, and this can result in erroneous tint strengths. It is possible to equalize gloss differences between specimens by top coating them with a clear coating. Evaluation through the clear coating reveals the true tinting strength differences between sample and standard. Tinting strength results can also be affected if the lightness, chroma, and saturation of the sample differ significantly from those of the standard, since the measurements involve matching two color variables—either lightness and chroma or lightness and saturation—by adjusting only the amount of pigment used [10,11].
Chromatic Paints
ASTM D4838 [12] is a method used for determining the absorption tinting strength of a chromatic test paint relative to that of a standard or reference paint of the same chemical type. The procedures in this method are based on dilution of paints with a standard mixing white paint followed by instrumental measurement and calculation. Provision is made for correcting the results for small differences in hue or chroma (or both) between the test and reference chromatic paints. The method is meant for comparison of paints that contain the same chemical type vehicle (acrylic, alkyd, or oil) and single-pigment colorants of the same Colour Index name and number. It is unnecessary to have information about the amounts of pigments or other components in the paint. The color-measuring instrument can be either a spectrophotometer that provides 1931 CIE tristimulus values, X, Y, and Z for CIE standard illuminant C or a tristimulus colorimeter providing either such tristimulus values or colorimeter readings R, G, and B. Other test methods or practices useful in following the results of this test method can be found in the literature [13–15]. ASTM E1347 [16] deals with a test method for the use of a tristimulus colorimeter to evaluate specimens and provide color coordinate and color difference values. The device is also known as a tristimulus filter colorimeter or a color-difference meter.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
594
PAINT AND COATING TESTING MANUAL
Q
ASTM D387 is used to compare the color and strength of a pigment that is being tested with a reference standard of the same type and grade [17]. The method specifically points out that it does not apply to white pigments. The pigments are dispersed in a suitable vehicle with a mechanical muller. Opaque drawdowns are made on white paper charts that have a black band and a surface that is impervious to paint liquids. These drawdowns are compared either visually or instrumentally for color and strength differences. ASTM D3022 is similar to this method, but it utilizes a miniature sandmill rather than a mechanical muller to disperse the color pigment [18]. ASTM D2066 is a related test method that is applicable to the relative tinting strength of paste-type printing ink dispersions [19] and is similar in nature to other referenced ASTM test methods [12, 17]. The strength may be determined by instrumental or visual observation of manually or automated mixed tints. The methods are applicable to paste-type printing inks, flushed pigments, and other pigment dispersions that are essentially nonvolatile under ordinary room conditions and for which there is a wet reference standard of the same pigmentation and consistency. If the proper choice of tinting base is used, the methods are applicable to dispersions of any color, including black and white.
White Paints
ASTM D2745 is a standard method for determination of the relative tinting strength of white pigments by reflectance measurements made on black tints [20]. It is only applicable for comparing the test pigment with a reference standard of the same type and grade. The method is carried out by dispersing the pigment in an agreed-on, solvent-free vehicle and then letting it down with additional vehicle that has been tinted with a lampblack that has been predispersed in a vehicle similar in nature to the test vehicle. Refined or low-bodied linseed oil should not be used with this procedure. Both dispersion and let-down are done with an automatic, mechanical muller. Tristimulus values are determined with a colorimeter. ASTM D332 is a test method for determination of the relative tinting strength of white pigments by visual assessment of blue tints [21]. It is only applicable for comparing a test pigment with a reference standard of the same type and grade. The test is conducted by dispersing specified amounts of a white pigment and a blue tinting pigment that conforms to specifications [22] in a refined linseed oil with an acid number of about 4 using a glass, hand muller, or an automatic muller. The pastes are drawn down on a specified panel and visually evaluated for tinting strength.
15TH EDITION
A numerical rating of tinting strength is obtained by preparing dispersions with the standard white pigment and more or less of the tinting pigment. These are compared until the lightness of the test paste is matched. The weight of the tinting pigment is used to calculate the relative tinting strength.
PIGMENT DISPERSION
Tinting strength and mass color require that pigments be well dispersed in the binder to achieve maximum tinting strength. Ideally, it would be desirable to break down pigment agglomerates into the smallest possible individual particles, i.e., to an ultimate dispersion state. The smaller the particle size the greater the surface area of a given weight or mass of pigment. This then results in more intense color production in the same volume of liquid. However, an ultimate dispersion state is impractical and effectively impossible to achieve. Therefore, pigments under investigation for these properties must be processed in the same manner and receive the same level of mechanical work. Mechanical mullers, which are instruments that have two circular, usually ground-glass grinding surfaces that contain the pigment and vehicle, are used to disperse the two components. A variety of these devices are commercially available. Development of tinting strength is dependent on the force applied to the glass plates, the number of revolutions used, and the mass of pigment and vehicle used. If muller conditions, pigment, and vehicle have not been agreed on by purchaser and seller, the mandatory dispersing conditions given in the Annex of ASTM 387 should be used to attain a consistent level of tinting strength [17]. These conditions include: t Determination of the appropriate ratio of color pigment to dispersing vehicle for the standard and test pigments. t Determination of appropriate masses of pigment and vehicle to use. t Preparation of a standard tint by application of 100 lbf (N) to the muller plates, introducing the appropriate mass of pigment/vehicle, and mulling the paste for 100 revolutions in two stages of 50 revolutions each. This is then repeated on three more specimens of the standard mixture except the mulling is carried out for 200, 300, and 400 revolutions in stages of 50 revolutions. t Each of the four specimens is compared one to the other for tinting strength and the minimum number of revolutions needed to develop maximum or full tinting strength is ascertained. The parameters and dispersing conditions used for three pigments that were investigated in an interlaboratory test to determine the precision of this standard method are given in Table 1.
TABLE 1—Interlaboratory pigment dispersing parameters and specific conditions obtained for maximum tinting strength [17] Pigment Type Parameter
Yellow Iron Oxide
BON Red
Phthalocyanine Green
Force applied, Ibf (N)
100 (440)
100 (440)
100 (440)
Total No. revolutions
100 (2 × 50)
200 (4 × 50)
400 (8 × 50)
Mass of color pigment, g
1.0
0.6
0.75
Mass of dispersing vehicle, g
1.7
1.4
1.8
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 46
Q
MASS COLOR AND TINTING STRENGTH OF PIGMENTS
595
PIGMENT DISPERSION TECHNIQUES
It should be kept in mind that the techniques described in this chapter pertain to preparing specimens for determination of tinting strength and mass color, and they are not meant for pigment dispersion in general. The general topic of pigment dispersion is discussed elsewhere in this manual as well as in a variety of references [23–31]. There is a vast amount of literature that deals with the surface treatment of organic pigments to improve ease of dispersibility, and interested readers are encouraged to seek such information in the classic work of Hayes [32], the above listed references, and others [33–35]. Detail regarding the surface treatment of inorganic pigments is also in the literature [36–38].
test grinds, specified media must be used. For small grinds, a media mill can be simulated with a laboratory dispenser equipped with a fiber or Telfon 1 5/8 in. disk, a 200 mL tall-form beaker, and media. Equal volumes of millbase to media are used. Grinds may be 60.0 g for carbon black and some organic pigments to 160.0 g for inorganic pigments. The millbase must be carefully prepared to eliminate gross and oversized agglomerates. Peripheral impeller speed should be 2000 feet (610 m) per minute, and the mixture should be ground for a set time or to a set dispersion level such as a Hegman value of 7.0+. Advantages of media mills include development of the highest tinting strength and mass color, simulation of actual factory grinding conditions, and low cost.
Spatula and Hand Mullers
LABORATORY ROLLER MILL
Grinding pigment/vehicle combinations can be carried out with a spatula or a hand muller4 by rubbing or mulling the materials over a 3 × 12 in. (8 × 30 cm) strip area on a glass plate. The rubbing is done by pushing the muller up one side and pulling it down the other side of the strip area so that all color particles receive the same amount of rubbing. One rub is one up and down course on the strip area. Early studies by Ayers [39] indicated that the muller gave more reliable results than a spatula and that the rubbing surface may vary a great deal without affecting the results. Stutz’s [40] results during investigation of the tinting strength of white pigments also found a muller was superior to a spatula. It was also found that a weighted or un-weighted muller could be used without affecting the results.
Automatic Mullers
Automatic5 or mechanical mullers have two circular glass grinding surfaces that contain the pigment/vehicle mixture. The grinding surfaces are usually constructed of ground glass with one surface stationary and weighted to exert a pressure of 100 psi (440 N) and the other surface rotary with rotation effected with a motor. Because rotation is about the diskcenters, paste located at the center can receive less mulling than paste located near the edges. To compensate for this effect, it has been found helpful to spread the paste in a ring approximately halfway between the edge and the center. The revolutions per mulling cycle can be adjusted in increments of 1 to 999. Mechanical muller advantages include very good development of tinting strength, the possibility to rapidly mull small quantities of materials, and efficient processing of a large number of samples. The specific way to operate a mechanical muller when determining tinting strength or mass color can be found in the annex to Ref. [17].
LABORATORY MINIATURE MEDIA MILLS
Commercial horizontal or vertical laboratory media mills that can process up to one quart of millbase are available. In a general sense, media can be nonferrous or ferrous as, for example, flint pebbles, sand, ceramic alumina, porcelain, stainless steel, high carbon steel spheres, etc. but for Hand mullers are implements made of a hard substance such as glass, stone, or similar material and used as a pestle to grind pigment/vehicle (i.e., paint) combinations. These are seldom, if at all, used today. 5 Automatic in this area of technology means “motor driven.” 4
Small, three-roller mills have been found useful for grinding small, laboratory-size batches of paint. Rolls of such mills are about 4 in. (10.16 cm) in diameter and 8 in. (20.82 cm) in length. Batches as small as 5 g have been prepared in such mills.
PALL GLASS MILL
Small quantities of metal-free pigment paste can be prepared with the Pall Glass Mill [41]. The mill is a heavy ground-glass stopper in a heavy glass joint. The mixed, un-ground materials (1–8 g) are placed in the joint and the stopper/plunger is inserted. The stopper is rotated with a small motor at about 150 rpm. In this mill grinding pressure ranges from 20 to 30 psi. The Pall Glass Mill is said to be an improvement over hand mulling because it saves time and because it results in greater development of tinting strength.
PIGMENT CONCENTRATION
Paste viscosity has an effect on grinding efficiency, and it is a property that determines the level of tinting strength and mass color that is developed. Ayers [39] investigated iron oxides, and his results indicated that color developed faster as paste viscosity increased. A low-viscosity paste had a reflectance of 26 % at a wavelength of 700 nm, whereas a high-viscosity paste was darker and redder with a reflectance of 23 %. It should be pointed out that there is a point above which viscosity has no effect on grinding efficiency.
MIXING TIME OF LIQUID COLORS
An important property of oil or universal liquid colorants is the ease with which they can be incorporated into white paints. A method for testing the speed of incorporation of such colorants has been described by Paul and Diehlman [42]. Their method involves use of a mechanical rotating bottle that contains a white paint, the liquid colorant, and a grinding media. When the test was first developed, No. 11 lead shot was used as the media; however, today glass beads, zirconia grinding media, and steel shot are among the media used to avoid lead contamination of the paint. The test is conducted by charging the bottle with 550 g of grinding media, 2 mL of the liquid colorant, and 75 mL of white paint. The bottle is then closed with a cork that is concave on the inner end to match the glass end of the bottle and placed in the holder of the rotating device. The
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
596
PAINT AND COATING TESTING MANUAL
Q
TABLE 2—Related ASTM Documents ASTM Designation
Title
D282
Methods of Test for Mass Color and Tinting Strength of Pigments (withdrawna)
D2244
Practice for Calculation of Color Tolerances and Color Differences from Instrumentally Measured Color Coordinates
D3265
Standard Test Method for Carbon Black-Tint Strength
D5326
Standard Test Method for Color Development in Tinted Latex Paints
D6531
Standard Test Method for Relative Tinting Strength of Aqueous Ink Systems by Instrumental Measurement
E179
Guide for Selection of Geometric Conditions for Measurement of Reflection and Transmission Properties of Materials
E284
Terminology of Appearance
E805
Standard Practice for Identification of Instrumental Methods of Color or Color-Difference Measurement of Materials
E1164
Practice for Obtaining Spectrometric Data for Object Color Evaluation
E1345
Practice for Reducing the Effect of Variability of Color Measurement by Use of Multiple Measurements
E1347
Standard Test Method for Color and Color Difference Measurement by Tristimulus Colorimetry
This particular method was withdrawn in 1929 and has not been replaced. It will not be found in ASTM’s listing of standards. It is included for historical purposes.
a
bottle is then tilted and rotated slightly by hand to wet the container walls. This is followed by mechanical rotation of the bottle until the paint is homogeneous in that no streaking of the color is seen. The time in seconds for mixing is determined and recorded.
OTHER RELATED TEST PROCEDURES
While the various test procedures cited above each contain reference to a number of ASTM documents, it is worthwhile to list some of these. Table 2 is a partial listing of such potentially useful documents.
References [1] Rich, D. C., “Artificial Intelligence in Today’s Colorant-Management Systems,” Paint & Coating Industry, Vol. XIV, No. 9, Sept. 1998, p. 48. [2] Mitton, P. B., Chapter 1.4 “Mass Color and Tinting Strength,” in Paint Testing Manual, Thirteenth Edition, G. G. Sward, Ed., ASTM International, West Conshohocken, PA, 1972.
15TH EDITION
[3] Aviles, J. I., Chapter 43, “Mass Color and Tinting Strength,” in Paint and Coating Testing Manual, Fourteenth Edition, J. V. Koleske, Ed., ASTM International, West Conshohocken, PA, 1995. [4] Völz, H. G., Industrial Color Testing, 2nd Ed., Wiley-VCH Verlag GmbH & Co. KgaAm, Weinheim, Germany, 2002. [5] “Color Handbook,” SpecialChem’s Coating & Inks Formulation Bulletin, Issue No. 106, August 2, 2007; www. specialchem4coatings.com/tc/color/. [6] ASTM D16-11, “Standard Terminology for Paint, Related Coatings, Materials, and Applications,” Annual Book of ASTM Standards, Vol. 06.01, ASTM International, West Conshohocken, PA. [7] ASTM E284-09A, “Standard Terminology of Appearance,” Annual Book of ASTM Standards, Vol. 06.01, ASTM International, West Conshohocken, PA. [8] Colour Index International, Fourth Edition Online in two parts, http://www.colour-index.org/newusr1.asp, 2002. [9] ASTM D4303-10, “Standard Test Methods for Lightfastness of Colorants Used in Artists’ Materials,” Annual Book of ASTM Standards, Vol. 06.02, ASTM International, West Conshohocken, PA. [10] Zeller, R. C., “The Meaning of Tint Strength,” Color Res. Appl., Vol. 3, 1978, p. 34. [11] Vernardakis, G., American Ink Maker, Vol. 62, No. 2, 1984, p. 24. [12] ASTM D4838-88(2010), “Standard Test Method for Determining the Relative Tinting Strength of Chromatic Paints,” Annual Book of ASTM Standards, Vol. 06.02, ASTM International, West Conshohocken, PA. [13] ASTM E308-08, “Practice for Computing the Colors of Objects by Using the CIE System,” Annual Book of ASTM Standards, Vol. 06.01, ASTM International, West Conshohocken, PA. [14] ASTM D1640-03(2009), “Test Methods for Drying, Curing, or Film Formation of Organic Coatings at Room Temperature,” Annual Book of ASTM Standards, Vol. 06.01, ASTM International, West Conshohocken, PA. [15] ASTM E1164-09a, “Standard Practice for Obtaining Spectrometric Data for Object-Color Evaluation,” Annual Book of ASTM Standards, Vol. 06.01, ASTM International, West Conshohocken, PA. [16] ASTM E1347-06, “Standard Test Method for Color and ColorDifference Measurement by Tristimulus Colorimetry,” Annual Book of ASTM Standards, Vol. 06.01, ASTM International, West Conshohocken, PA. [17] ASTM D387-00(2008), “Standard Test Method for Color and Strength of Chromatic Pigments with a Mechanical Muller,” Annual Book of ASTM Standards, Vol. 06.01, ASTM International, West Conshohocken, PA. [18] ASTM D3022-84(2005), “Test Method for Color and Strength of Color Pigments by Use of a Miniature Sandmill,” Annual Book of ASTM Standards, Vol. 06.01, ASTM International, West Conshohocken, PA. [19] ASTM D2066-07, “Standard Test Methods for Relative Tinting Strength of Paste-Type Printing Ink Dispersions,” Annual Book of ASTM Standards, Vol. 06.02, ASTM International, West Conshohocken, PA. [20] ASTM D2745-00(2008), “Standard Test Method for Relative Tinting Strength of White Pigments by Reflectance Measurements,” Annual Book of ASTM Standards, Vol. 06.01, ASTM International, West Conshohocken, PA. [21] ASTM D332-87(2008), “Standard Test Method for Relative Tinting Strength of White Pigments by Visual Observation,” Annual Book of ASTM Standards, Vol. 06.01, ASTM International, West Conshohocken, PA. [22] ASTM D262-81(1999), “Specification for Ultramarine Blue Pigment,” Annual Book of ASTM Standards, ASTM International, West Conshohocken, PA. (This ASTM specification is included for historical purposes—it was withdrawn in December 2005 because the specifications for this pigment are no longer in the United States). [23] Pal, R., Rheology of Particulate Dispersions and Composites, CRC Press, Taylor & Francis Group, New York, 2006, p. 440. [24] Wissling, P., Metallic Effect Pigments: Basics and Applications, William Andrew Publishing, New York, 2006, p. 300.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 46
Q
MASS COLOR AND TINTING STRENGTH OF PIGMENTS
[25] Kunjappu, J. T., Essays in Ink Chemistry (For Paints and Coatings Too), Nova Science Publishers, Inc., New York, 2001, p. 136. [26] Lewis, P. A., Pigment Handbook, Properties and Economics, 2nd ed., John Wiley & Sons, Inc., New York, 1988, p. 976. [27] Varley, D. M., and Bower, H. H., Journal of the Oil and Colour Chemists Association, Vol. 62, 1979, p. 401. [28] Patton, T. C., Paint Flow and Pigment Dispersion, 2nd ed., Wiley-Interscience, New York, 1979. [29] Carr, W., Journal of the Oil and Colour Chemists Association, Vol. 61, 1978, p. 397. [30] Hafner, O., Journal of the Oil and Colour Chemists Association, Vol. 57, 1974, p. 268. [31] Parfitt, G. D., Dispersion of Powders in Liquids, 2nd ed., WileyInterscience, New York, 1973. [32] Hays, B. G., American Ink Maker, Vol. 62, No. 6, 1984, p. 28. [33] Hampton, J. S., and MacMillan, J. F., American Ink Maker, Vol. 63, No. 1, 1985, p 16. [34] Topham, A., Prog. Org. Coat., Vol. 5, 1977, p. 237. [35] Merkle, K., and Schafer, H., Pigment Handbook, Vol. III, T. C. Patton, Ed., Wiley-Interscience, New York, 1973, pp. 157–167.
597
[36] Linden, H., Rutzen, H., and Wegemund, B., United States Patent 4, 167,421 (1979). [37] Hauxwell, F., Stansfield, J. F., and Topham, A., U. S. Patent 4,042,413 (1977). [38] Franklin, M. J. B., Goldbrough, K., Parfitt, G. D., and Peacock, J., J. Paint Technol., Vol. 42, 1970, p. 740. [39] Ayers, J. W., “A Discussion of the Accuracy and Utility of Methods of Test for Mass Tone and Tinting Strength,” Proceedings, ASTM International, West Conshohocken, PA, Vol. 34, Part II, 1934, p. 497. [40] Stutz, G. F. A., “Tinting Strength of White Pigments,” Proceedings, ASTM International, West Conshohocken, PA, Vol. 34, Part II, 1934, p. 521. [41] Pall, D. B., “A New All-Glass Mill,” Ind. Eng. Chem. Anal. Ed., Vol. 14, 1942, p. 346. [42] Paul, M. R., and Diehlman, G., “Method and Apparatus for Determining the Interval Required to Disperse Oil Colors Throughout a Paint Medium,” Proceedings, ASTM International, West Conshohocken, PA, Vol. 34, Part II, 1934, p. 490.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
Part 11: Physical and Mechanical Properties
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
47
MNL17-EB/Jan. 2012
Adhesion Gordon L. Nelson1 PREFACE
IN PREPARATION OF THIS CHAPTER, THE CONTENTS of the fourteenth edition were drawn upon. The current edition will review and update the topics as previously addressed, introduce new technology that has been developed, and include additional references.
INTRODUCTION
Organic coatings are applied to a variety of substrate materials (woods, metals, plastics, ceramics) for decorative, protective, and functional applications. In each case, it is imperative that the coating adheres well to the substrate. Accordingly, adhesion assessments should be an integral part of the coating development. This may seem a straightforward task, but coating adhesion is, in fact, extremely complex and often poorly understood. The growing use of plastics to replace metals and other “traditional” materials renders the issue even more complex [1,2]. The objectives of this chapter are to briefly review salient concepts of the adhesion process and to discuss currently accepted standard test methods.
FUNDAMENTAL CONCEPTS
While the very definition of “adhesion” is of some controversy [1], adhesion may be loosely defined as the attraction between dissimilar bodies for one another. ASTM D907 on Terminology of Adhesives defines adhesion as “the state in which two surfaces are held together by interfacial forces which may consist of valence forces or interlocking action or both.” In discussing adhesion assessment, one may consider the issue from two different aspects: basic adhesion and practical adhesion. Basic adhesion signifies the summation of all interfacial, intermolecular forces, whereas practical adhesion is used to represent the forces or work required to disrupt the adhering system [3]. The next section will be devoted to theories and concepts of basic adhesion and will out of necessity be brief. The perspective will be from that of a polymeric coating on a plastic substrate [2]. Metal and metal oxide substrates will be discussed where appropriate. The reader is referred to Refs. [4–7] for a more thorough discussion.
BASIC ADHESION
Work of Adhesion
Bonding between polymeric coatings and substrates may be viewed as the union of two contiguous polymer phases, one a solid and the other a liquid that solidifies to form a
1
thin film. The reversible separation of the two phases may be expressed by the work of adhesion Wa = Y1 + Y2 – Y12
(1)
where Wa is the work of adhesion, and Y1 and Y2 are the surface tensions of the two phases. The maximum force per unit area, σ2, to effect this process is the ideal adhesive strength [8]
2 = [16 / 9(3)12 ](Wa / Z0 )
(2)
where Z0 is the equilibrium separation between the two phases, usually about 5 Å. The average value Wa for polymers is typically 50 ergs/cm2, yielding a theoretical value for σ2 of 15 000 psi (103 mPa). For practical purposes this value is never attained, primarily due to the fact that perfect inter-molecular contact is most unlikely. In fact, this is at least an order of magnitude higher than the practical adhesive strength usually observed. This deviation from ideality has led to the promulgation of several theories of adhesion, none of which are universally recognized [9]. This is not unexpected, since most theories deal exclusively with the mechanisms of bond formation and disregard the fact that bond strength is ultimately a function of both the degree of bond formation, the nature of the bond (chemical and physical), and the rheological properties of the bonding phases. The strength of an adhesive bond is, in fact, a function of all of these factors. Summary paragraphs about basic theories of adhesion follow below.
Fracture Theory
The area of interfacial bonding between coating and substrate will, in most instances, contain voids or defects. The result is deviation between the ideal adhesive strength and the practical limit. Good [10] and Williams [11] have applied the theory of cohesive fracture to coating fracture. The concept that fracture propagates from the weakest point, a defect, is fundamental to fracture mechanics. The strength of a bond, in terms of the energy required to induce fracture, is described as a function of the defect size and the energy dissipated by irreversible processes (plastic deformation, light emission, and electric discharge). The general equation given is f = k(EF /d )1/2
Dean, College of Science, Florida Institute of Technology, 150 West University Blvd., Melbourne, FL 32901-6975.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed byASTM International This standard is for EDUCATIONAL USE ONLY. Copyright © 2012 by www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
(3)
601
602
PAINT AND COATING TESTING MANUAL
Q
where f is the fracture stress, k = (4/π)1/2, E the elastic modulus of the material, d the defect length, and F the fracture energy or total work per unit area of fracture surface, which is dissipated. Accordingly, the fracture energy f = Wa + Wi
(4)
where Wa is the work of adhesion and Wi is the total work for the irreversible processes. For all practical purposes Wi is much greater than Wa therefore [12,13] f = Wi
(5)
Weak Boundary Layer Theory
Proposed by several workers [14–16], the weak boundary layer theory maintains that true interfacial fracture does not occur, and that fracture usually occurs cohesively in a weak boundary layer, which may be near the interface between coating and substrate. Experimental evidence has perhaps disproved the first point [10], that true interfacial failure can occur, but it has been shown that the second is valid in some instances [17,18]. Of practical importance to WBL investigations is the locus of fracture, which can occur in one or more of the zones in Fig. 1. For a strong bond, the boundary layer (layers) must be rheologically sound and chemically durable. The zone of failure may, of course, be studied by scanning electron microscopy (SEM), transmission electron microscopy, and Fourier transform-infrared [19]. In fact, thin layers of coating have been noted where interfacial failure was only thought to occur on visual examination.
Wetting-Contact Theory
The wetting-contact theory states that van der Waals attractive forces alone provide sufficient coating/substrate bond strength given perfect molecular contact, and that the extent of contact and resulting bond strength are functions of wetting energetics [20–28]. No one denies the importance of wetting in adhesion, i.e., the lower the contact angles, the more the interfacial area of contact, which generally yields improved adhesion (Fig. 2). However, thermodynamic wetting is a necessary but not sufficient condition for the establishment of coating film adhesion. Wetting is a kinetic phenomenon as well. Furthermore, this model does not consider the effects of weak boundary layers or the effects of defects or fracture
Fig. 1—Weak boundary layer theory. Possible zones of failure (after Good [19]).
15TH EDITION
mechanics. What is the effect of surface contamination on the contact angle? For guidance on the measurement of contact angle, see ASTM D7334 (WK11928) Standard Practice for Surface Wettability of Coatings, Substrates and Pigments by Advancing Contact Angle Management.
Diffusion Theory
Voyutskii and others [29–33] have proposed that coating adhesion between high polymers arises from interdiffusion. This has been validated somewhat by the observation that adhesive strength is a function of polymer molecular weight, structure, and contact time. Key to this theory is the concept that no clear-cut inter-facial boundary exists, but rather that an interphase exists which consists of polymer chain segments from both contiguous phases. The theory has been criticized for giving insufficient weight to the contribution of van der Waals attractive forces. Nevertheless, for the case of solvent-borne organic coatings applied to plastic substrates, this phenomenon is intuitively appealing. Interdiffusion of coating and substrate polymer molecular segments is a function of polymer-polymer compatibility. An interphase forms as a result of the blending of the two phases. Although polymer pairs are generally incompatible, Helfand and others [34–38] have utilized statistical thermodynamics to predict that interdiffusion also occurs as a result of the tendency for free energy at the interface to minimize. A useful equation that describes interfacial thickness as a function of the Flory–Huggins interaction parameter as derived from Hansen solubility parameters is [36,38] ai = 2(m / x)1/2
(6)
where ai is the interfacial thickness expressed as the crosssectional area of a lattice cell, m is a lattice constant directly proportional to nearest-neighbor contacts, and x is the Flory– Huggins interaction parameter. By this equation, interfacial thickness increases by interdiffusion as the solubility parameter difference between the two phases decreases. Although experimental verification is scarce, adhesive strength has been shown to decrease with increasing disparity in the solubility parameters of the two phases [39]. It is important to note that diffusion should occur in the latter stage of bond formation, the first stage being wetting to establish contact. Additional support for the theory comes from the observation that adhesive strength between certain high polymers increases with time [30]. This theory, however, cannot be applicable to systems involving one or more hard solids (metals, glass, or metal oxides).
Fig. 2—Contact angle.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 47
Chemical Adhesion
The bond strength of covalent bonds is 1 to 2 orders of magnitude greater than van der Waals attractive forces. There have been numerous applications where interfacial chemical bonds have been employed to promote adhesion. Dramatic increases in bond strength have been reported [40]. Coupling agents such as chrome complexes [41], silanes [42], and titanates [43] have been used effectively, Organic functionalities, including isoeyanates, carboxyls, amides, amines, hydroxyls, and epoxides [43,44], have been reacted interfacially to enhance adhesion. Adhesive strength has been shown to increase with functional group concentration as expressed by f = kC n
ADHESION
603
modification of surface acidity or basicity of the substrate should increase adhesion. Modification of surface acidity or basicity of inorganic solids can increase mechanical properties (modulus, extension to break, and toughness) of coatings on these substrates. It has also been claimed that the reverse is true, however [51,52].
Combination of Phenomena
The preceding theories interpret adhesion in terms of single phenomena, each of which undoubtedly plays some role in interfacial bonding between coating and substrate. One should be cautioned, however, against exclusive use of a single theory to explain the adhesion of a given system. A more logical approach has been proposed by Allen [53].
(7)
where f is the adhesive strength, C the functional group concentration, and k and n are positive constants [45]. There, however, appears to be an upper limit for functional group concentration above which adhesion may tend to decrease [46,47].
Mechanical Adhesion
In mechanical adhesion substrate roughness is thought to provide a mechanical locking of the coating to the substrate. However, if there is not intimate contact between the coating and the substrate, then increased roughness should lead to a decrease in adhesion by producing uncoated voids. In the practical application of electroless metals on polymer substrates, etchants are used that create deep channels, increasing adhesion. Other examples are anchor coats for polytetra fluoroethylene, adhesion to porous anodic films on aluminum, and hot melt polyolefin coatings on metals. Mechanical adhesion is also important for porous substrates such as wood, cloth, and paper. Since good adhesion can be obtained on smooth surfaces such as glass, questions can be raised about the general validity of the mechanical adhesion mechanism. It should also be noted that mechanical abrasive treatment of a solid surface may also yield macroradicals and active secondary chemical products, which if they do not come into contact with atmospheric oxygen, may interact with components of the coating. This has been shown for certain adhesives. Active radicals or functionality would yield adhesion, better described as mechanico-chemical adhesion than just mechanical [7,48,49].
Electrostatic Adhesion
In the theory of electrostatic adhesion, when two dissimilar materials are brought into contact, a charge transfer takes place, which results in the formation of an electrical double layer, much like a capacitor. Work would then be required to separate the two charged layers. This is thought to be particularly applicable to metal-polymer bonds. Indeed, ionizing discharge has been shown to affect a copper to acrylic bond but not copper on salt (NaCl) or glass. This theory would not be applicable to two nonpolymer systems [49,50].
Acid-Base Adhesion
Q
In the acid-base adhesion theory, it is said that the strength of the adhesive bonds is increased significantly by acid base interactions between coating and substrate. Appropriate
" = a" A + b" B + c" C + …
(8)
which suggests that a combination of phenomena is more realistic, that is, that basic adhesion is the summation of all interatomic or intermolecular interactions at the interface. One would think that systematic studies could be made to assess the contribution of a given variable to the adhesion process, with the others being held constant. This would reveal useful information as to factors critical to coating adhesion for a particular system. Unfortunately, being able to accomplish that task is questionable given that adhesion assessment not only involves factors of basic adhesion but variation in application of the applied external stress (tensile, shear, or peel) and many other factors. It can be concluded that adhesion is an interfacial phenomenon in which both physical and chemical forces operate when surfaces develop to form an interface. Adhesive strength is a measure of the degree to which the two surfaces are attracted. This is a function of wettability, relative surface energetics of both phases, and of the kinetics of wetting. For integrity of a bond at the interface between a coating and a substrate, one needs to consider the following factors: 1. Thermodynamics and kinetics of the formation of the bond. 2. The forces acting near the interface in both the coating and the substrate. 3. The cohesive forces within the coating layer. 4. Internal stresses in the coating layer. 5. The behavior of the coating layer under stress. To understand basic adhesion one must understand the surface chemistry, surface physics, surface architecture, coating polymer chemistry and physics, polymer rheology, coating internal stresses, fracture mechanics, and effects of changes in the environment. In fact, it has been noted that spontaneous loss of adhesion can occur due solely to internal strain of the coating [54–56].
Effects of Substrates ADDITIONAL CHEMISTRY
The adhesion of organic coatings to metals is at a high level of development in the practical sense. The contribution of surface energy, chemical functionality, surface irregularities, and contaminants (oxides, adsorbed water, etc.) have been identified. On the other hand, coating adhesion to plastic substrates has presented additional complexities. Polymer
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
604
PAINT AND COATING TESTING MANUAL
Q
surfaces are often more difficult to wet and bond because of low surface energy, incompatibility, chemical inertness, or the presence of contaminants (oils, lubricants, plasticizers, etc.), and weak boundary layers. Surface modification techniques have been developed to enhance adhesion. In the absence of surface preparation, coating adhesion is felt to be a function primarily of van der Waals and/or polar-dipolar attractive forces, mechanical adhesion (from surface irregularities), and interdiffusion. This latter contribution to adhesion has been studied using SEM. In Fig. 3 is presented a scanning electron micrograph of a thermoplastic acrylic coating applied to modified polyphenylene oxide (a blend of PPO and polystyrene), an important engineering thermoplastic. The first coating contains 60 % of the recommended amount of solvent. The interface formed is sharp and well-defined. In this instance, adhesion appears to be due largely to mechanical and attractive forces. Fig. 4 shows the same substrate and coating, which in this case contains 100 % of the recommended amount of solvent. The sharp interface seen in Fig. 3 is no longer present, an intermediate zone or interphase having formed in its place. Diffuse interface formation of this magnitude is undoubtedly unique to coatings applied to plastics. Bond strength will be a function of formulation, solvent content, and drying time [2]. Organic coatings are complex formulations, thus, the polymer chemistry at the developing coating substrate interface will clearly be impacted by the solid surface, whether plastic, metal, or inorganic, that is, the polymer
Fig. 3—SEM of sharp interface.
15TH EDITION
interface may be somewhat different than the bulk coating (i.e., a boundary layer). Fully developed coatings may also be porous or permeable, thus aging and weathering may impact adhesion, as water, oxygen, and other agents penetrate to the polymer coating solid interface [7,57]. Indeed, some coatings that have acceptable adhesion while dry may fail badly when tested under high humidity or after immersion in water for several hours. It has been shown for polyolefin-metal adhesion that oxidation of the polymer attached to the surface occurs through action of oxygen absorbed by the metal and the polymer. The appearance of oxygen-containing groups in the otherwise inert polymer should promote an increase in the extent of interaction with the metal. Adhesion in this case would also be increased by oxygen donor fillers [7,57]. All metals (except gold) are known to exist with an oxide film on their surfaces. The volume of oxide formed may be smaller than the metal reacted, and consequently the oxide film is porous and nonprotective (alkali and alkaline earth metals) or the volume may be larger and therefore protective (transition metals and aluminum). However, in the latter case, migration of metal cations from the surface leaves vacancies, which aggregate to form cavities. Treatment of a metal before coating to produce a preferred or controlled bonding surface is therefore common [7,57]. From the above discussion, it is clear that chemical interaction between coating and solid can occur even when not recognized or anticipated. Finally, a change in substrate may impact adhesion. Weathering failures such as blistering and scab corrosion are often regarded as adhesion failures by coatings development chemists. Delamination of coatings in the absence of substrate corrosion can also be produced by weathering. A common problem involves the interfacial chalking of an epoxy primer and a topcoat that has a high UV light transmission. In the presence of UV light, moisture, and oxygen, the epoxy primer is degraded at the interface between the primer and topcoat, leading to delamination of the topcoat from the primer. Delamination of clear or semitransparent exterior wood coatings can also occur by UV, water, and oxygen attacking the wood substrate. Delamination of the intact coating results when the wood substrate coating interface is destroyed. The solution to this problem is to add organic and inorganic UV absorbers to the coating to protect the wood substrate from degradation.
PRACTICAL ADHESION
Fig. 4—SEM of diffuse interface.
Given the complexities of the adhesion process, can adhesion be measured? As Mittal [3] has pointed out, the answer is both “yes” and “no.” It is reasonable to state that at the present time no test exists that can precisely assess the actual physical strength of an adhesive bond. But it can also be said that it is possible to obtain an indication of relative adhesion performance. Practical adhesion test methods are generally of two types: implied and direct. Implied type tests include indentation or scribe techniques, rub testing, and wear testing. Criticism of these tests arises when they are used to quantify the strength of adhesive bonding. But this, in fact, is not their purpose. An implied test should be used to assess coating performance under actual service conditions. Direct measurements, on the other hand, are intended expressly
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 47
to measure adhesion. Peel, lap-shear, and direct tensile are common examples. Meaningful tests of this type are highly sought after, primarily because the results are expressed by a single discrete quantity, the force required to fracture the coating/substrate bond under prescribed conditions [2].
Test Methods
In practice, a battery of tests is used to evaluate adhesion by inducing bond rupture by different modes. Criteria deemed essential for a test to warrant large-scale acceptance are: use of a straightforward and unambiguous procedure, relevance to its intended application, reproducibility, and quantifiability, including a meaningful rating scale for assessing performance. Test methods used for coatings on metals are peel adhesion or “tape testing,” Gardner impact flexibility testing, and adhesive joint testing including shear (lap joint) and direct tensile (butt joint) testing. These tests do not, in fact, strictly meet the criteria listed, but an appealing aspect of the above tests is that in most cases the equipment/instrumentation is readily available or can be obtained at reasonable cost [2]. A wide diversity of test methods has been developed over the years. In this chapter only selected test methods developed through the consensus process will be discussed in detail. The reader should recognize, however, that numerous test methods have been developed that measure aspects of adhesion [7,58–61] and that there generally is difficulty in relating these to basic adhesion phenomena (Fig. 4).
The Tape Test
By far the most used test to access coating “adhesion” is the peel test. In use since the 1930s, in its simplest version, a piece of adhesive tape is pressed against the paint film. The test consists of observing whether the film is peeled off when the tape is removed. The method can be refined to measure the force required for film removal. In other tests, crosses or a cross-hatched pattern are cut into the coating, a tape applied and removed, and the coating removal assessed against an established rating scale. The current widely used version was first published in 1974 as ASTM D3359, Standard Test Methods for Measuring Adhesion by Tape Test. Two test methods are covered. This is one of the top-selling ASTM test methods, regardless of field. These test methods cover procedures for assessing the adhesion of coating films to metallic substrates only, by applying and removing pressure-sensitive tape over cuts made in the film. Method A is primarily intended for use at job sites, while Method B is more suitable for use in the laboratory. The cross-hatch test, Method B, is not considered suitable for films thicker than 5 mils (125 μm). Both test methods are used to establish whether or not the adhesion of a coating to a substrate is at a generally adequate level. They do not distinguish between higher levels of adhesion for which more sophisticated methods of measurement are required. In multicoat systems, adhesion failure may occur between coats so that the adhesion of the coating system to the substrate is not determined. In Test Method A an X-cut 1.5 in. (3.8 cm) long is made in the film (to the substrate) with a sharp cutting device. A 1-in. (2.5-cm)-wide pressure sensitive tape is applied over the cut and firmly adhered with a pencil eraser and then removed, and adhesion is assessed qualitatively on a 0 to 5 scale.
Q
ADHESION
605
In Test Method B, a lattice pattern with either 6 or 11 cuts in each direction is made in the film (to the substrate), pressure-sensitive tape is applied over the lattice and then removed, and adhesion is evaluated by comparison with descriptions and illustrations. For Test Method A, the following rating scale is used: 5A = No peeling or removal. 4A = Trace peeling or removal along incisions. 3A = Jagged removal along incisions up to 1/16 in. (1.6 mm) on either side. 2A = Jagged removal along most of incisions up to 1/8 in. (3.2 mm) on either side. 1A = Removal from most of the area of the X under the tape. 0A = Removal beyond the area of the X. For Test Method B, 3/4 in. (1.9 cm) cross-cuts are made. For coatings having a dry film thickness up to and including 2.0 mils (50 μm), 11 cuts are spaced 1 mm apart. For coatings having a dry film thickness between 2 mils (50 μm) and 5 mils (125 μm), 6 cuts are spaced 2 mm apart. For films thicker than 5 mils (125 μm), Method A is used instead of Method B. For Test Method B, adhesion is rated according to the following scale (as illustrated in Fig. 5): 5B = The edges of the cuts are completely smooth; none of the squares of the lattice is detached. 4B = Small flakes of the coating are detached at intersections; less than 5% of the area is affected. 3B = Small flakes of the coating are detached along edges and at intersections of cuts. The area affected is 5 to 15 % of the lattice. 2B = The coating has flaked along the edges and on parts of the squares. The area affected is 15 to 35 % of the lattice. 1B = The coating has flaked along the edges of cuts in large ribbons, and whole squares have detached. The area affected is 35 to 65 % of the lattice. 0B = Flaking and detachment worse than Grade 1. Repeatability within one rating unit is generally observed for coatings on metals for both methods, with reproducibility of 1 to 2 units. The method is widely used and is viewed as “simple” as well as low in cost.
Peel Adhesion Testing on Plastic Substrates
ASTM D3359 has drawn fire when used for substrates other than metal, such as plastics. The central issues are that the test may lack reproducibility and does not relate to its intended application. Poor reproducibility is a direct result of several factors intrinsic to the materials employed and the procedure itself. More importantly, in this instance the test is being applied beyond its intended scope. ASTM D3359 was designed for relatively ductile coatings applied to metal substrates, not for coatings (often brittle) applied to soft plastic parts [1]. Nevertheless, the tape test enjoys widespread popularity. The tape test is economical to perform, lends
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
606
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 5—Classification of adhesion test results. From ASTM D3359.
itself to job site application, and most importantly, after decades of use, people feel comfortable with it [42]. Unfortunately, the often unique functional requirements of coatings on plastic substrates dictate that the tape test as written may not be a satisfactory measure of practical adhesion performance. When a flexible adhesive tape is applied to a rigid coated substrate surface and then removed, the removal process has been described in terms of the “peel phenomenon,” as illustrated in Fig. 6. Peeling begins at the “toothed” leading edge (at the right) and proceeds along the coating-adhesive interface. It is reasonable to assume that any coating removal is due to a tensile force generated along this interface, which would be a function of the rheological properties of the backing and adhesive layer materials and the strength of the bond between the adhesive layer and the coating surface. Note, however, that in actuality this force is distributed over a discreet distance (0 to A in Fig. 6), which relates directly to
Fig. 6—Peel profile [62].
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 47
the properties described [63], not concentrated at a point (0 in Fig. 6) as in the theoretical case, though tensile force is greatest at the origin for both. It is worthwhile to note the significant compressive force which arises from the response of the tape-backing material to being stretched, which precedes the tensile force. Both tensile and compressive forces are involved in tape testing. Close scrutiny of the tape test with respect to the nature of the tape employed and certain aspects of the procedure itself reveal several factors, each or any combination of which can dramatically affect the results of the test as discussed below.
The Tape Controversy
Beginning in 1991, given the withdrawal of the originally specified tape, 3M-710, ASTM D3359 no longer specified a specific tape. Differences in tapes can lead to different results for reasons noted above since small changes in backing stiffness and adhesive rheology cause large changes in the tension area. It is also important to note that tapes, like most products, are manufactured to meet minimum standards. A given lot may surpass these criteria and be suitable for general market distribution, but a source of serious and unexpected error for tape testing [63]. There was, in fact, a commercially available tape test kit that included a tape with adhesion variations of up to 50 % claimed by the manufacturer [64]. And, of course, tapes age on storage. Bond strengths change over time. While there are tapes available that would appear to deliver consistent performance, a given tape does not adhere equally well to all coatings. For example, the peel removal force of the tape previously recommended by ASTM (3M-710) was examined with seven coatings. It was found that while peel was indeed consistent for a given coating, it varied by 25 % between the highest and lowest ratings among coatings. This observation could be the result of several factors, notably coating composition and topology, but the bottom line is that no single tape is likely to be suitable for testing all coatings. It is also useful to note the tape test does not give an absolute value for the force required for bond rupture, but serves only as an indicator that some minimum value for bond strength was met or exceeded [1,2].
Procedural Problems
The tape test is operator intensive. By design it was made as simple as possible to perform and requires a minimum of specialized equipment and materials that must meet certain specifications. Therefore, the burden of accuracy and reproducibility relies largely upon the skill of the operator and his/her ability to perform the test in a consistent manner. Key steps that directly reflect the importance of operator skill include the angle and rate of tape removal and the visual assessment of the tested sample. It is not unexpected that different operators might obtain different results [1,2].
PEEL ANGLE AND RATE
The standard requires that the free end of the tape be removed rapidly at as close to a 180° angle as possible. But if peel angle and rate vary, the force required to remove the tape can change dramatically. Nearly linear increases were observed in peel force approaching 100 % as peel angle was changed from 135 to 180°, and similar large differences
Q
ADHESION
607
can be expected in peel force as peel rate varies. These effects are related in that they both reflect certain rheological properties of the backing and adhesive, which are molecular in origin, but the most useful conclusion is that these phenomena can make large contributions and must be minimized to assure reproducibility [65].
VISUAL ASSESSMENT
The final step in the test is visual assessment of coating removed from the specimen, and this can be subjective in nature. This assessment can vary among individuals evaluating the same specimen [65]. Performance in the tape test is based on the amount of coating removed compared to a relative scale. But it was found that the exposure of substrate can be due to factors other than coating adhesion, arising from the requirement that the coating be cut (hence the synonym “cross-hatch adhesion test”). Justification for the cutting step is reasonable; cutting provides a free edge from which peeling can begin without having to overcome the cohesive strength of the coating layer. This might be suitable for coatings applied to metal substrates, but for coatings applied to plastics, the cutting process can lead to false indications of poor adhesion. This is due to the unique interfacial zone mentioned earlier. For coatings on soft plastics, how deep should this cut penetrate? Is it possible to cut only to the interface? If microscopic examination of panels is included, in several instances it is clearly evident that coating removal results from substrate failure at or below the interface, not from adhesive failure between coating and substrate. At the same time, it is also observed that cohesive failure within the coating layer is a frequent occurrence. The latter observation is significant in that the tape test assessment criteria make no provision for it [1,2].
Mechanized Tape Test
A mechanized tape test was published by ASTM Committee B 8 in 2000. The scope of ASTM B905, Test Methods for Assessing the Adhesion of Metallic and Inorganic Coatings by the Mechanized Tape Test, includes the assessment of adhesion of metallic and inorganic coatings and other thin films to metallic and nonmetallic substrates. A pressuresensitive tape is applied to the coated surface and then a mechanized device is utilized to remove the tape at a regulated, uniform rate and constant angle while the removal force is simultaneously recorded (Fig. 7). Four methods are described: A1 and A2, which do not involve cuts made to the coating (for use on parts), and B1 and B2, which use a crosshatch pattern of six lines (plus six at 90°) for laboratory use. Methods 1 and 2 refer to peel angles of 90° and 180°, respectively. Methods B1 and B2 are not recommended for polymeric substrates. For Method A coating detachment (failure) or no coating detachment (pass) are reported versus the maximum registered peel force. For Methods B a classification similar to ASTM D3359 is reported versus the maximum registered peel force. The spacing of cuts, however, is 4 mm versus 1 or 2 mm in D3359. In B 905 a 90° peel angle is preferred “due to the stress strain behavior of the tape,” versus the 180° peel angle of D3359. An interlaboratory study during the development of the test method revealed that the maximum peel force of the tape depends significantly on the type of coating, the type of tape, the peel rate, and the peel angle. The pressure used
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
608
PAINT AND COATING TESTING MANUAL
Q
Fig. 7—Mechanized tape tester.
for adhering the tape (a roller is used) was not significant. Results for the four test methods should not be compared. While the methods provide a regulated uniform peel rate and constant angle of test, and most importantly provide a way of measuring peel force, the other underlying issues in tape testing are not overcome. If two different coatings have different peel forces despite a single tape, peel rate, and angle of peel, then the adhesion rating/results are still not directly comparable.
Other Peel Tests
The corresponding test in the International Standards Organization to ASTM D3359 is ISO 2409. The crosshatch test is used in ISO 2409 up to a coating thickness of 10 mils (250 μm). A single crosscut test is used above 10 mils, versus 5 mils (125 μm) for D3359. A 1-mm crosshatch spacing is used for a coating thickness up to 60 μm for hard substrates, a 2-mm spacing for soft substrates, a 2-mm spacing for coatings 61–120 μm, and a 3-mm spacing above 120 μm. In ISO 2409 six cuts for the 90° cross-hatch are made in each direction. In D3359 a 1-mm spacing is used for coatings up to a thickness of 50 μm (11 cuts in each direction) and a 2-mm spacing for coatings 50–125 μm (6 cuts), with the single crosscut test applied at higher coating thickness. In ISO 2409 the tape is removed at an angle of 120° (angle away from operator) versus 180° in D3359. The classification scheme is 0 to 5 with 0 meaning no coating detachment in ISO 2409. This is the opposite in D3359, with ratings 5B to 0B, with 0B signifying greater than 65% coating removal. Thus ISO 2409 and ASTM 3359, while both being tape peel tests, differ in many significant ways, not the least of which being that the classification schemes signify opposite results. Numerous tape peel tests exist in the industry. Examples include the General Motors Engineering Standard GM 9071P-Tape Adhesion Test for Paint Finishes, which is similar to ASTM D3359, and Boeing Specification Support Standard (BSS) 7225, which provides more varied crosshatch patterns than D3359 and provides specific procedures for tape testing after wet immersion.
Direct Tensile Testing
A long-used approach to coating adhesion testing is the direct tensile test, perhaps “conceptually” the simplest of
15TH EDITION
all methods for measuring adhesion. A dolly or stud is bonded to the coating film. The normally applied force that is required to remove the film is measured. If failure occurs at the substrate-film interface, this force is taken to be the “force of adhesion.” An obvious limitation is, of course, the strength of the adhesive bond of the stud to the cured coating. Such methods have been available since the 1930s. Many of these test methods have unfortunately suffered from their own lack of reproducibility. This is not unexpected since the forces involved are not quite as simple as appearance would have it [5,66]. It is essential that the force is applied strictly in the direction normal to the sample and that no bending moment is active across the test area. Deviations from symmetry in the test arrangement, poor alignment, deviations from homogeneity and of thickness of the adhesive (coating), and random variations in the strength of the bond between film and substrate affect test results [5,66]. The stress at locations where the adhesive film is thinner will be higher than the average stress and will be transmitted to the film under test. Another factor may be peeling during the test, which is not easily identified or analyzed. The adhesive used to bond a stud to the coating has the potential to influence the coating film properties by penetration through the film into microcracks and possibly into the substrate [66]. Test adhesive flexibility may also be an issue, as well as the flexibility of the substrate, if the sample is unrestrained. There exists now within ASTM both laboratory and field versions of direct tension tests for coatings. Test Method for Measuring Adhesion of Organic Coatings to Plastic Substrates by Direct Tensile Testing. ASTM D5179, while limited to organic coatings on plastics, uses a restrained sample and commonly available tensile test apparatus. The second, Method for Pull-Off Strength of Coatings Using Portable Adhesion-Testers, ASTM D4541, defines a class of portable pull-off adhesion testers for field evaluation of coating adhesion. ASTM D5179 is the successor to numerous attempts to develop a reproducible coating tension test and was approved in 1991. It will be discussed first.
ASTM D5179
This test covers the laboratory determination of adhesion of organic coatings to plastic substrates by mounting and removing an aluminum stud from the surface of a coating and measuring the force required to break the coating/substrate bond with a tensile tester. The test method provides an inexpensive test assembly, which can be used with most tensile test machines. The method is used to compare the pulloff strength (commonly referred to as adhesion) of coatings to various plastic substrates, thus allowing for a quantitative comparison of various coating/substrate combinations. A carefully prepared 2-cm diameter aluminum stud is bonded directly to a coated, cured panel using a cyanoacrylate adhesive. The adhesive is allowed to cure for 2 h at room temperature. Adhesive buildup is removed from around the stud. The specimen is then subjected to test on a tensile tester equipped with an upper coupling adapter and a restraining device (Fig. 8) to provide for sample alignment and minimal substrate flexing. The sample bearing the stud is installed in the restraining device, with only the stud protruding. The tensile machine crosshead is lowered so the upper coupling adaptor can be attached to the specimen.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 47
Q
ADHESION
609
Fig. 8—Direct tensile restraining device. From ASTM D5179.
When testing thin substrates, a piece of plastic is placed in the restraining device behind the specimen to insure a rigid assembly. The tension test is conducted, and pull strength recorded. Each specimen is rated according to type of coating failure, as follows: A = Adhesive failure of the coating from the substrate. C = Cohesive failure in the coating. AC = Combination of adhesive failure at the coating/substrate interface and cohesive failure in the coating. S = Adhesive failure at the stud. CS = Combination of adhesive failure at the stud and cohesive failure in the coating. For multilayer coatings, failure between the layers is noted and labeled as CM. Five specimens of each coated substrate are tested one day and five on a second day. If one specimen differs significantly from the other four tested at the same time, fails because of an uneven (nonplanar) stud, or for another reason performs unlike the other four, a replacement specimen is tested. The stud and specimen are carefully examined. The adhesive should have been applied uniformly to the entire stud surface. Coating should have pulled off uniformly over the entire stud surface either with adhesive failure from the substrate (A) or cohesive failure in the coating (C). If failure is less than 90 % A or C (or CM), if the adhesive has failed at the stud, a retest, exercising particular care in the specimen and stud preparation is performed. Pull strength for the ten runs on each coating substrate combination are averaged and reported. The precision and bias are primarily dependent upon the accuracy of the force measurements, the alignment of the device, and the care exercised in stud and specimen preparation, and in the care in testing. A ten-laboratory
round robin on ten samples gave an average standard deviation of 29 % for reproducibility and 22 % for repeatability. A range of pull strengths of 2 orders of magnitude has been observed (Table 1) for diverse coating-plastic combinations.
ASTM D4541
This test method defines a class of portable adhesion testers for measuring the pull-off strength of coatings. The method covers a procedure and apparatus for evaluating pull-off strength by determining either the greatest perpendicular force (in tension) that a surface area can bear before a plug of material is detached or whether the surface remains intact at a prescribed force (pass/fail). Failure will occur along the weakest plane within the system comprised of the test fixture, adhesive, coating system, and substrate and will be exposed by the fracture surface. The method maximizes tensile stress as compared to the shear stress applied by other methods such as scratch or knife adhesion, and results are not comparable. It is recognized that the pull-off strength reflects both material and instrumental parameters and therefore provides a relative, not absolute, measure of adhesion. The pull adhesion testers defined are portable and capable of applying a concentric load and counter load to a single surface so that coatings in the field can be tested even though only one side is accessible. Measurements are limited by the strength of adhesion bonds between the loading fixture and the specimen surface or the cohesive strength of the substrate. The pull-off test is performed by securing a “loading fixture” (dolly or stud) normal (perpendicular) to the surface of the coating with an adhesive. After the adhesive is cured, the testing apparatus is attached to the loading fixture and aligned to apply tension normal to the test surface. The force applied to the loading fixture is then gradually increased (in less than 100 s) and monitored until either
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
610
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
TABLE 1—Representative pull strength for organic coatings on plastics Failure Mode
Pounds Per Square Incha
Polyester/ Polycarbonate
C
58
Lacquer
Polycarbonate
A
36
3
Lacquer
ABC
C
82
4
Lacquer
Polycarbonate
A
37
5
Lacquer
ABS
A
93
6
Lacquer
PVC
AC
103
7
Enamel
Polycarbonate
A
308
8
Urethane
ABC
A
639
9
Urethane
ABS
A
476
10
Urethane
Metal
S
666
11
Enamel
Metal
75% A
342
12
Lacquer
PPO/Nylon
C
226
13
Enamel
PPO/Nylon
C
226
14
Enamel
PPO/Nylon
A
242
Sample No.
Coating
Substrate
1
Lacquer
2
Fig. 9—Portable adhesion tester. From ASTM D4541.
Errors in measurement in this test result from alignment of the apparatus that is not normal to the surface, poor definition of the area stressed due to improper application of the adhesive, poorly defined glue lines and boundaries, holidays in the adhesive caused by voids or inclusions, improperly prepared surfaces, and sliding or twisting of the fixture during the initial adhesive cure. Also, scratched or scored samples may contain stress concentrations leading to premature fractures. Interlaboratory data have been obtained for four commercial test apparatus and are presented in Table 2.
To convert to metric, multiply by 6.89 kPa. NOTE: ABC = acrylonitule-butadiene-styrene, PVC = polyvinyl chloride, PPO = polyphenylent ether.
a
ISO 4624
a plug of coating material is detached or a specified value is reached. The nature of the failure is assessed as to the percent of adhesive or cohesive failure, and the actual interfaces and layers involved are identified. The pull-off strength is computed based on the maximum indicated load, the instrument calibration data, and the original surface area stressed (Fig. 9). This method is general and is applicable to any portable apparatus meeting the standard’s basic requirements for determining the pull-off strength of coatings. Five apparatuses are commonly recognized to meet the requirements of the standard. It is known that the rigidity of the substrate affects pull-off strength results and is not a controllable test variable in field measurements as defined by this standard. For example, steel substrates of less than 1/8 in. (3.2 mm) thickness show reduced pull-off strength results compared to 1/4-in. (6.4-mm) thick panels.
Similar pull-off test methodology has been approved through the International Standards Organization, and the parameters involved have been carefully studied. ISO 4624, “Pull-off Test for Adhesion,” was approved in 1978. ISO 4624 specifies two different tests assemblies (A and B of Fig. 10). Test Assembly A (Sandwich Method) consists of a substrate painted on one or both sides, with test cylinders (studs) with a specified diameter bonded coaxially to the coated test surface and on the reverse. Test Assembly B consists of a rigid substrate coated on one side, with one test cylinder (stud) applied to that side only. While the latter is the more practical, Test Assembly A on stress analysis shows a smoother stress distribution than B. Test Assembly B shows strong stress peaks at the coating layer periphery. The result is that Test Assembly B shows 20 to 60 % lower breaking strength results than the sandwich method for several organic coatings on metal [67].
TABLE 2—ASTM D 4541 interlaboratory data (psi)a Instrument Loading Fixture Diameter Paint Sample
Patti 13 mm
Elcometer 20 mm
Hate 19 mm (mean of 3 results)
Dyna 50 mm
A
1160
586
1185
201
B
1099
624
1157
185
C
1333
827
1245
190
D
1678
888
1686
297
a
To convert to metric, multiply by 6.89 kPa.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 47
Fig. 10—ISO 4624 test assemblies.
For Test Assembly B, breaking strength values have been shown to be a function of the diameter of test cylinders (studs) as well as their geometry. On the range of 10 to 28 mm diameters, a factor of 2 in breaking strength was seen with a peak at about 14-mm diameter. It was observed that tensile forces only affected the central part of the test area. Even test cylinders of the same diameter, but different shapes, gave different results, up to 45 % lower. It has been noted that while uniform tensile stress can only be approached, the length of the test cylinder (stud) should be no less than half its diameter. When this is done, similar results are achieved regardless of cylinder diameter [67]. Perhaps it is only a coincidence, but it is interesting to note in Table 2 for apparatus meeting ASTM D4541 that the loading fixture diameter and geometry differ significantly and at least partially account for the greater than factor of 4 variation in test results, with the apparatus having a 50-mm loading fixture diameter showing the lowest numerical results. Substrate thickness (flexibility) has a significant effect upon results. In a study of steel panels in the range of 2 to 30 mm thick, breaking strength increased to a thickness of 15 mm using Test Assembly B. A comparison of results with the sandwich method for substrates above 15-mm thickness, however, still showed breaking strength values about one third higher for Test Assembly A. These observations can be explained in terms of differences in stress distribution. Thinner substrates show increased stress intensity at the periphery, which is distinguishable up to 15-mm thickness [67]. ISO 4624 specifies that tensile stress shall not be increased at a rate greater than 1 MPa/s, so that failure occurs within 90 s of initial stress application. An investigation of the rate of tensile stress increase in the range 0.15 to 1.2 MPa/s yielded indistinguishable results [67]. Sample parameters are important. Coating thickness (and adhesive thickness) is a key factor. The energy stored in a 5-mil (125μm) coating, by virtue of its internal strain, increases as the coating thickness increases and at a particular thickness can be sufficient to overcome the work of adhesion at the interface (spontaneous peeling). The variation is largest at thicknesses below 125 μm (5 mils), which of course is the area of most practical application. It was also observed that at the lowest coating thickness for the coatings studied, cohesive failure occurred in the uppermost parts of the coating, leaving a very thin film of coating on the test cylinder (stud). With thicker coatings, the fracture propagated deeper into the coating. Through study of the locus of failure, one can also study the effects of environmental conditions (aging, solvents, moisture) and of coating resin cure [67].
Q
ADHESION
611
More recent studies have also looked at instrument and sample parameters. Rocke et al. studied the influence of measured pull-off strength by stud area, adhesive thickness, the thickness of the substrate, as well as coating composition [68]. A comprehensive Norwegian study involved some 1400 pull tests using three different instruments on a three-coat paint system and on thermal arc-sprayed ALMg5 on blast-cleaned steel, with the blastcleaned steel also assessed as a standard. The parameters investigated included the effect of panel thickness on pull strength (5, 8, and 12 mm), the effect of cure time of the paint system on pull-off strength (8, 15, 28, 37, and 76 days after application of the top coat), the effect of four different brands of cyanoacrylate adhesive on pull-off strength, the effect of different temperatures on adhesive cure or pull-off on pull-off strength (8.5°C, 23°C, and 70°C), the effect of stud diameter on pull-off strength, and the effect of cutting around the edge of the stud on pull-off strength. The data reported were the averages of 10 to 30 runs [69]. Adhesion on blast-cleaned steel using an epoxy adhesive showed a 24–58 % increase in mean pull-off strength in going from 5 to 12 mm thick panels. The increase in pull-off strength for the three-coat paint system increased 30–34 % from 8- to 76-day coating cure time. Four cyanoacrylate adhesives were cured for 1, 2, 5 at 23°C or 8.5°C at 50 % relative humidity on bare blast-cleaned steel. At 23°C (1h) pull-off strength ranged from 25.6 to 35.4 MPa, while at 5 h pull-off strength ranged from 29.1 to 38.2 MPa. Corresponding data at 8.5°C were 21.5 to 27.0 and 22.2 to 30.6, respectively. Testing of arc spray coated samples at different temperatures showed an increase in pull-off strength for one of two cyanoacrylate adhesives and not the other and also a difference in locus of failure was also noted. Some effect of stud size was seen, with higher mean pull-off strength for studs at 1.57 cm2 versus 3.14 cm2. Cutting around the stud showed a slight lowering of pull-off strength (under 10 %). A number of experimental factors affect test results. A 5-mm rigid steel panel is not “infinitely thick.” Cured coatings may still be undergoing changes over long time periods. Such changes can effect adhesion. Temperatures of cure coatings and adhesives may affect pull-off strength [69]. Clearly, stress distribution changes, as altered by test apparatus and sample parameters, can alter results dramatically. Pull-off strength (breaking strength) is therefore a relative number. Comparisons must be made carefully in concert with an examination of the locus of failure. Practically speaking, while absolute values are only approached, relative values and studies of the locus of failure are sufficient for most purposes.
Scrape Adhesion Testing
Used to a lesser extent than the preceding methods is ASTM D2197, Test Methods for Adhesion of Organic Coatings by Scrape Adhesion. This test method covers the determination of the adhesion of organic coatings such as paint, varnish, and lacquer when applied to smooth, flat (planar) panel surfaces. The materials under test are applied at uniform thickness to planar panels, usually sheet metal of uniform surface texture. After drying, the adhesion is determined by pushing the panels beneath a rounded stylus or loop loaded in
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
612
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
TORQUE TESTS
Fig. 11—Balanced-beam scrape adhesion tester.
increasing amounts until the coating is removed from the substrate surface. The method is most useful in providing relative ratings for a series of coated panels exhibiting significant differences in adhesion. The balanced-beam, scrape-adhesion tester (Fig. 11) consists of a balanced beam to which is secured a platform for supporting weights and a rod at an angle of 45° that holds the scraping loop. The rod is set so that the scraping loop contacts test surfaces directly below the weights. The loop is a 1/16-in. (1.6-mm)-diameter rod bent into a “U” shape with an outside radius of 0.128±0.002 in. (3.25±0.05 mm) hardened to Rockwell HRC 56 to 58 and chromium plated and polished. At least 1/2 in. (1.3 cm) at the end of a test panel is un-coated. The surface of the substrate must be hard enough so that it is undamaged by the scraping loop. The test procedure is as follows. A test panel on the sliding platform is placed so that it may be moved away from the operator and the uncoated portion is toward the main beam support. Weights are placed on the weight support using an initial amount estimated to be appropriate for the particular coating. The beam is lowered until the loop rests on the uncoated portion of the specimen and the full load is applied; then the sliding platform is slowly pushed away from the operator 1 to 2 s/in. for a distance of at least 3 in. (76 mm). If the coating is removed, the testing is continued using successively smaller loads (0.5-kg increments) until the coating is not removed. If the coating is not removed by the initial scrape, the testing is continued, using successively larger loads (0.5-kg increments) until the coating is removed or until the maximum load of 10 kg has been applied. A new area of the test surface is used each time a scrape is made. When the critical load has been approximately located, the test is repeated five times at each of the three loadings: above, below, and at the load determined in the first trial. For each applied load, the number of times the coating was removed or adhered is tabulated. The load where the scrape results change from mainly adhering to mainly removed, ignoring the first 1/2 in. (13 mm) of the scratch if the coating was removed, is the adhesion failure end point. Perhaps the most “practical” approach to paint adhesion is assessment by the use of a paint knife. A test method that provides guidance for such evaluations is ASTM D6677, Test Method for Evaluating Adhesion by Knife.
While the scope of ASTM D3359 provides for the testing of coatings on metal substrates only, it and other tests discussed previously have been used to assess the adhesion of coatings on wood. Little data have, however, been published with wood as the substrate. Several papers using a torque technique have been published [70–73]. While such methods have not been standardized, it was felt that they might be of interest to readers. In this shear technique, studs are glued on the painted samples using a solvent-free, dual-component epoxy glue. After the glue has dried, a slot is drilled in the paint film around the studs until the wood surface is reached. The studs are removed using a recording torque wrench. The failure load is noted and the locus of failure determined. The shear stress (adhesion value) is calculated using the equation T = (167I)/(Πd3) where I is the measured torque (Nm), and d is the diameter of the stud (in this case, 12.5 mm). A minimum of six replicates is used for the calculation of the mean adhesion value. Adhesion values are given in MNm−2. Painted samples have been tested dry and after water immersion [74,75].
UNIFORM DOUBLE-CANTILEVER BEAM TESTS
While not standardized, workers have found success with a uniform double-cantilever beam technique for the testing of the adhesion of coatings on wood [74,75].
Other ASTM Standards
Other ASTM standards that pertain to aspects of adhesion measurements of films and coatings include the following: ASTM B533 – Standard Test Method for Peel Strength of Metal Electroplated Plastics. ASTM B571 – Standard Practice for Qualitative Adhesion Testing of Metallic Coatings. (This method includes bend, burnishing, chisel-knife, draw, file, grind-saw, heat-quench impact, peel, push, and scribe-grind tests.) ASTM C633 – Standard Test Method for Adhesion or Cohesion Strength of Thermal Spray Coatings. ASTM D3730 – Guide for Testing High-Performance Interior Architectural Wall Coatings. (This guide includes a pull-strength test for adhesion assessment.) ASTM D4145 – Standard Test Method for Coating Flexibility of Prepainted Sheet. (This method includes a tape pull-off test.) ASTM D4146 – Standard Test Method for Formability of Zinc-Rich Primer/ Chromate Complex Coatings on Steel. (This method is a tape test after dome-shaped deformation.) ASTM D7234 – Standard Test Method for Pull-Off Adhesion Strength of Coatings on Concrete Using Portable Pull-Off Adhesion Testers.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 47
ASTM D7334 – Standard Practice for Surface Wettability of Coatings, Substrates and Pigments by Advancing Contact Angle Measurement. ASTM F692 – Standard Test Method for Measuring Adhesion Strength of Solderable Films to Substrates. ASTM F1044 – Standard Test Method for Shear Testing of Calcium Phosphate Coatings and Metallic Coatings. ASTM F1147 – Standard Test Method for Tension Testing of Calcium Phosphate and Metal Coatings. ASTM F1842 – Standard Test Method for Determining Ink or Coating Adhesion on Plastic Substrates for Membrane Switch Applications. (This test is based upon D3359.) Additional tests exist for electrodeposits, and bending, burnishing, and wrapping tests for coatings on wire. Tests for adhesives have also been applied to coatings as well. Use of reverse impact tests, lap and butt joint tests, and tensile shear tests have been reported [1].
CONCLUSIONS
Basic adhesion is the summation of multiple phenomena. Ideal adhesion is probably neither obtainable nor measurable experimentally. Practical techniques do, however, allow sufficient assessment of relative adhesion for most purposes, if used with care and knowledge. Workers need to understand both basic adhesion concepts and the factors affecting practical adhesion for systems of their interest if they are to make improvements in real world products.
References [1] Nelson, G. L., Gray, K. N., and Buckley, S. E., Modern Paint and Coatings, Vol. 75, No. 10, 1985, pp. 160–172. [2] Nelson, G. L., and Gray, K. N., “Coating Adhesion to Plastics,” Proceedings, Waterborne and Higher-Solids Coatings Symposium, Vol. 13, New Orleans, LA, 5–7 Feb. 1986, University of Southern Mississippi, Hattiesburg, MS, pp. 114–131. [3] Mittal, K. L., “Adhesion Measurement: Recent Progress, Unsolved Problems, and Prospects,” Adhesion Measurement of Thin Films, Thick Films, and Bulk Coatings, ASTM STP 640, K. L. Mittal, Ed., ASTM International, West Conshohocken, PA, 1978, pp. 7–8. [4] Mittal, K. L., Adhesion Aspects of Polymeric Coatings, Plenum Press, New York, 1983. [5] Wu, S., Polymer Interface and Adhesion, Marcel Dekker, Inc., New York, 1982. [6] Patrick, R. L., Treatise on Adhesion and Adhesives, Vol. 6, Marcel Dekker, Inc., New York, 1967–1989. [7] Basin, V. E., Prog. Org. Coat., Vol. 12, 1984, pp. 213–250. [8] Good, R. J., in Treatise on Adhesion and Adhesives, R. L. Patrick, Ed., Vol. 1, Marcel Dekker, New York, 1957, pp. 9–68. [9] Huntsberger, J. R., in Treatise on Adhesion and Adhesives, R. L. Patrick, Ed., Vol. 1, Marcel Dekker, New York, 1967, pp. 119–150. [10] Good, R. J., in Recent Advances in Adhesion, L. H. Lee, Ed., Gordon and Breach, New York, 1973, pp. 357–380. [11] Williams, M. L., in Recent Advances in Adhesion, L. H. Lee, Ed., Gordon and Breach, New York, 1973, pp. 381–422. [12] Andrews, E. H., Fracture in Polymers, Elsevier, New York, 1968.
Q
ADHESION
613
[13] Mostovoy, S., Ripling, E. J., and Bersch, C. F., J. Adhes., Vol. 3, 1971, pp. 125–145. [14] Hansen, R. H., and Schonhorn, H., J. Phys. Chem. Ref. Data Suppl., Vol. B4, 1966, p. 203. [15] Schonhorn, H., and Hansen, R. H., J. Phys. Chem. Ref. Data Suppl., Vol. 11, 1967, p. 1461. [16] Bickerman, J. J., Ind. Eng. Chem., Vol. 59, No. 9, 1967, p. 40. [17] Bickerman, J. J., in The Science of Adhesive Joints, 2nd ed., Academic Press, New York, 1978. [18] Schonhorn, H., and Hansen, R. H., J. Appl. Polym. Sci., Vol. 11, 1967, p. 1461. [19] Good, R. J., “Locus of Failure and Its Implications for Adhesion Measurements,” Adhesion Measurement of Thin Films, Thick Films, and Bulk Coatings, ASTM STP 640, K. L. Mittal, Ed., ASTM International, West Conshohocken, PA, Fig. 1, p. 20. [20] Wu, S., J. Adhes., Vol. 5, 1973, p. 39. [21] Sharpe, L. H., and Schonhorn, H., Adv. Chem. Ser., Vol. 43, 1964, p. 189. [22] Zisman, W. A., Adv. Chem. Ser., Vol. 43, 1964, p. 1. [23] Kitazaki, Y., and Hata, T., in Recent Advances in Adhesion, L. H. Lee, Ed., Gordon and Breach, New York, 1973, pp. 65–76. [24] Dahlquist, C. A., in Aspects of Adhesion, D. J. Alner, Ed., Vol. 5, CRC Press, Cleveland Ohio, 1969, pp. 183–201. [25] Barbarisi, M. J., Nature (London), Vol. 215, 1967, p. 383. [26] Boucher, E.A., Nature (London), Vol. 215, 1967, p. 1054. [27] Smarook, W. H., and Bonotto, S., Polym. Eng. Sci., Vol. 8, 1968, p. 41. [28] Levine, M., Illka, G., and Weiss, P., Polym. Lett., Vol. 2, 1964, p. 915. [29] Voyutskii, S. S., and Vakula, V. L., J. Appl. Polym. Sci., Vol. 7, 1963, p. 475. [30] Voyutskii, S. S., J. Adhes., Vol. 3, 1971, p. 69. [31] Anand, J. N., J. Adhes., Vol. 5, 1973, p. 265. [32] Voyutskii, S. S., and Deryagin, B. V., Kolloidn. Zh., Vol. 27, 1965, p. 724. [33] Sharpe, L. H., and Schonhorn, H., Kolloidn. Zh., Vol. 28, 1966, p. 766. [34] Helfand, E., and Tagami, Y., Protoplasma, Vol. 9, 1971, p. 741. [35] Helfand, E., and Sapse, A. M., J. Chem. Phys., Vol. 62, 1975, p. 1327. [36] Helfand, E., Acta Crystallogr., Vol. 8, 1975, p. 295. [37] Helfand, E., J. Chem. Phys., Vol. 63, 1975, p. 2192. [38] Roe, R. J., J. Chem. Phys., Vol. 62, 1975, p. 490. [39] Voyutskii, S. S., Yagnyatinskaya, L., Kaplunova, Y., and Garetovskya, N. L., Rubber Age, Vol. 2, 1973, p. 37. [40] Ahagon, A., and Gent, A. N., J. Polym. Sci., Polym. Phys. Ed., Vol. 13, 1975, p. 1285. [41] Yates, P. C., and Trebilcock, J. W., SPE Transactions, October 1961, p. 199. [42] Plueddemann, E. P., in Interfaces in Polymer Matrix Composites, E. P. Plueddemann, Ed., Academic Press, New York, 1974, pp. 174–216. [43] Monte, S. J., and Bruins, P. F., Modern Plastics, December 1964, p. 63. [44] Mao, T. J., and Reegen, S. L., in Adhesion and Cohesion, P. Weiss, Ed., Elsevier, Amsterdam, 1962, pp. 209–217. [45] Hofrichter, C. H., Jr., and McLaren, A. D., Ind. Eng. Chem., Vol. 40, 1948, p. 239. [46] McLaren, A. D., and Seiler, C. J., J. Plasma Phys., Vol. 4, 1949, p. 63. [47] Brown, H. P., and Anderson, J. F., in Handbook of’ Adhesives, I. Skeist, Ed., Van Nostrand-Reinhold, Princeton, NJ, 1962, pp. 255–257. [48] Packham, D. E., “The Adhesion of Polymer to Metals: The Role of Surface Topography,” in Adhesion Aspects of Coatings, K.L. Mittal, Ed., Plenum Press, New York, 1983, pp. 19–44. [49] Mittal, K. L., Polym. Eng. Sci., Vol. 17, No. 7, 1977, pp. 467–472. [50] Derjaguin, B. V., and Smilga, V. P., in Adhesion Fundamentals and Practice, Maclaren and Sons, London, 1969, p. 152.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
614
PAINT AND COATING TESTING MANUAL
Q
[51] Fowkes, F. M., J. Adhes. Sci. Technol., Vol. 1, No. 1, 1987, p. 7. [52] Massingill, J. L., J. Coat. Technol., Vol. 63, No. 797, 1991, pp. 47–54. [53] Allen, K. W., in Aspects of Adhesion, Vol. 5, D. J. Alner, Ed., University of London Press, 1969, p. 11. [54] Paul, S., J. Coat. Technol., Vol. 54, No. 692, 1982, pp. 59–65. [55] Croll, S. G., “Adhesion and Internal Strain in Polymeric Coatings,” in Adhesion Aspects of Coatings, K. L. Mittal, Ed., Plenum Press, New York, 1983, pp. 107–129. [56] Sato, K., Prog. Org. Coat., Vol. 8, 1980, pp. 143–160. [57] Lewis, A. F., and Forrestal, L. J., “Adhesion of Coatings,” in Treatise on Coatings, Vol. 2, Part I, R. R. Myers and J. S. Long, Eds., Marcel Dekker, New York, 1969, pp. 57–98. [58] Gardner, H. A., and Sward, G. G., Chap. 7, Paint Testing Manual, 12th ed., Gardner Laboratory, Bethesda, MD, 1962, pp. 159–170. [59] Corcoron, E. M., Adhesion, Chap. 5.3 in Paint Testing Manual, ASTM STP 500, 13th ed., G. G. Sward, Ed., ASTM International, West Conshohocken, PA, 1972, pp. 314–332. [60] Mittal, K. L., J. Adhes. Sci. Technol., Vol. 1, No. 3, 1987, pp. 247–259, bibliography. [61] Stoffer, J. O., and Gadodia, S. K., American Paint and Coatings Journal, Vol. 70, No. 50, 1991, pp. 36–40, and Vol. 70, No. 51, 1991, pp. 36–51. [62] Wu, S., Polymer Interface and Adhesion, Marcel Dekker, Inc., New York, 1982, p. 531. [63] Marion, E. Wolters, 3M memorandum, 1984.
15TH EDITION
[64] Product Bulletin, 3M. [65] Proceedings, Symposium on Adhesion Aspects of Polymeric Coatings, K. L. Mittal, Ed., The Electrochemical Society, 1981, pp. 569–582. [66] Jacobsson, R. J., Thin Solid Films, Vol. 34, 1976, pp. 191–199. [67] Stichfeld, J., “Pull-Off Test, An Internationally Standardized Method for Adhesion Testing—Assessment of Relevance of Test Results,” in Adhesion Aspects of Polymeric Coatings, K. L. Mittal, Ed., Plenum Press, New York, 1983, pp. 543–567. [68] Rocke, P., Dole, M., and Bouzziri, J., J. Adhes. Sci. Technol., Vol. 8, 1994, p. 587. [69] Scheie, J., and Nilsen, R., “Do the Adhesion Instruments Measure Accurately? What effect does the type of adhesive curing of the paint, etc., have on the pull strength?,” Report 18601, National Institute of Technology, Norway, 1996. [70] Holloway, M. W., and Walker, P. A., Mater. Struct., Vol. 47, 1964, pp. 812–823. [71] Ahola, P., Materials and Structures, Vol. 28, 1995, pp. 350–356. [72] Ahola, P., J. Oil Colour Chem. Assoc., Vol. 74, 1991, pp. 173– 176. [73] Bardage, S. L., and Bjurman, J., J. Coat. Technol., Vol. 70, No. 878, March 1998, pp. 39–47. [74] Knaebe, M. T., and Williams, R. S., J. Test. Eval., 21(4), 1993, pp. 272–275. [75] Knaebe, M. T., Williams, R. S., and Spence, J. W., J. Coat. Technol., Vol. 68, No. 856, May 1996, pp. 27–30.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
48
MNL17-EB/Jan. 2012
Abrasion Resistance* Daniel K. Slawson1 PREFACE
IN PREPARATION OF THIS CHAPTER, THE CONTENTS of the fourteenth edition were drawn upon. The author acknowledges the authors of the fourteenth edition. The current edition will review and update the topics as addressed by the previous authors, introduce new technology that has been developed, and include up-to-date references. The problem of resistance to abrasion exists all around us: on floors, walls, furniture, automobiles, farm implements, construction equipment, military gear, appliances, shoe soles, and so on. Because resistance to abrasion is a basic factor in the durability of a coating, its measurement is of practical importance to both producer and consumer. Abrasion is caused by mechanical actions such as rubbing, scraping, or erosion from wind and water. It can take two general forms: marring or wearing.
DEFINITIONS
Mar abrasion consists of permanent deformations that have not ruptured the surface of a coating. The resistance of a coating to marring is its ability to withstand scuffing actions that tend to disfigure or change the appearance of its surface. Some examples of potential causes of marring of organic coatings are (a) sliding an object across the surface of furniture; (b) rubbing of a belt buckle, button, zipper, or rough fabric on an automobile finish; (c) sliding a toy across a wall or a refrigerator door, etc. Wear abrasion is caused by mechanical action that removes material from the surface of a coating. In many cases, the removal is gradual or progressive due to repetitive mechanical action.
RELATIONSHIP TO OTHER PHYSICAL PROPERTIES
Abrasion resistance is not a unique or isolated property of a material but is rather related to other physical characteristics such as hardness, cohesive and tensile strength, elasticity, and toughness. Also, from the standpoint of retaining its protective or decorative function, the thickness of a coating can be an important factor.
Hardness, Elasticity, and Tensile Strength
Abrasion resistance is related to hardness, yet the relationship is not simple. A first thought might be that the harder a coating, the better would be the abrasion resistance; however, this is not always true. Steel is much harder than
rubber, for example, but steel “tires” on an automobile, in addition to not giving a smooth ride, would not last long on concrete roads in comparison to the service life of a good rubber tire. The ability of a material, such as a rubber tire, to undergo elastic deformation and recover or “to ride with the blow” is associated with good abrasion resistance. The energy transferred to an elastic material by an impacting object is largely returned to the object, though redirected, instead of being expended in the destruction (separation and removal of material) of the impacted surface. From a fundamental viewpoint, this is a consequence of the smaller deceleration and hence smaller force generated, since force is equal to the product of mass and acceleration, when the impact is with a material that will deform or “give” with the impacting object. If the deformation caused by the object is not elastic, the material will yield and flow, causing damage. Therefore, soft materials with a low tensile strength are not abrasion resistant. The fact that elastic materials are often abrasion resistant does not mean that hard materials are not abrasion resistant. Theoretically, however, if one is given two materials of equal tensile strength, the material of lower modulus should have the best abrasion resistance. The deceptive factor here is that a hard material usually has a much higher tensile strength than a soft material. Thus, when a rubber is compared with steel, materials with orders of magnitude difference in tensile strength are being considered. The fact that rubber is abrasion resistant emphasizes that the value of a low modulus of elasticity and adequate tensile strength are factors that play a role in obtaining good abrasion resistance. In theory, a very hard material that has a hardness and cohesive strength adequate to completely resist any impact force it might encounter would not be dependent on rubberlike elasticity to reduce or dissipate impact stresses. It would be more abrasion resistant than the best elastic but weaker material [1].
CORRELATION WITH END-USE PERFORMANCE
From the discussion above, it is apparent that the measurement of abrasion resistance involves measuring a complex combination of interrelated properties among which there is no direct relationship. The task of devising a test methodology that will correlate with end-use performance is, therefore, complex and difficult but not impossible to
This chapter is an abridged and modified version of the chapter with the same title, written by Mark P. Morse, found in the previous edition of this manual. 1 Taber Industries, 455 Bryant St., North Tonawanda, NY 14120 *
615 Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed byASTM International This standard is for EDUCATIONAL USE ONLY. Copyright © 2012 by www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
616
PAINT AND COATING TESTING MANUAL
Q
develop. If the test method subjects the material under test to a composite of destructive forces similar to those met in service, then the test method will correlate—or predict—the service performance of the material in at least a qualitative or relative ranking respect. When accelerated tests are being considered, even tests that rank materials in the same sequence as actual service tests, a quantitative correspondence with actual service is seldom expected or obtained. Actual end-use tests, while the most reliable in providing an indication of a material’s probable long-term durability, suffer from the difficulties of ensuring equivalent usage and measurement, especially when the comparison of different materials is attempted. Because of such difficulties and because service tests are usually very time consuming, a wide variety of test machines have been developed to provide an accelerated indication of the abrasion resistance of coatings and related materials such as laminate flooring, plastics, linoleum, and wall coverings. In a comprehensive review article by Harper [2], there is a list of no less than 49 different abrasion-causing machines. An investigation by the International Study Committee for Wear Tests of Flooring Performance [3], in which seven commercial, organic flooring materials were systematically tested on 21 abrasion machines of 17 different types, indicated that very few of these machines were capable of providing a reliable comparison of the abrasion resistance of widely different materials that could be correlated with end-use performance. In addition, the different machines did not correlate very well with each other. On the basis of a round robin conducted in 1956 by ASTM with six different clear floor coatings evaluated by six different abrasion test methods [4], only two of the methods were found to correlate with actual end-use performance and to have the reproducibility necessary for acceptance as ASTM standards. These were ASTM Test Method for Abrasion Resistance of Organic Coatings by Falling Abrasive (ASTM D968) and ASTM Test Method for Abrasion Resistance of Organic Coatings by Air Blast Abrasive (ASTM D658, withdrawn in 1996). The jet abrader (ASTM D658, withdrawn in 1996) offers greater speed and precision of measurement [1]. More recent comparative testing (see Comparison of Wear Abrasion Testers) indicates ASTM D658 is superior to ASTM D968 and that ASTM Test Method for Abrasion Resistance of Organic Coatings by the Taber® Abraser (ASTM D4060) is superior to both of these tests regarding precision and sensitivity in differentiating between coatings.
MECHANISM OF ABRASION
The success of the particle impingement types of abrasion testers in correlating with service performance is perhaps not surprising when one considers the abrading mechanism. Whether or not a particular type of abrasion test correlates with end-use performance depends not only on a similarity of abrading mechanisms and the abradants used in both cases, but also on the extent to which that mechanism is maintained during the course of the abrasion test. It is on the latter factor that many methods fail. Rubbing (friction) and scraping methods obviously wear away the test surface in a different manner than methods in which abrasive particles are contacted with the surface. One aspect of the mechanistic difference lies in the
15TH EDITION
angle of contact with the surface. Abrasive particles striking the surface of a coating at nearly normal incidence tend to compress, scar, and cut into the coating. As a result, minute portions of the coating eventually should be crosscut and displaced. On the other hand, the rubbing and scraping types of abrasion that take place at near grazing incidence would tend to undercut and to shear through very thin layers of the coating in successive, irregular slices that ultimately wear it away. Different devices might incorporate various degrees of these basic processes, depending on the angle and force of particle attack. Apart from the nature of the above mechanisms, it should be apparent that whatever the mechanism, it is not maintained uniformly in friction methods. Such methods suffer from changes in the abrading conditions as the testing proceeds, either because of heating of the specimen or clogging of the abradant or other. If an abrasion test using the direct impingement of particles is chosen so the undesirable frictional effects are avoided, there may be doubt about whether the method correlates well with a type of service in which the coating is, for example, walked upon. Yet, the previously mentioned ASTM round robin of abrasion tests carried out with floor coatings clearly established the validity of particle-impingement-type tests for evaluating this type of end-use service life. Until the mechanism of abrasion is more completely studied and understood, industry must continue to rely to a large extent on experimentation and even intuition when devising abrasion test methods. In addition, properly designed test programs must be employed to determine the extent to which a given method correlates with actual enduse performance [1]. A side variety of useful abrasion testing procedures are described in sections that follow.
ABRADANTS
The most sophisticated and precise abrasion testing instruments are only as good as the contact point between the test specimen and the abradant used to wear the specimen. A consistent abrasive material to evenly wear the specimen is key to accurate abrasion testing. The abradant must be consistent from lot to lot. The Taber Abraser uses a variety of resin (Calibrase®) and vitrified (Calibrade®) abrasion wheels containing a wide range of silicon carbide and aluminum oxide grains. The scrub abrasion testers use black nylon or black hog bristle brushes, sponges, abrasive paper, and stainless steel/nylon balls. The Linear Abraser uses abrasive plugs (Wearasers®) containing the same abrasive material as the Taber Abraser wheels. The Falling Abrasive Test uses Ottawa silica sand and silicon carbide. The oscillating sand tester uses silica or alumina zirconia abrasive.
Falling Abrasive Test
This widely used abrasion test method has both ASTM and federal counterparts. Originally developed at Gardner Laboratory the method has been studied and further developed by others [5]. ASTM D968 employs an apparatus, Fig. 1, that is simple and inexpensive compared with other more complicated instruments, and the test results correlate reasonably well with various types of service [4]. However, the method is laborious and time consuming since large
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 48
Q
ABRASION RESISTANCE
617
D968, which specifies that a natural silica sand graded to a particular sieve size and known as Ottawa sand is the standard sand abrasive and a silicon carbide of particular sieve size is the other standard abrasive. Abrasion resistance is expressed in terms of the volume of abrasive required to wear through a unit thickness of the coating with the abrasive falling from a specified height through a guide tube. The substrate is supported at a 45° angle to the vertical.
ABRASIVE BLAST METHODS
Air Jet Erosion Tester and Micro-Abrader
These testers shown in Fig. 2 are used to test the erosion resistance of solid materials to a stream of gas containing abrasive particulate. Micro-sized abrasive particles are mixed with pressurized air to generate an abrasive flow. The abrasive flow is propelled and directed through a small nozzle toward the test sample. Material loss, in this case, is achieved via the impingement of small abrasive particles upon the surface of the test sample. Materials such as metals, ceramics, minerals, polymers, composites, abrasives, and coatings can be tested with this instrument. Depending on the test instrument used; the test specimen, temperature, angle of incidence of the jet stream, abrasive particulate speed and flux density, can be varied to best simulate actual conditions.
Fig. 1—Falling sand abraser (courtesy of Paul N. Gardner Co.).
quantities of abrasive must be handled due to the slow rate of abrasion, particularly when the material under test is abrasion resistant. Over the years a variety of abrasives have been used in the basic apparatus. These abrasives include sand, emery, and various grades of silicon carbide. Sand, although having the disadvantage of a slow abrasion rate, is readily available and has given reproducible results at low cost. Therefore, it has been the abrasive of choice in the standard methods. Only sand and silicon carbide are used in ASTM
Fig. 2—Micro-abrader (courtesy of Taber Industries).
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
618
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 3—Interior of a gravelometer (courtesy of the Q-Panel Company).
Gravel Projecting Machine
This machine, which is commonly termed a gravelometer, is a device designed to evaluate the resistance of automotive and railway finishes to abrasion by flying gravel and road ballast. ASTM Test Method for Chip Resistance of Coatings (ASTM D3170) has a description of such a device and provides a procedure for its use. Fig. 3 illustrates the interior components of a device that meets specification given in ASTM D3170. A test chamber is provided in which a coated panel is supported vertically and blasted with a stream of particular gravel. One pint of this gravel is introduced into an air stream having a pressure of 70±2 psi (4.92±0.14 kg/cm2) over a 10 s period. Tests are usually conducted in a cold room to simulate winter-like driving conditions. At the completion of the test, the coated panel is visually rated for degree of chipping by using a photographic reference standard provided with D3170. SAE Method J400 describes the test procedure used for evaluating the chip resistance of automotive coatings.
METHODS USING ROTATING DISKS Schiefer Abrasion Testing Machine
(This machine is available from Frazier Precision Instrument Company, Inc., 925 Sweeney Dr., Hagerstown, MD 21740. www.frazierinstrument.com) The machine was designed for measuring wear resistance of textiles such as rugs and fabrics [6–9] but it offers a potentially useful means for evaluating the wear abrasion of organic coatings. It has two unique features: (a) a uniform abrasion pattern and (b) interchangeable steel and Carboloy abradant disks with crosscut and rod patterns that reduce heating and clogging (Fig. 4). To conduct a wear abrasion test, the specimen is mounted on a motor-driven rotating plate where it is abraded uniformly in all directions by the motion of the offset rotating abradant mounted directly above it. Both specimen and abradant rotate in the same direction with approximately the same angular velocity, 250 rpm, each about its own axis. The axes are spaced 1 in. (2.54 cm) apart and are parallel. The abradant is loaded with a standard weight.
Fig. 4—Schiefer abrasion testing machine (courtesy of Frazier Precision Instrument Company).
METHODS USING ROTATING ABRASIVE WHEELS Taber® Abraser
(This instrument is available from Taber Industries, 455 Bryant Street, North Tonawanda, New York 14120. www. taber-industries.com) This apparatus, Fig. 5, is widely used for evaluating the wear abrasion resistance of organic coatings. A procedure for its use is given in ASTM Test Method for Abrasion Resistance of Organic Coatings by the Taber Abraser (ASTM D4060), in ISO/DIN Methods 3494 and 4584. ASTM D4060 utilizes a specimen in the form of a 4-in. (10.2-cm)-diameter disk or a 4-in. (10.2-cm)-square mounted on a turntable that is rotated at a fixed speed under a weighted abrading wheel. A wide variety of resin and vitrified abrasion wheels containing a wide range of aluminium oxide and silicon carbide grits create different types of abrasion. An important feature of the Abraser is the wheels traverse a complete circle on the specimen surface. This reveals abrasion resistance at all angles relative to the grain of the specimen. A vacuum pickup is used to remove
Fig. 5—Taber Abraser (courtesy of Taber Industries).
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 48
any loose particles generated during actual tests. Wear abrasion resistance is expressed in terms of (a) Wear Index, which is the weight loss per specified number of revolutions (usually 1000) under a specified load (500 or 1000 g), and/ or (b) wear cycles per mil, which is the number of cycles required to wear through a 1-mil thickness of coating and is reported as the number of revolutions per mil. ASTM D6037 Standard Test Methods for Dry Abrasion Mar Resistance of High Gloss Coatings, utilizes two test methods to assess mar resistance by measuring the gloss of abraded and unabraded areas. Mar resistance is directly related to the coating’s ability to retain gloss in abraded areas. Test Method A uses a device that contains an abrasive wheel (Taber Calibrase® CS-10 Wheel unless otherwise specified or agreed). Test Method B uses a device that contains a wheel that has been fitted with abrasive paper. ASTM D1044 Standard Test Method for Resistance of Transparent Plastics to Surface Abrasion describes a procedure for estimating the resistance of transparent plastics to one kind of surface abrasion by measuring the change in optical properties. This test uses Calibrase®, CS-10F Wheels resurfaced on a ST-11 Refacing Stone to create the wear. The abrasive damage is visually judged and numerically quantified by the difference in haze percentage in accordance with Test Method D1003 between abraded and unabraded specimen.
Fig. 6—Taber Grit Feeder Attachment (courtesy of Taber Industries).
Q
ABRASION RESISTANCE
619
Taber® Grit Feeder Attachment
When used with the Taber Abraser, the Grit Feeder, seen in Fig. 6, provides a unique method to evaluate three-body abrasion resistance on a variety of materials. This instrument was originally designed to simulate abrasion caused from grit embedded in leather soled shoes and is typically used to test flooring. Aluminum oxide grit particles are evenly distributed onto the specimen wear path and pass under a pair of leather wheels. This loose grit acts as an abradant and aids in the rolling action that contributes to the physical breakdown of materials. The Abraser Vacuum is attached to the grit feeder, and continuously removes both abraded material and used grit. In an ASTM round robin, Taber Abraser tests with a series of organic coatings produced abrasion resistance values that exhibited a within-laboratory coefficient of variation of 10 % and an interlaboratory coefficient of variation of 30 %. Another study of the reliability of test results was conducted by Hill and Nick in 1966 [10].
TESTS BASED ON LINEAR MOTION
Straight-Line Reciprocating Machines (Scrub Abrasion Testers)
These machines, as shown in Fig. 7, pull a sled or boat back and forth over the surface of a coated panel. The sled surface can be a brush, a sponge, rubber, or sandpaper. A sled travel of at least 10 in. (25.4 cm) is provided by the machines, and they can provide reciprocating cycles ranging from 35 to 60 per minute. Both dry and wet surface tests may be performed with these machines. For dry surface tests, wear abrasion resistance is reported as: (a) number of cycles to reach a certain visual end point, (b) degree of abrasion observed after a specified number of cycles, and (c) number of cycles required to abrade through the coating to the substrate. Procedures for conducting wet adhesion (scrub resistance) tests on interior paints are described in ASTM Test Method for Wet Abrasion Resistance of Interior Paints (ASTM D4213), and ASTM D2486 Standard Test Methods for Scrub Resistance of Wall Paints. Other methods include DIN 52778 and ISO 11998. In these test procedures, coating films are applied to a plastic substrate and allowed to dry. A sponge surface is mounted on the sled. Both the sled and the coating surface are wet with a soap solution of specified composition. Wet abrasion resistance is reported as: (a) computed rate of
Fig. 7—Scrub Abrasion Tester (courtesy of Byk-Gardner).
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
620
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 10—Taber Reciprocating Abraser (courtesy of Taber Industries).
Taber® Large Linear Abraser Fig. 8—Taber Linear Abraser (courtesy of Taber Industries).
erosion of the wet coating film and (b) number of cycles required to wear through the coating film to the substrate or to produce breaks in the film.
Taber® Linear Abraser
The Taber® Linear Abraser, shown in Fig. 8, utilizes the same abrasive material contained in the Taber Abraser wheels in the form of an abrasive plug (Wearaser®). This instrument is designed to test virtually any size or shape specimen in a wet or dry environment. This instrument can also be modified to evaluate scratch resistance and color transfer. A precision bearing on the spline shaft creates a “free-floating head.” As the arm strokes, the abradant follows the contour of the specimen, curved or flat. The user can customize the abradant, stroke length, speed, and weight load. A procedure for its use is given in ASTM Test Method D6279 Test Method for Rub Abrasion Mar Resistance of High Gloss Coatings.
The Large Linear Abraser, shown in Fig. 9, combines a longer test stroke and heavier weight load, than the Taber® Linear Abraser, with a scouring pad to simulate the wear bakeware and cookware receive in extensive home use as it is cleaned.
Taber® Reciprocating Abraser
The Reciprocating Abraser, shown in Fig. 10, referenced in “ISO 1518 Paints and Varnishes—Scratch Test,” is used for evaluating resistance to abrasion, scratch, and mar on most flat or slightly contoured objects. A sliding specimen platform moves in a horizontal, reciprocating motion under a stationary tool, generating a linear wear, scratch/mar path.
RCA Abrasion Wear Tester
(This device, is available from Norman Tool and Stamping Co., 15415 Old State Rd., Evansville, IN 47711. www. normantool.com). It is being used for evaluating the abrasion resistance of appliance finishes and of coatings on television set controls. The machine abrades a 2 in. by 2 in. (5.1 cm by 5.1 cm) coated panel surface by passing computer or polyester tape over the surface. A fresh tape surface is presented for each cycle of abrasion. Weights ranging from 55 to 275 g are applied to the tape depending on the nature of the coating under test. Abrasion resistance is reported either as the number of cycles required to produce the onset of visible scuffing or as the number of cycles required to wear through to the substrate.
Oscillating Sand Tester
Fig. 9—Taber Large Linear Abraser (courtesy of Taber Industries).
This testing device, seen in Fig. 11, is used to determine the resistance of transparent plastics and transparent coatings utilized in windows or viewing ports to surface abrasion. The test specimen is mounted in a tray (so that it forms part of the bottom of the tray) and covered with abrasive media to a depth of 12.7 mm. The tray reciprocates in a linear motion at a speed of 300 strokes per minute over a distance of 100 mm. A typical test utilizes a specimen that is 100 nm by 100 mm and is tested for 600 strokes. The results of the test consist of measuring and recording the change in haze
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 48
Q
ABRASION RESISTANCE
621
Fig. 12—Balanced Beam Tester (courtesy of Byk-Gardner). Fig. 11—Taber Oscillating Sand Tester (courtesy of Taber Industries).
and light transmission of the test specimen. This tester is referenced in ASTM F735, Standard Test Method for Abrasion Resistance of Transparent Plastics and Coatings Using the Oscillating Sand Method.
COMPARISON OF ABRASION TESTERS
ASTM Subcommittee D-01.23 conducted a round robin to determine the comparative precision, sensitivity, and correlation of four wear abrasion testing procedures. The procedures were (a) air blast silicon carbide (ASTM D658— discontinued in 1996), (b) falling sand (ASTM D968), (c) falling silicon carbide (ASTM D968), and (d) Taber Abraser (ASTM D4060). The wear abrasion resistance of four coatings with significantly different apparent resistances to abrasion were tested with each device. The results obtained from these tests are recorded in ASTM Report RR D01-1037, which is available at ASTM Headquarters. From the results, it was concluded that the air blast silicon carbide test and the cycles per mil Taber Abraser test had better sensitivity in differentiating coating abrasion resistances than the other test procedures. The falling sand test, air blast silicon carbide test, and Taber Abraser cycles per mil test ranked the coatings in the same order as the expected performance. The falling silicon carbide test reversed the ranking of two of the coatings. The precision exhibited by the four test procedures are as given in Table 1.
TESTS FOR MAR ABRASION RESISTANCE Balanced Beam Tester
A useful test for determination of mar resistance involves the use of a balanced beam scrape tester, seen in Fig. 12. ASTM D5178 was developed to determine the mar resistance on smooth, flat surfaces. Results are expressed in terms of force-to-mar films of organic coatings such as paint, varnish, and lacquer when applied to smooth, flat planar panel surfaces. This test is carried out by placing a coated test panel on a movable platform and a weight on the beam. The stylus is lowered gently onto the coated surface and then the platform is pushed against the stylus at a rate of 1/4 in. (6 mm) per second for a distance of at least 3 in. (75 mm). At the end of each stroke, the stylus is raised off the coated panel and the panel is moved slightly to the side to provide a new test surface. The coating is examined for marring. If none is visually observed in the initial scrape, successively greater weights are added to the beam until marring is apparent. If marring is produced in the initial scrape, testing is continued using successively lighter weights until the coating is not marred. The weight required to just produce visible marring is taken as the mar resistance value. Details about the balanced beam tester and its operation can also be found in ASTM D2197, Test Methods for Adhesion of Organic Coatings by Scrape Adhesion, ASTM D2248, Practice for Detergent Resistance of Organic Finishes.
Taber® Shear/Scratch Tester
TABLE 1—Comparison of abrasion resistance test method precision Coefficients of Variation Within Laboratory
Between Laboratories
Taber Abraser (ASTM D4060)
4%
16 %
Air Blast Silicon Carbide (ASTM D658)
7%
10 %
Falling Sand (ASTM D968)
9%
35 %
Falling Silicon Carbide (ASTM D968)
11 %
45 %
The Taber Shear/Scratch Tester, seen in Fig. 13, is used to evaluate the scratch and mar resistance of organic coatings, slate, high pressure decorative laminates, and polymeric substrates. A specimen rotates at 0.5 or 5 rpm and the scratching/marring occurs as the specimen comes in contact with either the diamond or carbide tools. An adjustable load is applied to the scratch tool. Mar resistance is reported as the amount of weight required to just produce a barely visible scuffing or loss of gloss. This instrument is referenced in ASTM C217, Weather Resistance of Slate, EN 438-2 Decorative High Pressure Laminates— Resistance to Scratching, ISO 4586-2 High Pressure Decorative Laminates—Resistance to Scratching.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
622
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
observed. The procedure can be used as a “pass/fail” test or for comparing the mar resistance of coatings on a relative basis.
Paperclip Mar Test
This procedure is used to determine the ability of a surface to resist scuffing or marring. Scuffing is defined as permanent damage to the coating. An example of scuffing would be a break in the top-coat with visible flakes of coating caused by scraping by the paperclip. A paperclip remains in contact with the test specimen as the specimen is moved in four directions that are parallel to each of the edges of the specimen. The appearance is evaluated for evidence of scuffing or marring. Fig. 13—Taber Shear/Scratch Tester (courtesy of Taber Industries).
Multi-Finger Scratch Tester (5-Finger Scratch Tester)
This testing device, made popular by the automotive industry, can be used for determining the scratch and mar resistance of molded-in-color automotive plastics by use of a linearly oriented, multi-fingered scratching device. Operation involves 1 or 7 mm, individually loaded, hemispherical tips scratching the surface at a constant speed of 100 mm per second. This testing device is seen in Fig. 14.
Coin Mar Test
This test consists of dragging the edge of a coin across the surface of a coated panel and visually determining the degree of marring produced. The procedure can be used as a “pass/fail” test or for comparing the mar resistance of coatings on a relative basis. Results will vary from laboratory to laboratory due to the particular coin used and the pressure used for contact.
Fingernail Test
In this test, the back of a fingernail is dragged across the surface of a coating and the degree of marring is visually
Nanoscratching
ASTM D7187 Standard Test Method for Measuring Mechanical Aspects of Scratch/Mar Behavior of Paint Coatings by Nanoscratching, covers the nanoscratch method for determining the resistance of paint coatings on smooth flat surfaces to scratch/mar. The loss of appearance is mainly due to surface damages created. This method uses a two step process to quantitatively and objectively measure scratch/ mar behavior. Step one is to find the relationship between damage shape and size. Step two is to relate damage shape and size to visual loss of luster.
RAIN OR WATER EROSION
Rain erosion resistance of aircraft coatings has been studied in two types of testing machines [11]—whirling arm and jet. In the whirling arm tester, specimens having an airfoil contour are fastened to the leading edges of the two arms of a propeller-like blade that is rotated at angular velocities equivalent to flying speeds of up to 700 mph. Simultaneously drops of water fall on the whirling arms. Both the simulated air speed and the quantity of “rainfall” can be varied and closely controlled. In the jet-type tester, a highpressure jet of water impinges on the specimen. A rotating slotted disk in the path of the water jet breaks the stream of water into individual drops to simulate rainfall. In both of the above methods, having the water in the form of individual drops is a basic principle of the test. This is because the erosion that occurs on the leading edges of the airfoils on high-speed aircraft flying through rain is often the result of cavitation produced by the high-energy impacts of a multitude of individual droplets. At very high velocities, the water droplets, in effect, behave as tiny solid projectiles.
MISCELLANEOUS METHODS PEI Abrasion Tester
Fig. 14—Taber Multi-Finger Scratch Tester (courtesy of Taber Industries).
Designed and developed by the Porcelain Enamel Institute to measure the abrasion resistance of porcelain enamels, this tester has also proven to be useful in determining the wet abrasion resistance of organic coatings. It is specified to be used in ASTM C448, Test Methods for Abrasion Resistance of Porcelain Enamels. The tester, which is pictured in Fig. 15, has a gyrating table with positions for nine specimens. Abrasive-retaining rings and lids are clamped over the specimen to form individual test chambers. Each chamber is charged with glass or stainless steel spheres, and a slurry of abrasive particles is added through a
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 48
Q
ABRASION RESISTANCE
623
abrasion test results can be meaningless. The most common methods for determining endpoint are as follows; weight loss of the specimen after the abrasion test, visual loss by inspecting the specimen either with the naked eye or digitally. Those less common methods include gloss loss, haze loss, and change in coating thickness.
Calibration
An important aspect of today’s quality systems is the traceability of testing instruments. Verification that the instruments, abradants, and scratch tools are within specification is the core to ensuring repeatable and reproducible test results. Precision testing instruments should be calibrated regularly to ensure accurate results.
References
Fig. 15—PEI abrasion tester. (courtesy of National Bureau of Standards).
filling aperture located in the lid. The table gyrates at 300 cycles per minute to cause abrasion of the specimens. The amount of abrasion is evaluated by measuring the loss of gloss or loss of weight in the specimens after a specified time period [12].
Sand on Wheel Tester
This test instrument is used for determining the resistance of metallic materials to abrasion by means of the dry sand/ rubber wheel apparatus. The test specimen is pressed against a rotating rubber-rimmed wheel [13]. The test specimen is exposed to three-body scratch abrasion as abrasive grit is introduced between the specimen and rubber-rimmed wheel under a specified load. The intent of this method is to produce data that will reproducibly rank materials in their resistance to abrasion under specified conditions. Test results are reported as volume loss in cubic millimeters for the particular test procedure specified. Materials of higher abrasion resistance will have a lower volume loss. ASTM G65 Standard Test Method for Measuring Abrasion Using the Dry Sand/Rubber Wheel Apparatus.
END POINT MEASUREMENT
[1] Roberts, A. G., Chap. 5.2, “Abrasion Resistance,” Paint Testing Manual, ASTM STP 500, ASTM International, West Conshohocken, PA, 1972. [2] Harper, F. C., “The Abrasion Resistance of Flooring Materials: A Review of Methods of Testing,” Wear, Vol. 4, 1961, pp. 461–478. [3] International Study Committee for Wear Tests of Flooring Performance, “Performance of Abrasion Machines for Flooring Materials,” Wear, Vol. 4, 1961, pp. 479–494. [4] Interim Report, “Abrasion Resistance of Floor Coatings,” Group 20 Subcommittee IX on Varnish, ASTM Committee D-1, 1956. [5] Hipkins, C. C., and Phair, R. J., “The Falling Sand Abrasion Tester,” ASTM Bull. No. 143, ASTM International, West Conshohocken, PA, December 1946, pp. 18–22. [6] Schiefer, H. F., J. Res. Natl. Bur. Stand., Vol. 39, RR 1807, 1949, p. 1. [7] Schiefer, H. F., Crean, L. E., and Krasny, F. F., “Improved Single-Unit Schiefer Abrasion Testing Machine,” J. Res. Natl. Bur. Stand., Vol. 42, RR 1988, 1949, pp. 481–497. [8] Schiefer, H. F., and Werntz, C. W., Text. Res. J., Vol. 22, 1952, pp. 1–12. [9] ASTM D4158, 2008, “Test Method for Abrasion Resistance of Textile Fabrics (Uniform Abrasion),” Annual Book of ASTM Standards, Vol. 07.01, ASTM International, West Conshohocken, PA. [10] Hill, H. E., and Nick, D. P., “Study of the Reliability of Taber Abrasion Results,” J. Paint Technol., Vol. 38, No. 494, 1966, pp. 123–130. [11] Grace, J. K., and Frey, G. C., “Laboratory Testing of the RainErosion Resistance of Aircraft Finishes,” ASTM Bull. No. 168, ASTM International, West Conshohocken, PA, September 1950, pp. 56–61. [12] Outbridge, R., Proceedings of Rubber Technology Conference, 4th, London, 1962, Preprint No. 21. [13] Marks, M. E., and Conrad, P., “Resistance of Plastics to Abrasive Particles,” Mod. Plast., Vol. 23, 1946, pp. 165–168.
Any abrasion test is useless unless an accurate and repeatable endpoint can be reached. Without a precise end, the
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
49
MNL17-EB/Jan. 2012
Dynamic Mechanical and Tensile Properties Loren W. Hill1
DYNAMIC MECHANICAL AND TENSILE PROPERTIES are determined in all branches of materials science. There is a large body of published structure/property information that can be integrated with coatings research and development. By using structure/property information, coatings chemists can design and optimize chemical structures of the binder components of coatings. Purposeful and enlightened formulation with well-designed components makes it possible to obtain desirable coatings performance in many cases. Determination of dynamic mechanical and tensile properties requires the use of free films. This requirement is a serious limitation because many, if not most, of the performance properties of coatings are influenced by coatingsubstrate interactions. Therefore, tests of coatings intact on their end-use substrates must be thoughtfully coupled with free film determinations. The practical utility of basic methods described in this section is greatly enhanced when results are interpreted in relation to results of adhesion, abrasion, hardness, flexibility, toughness, and internal stress tests as described elsewhere in the manual. Dynamic mechanical analysis (DMA) and stress-strain analysis (SSA) of tensile properties are complementary methods in several ways. DMA involves very small strains, whereas SSA involves the maximum strain that the sample can withstand. Since the small strains used in DMA usually do not exceed the tensile strength or yield strength of the sample, the method is nondestructive. This feature facilitates property determination over a wide temperature range on a single sample, that is, DMA is often used as a temperature-scanning method. In contrast, SSA data are usually obtained at a single temperature, preferably on an instrument located in a controlled temperature and humidity room. Since the sample is broken in each test, it is very arduous to carry out SSA over a wide temperature range, and SSA is not amenable to temperature scanning.
DEFINITIONS
Tensile Versus Shear Tests
Two types of deformation of a block-shaped sample are depicted in Fig. 1. These deformations are used frequently in property determinations because they can be carried out reproducibly and treated by simple mathematics. In a tension test (Fig. 1 (A)), the sample is pulled apart with straight line application of force, called “uniaxial extension.” In a shear test (Fig. 1 (B)), one face of the cube is held stationary and the sample is pushed sideways by application of force at the opposite face. Note that different symbols (ε epsilon
1
or γ) are used for strain in these two tests. A single symbol is used for stress (σ sigma), but subscripts indicate the type of test. In Fig. 1, F is force and A, B, and C are initial sample dimensions. The product A × B is the initial cross-sectional area. It is evident in Fig. 1 that strain is defined quite differently in tension and shear tests. In a tension test, strain is the fractional increase in sample length. In a shear test, strain is the distance moved by the movable face divided by sample thickness, i.e., the distance between the stationary and movable faces. In both tests, stress is force divided by cross-sectional area, and modulus is stress divided by strain. Since strain is unitless, stress and modulus have the same units (force/area). It is evident in Fig. 1 (A) that crosssectional area will decrease as ∆C increases. If the initial cross-sectional area is used to calculate σt, the resulting E is called “engineering” modulus. If the change in crosssectional area is incorporated in the calculation, the resulting E is called “true” modulus. The relationship between tensile modulus and shear modulus is E = 2(1 + )G
where μ is Poisson’s ratio [1,2]. For materials that do not undergo change in volume with strain, μ = 0.5, and Eq (1) becomes E = 3G. Experimentally, μ is very close to 0.5 for rubbery materials and slightly less than 0.5 for many thermoplastic polymers [1,2].
Definitions of Dynamic Properties
In dynamic testing, an oscillating strain is applied, and the resulting oscillating stress is measured, or conversely an oscillating stress is applied, and the resulting oscillating strain is measured. Definitions and mathematical treatments do not depend on which of these modes of operation is used. Definitions and symbols used here correspond to ASTM Standard D4092, Terminology Relating to Dynamic Mechanical Measurements in Plastics. Relationships between strain, stress, and time are sketched in Fig. 2 for tensile DMA with application of strain and measurement of stress. The maximum applied strain is ε0. The maximum resulting stress is σt,0. Oscillation is depicted as a sine wave, but whether or not the driver of the instrument in use actually delivers a sine wave oscillation may depend on the particular instrument. The sample is held under sufficient tension so that it remains taut (not slack) even when the oscillating strain is at a minimum.
9 Bellows Rd, Wilbraham, MA 01095.
624 Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 This standard is for EDUCATIONAL USE ONLY. www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
Downloaded/printed byASTM International Copyright © 2012 by
(1)
CHAPTER 49
Q
DYNAMIC MECHANICAL AND TENSILE PROPERTIES
625
Fig. 1—Deformations of test samples: A. tensile, B. shear. The two types of deformation have different definitions and different symbols for strain, stress, and modulus.
The sine waves for strain and stress have the same frequency, but for viscoelastic samples the waves are out of phase by an amount, δ, called the phase lag. Theoretically
and experimentally, δ is zero for an ideal (Hookean) elastic solid. If an ideal (Newtonian) liquid could be tested in this way, δ would be 90°. For viscoelastic materials, δ lies between 0 and 90°, and the value of δ is a rather direct indication of viscoelastic character [1–4]. Definitions of dynamic properties depend on the concept of resolving the stress wave of Fig. 2 into two waves, one that is in phase with strain and one that is 90° out of phase with strain. The in-phase resolved plot represents elastic response, and the 90° out-of-phase resolved plot represents viscous response. In terms of modulus, the separated responses result in the following definitions Tensile Storage Modulus = E′ =
Tensile Loss Modulus = E′′ = Fig. 2—Applied oscillating strain (ε) and resulting oscillating stress (σt) in a dynamic mechanical analysis experiment with tensile deformation. The phase lag (δ) and maximum values of strain (ε0) and stress (σt,0) are indicated.
Loss Tangent =
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
t ,0 cos ε0
t ,0 sin ε0
E′′ tan E′
(2)
(3)
(4)
626
PAINT AND COATING TESTING MANUAL
Q
The term “storage” is associated with the elastic part of the response E because mechanical energy input to elastic materials is “stored” in the sense of being completely recoverable. The term “loss” is associated with the viscous part of the response E″ because mechanical energy input to ideal liquids is totally lost through viscous heating. The ratio E″/E′ is viscous response expressed relative to elastic response. This ratio reduces to sin δ/cos δ, which is tan δ. Thus, the name “loss tangent” is appropriate. A check of limiting values of δ is consistent with assignment of elastic and viscous responses in Eqs (2) and (3). When δ = 0°, cos δ = 1.0 and sin δ = 0. By Eq (3), E″ is zero and all of the response is elastic, i.e., E′ from Eq (2). When δ = 90°, cos δ =0 and sin δ = 1.0. Now E′ is zero by Eq (2) and all of the response is viscous, i.e., E″ from Eq (3). DMA relationships in Eqs (2) and (3) relate directly to characterization of samples that are solid or semi-solid. Other objectives of DMA include following viscosity and elasticity changes as the cure of thermoset coatings takes place and determining the melt-flow properties of powder coatings before the onset of cross-linking. When viscous response is the main interest, DMA is often carried out in shear. The treatment for shear DMA is identical, but with selection of appropriate shear symbols from Fig. 1, the definitions are: Shear Storage Modulus = G′ =
Shear Loss Modulus = G′′ =
Loss Tangent =
s,0 cos
0
s,0 sin
0
G′′ = tan G′
(5)
(6)
(7)
Once the experimenters have values of σs,0, γ0, and δ in hand, it is their choice whether to express results in terms of modulus values, Eqs (5) and (6), or in terms of “dynamic viscosity.” For dynamic viscosity, the frequency of oscillation, ω in radians per second, is required, but this frequency is usually known. The frequency is required because viscosity is (shear stress)/(rate of shear strain). In a dynamic shear test, rate of shear strain is the product ω × γ. Definitions of dynamic viscosity are: Dynamic Loss Viscosity = ′ =
G′′ s,0 sin = 0
Dynamic Storage Viscosity = ′′ =
G′ s,0 cos = 0
′ Loss Tangent = = tan ′′
(8)
(9) (10)
In the Newtonian liquid limit (δ = 90°), η″ is zero by Eq (9), and all of the response is viscous, η′ from Eq (8). Furthermore in this limit, G′ is zero by either Eq (9) or Eq (5), and the viscous response could alternatively be expressed in terms of G″ from Eq (6). Thus, in dynamic shear tests, modulus and viscosity determinations are one and the same experiment. Usually experimenters will choose Eqs (5) and (6) when the sample has a lot of elastic character and only a little viscous character, whereas the logical choice for
15TH EDITION
mainly viscous materials having only a little elastic character will be Eqs (8) and (9). In principle, it would be valid to express tensile DMA results in terms of dynamic viscosities as well. This is not done very often, probably because viscosity is strongly associated with shearing experiments, not with tensile experiments. The close association of viscosity with shear is the basis for omitting the word “shear” at the left in Eqs (8) and (9). The relationships between modulus values defined in Fig. 1 and the dynamic modulus values are [1–4]: E 2 = E 2 + E′′2 2
2
2
G = G′ + G′′
(11) (12)
Use of complex numbers and quantities such as i = −1 has been avoided here. If readers would like definitions of such quantities as the “complex tensile storage modulus,” E*, they should consult Refs. [1–4]. Complex notation may be convenient for mathematical derivations, but complex modulus values, with their imaginary parts, add little or nothing to the interpretation of structure/property relationships.
Definitions of Tensile Properties
The term “tensile properties” logically refers to all properties that can be determined in tests that involve tensile deformation as depicted in Fig. 1 (A). Common tests that involve tensile deformation include stress-strain tests, creep tests, and stress relaxation tests. Stress-strain tests are used much more frequently than the others. Therefore, the terms “tensile properties” and “stress-strain properties” are often used interchangeably. Creep and stress-relaxation tests are sometimes referred to as “transient tests” because responses, either elongation or stress, change with time and are determined as a function of time [5]. Terminology, definitions, and symbols for stress-strain tests have been taken from earlier editions of the Paint Testing Manual. Additional terminology was taken from ASTM standards for various polymeric materials: D2370
Test Method for Tensile Properties of Organic Coatings
D638
Test Method for Tensile Properties of Plastics
D882
Test Method for Tensile Properties of Thin Plastic Sheeting
D412
Test Methods for Vulcanized Rubber and Thermoplastic Rubbers and Thermoplastic Elastomers—Tension
D883
Terminology Relating to Plastics
In a stress-strain test, the sample is elongated at constant rate. The force, also called “load,” required to maintain constant rate of elongation is determined. Force is converted to tensile stress (σt) by division by the initial cross-sectional area (A × B in Fig. 1 (A)). Results are presented as a plot of stress (σt) on the vertical axis versus strain (either ε or 100 × ε = % elongation) on the horizontal axis. A hypothetical example is shown in Fig. 3 [5,6]. The tensile modulus (E) is the slope of the initial, linear portion of the plot (see Fig. 3). If the initial part of the plot is not linear, several calculations for estimating E have been suggested in ASTM D638. Use of the slope for E amplifies the simple definition of tensile modulus given in Fig. 1 (A).
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 49
Q
DYNAMIC MECHANICAL AND TENSILE PROPERTIES
627
tion of an ideal Hookean spring of modulus, E, and a dashpot that contains an ideal Newtonian liquid of viscosity, η. As indicated in Fig. 4, τ is the viscosity of the liquid in the dashpot divided by the modulus of the spring. τ has units of time. Results of the creep experiment for the Maxwell model can be expressed as
ε ( t ) = ε (0 ) +
Fig. 3—A hypothetical stress-strain curve for a ductile film. Tensile properties are defined: tensile modulus (E), elongation at yield (εY), elongation at break (εB), yield stress (σY), and tensile strength (σB).
Other terms used for tensile modulus include “elastic modulus,” “Young’s modulus,” and “stiffness.” The first point on the plot of Fig. 3 where the slope is zero is called the “yield point.” Strain at the yield point is called “elongation at yield” (εY). Stress at the yield point is called “yield strength” (σY). Elongation is continued until the sample breaks. Strain at the break point is called “elongation at break” (εB). Stress at the break point is called “tensile strength” (σB) as shown in Fig. 3. However, in some cases (not shown) the stress is higher at the yield point than it is at the break point. In such cases, ASTM standards specify that the “tensile strength” be indicated as the higher value of stress and be designated as “tensile strength at yield.” Practice is not uniform with regard to this latter “tensile strength” terminology. Results of transient tests have not frequently been published for coatings. Such tests clarify viscoelastic character quite directly. Possibly unexpected field failures of coatings could be avoided in some cases if more attention were given to viscoelasticity. Only the most simple form of retardation and relaxation concepts are treated here. In a tensile creep experiment, the sample is subjected to constant stress, σt, and elongation is determined as a function of time, ε(t). Analysis of dependence of elongation on time yields “retardation time,” τ. The simplest mechanical model that permits definition of τ is the Maxwell model as shown in Fig. 4. This model consists of a series connec-
Fig. 4—A mechanical model consisting of a spring and a dashpot permits definition of relaxation time and retardation time.
ε (0 ) t
(13)
When stress is first applied, the spring extends instantaneously by an amount ε(0). Then retarded further elongation takes place due to flow in the dashpot. It is evident in Eq (13) that the retarded elongation is linear with time. The value of τ can be obtained from the product (reciprocal of the slope) × (intercept) [4,5]. In a tensile stress relaxation experiment, the sample is elongated instantaneously by an amount ε, and thereafter ε is held constant. Stress is determined as function of time, σ(t). Analysis of the dependence of stress on time yields “relaxation time,” τ. For the Maxwell model, τ values are the same whether from creep or relaxation. For real materials, experimentation is required to determine whether or not retardation and relaxation values are equal. Results of the stress relaxation experiment for the Maxwell model can be expressed as
( t ) = (0 ) e − t /
(14)
When the instantaneous elongation is applied, the time zero response is entirely in the spring. Then the dashpot extends with time relieving stress on the spring. It is evident from Eq (14) that the value of τ can be obtained as the time at which stress has been reduced to 1/e (0.368) of its initial value. Alternatively, one can obtain the value of τ from the negative of the reciprocal of the slope of a plot of ln σ(t) versus t [4,5]. To represent mechanical response of viscoelastic polymeric materials, it is usually necessary to use more elaborate mechanical models and to replace a single value of τ by “a spectrum of relaxation times” [1,2].
PREPARATION OF FREE FILM SAMPLES
Methods for preparation and cure of adherent films are described elsewhere in the manual and in ASTM Standard Practice for Producing Films of Uniform Thickness of Paint, Varnish, and Related Products on Test Panels (D823). Since dynamic mechanical and tensile property determinations require free films, ASTM Practice for Preparation of Free Films of Organic Coatings (D4708), is also very useful. Details concerning thickness measurements, which are required for calculation of cross-sectional area, are given in ASTM Test Method for Measurement of Dry-Film Thickness of Organic Coatings Using Micrometers (D1005). The most widely used method for free film preparation involves application on release substrates, i.e., low surface energy substrates. Four release substrates are described in ASTM D4708. Low surface energy results in poor adhesion so that the coating can be stripped from the release substrate after it is cured. Surface tension differences between the coating and release substrate must be carefully balanced. If the surface tension of the liquid coating is higher than the critical surface tension of the release substrate, the coating will crawl inward from the edges to
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
628
PAINT AND COATING TESTING MANUAL
Q
give nonuniform thickness. In extreme cases of crawling, the coating will break up into unconnected puddles. There are many types of release paper and many surface treating agents to convert glass or metal panels into low-energy surfaces. It is worthwhile to try several release surfaces to find the balance that will avoid crawling but will still permit separation. Tendency to crawl can be reduced by use of high-viscosity formulations. Viscosities for draw-down application (see ASTM D823) can be quite high compared to those required for spray application. Thickness is an important consideration in preparation of samples for dynamic mechanical and tensile property tests. Usually tests are more reproducible if samples are thicker than normal coatings thicknesses. However, film formation seldom occurs in exactly the same way for thick films as for thin ones. Several reasons for dependence of cured film properties on thickness have been discussed [5]. In our laboratory, films of 1 mil (25 μm) and up have been analyzed routinely. It has not been possible to analyze or even handle very thin free films such as beverage can coatings, which are approximately 3 to 5 μm thick. Coatings and thin plastic film samples are usually prepared as rectangular strips, whereas thicker materials are prepared as dogbone-shaped samples. The narrowing in the middle of the dogbone tends to control where failure occurs in tension tests, but for thin samples this narrowing causes the cross-sectional area to be too small for many load measuring cells. The dogbone shape also provides more area for clamping, which facilitates the balance between slip-free clamping and avoidance of rupture at the clamp. There is extensive literature on the notch sensitivity of polymeric materials in stress-strain tests [7]. The challenge in preparation of high Tg coating samples is to avoid undesired notches, nicks, or cracks along the edges. Small edge cracks, which are very difficult to detect even with magnification, can cause premature failure in tension tests. Usually samples are die cut, and the sharpness and condition of cutting edges of the die affect uniformity of sample edges.
DMA
Instruments and Methods Used for DMA
Many DMA practices used for plastics have been applied to coatings as well. The broadest standard practice is: ASTM D4065 Standard Practice for Plastics— Dynamic Mechanical Properties: Determination and Report Procedures Thirteen procedures for applying oscillating strain are diagrammed in D4065, and equations are given for calculating dynamic properties for each. Other methods are specific to the type of oscillating deformation used: Test Method for Plastics: DMA D5023
Flexure—Three Point Bending
D5024
Compression
D5026
Tension (most common for coatings)
D5279
Torsion (sometimes called shear)
D5418
Flexure—Dual Cantilever Beam
One of the reasons for rapidly expanding use of DMA for coatings and other research is a rather long history of
15TH EDITION
TABLE 1—DMA instruments Cited, 2002–2006 Company
Model Number
Number of Citations
References
Rheometrics Scientific
DMTA—3
Four
[29–31,51]
Perkin Elmer
DMTA—7
Three
[35,54,55]
TA Instruments
2980
Two
[32,56]
Polymer Laboratories
PL & (England)
Two
[48,57]
improvements in the quality of automation and computercontrol of instruments. DMA instrumentation was undergoing rapid change and development in 1995 when the last edition of this manual was published. Eighteen DMA instruments were tabulated for coatings used prior to 1995. A great deal of consolidation has occurred in terms of instruments offered and instrument companies active in DMA. Table 1 lists the instruments used in eleven papers published from 2002 to 2006. A common feature of all these instruments is oscillatory deformation, an example of which is shown in Fig. 2. Variable features include: the type of deformation (tensile, shear, etc. ), free versus forced oscillation, frequency scan versus temperature scan versus either, sensitivity for thin film analysis, capability of traversing the entire range of property behavior (glassy to transition to rubbery) during a single temperature scan, breadth and rate of temperature scan, breadth and rate of frequency scan or range and number of frequency settings, versatility of sample holding devices, ruggedness versus flimsiness, amount of attention required once a run has been started, accuracy and versatility of the associated software for control during the run and data treatment and plotting after the run. In several cases, newer models permit determination of properties at several frequencies during a single temperature scan. The procedure for carrying out a DMA run on an automated instrument is rather simple, with details depending on the particular instrument. Usually the associated software includes a “run” program that prompts the operator to input sample data (e.g., thickness) and settings for the run such as initial and final temperatures, frequency of oscillation, heating rate, etc. After the input steps, there is usually a cooling period. The instrument takes over when the preset initial temperature is reached. Thereafter nothing is required of the operator until the run is finished. Usually initial and final temperatures are selected to span the glassy region, transition region, and rubbery plateau region. For direct tensile DMA, the run program usually contains a tensioning sub-routine, which provides constant static tension sufficient to avoid slack in the sample in the glassy region and then decreasing static tension as the sample softens in the transition region. Modification of the tensioning sub-routine is often necessary. If tension is too high, films break in the glassy region or are pulled apart in the transition region. If tension is too low, slack results or the oscillatory stress falls below measurable values. Skilled operators soon develop several modified run programs with different tensioning parameters that are suitable for samples of various properties and dimensions. Calibration of DMA instruments is described in E2254 Storage Modulus Calibration of DMA.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 49
Q
DYNAMIC MECHANICAL AND TENSILE PROPERTIES
629
Fig. 5—DMA plots for a clearcoat prepared from an ACR and an etherified MF resin, ACR/MF 70/30. The film was cured before DMA for 30 min at 120°C with 0.5 % para-toluenesulfonic acid. DMA was carried out at 11 Hz. See Table 2 for dynamic properties.
Various support materials, such as glass braids and metal springs or shims, are used when the sample is a viscous liquid and the objective is to follow cure as the coating formulation is converted to a solid. Use of supports makes it difficult to obtain absolute values of E′, but relative values are often sufficient for component optimization. The position of peaks in E″ or tan δ plots are usually not shifted when a support is used. When coating samples do not contain solvents, e.g., powder coatings and 100 % active coatings, parallel plate, and cone and plate geometries can be used. In some cases, DMA is treated as an adjunct to thermomechanical analysis (TMA). The TMA probe is driven up and down in an oscillatory manner as temperature is scanned. General indications of liquid-like character during transitions can be obtained, but quantitative DMA data are not often obtainable in this manner.
Interpretation of DMA Plots
DMA plots are shown in Fig. 5 for a clear film prepared from an acrylic polyol (ACR) and an etherified melamine formaldehyde (MF) cross-linker. Plots are labeled according to Eqs (2)–(4). The storage modulus level at the left is typical of amorphous, unpigmented films in the glassy state. The 25°C data are of interest for comparison with modulus values obtained from the slope of stress-strain plots because SSA is usually carried out at 25°C. The value of E′ (25°) from the computer printout corresponding to Fig. 5 is 1.38 × 1010 dynes/cm2. The same value expressed in other units is 1.38 × 109 Pa or 2.00 × 105 psi (1 pascal = 1 newton/m2 = 10 dynes/cm2 = 1.45 × 10–4 pounds-force per square inch). The recommended SI unit is Pa. The value of E″ (25°) is 6.86 × 108 dynes/cm2. Inserting E′ and E″ (25°) values into Eq (11) and solving for tensile modulus, E, we obtain E = 1.382 × 1010. In this case E = E′ (25°) to a very close approximation, and the contribution of E″ (25°) to E is negligible. Hard, tough coatings often have tensile modulus
values from SSA ranging from 1 × 1010 to 3 × 1010 dynes/cm2 [5,7,8], in agreement with the DMA value. For a quantitative comparison of E from SSA and E from DMA via Eq (11), the strain rate from SSA would have to be matched with the oscillating frequency from DMA. The viscous response is not always negligible relative to elastic response, of course. The E″ contribution is highest at the temperature of the maximum in tan δ, 79°C in Fig. 5. Note that E′ (79°) is 1.04 × 109 dynes/cm2, and E′ (79°) is 5.15 × 108 dynes/cm2. From Eq (11), E = 1.16 × 109 dynes/ cm2, and E > E′ (79°) for this case. If we actually carried out SSA at 79° and a strain rate corresponding to 11 Hz, we would expect to find E = 1.16 × 109 dynes/cm2. During the elongation in SSA, it would not be evident that a significant fraction of the resistance was viscous in nature. However, after the sample broke or was released from the grips, retraction would be delayed (not instantaneous) and part of the deformation would be permanent. The middle portion of the plots in Fig. 5 represents the transition region where E′ drops sharply and both E″ and tan δ go through a maximum. In coatings papers the glass transition temperature is usually taken as the temperature of the maximum in the tan δ plot, and the symbol used for this temperature is Tg. This definition will be used in this chapter. Different assignments and symbols of various transition temperatures are described in ASTM E1640 Standard Test Method for Assignment of the Glass Transition Temperature by Dynamic Mechanical Analysis: Peak temperature of the loss modulus plot = T1 (55°C in Fig. 5) Peak temperature of the tan δ plot = Tt (79°C in Fig. 5) Intercept of two E′ tangents= Tg (52°C in Fig. 5) The two tangents in the last assignment are drawn (1) along the glassy portion and (2) through the inflection point in the transition region. The lower transition values
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
630
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
TABLE 2—Dynamic properties of acrylic/MF clearcoat Frequency, Hz
Tg, °C
Tan δ, max
PW,a ∆,°C
E′ (min), dynes/cm2
Temperature of E′ (min),°C
3.5
76
0.49
45
3.3 × 108
108
11.0
79
0.49
50
3.6 × 10
8
112
35.0
83
0.52
55
4.2 × 10
8
116
110.0
88
0.52
59
4.7 × 10
8
121
PW is the peak width at half height.
a
agree more closely with that obtained from differential scanning calorimetry, but the highest value can be obtained with better reproducibility because the tan δ peak is prominent. Coatings often have several binder components, and many coatings are cross-linked. These structural features broaden the glass transition and favor use of the peak in tan δ for assigning Tg. Regardless of the data point selected to represent Tg, it is important to remember that Tg depends on rate of testing for viscoelastic materials. The effects of changing rate of testing, i.e., frequency in DMA, are shown in Table 2. The 11 Hz data of Table 2 correspond to the run depicted in Fig. 5. A ten-fold increase in frequency results in a 7° to 9° increase in Tg with a slightly stronger dependence on frequency at the higher frequencies. Hartmann [9] noted that a 7°C Tg change per decade change in frequency is used as a “rule-of-thumb.” Results of Table 2 are consistent with this generalization. The height of the tan δ peak is nearly independent of frequency, but the width increases with increasing frequency. Very similar dependence of height and width on frequency was observed for lightly cross-linked epoxy films [9]. Values of E′ (min) increase slightly with increasing frequency (Table 2). The structural implications of E′ (min) will be discussed in the next sub-section. Height and width of tan δ peaks reflect structural homogeneity and cross-link density. Homogeneous, uncross-linked, noncrystalline polymeric materials of narrow molecular weight distribution usually have tan δ (max) values greater than one and sometimes greater than two. Such tan δ peaks are very narrow. A broader molecular weight distribution results in a wider peak and a lower value of tan δ (max). Introduction of cross-links invariably reduces tan δ (max) and usually increases peak width. For homogeneous cross-linked samples, peak width reflects the broadness of the distribution of lengths of chains between junction points in the network [2,3]. For sound and vibration damping, materials with both high and wide tan δ peaks would be desirable, but height and width cannot be adjusted independently [9]. High peaks tend to be narrow, and wide peaks tend to be low. These observations have resulted in speculation concerning a general comparability of peak areas of the glass transition. Eventually theoretical treatment of peak areas may prove to be useful for structure/property correlations. Samples that undergo partial phase separation during molecular weight buildup and cross-linking often have very broad transitions [4,9,10]. Manson and Sperling [10] have described the use of interpenetrating polymer networks to limit and control the extent of phase separation. When Tg values of the separate domains are close to one another, a
general broadening is observed. When Tg values of the separate domains are considerably different, tan δ plot shapes provide considerable structural information. Observation of two narrow peaks with low tan δ values between them indicates distinct phase separation with little mixing at domain boundaries. Observation to two broad and indistinct peaks with high tan δ values between them indicates diffuse domains with extended regions of varying composition at the boundaries [4,10]. Diffuse domains have also been created in multiblock uncross-linked polymers, and block design has resulted in controlled broadening of tan δ peaks [11]. The relationships between transition width and coating end-use performance have been presented for radiationcured coatings [12], polyol/melamine thermoset coatings [13,14], gel coats [14], and can coatings [14]. Although the glass transition is emphasized in this section, many polymers are known to have multiple transitions. Low-temperature transitions are observed in DMA as tan δ or E″ peaks that are quite small compared to the glass transition peaks. As discussed in the section entitled “Relationship to Other Mechanical Properties,” good impact resistance is often attributed to transitions that produce low-temperature loss peaks. Interpretation of tan δ peak broadening in terms of structural nonuniformity requires considerable restraint because in some cases broadening is caused by physical or chemical changes that take place during the temperature scan. In some cases, DMA has its own uncertainty principle; the structural features that are under study are changing during the determination. Physical changes include loss of plasticizer or absorbed water and morphological changes such as partial crystallization. Chemical changes include additional cross-linking of thermoset coating during the scan and oxidative or thermal degradation. In general, if the temperature of processing the samples is well above the temperature of the tan δ peak under analysis, there is much less chance that properties are changing during the scan.
Determination of Cross-Link Density
For unpigmented, cross-linked coating films, the level of storage modulus, E′, in the rubbery plateau region above Tg is an indication of cross-link density. A wide range of variation has been observed from approximately 4 × 107 dynes/ cm2 for lightly cross-linked films to approximately 2 × 109 dynes/cm2 for every highly cross-linked films. Increases in E′ values in the rubbery plateau have been attributed to increase in cross-link density for many types of studies: increasing cure temperature and cure time [13,16–18], increasing radiation dose in electron beam curing [15],
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 49
Q
DYNAMIC MECHANICAL AND TENSILE PROPERTIES
increasing the stoichiometric balance in epoxy films [2,19] and in polyol/melamine films [13,17,18], increasing the functionality of the cross-linker in radiation cure films [20] and in powder coating [21,22], and increasing the molecular weight of the main film former in acrylic clearcoats [23]. Quantitative aspects of the relationship between E′ in the rubbery plateau and cross-link density have been clarified [18]. Cross-link density is defined as ve = moles of elastically effective network ch hains per cubic centimeter of film
(15)
Cross-link density can be calculated using Eq (16) G′ E′ ve = = RT 3 RT
(16)
where the storage modulus values, G′ or E′, are obtained in the rubber plateau, T is temperature in degrees K corresponding to the storage modulus value, and R is the gas constant (8.314 × 107 dynes/degrees K · mole in the cgs unit system). Inserting the 3.5 Hz data from Table 2 (E′ = 3.3 × 108 dynes/cm2 and T = 108°C = 381°K) into Eq (16) giνes ve = 3.47 × 10–3 moles/cm3. Since Eq (16) has no correction for frequency dependence of E′, a more accurate value would be obtained if frequency were reduced until E′ no longer depended on frequency. Extrapolation to zero frequency was used in cross-link density calculations for powder coatings [21]. For a polyurethane film of low Tg but relatively high cross-link density, E′ plots were identical in the rubbery plateau at 11 and 110 Hz [18]; thus, no frequency extrapolation was necessary for the is case. Equation (16) has been called “the ideal network law” with an analogy implied to the ideal gas law [18]. In an ideal network, all chains are elastically effective. Conversion in the network forming reaction is complete, and there are no small loops or dangling ends. For ideal networks formed by functional group reactions of terminally functional (telechelic) starting material, the value of ve can be calculated directly from Eq (15). The calculation requires only a balanced chemical equation and an experimental determination of cured film density. For example, consider a tetrafunctional coreactant B4 (Mn = 3000) cured by a trifunctional cross-linker A3 (Mn = 300) 4A3 + 3B4→ Ideal Network (density = 1.10 g/cm3). The mass of the network formed by this reactions is 10,200 g (3 × 3000 + 4 × 300), and the volume is 10,2001/.10 = 9273 cm3. This volume of cured film contains 24 mol of chain ends (4 × 3 + 3 × 4) coming into junction points in the network. By definition, a chain has two ends; therefore, there are 12 mol of elastically effective chains in 9273 cm3 of ideal network film. From Eq (15), ve = 12/9273 = 1.29 × 10–3 mol/cm3. Equation (17), attributed to Scanlon [18,24], formalizes this type of calculation ve =
3 4 5 1 ∞ C3 + C4 + C5 + … = ∑ fCf 2 2 2 2 f =3
(17)
where Cf is the concentration of “f” functional reactant expressed in unusual units, i.e., moles of reactant per cubic centimeter of final film. Difunctional reactants contribute to volume but do not increase the number of moles of chains. Therefore, there is no C2 term in Eq (17).
631
Ideal network calculations have been carried out for several types of coating such as polyester polyol/melamine films [17], epoxy/diamine films [19], and powder coating films [21,22]. The ve values calculated from reactant structures agree remarkably well with experimental ve values from Eq (16). Although ve is useful for network characterization, most coating chemists can more easily visualize a network based on the value of Mc Mc = weight of sample in grams that contains one mole of elastically effective chains
(18)
If chain lengths in the network vary, one can place a bar over Mc and refer to this quantity as “number average molecular weight of effective network chains.” Based on Eqs (15) and (18), the relationship between vc and Mc is Mc = (19) ve where ρ is film density in g/cm3. For the 4A3 + 3B4 example, Mc =1.10/1.29 × 10–3 = 853 g/mol. For the film of Table 2, the experimental density is 1.12 g/cm3, and Mc is 1.12/3.47 × 10–3 = 323 g/mol. Or course, a high value of Mc corresponds to a loose network and a low value to a tight network. Equation (16) can be considered empirical or the result of kinetic theory of rubber elasticity [20,24–28]. The theory was developed for networks that have very long chains between junction points. For long chains, conformations can be treated by statistical-mechanics concepts. The chains in networks of greatest interest for coatings are much too short for such treatments. For long chains between junction points, results of the theory are of ten expressed as [20] G ve = vc + vp = (20) gRT where ve is sum of a chemical contribution, is vp. The physical contribution is attributed to chain entanglements [28] or physical constraints [25–27]. The factor, g, is related to junction point displacement under stress and is reported to depend on functionality [20,25]. If the reader chooses to use rubber elasticity theory, Eq (16) is obtained from Eq (20) with g = 1.0, G = G′, and vp = 0 (relative to vc). It has been shown that Eq (16) can be derived rather directly for networks made up of short chains such that chain entanglements are minimal [18]. For many years DMA has been used to determine the extent of cure of thermoset coatings. Cross-link density (Eq (16)) and glass transition temperature (from the maximum in the plot of tan δ versus temperature) are used as measures of extent of cure. Recent examples include: Acrylates cross-linked by hydroxyl alkyl amides [29] (2006) Catiionic UV curable coatings [30] (2004) UV cured acrylate/hyperbranched polyester coatings [31] (2003) Urethane/acrylic hybrid coatings [32] (2005) Alternative UV cure methods for automotive clearcoats [33] (2004) High performance all-acrylic coatings [34] (2003) Pigmented or nanoparticle-reinforced coatings, of course, do not have a direct relationship between E′ (min)
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
632
PAINT AND COATING TESTING MANUAL
Q
and cross-link density. Finely divided particles, as well as cross-links, cause an increase in E′ (min). Rodriguez et al. [35] have reported on the effects of varying Pigment Volume Concentration (PVC) on dynamic mechanical properties of epoxy coatings. Soucek and Ni [36] have reported the effect on dynamic properties of varying the content of oligosiloxane ceramer in a polyurethane coating for aircraft.
RELATIONSHIP TO OTHER MECHANICAL PROPERTIES
Knowledge of dynamic properties is useful for optimizing the chemical structure of coating components. When the binder is a copolymer, monomer ratios can be altered to control thermosets, increasing the functionality of the cross-linker is expected to cause an increase in cross-link density. DMA is a very direct method for determining whether purposeful structural changes have actually had the desired effects. In addition to general structure/property use of DMA, progress is being made in establishing relationships between dynamic properties and the results of paint test methods for hardness, flexibility, impact resistance, and solvent resistance. DMA has some inherent limitations in the establishment of these property/property relationships. The oscillatory deformation, e.g., ε0 in Fig. 2, is very small. If the paint performance property depends critically on large deformations, a property/property correlation should not be expected. Results of paint performance tests usually depend to some extent on interactions between coating and substrate. When dependence on coating-substrate interactions is dominant, a free film method such as DMA should not be expected to correlate with results of paint tests. DMA has helped elucidate the causes of reversals in rank orders of hardness among films when different tests are used [5,37]. As described elsewhere in the manual, hardness is determined by penetration, pendulum, and scratch test methods. The most important property for influencing hardness is Tg. However, even for a series of films that have the same Tg, considerable differences are sometimes observed in hardness test results. We have observed that penetration hardness depends more directly on the E′ (25°) value than on Tg. Films of the same Tg can have quite different values of E′ (25°). Pendulum hardness depends more directly on E″ values than on E′ values. This result is expected because damping of the swings of a pendulum depends on conversion of mechanical energy into heat through viscous effects in the coating. As noted earlier, E″ is a measure of viscous response. Films that have nearly the same E′ (25°) values and the same penetration hardness can have quite different E″ (25°) values and quite different values of pendulum hardness. The back and forth rolling motion of a Sward Rocker is also damped by viscous effects, and therefore depends strongly on E″ values. In this case, there is also a contribution from sliding friction. Usually pendulum hardness is determined at a single temperature, but Sato [37] describes studies of damping time as a function of temperature. Plots of the reciprocal of damping time versus temperature have exactly the same shape as the tan δ plot for the same films. Both glassy and rubbery materials have low tan δ values as shown in Fig. 5. If a pendulum hardness test is carried out at room temperature on a material that is in its rubbery region at room temperature,
15TH EDITION
a careless interpretation of results is likely to conclude that rubber is very hard. This example illustrates that DMA can be used to understand the results of paint tests more fully. Among the many scratch tests that have been devised, the most widely used is pencil hardness. Although pencil hardness results are quite reproducible when carried out by one skilled in the art, these results sometimes do not correlate with either penetration or pendulum hardness results. Furthermore, there is no known dynamic property that correlates well with pencil hardness. The plowing action in pencil hardness failures involve large localized deformations. A relation to stress-strain analysis, which also involves large deformation, is more likely. Flexibility of coatings is often measured by mandrel bend tests and falling weight impact test (see a previous chapter in this book). In thermoplastic polymer studies, good flexibility and impact resistance are often associated with low temperature peaks in E″ and tan δ plots [38]. Rubber-toughened epoxy coatings clearly show the low temperature peak attributable to the rubber phase [39]. Polyester/melamine films have much better mandrel bend performance than do acrylic/melamine films of similar Tg and cross-link density. DMA scans beginning at –100°C show a weak tan δ peak at –70°C for the PE/MF films, but no such peak is observed for the ACR/MF films [18]. In some cases [12,13], extreme broadening of the peak in the E″ plot has been associated with improvements in flexibility of coatings. Pigments often increase the toughness of coatings without broadening the transition or introducing lowtemperature loss peaks [15]. The main effect of increasing pigment volume content (PVC = 0 to 55 %) on DMA plots is to increase E′ values moderately in the glassy region and strongly in the rubbery plateau [15]. Values of Tg often increase by about 5 to 10°C at PVC = 0.4 relative to the corresponding clear coatings [15]. Solvent resistance as measured in methyl ethyl ketone (MEK) double rub tests is related to E′ values in the rubbery plateau, i.e., to cross-link density [5,13]. MEK resistance also depends on the solubility parameter of the coating. Although the double rub test is widely used to determine the degree of cure of thermoset coatings, the only ASTM method adopted relates to zinc-rich primers (D4752). Relationships between the number of double rubs and E′(min) values are quite reproducible within a coating type but not from type to type. For example, with acrylic clearcoats, 200 double rubs are obtained when E′(min) = 2 × 108 dynes/cm2 [13]. Films of this type having 50 to 100 double rubs typically have E′ (min) values in the range of 5 × 107 to 1 × 108 dynes/cm2. In contrast, polyester polyol powder coatings, cured with several types of cross-linkers, yielded 200+ double rubs despite having E′ (min) values as low as 5 × 107 dynes/cm2 [22]. The lack of generality in the relationship of MEK resistance to E′ (min) for various coatings types is believed to result in part from differences in solvent-polymer interactions. There may also be a weak dependence of MEK resistance on Tg as well as a strong dependence on E′ (min).
DETERMINATION OF TENSILE PROPERTIES Description of SSA
Instrumentation for SSA is described in detail in the ASTM test methods listed in the section of this chapter entitled
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 49
Q
DYNAMIC MECHANICAL AND TENSILE PROPERTIES
“Definitions of Tensile Properties.” In most cases, specific instruments are not identified, apparently to avoid commercial implications, but it is generally known than Instron® instruments (Instron Engineering Corp.) are widely used. For organic coatings, the most suitable test method is usually ASTM D2370. However, other tension test methods contain additional useful information. ASTM D638 is useful for selection of metric units and units consistent with SI recommendations. ASTM D882 has rather extensive data on statistical treatments for plastic sheeting, but the statistical methods could be beneficially applied to coatings films. For certain types of coatings, such as flexible primers or coatings for plastic substrates, rubbery behavior is expected, and ASTM D412 provides useful information such as data treatment when a crack moves slowly across the sample as opposed to the more usual instantaneous failure mode. Often stress-strain curves do not start out with constant slope as shown in Fig. 3, but instead have an initial curvature caused by takeup of slack. The initial curved portion is called a “toe,” and toe compensation is described in ASTM D638 and in ASTM D882. Some stress-strain curves have no linear (i.e., no Hookean) region from which to calculate the slope for modulus values. In such instances, the “toe corrected” origin and another point on the curve are joined by a straight line. The slope of the line is reported as the “secant modulus.” The elongation percentage at the second point selected is always reported along with the value of secant modulus. Graphical illustrations of secant modulus determinations are given in ASTM D638 and in ASTM D882. Among the various test methods, D638 contains the most complete list of definitions and symbols. Determination of “tensile energy to break” is described in ASTM D882. The area under the stress-strain curve, e.g., see Fig. 3, is divided by sample volume to obtain this quantity, which has also been called “work-to-break” or “toughness” [5]. Recommended units are megajoules per cubic meter. ASTM D882 describes how an integrated chart paper area (distance2) can be converted to energy/volume by using the ordinate setting (force/distance of chart paper), the abscissa setting (distance of elongation/distance of chart paper), and the sample dimensions. Now that computers are used to control operation and to treat data for SSA [40], numerical integration is nearly instantaneous, and tensile energy to break is likely to be reported more often. This SSA property may prove to be useful for correlations with paint test results.
Interpretation of Stress-Strain Curves
Schematic stress-strain curves for various types of polymeric materials are shown in Fig. 6 [2]. This extremely wide range of property variation is represented in coatings of various types. Scales on the graphs give an order-of-magnitude indication of property values. Brittle materials (6A) have high modulus values (initial slope), tensile strengths up to about 8000 psi (5.5 × 107 Pa = 5.5 × 108 dynes/cm2), and elongations below 10 %. Ductile materials (6B) usually have lower initial slopes, tensile strengths in the 4000 to 6000 psi range, and elongations of about 100 %. The upper plot in 6B corresponds to yielding with uniform sample deformation between the grips. The lower plot in 6B corresponds to a ductile sample that necks down at the yield point, and further elongation (~40 % to 110 % in Fig. 6) occurs with increase in length of the necked part of the
633
Fig. 6—General kinds of stress-strain curves for various types of coatings. Scales indicate order-of-magnitude values.
sample. Elastomeric (rubbery) samples (6C) have much lower initial slopes than brittle materials, tensile strengths of about 2000 psi, and elongations of the order of 400 % to 500 %. The upward curvature near the end of the lower plot in Part C is attributed to strain-induced crystallization [2]. Materials represented in 6A and 6B of Fig. 6 have Tg values above the test temperature. Therefore, modulus values depend on secondary interactions of polymer chain segments and partial crystallinity, if any exists. The elastomeric material (6C) has Tg well below the test temperature. Therefore, modulus values depend in part on cross-link density, chain entanglements, or both. Stress-strain curves are carried out at a constant rate of strain and results depend on strain rate selected. In general, a higher rate results in higher modulus. The two curves in 6A could represent the same material strained at different rates. In fact a large increase in strain rate could cause the plots of 6B to be converted to the plots of 6A. Dependence on strain rate is evidence of viscoelasticity. Strain rate dependence in SSA, therefore, has the same origin as dependence on oscillatory frequency in DMA (see Table 2). A recent study of silicone-epoxy resins cross-linked with amines of various functionality [40] illustrates the extreme range of stress-strain properties exhibited by coating films. Tensile strength ranged from 31 to 4418 psi. Percent elongation ranged from 7 % to 177 %. Unfortunately, neither modulus values nor tensile energy to break were reported. Property variations were attributed to differences in degree of entanglement and to partial phase separation. In a network-forming thermoset system, it would be impossible to get a tensile strength as low as 31 psi unless incompatibility had prevented occurrence of the networkforming reaction. Examples of all types of behavior shown in Fig. 6 are represented in a single figure for coatings used on naval aircraft [41] (see Fig. 4, Ref. [41]). One plot shows results for a polysulfide sealant that has a tensile strength of about 250 psi and an elongation >130 %. Results for a flexible polyurethane primer are a tensile strength of 3500 psi, a
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
634
PAINT AND COATING TESTING MANUAL
15TH EDITION
Q
step [34], which permits use of a wide range of epoxy:polyol ratio without need for stoichiometric balance. Selection of flexible polyols with this wide formulating latitude permits preparation of cross-linked films with an extremely wide range of mechanical properties. Numerous stress-strain curves have been presented for this type of system with oligomeric propylene oxide polyols [43] and with oligomeric caprolactone polyols [44]. Examples of stress-strain curves for a cycloaliphatic diepoxide (CYRACURE® UVR6110, Union Carbide) and a propylene oxide triol (MW = 702) are shown in Fig. 7 [43]. The corresponding tensile properties are given in Table 3. Films from Compositions 1 through 4 all give brittle failure (compare Fig. 6(A)) and only moderate changes in tensile properties (Table 3) despite a large change in composition. At Composition 5 (70/30 epoxy:polyol, see Table 3) a yield point is first noted, and thereafter very large property changes occur despite relatively small changes in composition. Compositions in the 4 to 6 range (see Table 3) are reported to give excellent post formability, as required for coatings on beverage can ends, while maintaining adequate hardness and solvent resistance [42–44].
Relationship to Other Mechanical Properties Fig. 7—Stress-strain curves for UV-cured cycloaliphatic epoxide films flexibilized with oligomeric propylene oxide triol. Strain rate is 40 % per minute. See Table 3 for tensile properties.
yield strength of 3200 psi, an elongation at break of 90 %, and an elongation at yield of 40 %. The plot shows that a polyurethane topcoat yields and breaks at about the same point: 4000 psi and 22 % elongation. The plot for an epoxy primer shows brittle behavior (no yield point), a tensile strength of 2300 psi, and an elongation at break of 7 %. The authors [41] report that replacing the epoxy primer by the flexible polyurethane primer eliminated the need for the sealant coat. SSA has been used extensively to characterize cationic UV-cured cycloaliphatic epoxy/polyol coatings [42–44]. The mechanism of introduction of polyol is a chain transfer
SSA is used in a general way to assess suitability of a binder for various coating end uses. Most coating chemists associate modulus with coating hardness and percentage elongation at break with coating flexibility. Quantitative correlations of these properties are not published very often, however. The lack of published correlations may result from the fact, noted above, that paint tests of adherent coatings depend on coating-substrate interactions, whereas SSA is carried out on free films. ASTM Test Method for Elongation of Attached Organic Coatings with Conical Mandrel Apparatus (D522) describes how to calculate percent elongation from the crack length in a conical mandrel bend test. Comparison of elongation of adherent coatings by the conical mandrel method and elongation of the same coating as a free film from SSA would certainly be of interest, but such comparisons were not found in the literature.
TABLE 3—Tensile properties (strain rate, 40 %/min) of UV-cured cycloaliphatic epoxide films flexibilized with oligomeric propylene oxide triol Compositionb Film Numbera
Epoxide, wt %
Triol, wt %
Tensile Modulus,c psi
Tensile Strength, psi
Elongation, %
1
90.0
10.0
3.88 × 10
5
9.5 × 10
3
6.6
2
85.0
15.0
3.72 × 10
5
8.9 × 10
3
6.6
3
80.0
20.0
3.33 × 10
5
8.4 × 10
3
7.4
4
75.0
25.0
2.95 × 10
5
7.0 × 10
3
8.1
5
70.0
30.0
2.05 × 10
5
4.3 × 10
3
16.2
6
66.7
33.3
1.48 × 10
5
3.7 × 10
3
24.3
7
63.4
36.6
0.70 × 10
5
2.5 × 10
3
54.0
8
60.0
40.0
0.26 × 10
5
2.0 × 10
3
88.4
Keyed to the plots in Fig. 7. Weight % of polymeric binder. (Films also contain 2.9 wt % photoinitiator and 0.5 wt % flow agent.) c 1% secant modulus. (The modulus range expressed in pascals is 2.68 × 109, No. 1, to 1.79 × 108, No. 8.) a
b
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 49
Q
DYNAMIC MECHANICAL AND TENSILE PROPERTIES
The logic of associating yield behavior in SSA with post formability of coil coated metal was noted in 1977 [45].In 1987, Koleske [43,44] confirmed that compositions that exhibited a yield point performed well in the demanding post-forming operations carried out on beverage can ends. Evans and Fogel [46] provided convincing evidence that gloss retention during abrasion of floor coatings is related to the area under stress-strain curves. This area, divided by sample volume, is called “work-to-break” or “toughness” as noted previously. The authors provide a clear example of the need to match strain rates when attempting to correlate SSA results with paint test results. The field of Fracture Mechanics was well established for analysis of cracking of plastics by the 1980s, but coatings publications using Fracture Mechanics concepts began in the 1990s. In 1992 Gregorovich and McGonigal [47] reported that “the essential work to break,” obtained from SSA plots with notched samples, correlated with resistance to car wash marring of automotive topcoats. In 2005, Sadati et al. [48] analyzed scratch resistance of clear coats based on fracture energy. In 1998 Nichols et al. [49] analyzed the fracture energy of automotive clearcoats using Fracture Mechanics, and in 2002 Nichols [50] used Fracture Mechanics concepts for anticipating paint cracking during weathering. In 2006 Skaja et al. [51] combined DMA and SSA data in analysis of mechanical property changes during accelerated weathering of polyester-urethane coatings. In 2001 Soucek et al. [52] analyzed fracture toughness of inorganic-organic hybrids, and in 2002 Soucek and Ni [36] used SSA and Fracture Mechanics to determine mechanical properties of nanostructured polyurethane ceramer coatings for aircraft. Failure of attempts to correlate pencil hardness with penetration hardness, e.g., Tukon Hardness, probably result because the former has a strong requirement for toughness, whereas the latter is more dependent on the modulus value at room temperature. DMA is much more generally applicable to determination of cross-link density (see the chapter subsection entitled “Determination of Cross-Link Density”) than is SSA. If the Tg of a coating binder is well below the temperature at which SSA is carried out, then the modulus from the initial slope of the stress-strain curve is a rubbery plateau modulus and Eq (16) is valid at least in principle. In practice, curvature in stress-strain curves and permanent deformation usually result in inappropriate modulus values. An innovative approach to avoiding the permanent deformation problem consists of reversing the extension mode of SSA so that a retraction plot is also obtained. Hergenrother [53] has applied this tensile retraction method for determination of cross-link density of elastomeric polyurethanes.
CONCLUSIONS
A wide range of automated and computer-controlled instruments is available for determination of dynamic mechanical and tensile properties. Careful review of variable features is necessary to insure suitability for property determinations on coating samples of normal thickness. Determination of basic physical properties makes it possible to integrate structure/property knowledge from many polymer fields with coatings research and development. Free film coating data are much more useful when thoughtfully interpreted in relation to results from tests carried out with films intact
635
on their end-use substrates. This review includes many examples that illustrate the benefits of combining DMA or SSA data with results from well controlled and documented tests as provided by the ASTM. The goal of much of the discussion provided here is better understanding of hardness, flexibility, post-formability, solvent resistance, and abrasion resistance. DMA and SSA are often complementary because strains imposed on test samples are very different. SSA provides information on yield behavior and failure at high strains. DMA provides low strain properties and reveals the viscoelastic nature of coatings very directly and quantitatively. For unpigmented thermoset coatings, values of storage modulus, E′, in the rubbery plateau can be used to calculate cross-link density. Determination of cross-link density usually makes it possible to confirm or deny that purposeful reformulation or changes in resin structure have had the desired effects.
References [1] Aklonis, J. J., and MacKnight, W. J., Introduction to Polymer Viscoelasticity, 2nd ed., Chap. 2, Wiley Interscience, New York, 1983. [2] Nielsen, L. E., Mechanical Properties of Polymers and Composites, Vol. I, Marcel Dekker, New York, 1974. [3] Murayama, T., Dynamic Mechanical Analysis of Polymeric Material, Elsevier, New York, 1978. [4] Sperling, L. H., in Sound and Vibration Damping in Polymers, R. D. Corsaro, and L. H. Sperling, Eds., Chap. 1, ACS Symposium Series 424, American Chemical Society, Washington, DC, 1990. [5] Hill, L. W., Mechanical Properties of Coatings, D. Brezinski, and T. J. Miranda, Eds., Federation of Societies for Coatings Technology, Philadelphia, 1987. [6] Schurr, G. G., “Tensile Strength and Elongation,” Paint Testing Manual, 13th ed., G. G. Sward, Ed., Sec. 5.5, ASTM International, West Conshohocken, PA, 1972. [7] Takano, M., and Nielsen, L. E., “The Notch Sensitivity of Polymeric Materials,” J. Appl. Polym. Sci., Vol. 20, 1976, p. 2193. [8] Wicks, Jr., Z. W., Jones, F. N., and Pappas, S. P., Organic Coatings Science and Technology, Film Formation, Components and Appearance, Vol. 1, Wiley, New York, 1992; Applications, Properties, and Performance, Vol. 2, 1994. [9] Hartman, B., Sound and Vibration Damping in Polymers, R. D. Corsaro and L. H. Sperling, Eds., Chap. 2, ACS Symposium Series 424, American Chemical Society, Washington, DC, 1990. [10] Manson, J. A., and Sperling, L. H., Polymer Blends and Composites, Plenum Press, New York, Chap. 3, 8, and 13, 1976. [11] Cooper, S. L., and Estes, G. M., “Multiphase Polymers,” ACS Advances in Chemistry Series 176, American Chemical Society, Washington, DC, 1979. [12] Roller, M. B., “The Glass Transition: What’s the Point?,” J. Coat. Technol., Vol. 54, No. 691, 1982, p. 33. [13] Hill, L. W., and Kozlowski, K., “The Relationship Between Dynamic Mechanical Measurements and Coatings Properties,” Advances in Organic Coatings Science and Technology, Vol. 10, Proceedings of the Twelfth International Conference in Organic Coatings Science and Technology, A. V. Patsis, Ed., Technomic, Inc., Lancaster, PA, 1986, p. 31. [14] Provder, T., Holsworth, R. M., and Grentzer, T. H., “Dynamic Mechanical Analyzer for Thermal Mechanical Characterization of Organic Coatings,” in Polymer Characterization, C.D. Craver, Ed., Chap. 4, ACS Advances in Chemistry Series 203, American Chemical Society, Washington, DC, 1983. [15] Zosel, A., “Mechanical Behavior of Coating Films,” Prog. Org. Coat., Vol. 8, 1980, p. 47. [16] Skrovanek, D. J., “The Assessment of Cure by Dynamic Thermal Analysis,” Prog. Org. Coat., Vol. 18, 1990, p. 89.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
636
PAINT AND COATING TESTING MANUAL
Q
[17] Hill, L. W., and Kozlowski, K., “Crosslink Density of High Solids MF-Cured Coatings,” J. Coat. Technol., Vol. 59, No. 751, 1987, p. 63. [18] Hill, L. W., “Structure/Property Relationships of Thermoset Coatings,” J. Coat. Technol., Vol. 64, No. 808, 1992, p. 29. See also: Hill, L. W., “Calculation of Crosslink Density in Short Chain Networks,” Prog. Org. Coat., Vol. 31, 1997, 235. [19] Grillet, A. C., Galy, J., Gerard, J.-F., and Pascault, J.-P., “Mechanical and Viscoelastic Properties of Epoxy Networks Cured with Aromatic Diamines,” Polymer, Vol. 32, No. 10, 1991, p. 1885. [20] Yeo, J. K., Sperling, L. H., and Thomas, D. A., “Rubber Elasticity of Poly (n-butyl Acrylate) Networks Formed with Multifunctional Crosslinkers,” J. Appl. Polym. Sci., Vol. 26, 1981, p. 3977. [21] Scholtens, B. J. R., Tiemersma-Thoone, G. P. J. M., and van der Linde, R., “Thermoviscoelastic and Thermoanalytic Characterization of Some Reactive Polyester Powder Coatings Systems,” Verfkroniek, Vol. 62, 1989, p. 238. [22] Higginbottom, H. P., Bowers, G. R., Grande, J. S., and Hill, L. W., “Structure/Property Studies of MF-Cured Powder Coatings,” Prog. Org. Coat., Vol. 20, 1992, p. 301. [23] Oshikubo, T., Yoshida, T., and Tanaka, S., “Studies on Acrylic Resins and Melamine Formaldehyde Resins for High Solids Coatings,” Proceedings Tenth International Conference in Organic Coatings Science and Technology, Athens, Greece, July 9–13, 1984, p. 317. [24] Scanlon, J., “The Effect of Flaws on the Elastic Properties of Vulcanizates,” J. Polym. Sci., Vol. 43, 1960, p. 501. [25] Flory, P. J., “Molecular Theory of Rubber Elasticity,” Polym. J. (Tokyo, Jpn.), Vol. 17, No. 1, 1985, p. 1. [26] Flory, P. J., and Erman, B., “Theory of Elasticity of Polymer Networks,” Macromolecules, Vol. 15, 1982, p. 800. [27] Erman, B. and Flory, P. J., “Relationships Between Stress, Strain and Molecular Constitution of Polymer Networks. Comparison of Theory and Experiments,” Macromolecules, Vol. 15, 1982, p. 806. [28] Graessley, W. W., “The Entanglement Concept in Polymer Rheology,” Advances in Polymer Science, Vol. 16, Springer-Verlag, New York, 1974, p. 1. [29] Palansiamy, A., and Rao, B. S., “Tetrafunctional Acrylates, based on β-Hydroxy Alkyl Amides as Crosslinkers for UV Curable Coatings,” Prog. Org. Coat., Vol. 56, 2006, p. 297. [30] Nash, H., Docktor, H. J. and Webster, D. C., “Effect of Composition on performance properties in Cationic UV-Curable Coatings,” J. Coat. Technol., Vol. 1, 2004, p. 153. [31] Huanyu, W., Hulguang, K., and Wenfang, S., “Thermal and Mechanical Properties of UV-Cured Acrylated Hyper-branched Polyester and Its Blends with Linear Polyurethane Acrylate,” J. Coat. Technol., Vol. 75, No. 939, 2003, p. 37. [32] Brown, R. A., Coogan, R. G., Fortier, D. G., Reeve, M. S., and Rega, J. D., “Comparing and Contrasting the Properities of Urethane/Acrylic Hybrids with Those of Corresponding Blends of Urethane Dispersions and Acrylic Emulsions,” Prog. Org. Coat., Vol. 52, 2005, p. 73. [33] Seubert, C. M., and Nichols, M. D., “Alternative Curing Methods of UV Curable Automotive Clearcoats,” Prog. Org. Coat., Vol. 49, 2004, p. 218. [34] Shalati, M. D., McBee, J. H., DeGooyer, W. J., Thys, F., van der Ven, L., and Leijzer, R. T. M. “High Performance Accelerated All-Acrylic Coatings—Kinetics and Mechanistic Aspects of Non-Isocyanate Coatings,” Prog. Org. Coat., Vol. 48, 2003, p. 236. [35] Rodriguez, M. T., Gracenea, J. J., Kudama, A. H., and Suay, J. J., “The Influence of Pigment Volume Concentration (PVC) on the Properties of an Epoxy Coating, Part I Thermal and Mechanical Properties,” Prog. Org. Coat., Vol. 50, 2004, p. 62. [36] Soucek, M. D., and Ni, H., “Nanostructureed Polyurethane Ceramer Coatings for Aircraft,” J. Coat. Technol., Vol. 74, No. 933, 2002, p. 125.
15TH EDITION
[37] Sato, K., “The Hardness of Coating Films,” Prog. Org. Coat., Vol. 8, 1980, p. 1. [38] Heijboer, J., “Dynamic Mechanical Properties and Impact Resistance,” Polym. Sci., Part C: Polym. Symp., Vol. 16, 1968, p. 3755. [39] Roller, M. B., and Gillham, J. K., “Application of Dynamic Mechanical Testing to Thermoset Coatings Research and Development,” J. Coat. Technol., Vol. 50, No. 636, 1978, p. 57. [40] Ryntz, R. A., Gunn, V. E., Zou, H., Duan, Y. L., Xiao, H. X., and Frisch, K. C., “Effect of Siloxane Modification on the Physical Attributes of an Automotive Coating,” J. Coat. Technol., Vol. 64, No. 813, 1992, p. 83. [41] Hegedus, C. R., Pulley, D. F., Spadafora, S. J., Eng, A. T., and Hirst, D. J., “A Review of Organic Coating Technology for U. S. Naval Aircraft,” J. Coat. Technol., Vol. 61, No. 778, 1989, p. 31. [42] Koleske, J. V., “Cationic Radiation Curing,” Federation Series on Coatings Technology, D. Brezinski, and T. J. Miranda, Eds., Federation of Societies for Coatings Technology, Philadelphia, 1991. [43] Koleske, J. V., “Mechanical Properties of Cationic Ultraviolet Light-Cured Cycloaliphatic Epoxide Systems,” Proceedings Radcure Europe ‘ 87, Munich, W. Germany, May 4–7, 1987. [44] Koleske, J. V., “Copolymerization and Properties of Cationic, Ultraviolet Light-Cured Cycloaliphatic Epoxide Systems,” Proceedings RadTech ‘88—N. America, New Orleans, April 24–28, 1988, p. 353. [45] Hill, L. W., “Stress Analysis: A Tool for Understanding Coating Performance,” Prog. Org. Coat., Vol. 5, 1977, p. 277. [46] Evans, R. M., and Fogel, J., “Comparison of Tensile and Morphological Properties with Abrasion Resistance of Urethane Films,” J. Coat. Technol., Vol. 49, No. 634, 1977, p. 50. [47] Gregorovich, B. V., and McGonigal, P. J., “Mechanical Properties of Coatings Needed for Good and Scratch and Mar,” Proceedings of the Advanced Coatings Technology Conference, Chicago, Nov. 3–5, 1992, ASM/ESD. [48] Sadati, M., Mohammadi, N., Qazvini, N. T., Tahmasebi, N., and Koopahi, S. “Evaluation of Scratch Resistance of an Acrylic-Melamine Clear Coat Based on Its Fracture Energy,” Prog. Org. Coat., Vol. 53, 2005, p. 23. [49] Nichols, M. E., Darr, C. A., Smith, C. A., Thoules, M. D., and Fischer, E. R., “Fracture Energy of Automotive Clearcoats—I. Experimental Methods and Mechanics,” Polym. Degrad. Stab., Vol. 60, 1998, p. 291. [50] Nichols, M. E., “Anticipating Paint Cracking: The Application of Fracture Mechanics to the Study of Paint Weathering,” J. Coat. Technol., Vol. 74, No. 924, 2002, p. 39. [51] Skaja, A., Ferando, D., and Croll, S., “Mechanical Property Changes and Degradation During Accelerated Weathering,” J. Coat. Technol., Vol. 3, 2006, p. 41. [52] Ballard, R. L., Sailer, R. A., Larson, B., and Soucek, M. D., “Fracture Toughness of Inorganic-Organic Hybrid Coatings,” J. Coat. Technol., Vol. 73, No. 913, 2001, p. 107. [53] Hergenrother, W. L., “Determination of the Molecular Weight Between Cross-links of Elastomeric Stocks by Tensile Retraction Measurements. II Polyurethanes,” J. Appl. Polym. Sci., Vol. 32, 1986, p. 3683. [54] Korhonen, M., Starck, P., and Lofgren, B., “Study of PolyesterBased Coil Coatings by Using Thermal Methods,” J. Coat. Technol., Vol. 75, No. 937, 2003, p. 67. [55] Osterhold, M., and Wagner, G., “Methods for Characterizing the Mar Resistance,” Prog. Org. Coat., Vol. 45, 2002, p. 365. [56] Spinks, G. M., Liu, Z., Brown, H., Swain, M., See, H., and Evans, E., “Paint Layer Thermomechanical Properties Determined by In Situ Dynamic Mechanical Analysis in 3-Point Bending,” Prog. Org. Coat., Vol. 49, 2004, p. 95. [57] Mafi, R., Mirabedini, S. M., Attar, M. M., and Moradian, S. “Cure Characterization of Epoxy and Polyester Clear Powder Coatings Using Differential Scanning Calorimetry (DSC) and Dynamic Mechanical Analysis (DMTA),” Prog. Org. Coat., Vol. 54, 2005, p. 164.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
50
MNL17-EB/Jan. 2012
Flexibility and Toughness John Fletcher1 and Joseph Walker2 PREFACE
IN PREPARATION OF THIS CHAPTER, THE CONtents of the fourteenth edition were extensively drawn upon. The authors acknowledge the author of the fourteenth edition, Mark P. Morse. The current edition will review and update the topics as addressed by the previous author, introduce new technology that has been developed, and include up-to-date references. This chapter is an abridged and modified version of the chapter entitled “Flexibility,” written by Garmond G. Schurr, found in earlier editions of this manual.
DEFINITIONS
For a coating to meet its service requirements it must exhibit appropriate properties of flexibility and toughness to withstand cracking when it is subjected to stresses produced by shrinking or swelling, forming, mechanical abuse, and weathering. Flexibility is the ability of a material to be bent or flexed without cracking or undergoing other failure. Toughness is the strength and resilience of a material; it is the material’s ability to withstand great strain imposed in a short time period, such as an impact, without tearing, breaking, or rupture.
INTERPRETATION
The flexibility of a coating applied to a substrate depends not only on its ability to expand by pressure from within the film, distensibility, but also on the coating thickness and on the adhesion between coating and substrate. Good adhesion tends to give better apparent flexibility than does poor adhesion. The toughness of a coating is dependent on its hardness, stiffness, resiliency, distensibility, and the existence of an energy dissipation mechanism that operates at temperatures far below room temperature and is discernable by dynamic mechanical measurements made over a broad temperature or frequency range. Generally, the bend and impact tests used to evaluate flexibility and toughness are much more severe than actual service conditions. This is because the tests are usually performed on relatively fresh, un-aged coating films. Since coating films tend to lose flexibility during use due to the loss through volatilization of free plasticizing components and chemical changes such as degradation, cross-linking, and the like, these severe tests that exceed normal operational expectations are useful in predicting long-term serviceability [1]. 1
Technical Support Manager Elcometer Instruments Ltd.UK
2
VP Sales & Marketing Elcometer Inc, Rochester Hills, Michigan
BASIC PROPERTIES AFFECTING COATING PERFORMANCE
Both flexibility and toughness depend on very basic properties: the viscoelastic behavior of the coating and its physical transitions and relaxations. The following is a discussion of these properties taken from a paper by Skrovanek and Schoff [2]. Coatings, as the polymers from which they are prepared, are viscoelastic in nature, that is, they behave both as viscous liquids and as elastic solids. The coatings have elastic recovery and yet will flow with time when placed under a stress. In general, viscoelastic behavior and mechanical properties are markedly affected when a coating enters the glass transition, softening point, or other relaxation. To be certain that the properties of a coating will fulfill the needs of its intended use, the viscoelastic behavior of the coating should be measured, controlled, and designed to meet the particular end use. The softening point of a coating can be used as an index of flexibility. The softening point is between the temperature where the coating changes from being hard and glassy and the temperature where it is leathery or rubbery. For example, if a coating has a softening temperature region near the temperature of the forming operation, the coating is less susceptible to failure by cracking or a similar mechanism than if the softening region is above the forming temperature. Measurement of energy storage (related to elasticity) and energy loss (related to viscous losses) as a function of temperature is a means of predicting impact resistance. Impact resistance of a paint film can be considered as energy dissipation by vibration or rotation of various molecular segments so that at no time will sufficient energy be focused to cause fracture. Since the impact tests performed on paint films often produce deformations beyond the elastic limit of the films, flow within the films must take place or fracture will occur [3]. To obtain good impact resistance, the paint film must consist of a polymer that has a sufficiently high molecular weight to have strong intermolecular entanglement (and therefore, high tensile strength), but sufficiently low viscosity (by choice of proper molecular constituents and limiting molecular weight) that flow and accompanying energy dissipation will take place. Polymer viscosity increases as molecular weight increases so that polymers with very high molecular weights will have greater flexibility than those polymers
637 Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed byASTM International This standard is for EDUCATIONAL USE ONLY. Copyright © 2012 by www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
638
PAINT AND COATING TESTING MANUAL
Q
with intermediate or low molecular weights. At the same time, molecular weights below the critical molecular weight for entanglement lead to very low tensile strengths and the mechanical behavior observed is brittleness. It has been found that modulus is the dominant factor in the relationship between the tensile properties of a coating and its impact resistance [4]. In addition to dynamic mechanical behavior, the relaxation behavior as measured by dissipation or damping of coatings has been determined by application of dynamic electrical tests [5]. In a dielectric relaxation test, a periodic electrical potential is applied to the sample coating situated between a pair of electrodes. The dielectric constant and dissipation factor are measured as a function of frequency and temperature.
TECHNIQUES FOR MEASURING BASIC VISCOELASTIC PROPERTIES Thermal Mechanical Analyzer (TMA)
This instrument employs transducers to sense the position of a vertical rod that rests on the surface of a coating sample. The instrument is usually equipped with a furnace and program planner so that heating, cooling, and isothermal temperature operations can be employed. With its use, softening points and glass transitions can be determined from plots of coating indentation as a function of temperature. Also, changes in stresses within a coating at a constant temperature (creep) can be determined from plots of indentation versus time [2].
Dynamic Mechanical Thermal Analyzer (DMTA)
This instrument produces vibrations in a coating film over a wide frequency range and/or temperature range. It can scan a wide range of sample temperatures at different rates. The resulting deformations from the sinusoidal applied stresses are analyzed to compute values related to energy storage and energy loss [2].
Tensile Testing
Tensile testing of organic films can measure tensile strength, modulus and elongation of the coating materials. The area under the stress-strain curve can be used as a measure of the toughness of the coating.
EXTERNAL FACTORS AFFECTING FLEXIBILITY AND TOUGHNESS
Flexibility and toughness are not constant characteristics of a specific coating. A number of external factors affect these properties.
Humidity
Water is a good plasticizer for almost all paint films in that it preserves the flexibility and adhesive power of the film. A change in relative humidity of as little as 2 % can be detected in flexibility measurements. Some paint films, such as those based on latexes, imbibe moisture very rapidly, whereas others reach equilibrium with the atmosphere very slowly. It is imperative that tests for flexibility and toughness are conducted in an atmosphere of controlled relative humidity and that the specimens are conditioned in that atmosphere for a day or more before the tests are performed. Generally, flexibility and toughness tests are carried out at a relative humidity of 50 ± 5 %. The 10 %
15TH EDITION
tolerance is needed because of the difficulty in more accurately controlling relative humidity in most laboratories. If the environment cannot be controlled at this recommended level, then the relative humidity should be measured and reported along with the mechanical properties.
Temperature
The flexibility and toughness of coatings are dependent on temperature. This is particularly true of thermoplastic coatings, but it also is a factor for thermoset coatings. These coatings have a definite second order transition temperature known as the glass transition temperature, Tg. Coatings at a temperature below Tg are hard and brittle with poor flexibility and impact resistance unless there is an auxiliary loss mechanism below Tg or below the temperature at which the coating is used, as exists in polycarbonates that have a high Tg of about 160°C (at 1 Hz) and yet have excellent impact resistance because of a relaxation that occurs at about −90°C (at 1 Hz). If coatings do not have this type loss mechanism, at temperatures just above Tg they are flexible, and at temperatures substantially above Tg they tend to develop viscous rather than elastic properties. There is a tendency for all thermoplastic coatings to have identical flexibility properties if these properties are measured at the same temperature relative to Tg, for example, at 10°C above Tg [1,6]. Flexibility and toughness measurements are usually made at a temperature of 25 ± 1°C after the coatings are conditioned at that temperature. However, there are many instances when test are performed at lower temperatures as might be encountered in cold climates.
Strain Rate
Strain rate is the rate at which a coating specimen is elongated and is usually expressed in percent per minute, in./ in./ min or cm/cm/min. This is the rate of extension relative to specimen size. That is, if a specimen 10 cm long is elongated at a rate of 1 cm/min, it is the same as a specimen 1 cm long being elongated at a rate of 0.1 cm/min (1 mm/ min). In both cases, the strain rate is 10 % per min. Strain rate has a great influence on the flexibility and toughness of a coating. In general, the effect of increasing the strain rate is similar to decreasing the coating temperature, that is, as the strain rate is increased, flexibility and toughness decrease. There can be critical strain rates where flexibility has sharp changes, which are very similar to the effects produced at the glass transition temperature [7]. This means that the strain rate used in a test must be closely controlled. In some tests, such as a bend test, this is difficult to do. This also means that tests performed at a low strain rate (cupping test) are likely to produce different flexibility ratings than those produced by a high strain rate (conical mandrel test) [1,7]. It should also be noted that the temperature and humidity of the test are significant in the results that are obtained and should be noted as part of the test result.
FLEXIBILITY AND TOUGHNESS MEASUREMENTS Mandrel Bend Tests
Both conical and cylindrical mandrels are used for evaluating the flexibility of coatings. Even though it is difficult to control the strain rate in these manually operated tests,
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 50
Q
FLEXIBILITY AND TOUGHNESS
639
relative humidity before performing the test, which is conducted under the same conditions. Film thickness will also influence the results of the test and therefore the thickness of the film should be reported along with the temperature and humidity conditions. The crack resistance value of a coating is obtained by measuring the distance from the furthest end of the crack to the small end of the mandrel. This distance is converted to cone diameter by means of a plot given in ASTM D522. The mandrel diameter at which cracking occurs is taken as the crack resistance value. If the elongation of the coating at the onset of cracking is to be reported, a bend time of 15 s is used and the diameter at which the onset of cracking occurred is converted to percent elongation from a plot given in ASTM D522 Test Methods for Mandrel Bend Test of Attached Organic Coatings.
Cylindrical Mandrel Bend Tests
Fig. 1—A conical mandrel bend tester (Reprinted by permission of Elcometer Inc.).
they can provide very useful flexibility ratings and are often used to evaluate coatings applied to metal test panels.
Conical Mandrel Tests
A conical mandrel test consists of manually bending a coated metal panel over a cone. As described in ASTM D522, Standard Test Methods for Mandrel Bend Test of Attached Organic Coatings, Test Method A, a conical mandrel tester consists of a metal cone, a rotating panel bending arm, and panel clamps. These items are all mounted on a metal base as illustrated in Fig. 1. The cone is smooth steel 8 in. (203 mm) in length with a diameter of 1/8-in (3.2 mm) at one end and a diameter of 1.5 in. (38 mm) at the other end. When a coating is applied on a 1/32-in (0.8 mm)thickcold-rolled steel panel, as specified in ASTM D522, a bend over the mandrel produces an elongation of 3 % at the large end of the cone and of 30 % at the small end of the cone. The coated panel is bent 135° around the cone in approximately 1 s to obtain a crack resistance rating under simulated abuse conditions. In some instances, longer bend times have been found to be useful. For example, if the percent elongation of the coating at the point of cracking is to be determined, the method specifies a bend time of 15 s. ASTM D6905, Standard Test Method for Impact Flexibility of Organic Coatings, describes a procedure for determining the ability of a coating and its substrate to resist shattering, cracking, or chipping when the film and the substrate are distended beyond their original form by impact. This test method references D522 but states that D6905 is perhaps a better way to apply the test using a rapid application of the force. Since the impact is almost instantaneous the problems associated with time variables in the mandrel tests are eliminated. Since variations in temperature and humidity can affect mandrel bend tests, it is imperative that the coated panels be conditioned at a standard temperature and
When executing cylindrical mandrel flexibility tests, a coated panel is bent manually over one or more cylindrical rods or surfaces of different diameters. ASTM D522 Method B states that the testing device should include mandrels with 1-in. (25 mm), 3/4-in. (19 mm), 1/2 -in. (12.7 mm), 3/8 -in. (9.5 mm), 1/4-in. (6.4 mm), and 1/8-in (3.2 mm) diameters. Examples of cylindrical mandrel testers are given in Figs 2 and 3. The panel should be bent over a mandrel with the uncoated side of the panel in contact with the mandrel surface. The panel should be bent approximately 180° around the mandrel at a uniform velocity in a time of 1 s. If cracking has not occurred, the procedure is repeated using successively smaller and smaller diameter mandrels until cracking is apparent. The cracking-resistance value of a coated panel is the minimum diameter at which cracking does not appear.
Fig. 2—A cylindrical mandrel bend tester (Reprinted by permission of Elcometer Inc.).
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
640
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 4—T-Bend test bending the coated sheet on itself (ASTM D4145: Standard Test Method for Coating Flexibility of Prepainted Sheet). Fig. 3—Simple cylindrical mandrels test apparatus. (Reprinted by permission of Elcometer Inc.)
This testing procedure can be applied as a “pass/fail” test by determining whether cracking is produced by bending over a specified mandrel diameter. A table for converting mandrel diameter to percent elongation is given in ASTM D522. Schuh and Theuerer [8] derived the relationship between the diameter of a mandrel and the elongation of a coating to be as follows: Persent Elongation = 100(t/(2r + r))
(1)
where t is the thickness of the coated panel and r is the radius of the mandrel. Observed elongations are greater than values calculated from the above expression and vary with different types of metal substrates. Crack resistance of a coating is dependent on its thickness. That is, the thicker the film, the lower the crack resistance. Values of crack resistance obtained by the mandrel bend tests should be corrected for film thickness when comparisons are made between different coatings. ASTM D522 contains corrections to be added to elongation values obtained with coatings having a thickness greater than 1 mil (0.03 mm) when applied to 1/8-in (0.8-mm)-thick steel panels. Conical mandrel bend test procedures, similar to those given in ASTM D522, are found in ISO 6860 and BS 3900 E11. Cylindrical mandrel bend test procedures, similar to those given in ASTM D522, are found in ISO 1519, DIN 35 152, and BS 3900 E1.
repeated successively to produce a 1T (pronounced one T), 2T, 3T, etc. bends. These successive bends result in two, three, etc. thickness of the metal around the first bend. It should be apparent that the greater the number of thicknesses around which the coated metal is bent, the less severe the test. Test results are reported as passing the smallest T-bend on which cracks are observed. In some cases, cracking must be detected by removal of the coating using a pressuresensitive tape placed on the bend edges and observing the degree of removed coating particles. ASTM D4145, Test Method for Coating Flexibility of Prepainted Sheet, describes this test procedure. A triangular-shaped specimen has been found to be convenient for making T-bend tests as this shape leaves a portion of each bend exposed for later examination and for permanent record. Alternatively the T-bend test described in D4145 can be carried out using a die around which the deformation is
T-Bend Tests
T-bend tests are a means of evaluating the flexibility of coated strip metal that is to be formed during a fabrication process (Fig. 4). Multiple 180° bends of the coated metal are made, and the amount of cracking produced at each bend is visually determined. Ratings are classified as 0T, 1T, 2T, 3T, and so on. The 0T (pronounced zero T) bend consists of making a 180° bend with the paint on the outside of the bend and pressing the bend flat so there is no space between the metal surfaces. This operation is
Fig. 5—T-Bend test using a die around which the specimen is bent (ASTM D4145: Standard Test Method for Coating Flexibility of Prepainted Sheet).
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 50
Q
FLEXIBILITY AND TOUGHNESS
641
can produce a sufficiently high pressure for pressing a 1 5/8-in (41-mm)-diameter indenter ball into the coated metal is provided. The rate of forming can be adjusted over a range of 0.2 to 1.0 in./min (4.8 to 25 mm/min). A dial gage monitors the movement of the indenting ball. Adhesive tape is applied over the dome formed in the metal, and the tape is rapidly removed. The amount of coating removed is given a rating by comparing it with a set of photographic standards.
Impact Resistance Tests
Fig. 6—Motorized Cupping Tester fitted with a web-cam to record the onset of cracking in the coating. (Reprinted by permission of Elcometer Inc.)
made (Fig. 5). Progressively larger dies are used to progress the test to the point where no damage to the coating results from the bending.
Cupping Tests
A test using a relatively slow rate of forming can be conducted with a Cupping Tester (Fig. 6) that pushes a punch into the unpainted side of a coated panel until the increasing deformation produces cracks in the coating. Test procedures are given in ISO TC 35, BS 3900 E4, NFT 30–019, SIS 18 41 77, DIN 50 101, and DIN 50 102. Cupping testers can be used to simulate different forming operations. The motorized Cupping Tester uses a spherical punch and a range of cupping speeds can be achieved. The maximum cupping depth is approximately 18 mm (0.7 in.). The cupping action is stopped when cracking in the coating is detected visually. The depth of cupping at that point is indicated on a dial gage and is considered to be the flexibility rating. The cupping tester can be equipped with a magnifier, a stereomicroscope, or a web-cam for observing the onset of cracking.
Forming Tests
In many industrial operations, metal is coated flat and then formed into various shapes by drawing the coated metal. This can be simulated directly or by elongating a coated metal sheet. Any tension-testing instrument capable of rapidly elongating a metal strip can be used for determining the ability of a coating to be drawn (drawability). A coated metal strip is elongated at a high rate of strain until cracking occurs. The elongation can be measured with an extensiometer [1]. Drawability would be reported as the percent elongation obtained just before cracking is observed. Since elongation is rate dependent, the rate of elongation used in the test should be reported. ASTM D4146 Test Method for Formability of ZincRich Primer/Chromate Complex Coatings on Steel provides a procedure for determining the formability of coated strip metal. An outline of a testing machine that
The most commonly used impact testers drop a weight onto an indenter resting on the surface of a coated panel that is resting on a platform (Fig. 7). A die in an opening in the platform allows the panel to be pushed down by the indenter to form a dimple in the panel. The weight is dropped through a guiding tube whose height is marked in increments. There are a number of possible combinations of weights, indenter sizes, die sizes, and weight heights that can be used in performing impact tests. The tests can be performed by impacting either the coating directly (coating facing upward) or indirectly (coating facing downward). Cracking observed on or around the impact-produced dimple is considered failure, and the force to produce the cracking is given in inch-pounds (kilograms-meters), that is, weight times height. The test can be performed either to determine the inch-pounds required to produce cracking or to determine whether a coating passes or fails at a specified inch-pound value. ASTM D2794, Standard Test Method for Resistance of Organic Coatings to the Effects of Rapid Deformation (Impact), describes such a test procedure and offers three procedures for determining the degree of cracking produced in an impact deformation: (a) visual inspection with a magnifier, (b) visual inspection after application of an acidified copper sulfate solution, and (c) use of a pin-hole detector. The General Electric Impact Flexibility Tool is used for simultaneously making several indentations of different sizes. From these indentations, conclusions can be made regarding crack resistance and the amount of draw that a coating applied to sheet metal can tolerate. This tester consists of a steel cylinder that has knobs (segments of spheres) of different radii machined on each end. The cylinder is dropped onto a coated panel that is supported coating side down by a rubber pad. The height of drop is adjusted so that the boundary of the cylinder is just discernible. This procedure assures that each knob is used to its limit. Eight knobs cover a range of 0.5 to 60 % elongation. See Federal Test Method Standard 141C, Method 6226. ASTM G14, Standard Test Method for Impact Resistance of Pipeline Coatings (Falling Weight Test), describes a test procedure for determining the impact resistance of pipe coatings. A fixed weight of 3.0 lb (1.36 kg) and having a 5/8-in. nose diameter is dropped through a guide tube onto a coated pipe specimen. The height of the weight is adjusted until the minimum height at which cracking appears is attained. A wet-sponge pinhole detector is used to determine the presence of cracks in the coating on the impacted pipe. An equation is given for calculating the impact resistance from the weight and its height of drop required to just produce cracking.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
642
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
A different type of impact tester was developed and is being used at the Bell Laboratories of AT&T. A coated panel is subjected to repeat glancing blows by a casehardened steel ball at the end of a short arm that is pivoted to another arm connected to a rotating shaft. During the test, the coated panel is mounted on a platform that moves so that successive blows do not strike the same spot. The energy level of the blows may be held constant, as in a “pass/fail” test, or it can be adjusted by changing the speed of the rotating shaft to determine the onset of cracking. If the hammer energy level required to destroy the coating is desired, a transparent, calibrated scale of shaft speed in revolutions is superimposed over the impact pattern. This tester is not commercially available. There are a number of other impact testers that have been developed over the years and used to some extent. These include the Parlin-duPont Tester, Camp Impact Test, Hart Impact Tester, Ball Punch, General Electric Ball Drop, and Navy Falling Ball test. None of these testers are commercially available now. The ISO 6272 standard, Rapid-deformation (impact resistance) tests is made up of two parts, part 1, Fallingweight test, large-area indenter and part 2 Falling-weight test, small-area indenter. Part 1 uses a 1 kg weight with the 20 mm diameter spherical indenter attached, whereas part 2 is based on ASTM D2794-93 (2010) and uses either a 12.7 mm or 15.9 mm indenter.
Testing of Free Films
ASTM D2370, the Standard Test Method for Tensile Properties of Organic Coatings, covers the elongation, tensile strength and stiffness (modulus of elasticity) of organic coatings when tested as free films. The method for preparing the uniform free films of organic coatings is given in ASTM D4708, Standard Practice for Preparation of Uniform Free Films of Organic Coatings using substrates that can be peeled from the cured film. Procedures are given for preparing free films on three alternative substrates. These substrates are treated FEP (fluorinated ethylene-propylene) sheet, silicone coated paper, and halo-silane coated glass plates. (Note: The dental foil that used to be included as one of the substrate has been removed from this practice because it contains mercury and is therefore hazardous.)
Cold Crack Resistance Tests
Fig. 7—Falling weight impact tester. (Reprinted by permission of Elcometer Inc.)
Tests in which coatings on substrates are cycled through elevated temperature, low temperature, and room temperature environments are called cold crack tests. They have been used in the coatings industry for many years as an indication of the ability of a coating to resist cracking in service and therefore are considered to be tests of coating flexibility. ASTM D1211 Standard Test Method for TemperatureChange Resistance of Clear Nitrocellulose Lacquer Films Applied to Wood is an example of such a cold crack test, however this Standard Test Method was withdrawn without a replacement in 2006. It described a procedure for testing lacquer coatings applied on wood. The testing cycles consist of 1 h at 120°F (49°C), 1 h at −5°F (−21°C), and ½ h at room temperature. The results were reported as the number of cycles required to produce visible cracking in the coating.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 50
Automotive coatings are subjected to cold crack cycle tests. A typical test for exterior coatings on metal panels consists of (1) equilibration at room temperature, (2) exposure in a humidity cabinet at 100°F (38°C) and 100 % relative humidity for 20 h, and (3) exposure in a freezer at −22°F (−30°C) for 4 h. After removal from the freezer, the coated panels are allowed to stand at room temperature for 2 h. Then the coatings are visually rated for cracking. In some cases, an exposure of 2 h in an oven at 150°F (65°C) is introduced into the above cycle conditions. A rapid cold crack test that has been developed is based on the use of cooled air entering a transparent, insulated box. Cold air that has been cooled at a rapid controlled rate is introduced into the box, and the coatings are observed for cracking. The coatings are then rated by determining the temperature decrease from room temperature that is required to produce visual cracks in the coatings [4].
Effects of Aging and Weathering
The ultimate measure of satisfactory flexibility and toughness of a coating applied to a substrate is performance under service conditions. Most flexibility and toughness tests are performed on relatively fresh-coated panels, that is, tests are usually performed after the panels have been conditioned in a specified atmosphere for a specified period between 24 h and 7 days. The results obtained are applicable to service conditions if these are concerned with post forming or service indoors without a degrading atmosphere, since most coatings do not change appreciably in their physical service properties under such conditions.
Q
FLEXIBILITY AND TOUGHNESS
643
However, if the service conditions include exposure to weathering, this factor can cause appreciable changes to occur in the coatings properties. The effects of moisture, temperature changes, and exposure to sunlight (ultraviolet wave lengths) encountered in outdoor exposure generally reduce the flexibility and toughness of organic coatings. Therefore, it often is desirable to conduct tests for flexibility and toughness after periods of weathering to determine how a coating will perform under actual weather conditions [1].
References [1] Schurr, G. G., “Flexibility,” Paint Testing Manual, ASTM STP 500, 13th ed., H. A. Gardner and G. G. Sward, Eds., ASTM International, West Conshohocken, PA, 1972, pp. 333–337. [2] Skrovanek, D. J., and Schoff, C. K., “Mechanical Analysis of Organic Coatings,” Prog. Org. Coat., Vol. 16, 1988, pp. 135–163. [3] Moore, R. J., “Molecular Basis for Impact Resistance of Epoxy Paint Films,” J. Paint Technol., Vol. 43, No. 554, March 1971, pp. 39–46. [4] Morse, M. P., “Physical Properties of Paint Films Relating to Service,” presented at Gordon Research Conferences, Organic Coatings Section, 15–19 Aug. 1955. [5] Varadarajan, K., “Review of Dielectric and Dynamic Mechanical Relaxation Techniques for the Characterization of Organic Coatings,” J. Coat. Technol., Vol. 55, No. 704, September 1983, pp. 95–104. [6] Tordella, J. P., “Mechanical Properties of Amorphous Polymers,” Official Digest, Vol. 37, 1965, p. 349. [7] Supnik, R. H., “Rate Sensitivity: Its Measurement and Significance,” Mater. Res. Stand., Vol. 2, 1962, p. 498. [8] Schuh, A. E., and Theuerer, H. C., “Measurement of Distensibility of Organic Finishes,” Ind. Eng. Chem., Vol. 9, 1937, p. 9.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
51
MNL17-EB/Jan. 2012
Understanding Osmotic Activity in Paint Films and Determining Cause by Systematic Analysis of Blister Fluids and Blistered Coatings George Mills1
INTRODUCTION
A PROBLEM OFTEN OBSERVED WITH PAINT FILMS is the development of domed blisters. Usually these “bubbles” in the paint film are the result of osmotic activity. Understanding the osmotic process will be very helpful in the investigation of the cause(s) or driving force with this mode of paint failure. Osmotic blisters may develop between the primer paint and base substrate, between coats of applied paints, or within the matrix of a single layer of the composite (single or multi-coat) paint film. Although in most cases the blisters are easily seen, it is possible to experience osmotic activity under paint films when a domed blister is not elevated enough to be visible. When this occurs on a steel substrate, typically, corrosion will develop under the film. The osmotic process may be represented by the following graphic. It is most important to appreciate that potential solvent molecules will migrate through the paint film “one molecule at a time.” One will remember that a collection of single, individual molecules is defined as a gas. For this reason, osmosis follows, and is controlled by, the general gas laws. A common “definition” of osmotic activity in paint films describes the migration of solvent molecules through the film and dissolving one or more solutes forming the blister fluid solution. This process then continues in an effort to dilute the blister fluid solution to the point that it has the same concentration (actually, same vapor pressure) as the outer contact solvent liquid. It is this differential in vapor pressures that drive the osmotic activity. Osmosis involves a semi-permeable membrane (as with a paint film), a solvent, and one or more solutes. To understand the potential interventions and investigation of the driving forces, knowledge of the basics of osmosis are extremely useful. Within this chapter we will review the chemistry and physics controlling the osmotic phenomena and present the analytical techniques that will allow one to determine the source, or driving force, of the osmotic activity. After determining the driving force one then has the ability to modify the paint formulation, application procedures, surface preparation, or service restrictions if any of these are shown to be the cause (source). If a coating failure has occurred one may determine where the
1
liabilities lie as well as the extent of remediation that will be required to allow a true long-term fix of the problem. Problems associated with faulty application such as improper surface preparation are frequently claimed as the culprit. But, the inclusion of a water soluble thinner or chemical reaction under, or within, the coating matrix that produce water soluble materials are possible. Examples of some of these will be discussed later. Other potential sources of osmotic problems include deficiencies in the paint specification and unanticipated service conditions. Paint formulations that include pigments, resins, or additives that are soluble, or convert to a form that is soluble in service, can cause osmotic problems. On rare occasions manufacturing anomalies may lead to osmotic activity. The investigation of osmotic activity in paint films should involve the systematic collection of samples from the site along with attentive observation of the blister location, size, and shape. If within a tank, note if blister predominance is in the liquid or vapor phase of the tank (or at the vapor/liquid interface). Are there temperature differentials that would cause condensation? Data relating to cargo history may be needed in tank lining failures to recognize and explain unusual solutes found in the blister fluid and/ or coating dome analysis. The collected samples should include blister fluid withdrawn from blisters, recovered paint chip film blister domes, paint film samples collected away from blistered areas, and, for internal tank lining failures within non-aqueous tank lining failures, a sample of the material within the tank leading to the osmotic problem. Sometimes, solutes found in the blister fluid are the result of chemical reaction products. These reaction products may be generated in the blister fluid over time or generated within the coating matrix itself producing a solute in situ. Examples of these include the highly alkaline fluid on an underground pipeline or ship’s ballast tank caused by the reduction of oxygen in systems under cathodic protection (CP). This high pH fluid may hydrolyze bound chlorine from an epoxy resin or bromide from the blue-green pigment in the fusion bonded epoxy (FBE). Tank linings exposed to chlorinated solvents will experience generation of hydrochloric acid (HCl) from hydrolysis of the material within the coating film. Hydrolysis of acetate solvents
George Mills and Associates International, Inc., Foley, AL 36535
644
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 This standard is for EDUCATIONAL USE ONLY. www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
Downloaded/printed byASTM International Copyright © 2012 by
CHAPTER 51
Q
UNDERSTANDING OSMOTIC ACTIVITY IN PAINT FILMS
Fig. 1—Osmotic blisters can form at various interfaces. This includes under the initial primer coat at the substrate or between intermediate coats. Individual molecules of the solvent migrate through the polymer forming the paint film. Some of these molecules of solvent coalesces to a liquid forming the blister and disolves one or more solutes. The difference in vapor pressures between the fluid in the blister and the migrating solvent controls the direction of solvent flow. (Graphics done by George Mills, Jr.; Building Design Solutions, LLC.)
generates acetic acid. When blister fluids are low pH, often there will be an iron corrosion product as a solute as well. These will be discussed later (Fig. 1). Chemical analysis of the blister fluid along with thermal out-gassing of the blister dome film, usually, will define the cause of the osmotic activity. When “strange” or unexpected solutes are seen, the investigator may be required to review some of the chemical reactions that could produce these solutes in the blister fluid. While the solvent is often water, tank linings and internal pipe coatings in non-aqueous service may experience osmotic activity wherein the solvent of the osmotic blister is the product of the contact service media. The solutes may be occluded previous cargo materials or lower molecular weight polymer fractions generated by the cargo material. These will be discussed in detail later. Most domed paint blisters will be present with the liquid that caused the blister, although some blister domes may be found after the liquid has escaped from the coating. The disruption in the paint film often leads to severe corrosion of the substrate as seen in Fig. 2 within a ballast tank of a large ocean going tanker.
Fig. 2—Osmotic blisters in the ballast tank of a large marine tanker. Note that many blisters have broken open and corrosion of the substrate is substantial. Some of these may be driven by cathodic protection anodes installed in the tanks that are operative only when the tank is with ballast water.
645
As discussed previously the cause (driving force) of osmotic blister formation may originate as a result of numerous sources. While the applicator often is accused of improper surface preparation, the service conditions encountered and the paint’s formulation may be responsible. The real driving force, and hopefully a correction in the future, may be determined by systematically analyzing the blister fluids and understanding some of the complex chemistry that may be operative. The physical development of osmotic blisters in a paint film requires that a solvent be available having marginal transmission within the paint film. This solvent will dissolve one or more solutes under the film at the substrate, between applied coats, or within the matrix of a single layer of the paint film. Usually the source of the solvent is not a mystery. Typically, it is water from condensation or immersion when water contact is possible. For tanks in non-aqueous service, the linings may experience osmotic blister formation wherein the materials in the tank constitute the solvent. For these situations, the contact solvent will be known or discovered upon chemical analysis. The identification of the solutes, and the source of these solutes, may present the investigator with a challenge. While in some cases, salts over-coated during application have been cited as the driving force, the real picture, usually is considerably more complicated. The solutes may involve the generation of ions, with time, under the film. A good example is the osmotic blisters often discovered in the lining of potable water tanks at the “one year” guarantee inspection after use of hypochlorite as the purification agent. Through multiple chemical steps, hydrochloric acid is generated at the steel interface leading to iron chloride salts under the blisters. Fig. 3 depicts the blisters formed in such a way within the potable water tank onboard a very large passenger cruise ship. The blister fluid may have become strongly acid or strongly alkaline by generating an acid or alkali under the film after gas phase transmission of the acid or alkaline precursors. Acids and alkalis are both strong ionic solutes and may drive osmotic blister enlargement once initiated. Water soluble solvents included in the paint formulation by the manufacturer, or added at application, are another source of solutes in which there is no additional ionic activity.
Fig. 3—Osmotic blisters in a potable water tank within a large cruise ship. Notice rust staining coming from broken blisters. Often these initiate from water soluble solvents formulated into the epoxy or added during application. Analysis of the blister fluid will indicate the source of the osmotic blister formation.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
646
PAINT AND COATING TESTING MANUAL
OSMOSIS IN PAINT FILMS
Q
When a practitioner is asked to define an “Osmotic paint failure,” the typical response is “… the passage of water through a semi-permeable membrane to dissolve salts under the paint thereby forming blisters under the film.” While this might describe a portion of such failures, the picture is much broader than this description might imply. The term “semi-permeable” is in common usage although, often, there is little, or no, thought as to the considerable significance of the numerical value of the “semi-” in this common definition. This is true as it relates to the material acting as the solvent as well as the materials acting as the solutes driving the osmotic activity. It is important to appreciate that this variability in migration rate/ability through the paint film applies to both the solvent and (the often neglected) solute(s), when these solutes are water-soluble solvents included in the paint formulation, added during application, or generated in service. In this chapter on methods of determining the source of osmotic activity in paint films, the terms “blister fluid” and “blister solution” represent the mixture of the solvent and one or more solutes found in the blister. By convention, the material present in this solution in the largest amount is considered to be the solvent. Generally speaking, the value of the vapor pressure of the solvent in the blister fluid solution as compared to the vapor pressure of the solvent outside the blistered film dictates the direction of the solvent flow. One must remember that a solvent’s vapor pressure will always be depressed when the solvent has any solutes added. In the production of pickles, the cucumber is soaked in a strong salt brine solution wherein the water of the brine has a lower vapor pressure than the water inside the cucumber. In this case, the water flows out of the cucumber causing it to shrivel to a much smaller size. Within the blister fluid, all of the dissolved materials are considered as the solutes and will act to depress the solvent’s vapor pressure. Many only think of salts, over-coated at the time of application or included in the formulation, as potential solutes and water from either surface condensation or immersion service as the solvent. While water soluble salts are a common and strong osmotic driving force, water soluble solvents, such as oxygenated alcohols, have been shown to present a serious osmotic driving force. Oxygenated alcohols are alcohols (already containing at least one oxygen atom) that contain additional oxygen in the form of ethers or ketones. These materials are excellent solvents for the epoxy coatings as they cure and may be employed in the paint formulation, but they are typically not very volatile and remain within the coating matrix for long periods of time. Their added oxygen in the molecule as ether linkages (–COC–) usually render them water soluble. Their water solubility also aides in displacing moisture at the interface during application over damp surfaces. Unfortunately, since most are totally water soluble, they can become the osmotic driving force when liquid water condenses on, or comes into contact with the coating, especially when insufficient time is available to allow escape during the drying stage [1]. Water soluble reaction products such as acetic acid remaining after hydrolysis of acetate solvents or microbial digestion of sugars have also been seen as solutes. Acid solutes usually cause additional corrosion of steel substrates under coatings with the initially formed iron salts
15TH EDITION
becoming solutes in the osmotic process. The ferrous (Fe+2) corrosion product soon converts to an insoluble ferric (Fe+3) corrosion product after reaction with oxygen in the air. If the adhesion is good and the blister dome breaks from internal pressure, osmosis cannot continue at that blister location since osmosis requires movement of one molecule at a time. There are a number of factors that dictate the degree and speed of damage to a coating system that is experiencing osmotic activity. This communication is intended to present some important aspects useful in understanding the process of osmosis, its damaging effects upon paint films, and a methodology that has proven useful in the systematic elucidation of the source of the osmotic driving force. Osmotic activity in paint films is a very common failure mode. This chapter is intended to assist the coatings formulator, the application contractor, the failure analysis practitioner, and the end user in understanding the operative driving forces and demonstrate methodology useful in determining the source of the osmotic activity after coating failure. Osmotic activity in paint films will be addressed in the following steps: (1) discuss the chemistry/energetics of the osmotic process and other chemical processes that may accompany and exacerbate this coating failure mode; (2) discuss some of those factors causing, or accelerating, osmotic activity in paint films; (3) discuss a proven methodology employed in the systematic chemical analysis of the osmotic blister fluids to identify the source of the activity and the significance of the analytical data as it relates to potential problems with the paint formulation, the specific application, and/or the service leading to the paint failure; and (4) present some case history examples to demonstrate the data developed in identifying the source of the osmotic activity. Water is the most common solvent encountered in paint film osmotic activity. For this reason, water systems, primarily, will be discussed in this communication. It is important to appreciate that other solvent/solute systems may be encountered and are just as damaging. A common example exists with marine epoxy tank linings containing benzyl alcohol or nonyl-phenols and cargo materials having marginal transmission coefficients such as the fuel additive methyl-t-butyl ether (MTBE) and some lower molecular weight alcohols sequenced with the presence of water.
THE CHEMISTRY OF THE OSMOTIC PROCESS
In simpler terms, osmosis is the natural process of individual solvent molecules migrating through the film, condensing, and dissolving some solute to form a solution. For osmosis to be possible, molecules of solvent MUST travel through the paint film ONE MOLECULE AT A TIME. By definition, this makes the process a “Gas Phase Transition” since a “gas” is a collection of individual molecules. For this reason osmotic activity follows the general gas laws of chemistry. It is more understandable to say that the vapor pressure of the pure solvent (outside the coating film) is greater than the vapor pressure of the diluted solvent in the blister fluid solution. This difference in vapor pressure thereby forces the solvent towards the osmotic solution. This is due to the depression in vapor pressure of the blister fluid caused by
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 51
Q
UNDERSTANDING OSMOTIC ACTIVITY IN PAINT FILMS
the solutes present as opposed to the pure solvent. Just as dissolved solutes will depress the freezing point, or elevate the boiling point of a solvent (by depressing its vapor pressure), the energy required to perform the chemical work, or, simply, the osmotic driving force, has its origins in the similar basic laws of chemistry [2]. This then leads one to the more common “definition” of osmotic activity. The solvent molecules migrate through the film and dissolve into the blister fluid solution in an effort to dilute that solution to the point that it is the same concentration (same vapor pressure) as the outer contact liquid. More precisely, the process continues until the vapor pressure of the blister fluid solution equals the partial vapor pressure of the solvent outside the blister. (The physical chemistry purist will speak of equating the chemical potential of the two.) Because osmotic activity is a “colligative property” of the solution, the driving force is based on the number of individual species dissolved in the solution rather than the individual solute’s molecular weight, size, or chemical structure. In general, this means that the effect of one mole (90 g) of 2-butoxy ethanol/liter has the same damaging effect as a half “mole” (about 29 g) of sodium chloride/liter since the sodium and the chloride ions are each separated in the osmotic solution and the total sum of the number of individual species dissolved (ions) are equal to the number of the 2-butoxy ethanol molecules in solution. Individual movement of molecules of the solvent through the paint film is an important concept to appreciate. By definition, this is a gas phase transition since a gas, also, exists as individual molecules. For osmotic blisters to form in the first place there must be sufficiently compromised adhesion for a condensed (liquid) phase to exist, even if over an extremely minuscule area. Once a liquid solution is present, at an interface (base substrate, inter-coat interface, particle inclusion, etc.), the volume and, usually, the pressure of the condensing fluid will continue to increase until two conditions exist collectively. In osmotic activity, the first condition is that the concentration of the solute within the blister fluid decreases as the volume of pure solvent passing through the film increases. When this concentration of driving solute in the blister fluid equals that in the external contact fluid, the process will stop. More precisely stated, the osmotic driving force diminishes as the partial pressure (chemical potential) of the solution in the blister and the partial pressure of the liquid outside the film approach the same value. The second condition occurs when the actual pressure that is gradually created in the osmotic blister reaches the value that will force the pure solvent (or solutes) back through the film. This is the process of reverse osmosis used to purify water in certain desalination systems. When the adhesion of the paint film is good and the film cannot be peeled easily from the substrate by the internal osmotic pressure generated, the pressure may become sufficient to rupture the blister. When this occurs, if the crack remains open and liquid solvent can pass, osmotic activity is no longer possible. This is because liquid phase movements (as opposed to gas phase) are now operative since there is no semi-permeable membrane. Osmosis requires that individual molecules (i.e., separated from each other) move through the semi-permeable membrane (paint film) thus acting as a vapor phase transmission.
FACTORS CAUSING VARIATION IN OSMOTIC ACTIVITY IN PAINT FILMS
647
While most osmotic activity in paint films is clearly visible, osmotic activity can, and does, occur without the appearance of visible domed blisters. The development of osmotic blisters at the coating/substrate interface, whether visible or not, depends on a large number of factors. A few of these include: (1) the value of the paint’s adhesion force at that location; (2) the degree of the depression of the glass transition temperature, Tg, and/or plasticity of the paint film caused by the water (or other solvent) and any migrating solutes; (3) the concentration, and location, of available solutes; (4) the transmission coefficients of the solvent, and solutes, within the paint film at that location and under the local ambient cure conditions. For coatings like epoxies, the actual variation in transmission rates are possible within a single formulation and depend on numerous application variables. Solvent and molecular solute transmission coefficients within a paint film will depend on variations in the degree of mixing prior to application, closeness to true mixed chemical stoichiometry of the base and converter resins, temperature of the substrate (and, therefore, the reacting paint mixture) during curing, ultimate cross-link density of the formulation, amount of carbon dioxide (CO2) stripped from the air by the amino functional converter during mixing and application, and a myriad of other variables. Although rare, osmotic activity can occur within the matrix of the paint film as opposed to an interface. When this happens, bubbles of liquid will form within the coating matrix, usually from condensation of water (or the solvent) on the surface of some occluded particle. This requires the osmotic pressure to develop to a value greater than the tensile strength of the paint matrix at the location of the formed bubble. Osmotic activity in this form is typically present as very small bubbles within the paint film. The bubbles are not spherical in nature, and they are best viewed under a microscope. It is important to emphasize that osmotically driven bubbles have nothing to do with any air or gas bubbles or voids that may have been trapped during application and are present in the applied film [3]. Osmotically generated voids or bubbles within the coating matrix will show, microscopically, a fractured edge as opposed to a smooth surface for air trapped bubbles. The size of the bubbles formed from osmotic activity that occur within the coating matrix depends largely on the tensile strength of the coating matrix at the inclusion and the distribution of water soluble inclusions, either salts, pigments, or converter fractions. Water soluble low molecular weight amine converters have been seen in blister fluid when application occurred at cold temperatures. Typically, these are usually at an interface as opposed to within the film matrix and occur as a result of reaction of the alkaline amine functional cross-linker and the acid gas carbon dioxide. From the above discussion, it is evident that for osmotic activity to occur at an interface where adequate solute is available, the value (strength) of adhesion is the major factor that will dictate the size of the blisters formed. When blisters form, areas having very good adhesion will yield small blisters. Areas having compromised or lower values of adhesion will yield larger blisters. Very large blisters (16 in. diameter and greater) have been observed when CP produced alkali salts in situ under FBE pipe coatings.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
648
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
2.
3. Fig. 4—Blisters as large as 20 in. have been seen on cathodically protected pipelines. The fluid from the blister on the FBE coated pipe above was highly alkaline from the reduced oxygen production of hydroxide.
These coatings were applied over salt contaminated areas of a pipe and developed within a few years time. This very common example typically is the result of applied cathodic protection and the production of hydroxide ion (salts) from reduction of oxygen gas passing through the coating and dissolving into the blister fluid. See Fig. 4. All amorphous (i.e., non-crystalline) organic coatings and plastics will allow small molecules to migrate within, or diffuse through, the plastic matrix. This is the same mechanism that allows a freshly applied coating, after gelling, to lose most of its included solvents as it cures and dries. The rate, or speed, of passage of these individual molecules through a coating film is called its transmission coefficient for that material and depends on numerous factors. These include the solute size and molecular weight. Ficks Law dictates that the diffusion rate through a polymer matrix is inversely proportional to the square root of the molecular weight of the migrating molecule. Diffusion rate is also impacted by the cross-link density of the film. The transmission coefficient of migrating molecules is often related to the glass transition temperature, Tg. Migration rates are also impacted by the chemical and structural nature of the matrix as well as the migrating specie’s polarity, polarizability, and chemical stability. This migration within a coating matrix is normal and to be expected, although sometimes it is undesirable. When sufficient volumes are lost after reaching the gelled state in its curing process, cracking of the coating from shrinkage stress development may occur. No coating stresses can develop from volume loss before the gel point is reached since liquid mechanics are operative within the freshly applied paint. The inclusion of these migrating species within a coating matrix causes a decrease in the cured film’s glass transition temperature, Tg. This dynamic, constant movement within the coating film is the same as experienced with migrating plasticizers, chemically blocked cross-link accelerators/enhancers, molecular catalysts, as well as binder extenders such as benzyl alcohol, nonyl-phenols, etc. The past experience of the author has shown that the water-soluble solutes causing osmotic blisters can come from a number of sources. Some typical sources include the following: 1. Water-soluble solvents purposefully formulated into the paint by the manufacturer. The most common are
4.
oxygenated alcohols such as 1-methoxy-2-propanol and 2-butoxy ethanol. Some are listed in the product’s material safety data sheets (MSDS) while some paint manufacturers are reluctant to identify their presence. Water-soluble solvents added at application (thinners). While the paint or the manufacturer’s recommended thinners may not contain the oxygenated alcohols, the manufacturer’s recommended wash thinners may contain these solvents. These have been detected in applied paint films in past blister fluid analyses, apparently from inadvertent inclusion by the applicator when the wash thinner was used to thin the paint at application. Salts remaining on the substrate at the time of application. All water-soluble salts, whether sodium chloride, sodium/potassium/lithium hydroxide, or whatever the ionizable water-soluble salts may be, these can become solutes in the osmotic blister fluid. Most water based zinc silicate formulations use a binder that originates as a water glass. This is a silicate dissolved by alkali hydroxide. Shop primers often are not formulated above the critical pigment volume concentration (CPVC) and allow the hydroxide salts to become trapped within the matrix of the shop-primer coating. These are slowly released when used in immersion or damp, water-condensing service. Usually, efforts are made to water rinse any soluble salts from the primers prior to over-coating but with systems formulated below the CPVC, this cannot be done effectively. Even with systems formulated above the CPVC, extended curing times are always required to minimize the availability of the soluble alkali hydroxides. Rinsing should continue until the dampened surface has a pH equal to the rinse water. The lack of porosity makes this task difficult (or impossible). In these cases, the blister fluid will be of a high pH, high conductivity, and the ion analysis will indicate the presence of the alkali metal (Na, K, or Li) from the water based zinc silicate primer. Ethyl silicate based zinc primers are not as prone to the blistering in this way, although the mineral acids used to stabilize the silicate binder in the storage container may present a problem if not rinsed away prior to over-coating and immersion. Hydroxide salts generated in the blister fluid. The continuous generation of hydroxide ion in blister fluid will occur when two conditions are met: (a) a liquid water filled area exists at the steel or zinc primer interface irrespective of how small (as opposed to intercoat delaminated areas) allowing oxygen to migrate in; (b) a cathodic potential exists of sufficient voltage to reduce any oxygen that migrates into the blister fluid. The oxygen is converted to hydroxide as it migrates into the blister fluid. 2H2O + O2 → 4OH–
5.
This represents the in situ production of a salt that will drive osmotic blister activity. The blister fluid will present with high conductivity and a high pH but may also contain dissolved organic and inorganic moieties from the paint. Benzyl alcohol is more soluble in alkaline blister fluid than in neutral pH water and has been seen in high pH blister fluid. Hydrolysis of formulated, or cargo absorbed, materials such as ester solvents or chlorinated solvents. Esters are the reaction product of an acid and an alcohol.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 51
Q
UNDERSTANDING OSMOTIC ACTIVITY IN PAINT FILMS
Hence, butyl acetate is the reaction product of butanol and acetic acid. When this has been a cargo in a lined tank, after washing with water we have seen hydrolysis within the coating matrix. The butyl alcohol and acetic acid are regenerated by soponification reaction with a water molecule. Acetic acid that migrates to the steel interface may produce a corrosion product with iron that is itself somewhat soluble and becomes one of the osmotic driving solutes. Chlorinated solvents hydrolyze to HCl. The rate of hydrolysis is a function of the specie undergoing hydrolysis and the alkaline strength of the nucleophilic centers (nucleophilicity) within the coating matrix that catalyze the hydrolysis. The blister fluid will be of a low pH (less than 7 depending on acid content). Fermentation of sugars, as in molasses cargoes aboard ships, also produces acetic acid. Acetic acid will migrate through the coating and initiate corrosion at the steel interface. Analyses of the solid iron corrosion product recovered from under ruptured blisters will give a strong indication of acetic acid when utilizing the pyrolysis function of the gas chromatograph mass spectrometer (GC/MS) if this has occurred. 6. Production of sulfuric acid (H2SO4) within the slime layer of an internal lined sewer pipe caused by the biooxidation of hydrogen sulfide (H2S). This leads to low pH blister fluids, and often perforation of the pipe. 7. Low molecular weight amine functional curatives that have been blocked from the epoxy reaction by carbon dioxide (CO2) or other blocking agents. For CO2 blocked converters, this is typically the result of cold temperature application wherein the epoxy-amine reaction slows or stops and the acid gas CO2-amine reaction continues reasonably fast. During mixing of the amino side of the epoxy coating or the “A” + “B” mix, if a vortex is drawn on the mixing paint and air is mixed into the paint, most of the CO2 in the air will be stripped and the resulting blocked curative will not be available for reaction with the epoxy. This is one of the sources of variability in the ultimate Tg of the cured coating. Inclusion of ketones in the amino side of an epoxy curatives is another potential source of failures. This happens when the amino functionality is blocked by the formation of ketimines in the can during storage. In this case, the length of time since curative manufacturer and the past thermal history will influence the amount of ketimine that has developed in the can. Ketimines require moisture from the atmosphere to cure and are the curatives used in “moisture cured” epoxy systems. The data developed by the blister fluid analysis, on occasion, will indicate additional tests useful in confirming the source of the water soluble solutes found. This is true when their presence cannot be explained by the MSDS of one of the paint products involved. The MSDS for the coating material does not always include non-hazardous watersoluble solvents. Evidence developed from work by this laboratory on ship-bottom paints has shown that solvents from any coat of applied paint will migrate throughout all prior applied old layers allowing blisters to form at the steel interface. This was demonstrated after the fresh application of a topcoat containing an oxygenated alcohol was applied to a ship’s hull over three layers of five year old, pre-existing paint that never blistered. The analysis of the blister fluid
649
collected from below the original paint system indicated the water-soluble alcohol that was present in the new paint application. On occasion, chemical reactions in or about the osmotic blister area will be producing the solutes. For this reason, to establish liability, the investigator must be prepared to accept the data from the blister fluid analysis and seek the true chemical pathways to these solutes, even if he must revisit his chemistry textbooks.
THE SYSTEMATIC CHEMICAL ANALYSIS OF OSMOTIC BLISTER FLUIDS AND BLISTERED COATINGS
A systematic analysis of the fluids removed from osmotic blisters will provide the answers as to what drives the osmotic activity and therefore assist in establishing liability. In addition, the analysis often will indicate ancillary chemistries occurring within the blister fluid itself and/or within the coating matrix forming the blister dome. The following analytical methodologies have been found to provide sufficient data to identify the material(s) initiating the activity and any additional driving forces that might be developed after blister formation. The collection of blister fluids is best done using a hypodermic syringe after first producing a small hole through which the needle may be placed as seen in Fig. 5. Usually, the coatings are too hard and brittle to force the needle through the coating film itself. For this reason, to minimize the chance of destroying, bending, or plugging the needle, drilling or puncturing with a sharp nail at the edge of the blister may be best. The solution should be collected and stored in glass vials since many of the solvents will migrate through plastic containers if the chemical analysis is not done timely. Ideally, many answers are provided by using a GC/MS having the ability, first, to out-gas and identify the lower molecular weight volatiles followed by a separate, second, analysis of the fragments after pyrolysis at elevated temperature of the residues. The pyrolysis, then, is done without opening the thermal extraction (TE)/pyrolysis (Pyro) oven. This is useful in solving some of the mysteries when the solutes are organic in nature. Higher molecular weight water soluble materials may be identified that are not seen on direct injection. When this type of equipment is not available, gas chromatograph/mass spectrometers utilizing
Fig. 5—Blister fluid is being removed under water using a small hypodermic syringe. This is in the coating within a large concrete pool at a state aquarium. Note sharp wood chisel used to first produce a very small hole in the coating at the edge of the blister.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
650
PAINT AND COATING TESTING MANUAL
Q
direct injection of the blister fluid is an alternative that will provide good definitive data for some of the cases, but the solvent portion is always much greater than the molecular solutes. The ionic solutes will not volatilize and cannot be seen by the GC. These are identified by ion chromatography or by scanning electron microscope (SEM) with energy dispersive x-ray (EDX) after drying the blister fluid residues (Fig. 6). Knowing potential sources of osmotic activity allows narrowing the search. The sequence of procedures followed in the blister fluid analysis yielding the most data will include the following steps. 1. GC/MS analysis by low temperature TE/desorption (water soluble solvents). 2. GC/MS analysis by pyrolysis analysis of non-volatiles remaining after TE (higher molecular weight, water soluble reaction products); Done without opening TE/ Pyro oven. 3. Determination of electrical conductivity and pH 4. Ion chromatography to identify the specific ions other than OH− and H+ ions seen in the pH determination. When not available, SEM/EDX analysis of the dried blister fluid residue will reveal the inorganics present. While the mass spec detector suffers in its being more difficult to perform finite concentration determinations, its ability to provide the identity of unknown water soluble organic solutes from their specific mass fragment fingerprint is excellent. Although it may be useful in some cases, the absolute concentration of the specific solutes in the blister fluid is not required to determine the source of the osmotic activity. After analysis of the blister fluid, additional analyses of related coating materials will allow determining the source of the water soluble solvents. These analytical techniques
Fig. 6—The ideal analytical instrument configuration for identifying the source of osmotic activity in paint films is a gas chromatograph having a mass spectrometer as a detector, a thermal out-gassing/pyrolysis attachment, and a liquid nitrogen trapping section at the head of the GC column. To the right is the pyrolysis and thermal out-gassing attachment. The liquid nitrogen cryogenic focusing on capillary column is within the GC oven as seen in Fig. 6. The components of the analytical system are as follows: 1. Gas chromatograph; 2. Mass spectrometer; 3. Pyrolysis-outgassing controller for solid samples; 4. Liquid nitrogen cryogenic controller; 5. Heated crossover line from pyrolysis oven to cryogenic trap in the GC oven; 6. Liquid nitrogen supply line from outside LN2 storage.
15TH EDITION
include thermal extraction and pyrolysis analysis of the coating materials forming the dome of the blister, as well as coating material in the near vicinity of the blister. Testing wet, un-applied, un-mixed coating materials from the can will pinpoint specific problems if these are from the paint formulation. If a water soluble solvent is found to be the osmotic driving force, the comparison of the analyses of the applied coating and the unapplied coating material will indicate whether this bad solvent was added during application or was included by the manufacturer in the paint formulation. Occasionally a review of the MSDS for the coating product will indicate the water-soluble solvent, though from experience we have seen that the paint manufacturer often does not identify these solvents (Fig. 7).
GC/MS Analysis by Low Temperature Thermal Extraction/Desorption (TE/GC/MS)
Often, the amount of available blister solution is small. Because of the large amount of data provided by the GC/ MS thermal desorption and pyrolysis techniques, these analyses should be performed first if only one drop of blister fluid solution is available. When salts are the primary osmotic driving force, usually there will be ample blister fluid sample available (Fig. 8). The ideal GC/MS analytical instrument used for the thermal extraction and pyrolysis analysis is a combination of four instruments linked together. It has been described in detail previously [4,5]. Briefly, it is a late-model GC with a mass spectrometer used as the detector. Integrally attached to the GC inlet is a heated transfer line of deactivated fused silica maintained at about 325°C (620°F) connecting the computer controlled TE/ Pyro oven. The very small TE/ Pyro oven has a stream of Helium sweeping any volatilized materials into the cryogenic trap in the GC. The separation
Fig. 7—Oven of gas chromatograph housing quartz capillary column (30 m) and quartz cryogenic focusing chamber. The components are: 1. 30 m capillary column passing through liquid nitrogen (LN2) cryogenic trap; 2. LN2 inlet and outlet fed by thermostat control of LN2 valve; 3. Quartz LN2 cryogenic chamber with electrical resistance heater to flush trapped components onto the column at the start of the GC analysis. Capillary column passes through the LN2 cooled cryogenic trap. An alternative “home-made” cryogenic trap is a simple styrofoam cup containing liquid nitrogen. This is placed in the GC oven allowing a short section of the capillary column to be manually submerged during sample out-gassing. The LN2 is removed at the start of the GC analysis.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 51
Q
UNDERSTANDING OSMOTIC ACTIVITY IN PAINT FILMS
651
Fig. 9—A pH indicating a paper strip is used to determine the pH of the blister solution. The fluid from the conductivity meter is adequate for this determination. If there is minimal blister fluid available, a very small spot from the hypodermic syringe will suffice. Fig. 8—A conductivity meter requiring only one drop of blister fluid is used to determine electrical conductivity. Conductivity of the blister fluid solution is a measure of the total number of ions or salts present. Specific identification of the ions is provided by SEM/EDX or ion chromatography. (Photo provided by HORIBA Instruments, Inc., www.horiba.com)
column used is a 30 m DB-5 capillary column. About 4 in. of the front end of the column passes through a quartz chamber plumbed to receive liquid nitrogen. This chamber also contains a wire resistance heater capable of heating the frozen column to 275°C (527°F) very quickly. A separate computer controller is programmed to maintain −125°C (−193°F), initially to cryogenically trap the volatilized components or thermally degraded fragments from the heated polymer. The thermal desorption/pyrolysis oven operates for about 6.5 min, going from ambient to 125°C (257°F) for TE or ambient to 500°C (932°F) for pyrolysis. This drives off any materials volatile at these conditions. The cryogenic trap maintains its cold −125°C (−193°F) temperature for 10 min. At the end of the 10 min. collection period, the controller shuts off the LN2, the resistance heater quickly takes the cryogenically cooled trapped section of the column to 275°C (527°F), the GC is automatically switched on, and the mass spectrometer starts to collect data. An alternative cryogenic trap that works very well is a simple styrofoam cup placed in the GC containing liquid nitrogen. A loop of the GC column is then submerged under the LN2. To start the analysis the cup of LN2 is removed and the GC started after the out-gassing/ pyrolysis. This provides −190°C (−310°F) cryogenic trapping. At this temperature, all products of interest will be stopped, including carbon dioxide and acetone. Frequently, solutes identified by the low temperature TE analysis include the oxygenated alcohols (2-butoxy ethanol, 1-methoxy-2-propanol, etc.), the butanols, benzyl alcohol, the xylenes, ketones, acids (both organic and mineral), and any solutes volatilized at the 125°C (257°F) TE condition. If the blister is being driven by hydroxyl ion being leached from the substrate (as with some water based zinc silicates) or generated from CP impressed currents, the pH will be high. For high pH blister fluids, the solubility of any benzyl alcohol binder extender and other alcohols become much greater. These may then be seen. Typically, their concentrations will not be sufficiently great to continue the
development of the blisters in neutral (water) blister fluid. Tank linings used in organic solvent service can generate blisters from the alcohol extenders depending on the coating and service (Fig. 9).
GC/MS High Temperature Pyrolysis (PYRO/GC/MS)
To identify the higher molecular weight solutes, without opening the TE/Pyro oven system after the TE run, the TE/ Pyro oven controller is set for the pyrolysis programmed conditions. The cryogenic trap is brought back down to −125°C (−193°F) or the cup of LN2 is re-inserted. When the cryogenic section of the separation column reaches the low trapping temperature, the pyrolysis oven containing the dried blister fluid sample is heated to 500°C (932°F). Any organic materials that were dissolved in the blister fluid but did not volatilize during the warming thermal extraction procedure are now subjected to the potential thermal degradation temperature of 500°C (932°F). Any heavy organic solutes will degrade to volatile components. The identification of these components allows one to visualize their precursors.
Determination of Electrical Conductivity and pH
If there is sufficient blister fluid, one drop is used to determine the electrical conductivity and pH. When the conductivity is high, there is ionic activity. If the pH is alkaline, the ionic source could be from applied CP, or a source of stray current potential sufficient to reduce molecular oxygen to the hydroxide. The coating may have been applied over a water based zinc shop primer wherein all of the sodium or potassium hydroxide could not be rinsed away. There is a very low probability that some of the amine curative may have been solubilized. If the conductivity is high and the pH is neutral, as may be the case between coating layers, the salts may be organic acid or carbonated amine curatives. An exception is in areas where wind blown salts are available, as on offshore facilities and in marine applications. If the conductivity is high and the pH is low, there is some source of acid causing the pH depression. Unlike the clean, non-corroded metal seen under high pH blister
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
652
PAINT AND COATING TESTING MANUAL
Q
fluids, typically, the low pH fluids cause the steel to corrode. The source of acid must be determined from past history or pyrolysis of the iron corrosion product. For cargo tank linings, this is a common problem and may be deduced from cargo histories that are available. Acetic acid is especially problematic. Cargoes of molasses and natural sugar products often contain acetic acid that is the result of bio-degradation. The transmission coefficients for acetic acid through most tank linings are such that it often becomes the primary solute in the osmotic activity. After desolution of iron, additional soluble iron (II) complexes add to the solute mix dissolved in the blister fluid. As oxygen transports across the paint film to the blister fluid, the soluble iron (II) complex is converted to iron (III) and precipitates out of solution. Analysis of the rusty iron scrapings from under the blister area by pyrolysis GC/MS will often reveal which acid has been operative, even if it is hydrochloric acid. Hydrochloric acid results from the hydrolysis of chlorinated solvents. There are many different chlorinated organic solvents stored in coated tanks as well as transported in coated marine tankers. All will hydrolyze to some degree to produce the HCl. The most problematic chlorinated solvent seen by this laboratory is the cargo material allyl chloride, CH2苷CH2CH2Cl. This material absorbs into the tank lining and within hours after water washing, liberates HCl. Hydrogen sulfide, H2S, may be generated within sewer systems that are not properly treated to control the condition. Since this is an anaerobic action (occurs in the absence of oxygen), to stop this H2S formation, sewer operators will sometimes attempt to maintain a sufficient amount of dissolved oxygen (DO) in the effluent to disallow attack of the sulfates. Another practice is the addition of nitrates into the effluent, since the bacteria prefer nitrates to sulfates as their hydrogen acceptor in the energetics of their life cycle. If the sewer system has not been properly controlled for the production of hydrogen sulfide, sulfuric acid will be produced from any H2S after reaching aerobic conditions. This occurs readily in the crown of partially filled pipes where air is available above the wastewater. The resulting corrosion is called “Crown Corrosion” for this reason [6]. This can destroy both steel and concrete pipe systems after loss of coating by osmotic blister formation.
Ion Chromatography
The use of ion chromatography usually will identify the specific ions causing the high conductivity in blister fluids. Past experience of this laboratory has shown that the presence of hydroxyl ions from a CP induced high pH blister fluid often will cause an increase in additional ionic species in the blister fluid. Under blue or green coatings, bromine may be hydrolyzed from the phthalo blues and greens pigments. Chlorides may be hydrolyzed from epoxy resins in high pH blister fluids. For this reason, the presence of the chloride or bromide ion in the blister fluid is NOT, conclusively, indicative of salts over-coated during application although these may have contributed to the diminished adhesion exacerbating the osmotic activity. Sulfates will be seen in oxidation of the H2S in sewer pipes. Sea water evaporation and blown sea salts will contain chlorides, bicarbonates or carbonates (depending on pH), sulfates, and all other ions typically seen in seawater.
15TH EDITION
CASE HISTORIES OF OSMOTIC ACTIVITY PAINT FAILURES
Case Number 1: A series of barges employed in fresh water river trade developed blisters after being in service a number of months. The time to discovery was about a year. The paint system included the use of a water based potassium silicate zinc shop primer applied to the steel plate upon arrival at the fabrication facility. After application of the shop primer, the steel was fabricated into barges within about a month. After fabrication of the barge, and just a few days before launch, the finished barge was high-pressure fresh water rinsed, allowed to dry, and over-coated with an epoxy coating. Blisters were discovered during dry-docking about a year later. There were no cathodic protection anodes used. The declaration was made that there had been no cathodic protection potentials applied to the vessel during its moorings. Grounding a vessel to the CP system of a dock is a practice that often occurs with vessels that trade in flammable cargo. These barges traded in dry cargo. Fluid was collected from a number of the blisters on the bottom of a barge. The blister fluid was clear, the pH of the fluid was about 13, the conductivity was measured at 19 200 μS (microsemins). Chemical analyses of the fluid by GC/MS TE followed by PYRO indicated that a small amount of organic solubles was present in the fluid. These were all alcohols that had migrated from the epoxy paint topcoat. There were no heavier molecular weight materials seen during the pyrolysis analysis. The analyses indicated that the water based zinc silicate still contained sufficient water-soluble potassium hydroxide to become the solute in the osmotic blisters. It is extremely difficult to extract all alkali from these types of silicate films in one short rinsing. Additionally, the curing of the silicate film may require months to fully bind the potassium and allow loss of the excess alkali to become minimally solublesalt free. (It cannot become salt-free.) The driving force to produce blisters is just as strong for the salt potassium hydroxide as it is for the salt sodium chloride. One would never consider over-coating high concentrations of sodium chloride salt. Because zinc metal is amphoteric (i.e., can be dissolved by either acid or base) the tetrahydroxy zinc (II) ion is just another solute driving the osmotic activity. The organic solutes seen in the GC/MS analysis are, in part, because of their increased solubility as the pH of the blister fluid increases. The butanol and benzyl alcohol have limited solubility in neutral water. The 1-methoxy-2-propanol is 100 % water-soluble and can be a problem in certain scenarios. This water soluble solvent is sensitive to length of drying time and ventilation. The transmission coefficient for this oxygenated alcohol is greater than the 2-butoxy ethanol and often will not be sufficient to generate visible osmotic blisters. The exception is in well cross-linked epoxy systems where its mobility is more restricted. This is an example of the importance of the numerical values of the “semi” in the “semi-permeable membrane” part of a typical definition of “osmosis.” In this case, although the concentrations were relatively small for all alcohols, the 1-methoxy-2-propanol exacerbated the blister formation and the hydroxide ion provided the driving force for the continued osmotic activity (see Fig. 10 and Fig. 11). Case Number 2: Blisters developed within the coating system of a large floating offshore drilling platform. The
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 51
Q
UNDERSTANDING OSMOTIC ACTIVITY IN PAINT FILMS
653
Fig. 10—GC/MS response from the thermal extraction of blister fluid collected from the bottom of a barge. Alcohols that normally are not appreciably soluble in water are more soluble in alkaline blister fluid. The scan is magnified to reveal the small amounts of organic solutes.
Fig. 11—Fragment pattern collected by the mass spectrometer at 13.677 min. The computer matched this pattern to 1-Methoxy-2Propanol, a water soluble epoxy thinner. This “oxygenated” alcohol can cause osmotic blisters when condensed water is possible.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
654
PAINT AND COATING TESTING MANUAL
Q
paint system was similar to Case Number 1 with the exception that the zinc silicate was applied as a holding primer after construction. It was a solvent based, ethyl silicate zinc coating. The two-coat epoxy topcoat was applied during colder months in Europe. Severe blistering occurred after flooding the tanks for testing. The drilling platform was delivered to the US for completion and an investigation of the cause of the osmotic activity was done [7]. Blister fluids were collected from two types of blisters based on their development location. Type one were blisters that originated at the metal interface and were under the full thickness of epoxy paint. Type two blisters were between the layers of the epoxy coating. Both blister fluids were clear as they were removed using hypodermic syringes. These fluids were analyzed separately. The blister fluid removed from the metal interface had a relatively high pH of greater than 12. The pH of the fluid extracted from blisters between layers of epoxy was neutral, pH of 7. The GC/MS chemical analysis of the two blister fluids was very similar even though there was a large difference in their pH. There was a very large peak of 2-butoxy ethanol in each. It was very evident that the driving force for the blister formation was the oxygenated alcohol, 2-butoxy ethanol. A review of the MSDS for the reported coating material did not indicate the presence of this alcohol in the coating formulation. For this reason, it was first suspected that the applicator might have used this water-soluble solvent during application. This was shown to not be the case when retained samples of the actual coating materials used became available for GC/MS analysis. These were tested and found to contain the 2-butoxy ethanol as part of the paint formulation. The coating manufacturer was able to determine that there was a difference in the solvent system of the European produced material as opposed to that manufactured in the US. The high pH of the blister
15TH EDITION
fluid removed from contact with the metal substrate was the result of sacrificial anodes installed during the construction of the platform. Oxygen that migrates into the blister fluid at the metal interface experienced sufficient electrical potential to cause the oxygen to be converted to hydroxide ion producing the high pH. For blisters formed between paint layers, there could be no electrical potential felt between layers, although there was compromised adhesion that allowed the delamination between coats of epoxy at the blister site [8].
References [1] Storfer, S. J., and Yuhas, S. A., “Mechanism of Blister Formation in Organic Coatings; Effect of Coating Solvents,” Material Performance, July 1989, pp. 35–41. [2] The Energetics Driving Osmotic Activity: Physical Chemistry, 3rd Ed., Alberty/Daniels, pp. 170–172. [3] Mills, G. D., “Sources of Bubbles, Foam, and Voids in Coatings,” Materials Performance, March 2000, National Association of Corrosion Engineers, Houston, TX. [4] Mills, G. D., Sansum, A., and Cox, G., “Instruments modernes d’analyse: Possibilites et limites,” European Coatings Journal, p. 960, Hanover, Germany, December, 1994. [5] Mills, G. D., “Tool of the Trade—Analyzing Paints and Raw Materials with a Specially Configured Hyphenated Gas Chromatograph/Mass Spectrometer,” Modern Paint & Coatings, pp. 24–29, December, 1999. [6] Design Manual: Odor and Corrosion Control in Sanitary Sewerage Systems and Treatment Plants; EPA/625/1-85/018. October 1985. [7] Mills, G. D., “Using a Specially Configured Composite Instrument Gas Chromatograph/Mass Spectrometer (GC/MS) for the Analysis of Paints and Raw Material,” Proceedings of the 77th Annual Meeting Technical Program, FSCT, Dallas, Oct. 1999, p. 214. [8] Mills, G. D., and Eliasson, J., “Factors Influencing Early Crack Development in Marine Cargo and Ballast Tank Coatings,” SSPC/PACE 2005, Tampa, Florida, Feb. 1, 2006.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
52
MNL17-EB/Jan. 2012
Stress Phenomena in Organic Coatings Dan Y. Perera1 PREFACE
THIS CHAPTER REVIEWS AND UPDATES THE TOPICS addressed by the previous authors, including up-to-date references and expanded to include stress in multi-layer systems and stress in physical aging. Interest in the stress phenomena in organic coatings is relatively new. The importance of understanding and measuring this stress became evident as facts accumulated about its role in coating degradation [1–10]. This is also supported by the fact that many coatings used today (e.g., thermosets) are more susceptible to developing high stresses than traditional ones (e.g., alkyd paints). It is now quite clear that stress can affect coating adhesion and/or cohesion and provoke delamination and/or cracking. Since the development of stress is involved in practically all stages of coating life (film formation, exposure to various climatic conditions), its measurement is essential for a better understanding of coating behavior. In addition, the measurement of stress can be used to evaluate other important characteristics of a coating, such as the glass transition temperature (Tg) and the critical pigment volume concentration (CPVC).
ORIGINS OF STRESS IN ORGANIC COATINGS
Stresses originate in organic coatings as a result of their adhesion to the substrate. Good adhesion is, on the one hand, indispensable for adequate substrate protection, but on the other hand prevents the normal movement of a coating. The three main causes which provoke stress in an organic coating are [11]: 1. Film formation. 2. Variation in temperature. 3. Variation in relative humidity (RH). The stresses induced by film formation, variation in temperature, and variation in RH are known, respectively, as internal, thermal, and hygroscopic. The latter two stresses are also referred to as hygrothermal (SHT) It is important to realize that, due to the coating adhesion to its substrate, stress exerts its action mainly in a plane parallel to the substrate [6,12,13]. Therefore, one can write Eε S= (1) 1− v where S = stress, E = elastic modulus,
1
ε = strain, v = Poisson’s ratio. This relationship represents the stress equation for an uni-axial deformation adapted for a bi-axial deformation by dividing it with 1 – v. It is therefore considered that the coating can contract or expand only through its thickness. In practice this might not always be so, i.e., in addition to tangential stresses, the coating can also be subjected to normal stresses. Stresses are especially damaging at edges [14] where they cannot cancel each other or in the middle of the plate if defects or heterogeneities are present.
Film Formation
During film formation and regardless of the mechanism involved (evaporation of solvent, coalescence, chemical reaction, or their combination), in almost all cases the coating tends to contract. If this contraction is prevented by coating adhesion to its substrate and/or the mobility of macromolecular segments is hindered, a tensile stress will develop in the coating. In the literature certain confusion reigns about the denomination of the stress arising during film formation. For example, for solvent-based coatings an author may use these terms: cure stresses, solvent removal stresses, residual stresses, shrinkage stresses, internal stresses. For simplification, I suggest referring to the stress arising during film formation as internal stress (SF). If a liquid paint is applied on a substrate [8] and the development of stress is measured as a function of time, a number of stages can be observed (Fig. 1). In Stage 0, the coating is still liquid and is mobile enough to permit volume contraction, and consequently no stress develops. In Stage 1, the film starts to form, the volume contraction is restricted, and stress develops. In Stage 2, a number of processes can occur. Depending on the coating composition, the evolution of stress can include various combinations of increasing, decreasing, and stationary stress levels. If no damage occurs to the film (cracks, microfissures, loss of adhesion), a decrease in stress results from relaxation processes which occur from the moment the stress develops but only become evident at this stage. It also follows that the measured stress values are a result of two opposite processes, one tending to develop stresses in the coating (in this case due to the volume shrinkage) and the other tending to decrease them (due to stress relaxation).
Scientific Adviser, Coatings Research Institute (CoRI), B-1342 Limelette, Belgium.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed byASTM International This standard is for EDUCATIONAL USE ONLY. Copyright © 2012 by www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
655
656
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 1—Schematic description of stress (S) dependence on time.
Examples of stress development as a function of time for latex coatings [15] above and below the CPVC are shown in Fig. 2. It is important to add that stress starts to develop in a coating when its Tg is at least equal to the experimental temperature [6]. For solvent-based coatings (Fig. 3), this is also confirmed by the fact that stress appears when Phase 2 of the evaporation kinetics (the diffusion process) starts [16]. At this point in the evaporation, referred to as solidification, the solvent volume fractions are equal to the solvent concentration necessary to bring the Tg of the coating to a temperature equal or higher than the experimental one [6,17]. The stresses arising in the different stages of the film formation of the polymer lattices are also discussed in Ref. [18]. The paper provides evidences of lateral dilatation stress occurring during the initial stage that is attributed to the capillary pressure. If Eq (1) is extended to accommodate the specific effect of internal strain, εF [6], then
εF ≈
Vs − Vt ΔV = 3Vs 3Vs
(2)
where Vs = volume of coating at solidification, and Vt = volume of coating at time t after solidification.
Fig. 3—Schematic representation of the dependence of VS/VF and stress (S) on time, VS = volume of solvent present in the film; VF = volume of the dry film.
One can write [19] that SF =
∫
VS
Vt
E 1 dV 1 − v 3Vs
(3)
Despite the limitations of Eq (2), which assumes that the strain is isotropic and linear, Eqs (2) and (3) enable one not only to better understand the factors affecting the SF but eventually to calculate it approximately.
Variation of Temperature
When coated substrates are exposed to variations in temperature, dimensional changes are induced. If the thermal expansion coefficients of the coating ( FT ) and the substrates ST are different, which is usually the case, a thermal stress (ST) will develop in the coating [2,7,11,20,21]. Since the thermal strain, εT is given by
( )
ε T ≈ ( FT − ST ) ΔT
(4)
the combination of Eqs (1) and (4) gives ST =
∫
T1
T2
E (FT − ST ) dT 1− v
Determination of Tg
Fig. 2—Schematic description of stress (S) dependence on time for latex coatings: (a) and (b) PVC < CPVC; (c) and (d) PVC > CPVC; (e) PVC < CPVC in the presence of a poor coalescent; (f) PVC > CPVC in the presence of a poor coalescent.
(5)
A schematic description of the stress dependence on temperature [19] is represented in Fig. 4. Such dependence indicates the possibility of determining the Tg of a coating from the stress measurement. This is due to the fact that E, FT , and v (see Eqs (1) and (5)) also show a profound change at Tg. Below temperature a the coating is in the glassy region, and above temperature b the coating is in the rubbery region. The linear dependence of S on temperature in the glassy region greatly facilitates the measurement of Tg. This linearity is due to the fact that in this region E, FT , and v are practically independent of temperature.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 52
STRESS PHENOMENA IN ORGANIC COATINGS
Q
657
giving ve =
Er Sr = 3 RTr 6 RTr ε r
(8)
where R, Sr, Er, and Tr are, respectively, the gas constant, the stress, the elastic modulus, and the temperature at the beginning of the rubbery region. If Sr is measured and εr is known or determined from separate measurements (e.g., by thermomechanical analysis), ve can be calculated. Others have used the evaluation of Tg by stress measurements to investigate the effect of a pretreatment on certain pigments [25], the state of cure of baking enamels [22], and the modification of epoxy coatings [20,26].
Variation of Relative Humidity
Fig. 4—Schematic representation of dependence of S, E, α, and v on temperature (T).
Three examples of the stress dependence on temperature and Tg determination by thermal stress measurements [19] are shown in Fig. 5. When the stress magnitude in the rubbery region is small (e.g., elastomers, coatings with a low degree of crosslinking), the Tg can be determined with fair accuracy just by carrying out a few measurements in the glassy region and then extrapolating the straight line to ST = 0. For coatings subjected to a significant stress in the rubbery region (e.g., highly cross-linked thermosets), the measurement of stress in this region might provide a way to approximately determine the cross-link density, ve [22]. Since in accordance with the molecular theory of rubber elasticity in its simplified form [23,24] Er = 3 ve RTr
(6)
Sr = 2 Er ε r
(7)
and
Dimensional changes induced by absorption and desorption of water as a result of variation in RH is another cause of stress development in a coating [11,27,28]. As in the case of temperature, if a mismatch between the expansion coefficients of the coating and the substrate exists, a hygroscopic stress (SH) will arise in the coating. Since the hygroscopic strain εH is given by
ε H ≈ ( FH − SH ) ΔH
(9)
one can write that SH =
∫
RH2 RH1
E (FH − SH ) dRH 1− v
(10)
where FH and SH are, respectively, the hygroscopic expansion coefficients of the coating and the substrate. Some examples of the SH dependence on RH are given in Fig. 6.
INTERDEPENDENCE OF STRESSES
While in previous sections the various stresses (SF, ST, SH) were discussed separately, in practice they can act together in such a way that the total stress (Stot) is small or, as in many cases, very important [11] Stot = S F ± ST ± S H
(11)
Fig. 5—Stress (S, MPa) dependence on temperature (T, °C) for three coatings at RH ≅ 0 % [an epoxy (1), a polyurethane (2), and an epoxy/melamine system (3)]. Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
658
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 7—Schematic description of the stress (S) dependence on time at two experimental conditions (water and 50 % RH). X = the initial stress.
Fig. 6—Stress (S, MPa) dependence on relative humidity (RH, %) for a polyester powder coating (1); an epoxy (2); a polyurethane (3); and a latex coating (4).
The positive and negative signs are arbitrarily chosen. The positive sign denotes a coating tendency to contract (tensile stress) and the negative sign a coating tendency to expand (compressive stress). SF is practically always positive. Equation (11) indicates the existence of two climatic conditions which might provoke a high stress in a coating: 1. Low temperatures and RH’s induce high tensile stresses (e.g., a dry, cold winter). 2. High temperatures and RH’s induce high compressive stresses (e.g., a humid, warm summer). According to Eq (11) and depending on the type of coating and the way it was cured, a number of situations can arise. For example, for a thermosetting coating cured at a high temperature (i.e., at T > Tg, RH = 0 %) and then exposed to different RH’s, since S F ≈ 0 (except for highly cross-linked coatings), the Stot is given by Stot ≈ ST − S H
Stot = ST − S H
(14)
The results obtained [11] show that in certain cases (e.g., for an epoxy coating), the Stot can be very different [Stot (Eq (13)) = 5 MPa; Stot (Eq (14)) = 0.4 MPa]. Interesting cases are those where the Tg of a coating is close to or below the experimental temperature once they are immersed in water or exposed to a high RH [29]. Under such conditions SF or ST can relax and therefore Stot ≈ S H. Such a situation is illustrated schematically in Fig. 7 and for a particular epoxy coating in Fig. 8. Once a coated substrate is immersed in water one can observe first the development of a hygroscopic compressive stress followed by its decrease. The time necessary to reach zero stress is mainly dependent on the type of coating. The withdrawal of the coated substrate from water provokes first the development of a relatively high tensile hygroscopic stress followed by its decrease. The decrease in stress (Curves 2 and 4 in Fig. 7) is due to relaxation processes facilitated by the low Tg of the coating. The much higher stress values attained after the immediate withdrawal of the
(12)
In Fig. 6, for one coating, the stress is positive regardless of RH, meaning that ST is always higher than SH; but cases where negative stress values are obtained are not rare. It could be demonstrated [11] that the same coating conditioned in an identical environment but with different previous histories can develop different values of stress. Thus, a coating cured under isothermal and constant RH conditions (e.g., 21°C and 50 %) will develop a total stress given by Stot ≈ S F − S H
(13)
Now, if the coated substrate is first heated at T > Tg for sufficient time to enable a maximum stress relaxation and then brought back to the initial conditions (i.e., 21°C and 50 % RH)
Fig. 8—Epoxy coating. Stress (S) dependence on time (t) at two experimental conditions (water and 50 % RH).
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 52
Q
coated substrate from water (Curve 3 in Fig. 7), in comparison with the initial stress (Stot ≈ S F − S H or Stot ≈ ST − S H) are due to the high hygroscopic tensile stress Stot ≈ S H . The above findings are important not only for understanding the mechanism of the stress development in organic coatings but also for practical reasons such as choosing the experimental conditions of natural or accelerated weathering tests, i.e., the magnitude of stress developed in a coating might depend as much on natural climatic conditions as on the type and order of cycle selected in the laboratory.
(
STRESS PHENOMENA IN ORGANIC COATINGS
659
)
STRESS MEASUREMENT
Equations (1), (3), (5), and (10) show that, if one knows the values of E, ε, and v of a coating, then in principle the various stresses can be calculated. However, except for relatively simple cases, this is difficult because the above coating characteristics can be time, temperature, and relative humidity dependent. Therefore, efforts were made to directly determine the stress arising in a coating. Among the methods one can find in the literature are: optical [30–33], strain gages [20,34,35], brittle lacquer materials [36], x-ray diffraction [37], and cantilever (beam) [2,4,8,11,27,38–45]. The cantilever (beam) method appears to be the most widely used and is suitable for determining the stress in an organic coating. This method makes use of the fact that for a coating under stress, applied on a substrate, the coated substrate will deflect in the direction which relieves the stress. Since the deflection can be measured and the elastic properties of the substrate are known from separate determinations, the stress can be calculated. Two variations of the cantilever (beam) method are described in the literature: a one-sided coated substrate either (1) fixed vertically at one end [4,8,27,38] (Fig. 9) or (2) freely supported on two knife edges [2,11,43,45] (Fig. 10). The stress analysis of (1) is more complicated and shows that to eliminate the effect of clamping on the coated substrate deflection, its measurement should be made at a distance higher than 80 mm from the clamping point [8]. Variation (2) is much simpler to analyze and can be designed to eliminate the effect of weight loss on the coated substrate deflection by choosing the right distance between the two knife edges [45]. Each variation has its advantages, but if correctly used they should give identical results. For example (1) is more suited to evaluate stress in water [29] and (2) to determine the effect of temperature [2,11]. Among the techniques used to measure the deflection of coated substrates, one can mention: capacitive transducers [44,45], laser [46], traveling microscope [4,8,27], automatic micrometer [11]. A variety of flexible substrates can be used, such as cold laminated steel, stainless steel, aluminum, copper, and silicon wafers.
Fig. 9—Schematic description of the vertically fixed at one end cantilever (beam) method.
Fig. 10—Schematic description of the freely supported beam method.
A commercial apparatus, the CoRI Stressmeter (Fig. 11), is based on variation (2) and the mathematical analysis described in Ref. [45]. This apparatus is almost completely automatic and enables one to measure the stress from about −5 to 110°C under a variety of RH’s. This instrument requires the use of substrates that are electrically conductive, e.g., cold laminated steel, stainless steel, aluminum, and copper. In certain cases, such as for investigating the stress arising in ultraviolet cured coatings, the stress measurement might necessitate an elaborate set-up. In this case the set-up might include, among others, a UV lamp, a controlled flow gas and temperature device, an application chamber and a video camera monitoring the deflection of the coated substrate [47].
Mono-Layer Systems
A number of mathematical equations are proposed in the literature to calculate the stress in a mono-layer system [41,42], but in the author’s opinion those proposed by Corcoran [41] are the closest to the real situation (e.g., considers the fact that stress develops in two directions)
S=
dES t 3 dE(t + c) + 3l c(t + c)(1 − vS ) l 2 (1 − vS )
(15)
S=
4 d′ES t 3 4 d′E(t + c) + 2 3l c(t + c)(1 − vS ) l (1 − vS )
(16)
2
2
Fig. 11—CoRI Stressmeter apparatus (Courtesy of Elcometer, Liège, Belgium).
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
660
PAINT AND COATING TESTING MANUAL
Q
where d = deflection of the substrate (Fig. 9), d = deflection in the middle of the substrate (Fig. 10), Es = elastic modulus of the substrate, E = elastic modulus of the coating, vs = Poisson’s ratio of the substrate, t = thickness of the substrate, c = thickness of the coating, l(Eq 15) = length of the coated substrate between the point at which it is clamped and the point at which the deflection is measured (Fig. 9), and l(Eq 16) = distance between the two knife edges (Fig. 10). Equations (15) and (16) assume, among other things, good adhesion between the coating and the substrate, isotropic elastic properties of the coating and the substrate, the elastic limit of the substrate is not exceeded, and the stress is constant throughout the coating thickness. If the last condition is not fulfilled it means that the coating is heterogeneous through the film thickness. Therefore, the measurement of stress as a function of film thickness might enable one to determine, depending of the film composition, the distribution of the solvent, cross-link density, pigmentation, etc., through the film thickness [8,15,16,68,76(a),76(b),77,79–81]. The second term in Eqs (15) and (16), which contains a number of coating properties difficult to determine, can be neglected if ES E and t c. Since the substrate is flexible and deflects during a stress measurement, in principle, part of the stress relaxes. This relaxation is however negligible with respect to the stresses arising in the coating. An approximate calculation, by considering even a high elastic modulus ( E ≈ 2 MPa ) and −6 strain (≈ 6 × 10 ) values, induces a negligible stress relaxation (0.01 MPa ) with respect to stresses usually arising in the coating. Most commonly, stainless steel or cold laminated steel [4,8,11,27,44,45] shims are used. Other substrates such as aluminum and copper can also be used. The elastic modulus, Es, of each shim can be determined prior to use with the CoRI Stressmeter by applying Eq (17) ES =
Pl 3 4 d ′t 3 b
(17)
where P = weight placed in the middle of the substrate, b = width of the substrate, and d, l, t = as in Eq (16). It is important to add that, although it is not difficult to make the measurements, nevertheless great care is necessary. One should always use the correct substrate thickness, adequately condition it, and precisely calibrate the apparatus.
Multi-Layer Systems
In many cases multi-layer systems are applied on substrates to protect them against a variety of aggressive factors. Since the mechanical characteristics of each layer
15TH EDITION
are usually different, one can expect that the stresses arising in each layer be also different. Furthermore, a monolayer system might become a multi-layer one if part of it acquires mechanical characteristics different from that of the bulk. This is typical of coatings subjected to chemical changes when submitted to natural or accelerated weathering. To the author’s knowledge, all the relationships for calculating the stress in multi-layer systems found in the literature [48–52] necessitate the determination of elastic properties of each layer, coating characteristics not always easily measurable. To enable the determination of stress in each layer of multi-layer systems mathematical equations that do not necessitate the knowledge the elastic characteristics of each layer were derived [53]: S2 = S3 =
4 ES t 3
3l 2 c2 ( t + 2c1 + c2 ) (1 − vS ) 4 ES t 3
(d
3l 2 c3 ( t + 2c1 + 2c2 ) (1 − vS )
2
− d1 )
(d
3
− d2 )
(18)
(19)
where c1, c2, c3 = thickness of individual layers, d1, d2, d3 = deflection of the coated substrates (d in Fig. 10) with one, two, and three layers, S2 (Eq (18)) and S3 (Eq (19)) = stress in a top coat of two- and three-layer systems, respectively. For a three-layer coating system whose composition is unaffected by multiple applications and curing, the determination of stress in each layer is as follows: —application of the first layer and its curing, —measurement of the deflection d1 and of the thickness c1, —calculation of stress S1 from Eq (16), —application of the second layer and its curing, —measurement of the deflection d2 and of the thickness c2, —calculation of stress S2 from Eq (18), —application of the third layer and its curing, —measurement of the deflection d3 and of the thickness c3, —calculation of the stress S3 from Eq (19). Procedures for coatings affected by successive applications and curing are described in Ref. [53]. Two examples by applying Eqs (18) and (19) are presented in Figs. 12 and 13, respectively. Fig. 12 shows that for a three-layer system the highest thermal stresses are arising in the primer. Fig. 13 illustrates the case when a single layer system of 77 μm becomes a multi-layer one, due to its exposure in an accelerated weathering device. After seven days of weathering the thermal and hygroscopic stresses arising in the first 27 μm are much higher than in the bulk.
STRESS AND PHYSICAL AGING
We are associating the stress development with physical aging because both phenomena are mainly occurring in the glassy region, and they are therefore unseparated.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 52
Q
STRESS PHENOMENA IN ORGANIC COATINGS
661
Fig. 14—Schematic description of the dependence of volume (V), enthalpy (H), and entropy (S′) on temperature. Ta = physically aging temperature.
Fig. 12—Stress (S) dependence on temperature (T) for a three layer acrylic/melamine system. S1, S2 and S3 indicate the stresses arising in the primer, middle, and top layers, respectively [53]. (With kind permission of Springer Science + Business Media.)
Physical aging is a general term that includes relaxation processes of a material in the glassy state, caused by its non-equilibrium state [54]. This process is easily explained by considering the dependence of volume (V), enthalpy (H), and entropy (S′) on temperature (T) above and below Tg (Fig. 14) [54–58]. During cooling at T > Tg, due to the great mobility of polymer chain segments, V, H, and S′ can follow the decrease in temperature. Below Tg the
mobility of polymer chain segments is reduced, and V, H, and S′ cannot follow any more the decrease in temperature, meaning that the material departs from its equilibrium state. The values of these thermodynamic quantities being higher than in equilibrium state they will continuously decrease until the equilibrium state is reached. The closer the aging temperature to Tg the higher the aging rate and the smaller the change in property. Since the studies made by Struik [54], this relaxation toward equilibrium is known as physical aging. In contrast with chemical aging that induces irreversible changes, physical aging is a reversible process. It can be eliminated by heating the samples to T > Tg for sufficient time. More than 800 studies have been dedicated to various aspects of physical aging. While certain phenomenological models (e.g., Kohlrausch—Williams—Watts) describe
Fig. 13—Stress (S) dependence on temperature (T) (left) and on relative humidity (RH) (right) for an acrylic/melamine system after an exposure of 7 days in a Q-Lab Corporation (QUV) device. Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
662
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
B are typical of physically non-aged and physically aged coatings obtained during their heating. The U-shape curve is mainly due to the increase of E with physical aging which in turn induces higher compressive thermal stresses (see Eq (5)). The right side of the trough represents the decrease of this stress with temperature due to the stress relaxation in the rubbery region. Studies made with thermoplastic (e.g., latex paints) and thermosetting (e.g., waterborne, powder coatings) systems show that the coating composition (type of binder, presence of coalescents, degree of film cross-link density, presence of pigments) affects the physical aging process [61,62]. Physical aging also influences the coating durability by increasing the stress magnitude and by decreasing the stress relaxation rate. There are indications that physical aging affects the late stages of coalescent process.
EFFECT OF COATING COMPONENTS
Fig. 15—Schematic description of the stress (S) dependence on temperature (T), for physically unaged (A) and aged (B) organic coatings. Segments 1 and 3 represent the glassy and the rubbery regions, respectively. Segment 2 represents the glass transition region [59].
well its kinetics, this process is not very well understood on a molecular scale. Nevertheless, it is associated with increased molecular packing and densification, processes that can greatly affect important coating properties like mechanical, thermal, and electric. Concerning mechanical properties, which directly affect coating integrity, densification causes the coating to become harder and more fragile. To describe the interdependence between stress and physical aging the thermal stress will be considered. This interdependence was discussed in detail in Refs. [59,60] and is schematically described in Fig. 15. Curves A and
Since E, ε, and v (Eqs (1), (3), (5), and (10)) are known to be affected by the coating components, one should expect the same to hold for stress development. This section will briefly review the influence of pigmentation, solvents, and binder.
Pigmentation
It has been shown that pigmentation, both the pigment volume concentration (PVC) and the type of filler (i.e., pigments and extenders), affects the development of internal stress [7,43,63–68]. To illustrate this, examples are presented in Figs. 16–21. Figs. 16 and 17 show, respectively, the stress dependence on time for a thermoplastic binder in solution and a latex, filled with a titanium dioxide (TiO2). Some PVCs are above the CPVC, and some are below the CPVC. The different stages occurring during the film formation, discussed previously, can be recognized. In Stage 1 stress increases relatively rapidly. For latex coatings, this stage corresponds to the transition phase of the evaporation kinetics when the greatest part of the coalescence occurs.
Fig. 16—Stress (S, MPa) as a function of time [hour (h) and day (d)] for a polyisobutyl methacrylate filled with TiO2. The numbers in the figure indicate the different PVCs (%) investigated. CPVC = 51 %. Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 52
Q
STRESS PHENOMENA IN ORGANIC COATINGS
663
Fig. 17—Stress (S, MPa) as a function of time for a latex (vinyl acetate/vinyl versatate copolymer filled with TiO2):45 % (○); 50 % (□); 55 % (X); 60 % (∆). CPVC = 52 %.
The plot of the maximum internal stress (Sm) as a function of PVC enables one to determine the CPVC of a coating. Some examples are presented in Figs. 18 and 19.
These figures, as well as other results presented in Refs. [66,67], clearly indicate that Sm is a function of PVC. Sm increases with PVC up to a certain PVC and then decreases. This PVC corresponds to the CPVC, indicating the possibility of accurately determining this characteristic [69,70] from stress measurements, and agrees well with the CPVCs calculated or determined by other methods (density, various mechanical properties). The dependence of Sm on PVC can be understood if one considers Eq (1) and the way E, ε, and v are affected by the PVC.
Fig. 18—Maximum internal stress (Sm) as a function of PVC (%) for a solvent-based thermoplastic binder filled with a TiO2 (X),a red iron oxide (■), a yellow iron oxide (∆), and a talc (○).
Fig. 19—Sm as a function of PVC (%) for a styrene acrylic copolymer filled with a TiO2 (■), a calcium carbonate (X), and talc (□).
In Stage 2 (which corresponds to Phase 2 of evaporation kinetics), depending on the PVC and the type of filler, the stress can decrease or first decrease and then increase. For PVC < CPVC, this decrease is mostly due to the relaxation process, but for PVC > CPVC is mostly due to relief processes such as filler/binder dislocations and/or formation of microfissures.
DETERMINATION OF THE CPVC
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
664
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 20—Sm (MPa) as a function of PVC (%) for a thermoplastic binder in solution containing a yellow iron oxide (∆), a talc (□), and their mixture (n = 0.5) ( ); n = pigment volume fraction.
t
For PVC < CPVC, E increases with increasing PVC because the E of an inorganic filler is in general higher than that of an organic binder [71]. For PVC > CPVC, E decreases as a result of the increasing film discontinuity [23,72]. Since in general ε and v are decreased (or are little affected) by the PVC, it follows that the increase of Sm with PVC is mainly due to the effect of PVC on E. Figs. 18 and 19 also show that the magnitude of Sm is dependent on the type of filler. There are fillers which induce a higher stress (e.g., TiO2, red iron oxide) than others (e.g., CaCO3, talc). This is due to the filler/binder interaction (reinforcing effect), which is determined by the nature and, in particular, by the surface area and the acid/base character of the binder [73]. Examples of Sm = f(PVC) for coatings containing a mixture of fillers are presented in Figs. 20 and 21. An examination of the results obtained with binary and ternary filler coating systems indicates that the stress Smt can be calculated approximately by means of an additive rule x= i
Smt ≈ n1Sm1 + n2Sm2 + $ + ni Sm ≈ ∑ nx Smx
(20)
x =1
where n1, n2, ni = volume fraction of different fillers present in the mixtures, and Sm1, Sm2, SmI = maximum stress of different single filler systems at a given A below or equal to 1 (A = reduced PVC, PVC/ CPVC). For organic coatings cured at high temperatures, such as baking paints, the CPVC can be determined by measuring the stress as a function of temperature of paints formulated at different PVCs. The representation of stress arising at a temperature (in the glassy or rubbery regions) as a function
Fig. 21—Sm (MPa) as a function of PVC (%) for a latex binder containing a TiO2 ( ), calcium carbonate (X), and their mixture: TiO2 (n = 0.4)/CaCl3 (n = 0.6) (□).
t
of PVC enables one to determine the CPVC (Fig. 22). As in the case of coatings formed at room temperature, the CPVC corresponds to the PVC that develops the highest stress. In the literature one can also find simplified methods to determine the CPVC of latex coatings based on the same principle. They simply compare the force [74] or the deflection induced by the internal stress for flexible plastic substrates [75](Fig. 23) coated with paints of different PVCs. These methods can be useful, but one has to be aware that they are valid only if both the thickness of the various paints and the time necessary to reach the maximum stress are the same.
Solvents
The presence of solvents in a coating can affect the magnitude and especially the rate of development of internal stress. This is illustrated in Fig. 24. For thermoplastic binders in solution, the slower the evaporation of a solvent from a coating, the slower is the development of internal stress. The converse also is true. One should note that the coating cast from fast-evaporating solvents (Curve 1, Fig. 24) produces slightly higher stress values than those cast from more slowly evaporating ones (Curves 3 and 4, Fig. 24). The results obtained were explained by using Eq (1), the principle of plasticizing effectiveness of solvents and the stress relaxation favored by the presence in the coating of the slower evaporating solvent [16]. For coatings containing a mixture of solvents (Fig. 25), both the development rate and the stress magnitude are mainly determined by the presence in the film of the less volatile solvent(s). The situation may be different if the film formation is a result of solvent evaporation and cross-linking
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 52
Q
STRESS PHENOMENA IN ORGANIC COATINGS
665
Fig. 22—A polyester/melamine system filled with a mixture of pigments cured at 150°C for 30 min. Left: Stress (S) as a function of temperature (T) at 0 % RH of 7 PVC formulations. Right: Stress (S) as a function of PVC at different temperatures determined during heating or cooling. CPVC (%) = 57±0.5.
processes (e.g., epoxy and polyurethane coatings). Under such circumstances the coating containing faster evaporating solvents can develop smaller stress values [76]. For such coatings, the volume of solvent present in the film after most of the cross-linking has occurred (Eq (2)) and which determines the magnitude of εF will increase the slower the solvent evaporates. If the stress relaxation process is negligible, it follows that for such coating systems the faster the solvent evaporates the smaller the internal stress. The influence of solvents on internal stress is also evident for latex coatings where certain solvents, for example the coalescing agents, play an important role in the film formation process [15,52,77]. Examples of how the level and the type of coalescent affect internal stress are presented in Figs. 26 and 27, respectively. Fig. 26 shows that: (1) the coalescent level affects the time necessary to reach the maximum stress, and (2) for each formulation, there is an optimal coalescent level to obtain a tight continuous coating developing the lowest internal stress. In Fig. 27, the results obtained with three solvents used in latex coatings are given. One can see that the type of solvent influences both the value and the development rate of the internal stress.
As in the case of thermoplastic coatings, the influence of solvents on internal stress development in latex coatings was explained [15,68,75] by taking into consideration the plasticizing effectiveness [18,78], the molar volume, and the steric hindrance of solvents. The plasticizing effectiveness affects the internal stress magnitude, while the molecular dimensions affect the evaporation kinetics and consequently the rate of the stress development.
Binder
The binder is a key component of an organic coating with regard to stress development. To understand the role of the binder, one must consider the general Eq (1) (see also Eqs (3), (5), and (10)) and/or the Tg of a coating with respect to the experimental temperature. Equation (1) indicates that the stress is directly affected by the magnitude of E, ε, and v of the binder. The smaller the values of E, ε, and v, the smaller the magnitude of stress. With respect to Tg, it should be remembered that the binders having their Tg below or close to the film formation temperature (T) develop a negligible internal stress, while those having their Tg > T develop a significant one. This is due to the fact that at T > Tg, the mobility of the binder molecular segments is high and the stress arising during film formation can partially or totally relax. Moreover, it can be shown that for a thermoplastic binder in solution
Fig. 23—Determination of the CPVC by comparing the curvature of painted plastic substrates. Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
666
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 24—Stress (S) as a function of time in a thermoplastic varnish cast from methyl ethyl ketone and ethyl acetate (1); toluene (2); xylene (3), methyl isobutyl ketone and isobutyl acetate (4), at 52 % RH and 21°C.
Fig. 26—Stress (S, MPa) as a function of time (h, hour; d, day) for a styrene acrylic latex paint containing different amounts of Texanol 5 % (X);10 % (○);15 % (+); percent of Texanol calculated by weight of binder solids.
having a Tg > T, the lower the Tg of the binder, the less will be the solvent in the film after its formation (see Eqs (2) and (3)) and therefore the smaller the internal strain and stress in the dry coating [16]. Studies carried out with polyimide coatings, materials usually developing high stresses, indicated that the stress level is determined by morphological and backbone structure [82], polymer chain rigidity and molecular orientation [79]. Ultraviolet-cured coatings, materials that also develop high stresses, were also investigated. It was found that the stress magnitude is mainly affected by photo-initiator concentration, light intensity [80], monomer functionality, monomer chain length [47,83], degree and distribution of cross-linking, and temperature at solidification point [81]. For multifunctional methacrylate coatings a lower stress can develop through lowering light intensity, lowering monomer functionality, [82] raising monomer chain length and plasticizer concentration, and using acrylate monomer instead of methacrylate [83]. Stress development in thermosetting coatings, below and above Tg, is discussed in Ref. [84]. Investigations designed to decrease the stress magnitude by forming low Tg polymer domains [85] or introducing soft polymeric inclusions [86] in the film have been performed. In brief, any change occurring in the molecular structure of a binder (e.g., cross-linking, crystallinity, molecular weight, steric hindrance) may induce a change in E, ε, v, and Tg and thus affect stress development.
STRESS VERSUS ADHESION AND COHESION
Fig. 25—Stress (S, MPa) as a function of time (day) for a thermoplastic varnish cast from toluene (X); isobutyl acetate (○), and their mixtures [1/1 (W/W) ( ) and 1/3 (W/W) (∆)].
t
It is accepted that the stress arising in a coating can reduce the adhesion and cohesion, two crucial properties of an organic coating for obtaining durable coatings [1–10,87–91]. The way the stress affects adhesion is described in detail in Refs. [87,88]. Equation (21) will be used to demonstrate how the stress affects adhesion. This relation describes the factors affecting a measurement of adhesion regardless of the method used (pull-off, peel, torsion, etc.)[91]. WT ≈ adh + WP −
(21)
where: WT = total measured work (energy) used to separate the coating from the substrate, where the substrate can be another coating. In the literature WT is also known as adherence and the measured or practical adhesion; γadh = the interfacial work of adhesion, a consequence of a variety of chemical and/ or physical interactions between the coating and the substrate. It represents the actual physical strength of adhesive bonds, and is known in the literature also as the adhesion or basic and fundamental adhesion;
Fig. 27—Stress (S) as a function of time (h, hour; d, day) for a styrene acrylic latex paint containing Texanol (○); Dalpad A (X); Dalpad A + propylene glycol ( ).
t
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 52
Q
Fig. 28—Stress (S) as a function of temperature (T, °C) for a primer/clearcoat system after different periods of weathering (4 h UV-B at 60°C and 4 h condensation at 50°C) [100].
WP = work (energy) expanded in plastic deformation and viscous dissipation occurring during the adhesion measurement. Depending on the viscoelastic character of the coating this work might be important; β = elastic energy acting against adhesion. It is shown that the application of an energy balance analysis [92] leads to the factor β
= cUr =
cEε 2 1− v
(22)
where Ur, c, and ε are the recoverable strain energy, the film thickness, and the strain.
Fig. 29—Stress (S, MPa) as a function of RH (%) for a primer/ clearcoat system after different periods of weathering (4 h UV-B at 60°C and 4 h condensation at 50°C) [100].
STRESS PHENOMENA IN ORGANIC COATINGS
667
Fig. 30—Stress (S) as a function of time of weathering (tQUV, hour) for a primer/clearcoat system at two climatic conditions [100].
To accommodate the effect of stress [19,28,93], Eqs (1) and (22) are combined to give:
= cSε
(23)
and the combination of Eqs (21) and (23) leads to WT ≈ adh + WP − cSε
(24)
Equation (24) shows, among others that: (1) that in most cases the measured adhesion (WT) is not identical with inter-facial work of adhesion (γadh) one normally expects to measure, and (2) when the product “c S ε” overcomes γadh the coating will spontaneously detach from the substrate; thus, if a high stress arises in a coating, in order to prevent its detachment, it should be applied in a thinner layer.
Fig. 31—Stress (S) as a function of temperature (T) at 5 % (left) and of RH at 21°C (right) for a polyester/melamine system after different periods of weathering (d, day) (4 h UV-B at 60°C and 4 h condensation at 50°C) [100].
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
668
PAINT AND COATING TESTING MANUAL
Q
Fig. 32—Stress (S) as a function of temperature (T, °C) at 5 % (left) and of RH at 21°C (right) for an acrylic latex paint containing coalescing agents after different periods of weathering (d, day) (4 h UV-B at 60°C and 4 h condensation at 50°C).
Because it is energy that expresses the effect of e, tensile or compressive strains are identical in their effect [87]. Since stress develops in most organic coatings, the product “c S ε” should also be considered in adhesion tests performed in laboratories. This fact is taken into consideration in mathematical equations (25) and (26) for pull off and peeling at 90°, respectively [87]. 1/ 2
⎡ 2K ≈⎢ ( − cSε )⎤⎥ c ⎣ ⎦ F ≈ − cSε b
15TH EDITION
Fig. 34—Stress (S, MPa) as a function of RH at 21°C for a polyester/TGIC powder coating after different after different periods of weathering (d, day) (4 h UV-B at 60°C and 4 h condensation at 50°C).
γ = interfacial work of adhesion, F = force applied to peel the coating, and b = width of the coating. Equation (27)(for pull-off measurements of coatings having v = 0.4) is interesting because it takes into consideration the work expanded in plastic deformation (WP) [94].
(25) (26)
where σ = stress applied to pull the coating from the substrate, K = bulk modulus of the coating,
Fig. 33—Stress (S) as a function of temperature (T, °C) at 5 % for a polyester/TGIC powder coating after different periods of weathering (d, day) (4 h UV-B at 60°C and 4 h condensation at 50°C).
=
c 2 + 2.572S2 4.286 E
(
)
(27)
In addition, the study demonstrates the difference between adhesion and adherence [94] and the wrong interpretation of data that might result if only one of them is considered. Equations (22)–(24) also indicate the possibility of determining the adhesion of a coating (i.e., factor γ) without applying any external force. This is due to the fact that at a particular film thickness a spontaneous detachment should occur [87]. Unfortunately, this method can only be used for badly adherent coatings. For all other coatings, extremely high film thicknesses, difficult to apply and cure, would be necessary. When necessary, the product “c S ε” can also be increased by submitting the coated substrate to high stresses, e.g., thermal stresses. If the adhesion forces exceed the cohesive strength of a coating the stresses arising in the coating can provoke cracking, Assuring rather than loss of the coating from the substrate. Such damage can occur at low stress values due to stress concentration in the film, a consequence of the presence of heterogeneity in the coating (e.g., bad pigment/ binder dispersion, presence of impurities, air voids, and variation in film thickness). The verification of this principle can be realized very approximately by determining the stress and the ultimate properties of a coating. Since most organic coatings are viscoelastic, it is essential that these properties be evaluated under the conditions (strain rate, temperature, RH) corresponding to the stress development in the coating. A more quantitative approach of the failure mechanism in organic coatings, but also more difficult to
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 52
Q
STRESS PHENOMENA IN ORGANIC COATINGS
669
Fig. 35—Schematic representation of processes causing crack formation in a coating.
realize, mainly due to relatively low thickness of the films, is the application of fracture mechanics principles [95]. According to certain publications these difficulties are partially solved both theoretically [96,97] and experimentally [98,99]. It is important to add that fatigue processes (e.g., due to variation of temperature and RH) are factors decreasing the overall stress at which the coatings are damaged (delaminates and/or cracks).
DURABILITY AND STRESS DEVELOPMENT Weathering
In most cases when organic coatings are exposed to weathering (accelerated or natural), they undergo chemical and physical modifications which are expressed in the change of Tg, E, α, v, and cross-link density. Under such circumstances and according to Eqs (1), (3), (5), and (10), one can also expect to see changes in stress development. A more detailed discussion of the role played by stress in the coating deterioration is given in Refs. [9,100]. It is shown that depending on the type of coating the weathering can affect the stress development in different ways. In one case weathering provokes: (1) a shift of the curves S = f(T) to higher temperatures indicating an increase of Tg, which in turn induced higher thermal stresses (Fig. 28), and (2) an increase of the slope of the curves S = f(RH) due to an increase of the coating sensitivity to moisture, which induced higher hygroscopic stresses (Fig. 29). The representation of the stress as a function of time of weathering for two experimental conditions similar to those present in the accelerating apparatus (Fig. 30) indicates that every time the experimental conditions changed (every 4 h) the coating was exposed to an increasing hygrothermal stress (tensile thermal stresses under dry conditions and compressive hygroscopic stresses under wet conditions). Another case often met in practice is represented in Fig. 31. In this case weathering induced a relatively small increase of Tg and a significant upward shift of the curves S = f(T) to higher stress values. This upward shift is a consequence of oxidative cross-link reactions the coating undergoes, a fact confirmed by infra red spectra measurements and the stress increase in the rubbery region (Eq (8)). The S = f(RH) curves are similar to the case described above, showing an increase of coating hydrophylicity with weathering.
Fig. 32 is often observed with (acrylic) latex paints containing coalescent solvents. The shift of the curve S = f(T) to higher temperatures, observed during the first days of weathering, is mainly due to loss of coalescent solvents present in the coating. The stress dependence on RH of the unweathered coating is negligible. This is due to its Tg (about 21°C at 0 % RH), close to the experimental temperature (21 ± 1.5°C), a fact that enables a fast relaxation of hygroscopic stresses induced by variation of RH. With coalescing solvents departing from the coating, Tg increases (to about 31°C at 0 % RH) and the hygroscopic stresses become measurable. For organic coatings possessing a good durability it is expected that S = f(T) and S = f(RH) to be little, or not at all affected by weathering. This is illustrated in Figs. 33 and 34. Contrary to other coatings, for this one (a polyester/ triglycidyl isocyanurate (TGIC) powder coating), the dependencies S= f(T) and S = f(RH) are not changing or changing very little with weathering. These results are in agreement with the behavior of the coatings mentioned above under tropical conditions. While the acrylic/polyurethane and the polyester/melamine coatings are cracking after a few years, for the same period, the powder coating is still undamaged. The processes thought to lead to fissuring of the clearcoat are described in Fig. 35. Briefly, the chemical degradations induced by UV, water, and oxygen, which decrease the coating cohesion, combined with the fatigue process at steadily increasing hygrothermal stress levels are the causes of the coating degradation.
Hygroscopic Stress and Corrosion
Due to the heterogeneous nature of most of organic coatings, which possess regions of different hydrophilicity [101], hygroscopic stresses are arising in a coating exposed to a high RH or immersed in water [102]. They participate in formation and/or enlargement of pathways through which water and electrolyte can reach the metallic substrate and thus provoke corrosion. This affirmation is based on the observation that the coatings developing low hygroscopic (compressive) stresses or able to relax them fast, protect well the substrate against corrosion. Among the coatings investigated in one study (Fig. 36), the powder coating and Plexigum developed the lowest hygroscopic compressive stresses and protected well the substrate against the aggressive action of 0.5M NaCl electrolyte. The polyurethane,
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
670
PAINT AND COATING TESTING MANUAL
Q
Fig. 36—Stress (S) as a function of immersion time in water at 21°C for five organic coatings [102]. (With permission from Marie-Odile Barkallah, AFTPVA-EUROCOAT.)
a system which developed high hygroscopic compressive stresses but which relaxed relatively fast, provided only an average protection.
References [1] Gusman, S., “Studies of the Adhesion of Organic Coatings,” J. Paint Technol., Vol. 27, No. 1, 1963, pp. 17–26. [2] Dannenberg, H., “Determination of Stress in Cured Epoxy Resins,” SPEJ., Vol. 21, 1965, pp. 669–675. [3] Prosser, J. L., “Internal Stress Studies,” Mod. Paint Coat., Vol. 67, No. 7, 1977, pp. 47–51. [4] Saarnak, A., Nilsson, E., and Kornum, L. O., “Usefulness of the Measurement of Internal Stresses in Paint Films,” J. Oil Colour Chem. Assoc., Vol. 59, 1976, pp. 427–432. [5] Hamburg, H. R., and Morgans, W. M., Eds., Hess’s Paint Film Defects: Their Causes and Cure, 3rd ed., Chapman and Hall, London, 1979. [6] Croll, S. G., “The Origin of Internal Stress in Solvent-Cast Thermoplastic Coatings,” J. Appl. Polym. Sci., Vol. 23, 1979, pp. 847–885. [7] Sato, K., “The Internal Stress of Coating Films,” Prog. Org. Coat., Vol. 8, No. 2, 1980, pp. 143–160. [8] Perera, D. Y., and Van den Eynde, D., “Considerations on a Cantilever (Beam) Method for Measuring the Internal Stress in Organic Coatings,” J. Coat. Technol., Vol. 53, No. 677, 1981, pp. 39–44. [9] Oosterbroek, M., Lammers, R. J., van der Ven, L. G. T., and Perera, D. Y., “Crack Formation and Stress in Organic Coatings,” J. Coat. Technol., Vol.63, No. 797, 1991, pp. 55–60; Perera, D. Y., and Oosterbroek, M., “Hygrothermal Stress Evolution During Weathering in Organic Coatings,” J. Coat. Technol., Vol. 66, No. 833, 1994, pp. 83–88. [10] Kamarchik, P., Jr., and Jurezak, E. A., “Post-UV Cure Phenomena in Radiation Curable Systems,” Proceedings, Radtech, Edinburgh, Scotland, Great Britain, 1991. [11] Perera, D. Y., and Van den Eynde, D., “Moisture and Temperature Induced Stresses (Hygrothermal Stresses) in Organic Coatings,” J. Coat. Technol., Vol. 59, No. 748, 1987, pp. 55–63. [12] Crackin, F. L., and Bersch, C. F., “Creep Behavior of Transparent Plastics at Elevated Temperatures,” SPEJ., Vol. 15, 1959, pp. 791–796. [13] Bierwagen, G. P., “Film Formation and Mudcracking in Latex Coatings,” J. Coat. Technol., Vol. 51, No. 658, 1979, pp. 117–126. [14] Bauer, C. L., Farris, R. J., and Vratsanos, M. S., “Determination of the Stresses and Properties of Polymer Coatings,” J. Coat. Technol., Vol. 60, No. 760, 1988, pp. 51–55. [15] Perera, D. Y., and Van den Eynde, D., “Internal Stress in Latex Coatings,” J. Coat. Technol., Vol. 56, No. 716, 1984, pp. 111–118.
15TH EDITION
[16] Perera, D. Y., and Van den Eynde, D., “Solvent Influence on theDevelopment of Internal Stress in a Thermoplastic Coating,” J. Coat. Technol., Vol. 55, No. 699, 1983, pp. 37–43. [17] Hansen, C. M., “Polymer Coatings: Concepts of Solvent Evaporation Phenomena,” Ind. Eng. Chem. Prod. Res. Dev., Vol. 9, No. 3, 1970, pp. 282–286. [18] Petersen, C., Heldmann, C., and Johannsmann, D., “Internal Stresses During Film Formation of Polymer Lattices,” Langmuir, Vol. 15, 1999, pp. 7745–7751. [19] Perera, D. Y., “Stress Development and Organic Coatings Performance,” Proceedings, 16th International Conference in Organic Coatings Science and Technology, Athens, Greece, 1990, p. 309. [20] Shimbo, M., Ochi, M., and Arai, K., “Internal Stress of Cured Epoxide Resin Coatings Having Different Network Chains,” J. Coat. Technol., Vol. 56, No. 713, 1984, pp. 45–51; “Effect of Solvent and Solvent Concentration on the Internal Stress of Epoxide Resin Coatings,” J. Coat. Technol., Vol. 57, No. 728, 1985, pp. 93–99. [21] Geldermanns, P., Goldsmith, C., and Bedetti, F., “Measurements of Stresses Generated During Curing and in Cured Polyimide Films,” Proceedings, First International Technical Conference on Polyamides, Society of Plastic Engineers, Ellenville, NY, November 1982. [22] Perera, D. Y., “Cure Characterization of Stoving Paints,” Proceedings, 19th FATIPEC Congress, Aachen, Vol. I, 1988, p. 1; Material Prëfung, Vol. 31, 1989, p. 57. [23] Zosel, A., “Mechanical Behaviour of Coating Films,” Prog. Org. Coat., Vol. 8, 1980, pp. 47–79. [24] Hill, L. W., and Kozlowski, K., “Crosslink Density of High Solids MF-Cured Coatings,” J. Coat. Technol., Vol. 59, No. 751, 1987, pp. 63–71. [25] Yamabe, H., and Funke, W., “Die Bestimmung der Glasumwandlungstemperatur von Beschictungen,” Farbe Lack, Vol. 96, No. 7, 1990, pp. 497–502. [26] Shimbo, M., Ochi, M., Inamura, T., and Inoue, M., “Internal Stress of Epoxide Resin Modified with Spiro Orho-Ester Type Resin,” J. Mater. Sci., Vol. 20, 1985, pp. 2965–2972. [27] Nilsson, E., “Inre Spanningar I Fargskikt,” Farbe Lack, Vol. 21, 1975, pp. 318–332; FALAAA, Vol. 23, 1977, p. 179–185; FALAAA, Vol.23, 1977, pp. 199–210. [28] Perera, D. Y., and Van den Eynde, D., “Utilisation de la Mesure de la Contrainte Interne Pour la Caracterisation des Revêtements Organiques,” Proceedings, 16th FATIPEC Congress, Liege, Belgium, Vol. 1, 1982, p. 129. [29] Perera, D. Y., and Van den Eynde, D., “Stress in Organic Coatings Under Wet Conditions,” Proceedings, 20th FATIPEC Congress, Nice, 1990, p. 125. [30] De Waard, R., Stock, C. R., and Alefrey, T., Jr., “Measurement of Residual Stress in Thermostting Plastics,” ASTM Bull, April, 1952, pp. 53–59. [31] Zubov, P. J., Lepilkina, L. A., Gilman, T. P., and Leites, A. Z., “Investigation of Internal Stresses on Hardening of Polyester Resin,” Colloid J., Vol. 23, No. 5, 1961, pp. 469–472. [32] Imamura, H., “The Internal Stresses in Epoxy Resin Cured on Wood,” Mokuzai Gakkaishi, Vol. 16, 1970, pp. 168–172; “Internal Stress in Paint Film,” Mokuzai Gakkaishi, Vol. 19, 1973, pp. 89–93 and 393–398; “Influence of Surface of WoodSubstrate on Internal Stress in Paint Film,” Mokuzai Gakkaishi, Vol. 22, 1976, pp. 325–330 and 331–336. [33] Theocaris, P. S., and Paipetis, S. A., “Shrinking Stresses in Three Dimensional Two Phase Systems,” J. Strain Anal., Vol. 8, 1973, pp. 286–293. [34] Shimbo, M., Ochi, M., and Shigeto, Y., “Shrinkage and Internal Stress During Curing of Epoxide Resins,” J. Appl. Polym. Sci., Vol. 26, 1981, pp. 2265–2277. [35] Association Belge pour l’Etude, l’Essai et l’Emploi des Matériaux (ABEM), “Cours D’initiation à L’analyse des Contraintes,” Bruxelles, 1973. [36] Kanno, A., and Murato, Y., “Study on Brittle Lacquer Coating of Air-Drying Type,” Proceedings, 15th Japanese Congress on Materials Research, Japan, 1972, p. 177. [37] Nakamae, K., Nishino, T., Airu, X., Matsumoto, T., and Miyamoto, T., “Studies on Mechanical Properties of Polymer
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 52
[38] [39] [40] [41] [42] [43] [44]
[45] [46] [47] [48] [49] [50] [51]
[52] [53] [54] [55] [56]
[57]
[58]
Q
Composites by X-Diffraction,” J. Appl. Polym. Sci., Vol. 40, 1990, pp. 2231–2238; Nishino, T., Airu, X., Matsumoto, T., Matsumoto, K., and Nakamae, K., “Residual Stress in Particulate Epoxy Resin by X-Ray Diffraction,” J. Appl. Polym. Sci., Vol. 45, 1992, pp. 1239–1244; Nishino, T., Koter, M., Inayoshi, N., and Nakamae, K., “Residual Stress in Microstructures of Aromatic Polyimide with Different Imidization Processes,” Polymer, Vol. 41, 2000, pp. 6913–6918. Sanzharovskii, A. T., “Internal Stresses in Coatings,” Vysokochist. Veshchestva, Vol. 2, No. 11, 1960, pp. 1698–1702, 1703–1708, 1709–1714. Gusman, S., “Studies of the Adhesion of Organic Coatings,” Paint Technol., January 1963, pp. 17–26. Simpson, W., and Boyle, D. A., “The Observation of Stress Development in Supported Films,” J. Oil Colour Chem. Assoc., Vol. 46, 1963, pp. 331–338. Corcoran, E. M., “Determining Stresses in Organic Coatings Using Plate Beam Deflection,” J. Paint Technol., Vol. 41, No. 538, 1969, pp. 635–640. Inoue, Y., and Kobtake, Y., “Effects of Fillers on Residual Stresses in Coatings,” Kolloid-Z., Vol. 159, 1958, pp. 18–24. Aronson, P. D., “Some Aspects of Film Formation in Emulsion Paints,” J. Oil Colour Chem. Assoc., Vol. 57, 1974, pp. 66–82. Croll, S. G., “Internal Stress in a Solvent-Cast Thermoplastic Coating,” J. Coat. Technol., Vol. 50, No. 638, 1978, pp. 33–38; “Residual Stress in Solventless Amine-Cured Epoxy Coating,” J. Coat. Technol., Vol. 51, No. 659, 1979, pp. 49–55. Croll, S. G., “An Overhanging Beam Method for Measuring Internal Stress in Coatings,” J. Oil Colour Chem. Assoc., Vol. 63, 1980, pp. 271–275. O’Brien, R. N., and Michalik, W., “Laser Interferometry Method for Internal Stress Measurement in Coatings,” J. Coat. Technol., Vol. 58, No. 735, 1986, pp. 25–36. Francis, L. F., McCormick, A. V., Vaessen, D. M., and Payne, J. A. “Development and Measurement of Stress in Polymer Coatings,” J. Mater. Sci., Vol. 37, 2002, pp. 4717–4731. Elsner, G., “Residual Stress and Thermal Expansion of SpunOn Polyimide Films,” J. Appl. Polym. Sci., Vol. 34, No. 1, 1987, pp. 815–828. Olsen, G. H., and Ettenberg, M., “Calculated Stresses in Multilayered Heteroepitaxial Structures,” J. Appl. Phys., Vol. 48, No. 6, 1977, pp. 2543–2547. Vilms, J., and Kerps, D., “Simple Stress Formula for Multilayered Thin Films on a Thick Substrate,” J. Appl. Phys., Vol. 53, No. 3, 1982, pp. 1536–1537. Townsed, P. H., Barnett, D. M., and Brunner, T. A., “Elastic Relationship in Layered Composite Media with Approximation for the Case of Thin Films on a Thick Substrate,” J. Appl. Phys., Vol. 62, No. 11, 1987, pp. 4438–4444. Suhir, E., “An Approximation Analysis of Stresses in Multilayered Elastic Films,” J. Appl. Mech., Vol. 55, No. 1, 1988, pp. 143–148. Boerman, A. B., and Perera, D. Y., “Measurement of Stress in Multi-Coat Systems,” J. Coat. Technol., Vol. 70, No. 881, 1984, pp. 69–75. Struik, L. C. E., Physical Aging in Amorphous Polymers and Other Materials, Elsevier, Amsterdam, Holland, 1978. Kovacs, A. J., “Transition Vitreuse Dans les Polymères Amorphes. Etude Phénoménologique,” Fortschr. Hochpolym.Forsch., Vol. 3, 1963, pp. 394–507. K. L. Ngai and G. B. Wright, Eds., Relaxations in Complex Systems, Naval Research Laboratory, Washington DC, 1984. (The book contains survey and review of relaxation phenomena in: viscous liquids and glasses; polymers; metallic glasses; spin glasses; electrical relaxations in ionic conductors; and theory and models.) Mangion, M. B. M., and Johari, G. P., “Relaxation of Thermosets. III—Sub-Tg Dielectric Relaxations of Bisphenol-A-Based Epoxide Cured with Different Cross-Linking Agents,” J. Polym. Sci., Part B: Polym. Phys., Vol. 28, 1990, pp. 71–83. Lee, A., and Mc Kenna, G. B., “Effect of Crosslink Density on Physical Ageing of Epoxy Networks,” Polymer, Vol. 31, 1990, pp. 423–430.
STRESS PHENOMENA IN ORGANIC COATINGS
671
[59] Perera, D. Y., and Schutyser, P., “Effect of Physical Aging on Thermal Stress Development in Powder Coatings,” Prog. Org. Coat., Vol. 24, 1994, pp. 299–307. [60] Perera, D. Y., and Schutyser, P., “Physical Aging in Organic Coatings,” Procedure, 22nd FATIPEC—Congress, Budapest, Hungary, Vol. 1, 1994, p. 25. [61] Perera, D. Y., Schutyser, P., de Lame, C., and Van den Eynde, D., “Film Formation and Physical Ageing in Organic Coatings,” ACS Symp. Ser., Vol. 648, 1996, pp. 210–221. [62] van der Linde, R., Belder, E. G., and Perera, D. Y., “Effect of Physical Aging and Thermal Stress on the Behavior of Polyester/TGIC Powder Coatings,” Prog. Org. Coat., Vol. 40, 2000, pp. 215–224. [63] Kris, G. J., and Sanzharovskii, A. T., “Influence of Pigmentation on Internal Stress,” Lakokrasochnye Materialy i Primenenie, Vol. 3, 1970, pp. 27–33. [64] Haagen, H., “Untersuchen uber Ursachen und Zusammenhang von Schwundvorgangen Mit Inneren Spannungen in Beschichtungen,” Farbe Lack, Vol. 85, No. 2, 1979, pp. 94–100. [65] Croll, S. G., “The Effect of Titania Pigment on the Residual Strain, Glass Transition and Mechanical Properties of PMMA Coating,” Polymer, Vol. 20, No. 11, 1979, pp. 1423–1430. [66] Perera, D. Y., and Van den Eynde, D., “Internal Stress in Pigmented Thermoplastic Coatings,” J. Coat. Technol., Vol. 53, No. 678, 1981, pp. 40–45. [67] Perera, D. Y., and Van den Eynde, D., “Effect of Pigmentation on Internal Stress in Latex Coatings,” J. Coat. Technol., Vol. 56, No. 717, 1984, pp. 47–53. [68] Perera, D. Y., and Van den Eynde, D., “Internal Stress and Film Formation in Emulsion Paints,” J. Oil Colour Chem. Assoc., Vol. 11, 1985, pp. 275–281. [69] Bierwagen, G. P., “CPVC Calculations,” J. Paint Technol., Vol. 44, No. 574, 1972, pp. 45–55. [70] Bierwagen, G. P., and Mallinger, R. G., “Comparison of Prediction and Experimental for the Critical Pigment Volume Concentration in Thermoplastic Coatings,” J. Coat. Technol., Vol. 54, No. 690, 1982, p. 73. [71] Sato, K., “The Mechanical Properties of Filled Polymers,” Prog. Org. Coat., Vol. 4, 1976, pp. 271–302. [72] Bierwagen, G. P., and Hay, T. K., “The Reduced Pigment Volume Concentration as an Important Parameter in Interpreting and Predicting the Properties of Organic Coatings,” Prog. Org. Coat., Vol. 3, 1975, pp. 281–303. [73] Toussaint, A., and D’Hont, L., “Ultimate Strength of Paint Films,” J. Oil Colour Chem. Assoc., Vol. 64, 1981, pp. 302–307. [74] Helmen, T., and Strauch, D., “Ermittlung der KPVK an Dispersionsanstrichen durch Messung der Inneren Spannung,” Farbe Lack, Vol. 96, No. 10, 1990, pp. 769–772. [75] Dörr, H., and Holzinger, F., “Le Dioxyde de Titane KRONOS dans les Peintures-Émulsion,” Kronos International, Inc., Leverkusen, Germany, 1990. [76] Croll, S. G., “Effect of Solvent on Residual Strain in Clear Epoxy Coatings,” J. Oil Colour Chem. Assoc., Vol. 63, 1980, pp. 230– 236; “Residual Strain due to Solvent Loss from a Crosslinked Coating,” J. Coat. Technol., Vol. 53, No. 672, 1981, pp. 85–92. [77] Perera, D. Y., and Van den Eynde, D., “Effect of Organic Solvents on Internal Stress in Latex Coatings,” J. Coat. Technol., Vol. 56, No. 718, 1984, pp. 69–75. [78] Hansen, C. M., “The Free Volume Interpretation of Plasticizing Effectiveness and Diffusion of Solvents and Plastizers in High Polymers,” Off. Dig., Vol. 37, No. 480, 1965, pp. 57–77. [79] Ree, M., Chu, C. W., and Goldberg, M. J., “Internal Stress Behavior in Polymer Thin Films: Effects of Polymer Chain Rigidity and Orientation,” Proceedings, ACS (PMSE), Vol. 69, 1993, p. 312. [80] Payne, J. A., Francis, L. F., and McCormick, A. V., “The Effects of Processing Variables on Stress Development in Ultraviolet— Cured Coatings,” J. Appl. Polym. Sci., Vol. 66, 1997, pp. 1267–1277. [81] Stolov, A. A., Xie, T., Penelle, J., and Hsu, S. L., “Stress Buildup in Ultraviolet—Cured Coatings: Sample Thickness Dependence,” Macromolecules, Vol. 34, 2001, pp. 2865–2869. [82] Chung, H., Joe, Y., and Han, “The Effect of Curing History on the Residual Stress Behavior of Polyimide Thin Films,” J. Appl. Polym. Sci., Vol. 74, 1999, pp. 3287–3298.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
672
PAINT AND COATING TESTING MANUAL
Q
[83] Wen, M., Scriven, L. E., and McCormick, A. V., “Differential Scanning Calorimetry and Cantilever Deflection Studies of Polymerization Kinetics and Stress in Ultraviolet Curing of Multifunctional (Metha) Acrylate Coatings,” Macromolecules, Vol. 35, 2002, pp. 112–120. [84] Lange, J., Toll, S., Manson, J-A. E., and Hult, A., “Residual Stress Build-Up in Thermoset Films Cured Above Their Ultimate Glass Transition Temperature,” Polymer, Vol. 36, No. 16, 1995, p. 3135–3141. [85] Nakamura, Y., Yamaguchi, M., Kazuo, I., Masayoshi, O., and Matsumoto, T., “Internal Stress of Epoxy Resin Modified with Acrylic Polymers Containing Functional Groups Produced by In Situ U. V. Radiation Polymerization,” Polymer, Vol. 31, 1990, pp. 2066–2070. [86] Piens, M., and De Deurwaerder, H., “Effect of Coating Stress Adherence and on Corrosion Prevention,” Prog. Org. Coat., Vol. 43, 2001, pp. 18–24. [87] Croll, S. G., “Adhesion Loss Due to Internal Strain,” J. Coat. Technol., Vol. 52, No. 665, 1980, pp. 35–43. [88] Croll, S. G., “Correlation of the Adhesion of Polystyrene Lacquer to Inorganic Substrates and Their Wetting Characteristics,” J. Oil Colour Chem. Assoc., Vol. 63, 1980, pp. 200–209. [89] Pierce, P. E., and Schoff, C. K., “Coating Film Defects,” Federation of Societies for Coatings Technology, Philadelphia, PA, 1988. [90] Schmid, E. V., “Glass-Transition of Coatings,” Polym. Paint Colour J., Vol. 180, No. 4258, 1990, pp. 212–217. [91] Farris, R. J., Maden, M. A., and Goldfarb, J., “The Determination of Residual Stresses and Peel Strengths in Polymer Coatings as an Aid in Predicting Failure by Delamination,” Proceedings, The Adhesion Society, 14th Annual Meeting, Clearwater, FL, 1991, p. 138; Farris, R. J., and Goldfarb, J., “An Experimental Partitioning of the Mechanical Energy Expended During Peel Testing,” J. Adhes. Sci. Technol., Vol. 7, No. 8, 1993, pp. 853–868.
15TH EDITION
[92] Kendall, K., “The Adhesion and Surface Energy of Elastic Solids,” J. Phys. D, Vol. 4, 1971, pp. 1186–1195; “Shrinkage and Peel Strength of Adhesive Joints,” Vol. 6, 1973, pp. 1782–1787. [93] Perera, D. Y., “Mieux Formuler par la Maitrise de la Contrainte Interne,” Proceedings of 17th FATIPEC Congress, Lugano, Switzerland, Vol. 1, 1984, p. 13. [94] Piens, M., and De Deurwaerder, H., “Effect of Coating Stress on Adherence and Corrosion Prevention,” Proceedings, 26th International Conference in Organic Coatings Science and Technology, Athens, Greece, 2000. [95] Kinloch, A. J., Young, R. J., Fracture Behavior of Polymers, Applied Science Publishers, Essex, England, 1993. [96] Beuth, J. L., “Cracking of Thin Bonded Films in Residual Tension,” Int. J. Solids Struct., Vol. 29, 1992, pp. 1657–1675. [97] Thouless, M. D., Olsson, E., and Gupta, A., “Cracking of Brittle Films on Elastic Substrates,” Acta Metall. Mater., Vol. 40, 1992, pp. 1287–1292. [98] Hashemi, S., “Fracture Toughness Evaluation of Ductile Polymeric Films,” J. Mater. Sci., 32, 1997, pp. 1563–1573. [99] Nichols, M. E., Darr, C. A., Smith, C. A., Thouless, M. D., Fischer, E. R., “Fracture Energy of Automotive Clearcoats: I, Experimental Methods and Mechanics,” Polym. Degrad. Stab., Vol. 60, 1998, pp. 291–299. [100] Perera, D. Y., “Role of Stress on Durability of Organic Coatings,” in Plastics and Coatings (Durability-Stabilization Testing), R. Ryntz, Ed., Hanser Publishers, Munich, Chap. 6, 2001. [101] Nguyen, T., Hubbard, J. B., and Pommersheim, J. M., “A Unified Model for the Degradation of Organic Coatings on Steel in a Neutral Electrolyte,” J. Coat. Technol., Vol. 68, No. 855, 1996, pp. 45–56. [102] Perera, D. Y., and Nguyen, T., “Hygroscopic Stress and Failure of Coating/Metal Systems,” Proceedings of Eurocoat, Genova, Italy, 1996, p. 1.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
53
MNL17-EB/Jan. 2012
Friction and Slip Resistance Joseph V. Koleske1 PREFACE
In preparation of this chapter, the contents of Chapter 50 Slip Resistance in the fourteenth edition were drawn upon. The author acknowledges the author of the chapter in the fourteenth edition, Paul R. Guévin, Jr. The current edition will review and update the topics as addressed by the previous author, introduce new technology, and include up-todate references.
INTRODUCTION
FRICTION IS AN INTERESTING PROPERTY THAT IS often taken for granted. For example, it would not be possible to walk if there were no friction between shoe soles and the ground or other surfaces. There is also an awareness that friction varies with the nature of the shoe bottoms— rubber, plastic, leather, etc.—and of the surface—ground, concrete, painted concrete, ice, wood floor, polished floor, steel, etc. Automobiles would not move if it were not for the friction between tires and road surface, would not slow down if it were not for the friction between brake discs or drums, and would not stop if it were not for the friction between tires and road surface. Without friction, dishes would not stay on a table and the slightest jostling would cause items to slide. Many other examples could be given of friction and slip. At times, conditions of high friction are desirable and other times low friction is needed. If a heavy object is to be slid, low friction is useful. Skiers want low friction between their skis and snow or ice and use waxes as a lubrication to decrease friction. In the cases of polished, wet, or icy surfaces, frictional forces are quite small and automobiles, people, and animals have a tendency to slip. In the case of dry concrete, neat or painted, or wood, the frictional forces are quite high and the surface is stable and with it less chance for slipping. It should be kept in mind that there is a need for material combinations that result in high friction, as in a painted or otherwise coated walkway area where a person walking or running on the surface does not want to slip and fall, and in low friction, as when an object is to move smoothly and evenly by another object of similar or different composition, as when sheets of paper move by one another in a copy machine. Friction can be increased by use of additives such as sand on an icy surface and decreased by the use of lubricants such as oil, water, graphite, and the like. Slip is a word that has many meanings [1–3], but it is obvious that slip occurs when friction is low. In certain areas of the coating industry, slip is considered to be the ability of an object to move in a relatively free or unencum1
bered but controlled manner as when one sheet of coated or uncoated metal passes over another in a coating-line feeding operation, when such sheets move along a conveyor system in coating, printing, fabrication, or packaging operations, and the like. The material or coated material must have proper “slip” or “lubricity” to allow passage through various sections of the coating fabrication or packaging section of the line. Slip is also important in fabricating operations wherein the coated object is in contact with a tooling system as in fabrication of bottle caps, in bending or forming operations, and so on. In various textile lines, proper slip is need for threads to flow smoothly and uniformly through fabrication systems and not cause thread or yarn breaks and defective final products or lost production.
CONCEPTS OF FRICTION
In a technical sense, friction is the force between contacting objects that resists the relative motion of the objects. Starting or static friction is the frictional force between stationary objects, and dynamic, kinetic, or sliding friction is the frictional force between objects that are sliding with respect to one another. Consider a block resting on a solid surface as in Fig. 1. The upward force, FU, of the solid surface, counterbalances the weight or mass of the block, FN, and the block is at rest. FN is usually referred to as the perpendicular or normal force. If a force, FA, is applied to the block and parallel to the solid surface, the force of friction, FF, resists it. If FA is slowly increased, at some point the block will begin to move and slide along the solid surface without further increase in FA and actually with a small decrease in it [4,5]. Thus, the force required to move an object from rest is greater than the force required to continue the movement at constant velocity. At this point, FA = FF
(1)
By sliding blocks of the same material with equally smooth surfaces but different sizes (i.e., different contact areas) across the same surface, it has been found that the ratio of the frictional force to the mass of the block or FN is constant. The proportionality factor between the frictional resistance force and the normal force is known as the coefficient of friction, μ, with a distinction made between the coefficient of static friction, μS, and coefficient of kinetic friction, μK, as indicated in the following relationships: FF = μSFN FF = μkFN
stationary
(2)
sliding
(3)
1513 Brentwood Road, Charleston, WV 25314-2307.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed byASTM International This standard is for EDUCATIONAL USE ONLY. Copyright © 2012 by www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
673
674
PAINT AND COATING TESTING MANUAL
Q
Fig. 1—A solid block resting on a solid surface with an upward force, FU, and a normal or downward force, FN, acting on it. A force, FA, acting parallel to the solid surface is applied to the block and a friction force, FF, opposes the force FA.
This points out that it should be stated whether a measured or cited coefficient of friction is for the static or the kinetic condition, however, quite often this is not done in the literature as will be illustrated below. The friction force is often discussed in terms of an object sliding down a solid inclined surface as indicated in Fig. 2. When the angle of incline, θ, is 0°, the situation is that of Fig. 1, which becomes Fig. 2 as the angle of incline is increased from zero to some finite value less than 90°. As the angle of the incline increases, the force component tending to move the object down the incline, FA, also increases. At some point, the slightest increase in the angle will sufficiently increase the force component parallel to the plane of the incline, FA, to a value such that FF is overcome and, if unrestrained, the object begins to move down the incline at constant speed. Thus, the object does not slide until the force parallel to the incline, FA, reaches a certain value that is equal to the frictional force, FF. The mass of the object represented by vector FM can be resolved into vectors FA and FN parallel and normal to the plane of the incline, respectively. At the described condition of sliding at constant speed, FA = FF. Since FF = μK FN, substitution leads to FA = μK FN or
μK = FA/FN
(4)
From the vector diagram in Fig. 2, it is apparent that FA/FN = tan θ = μK
(5)
θ = arc tan μK
(6)
and
Thus, if μK is known, the angle of incline, at which a block placed on the incline will slide regardless of the weight or mass of the block, can be determined. Note that the coefficient of friction can take on values from zero to extremely high values, but usually it has a value of between zero and
15TH EDITION
one unless special conditions are involved as, for example, in the case of gage blocks, which are ultrasmooth, and in contacting spectroscopically pure metals such as copper that were outgassed in a vacuum prior to admitting other gases such as nitrogen, oxygen, or hydrogen. In physics, the study of friction is called tribology. In a physics sense, friction is the resisting force that takes place when two surfaces contact and rub against each other. When investigating the causes of friction, it should be accepted that the surface of materials is microscopically irregular and is not flat on an atomic scale. Although surfaces may appear to be smooth, on a microscopic scale they are covered with asperities and are jagged and rough. The uneven or rough surfaces of the touching objects are thought to be one of the main causes of friction. As the surfaces move or tend to move, the rough surfaces interlock and offer resistance to being moved over one another until a certain value of applied force is reached. From this model of rough surfaces interlocking, friction would be expected to decrease as the surfaces are smoothened. This is the actual state of affairs—up to a point. If the surfaces are made increasingly smoother, at some point friction will increase.2 This phenomenon, which is another of the main causes of friction, is thought to be due to the bonding or interaction between surface atoms and these bonds must be broken before friction is overcome and sliding can take place. Such bonding will only take place where the surface atoms can come sufficiently close so as to be within each other’s force fields. Thus with usual surfaces, it is only asperities on one surface that make contact with asperities on the other surface that can “bond” and interact through van der Waals or other chemical forces with adhesion bonds. The entire normal force is carried on the tips of the asperities and the true or actual area of contact is much less than the apparent or measured area. Local stress on the contacting asperities may result in a flattening of the asperities and produce an area of minute but intimate contact. With metals, the deformation may be within the elastic limit, but with plastics the viscoelastic deformation may be permanent or a hysteresis effect may be involved. As the surfaces are made increasingly smoother, at some point the overall contact area between them increases and as a result, bonding and resultant friction forces will increase. To move a surface over another surface, the adhesive bonds at the interface must be broken in a shear plane near the surface of one of the materials. The shear fracture at the minute interaction areas or area of true contact between the dry contacting surfaces is a part of the mechanism of dry friction theory. If it is assumed that the interaction areas are more or less randomly distributed over the apparent contact area, the friction force can be expressed as Gage blocks are probably the best example of smooth surfaces that have high friction. These devices are used to provide linear dimensions to within given and very small tolerances—reference blocks that can have a high tolerance of ±0.000 002 in. or ±0.000 05 mm are built up to make the desired linear dimension. To develop a particular length, the appropriate blocks are slid together and held together by “forces” akin to the forces between molecules or atoms. Assembled blocks cannot be left together for long periods of time or they will become permanently joined through the forces. See www.claymore.engineer.gvsu.edu or www.ptb.de/en/org/5/54/543/ parendm_e.htm as well as other sites for detailed information.
2
Fig. 2—A solid block resting on an incline. As the angle θ increases, the force, FA, increases until the friction force, FF, is overcome and the block will begin to slide.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 53
FF = σG · A
(7)
where A is the actual area of contact and σG is the average shear strength of the interaction areas. Since it is the interaction areas that actually carry the normal force, FN, between the surfaces, the following relationship also exists. FN = ρm · A
(8)
where ρm is defined as the flow pressure or shear stress of the softer material in the vicinity of local areas of actual contact. If the last two equations are combined, the coefficient of friction is obtained as
μ = FF/FN = σG/ρm
(9)
and the coefficient is independent of the contact area. Friction coefficients involving true area of contact have also been expressed in accord with the following relationships [6]. For metals, the deformation of asperities to form the true area of contact is approximately proportional to the normal force, FN. In the case of polymeric materials, rubbers undergo an essentially elastic deformation and the area of contact increases with (FN)2/3 indicating that the coefficient of friction decreases with increasing FN. When polymers that deform in a viscoelastic manner, the friction force follows a relationship that has the form FF = k(FN)X
(10)
with k a constant. Since FF = μ/FN, the coefficient of friction may be expressed as
μ = k(FN)X–1
(11)
with 2/3 < X < 1. In the case of polytetrafluoroethylene, ithas been found that X has a value of 0.85 for over ten orders of magnitude change in FN. In a general sense, both mechanisms, the interlocking and the bonding mechanisms, play a major role in causing friction. Other mechanisms that may be involved in friction forces include, a gouging effect in which asperities on a hard material may gouge a grove in a softer material as moving objects drag and grind against each other; a roughness effect wherein the asperities interlock and there is a need to lift some asperities over other asperities; electrostatic effects in which work must be done to separate charged regions on the sliding surfaces—an effect that may be seen with electric insulators; and hysteresis effects in elastic and plastic deformations at the touching points or near them that take place with not all of the deformation work recoverable. Heat generation and dissipation account for the non-recoverable work. Values for the coefficient of friction for plastics will depend on test conditions such as humidity, material degree of cure, moisture content, and temperature. For example, the coefficient of friction for nylon will vary from 0.91 to 1.19 when the moisture content is varied from 0.2 to 10 % [7]. Table 1 has coefficients of static friction, μS for various synthetic and natural polymers when in contact with the same polymer and with steel. In general, μS is greater for these polymers when sliding on the same polymer than when sliding on steel. The values are abstracted
Q
FRICTION AND SLIP RESISTANCE
675
TABLE 1—Coefficient of Static Friction, μS, for various synthetic and natural polymers on self and on steel [8] Materials in Contact
Condition
μS
Poly(methyl methacrylate) on poly(methyl methacrylate)
Clean
0.8
Poly(methyl methacrylate) on steel
Clean
0.4–0.5
Polystyrene on polystyrene
Clean
0.5
Polystyrene on steel
Clean
0.3–0.35
Polyethylene on polyethylene
Clean
0.2
Polyethylene on steel
Clean
0.2
Polytetrafluoroethylene on polytetrafluoroethylene
Clean
0.04
Polytetrafluoroethylene on steel
Clean
0.04
Nylon on nylona
Clean
0.15–0.25
Silk on silk
Cleanb
0.2–0.3
Cotton thread on cotton thread
Cleanb
0.3
Type nylon not specified. Commercially clean.
a
b
from a handbook [8] that contains many more values of the coefficients of friction including the static coefficients for a variety of polymers when in contact with wet and dry snow at 0°C as given in Table 2. In general, the coefficients are lower on dry snow indicating that it should be easier to move an object on dry snow than on wet snow. As the temperature of dry snow decreases from 0°C to −32°C, Table 2, μS increases with the effect dependent on the nature of the polymer. A range of values for the coefficient of sliding friction for a number of polymers sliding on a variety of unnamed substrates is given in Table 3. Examination of the data in Tables 1–3 indicates that polytetrafluoroethylene has a very low coefficient of friction, and it is widely used to make slippery or low friction surfaces. But, the polymer can quite easily erode and wear when used in systems that have repetitious operation. The wear factor can be improved without compromising its frictional properties when it is filled with certain hard particulate solids. However, hard fillers can have a deleterious effect on an opposing surface. It has been found the improved wear while maintaining low friction or a high degree of slipperiness can be obtained without affecting the opposing surface if the polymer is irradiated with an electron beam [9]. Table 4 has values of the coefficients of kinetic friction for Nylon 6-6 in two states of smoothness (molded or machined surface) sliding against the same surface, against the opposite surface, or against mild steel as well as mild steel sliding against itself or the two nylon surfaces. The data indicate the machined nylon surface, μK = 0.46, is smoother than the molded nylon surface, μK = 0.63. It also appears that friction is less when the smoother machined nylon surface is moving on the rougher molded nylon surface than vice versa. There is very little, if any, difference in friction when either nylon surface is moving on mild steel, μK = 0.31 and 0.33, or when a mild steel surface in moving on either nylon surface, μK = 0.41. The strong effect of
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
676
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
TABLE 2—Coefficient of static friction, μS, for natural polymers under various conditions and for synthetic polymers at certain temperatures on wet and dry snow [8]. Material
Condition
Temperature, °C
μS
Leather on
Metala
. . .
0.6
Leather on
Greasy metala
. . .
0.2
Leather on
Wet metala
. . .
0.4
Leather on
Wood
. . .
0.3–0.4
Nylonb on
Wet snow
0
0.4
Nylonb on
Dry snow
0
0.3
Nylonb on
Dry snow
−10
0.3
Paraffin wax on
Wet snow
0
0.06
Paraffin wax on
Dry snow
0
0.06
Paraffin wax on
Dry snow
−10
0.35
Paraffin wax on
Dry snow
−32
0.4
Polyethylene terephthalate on
Wet snow
0
0.5
Polyethylene terephthalate on
Dry snow
0
0.35
Polyethylene terephthalate on
Dry snow
−10
0.38
Poly(methyl methacrylate) on
Wet snow
0
0.5
Poly(methyl methacrylate) on
Dry snow
0
0.3
Poly(methyl methacrylate) on
Dry snow
−10
0.34
Poly(methyl methacrylate) on
Dry snow
−32
0.4
Polytetrafluoroethylene on
Wet snow
0
0.05
Polytetrafluoroethylene on
Dry snow
0
0.02
Polytetrafluoroethylene on
Dry snow
−10
0.08
Polytetrafluoroethylene on
Dry snow
−32
0.1
Ski lacquer on
Wet snow
0
0.2
Ski lacquer on
Dry snow
0
0.1
Ski lacquer on
Dry snow
−10
0.4
Ski lacquer on
Dry snow
−32
0.4
Commercially clean, unspecified metal. Type nylon not specified.
a
b
lubricants, which will be discussed later in this chapter, is apparent from certain data in Table 4. Shearing processes as described above depend on the viscoelastic nature of polymers and therefore are a function of strain rate (sliding speed) and temperature. It has been demonstrated that the coefficient of friction for crystalline polymers such as polyethylene, polypropylene, and Nylon 6-6 depends markedly on sliding speed and passes through a maximum [6,11]. Amorphous polymers such as poly(methyl methacrylate) and polystyrene only show moderate increases in μK and only a slight maximum as strain rate increases. As the temperature is increased, the dependency of friction coefficient curves is shifted to the right or to higher sliding speeds. Grosch [12] found similar effects for μK when a variety of rubbers (natural, butyl, styrenebutadiene, and acrylonitrile-butadiene elastomers) sliding
on glass was examined as a function of sliding speed over a broad range of temperatures. The viscoelastic nature of friction is further demonstrated by subjecting the μK, sliding speed, temperature data to the Williams–Landel–Ferry reduced variables analysis [13]. The shift factor, log aT, fits a continuous master curve when reduced to a reference temperature. Others [14–16] have obtained similar results that demonstrate the frictional characteristics of elastomeric materials behave in a viscoelastic manner. Polymers are used in many facets of the electronics industry, and frictional characteristics can play an important role in their usage. Coefficients of friction for some polymers used in this industry and for two lubricants used are given in Table 5 [17]. Friction can cause tribolelectric charging of polymers by contact and rubbing. Since polymers have high resistivity, electrostatic charge can accu-
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 53
TABLE 3—Ranges of coefficients of sliding friction, μK,for various materials sliding on various unnamed surfaces [6]. Material
μK
Natural rubber
0.5–3.0
Nylon 6-6
0.15–0.40
Poly(ethylene terephthalate)
0.20–0.30
Poly(methyl methacrylate)
0.25–0.50
Poly(vinyl chloride)
0.20–0.90
Poly(vinyl fluoride)
0.10–0.30
Poly(vinylidine chloride)
0.68–1.80
Polybutadiene
0.4–1.5
Polyethylene—High Density
0.08–0.20
Polyethylene—Low Density
0.30–0.80
Polypropylene
0.67
Polystyrene
0.33–0.5
Polytetrafluoroethylene
0.04–0.15
Styrene-butadiene copolymer
0.5–3.0
mulate in large amounts and represent serious, hazardous consequences. For example, discharge between dissimilar pairs can cause a spark and a resultant fire or destruction of electronic equipment. A tribolelectric series for a number of widely used polymers is given in Table 6 [6]. Antistatic additives can decrease resistivity, and leakage currents can be used to dissipate an accumulation of electrostatic charge from friction.
ASTM DEFINITIONS/TERMINOLOGY
Table 7 contains a number of specific definitions or terminology that can be found in the ASTM publications indi-
Q
FRICTION AND SLIP RESISTANCE
677
cated in the table. Although the definitions may differ in the exact wording, they are in basic agreement with each other and with the information given above. Terminology related to that in Table 7 may be found in ASTM D1436 Test Method for Application of Emulsion Floor Polishes to Substrates for Testing Purposes in ASTM D2825 Terminology Relating to Polishes and Related Materials, as well as in ASTM and other standards. ASTM G115-10 Standard Guide for Measuring and Reporting Friction Coefficients contains schematic diagrams for more than 20 device/test configurations for friction testing, the applicable materials for the particular test, and the measured parameters for the standards listed in Table 8. It is a useful guide for those involved in a friction and its measurement. Two other definitions included in this standard are: triboelement—one of two or more solid bodies that comprise a sliding, rolling, or abrasive contact, or a body subjected to impingement or cavitation. Each triboelement contains one or more tribosurfaces; and tribosystem—any system that contains one or more triboelements, including all mechanical, chemical, and environmental factors relevant to tribological behavior.
SLIPPERINESS
In a general sense, slipperiness can be defined as the tendency of something to suddenly or involuntarily slide. In many cases, an organic polymer surface such as a coating is involved when slipperiness in considered. In terms of the flooring market area, for which the above definition was written, the flooring surface could be an alkyd or polyurethane enamel on a porch or deck, a vinyl- or polyurethanecoated gymnasium floor, a waxed vinyl composition floor, an epoxy-coated concrete floor, structural steel and coated structural steel, and so on. Slipperiness can be considered as involving of two factors—skid resistance and slip resistance. Skid may be defined as an act of sliding without rota-
TABLE 4—Coefficients of kinetic friction, μK, for Nylon 6-6 with various degrees of surface smoothness in contact with same material or steel with and without lubricants [10]. Moving Surface
Stationary Surface
Lubricant
μK
Nylon—molded surface
Nylon—molded surface
None
0.63
Nylon—molded surface
Nylon—machined surface
None
0.52
Nylon—molded surface
Mild steel
None
0.31
Nylon—machined surface
Nylon—molded surface
None
0.45
Nylon—machined surface
Nylon—machined surface
None
0.46
Nylon—machined surface
Mild steel
None
0.33
Mild steel
Nylon—molded surface
None
0.41
Mild steel
Nylon—machined surface
None
0.41
Mild steel
Mild steel
None
0.6–1.0
Nylon—machined surface
Nylon—machined surface
Graphite
0.28
Nylon—machined surface
Nylon—machined surface
Liquid paraffin
0.13
Nylon—machined surface
Nylon—machined surface
Water
0.24
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
678
PAINT AND COATING TESTING MANUAL
Q
TABLE 5—Coefficients of friction for various coatings used in the electronics industry and elsewhere [17]. Coating Material
Coefficient of Sliding Frictiona
Epoxy coating with Teflon™ fiber
0.15
Nylon
0.3
Polyethylene
0.6–0.8
Polyimide
0.17
Poly(methyl methacrylate)
0.4–0.5
Polystyrene
0.4–0.5
Polytetrafluoroethylene (Teflon™ TFE)b
0.05–0.1; 0.04; 0.016
Poly(vinyl chloride)
0.4–0.5
Polyxylylene Parylene N
0.25
Parylene C
0.29
Parylene D
0.31–0.33
Lubricants for comparison Graphite
0.18
Molybdenum disulfide
0.12
Not designated as static or kinetic coefficients. Values from different sources.
a
b
tion, and slip can be defined as a slide that occurs suddenly or involuntarily. The coefficient of friction is a measure of slip with a high coefficient of friction denoting poor slip and a low coefficient of friction denoting good slip [5]. Slip resistance is something that is not really measurable in common terms and is often defined by different people in
TABLE 6—Tribolelectric series for a number of widely used polymers [6]. Most Negative End of Series
Polytetrafluoroethylene Polyethylene Poly(vinylidine chloride) Poly(vinyl chloride) Polyacrylonitrile Poly(ethylene terephthalate)
Decreasing Negativity
Poly(vinyl alcohol) Poly(methyl methacrylate) Cellulose acetate Silk Cotton Cellulose Polyamide (nylons)
Most Positive End of Series
Wool
15TH EDITION
different ways— but everyone realizes that it does exist and the importance of testing for it remains regardless of the difficulty in defining the term [18]. The importance of slip and slipperiness is evidenced by a recent scientific symposium that was entitled “The Measurement of Slipperiness—an International Scientific Symposium” [19]. The seven papers in this symposium dealt with concepts and definitions [20], occupational slip injuries [21], biomechanics [22], human-centered approaches to slipperiness measurement [23], the role of surface roughness [24], and the role of friction in slipperiness [25,26]. For example, Chang and coworkers [26] studied friction mechanisms at the shoe-floor interface under a variety of conditions that included dry surfaces, liquid- and solid-contaminated surfaces, and icy surfaces. They concluded that static friction measurements determined by the traditional use of a drag-type device are only suitable for dry, clean surfaces. Dynamic and stick-slip measurements are needed to properly evaluate contaminated and icy surfaces and to estimate potential risk on such surfaces. Miller [27,28] has presented interesting discussions related to the measurement of slip resistance and to slip resistance standards. These discussions are readily available on the Internet. Di Pilla [29] has also discussed the measurement of slip resistance and the state of the technology, as it existed in late 2001.
DETERMINATION OF THE COEFFICIENT OF FRICTION (COF)
Three types of instruments are used to measure the COF, and these are illustrated in Fig. 3 [30]. These devices are drag-type meters that are based on μ = FF/FN, pendulumtype meters that measure the energy loss of the pendulum as an indirect indication of dynamic friction, and articulatedstrut devices that are based on the direct and fundamental principle of the resolution of forces that take place when as object slides down an incline as described above wherein μK = tan θ.3 Other instruments include torque (ASTM E510 and G99), ball-rolling (ASTM G123), roller (ASTM D3412 and G143), and other devices (see Table 8 for Standard titles). Short descriptions of various devices for studying friction and related properties that are available nationally and internationally can be found on the Internet [18]. Recently [31], sliding experiments were performed on amorphous polystyrene of various molecular weights using an atomic force microscope equipped with a nanotip probe. Friction coefficients were found to range between 0.2 and 0.45 and were similar to those found with other devices. Velocity effects on the dynamic coefficients were slight to nil, but there were marked effects involving the force at which the tip is unstuck as a function of polystyrene molecular weight. As molecular weight decreases, the applied force to overcome friction increases. Overall the results suggest that untangling of polymer molecules plays a role in describing breakage of the contact between interacting Although the COF from inclined plane testing is often referred to as a dynamic COF, μK, Section 10.1 of ASTM G115, Standard Guide for Measuring and Reporting Friction Coefficients, indicates that such testing only yields the static coefficient of friction. The standard recommends that the term static friction coefficient, μS, be used to describe a coefficient calculated using a breakaway force in a friction test rig that moves a specimen with a mechanism other than gravity. 3
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 53
Q
FRICTION AND SLIP RESISTANCE
679
TABLE 7—Definitions/Terminology for friction and slip properties taken from two ASTM publications. ASTM G115-10 Standard Guide for Measuring and Reporting Friction Coefficients
ASTM D2047 Test Method for Static Coefficient of Friction of Polish-Coated Floor Surfaces as Measured by the James Machine
Coefficient of friction—The dimensionless ratio of the friction force between two bodies to the normal force pressing these bodies together.
Coefficient of friction—The ratio of the horizontal (shear) component of force required to overcome friction, to the vertical (normal) component of the force applied.
Kinetic coefficient of friction—The coefficient of friction under conditions of macroscopic relative motion between two bodies.
Dynamic coefficient of friction—The ratio of the horizontal component of force required to cause a body to continue to slide at a constant velocity, to the vertical component of the force applied.
Friction force—The resisting force tangential to the interface between two bodies when, under the action of external force, one body moves or tends to move relative to the other.
Friction—The resistance to relative motion developed between two solid contacting bodies at, and parallel to, the sliding plane.
Static coefficient of friction—The coefficient of friction corresponding to the maximum friction force that must be overcome to initiate macroscopic motion between two bodies.
Static coefficient of friction—The ratio of the horizontal component of force required to a body that just overcomes the friction or resistance to sliding, to the vertical component of the force applied.
. . .
Slip resistance—The frictional force opposing movement of an object across its surface, usually with reference to the sole or heel of a shoe on a floor. A surface having a static coefficient of friction of 0.5 or greater as measured by this test method is considered to have adequate slip resistance. That is, it will provide the required traction for preventing or markedly reducing the probability of slipping while walking.
Stick-slip—A relaxation oscillation usually associated with decrease in coefficient of friction as the relative velocity increases
. . .
bodies. Scratch testing with a spherical indenter has been used to measure the friction coefficient of polydimethylsiloxane on stainless steel [15]. These results indicated that the dynamic friction coefficient decreased with increasing normal load suggesting that the viscoelastic character of the coating influenced its frictional behavior. Another technique involving scanning-probe microscopy and known as friction-force microscopy measures the lateral force on a sliding probe caused by sliding friction [32]. The technique allows frictional processes to be examined on a nanometer scale. It has been found that the velocity dependence of friction has a maximum in the friction-velocity interaction and this is related to the glass transition temperature, Tg. Hammerschmidt and coworkers [33] have summarized some of these studies when they used temperature-controlled friction-force microscopy to investigate viscoelastic relaxations of polystyrene, poly(methyl methacrylate), and poly(ethylene terephthalate) and their effect on friction. They concluded that for polymer films the main contribution to friction was viscoelastic mechanical loss.
SLIP RESISTANCE
One may think of slip as the antithesis of friction. While both high and low friction are important, in the sense of walkways, working surfaces, sidewalks, swimming pool, bathing, and other tiled areas, and similar surfaces, there must be sufficient friction to ensure that slipping and/or tripping with subsequent falling and potential injury will not occur. There are many different slip resistance testing specifications and standards. OSHA (Occupational Safety and Health Administration) has a standard [34] that deals
with the slip resistance of skeletal structural steel. A subpart of this standard, Standard Number 1926.754(c)(3), indicates that after July 18, 2006, workers will not be permitted to walk on the top surface of structural steel that has been coated with paint or similar material unless documentation or certification indicates that the coating has achieved a minimum average slip resistance of 0.5. This value must be achieved when measured with an English XL tribometer [35] or equivalent tester on a wetted surface at a testing laboratory. The certification or documentation is to be based on the appropriate ASTM standard test method in a laboratory that is capable of performing the test. The cited ASTM test methods [36] are: ASTM F1677-96 Standard Test Method for using a Portable Inclinable Articulated Strut Slip Tester (PIAST). This method is approved for dry and wet testing. The method, which simultaneously applied the horizontal and vertical forces, can be used on nearly all surfaces. The device used with this method is the Brungraber Mark II. ASTM F1679-00 Standard Test Method for using a Variable Incidence Tribometer (VIT). The literature suggests that this is an excellent testing method that is approved for both dry and wet testing on essentially all surfaces. As in the previous method, the horizontal and vertical forces are applied in a simultaneous manner. The device used with this method is the English XL tribometer [35], and operators using this equipment should be certified for such use. With the above said, it must be pointed out that ASTM has found that these methods lack a precision and bias statement and have a reference to proprietary apparatus
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
680
PAINT AND COATING TESTING MANUAL
Q
TABLE 8—ASTM Standards for measuring friction coefficients and related technology (other standards exist). ASTM
Title
C80875(2010e)
Guideline for Reporting Friction and Wear Test Results of Manufactured Carbon and Graphite Bearing and Seal Materials
C1028-07e1a
Test Method for Determining the Static Coefficient of Friction of Ceramic Tile and Other Like Surfaces by the Horizontal Dynamometer Pull-Meter Method
D1894-08
Test Method for Static and Kinetic Coefficients of Friction of Plastic Film and Sheeting
D2047-99
Test Method for Static Coefficient of Friction of Polish-Coated Flooring Surfaces as Measured by the James Machine
D2394-83
Methods for Simulated Service Testing of Wood and Wood-Base Finish Flooring
D2714—94
Test Method for Calibration and Operation of the Falex Block-on-Ring Friction and Wear Testing Machine
D3108-07
Test Method for Coefficient of Friction, Yarn to Solid Material
D3412-07
Test Method for Coefficient of Friction, Yarnto-Yarn
D410390(2009)
Practice for Preparation of Substrate Surfaces for Coefficient of Friction Testing
E303-93(2008)
Test Method for Measuring Surface Frictional Properties Using the British Pendulum Tester
E670-09
Test Method for Side Force Friction on Paved Surfaces Using the Mu-Meter
E707-90(1996)
Test Method for Skid Resistance Measurements Using the North Carolina State University Variable-Speed Friction Tester
F609-05
Test Method for Using a Horizontal Pull Slipmeter (HPS)
F695-01(2009)
Practice for Ranking of Test Data Obtained for Measurement of Slip Resistance of Footwear, Sole, Heel, or Related Materials
F732-00(2006)
Test Method for Wear Testing of Polymeric Materials Used in Total Joint Prosthesis
G40-99
Terminology Relating to Wear and Erosion
G65-04(2010)
Test Method for Measuring Abrasion Using the Dry Sand/Rubber Wheel Apparatus
G77-98
Standard Test Method for Ranking Resistance of Materials to Sliding Wear Using Block-onRing Test
G99-05(2010)
Test Method for Wear Testing with a Pin-onDisk Apparatus
G115-10
Standard Guide for Measuring and Reporting Friction Coefficients
G13305(2010)
Test Method for Linearly Reciprocating Ballon-Flat Sliding Wear
G14303(2009)
Test Method for Measurement of Web/Roller Friction Characteristics
Withdrawn, but these standards are referenced and therefore included in above listing.
a
15TH EDITION
that violates certain ASTM requirements [37]. A deadline of September 30, 2006 had been set for correcting the deficiencies, but it was felt that this deadline would not be met. As a result, a decision was made to withdraw ASTM F1677 and ASTM F1679 as well as two related test methods.4 At the same time the above was taking place, OSHA [38] reopened and sought comments and information regarding its Standard Number 1926.754(c)(3) and attendant Appendix B [34,36]. OSHA’s purpose of reopening this matter was to consider whether to retain, amend or revoke the provision regarding test methods for slip resistance on skeletal structural steel and whether slip-resistant coatings meeting the requirements could be available by July 18, 2006. Since OSHA ascertained that no progress would be made toward meeting these goals5 by the “to-be-effective” date, the group decided in early 2006 to revoke a provision within the Steel Erection Standard that addresses slip resistance of skeletal structural steel [39]. Readers interested in this matter would do well to follow proceedings that probably will be found in the Federal Register and/or ASTM literature. English [40] has a listing of slip resistance standards that are in wide use. Included in the compilation are the above test methods as well as other standards from ASTM, American National Standards Institute (ANSI), National Fire Protection Association (NFPA), and Underwriters Laboratories. The standards are separated by their suitability for wet or dry testing. English also has a description of other tribometers based on pendulum and drag-sled principles.
SENSOR MATERIALS
Materials that may be used as sensors or test pads include a broad range of known items [41]. Of course, sensors are used in context with the tests being conducted. For friction testing, two surfaces are required for any measurement— one is the surface of the specified material to be tested and the other is a sensor material or the surface against which the specified material is to be tested. The specified material and sensor material may be the same or different, but both must be defined and specified. To obtain meaningful and reproducible results, it is essential that the sensor material be selected to represent use conditions and be well defined. Properties such as uniformity (surface character including flatness, roughness, chemical composition, resilience, and shear modulus), permanence in that chemical and physical characteristics do not change with time, and availability in a usable form that does not require excessive preparation should be considered when selecting a sensor. Many feel that a major concern in sensor selection is choosing a material that will not change as a function of time. Both sensor material and
The “other two” test methods are ASTM F1678-96 Test Method for Using a Portable Articulated Strut Slip Tester and ASTM D5859 Test Method for Determining the Traction of Footwear on Painted Surfaces Using the Variable Incidence Tester. 5 Comments received by OSHA prior to the deadline suggested that no significant progress had been made regarding the suitability of the test methods referenced in the provision for testing slip resistance or the availability of coatings that would need the slip resistance requirements of the provision [39]. The final revocation rule was effective January 18, 2006. 4
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 53
Q
FRICTION AND SLIP RESISTANCE
681
Fig. 3—Schematics of different friction measurement devices (courtesy of the National Institute of Standards and Technology).
test material may be the same or different and both should be reported when the coefficient of friction is stated. Sensors might be exemplified by leather that is used as a primary source material when friction of flooring is studied. For interlaboratory and specification testing, shoe materials such as leather conforming to Federal Specification KK-L-165C6 [42] are used. When leather is used as a sensor or as a test material, it should be kept in mind that leather soaks up water and becomes somewhat sticky which will result in high coefficients of friction being measured. Also, leather is not homogeneous in character and can vary from piece to piece. It is important that the appropriate materials are used when specific testing is done. In the case of rubber, it is recommended that a rubber conforming to Section 7.1 of ASTM D1630-06 Test Method for Rubber Property—Abrasion Resistance (Footwear Abrader) Available from Standardization Documents Order Desk, Attn: NPODS, Bldg. 4 Section D, 700 Robbins Ave., Philadelphia, PA 19111-5094.
6
be used. Neoprene, nitrile, and styrene-butadiene rubbers have been used in many investigations. In another study [43], three leathers with three levels of oil content, two Kraton® thermoplastic elastomers, and 15 different rubbers were used to generate statistical data for ASTM F609-05 Test Method for Using a Horizontal Pull Slipmeter (HPS). Neolite® is also used as a test pad material that does not change regardless of moisture or wear conditions.
LUBRICANTS
For those instances where friction is to be decreased (slip increased), a variety of lubricants, slip agents, or slip additives are available. These compounds include micronized polyethylene, polytetrafluoroethylene as well as various waxes, oils, and silicones. Water can also act as a lubricant when it is present on many solids. Persson [44] and coworkers have studied the effect of water on the loss of braking power and rubber skidding on a wet highway (tire asphalt). They reported that rubber can seal water-filled pools or
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
682
PAINT AND COATING TESTING MANUAL
Q
TABLE 9—Examples of lubricants Esters Fatty alcohol esters Fatty esters Glycerol esters Wax esters Fatty acid amides Alkanolamides Monoamides Bisamides Metallic compounds Aluminum stearate Barium stearate Calcium stearate Graphite Molybdenum sulfide and disulfide Oils and greases Conventional and synthetic Polymeric and oligomeric materials Fluoropolymers Micronized polyethylene Micronized polytetrafluoroethylene Montan waxes Oxidized polyethylene waxes Paraffins, low melting Polyolefin waxes Silicones
pockets and effectively smooth the rubber surface. This effect reduces the viscoelastic dissipation that surface asperities induce in bulk rubber, and decreases friction. High molecular weight polyoxyethylene has remarkable properties for reducing the friction or drag characteristics of water [45] on objects such as pleasure/racing boats, naval craft, and torpedoes when used in various forms, for reducing the friction of ski waxes on snow when compounded into the wax, and increasing volume output of water under constant pressure through hoses and ducts. In some instances, a material of construction can have an internal lubricant as is the case with the acid- and chemical-resistant wood from the tree lignum vitae7 (Guai-acum Officinale), an extremely hard wood that find use in noiseless underwater bearings and as saw guides. McLaren and Tabor [46] found that in the dry state this self-lubricating wood has friction characteristics comparable to those of polytetrafluoroethylene. Experimental evidence indicated that the lubricity is due to wood waxes that exude from the wood during sliding operations. Formulation with solid lubricants has made great strides in recent years for indusA small tree that is native to the Netherlands Antilles and St. Johns in the Virgin Islands and that has been cultivated in Argentina. The wood from this tree is very dense and will sink in water. In addition to its self-lubricating frictional properties, the tree s sap and bark are said to have medicinal properties.
7
15TH EDITION
trial operations that impose severe sliding conditions [47]. Nanostructured or nano-composite forms are coupled with micro-texturing, micro-patterning, or other smart surface engineering strategies [48] to achieve very high levels or durability and performance. Many lubricants also function as abrasion resistance and mar-reduction agents. In certain instances, they can be used as antiblocking agents. They are often used in the plastics processing industry where they function as melt viscosity reducers and flow agents to improve flow onto metal surfaces and, at times, as co-stabilizers. The use of lubricants to reduce wear and power consumption is well known. Factors that should be considered in selecting a lubricant include absorptivity, melting point, polarity, solubility, effect on adhesion, potential reactivity from friction thermal energy developed, volatility, and similar characteristics. Common lubricant families as well as selected specific lubricants are listed in Table 9. Polyethylene and polytetrafluoroethylene are available in powdered or micronized form in a variety of particle sizes [49]. Silicones are available as formulated products designated for use in the coating industry [50,51]. The powdered and micronized products are incompatible and function as filler when incorporated into the coating where they act as tiny “ball bearings” that decrease friction and often improve abrasion and mar resistance.
BIBLIOGRAPHY Bowden, F. P., and Tabor, D., Friction and Lubrication of Solids, Parts I, Clarendon Press, Oxford, UK, 1950, Friction and Lubrication of Solids, Part II, Clarendon Press, Oxford, UK, 1964. Bowers, R. C., and Zisman, W. A., “Surface Properties,” in Engineering Design for Plastics, E. Baer, Ed., Reinhold Publishing Corp., New York, Chap. 10, 1964, pp. 719–741. Briscoe, B. J., and Tabor, D., “Friction and Wear of Polymers,” Polymer Surfaces, John Wiley & Sons, New York, Chap. 1, 1978, pp. 1–23. Physics of Plastics, P. D. Ritchie, Ed., D. Van Nostrand Co., Inc., Princeton, NJ, 1965. Rabinowicz, E., Friction and Wear of Materials, John Wiley & Sons, New York, 1965. Schallamach, A., “Abrasion and Tyre Wear,” in The Chemistry and Physics of Rubber-Like Substances, L. Bateman, Ed., John Wiley & Sons, New York, Chap. 13, 1963, pp. 355–416. Sextro, W. D., Dynamical Contact Problems with Friction: Methods, Experiments and Applications, 2nd ed., Springer, New York, 2007, 190 p. Stachowiak, G. W., “Wear: Materials, Mechanisms and Practice,” Tribology in Practice Series, John Wiley & Sons, Inc., 2006, 478 p. Tribol. Lett., A journal devoted to development of the science of Tribology and its applications, Kluwer Academic Publishers, Boston. “Tribology: Lubrication, Friction, and Wear,” Tribology in Action Series, I. V. Kragelsky, and V. V. Alisin, Eds, John Wiley & Sons, Inc., 2005, 948 p.
References [1] Koleske, J. V., Springate, R., and Brezinski, D., “Two Thousand Seven Additives Guide,” Paint and Coating Industry, Vol. 23, No. 5, 2007, pp. 42–121. [2] Additives for Plastics, Vol. 1, R. B. Seymour, Ed., Academic Press, New York, 1978. [3] Paint/Coatings Dictionary, S. LeSota, Ed., Federation of Societies for Coating Technology, 613 p., 1978. [4] Cramp, A. P., and Masters, L. W., “Preliminary Study of the Slipperiness of Flooring,” NBSIR 74-63, National Bureau of Standards (July 1974).
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 53
[5] Burwell, J. T., and Rabinowicz, E., “The Nature of the Coefficient of Friction,” J. Appl. Phys., Vol. 24, 1953, pp. 136–139. [6] Hall, C., Polymer Materials, John Wiley & Sons, New York, 1981, p. 198. [7] Shooter, K. V., “Sliding Friction of Some Metallic Glasses,” Plastics, Vol. 26, No. 281, 1961, p. 117. [8] Handbook of Chemistry and Physics, 76th ed., D. R. Lide, Ed., Sec. 15, pp. 28–30, 1995. [9] Brydson, J. A., Plastic Materials, D. Van Nostrand Co., Inc., Princeton, NJ, 1966, p. 576. [10] Blanchet, T. A., Peng, Y.-L., and Nablo, S. V., “Tribology of Selectively Irradiated PTFE Surfaces,” Tribol. Lett., Vol. 4, No. 1, 1998, pp. 87–94. [11] McLaren, K. G., and Tabor, D., “Visco-Elastic Properties and the Friction of Solids: Friction of Polymers: Influence of Speed and Temperature,” Nature (London), Vol. 197, 1963, p. 856. [12] Grosch, K. A., “The Relation Between the Friction and ViscoElastic Properties of Rubber,” Proc. R. Soc. London, Ser. A, Vol. 274, No. 1356 1963, pp. 21–39. [13] Williams, M. L., Landel, R. F., and Ferry, J. D., “The Temperature Dependence of Relaxation Mechanisms in Amorphous Polymers and Other Glass-Forming Liquids,” J. Am. Chem. Soc., Vol. 77, 1955, p. 3701; Ferry, J. D., Viscoelastic Properties of Polymers, 3rd ed., John Wiley & Sons, New York, 1980. [14] Golden, J. M., “A Molecular Theory of Adhesive Rubber Friction,” J. Phys. A, Vol. 8, 1975, pp. 966–979. [15] Zhang, S. L., Tsou, A. H., and Li, J. C. M., “Scratching Behavior of Elastomeric Poly(dimethyl siloxane) Coatings,” J. Polym. Sci., Part B: Polym. Phys., Vol. 40, No. 14, 2002, pp. 1530–1537. [16] Schallmach, A., “Abrasion and Tyre Wear,” in The Chemistry and Physics of Rubber-Like Substances, L. Bateman, Ed., John Wiley&Sons, New York, 1963, Chap. 13, p. 784. [17] Licari, J. J., Plastic Coatings for Electronics, Robert E. Drieger Publishing Co., Inc., Malabar, FL, 1980, pp. 381. [18] http://www.safety-engineer.com. [19] Courtney, T. K., Chang, W.-R., Grönqvist, R., and Redfern, M. S., “The Measurement of Slipperiness—an International Scientific Symposium,” Ergonomics, Vol. 44, No. 13, 2001, pp. 1097–1101. [20] Grönqvist, R., Chang, W.-R., Courtney, T. K., Leamon, T. B., Redfern, M. S., and Strandberg, L., “Measuremnt of Slipperiness: Fundamental Concepts and Definitions,” Ergonomics, Vol.44, No. 13, 2001, pp. 1102–1117. [21] Courtney, T. K., Sorock, G. S., Manning, D. P., Collins, J. W., and Holbein-Jenny, M. A., “Occupational Slip, Trip, and FallRelated Injuries—Can the Contribution of Slipperiness be Isolated?,” Ergonomics, Vol. 44, No. 13, 2001, pp. 1118–1137. [22] Redfern, M. S., Cham, R., Gielo-Perczak, K., Grönqvist, R., Hirvonen, M., Lanshammar, H., Marpet, M., Pai, C. Y.-C., and Powers, C., “Biomechanics of Slips,” Ergonomics, Vol. 44, No. 13, 2001, pp. 1138–1166. [23] Grönqvist, R., Abeysekera, J., Gard, G., Hsiang, S. M., Leamon, T. B., Newman, D. J., Gielo-Perczak, K. Lockhart, Thurman, E., and Pai, C. Y.-C., “Human-Centered Approaches in Slipperiness Measurement,” Ergonomics, Vol. 44, No. 13, 2001, pp. 1167–1199. [24] Chang, W.-R., Kim, I.-J., Manning, D. P., and Bunterngchit, Y., “The Role of Surface Roughness in the Measurement of Slip-periness,” Ergonomics, Vol. 44, No. 13, 2001, pp. 1200–1216. [25] Chang, W.-R., Grönqvist, R., Leclercq, S., Myung, R., Makkonen, L., Strandberg, L., Brungraber, R. J., Mattke, U., and Thorpe, S. C., “The Role of Friction in the Measurement of Slipperiness, Part 1: Friction Mechanisms and Definition of Test Conditions,” Ergonomics, Vol. 44, No. 13, 2001, pp. 1217–1232. [26] Chang, W.-R., Grönqvist, R., Leclercq, S., Brungraber, R. J., Mattke, U., Strandberg, L., Thorpe, S. C., Myung, R., Makkonen, L., and Courtney, T. K., “The Role of Friction in the Measurement of Slipperiness, Part 2: Survey of Friction
[27] [28] [29] [30]
[31] [32]
[33]
[34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51]
Q
FRICTION AND SLIP RESISTANCE
683
Measurement Devices,” Ergonomics, Vol. 44, No. 13, 2001, pp. 1233–1261. Miller, B., C., Lecture, “Measurement of Slip Resistance, a Legal and Practical Perspective,” http://www.safety-engineer. com/surfaces.htm, p. 20. Miller, B., C., “Slip Resistance Standards: Sorting it All Out,” Safety and Health (March 1999); http://safety-engineer.com/ standards.htm. Di Pilla, S., “Slip Resistance Measurement: The Current State of the Art,” By Design, Engineering Practice Specialty Newsletter, Vol. 1, No. 1,1, Fall 2001; www.asse.org. Adler, S. C., and Pierman, B. C., “History of Walkway SlipResistance Research at the National Bureau of Standards,” NBS Special Publication 565, National Bureau of Standards, Washington, DC, December 1979. Michel, D., Kopp-Marsaudon, S., and Aimé, J. P., “Tribology of a Polystyrene Film Investigated with an AFM,” Tribol. Lett., Vol.4, No. 1, 1998, pp. 75–80. Overney, R. M., Meyer, E., Frommer, J., Brodbedck, D., Luthi, R., Howald, L., Güntherodt, H. J., Fujihira, M., Takana, H., and Gotoh, Y., “Friction Measurements on Phase-Separated Thin Films with a Modified Atomic Force Microscope,” Nature (London), Vol. 359, 1992, p. 133. Hammerschmidt, J. A., Gladfelter, W. L., and Haugstad, G., “Probing Polymer Viscoelastic Relaxations with TemperatureControlled Friction Force Microscopy,” Macromolecules, Vol. 32, 1999, pp. 3360–3367. U.S. Department of Labor, OSHA Standards 29 CFR, “Steel Erection; Slip Resistance of Skeletal Structural Steel,” Federal Register No. 69:42379-42381, Proposed Rules (July 15, 2004). English, W., “Using the English XL Sliptester,” March 8, 2001, www.englishxl.com. OSHA, “Acceptable test methods for testing slip-resistance of walking/working surfaces: Non-Mandatory Guidelines for Complying with 1926.754(c)(3).—1926 Subpart R Appendix B”. Letter, Childs, W. H., Chairman ASTM Committee on Standards, to Mr. Stephen DiPilla, ASTM F 13 Chairman, April 5, 2004. www.safety-engineer.com/dipilla. Federal Register, “Steel Erection; Slip Resistance of Skeletal Structural Steel,” Vol. 69, No. 135, 42379–42381, July 15, 2004. Federal Register, “Steel Erection; Slip Resistance of Skeletal Structural Steel,” Vol. 71, No. 11, 2879–2885, January 18, 2006. English, W., “Current Slip Resistance Testing Standards,” Revised March 07, 2003, www.englishxl.com/stds. Moseley, P. T., and Crocker, A. J., Sensor Materials, Institute of Physics Publishing, London, 1996, p. 227. Federal Specification KK-L-165C Leather, Cattlehide, Vegetable Tanned and Chrome Retanned, Impregnated and Soles. Type I—Factory (for Shoe Making), Class 6—Strips. ASTM Research Report F13–1001,27 July 1979. Persson, B. N. J., Tartaglino, U., Albohr, O., and Tosatti, E., “Sealing is at the Origin of Rubber Slipping on Wet Roads,” Nature Mater., Vol. 3, 2009, pp. 882–885. Bailey, F. E. Jr., and Koleske, J. V., Alkylene Oxides and Their Polymers, Marcel Dekker, Inc., New York, 1990, p. 261. McLaren, K. G. and Tabor, D., “The Frictional Properties of Lignum Vitae,” Br. J. Appl. Phys., Vol. 12, 1961, pp. 118–120. Donnet, C., and Erdemir, A., “Solid Lubricant Coatings: Recent Developments and Future Trends,” Tribol. Lett., Vol. 17, No. 3, 2004, pp. 389–397. Podsiadlo, P., and Stachowiak, G. W., “Classification of Tribological Surfaces Without Surface Parameters,” Tribol. Lett., Vol.16, No. 1–2, 2004, pp. 163–171. “Innovation in Powder Technology,” Technical Data Brochure, Shamrock Chemicals Corporation, Newark, NJ. “Byk-Mallinckrodt Paint-Additives,” Technical Data Notebook, Byk-Mallinckrodt USA, Inc., Wallingford, CT. “Dow Corning Additives,” Technical Data Brochure 24-391 E-93, Dow Corning Corp., Midland, MI.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
Part 12: Environmental Resistance
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
54
MNL17-EB/Jan. 2012
Prevention of Metal Corrosion with Protective Overlayers William H. Smyrl1
THE PREVENTION OF CORROSION BY SURFACE processing enjoys significant economic leverage, and, as evidence, one may cite the widespread use of coatings, films, and inhibitors for metals and semiconductors in many service environments. All engineering metals used in modern technological societies are unstable with respect to corrosion, and the result is a loss of properties. Natural oxide films provide protection against continued attack for some metals, and alloying extends the life of other metals by developing highly stable passive films. Where metals may not be protected by oxide films, other modification methods have been developed to reduce corrosive attack. In reality, the improvement of corrosion resistance of metals by modification of the surface has been practiced since the invention of metal tools. Some of the earliest techniques to prevent corrosion involved coating with greases or natural oils. More modern methods were developed in the 19th and 20th centuries and include multiple coatings, zinc galvanizing, electroplating of other pure metals, and vacuum physical vapor deposition of mostly pure metal coatings by electron beam and sputtering techniques. The metal coatings are better barriers than organic films because of the lower permeability of the former to moisture, oxygen, and ions. Inhibitors or conversion coatings and primers for paints are cheaper than metal coatings and are used widely by paint manufacturers even though they remain highly proprietary in nature. The use of organic coatings to protect metal surfaces is practiced widely. Much of the use is for atmospheric exposure of motor vehicles as well as for structural units such as bridges and buildings. The successful implementation of existing technologies has greatly reduced the effects of corrosion of automobiles, for example, in the past decade in response to consumer demand. Despite many recent advances, coating technologists and scientists acknowledge that much is unknown and that new processes and understanding are the keys for further progress [1]. Defects in the metal substrate and in the overlayers are among the primary concerns because they are the source of localized corrosion phenomena. Defects may occur on length scales from atomic-level lattice vacancies to arrays of defects at grain boundaries (for crystalline materials) or to random pores or cracks (for example, in noncrystalline films). Avoiding such defects by proper quality control is a major concern in coatings science and technology.
In the discussion that follows, aspects of corrosion that involve thermodynamics and kinetics will be developed as a basis for the description of the specific nature of corrosion of metals under protective films and overlayers. Some emphasis will be given to protection of thin metal films and micro-structures that are particularly sensitive to corrosion and whose successful protection provides a basis for advancing protection technology in general.
CORROSION IN AQUEOUS SOLUTIONS
The driving force for a reaction is the change in Gibbs free energy, AG, for reactants to products. Mathematically, this may be expressed by ΔG =
G−
products
G
reactants
(1)
The summation signs are used as a general notation to indicate that all reactants and products are included in the calculation. From the nature of the free energy function, this calculation applies to initial (reactants) and final (product) states and is independent of intervening states. The reaction may be investigated under controlled reversible conditions such as in an electrochemical cell or under irreversible conditions such as in corrosion, and the same total free energy change (∆G) will be appropriate. A quite general predictive capability may be applied to specific corrosion reactions since all the available thermodynamic data may be used for corrosion calculations directly. This enables the position of final equilibrium of the corrosion system to be established. The thermodynamic calculations have the limitation that no information concerning the rate of the reaction is provided, only what the final state will be for the process. The value of ∆G for reactions of elements to form a compound, all in their standard states at a particular temperature, is the standard free energy of formation of the compound, ΔGf0,T . Here, the subscript T denotes the temperature. Description of the detailed calculations are beyond the scope of this discussion, but several excellent textbooks are available [2,3]. The most extensive tabulations of thermochemical data for chemical compounds in their standard state at 25°C are in a series of National Bureau of Standards publications [4]. These are NBS Technical Notes 270-3, 270-4, and 270-5, which supersede the older NBS Circular 500 for the elements they cover. These tabulations also update the older
Professor, Corrosion Research Center, Department of Chemical Engineering and Materials Science, University of Minnesota, Minneapolis, MN 55455.
1
687 Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed byASTM International This standard is for EDUCATIONAL USE ONLY. Copyright © 2012 by www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
688
PAINT AND COATING TESTING MANUAL
Q
data of Latimer [5]. The book by Latimer [5] remains a valuable reference because of the description of techniques to estimate thermodynamic quantities where reliable data are sparse. Lewis and Randall [3] tabulated thermodynamic data, including data for aqueous solutions of a number of electrolytic solutions that are valuable for corrosion calculations. There is overlap between this tabulation and the JANAF thermochemical tables [6], but the latter also tabulate the standard enthalpy and free energy of formation of chemical compounds at several temperatures along with 0 — (G 0 − H298 ) / T. Oxide and hydroxide solubility are strongly influenced by the pH of the aqueous phase. Pourbaix recognized this and summarized the thermodynamic stability of metalaqueous systems by the use of potential-pH diagrams. The thermodynamic data tabulations already quoted [3–6] should be utilized for detailed calculations, however. Usually when studying corrosion, one is not concerned with the conditions for thermodynamic stability, but rather with the rate of attack and how it may be altered in basically unstable conditions because only a limited number of systems have absolute or thermodynamic stability. As a practical matter it is necessary to accept some rate of corrosion and/or to control or mitigate the rate of attack. Thus kinetic stability is always relative and subject to interruption if control is not maintained. Controlling the rate of degradation may be accomplished, for example, by the use of cathodic or anodic protection, the use of inhibitors, the maintenance of protective surface films, or buffering the composition of an otherwise aggressive solution. All these techniques are used widely to extend the life of metallic structures with continuing improvement. Corrosion reactions are electrochemical in nature. The kinetics of the corrosion reactions are then related to the kinetics of the electrochemical reactions that occur during the corrosion process. The reactions involve not one but at least two electron transfer reactions, and the reactions are not in series but are in parallel. Coupling of parallel or simultaneous reactions is a fundamental feature of the corrosion process. Each of the simultaneous reactions may consist of multiple steps, respectively, as described above, but the simultaneous, independent reactions are coupled electrically. The independent reactions occur on the same surface at the same time, but also at the same potential. The reactions may be coupled chemically as well, e.g., through pH effects, but this is not essential. The specific relation that defines the coupling of simultaneous corrosion reactions on an isolated metal surface is I a = − Ic
anodic
cathodic
(2)
There will then be zero net current to the corroding metal electrode. The relationship is written in terms of currents (Ia and Ic ) rather than current densities for reasons which will be discussed. The potential at which the balance is satisfied is the mixed or corrosion potential. It is determined by the rates of the simultaneous reactions and is not defined by the state of the system in a thermodynamic sense. The corrosion potential always lies between the equilibrium potentials of the anodic and cathodic processes, respectively. As a summary, the general corrosion of metallic materials in aqueous solutions is well understood. The anodic
15TH EDITION
or oxidation reaction of the metal is superimposed on a cathodic reaction, and the two are balanced locally on a homogeneous surface. The rate of the reaction is a function of both the rate of metal dissolution and the rate of the cathodic (reduction) reaction. Each reaction may be influenced in general by the composition of the solution, especially the pH and the electrolyte anion, and by the nature of the (oxide) films, if any, which may be formed at the metal/ electrolyte interface. If several oxidizing species are present in the solution, each may act in parallel so that the total rate of metal dissolution is increased. For example, most metals will react directly to displace hydrogen from water and to produce an oxide of the metal or some other corrosion process. The addition of oxygen will increase the rate of corrosion of the metal, usually in direct proportion to the concentration of the oxygen added. The specific details will vary with each metal to reflect the thermodynamic, kinetic, and mass transfer driving forces that are acting [7]. Heterogeneous surfaces are commonly observed in corrosion situations and are of the three general classes: (1) the inclusion of foreign metal impurities on the metal surface, (2) the nonuniform coverage of the surface by a film, either an oxide film or an artificial coating in aqueous solutions, and (3) nonuniform conditions in the electrolyte environment. All these are of great importance because localized, or nonuniform, corrosion of metals may be caused by any of the three. A form of galvanic corrosion and pitting corrosion is caused by the first type of heterogeneity, while crevice and pitting corrosion are produced by both (2) and (3). Restrictions of geometry, e.g., in crevices and corners, prevent mixing of solutions everywhere, and local buildup of reaction products or the exhaustion of an oxidant may occur. The local kinetics will be relatively independent of that in other regions except that there may be coupling through the electric field and electrical current may flow between a localized corrosion site and the surrounding surface. This may lead to nonuniform corrosion, particularly where the buildup of products increases the aggressiveness of the local solution. In this case, corrosion will be most severe, not where the concentration or flux of the bulk solution oxidant is highest but where it is lowest. Crevice corrosion is considered to be an example of this type of attack, and the aggressive solution within a crevice or pit is one which is more acidic than the external solution. Anodic dissolution, plus hydrolysis of the product metal ion, causes an increase of hydrogen ion concentration. On the other hand, reduction of either hydrogen ions or dissolved oxygen reduces the hydrogen ion concentration. If the net corrosion reaction plus hydrolysis would lead to an increase of hydrogen ion concentration, the process may occur independently of any other process and would accelerate with time to some steady state where diffusion out of the occluded region would limit the buildup. If the corrosion reaction plus hydrolysis leads to no net change in H+ concentration, an acid solution in a crevice or pit could only be created by separation of the anodic and cathodic regions. Concentrating the cathodic reaction on the outer surface would occur naturally if dissolved oxygen, for example, were the primary bulk oxidant. Coupling this with a net anodic reaction (plus hydrolysis) in the inner region for an overall current balance would lead to a steady state crevice or pit. For separation to occur as described above, a quite general condition imposed on the corrosion kinetics must be
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 54
Q
PREVENTION OF METAL CORROSION WITH PROTECTIVE OVERLAYERS
obeyed. The outside surface must support a cathodic reaction, and it must be supported at a potential that is positive of the potential of the anodic reaction in the crevice. The direction of current flow through the solution establishes this criterion. A qualitative laboratory test may be used to identify metal solution combinations that could cause localized attack by the mechanism described above. The test involves the corrosion kinetics on the metal of interest. Cathodic currents must be observed on the metal in the exterior solution at potentials that are positive of the anodic region for the crevice conditions or the separated reactions will not support increased anodic dissolution in the isolated region. This is a very definitive test, and very few metalenvironment combinations match the criterion. Ohmic drop restricts the penetration of current into a small-gap, occluded region [8]. This causes the anodic reaction to be distributed over a relatively small region, which concentrates the attack. At greater depths in the gap, the metal is isolated from the external surface reactions. Newman [9] calculated the limited depth to which a reaction may penetrate inside a circular geometry, in this case a cathodic protection reaction. The reaction is concentrated near the opening. Composition gradients are considered to be important for pitting and differential aeration corrosion as well. For pitting corrosion, similar conditions to those for crevice corrosion are considered important. Pits may be initiated in ways that are different from crevice corrosion, e.g., at foreign metal inclusions. However, the propagation of pits depends largely on a locally aggressive solution. Stirring to eliminate concentration effects will stop the growth of pits. Differential aeration could also drive corrosion at locally variable rates under an electrolyte film of nonuniform thickness. The diffusion-limited flux of oxygen through the film would be directly proportional to the film thickness. If the local corrosion rate is limited by the oxygen flux, the attack will be most severe at low film thicknesses. For active/passive metals, increase of the oxygen flux may exceed the peak current for active dissolution and cause the metal to adopt the passive state. In this case, then, thin films of electrolyte will reduce the corrosion rate.
ATMOSPHERIC CORROSION OF METALS
Most atmospheric corrosion tests have been conducted in environments such as indoor atmosphere, outdoor atmosphere, and laboratory tests under simulated conditions. Indoor corrosion studies have been performed for the electronics, computer, and communication industries for the development of more durable materials with desirable structural, magnetic, and electrical properties. On the other hand, outdoor studies aimed at understanding corrosion behavior are highly dependent on atmospheric weather factors, especially in marine and urban areas. The latter studies have been performed in the automobile, marine, and aircraft industries. Laboratory tests attempt to use accelerated methods under simulated atmosphere or aqueous conditions. Electrochemical methods have been used extensively in such tests to analyze and monitor the corrosion behavior of metals. Several weather factors are known to influence outdoor corrosion [10–13]. Precipitation, ambient and dew-point temperatures, atmospheric pollutants, wind direction and wind velocity, and solar radiation can be considered as
689
weather factors in outdoor and/or urban corrosion tests. Among these factors, moisture or relative humidity, temperature, and pollutants such as sulfur dioxide and chlorides are the most important variables. Relative humidity is known to be the most important factor in determining the atmospheric corrosion rate. It has been reported that rapid acceleration of corrosion occurs beyond a certain value of relative humidity, defined as the critical humidity [14,15]. The period in which the relative humidity exceeds the critical humidity is called the time-ofwetness, and this factor is quite significant in determining atmospheric corrosion rate of metals [16]. In addition, in the presence of a pollutant such as sulfur dioxide, the critical humidity at which corrosion is enhanced to a significant extent will decrease with increasing pollutant concentration [17,18]. It has been reported that comparatively large aggregates of water are present on oxyhydroxide surfaces at humidities below 40 % [19]. Even on clean metal surfaces obtained under ultrahigh vacuum or reducing conditions, significant quantities of water are adsorbed on air-formed films when exposed to the environments containing only oxygen and water vapor for more than a microsecond [13]. As a result, monolayers of adsorbed water may provide the medium for electrochemical microcells that may drive a heterogeneous corrosion process. Water may also exist in the form of complex mixtures with oxides, hydroxides, and mixed oxyhydroxides [19,20]. The corrosion rate of metals is accelerated by the presence of air pollutants such as sulfur dioxide, nitrite, nitrate, hydrogen sulfide, chloride, and some kinds of salts [10,15]. These species may derive from gas-borne particles or from reactions at the surface. Reaction with adsorbed water monolayers yield electrolyte films that facilitate further corrosion processes. Among these pollutants, sulfur dioxide, chlorine gases, sulfur gases, and ozone are important species that promote corrosion in the presence of water [21]. The corrosion-accelerating effect of sulfur dioxide with humidity has been reported by many investigators [10,13,15]. Vernon [15] suggested that sulfur dioxide may change the pH in electrolyte films present on metal surfaces and enhance the corrosion rate. Rice et al. [13] also suggested that sulfur dioxide is readily soluble in water to form sulfurous acid; these local acidic regions accelerate oxide formation, and the corrosion rate is also enhanced by other electrochemical effects. It has been reported that wetting of the metal surface is promoted in the presence of ammonia, and the water droplets contain higher concentrations of sulfates than for the same concentration of sulfur dioxide with no ammonia [10,22]. The effect of chlorine gas or chloride on atmospheric corrosion has been reported [10,13]. In aqueous electrochemical corrosion studies, the chloride ions usually enhance pitting corrosion of many metals and also degrade the oxide surfaces. Rice et al. [13] reported that chlorine gases reduce the surface pH and yield hygroscopic corrosion products that influence the amount of adsorbed water. A direct relationship between the amount of chlorides in corrosion products and atmospheric corrosion rates was reported by Sereda [10]. The effect of ozone gas on copper and silver corrosion has been known to be significant, while cobalt, nickel, and iron are insensitive to ozone [13]. It has been reported that
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
690
PAINT AND COATING TESTING MANUAL
Q
ozone may enhance the corrosion rate of metals sensitive to H2S. The atmospheric corrosion rate can be measured either in field tests in different atmospheres or with accelerated tests in the laboratory. The field tests require long exposure times and yield complicated data that prevent detailed analysis. Accelerated tests are performed under simulated atmospheric conditions, and they are the easiest way to acquire more information with various setup conditions. However, it may not be possible to simulate practical service conditions. There are several methods to monitor and control the corrosion rates by means of either field or laboratory tests. The conventional method is the weight loss determination, which requires long-term exposure unless a continuous method is used that involves the quartz crystal microbalance [23,24]. Another method is the electrical resistance method and measurement using electrochemical cells. Electrical resistance methods use the changes in the electric resistance of thin wires or foils to monitor failure, but they cannot be used for determination of the instantaneous corrosion behavior [25–27]. Electrochemical methods have been developed to take advantage of the electrochemical basis for atmospheric corrosion [28,29]. Corrosion currents can be monitored electrochemically, and the instantaneous value of current can be detected. One way to monitor atmospheric corrosion with an electrochemical method is to design a cell that will work under thin electrolyte layers (less than 500 m) with consideration of the effects of corrosion products and dilute pollutants [30]. Electrochemical methods for monitoring atmospheric corrosion have been well reviewed by Mansfeld [30,31]. Most of the studies have been aimed at macroscopic measurement of time-of-wetness that is associated with electrochemical corrosion [10,12,16,17]. Galvanic cells with electrodes of different metals have been commonly used [16,32]. Sereda [10] has developed galvanic cells of platinum-iron and platinum-zinc couples to determine the time-of-wetness. Time-of-wetness was arbitrarily defined as the interval during which the external potential exceeded 0.2 V. This figure was the period during which the relative humidity was greater than 85 % [12,16]. Tomashov [33] has used sandwich-type galvanic cells of iron-copper, iron-zinc, iron-aluminum, and copper-aluminum. They concluded that the method was suitable for fast determination of the corrosivity of the atmosphere and that the direct measurement of corrosion rate for testing metals was possible. Several investigators [17,34–36] have used galvanic cells consisting of steel-copper and electrolytic cells consisting of individual metals (steel, zinc, or copper) to which an external potential was applied. They concluded that the cell current gave qualitative agreement with the weightloss data. Recently, extensive studies have been performed by Mansfeld and his coworkers [30,31,37–39]. They used galvanic cells and electrolytic cells which consisted of two electrodes and three electrodes. Galvanic cells such as copper-steel, copper-zinc, steel-zinc, steel-aluminum, and aluminum-zinc couples were used to study the effects of pollutants, relative humidity, and so on. They used the electrolytic cells such as two- and three-electrode cells for studies of the corrosion kinetics and for the measurement of corrosion currents. The polarization resistance method
15TH EDITION
was used to determine atmospheric corrosion kinetics under thin electrolytic layers. Mazza et al. [40] have used a galvanic cell that consisted of a sandwich formed of bronze covered by its artificial corrosion products on which a highporosity gold film was applied. They monitored the corrosion current with a zero resistance ammeter and obtained instantaneous and continuous information on the corrosion rate of the bronze. Tosto and Bruco [41] used galvanic cells of copper-steel to obtain the relation between the corrosion content and relative humidity. They found that the corrosion current depended on relative humidity (RH). As a rapid electrochemical method for monitoring atmospheric corrosion, measurements of electrode potential using a suitable reference electrode have been developed by several investigators [42,43]. Although the method gives no absolute data on corrosion rates, it is a fast and easy method for on-site investigations. Thin film methods to measure corrosion rates were discussed by Howard [44]. Pourbaix and his coworkers [42,45] developed an accelerated electrochemical wet and dry method that was designed to use alternate immersion cycles in an electrolyte bath. The electrode potential was monitored when the steel electrode was in the wet part of the cycle. They concluded that their method was selective and yielded reproducible data. Electrochemical cells designed to simulate thin electrolytic films formed during atmospheric corrosion have been developed by several investigators [46,47]. Fishman and Crowe [47] have studied the thin film of electrolyte with a potentiostatic polarization technique. The corrosion current increased with an increase of RH. They concluded that the resultant corrosion rates were consistent with those reported from long-term weight loss measurements. Fiaud [46] created a thin electrolytic film (80 m thickness) using the device similar to one developed in the field of thin layer electrochemistry [48]. Platinum and nickel were used as electrodes and sodium sulfate (Na2SO4) solution was used as the electrolyte with change of pH by addition of sulfur dioxide (H2SO4). SO2 gas was introduced into the electrolyte through a membrane. They observed the depolarization effect of SO2, oxidation of SO2, and reduction of SO2 with use of cyclic voltammetry and linear polarization techniques.
CORROSION OF THIN METAL FILMS AND MICROSTRUCTURES
Corrosion of a metal occurs by the same fundamental reactions whether it is used in a large structure like an automobile, a bridge, or a heat exchanger, or in a small structure characteristic of magnetic, optical, or microelectronic devices, or under a protective layer [49]. The uniqueness of each application is tied up in the definition of the environment to which the metal is exposed or which develops with time, as well as the definition of a characteristic size of the corroding material. Since the time to failure of a material (i.e., its lifetime) is normally inversely proportional to the corrosion rate and directly proportional to the thickness of the corroding material (its characteristic size), small dimensions are more susceptible to corrosion failure and loss of properties. For example (see Fig. 1), a 50 nm cobalt magnetic film may be corroded completely in about 38 h at a corrosion rate of 1 A/cm2. The desired lifetime is about five years, so a protective film (e.g., diamond-like carbon) must be used to moderate the rate of attack. The protective layer
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 54
Q
PREVENTION OF METAL CORROSION WITH PROTECTIVE OVERLAYERS
Fig. 1—Thin film materials for magnetic, optical, metal conductor lines and microelectronic contacts make them highly susceptible to have small dimensions that are highly susceptible to corrosion.
must be thin to read or write to the cobalt film with the magnetic head, and defects in the protective layer will lead to localized corrosion attack. Wear and friction are mechanical processes that result from the relative movement of the disk and head. The head is designed to fly very close to the disk to take advantage of the magnetic properties [50], but it comes to rest on the disk surface when the system is idle. Humidity and other factors affect wear and friction, and layers or films may be added to lubricate the magnetic film. Of more interest here, however, are the chemical effects that cause corrosion. Accelerated tests have been used to determine disk reliability [51], tests that measure the affects of wear, friction, atmospheric contaminants, humidity, oxygen diffusion, and galvanic corrosion. Also described by Antler and Dunbar et al. [51] is the comparison between field test experience and laboratory simulated corrosion test results. Earlier results on microelectronics failures are reviewed by Schnable et al. [52], Comizzoli et al. [53], Wood [54], and Stojadinovic et al. [55]. Whatever the mode, the result is a loss of information at the site of degradation and the loss of properties. Better preparation and processing, or better design, may reduce flaws and defects that cause mechanical failure, but they may not reduce corrosion that is the result of the natural instability of the metal in an aggressive environment. Rarely are the thin film metals stable in the environment, so techniques must be found to stabilize the structures and extend the lifetimes. In the other example of Fig. 1, aluminum interconnects in microelectronic devices have characteristic widths of 1 m or smaller. Ionic conduction along adsorbed water layers at the silicon dioxide (SiO2) surface can lead to electrochemical corrosion and “breaks” in the Al conductor.
691
If the corrosion rate were 1 A/cm2, the lifetime of the Al interconnects would be about 48 days, rather than the 10 to 20 year lifetime desired. A protective layer is required for this application as well. The corrosion phenomena of thin films chosen for magnetic, optical, or electrical applications have unique characteristics, but they are often similar to those observed for bulk materials [35,50,56]. Thin films have bulk metallurgical properties in thicknesses larger than 1–3 nm and have the same chemical reactions as well. Both observations lead to the general conclusion that both bulk metals and films 30 nm or thicker will have similar corrosion behavior. On the other hand, thin film materials have small grain size and are prepared for magnetic disk applications in “tracked” or grooved geometries. The small grain size causes films to have more homogeneous properties, with fewer inclusions and smaller chemical segregation effects than ordinary bulk materials. The tracks have sharp edges and dimensions to generate unique morphologies in the films. The homogeneous properties would make the films less susceptible to corrosion, but the defects generated at edges could be sites for enhanced attack. The dimensions and geometry of the tracks may lead to nonuniform chemical composition in the recesses, which would produce localized corrosion effects as well. Atmospheric corrosion has been studied under simulated conditions for thin magnetic films [35], and, as in other cases, it was found that the affect of humidity and atmospheric pollutants was synergistic. The level of humidity may influence the condensation of thin moisture films on surfaces, which will facilitate transport across surfaces and may cause the accumulation of water in microscopic domains. In the latter, the concentrations of dissolved contaminants may approach saturation conditions. The contamination may come from dust or inclusions of other layers [49]. The conditions are difficult to simulate in the laboratory because of the lack of knowledge of local conditions in the microscopic regions that are relevant to the problem. In addition, it has been difficult to make in situ measurements for conditions that simulate atmospheric-corrosion measurements, which would give a direct indication of the processes responsible for corrosive attack. Several standard tests have been developed to assess atmospheric corrosion damage ASTM Practice for Conducting Atmospheric Corrosion Tests on Metals (G50–76 (1990)); ASTM Practice for Rating of Electroplated Panels Subjected to Atmospheric Exposure (B537-70 (1992)) without addressing the mechanism of the attack. The second topic relates to protective layers and encapsulants. Pore-free conventional protective layers over magnetic films are too thick to be compatible with the magnetic properties of thin film disk materials. In addition, polymer films can change the adhesion properties of the surface and interfere with the operation of the magnetic head. Highly resistive but electroactive overlayers could lead to galvanic attack of the substrate through holes in the thin film. Sputtered diamond-like carbon films [51,57–60] could fall into this category (see Fig. 1). A protective layer plated or sputtered over an active metal may have pores and defects that will permit the corrosive medium to contact the active substrate metal and thereby promote galvanic corrosion. The holes or defects may be present on a heterogeneous surface in the geometry of either regular arrays or random arrays. The mathematical
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
692
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 3—Simulations of galvanic interactions in multilayer arrays have been carried out with regular patterns or with (more realistic) Voronoi tessellation representations.
Fig. 2—Multilayer “sandwich” arrays may have underlayers exposed through holes or holidays in the overlayer, and galvanic interactions may enhance the corrosion rate in such systems.
modeling of galvanic behavior in plating corrosion systems has been discussed by Smyrl and Newman [21], where earlier work was also reviewed. They determined the current and potential distributions of galvanic corrosion system, which consisted of anodic disks in a cathodic plane as shown in Fig. 2. They solved the Laplace equation for potential with nonlinear (Butler-Volmer) boundary conditions with the use of finite difference method. The numerical modeling of galvanic corrosion in which the geometry consists of various array forms has been performed by Morris and Smyrl [61–63] in this laboratory. Either regular or random arrays of disks in the cathodic plane were used for the simulation of a heterogeneous surface. Most treatments of the regular array use the symmetry element derived from symmetrical geometry of the system. For mathematical simplicity, a particular hexagonal symmetry element can be approximated by a circular geometry, thus eliminating any angular dependence. For random arrays, the arrays simulated using a Voroni tessellation of the plane into random polygons as shown in Fig. 3 were used for the disk-cathodic plane geometry. The Voroni tessellation has proven to be useful for modelling the transport and mechanical properties of disordered or composite media. The geometrical properties of the Voroni tessellation and algorithms for generating the tessellation have been described by Winterfeld [64]. The model established by Morris and Smyrl included the Laplace equation for potential distribution with nonlinear boundary conditions, and it was solved by a finite element method. The potential distribution of the system was obtained from numerical simulation of a regular array of disks over a cathodic plane. The disks were of alternating sizes (disks with two different diameters) distributed on
the surrounding plane. The models for the tertiary current distribution, which includes both potential distribution and mass conservation with use of a geometry of the random array, are in progress in our laboratory. Since the total anodic current must be equal to the total cathodic current, the area ratio between anodic and cathodic components of the total area is an important parameter in galvanic corrosion. If the currents to each area were uniform, the area ratio is the only parameter that would affect the galvanic interaction for a particular combination of metals. On the other hand, Smyrl and Newman [21] found that ohmic effects in the electrolyte may cause a nonuniform galvanic current distribution on the component areas, and this leads to the conclusion that under such conditions some parts of the cathodic area are not important in determining the total galvanic current. The effect is even more pronounced under circumstances where the electrolyte phase is very thin, that is, galvanic current from cathode to anode flows only near where they join, and more remote areas of each are relatively unaffected by the galvanic coupling. It has recently been found that the active perimeter measure of the interactions is more relevant than the area ratio, and the former may be used to correlate the behavior of several geometries [61–63]. The nonuniform current distribution is also obtained if the cathodic surrounding plane is highly resistive but electroactive. For example, resistive sputtered carbon films would cause the cathodic galvanic current to flow only to areas at the periphery of holes and defects [65], and the total area would not be important in determining the galvanic current. However, smaller holes would increase the galvanic current at a constant area ratio because the active perimeter would increase. In summary, investigations in bulk solution provide a basis on which to begin to analyze atmospheric corrosion behavior. As the electrolyte phase decreases in thickness, ohmic and diffusion effects become more dominant and galvanic coupling is strongly affected. The more remote areas will show the behavior expected for uniform exposure to an aggressive environment. Effects of local composition and local physical geometry then will become more dominant. Behavior in the local areas would be expected to be very similar to the behavior in bulk electrolyte at the same composition conditions. Further general comparisons must be developed as further research is conducted.
COATINGS AND OVERLAYERS FOR CORROSION CONTROL
In the past 15 to 20 years, an explosion of interest in surface modification techniques has mostly involved the deposition
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 54
Q
PREVENTION OF METAL CORROSION WITH PROTECTIVE OVERLAYERS
of thin films, the application of coatings, and the formation of surface alloys. The development of many of the techniques has been driven by the need for the semiconductor electronics industry to create improved processing procedures. As a spin off of the advancing technology, other fields, such as corrosion protection, have benefited from the new processes. A recent panel report [66] has summarized the general surface modification techniques that are used. The techniques are divided into three broad categories: t Low-energy inorganic coating techniques. For the most part these are mature technologies that have been used for many years. t Polymer coatings include traditional paints, thermoplastics, poly(vinyl chlorides), epoxies, urethanes, and poly (methylmethacrylate). t Techniques involving the use of energetic ions. The techniques have developed rapidly in the last 10 to 15 years; several have neither reached maturity nor found use for corrosion protection. Only those designated as low cost and scalable for widespread use are viable for corrosion protection, except in critical applications. In addition, most techniques that require vacuum processing are too costly for most applications. Inorganic sol-gel films are formed from a sol through continuous stages of increasing concentration of a solid precursor. Typically, the sol is a solution of polymeric species or a suspension of “oligomers” including particles in the colloidal size range. During deposition through states of increasing solids concentration, the sol might gel, but the gel state is often a fleeting transient that quickly empties of liquid. Nevertheless, the structures formed during this stage have varying amounts of porosity and influence the structure of the final film. This processing offers good control of composition and homogeneity at low temperatures. It is not directional nor equipment intensive. Complex shapes of arbitrary size can be coated with good uniformity. The cost of high-purity liquid precursors may be high, but for thinfilm applications the materials cost would be acceptable. Films deposited using energetic deposition techniques are dense, highly adherent, have few pinholes, and can be laid down at low temperatures. They are attractive for corrosion protection. Ion-beam-assisted deposition and ion implantation have the best adhesion properties, while RF sputtering has the best throwing power. Three important factors affecting the performance of films are porosity, adhesion, and stress. Although there are compressive stresses, in ion-implanted surfaces, for example, delamination by buckling is practically unknown. Effective porosity in the treated layer could exist due to shadowing of the surface from the beam by contaminating particles. The problem has not been observed, but the exact reason is not known. With the exception of ion implantation [67], only a few studies on corrosion have been done on films deposited using energetic deposition methods. Ion-beam-assisted deposition coatings are adherent and more ductile than bulk materials due to the microcrystalline or amorphous structures. The adhesion is better for the films deposited by energetic beam techniques as compared to films derived from physical vapor deposition. More details may be found in the cited report [66] or in the original literature. Polymeric materials are widely used as protective coatings because they are transport barriers which limit access of reactive species (i.e., water, oxygen, ions) to the substrate
693
surface. Leidheiser and Granata [68] have discussed the roles that each of these species may play in degradation processes on metal surfaces, and, in particular, the role of ion transport through bulk films and “ion channels” in films. Several techniques are discussed in this paper for characterizing ion transport: d-c measurements, electrochemical impedance analysis, under-the-coating sensing, and radio tracer measurements. Characteristic d-c resistances of 1011 ohm-cm2 are observed for films without continuous aqueous pathways through the coating, as first described by Asbeck and Van Loo [69]. The resistance drops to the order of 108 ohm-cm2 if continuous aqueous pathways exist where such pathways have high rates of transport. It is also clear that films and coatings are heterogeneous and the aqueous pathways are surrounded by regions of lower transport rates. The resistance of films may also decrease with time as the ion channels or pathways equilibrate with an external aqueous environment. For films with high resistance and no ion channels, the impedance of the film is dominated by its geometric capacitance. For films of lower resistance, the low-frequency impedance is dominated by the sum of the resistance of the film and the resistance of the electrolyte. If corrosion proceeds under the coating because of ingress of the aqueous environment, the low-frequency impedance decreases in value. It has been argued that there is a strong correlation between the sites for corrosion under the film and the intersection of the ion channels with the underlying surface, but it has been difficult to confirm the correlation with direct observations. The nature of the easy pathways for transport appears to be related to several factors. One of the factors is the presence of pigment and filler particles, which could facilitate the formation of aqueous pathways adjacent to the pigment or filler and would be influenced by the interaction of the particles with the polymer matrix [68]. The channels could also form by coalescence of voids or pores in the polymer matrix, and this would be influenced by the formation processes of the films. Aggregation of solvent in the film could be influenced by the prior history of the film, by the presence of impurities, and by retained solvent. The presence of channels has been demonstrated to be a function of the glass transition temperature (Tg) of films as well. That is, below Tg , the polymer will be brittle unless a secondary, low-temperature relaxation exists, and this will favor the formation of microcracks and defects. Above Tg , the film will be more flexible and less susceptible to formation of fracture channels. Armstrong et al. [70] have investigated the influence of Tg on ion transport and permeability in chlorinated rubber films. Pigment and filler particles can have a beneficial influence because of the reduced transport of water, oxygen, and ions. The effect will depend on the pigment volume fraction, the chemical composition, the geometry, and the dispersion as noted by Burns and Bradley [71]. Equilibrium water uptake may cause plasticization and subsequent depression of Tg, as well as swelling, which counteracts the effects of reduced transport rate caused by the solid particles [72]. Pigments that have oxidizing character can induce passivation of the underlying metal, as observed for chromate or vana-date additives [73]. Other pigments may inhibit the cathodic reaction and thus suppress corrosion as well [74].
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
694
PAINT AND COATING TESTING MANUAL
Q
De-adhesion of organic coatings is responsible for enhanced corrosion rates on one hand and is the result of corrosion on the other hand. Leidheiser [75] has discussed de-adhesion processes which include: loss of adhesion when wet, cathodic delamination, cathodic blistering, swelling of the polymer, gas blistering by corrosion, osmotic blistering, thermal cycling, and anodic undermining. With few exceptions, the loss of adhesion processes requires that reactive species such as water, oxygen, and ions penetrate through the coating. Bonds of the coating with the surface of a steel substrate may be attacked by high pH conditions, which are the result of corrosion reactions or imposed cathodic protection conditions. In either case, OH” ions are particularly aggressive and cause disbonding on steel. In a recent investigation by Stratman [76], disbonding was followed by monitoring the surface potential of a polymer-coated steel surface with a scanning Kelvin probe technique. A recent international meeting [1] reviewed the unsolved problems of corrosion protection By organic coatings, described the current understanding of the technology, and outlined some focus for further progress. In addition to the principles of barrier layer transport that have been described above, there was discussion on the effects of (1) pre-treatment of surfaces, (2) the contribution made by surface inhomogeneities of the substrate, (3) the critical size of a water phase which may be responsible for corrosive attack, (4) stress in the film and stress in the substrate, and (5) incorporation of corrosion inhibitors in protective films. Funke [77] later summarized the continuing uncertainties that exist in studying corrosion protection properties of organic coatings. Some scatter of behavior is caused by the age and history of the coating—fresh coatings are more susceptible to swelling and changes in composition. Disbonding may initiate at defects, but it may also occur in the absence of holidays or defects. The water that is associated with disbonding could be transported along the surface and not by permeation through the film. Ions may also move along the interface. All these considerations have considerable implications for electrochemical characterization techniques. A review of various types of organic coatings and their applications in various service conditions is provided by Tator [78].
SUMMARY
The atmospheric corrosion of metals is one of the most important single problems facing corrosion science and technology. From small nanostructures to large buildings and bridges, coating techniques are being developed to moderate the rate of degradation with some success. The use of low-cost coatings continues to increase as the coatings are made more impermeable and more adherent to the protected substrate. Higher-cost films applied by highenergy methods are finding wider use in critical applications where conventional coatings are inadequate. In all systems where protection is necessary, the early detection of corrosion is desirable in order to plan replacement and maintenance measures and to avoid catastrophic failures. Detection of the presence of corrosion can be accomplished in two ways: detection of the agent that causes corrosion or detection of the results of the corrosion process either on the material of interest or on a specimen of the material. Sensors and monitors are receiving greater attention in accelerated life testing of materials, and eventually they
15TH EDITION
will be developed more widely for operating systems or in portable monitoring systems. The savings to industry and the public at large would be in the billions of dollars if the onset of failure processes could be detected prior to their culmination in a catastrophic event [79].
References [1] Funke, W., Leidheiser, H., Jr., Dickie, R. A., Dinger, H., Fischer, W., Haagen, H., Herrmann, K., Mosle, H. G., Oechsner, W. P., Ruf, J., Scantlebury, J. S., Vogoda, M., and Sykes, J. M., “Unsolved Problems of Corrosion Protection by Organic Coatings: A Discussion,” J. Coat. Technol., Vol. 58, 1986, p. 79. [2] Guggenheim, E. A., Thermodynamics, North-Holland, Amsterdam, The Netherlands, 1959. [3] Lewis, G. N., and Randall, M., Thermodynamics, revised by K. S. Pitzer and L. Brewer, McGraw-Hill, New York, 1961. [4] National Bureau of Standards NBS Technical Notes 2710-3, 270-4, 270-5, U.S. Government Printing Office, Washington, DC, 1968–1971. [5] Latimer, W. M., Oxidation Potentials, Prentice-Hall, Englewood Cliffs, NJ, 1952. [6] JANAF Thermochemical Tables, NSRDS-NBS 37, U.S. Government Printing Office, Washington, DC, 1968–1971. [7] Smyrl, W. H., “Electrochemistry and Corrosion on Homogeneous and Heterogeneous Metal Surfaces,” Comprehensive Treatise on Electrochemistry, Vol. 4, J. O’M. Bockris, B. E. Conway, E. Yeager, and R. White, Eds., Plenum Press, New York, 1981. [8] Newman, J., and Tiedeman, W., “Flow Through Porous Electrodes,” Advances in Electrochemistry and Electrochemical Engineering, Vol. 11, H. Gerischer and C. W. Tobias, Eds., Wiley Interscience, New York, 1978, p. 353. [9] Newman, J., “Mass Transport and Potential Distributions in the Geometries of Localized Corrosion,” Localized Corrosion, R. Staehle, Ed., NACE, Houston, 1974. [10] Sereda, P. J., “Weather Factors Affecting Corrosion of Metals,” Corrosion in Natural Environments, ASTM STP 558, ASTM International, West Conshohocken, PA, 1974, p. 7. [11] Perez, F. C., “Atmospheric Corrosion of Steel in a Humid Climate-Influence of Pollution, Humidity, Temperature, Solar Radiation and Rainfall,” Corrosion (Houston), Vol. 40, 1984, p. 170. [12] Guttman, H., “Effects of Atmospheric Factors on the Corrosion of Rolled Zinc,” Metal Corrosion in the Atmosphere, ASTM STP 435, ASTM International, West Conshohocken, PA, 1968, p. 223. [13] Rice, D. W., Cappell, R. J., Phipps, P. B. P., and Peterson, P., “Indoor Atmospheric Corrosion of Copper, Silver, Nickel, Cobalt, and Iron,” Atmospheric Corrosion, W. H. Ailor, Ed., John Wiley & Sons, New York, 1982, p. 651. [14] Freitag, W. O., “Testing for Indoor Corrosion,” Atmospheric Corrosion, W. H. Ailor, Ed., John Wiley & Sons, New York, 1982. [15] Vernon, W. H. J., “A Laboratory Study of the Atmospheric Corrosion of Metals. Part 1. The Corrosion of Copper, with Particular Reference to the Influence of Sulfur Dioxide in Air of Various Relative Humidities,” Trans. Faraday Soc., Vol. 27, 1931, p. 255. [16] Chawla, S. K., Anguish, T., and Payer, J. E., “Microsensors for Corrosion Control,” Materials Performance, May 1980, pp. 68–74. [17] Kucera, V., and Mattson, E., “Electrochemical Technique for Determination of the Instantaneous Rate of Atmospheric Corrosion,” Corrosion in Natural Environments, ASTM STP 558, ASTM International, West Conshohocken, PA, 1974, p. 239. [18] Phipps, P. B. P., and Rice, D. W., “The Role of Water in Atmospheric Corrosion,” Corrosion Chemistry, ACS Symposium Series, Vol. 235, G. Brubaker and P. B. P. Phipps, Eds., 1979, American Chemical Society, Washington, D.C., p. 89. [19] Rice, D. W., Phipps, P. B. P., and Tremoureux, R., “Atmospheric Corrosion of Cobalt,” J. Electrochem. Soc., Vol. 126, 1979, p. 1459.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 54
Q
PREVENTION OF METAL CORROSION WITH PROTECTIVE OVERLAYERS
[20] Klier, K., Shen, J. H., and Zettlemoyer, A., “Water on Silica and Silicate Surface. Partially Hydrophobic Silicas,” J. Phys. Chem., Vol. 77, 1973, p. 1458. [21] Smyrl, W. H., and Newman, J., “Current and Potential Distributions in Plating Corrosion Systems,” J. Electrochem. Soc., Vol. 123, 1976, p. 1423. [22] Scott, W. D., and Hobbs, P. V., “The Formation of Sulfate in Water Droplets,” J. Atmos. Sci., Vol. 24, 1967, p. 54. [23] Smyrl, W. H., and Lien, M., “The Electrochemical QCM (Quartz Crystal Microbalance) Method,” New Methods and Experimental Approaches in Electrochemistry, T. Osaka et al., Eds., Kodansha, Tokyo, 1993. [24] Smyrl, W. H., and Naoi, K., “Corrosion Studies with the Quartz Crystal Microbalance,” Perspectives on Corrosion, AIChE Symposium Series 278, G. Prentice and W. H. Smyrl, Eds., Vol. 6, American Institute of Chemical Engineers, New York, 1990. [25] Burns, R. M., and Campbell, W. E., “Electrical Resistance Method of Measuring Corrosion of Lead by Acid Vapors,” Trans. Electrochem. Soc., Vol. 55, 1929, p. 271. [26] Sereda, P. J., “Atmospheric Factors Affecting the Corrosion of Steel,” Ind. Eng. Chem., Vol. 52, 1960, p. 157. [27] Enrico, F., Riccio, V., and Martini, B., “An Electrical Resistance Method for Measuring Rates of Corrosion of Electrodeposited Metals in Laboratory Tests,” Trans. Inst. Met. Finish., Vol. 41, 1964, p. 74. [28] Evans, U. R., “Mechanism of Atmospheric Rusting,” Corros. Sci., Vol. 12, 1972, p. 227. [29] White, H. S., “Corrosion Principles in Microelectronics,” Electronic Packaging and Corrosion in Microelectronics, M.E. Nicholson, Ed., ASM International, Metals Park, OH, 1987. [30] Mansfeld, F., “Electrochemical Methods for Atmospheric Corrosion Studies,” Atmospheric Corrosion, Vol. 139, W. H. Ailor, Ed., John Wiley & Sons, New York, 1982. [31] Mansfeld, F., “Evaluation of Electrochemical Techniques for Monitoring of Atmospheric Corrosion Phenomena,” Electrochemical Corrosion Testing, ASTM STP 727, ASTM International, West Conshohocken, PA, 1981, p. 215. [32] Guttman, H., and Sereda, P. J., “Measurement of Atmospheric Factors Affecting the Corrosion of Metals,” Metal Corrosion in the Atmosphere, ASTM STP 435, ASTM International, West Conshohocken, PA, 1968, p. 326. [33] Tomashov, N. D., Theory of Corrosion and Protection of Metals, MacMillan, New York, 1966. [34] Agarwala, V. S., “A Probe for Monitoring Corrosion in Marine Environments,” Atmospheric Corrosion, Vol. 183, W. H. Ailor, Ed., John Wiley & Sons, New York, 1982. [35] McKenzie, M., and Vassie, P. R., “Use of Weight Loss Coupons and Electrical Resistance Probes in Atmospheric Corrosion Tests,” Br. Corros. J., London, Vol. 20, 1985, p. 117. [36] Kucera, V., and Gullman, J., “Practical Experience with an Electrochemical Technique for Atmospheric Corrosion Monitoring,” Electrochemical Corrosion Testing, ASTM STP 727, ASTM International, West Conshohocken, PA, 1981, p. 238. [37] Mansfeld, F., and Tsai, S., “Laboratory Studies of Atmospheric Corrosion. I. Weight Loss and Electrochemical Measurements,” Corros. Sci., Vol. 20, 1980, p. 853. [38] Mansfeld, F., and Kenkel, J. V., “Electrochemical Monitoring of Atmospheric Corrosion Phenomena,” Corros. Sci., Vol. 16, 1976, p. 111. [39] Mansfeld, F., and Kenkel, J. V., “Electrochemical Measurements of Time-of-Wetness and Atmospheric Corrosion Rates,” Corrosion (Houston), Vol. 33, 1977, p. 13. [40] Mazza, B., Pedeferri, P., Re, G., and Sinigaggla, D., “Behaviour of a Galvanic Cell Simulating Atmospheric Corrosion Conditions of Gold Plated Bronzes,” Corros. Sci., Vol. 17, 1977, p. 535. [41] Tosto, S., and Brusco, G., “Effect of Relative Humidity on the Corrosion Kinetics of HSLA and Low Carbon Steel,” Corrosion (Houston), Vol. 40, 1984, p. 507. [42] Pourbaix, M., “Applications of Electro chemistry in Corrosion Science and in Practice,” Corros. Sci., Vol. 14, 1974, p. 25. [43] Vassie, P. R., and McKenzie, M., “Electrode Potentials for on Site Monitoring of Atmospheric Corrosion of Steel,” Corros. Sci., Vol. 25, 1985, p 1.
695
[44] Howard, R. T., “Environmentally Related Reliability in Microelectronic Packaging,” Electronic Packaging and Corrosion in Microelectronics, M. E. Nicholson, Ed., ASM International, Metals Park, OH, 1987. [45] Pourbaix, M., Muylder, J. V., Pourbaix, A., and Kessel, J., “An Electrochemical Wet and Dry Method for Atmospheric Corrosion Testing,” Atmospheric Corrosion, Vol. 167, W. H. Ailor, Ed., John Wiley & Sons, New York, 1982. [46] Fiaud, C., “Electrochemical Behavior of Atmospheric Pollutants in Thin Liquid Layers Related to Atmospheric Corrosion,” Atmospheric Corrosion, Vol. 161, W. H. Ailor, Ed., John Wiley & Sons, New York, 1982. [47] Fishman, S. G., and Crowe, C. R., “The Application of Potentiostatic Polarization Techniques to Corrosion Under Thin Condensed Moisture Layers,” Corros. Sci., Vol. 17, 1977, p. 27. [48] Hubbard, A. T., and Anson, F. C., “The Theory and Practice of Electrochemistry with Thin Layer Cells,” Electro analytical Chemistry, Vol. 5, A. J. Bard, Ed., Marcel Dekker, New York, 1971. [49] Comizzoli, R. B., Frankenthal, R. P., Lobnig, R. E., Peins, G. A., Psotakelt, L. A., Siconolfi, D. J., and Sinclair, J. D., “Corrosion of Electronic Materials and Devices by Submicron Atmospheric Particles,” Interface (USA), Vol. 2, No. 3, 1993, p. 26. [50] Lee, W., “Thin Films for Optical Data Storage,” J. Vac. Sci. Technol., Vol. A3, 1985, p. 640. [51] Antler, M., and Dunbar, J. J., “Environmental Testing of Materials for Indoor Exposure,” IEEE Transactions, Vol. CHMT-1, 1978, p. 17. [52] Schnable, G. L., Comizzoli, R. B., White, L. K., and Kern, W., “A Survey of Corrosion Failure Mechanisms in Microelectronic Devices,” RCA Rev., Vol. 40, 1979, p. 416. [53] Comizzoli, R. B., White, L. K., Kern, W., and Schnable, G. L., Report RADC-TR-80-236, July 1980, Final Technical Report, Contract F30602-78-C-0276, 1 Sept. 1978 to 31 Aug. 1979, Rome Air Development Center (RBRP), Griffiss AFB, NY. [54] Wood, J., “Reliability and Degradation of Silicon Devices and Integrated Circuits,” Reliability and Degradation: Semiconductor Devices and Circuits, M. J. Howes and D. V. Morgan, Eds., John Wiley, New York, 1981. [55] Stojadinovic, N. D., and Ristic, S. D., “Failure Physics of Integrated Circuits and Relationship to Reliability,” Phys. Status Solidi A, Vol. 75, 1983, p. 11. [56] Howard, J. K., “Thin Films for Magnetic Recording Technology: A Review,” J. Vac. Sci. Technol. A, Vol. 4, 1986, p. 1. [57] Jansen, F., Machonkin, M., Kaplan, S., and Hark, S., “The Effects of Hydrogeneration on the Properties of Ion Beam Sputter Deposited Amorphous Carbon,” J. Vac. Sci. Technol. A, Vol. 3, 1985, p. 605. [58] Nyaiesh, A. R., Kirby, R. E., King, R. K., and Garwin, E. L., “New Radio Frequency Technique for Deposition of Hard Carbon Films,” J. Vac. Sci. Technol. A, Vol. 3, 1985, p. 610. [59] Koeppe, P. W., Kapoor, V. J., Mirtich, H. J., Banks, B. A., and Bulino, D. A., “Summary Abstract: Characterization of Ion Beam Deposited Diamond-like Carbon Coating on Semiconductors,” J. Vac. Sci. Technol. A, Vol. 3, 1985, p. 2327. [60] Savvides, N., and Window, B., “Diamond-like Amorphous Carbon Films Prepared by Magnetron Sputtering of Graphite,” J. Vac. Sci. Technol. A, Vol. 3, 1985, p. 2386. [61] Morris, R. G., and Smyrl, W. H., “Galvanic Interactions on Random Heterogeneous Surfaces,” J. Electrochem. Soc., Vol. 136, 1989, p. 3237. [62] Morris, R. G., and Smyrl, W. H., “Current and Potential Distributions in Thin Electrolyte Layer Galvanic Cells,” J. Electrochem. Soc., Vol. 136, 1989, p. 3229. [63] Morris, R. B., and Smyrl, W. H., “Electrode Processes on Heterogeneous Surfaces. I. Galvanic Interactions on Regular Geometry,” American Institute of Chemical Engineers Journal, Vol. 34, 1988, p. 723. [64] Winterfeld, P. H., 1980, “Percolation and Conduction Phenomena in Disordered Composite Media,” Ph.D. thesis, University of Minnesota, Minneapolis, MN. [65] Kassimati, A., and Smyrl, W. H., “Galvanic Corrosion of Sandwich Structures,” J. Electrochem. Soc., Vol. 136, 1989, p. 2158.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
696
PAINT AND COATING TESTING MANUAL
Q
[66] Smyrl, W. H., Halley, J. W., Hubler, G., Hurd, A., MacDonald, D., Snyder, D., and Williams, J., “Corrosion Protection,” Crit. Rev Surf Chem., Vol. 3, 1993, p. 271. [67] McCafferty, E., Natishan, P. M., and Hubler, G. K., “Surface Modification of Aluminum by High Energy Ion Beams,” Interface (USA), Vol. 2, No. 3, 1993, p. 45. [68] Leidheiser, H., Jr. and Granata, R. D., “Ion Transport Through Protective Polymeric Coatings Exposed to an Aqueous Phase,” IBM J. Res. Dev., Vol. 32, 1988, p. 582. [69] Asbeck, W. K., and Van Loo, M., “Critical Pigment Volume and Permeation of Paint Films,” Ind. Eng. Chem., Vol. 41, 1949, p. 1470. [70] Armstrong, R. D., Handyside, T. M., and Johnson, B. W., “Factors Determining Ionic Currents in PVC Protective Coatings,” Corros. Sci., Vol. 30, 1990, p. 569. [71] Burns, R. M., and Bradley, W. W., Protective Coatings for Metals, 2nd ed., Reinhold, New York, 1955. [72] Mastronardi, P., Carfagna, C., and Nicolais, L., “The Effect of the Transport Properties of Epoxy Based Coatings on Metallic Substrate Corrosion,” J. Mater. Sci., Vol. 18, 1983, p. 197. [73] Yamamoto, T., Okai, T., Oda, M., and Okumura, Y., “A Novel Anti-Corrosive Pigment Containing Vanadate/Phosphate,” Advances in Corrosion Protection by Organic Coatings, D.
15TH EDITION
[74]
[75]
[76] [77]
[78] [79]
Scantlebury and M. Kendig, Eds., The Electrochemical Society, Pennington, NJ, 1989. Guest, N., Scantlebury, J. D., John, G. R., and Thomas, N. L., “Metal Complex Agents as Possible Film Forming Anti Corrosives on Mild Steel,” Advances in Corrosion Protection by Organic Coatings, D. Scantlebury and M. Kendig, Eds., The Electrochemical Society, Pennington, NJ, 1989. Leidheiser, H., Jr., “Mechanisms of De-Adhesion of Organic Coatings from Metal Surfaces,” Polymeric Materials for Corrosion Control, R. A. Dickie and F. L. Floyd, Eds., ACS Symposium Series 322, American Chemical Society, Washington, DC, 1986. Stratmann, M., and Streckel, H., “Monitoring the Disbonding of Organic Films by the Kelvin Probe Method,” Ber. Bunsenges. Phys. Chem., Vol. 92, 1988, p. 1244. Funke, W., “Electrochemical Measurements for Characterizing Corrosion Protective Properties of Organic Coatings,” Advances in Corrosion Protection by Organic Coatings, D. Scantlebury and M. Kendig, Eds., The Electrochemical Society, Pennington, NJ, 1989. Tator, K. B., “Organic Coatings and Linings,” Metals Handbook, 9th ed., Vol. 13, 1987, p. 399. Smyrl, W. H., and Butler, M. A., “Corrosion Sensors,” Interface (USA), Vol. 2, No. 4, 1993, p. 35.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
55
MNL17-EB/Jan. 2012
Types of Metal Corrosion and Means of Corrosion Protective by Overlayers Kenneth B. Tator1 and Cynthia L. O’Malley2 INTRODUCTION
THIS CHAPTER WILL PROVIDE A GENERAL OVERview of the cause of metal corrosion, and its prevention by the use of protective coatings, or overlayers. It must be understood, however, that a thorough discussion of corrosion and corrosion mechanisms can extend far beyond the summaries provided in this chapter. In fact, there are numerous books written solely on corrosion, manifestation of the different types of corrosions, and the reasons why such corrosion occurs. Similarly, the use of protective overlayers over a metal to prevent corrosion is likewise a quite varied and extensive topic. The literal interpretation of the word “overlayers,” as used herein, means anything applied, or overlaid on a metal to prevent, retard, or reduce corrosion. This would include monomolecular layers of gasses, oxides, or other materials that may slow or stop the corrosion process under very specific conditions. Moreover, it includes all of the possible metals and organics that can be applied to a surface. Some of the more common overlayers will be discussed herein. Accordingly then, the discussion regarding metal corrosion consists of the following sections: Definitions Extractive metallurgy in reverse Electrochemical corrosion Forms of corrosion Uniform corrosion Galvanic corrosion Sacrificial anode cathodic protection Impressed current cathodic protection Crevice corrosion Pitting corrosion Microbiologically influenced corrosion Intergranular corrosion Selective leaching Erosion corrosion Cavitation Fretting Stress corrosion Hydrogen damage The sections on corrosion protective overlayers consist of: Conversion coatings Phosphate Chromate
1 2
Chromate free Hot dip galvanizing Batch Continuous Cementitious linings Glass and porcelain enamels Electroplating Thermal spray Rubber linings Paints, coatings, and linings
DEFINITION OF CORROSION
Corrosion, as defined by NACE, the Corrosion Society, is “the destruction of a substance ‘usually a metal’ or its properties because of a reaction with its environment.” This definition does not make use of the terms chemical or electrochemical because such terms would define corrosion only as it relates to metals and would not allow its application to many other materials which disintegrate due to environmental exposure [1]. The American Society of Materials (ASM) defines corrosion as a “chemical or electrochemical reaction between a material and its environment that provides a deterioration (change) of the material and its properties” [2]. ASTM defines corrosion with the same definition. Skorchelletti (translated from Russian) [3] states “the spontaneous oxidation of metals, which is detrimental in industrial practice (reducing the service life of articles) is termed “corrosion” (from the Latin “corrodere,” which means to eat away. If we consider corrosion as a deterioration of a material under the influence of an environment, the discussion of corrosion mechanisms and the effects of such deterioration are many and varied. If, however, the definition is restricted to corrosion of metals, the subject becomes more manageable and limited in scope. The mechanisms of metallic corrosion are relatively well understood. The fundamentals of corrosion, as they relate to metallic corrosion, can be summarized in the following sections.
EXTRACTIVE METALLURGY IN REVERSE
Extractive metallurgy consist of mining metallic bearing ores from the earth, such as “iron” ore, “bauxite” (aluminum ore), and the ores of other metals, such as nickel, tin, copper, etc. These ores consist of the metal in combination with oxygen and other elements. They are relatively stable
Chief Excecutive Officer, KTA-Tator, Inc., 115 Technology Dr, Pittsburgh, PA 15275. Laboratory Manager, KTA-Tator, Inc., 115 Technology Dr, Pittsburgh, PA 15275, www.kta.com
697 Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed byASTM International This standard is for EDUCATIONAL USE ONLY. Copyright © 2012 by www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
698
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
TABLE 1—Metals in order of energy required for conversion from their ores. (Courtesy of NACE Publication.) Common Metals Potassium Magnesium Beryllium Aluminum Zinc Chromium Iron Nickel Tin Copper Silver Platinum Gold
Energy Required for Conversion Most
Fig. 1—Metallurgy in reverse. (Courtesy of McGraw-Hill.)
Least
[SOURCE: Corrosion Basics—An Introduction, LaQue, F. L, Chapter 2, Basics of Corrosion, NACE, Houston, TX, p. 23, 1984.]
and nonreactive in the ground, as the metal in the ore has already oxidized and reacted during the passage of time until no further reactions occur. After mining, and during refining purification, large amounts of energy, usually in the form of chemical dissolution, electrolytic separation, and smelting heat, are used to convert the ore into its pure, or alloyed metal. The energy used during the conversion process is indicative of the reactivity of the metal in the environment. The more energy used to extract the metal from its ore, the greater the propensity of the purified metal to react with the environment and return to its former oxidized ore-like state. Some metals exist in nature in a relatively pure metallic form and need little if any energy to extract them from nature. Gold, for example, in nature is found in pure nugget form, and after shaping (into jewelry, a tooth filling, or other product), it can be placed into a variety of aggressive environments and show little, if any, corrosion or oxidation. Similarly, platinum, silver, and copper are relatively pure metals that require little refining energy compared to other metals such as iron or steel. Table 1 entitled “Metals in Order of Energy Required for Conversion from Their Ores” [4], illustrates the amount of energy required for conversion of the ore to the pure metal. If the refining of metal from a metallic ore requires the input of energy, metallic corrosion can be described as the reverse of that process—the metal’s release of energy as it naturally oxidizes and reverts to its ore-like status. Using iron as an example, the process can be illustrated, as in Fig. 1. However, the corrosion of some metals is not necessarily bad. In fact some oxidation reactions are beneficial to a metal. For example, aluminum will quickly oxidize (corrode) forming a layer of aluminum oxide (Al2O3) on the aluminum surface. The aluminum oxide that is formed is a thin, impervious, tightly adherent layer that seals and protects the underlying aluminum from further oxidation and environmental reaction. Only if the oxide layer is broken down in aggressive environments will further reaction of the aluminum occur. Anodizing (the imposition of an electrical current on aluminum or other light metals in a bath to form an oxide layer) is used to provide a thicker, more
protective oxide layer coating on the metal surface than that which results from atmospheric oxidation. Copper is another example of a metal that forms a protective oxide layer—in this case, the oxide is a characteristic color that forms as the copper oxidizes by weathering to form a patina. The main constituent of patina is a mixture of basic copper carbonates, sulfates, and chlorides. In rural areas, or areas where there are no chemical contaminates, the patina is predominately basic copper carbonate (green). In industrial areas basic copper sulfate (blue) predominates. In coastal areas where chlorides are present, copper chlorides (brownish-yellow) prevail. Patina color variations from greenish to blue-gray occur, depending on differences in chemical composition. Regardless of color, the patinaoxide surface layer slows or prevents further reaction of the copper metal. Alloying constituents of stainless steel also enable the formation of a tightly adherent oxide layer that protects the underlying steel from further corrosion. The numerous varieties of stainless steel that exist are due to the variations of metal alloy constituents (carbon, manganese, copper, chromium, nickel, and others) that comprise the oxide layer, and composition of the alloy steel. Conversely, the oxidation of carbon steel, the common low-alloy steel most common for use in structural steel, automobiles, pipelines, and for most industrial and commercial applications, forms a brown or black oxide (rust) that is porous and nonadherent. The rust layer from carbon steel is not protective, and does not prevent further corrosion of the underlying ferrous metal as a result of atmospheric exposure. Progressive oxidative deterioration occurs in a weathering or chemical environment resulting in pitting, general corrosion, and deterioration of the steel substrate.
ELECTROCHEMICAL CORROSION
Electrochemistry involves an electric current in conjunction with a chemical reaction. Electrochemical corrosion will occur when four required elements are present: (1) anode, (2) cathode, (3) metallic pathway connecting the anode and cathode, and (4) electrolyte. Most metals already have three of the four required elements: anode, cathode, and metallic pathway. The anode and cathode form from the metal and this process is discussed subsequently. The metallic pathway is present in all metals as they are ready conductors of electrons. Only the electrolyte is missing. Once an electrolyte is present, corrosion will proceed. An electrolyte is a liquid that contains ions or “charged particles.” Most electrolytes consist of water containing dissolved inorganic salts. Inorganic salts, in water solution dissociate into their respective ions (for example sodium chloride dissociates into sodium cations and chloride anions). These electrically charged species provide an excellent pathway for
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 55
Q
TYPES OF METAL CORROSION
699
Fig. 3—Corrosion cell. (Courtesy of KTA-Tator, Inc.)
Fig. 2—Dry cell battery. (Courtesy of KTA-Tator, Inc.)
the transmission of electric currents generated during the corrosion process. Almost all water, including that in the oceans, rivers, and lakes contains dissolved mineral salts which dissociate. Rainwater, while relatively pure, forms around a small dust mote, or other particulate matter, and leaches ionic material from that particle. Furthermore, freshly condensed rainwater, which initially is “pure” water, will react with carbon dioxide from the air to form weak carbonic acid [5]. This weak acid dissociates into the hydrogen ion (H+) and carbonate (CO–2 ) ions. Pollutants in the air 3 such as nitrogen and sulfur oxides will absorb in water, and will result in further electrolytic dissociation. Thus, when a metal is wet with water, or most chemicals, it is virtually always wetted with an electrolyte. A common dry cell battery shown in Fig. 2 provides a good illustration of how the four elements of a corrosion cell work together to produce a beneficial form of corrosion to generate an electric current and provide power [6]. The battery is essentially a corrosion cell. The anode, and its counterpart, the cathode, are the positive and negative terminals of the battery, respectively. During the electrochemical reaction process, electrical current (or electrons) flows from the anode to the cathode via the metallic pathway or connection (see external circuit). The electrolyte carries ions from the cathode to the anode to complete the electrical circuit. The anode (negative terminal) decays during this process and generates the electron flow, while the cathode (positive terminal) remains intact and is not deteriorated. The only difference between a corrosion cell and a manufactured dry cell battery is that the battery reaction process is designed to produce an electrical current for a productive use. When the anode is depleted in a manufactured battery, the reaction will stop and the battery will “die” since it cannot produce more energy. The battery is replaced with a new one, and electric current is again generated. In a corrosion cell, the anode is often a material that cannot be replaced readily. Corrosion of an automobile
means deterioration of the metal body of the vehicle; structural steel rusting is deterioration of the structural member; tanks, pipelines, ships, offshore platforms, building steel and other structures that corrode cannot be as easily replaced as a dry cell battery! An illustration of a corrosion cell is shown above in Fig. 3 [7]. The corrosion cell also consists of anode and cathode coupled by a metallic pathway in an electrolyte. During the corrosion process, the anode deteriorates by going into ionic solution. The anodic reaction, in every case consists of the metal going into ionic solution as follows: M → M+ + e– where M is a metal, M+ is the metal in ionic form, and e is the number of electrons liberated by the metal going into ionic solution. For example, the ionic reaction for steel is Fe → Fe+2 + 2e– and that for aluminum is Al → Al+3 + 3e–. The cathodic reaction, however, consists essentially of the neutralization of the electrons that are created as the metal goes into ionic solution. These electrons can be neutralized through one of three reactions, depending upon the cathodic environment: Alkaline, neutral, and near-neutral chemical environments: 2H2O + O2+ 4e– → 4OH– Acidic, with oxygen: H+ + e– → H, then 2H + 1 O2 → H2O 2 Highly acidic: 2H+ + 2e– → H2 The electron movement in a corrosion cell (the chemical concept) is defined as moving from the anode through the metallic pathway to the cathode. At the cathode, electrons are neutralized by chemicals in the electrolyte. Conversely, current flow, used in the electrical industry and for cathodic protection (the electrical concept) is opposite, and by definition, electrons comprising the electric current flow from the anode through the electrolyte to the cathode. Fig. 4 “Chemical Concept of Electron Flow” and Fig. 5 “Electrical Concept of Current Flow” [8] illustrate the differences between the electron and current flow in the chemical and electrical concepts. While the illustrations of the dry cell battery (electrical concept) and corrosion cell (chemical concept), as shown above, depict the anode and cathode as separate entities, in practice, anodes and cathodes are not isolated entities. For example, all metals consist of crystals, or “grains” that precipitate when molten metal is cooled from a hot melt. Grain boundaries form between grains, and are anodic
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
700
PAINT AND COATING TESTING MANUAL
Q
Fig. 4—Chemical concept of electron flow. (Courtesy of NACE International.)
relative to the cathodic grain. Areas of stress concentration, in a formed metal, are anodic to unstressed areas. Where temperature differentials exist, hot areas are anodic relative to cool areas. A scratch, nick, or abraded area on a metal, or mechanically worked area is anodic to adjacent unscratched or unworked areas. Oxygen concentration differences can also result in the formation of anodes and cathodes (the lesser oxygen concentration generally becomes anodic, relative to an adjacent area of greater oxygen concentration). An example of this is corrosion at the waterline of a steel piling or on the wall of a dam gate. Cor-
15TH EDITION
rosion within a crevice occurs as an anode develops within the sheltered area within the crevice relative to the more exposed areas, which become cathodic. Accordingly, to a greater or lesser extent, all metals will develop anodic and cathodic areas, and will have a comparative potential to corrode. Moreover, as corrosion proceeds and pits and crevices develop in a metal substrate, the anodic and cathodic areas may switch, and the anodes will become cathodes, and vice versa. A metal in contact with another metal may also become an anode with the other metal becoming the cathode. The tendency of a metal to become an anode or cathode in combination with other metals in a given immersion environment is its corrosion potential. The comparative corrosion potential of a series of metals can be tabulated as the electromotive force (EMF) series. EMF corrosion potentials will vary from environment to environment, and the standard EMF series is that for sea water. Hydrogen is used as an arbitrary reference element, with elements above hydrogen more reactive, while those listed below hydrogen are increasing, inert with a lesser tendency to ionize or go into ionic solution. The EMF series for metals and solutions of their own salts is presented in Table 2 [9], while the EMF series in sea water is shown in Table 3 [10]. If two metals in the galvanic series are coupled, the greater the potential difference between the two metals, the greater the driving force for the more negative metal (anode) to corrode. Note the similarities between the EMF and galvanic series tables with that of the order of metals in order of energy required for conversion from their ores (see Table 1). Thus, all metals, when wetted, either by immersion, precipitation, condensation, or even high relative humidity (generally 50 % RH or more) have an inherent ability to corrode, as the anode, cathode, metallic pathway, and electrolyte are all present. Although, not a prerequisite for the initiation and early commencement of corrosion, oxygen is critical, as without it, the corrosion reaction will slow or stop. Oxygen is a depolarizer in that it removes hydrogen from the cathode and allows additional electrons to be neutralized, enabling the process to proceed. If oxygen is not present, the hydrogen accumulation at the cathode may result in a hydrogen surface film, essentially isolating the electrons migrating to the cathode from the electrolyte and slowing or stopping their electron neutralization process.
FORMS OF CORROSION
Fig. 5—Chemical concept of electron flow. (Courtesy of NACE International.)
The appearance or a manifestation of corrosion has been categorized in various schemes by a number of authors. Categorization by environmental condition (immersion, atmospheric, hot, wet, dry, etc.) is one means of grouping corrosion. Covino and Cramer [11] sub-categorize “forms of corrosion” into six subsections. Four of those forms, “uniform corrosion,” “localized corrosion,” “metallurgically influenced corrosion,” and “microbiologically influenced corrosion” fit under the classification of corrosion that is not influenced by any outside process. The remaining two, “mechanically assisted degradation” and “environmentally induced cracking” address corrosion as influenced by another process (Table 4). Fontana [12] describes eight forms of corrosion that can be identified by visual observation, in most cases with
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 55
TABLE 2—Electromotive force series Electrode Reaction
Standard Electrode Potential E (Vs), 25°C
K=K +e
–2.922
Q
TYPES OF METAL CORROSION
701
TABLE 3—Galvanic series in seawater flowing at 13 FTPS (temperature, approx. 25°C) Material
Steady-StateElectrode Potential, V (Saturated Calomel Half-Cell)
Zinc
–1.03
Aluminum 3003-(H)
–0.79
Aluminum 6061-(T)
–0.76
Cast Iron
–0.61
Carbon Steel
–0.61
–0.762
Stainless Steel, Type 430, active
–0.57
Cr = Cr+++ + 3e–
–0.71
Stainless Steel, Type 304, active
–0.53
Ga = Ga+++ + 3e–
–0.52
Stainless Steel, Type 410, active
–0.52
Fe = Fe++ + 2e–
–0.440
Naval Rolled Brass
–0.40
Cd = Cd++ + 2e–
–0.402
Red Brass
–0.33
In = ln+++ + 3e–
–0.340
Bronze, Composition G
–0.31
TI = TI + e
–0.336
Admiralty Brass
–0.29
Co = Co +2e
–0.277
90Cu 10Ni, 0.82Fe
–0.28
Ni = Ni++ +2e–
–0.250
70Cu30Ni, 0.47Fe
–0.25
Sn = Sn++ +2e–
–0.136
Stainless Steel, Type 430, passive
–0.22
Pb = Pb++ +2e–
–0.126
Bronze, Composition M
–0.23
Nickel
–0.20
Stainless Steel, Type 410, passive
–0.15
Titaniuma
–0.15
Silver
–0.13
Titaniumb
–0.10
Hastelloy C
–0.08
Monel-400
–0.08
+
–
Ca = Ca + 2e
–2.87
Na = Na+ + e–
–2.712
Mg = Mg++ + 2e–
–2.34
Be = Be++ + 2e–
–1.70
AI = AI+++ + 3e–
–1.67
Mn = Mn++ +2e–
–1.05
Zn = Zn++ + 2e–
++
+
-
–
++
–
H = 2H+ +2e–
0.000
Cu = Cu+++2e–
0.345
Cu = Cu+ + e–
0.522
2Hg = Hg2+ +2e–
0.799
Ag = Ag+ + e–
0.800
Pd = Pd++ +2e–
0.83
Hg = Hg++ +2e–
0.854
Pt = Pt++ +2e–
ca1.2
Au = Au+++ + 3e–
1.42
Stainless Steel, Type 304, passive
–0.08
Au = Au+ + e–
1.68
Stainless Steel, Type 316, passive
–0.05
Zirconiumc
–0.04
Platinumc
+0.15
[SOURCE: Encyclopedia of Chemistry, Clark and Hawley, Electrochemistry, Reinhold Publishing Co., p. 338, 1957.]
Prepared by power-metallurgy techniques, i.e., sheath compacted powder, hot rolled, sheath removed, cold rolled in air. b Prepared by iodide process. c From other sources. [SOURCE: Fink, F. W., et al., the Corrosion of Metals in Marine Environment, Battelle Memorial Institute, DMIC Report 254, Distributed by N.T.I.S. AD-712 585-S, pp. 7, 13 (1970), with permission from NACE.] a
the naked eye, but sometimes magnification is necessary. The listing is arbitrary, and the eight forms are more or less interrelated, but they cover practically all corrosion failures and problems. The eight forms of corrosion are described in summary below, with some modifications by the writers. The eight forms are: 1. Uniform corrosion; 2. Galvanic corrosion; 3. Crevice corrosion; 4. Pitting; 5. Intergranular corrosion; 6. Selective leaching; 7. Erosion corrosion; 8. Stress corrosion. Hydrogen damage—Fontana discusses this last form but does not consider it a corrosion phenomena.
Fig. 6 depicts each of the corrosion forms itemized above.
UNIFORM CORROSION
Uniform corrosion is the most common form of corrosion and it is characterized by a chemical or electrochemical reaction that proceeds relatively uniformly over a large area. The metal is relatively uniformly attacked, and the rateofdeterioration of relatively uniform over all areas where corrosion occurs.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
702
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
TABLE 4—Forms of corrosion—adapted from Covino and Cramer 1. Uniform Corrosion a. Aqueous corrosion b. Atmospheric corrosion c. Galvanic corrosion 2. Localized Corrosion a. Pitting corrosion b. Crevice corrosion c. Filiform corrosion 3. Metallurgically Influenced Corrosion a. Stainless steels i. Austenitic ii. Ferritic iii. Duplex iv. High Performance b. Nickel Alloys c. Aluminum (1xxx through 7xxx series alloys) d. Weldment Corrosion i. Steel ii. Stainless steel iii. Non-ferrous alloys 4. Mechanically Assisted Degradation a. Erosion b. Fretting corrosion c. Fretting fatigue d. Cavitation erosion e. Corrosion fatigue 5. Environmentally Induced Cracking a. Stress corrosion cracking b. Hydrogen damage c. Liquid metal embrittlement d. Solid metal embrittlement 6. Microbiologically Influenced Corrosion
GALVANIC CORROSION
This type of corrosion occurs when two metals are in contact, and there is a potential difference between the metals when they are immersed or in a wet environment. The less resistant metal becomes anodic and the more resistant metal becomes cathodic (see Table 3). The anodic metal goes into ionic solution as previously described. In a service environment where two metals are in contact in a corrosive environment, galvanic corrosion can be readily detected, as one of the metals (the anode) will be aggressively attacked, with the most extensive deterioration closest to the other metal (the cathode). If possible, it is best to avoid dissimilar metal contact in immersion or wet environments unless the consequences of galvanic corrosion are known and anticipated. However, there are many instances of inadvertent contact between dissimilar metals in tanks, pipelines, heat exchangers, and other fabricated items. To prevent unwanted corrosion, sometimes the dissimilar metals are both painted in an attempt to stop galvanic attack. However, this usually makes the situation even worse, because if any defect develops in the coating, the galvanic attack will be concentrated at the small defect in the coating on the anodic metal. This may lead to rapid
Fig. 6—Depiction of forms of corrosion.
pitting and perforation. Painting, however, may work if only the cathode, or more resistant metal, is painted. This reduces the cathodic area and with a small defect in the coating on the cathode the corrosion current will be distributed over the much larger unpainted anode. Accordingly, pitting does not occur due to the small cathode/large anode relationship. Cathodic protection is the beneficial use of galvanic corrosion. There are two types of cathodic protection: (1) use of the sacrificial anode and (2) use of impressed current. Both types of cathodic protection utilize an anode that is designed to deteriorate and protect the cathode, which is, by design, the metal that is to be protected.
Cathodic Protection by Sacrificial Anode
Zinc, in the form of a zinc-rich coating, or as a galvanize layer on steel, corrodes preferentially as the anode, and protects the steel, which becomes the cathode in the galvanic couple. The steel is usually a pipeline, tank, ship hull, or another steel item in immersion or in the soil. For pipelines, tank interiors, and ship hull exteriors, zinc, magnesium, or aluminum alloy anodes in the form of blocks or ribbons of metal, are attached or connected to the steel to be protected. Irrespective of whether the anode is a piece of metal or a coating, when it becomes exhausted due to sacrificial galvanic protection, it is replaced by either a new metal anode or a new sacrificial coating system. The costs of replacement of the sacrificial anode are designed to be substantially less than the cost of replacement of the structure being protected. Fig. 7 [13] illustrates a pipeline being protected by a sacrificial anode system.
Impressed Current Cathodic Protection
In this means of corrosion protection, the normal corrosion current is reversed beneficially to make the structure being protected the cathode and the anode an inert material.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 55
Q
TYPES OF METAL CORROSION
703
less than 1 %) of a coated structure’s surface, the current demand and the anode consumption is proportionally reduced.
CREVICE CORROSION
Fig. 7—Cathodic protection of a buried pipeline using anodes. (Courtesy of ASM International.)
Usually, an ordinary AC current is rectified to direct current (DC) with the negative terminal of the DC supply connected to the structure being protected, and positive terminal to the inert anode. The choice of the inert anode depends upon the service environment and the associated current demand for protection. Commonly used inert anodes consist of high silicon cast iron, graphite, conductive polymers, lead alloys, precious metals (platinized titanium or tantalum), and mixed metal oxides, or ceramic oxides. Fig. 8 [14] depicts a pipeline being protected by an impressed current cathodic protection system. Cathodic protection, using either a sacrificial anode or impressed current, is almost always done in conjunction with coating the object being protected. If the object being protected is uncoated, the current demand required to protect the entire surface would be quite high, and accordingly, the rate of consumption of the sacrificial anode, or the power demand for the impressed current protection system, would be quite high also. However, if the structure being protected is coated with an appropriate coating system, cathodic protection would be needed only at defects in the coating system, such as pinholes, scratches, impact, and other mechanical damage, and, as such defects usually make up only a very small percentage (generally much
Intensive, localized pitting can occur within crevices or other areas shielded by corrosion products, dirt, or other materials that enable the creation of a stagnant condition beneath the shield. This often occurs between metals and nonmetallic surfaces in contact. As with most forms of corrosion, the environment must be wet or highly humid, resulting in wetness. In general, the tighter the crevice, the more severe the crevice attack. Crevice attack exists principally on passive metals (stainless steel, nickel, titanium, and aluminum-based alloys). This is because there is a need for a large cathode (the passive surface of the metal) to support the high rate of anodic pitting activity within the shielded area. Initially, crevice corrosion initiates as general corrosion over all surfaces of the metal. The corrosion reaction consists of oxidative dissolution of the metal, and reduction of oxygen to hydroxide ions as shown below: Oxidation: Reduction:
M → M+ + e–
O2 + 2H2O + 4e– → 4OH–
After a short time, the oxygen within the crevice is depleted, and oxygen reduction ceases. However, the metallic oxidation process continues, building up an excess of positively charged metal cations. As these cations cannot be neutralized by oxygen reduction, they draw anions, often chlorides, a ubiquitous highly mobile anion, into the crevice. Chlorides are particularly aggressive compared to most other anions in breaking down protective oxide passivation and accelerating corrosion. This results in increased concentrations of metal chloride within the crevice. The metal chlorides hydrolyze in water, precipitating a metal hydroxide and forming hydrochloric acid, as shown below: M+ + Cl + H2O → MOH + HCl The metal hydroxide precipitation drives the metallic oxidation reaction, resulting in a rapidly accelerating, or autocatalytic corrosion process. Filiform corrosion (a special type of crevice corrosion occurring beneath protective coatings or films) and pitting corrosion of metals are related phenomena.
PITTING CORROSION
Fig. 8—Cathodic protection of a pipeline using impressed current. (Courtesy of ASM International.)
Pitting is an aggressive, insidious, destructive form of corrosion that results in holes, or pits in a metal that is otherwise unattacked, or has only general corrosion. Pits are generally small in diameter, with a depth generally exceeding the pit diameter, but in some cases (notably microbiologically induced corrosion) the pits may be broad and bowl shaped in appearance. Pits generally grow downward in the direction of gravity. Pitting on vertical surfaces is less common, and on overhead surfaces (where the growth would be upward), is rare. Pitting may occur within weeks, or may take years to form. Pits are hard to detect and measure because they are often hidden by general corrosion, and are filled with
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
704
PAINT AND COATING TESTING MANUAL
Q
corrosion product. The most common form of pitting failure is perforation. Structural strength of the member (plate, tube, pipe, etc.) is otherwise unaffected, except that the pitted item has a small hole in its cross-section. The mechanism of pit growth is essentially identical to that for crevice corrosion—the pit bottom is anodic, and rapid metal dissolution into ions occurs there, while the sides of the pit are cathodic where oxygen reduction takes place. As metal ions build up in the bottom of the pit, the excess of positively charged ions attract chloride ions and form metal chlorides. The metal chlorides hydrolyze and precipitate, as illustrated previously under for crevice corrosion, driving more metallic dissolution, and accelerating pit growth. Most pitting is associated with chlorides, bromides and hypochlorites. The best means of preventing pitting are to use a metal alloy resistant to pitting in the environment of service, or to isolate the metal from the environment by a resistant overlayer.
Microbiologically Influenced Corrosion
Bacterial induced corrosion can cause pitting in immersion conditions. Sulfate reducing and oxidizing bacteria are responsible for corrosion in sewage piping, and many petrochemical environments (in tanks, and downhole) where sulfur compounds are present. Sulfates are ubiquitous in the environment, as sulfur comprises 0.06 % of the earth’s crust. Sulfur is important in the building of plant and animal tissues. It is found in combination with many other elements, and is a constituent of coal and petroleum, and the cause of acid rain. Sulfates are found in most soils and waters. Anaerobic (oxygen deficient) sulfur reducing bacteria Desulfovibrio desulfuricans reduce sulfate to sulfide as follows:
15TH EDITION
grain matrix. However, under certain conditions, such as in highly acidic environments, the interface between the grains becomes very reactive, and aggressive intergranular corrosion results. With most forms of intergranular corrosion, only the grain boundary deteriorates, and there is little, if any, corrosion to the grain matrix. Intergranular corrosion can be caused by impurities at a grain boundary, enrichment of one of the alloy elements, or depletion of one of these elements in the grain-boundary area. Depletion of chromium in the grain-boundary regions of stainless steels results in intergranular corrosion. Austenitic (18-8) stainless steels are particularly susceptible, particularly when heated within a temperature range of 950 to 1,450°F. Chromium carbide precipitates from the solid solution, depleting a zone near the grain boundary. This leads to corrosion because the chromium-depleted zone does not have sufficient resistance in a given corrosive environment. The appearance of intergranular corrosion results in a “sugary,” granular appearance due to attack of the grain boundaries, leaving small, protruding “grains” (appearing like sugar granules) of metal. Intergranular corrosion sometimes occurs when austenitic stainless steels are welded and not annealed by water quenching after heating. Annealing, a rapid cooling by quenching, does not give sufficient time for chromium carbide precipitation.
SELECTIVE LEACHING
The sulfuric acid so formed will aggressively attack metals and concrete. The cycle of anaerobic sulfate reduction to sulfides, and sulfide oxidation to sulfuric acid is a major problem in the corrosive deterioration of concrete sewage piping, and buried steel piping. Metallic attack by this process will result in somewhat distinctive bowl shaped pits. There are other oxidizing and reducing families of bacteria, but those cited above are perhaps the best known and most prolific. Other metals pitted by bacterial action can include aluminum, stainless steels, and copper.
Selective leaching is the removal of one element from a metal alloy. The removal of zinc from brass alloys, or the removal of iron or steel from a cast iron alloy, are perhaps the two most common examples of selective leaching. The mechanism of dezincification of brass alloys is that initially, the brass dissolves in the corrosive environment; the zinc ions stay in solution; and the copper from the brass (approximately 30 % zinc and 70 % copper) plates back onto the brass, resulting in a copper-rich, zinc-deficient plug or layer. Dezincification of brass is usually prevented by using a less susceptible brass alloy (for example, 15 % zinc, 85 % copper) or adding small amounts of arsenic, antimony, or phosphorus as dezincification inhibitors. Graphitization of gray cast iron occurs because the iron is anodic to graphite and dissolves, leaving a porous mass consisting of graphite (a form of carbon) with little or no iron. There is no dimensional change to the cast iron, and graphitization is visually hard to detect, even though the metal has lost its ductile strength and becomes very brittle because the structure (often a pipe) looks the same as it did before graphitization occurred. Many old, buried sewer and water pipes are made from gray cast iron, which after graphitization, crack due to soil settlement, movement of heavy equipment over them, excavation, or impact. Graphitization may be prevented by using ductile (nodular) cast iron, because the graphite agglomerates in nodes and a graphite network is not present in the malleable cast iron to hold the residue together in the original form after loss of iron from the alloy.
INTERGRANULAR CORROSION
EROSION-CORROSION
SO4–2 + 4H2 → S–2 + 4H2O For steel, the corrosion product is iron sulfide, which precipitates when ferrous (steel) ions and sulfide ions react. The FeS precipitate film forms on exposed steel, and where continuous, is protective. However, FeS is cathodic to steel, and where the film is broken, aggressive galvanic corrosion may occur, resulting in pitting. Conversely, aerobic (oxygen present) bacteria Thiobacillus thioxidans can oxidize elemental sulfur or sulfur containing compounds to sulfuric acid, as follows: 2S + 3O2 + 2H2O → 2H2SO4
As previously mentioned, grain boundaries are anodic with respect to the grain. As a rule, this is not a problem, as the grain boundary is only slightly more anodic than the
Erosion-corrosion is the acceleration or increase in the rate of deterioration by the relative movement between a corrosive fluid and a metal surface.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 55
Fig. 9—Depiction of cavitation. (Courtesy of KTA.)
The appearance of erosion-corrosion is distinctive as grooves, gulleys, waves, and rounded holes and valleys appear on the surface in a directional pattern. An Illustration is shown above. Velocity, either of the metal through the corrosive fluid or the corrosive fluid over the metal, increases the rate of attack, as a general rule. However, in some cases, increased velocity will reduce corrosion if it prevents deposition of materials that may cause crevice corrosion, or exposes the metal to inhibitors (chromates and sodium nitrite) in the fluid. Many stainless steels have a tendency to pit and suffer crevice corrosion in relatively stagnant water, but with movement or increased fluid velocity, that same metal can be used successfully because the motion prevents formation of deposits on the surface that could cause the initiation of pits and crevice corrosion. Metals that form a passive layer on their surfaces, such as stainless steels, aluminum, lead, copper, and brass, are susceptible to erosion-corrosion, because the eroding fluid may wear away the passive layer, exposing the metal to the corrosive immersion environment. Erosion-corrosion can be prevented or minimized by the choice of more resistant materials; design to slow flow or turbulence; removal of entrained solids; the addition of inhibitors; and the use of coatings, including hard, abrasion resistant metallic coatings or soft, energy absorbing rubberlike coatings to isolate the metal from the environment. Cavitation is a special form of erosion-corrosion caused by the formation and collapse of vapor bubbles in the liquid adjacent to the metal surface. If pressure on water is reduced sufficiently, it will boil at room temperature. Similarly, any dissolved air or other gases, will form bubbles at reduced pressure. When these bubbles are formed on or adjacent to a metal surface, and then subjected to rapidly increasing pressure, the bubble will collapse and cause an explosive shock to the surface. The surface shock causes a localized deformation and pitting, resulting in removal of the surface metal and a general roughening of the surface. Cavitation produces rounded microcraters on the metal surface. Sharp pressure pulses caused by collapsing bubbles are highly localized and can remove weak or soft portions of microstructural phases of an alloy. Fig. 9, depicts cavitation corrosion. Examples of cavitation can occur when pumping water with a piston. When a cylinder is full of water
Q
TYPES OF METAL CORROSION
705
and the piston is raised away from the water, pressure is reduced and the water vaporizes, forming bubbles. If the piston is again pushed toward the water, pressure is increased and the bubbles condense or collapse. In a similar fashion, the turning of a pump impeller through a liquid draws the liquid into the flowstream, reducing its pressure. As the impellor pushes the liquid downstream, pressure on the liquid is increased at the impellor surface, collapsing any bubbles that may have formed. A ship’s propeller, pushing the ship through the ocean, is subject to cavitation for the same reason. Calculations have shown that rapidly collapsing vapor bubbles produce shock waves with pressures as high as 60,000 pounds per square inch [15]. Fretting corrosion is related to erosion-corrosion and occurs at contact areas between materials under load subject to vibration and slip. Fretting corrosion appears as pits or grooves in the metal surrounded by corrosion products. Fretting occurs in the atmosphere rather than under aqueous conditions.
STRESS CORROSION
Stress corrosion results in cracking caused by the simultaneous presence of tensile stress and a specific corrosive medium. Stress corrosion cracking (SCC) is particularly insidious, as it is not predictable in many cases, and may result in a catastrophic loss of metal strength. The metal surface is generally unattacked, but fine cracks progress through it as a result of the SCC. For SCC to occur, the metal must be stressed and be in a corrosive environment. Metallic stress alone may result in creep, metal fatigue, and ultimately tensile failure; and a corrosive environment alone will cause deterioration of a metal, the simultaneous presence of both stress and a corrosive environment will sometimes produce disastrous SCC. Both intergranular (between grains) and transgranular (through grains) occur with SCC. Cracking generally proceeds perpendicular to the applied stress. Increasing the stress decreases the time before cracking occurs. As is the case with most chemical reactions, SCC is accelerated by increasing temperature. Most metal alloys are subject to SSC in specific environments. Carbon steel, high-strength steels, austenitic stainless steels, high nickel alloys, aluminum, titanium, magnesium, and zirconium will all crack under stress in certain specific environments. Environments that cause SCC are usually aqueous, and an environment that causes SCC in one alloy may not cause SCC in other alloys. Changing the temperature, degree of aeration, and/or the concentration of ionic species within the corrosion solution may change an innocuous environment into one that causes aggressive SCC failure. Also, an alloy that is immune in one heat treatment may be susceptible in another. As a consequence, the listing of all possible alloy environment combinations that may cause SCC is virtually infinite. SCC can initiate and propagate with little evidence of corrosion until catastrophic failure occurs. Cracks frequently initiate at surface flaws or defects in the metal as a result of rolling processes, corrosion, or wear. The cracks, under stress, grow with little evidence, even in metals that are normally quite ductile. Often, microscopic examination is necessary to verify the presence of SCC.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
706
PAINT AND COATING TESTING MANUAL
Q
Corrosion fatigue is a related form of SCC and is the tendency of a metal to fracture under repeated cyclic stressing. Because SCC is hard to predict, it is also difficult to anticipate means for prevention. However, some generalized methods for reducing SCC are reducing the stress of the metal (perhaps by stress-relief annealing); changing the metal alloy to one more resistant in a given environment; the use of coatings (to isolate the metal from the corrosive environment); work hardening by shot-peening to produce residual compressive stresses on the metal surface; the use of inhibitors such as phosphates in water solution; application of cathodic protection, where it is positively known that SCC is the cause of failure.
HYDROGEN DAMAGE
The penetration of nascent hydrogen (atomic hydrogen) into a metal structure may cause blistering and embrittlement of that metal. Hydrogen penetration may be accelerated by cathodic protection using either impressed cathodic currents or sacrificial anodes. Acid pickling or corrosion in environments containing hydrogen and sulfide, typically that in vessels handling sour [hydrogen-sulfide containing hydrocarbons], may also increase the incidence of hydrogen damage to steel. Atomic hydrogen (H), because of its extremely small atomic size, is capable of diffusing through steel and other metals. The reduction of hydrogen ions involves the production of hydrogen atoms that can be in contact with a metal surface prior to their combination into a hydrogen molecule. The application of cathodic protection, electroplating, and other processes are sources for introduction of nascent or atomic hydrogen into a metal. Certain substances such as phosphorus, arsenic compounds, and sulfide ions reduce the rate of hydrogen ion reduction by decreasing the rate at which hydrogen combines to form molecules. Accordingly, in the presence of these materials, there is a greater concentration of atomic hydrogen on a metal surface. Because of its small size, the hydrogen atom can penetrate into a metal. Hydrogen pick-up can result from welding, heat treating, corrosion, or during melting of the alloy. Once in a metal, the hydrogen atoms may accumulate at voids or other internal surfaces within the alloy to form molecular hydrogen. As the concentration of molecular hydrogen increases at these micro-structural discontinuities, high internal pressure is created that can initiate cracking or blistering. The presence of atomic hydrogen in a metal matrix has been theorized to lower the cohesive force between metal atoms such that local tensile stress perpendicular to the plane of a crack becomes equivalent to or greater than the lattice cohesive force, causing fracture and cracking. At high temperatures (400 to 1,110°F; 200 to 595°C), hydrogen can react with carbon in the metal alloy to form a hydrocarbon—typically methane. This can result in a surface or internal decarburization, whereby carbon is removed from the alloy, leading to embrittlement of the metal. With all hydrogen permeation, higher strength alloys are more susceptible to hydrogen cracking, and higher stresses caused by nascent hydrogen penetration cause cracking to occur more rapidly. Hydrogen embrittlement (penetration of atomic hydrogen at low temperatures) is a reversible process, as opposed
15TH EDITION
to hydrogen attack (penetration of high temperature, highpressure hydrogen). Baking at 190°C (375°F) for four hours will usually suffice in removing hydrogen from most steels. Cadmium coatings on steel require a longer baking time because atomic hydrogen diffuses more slowly through that metal. Baking times up to 24 hours at 190°C (375°F) are common for cadmium-plated parts.
USE OF PROTECTIVE OVERLAYS TO PREVENT CORROSION
Corrosion Prevention with Protective Overlayers
A number of overlays on a metal surface (in the form of coatings) can be used to isolate the metal from a corrosive environment (barrier coating), provide galvanic sacrificial protection, or inhibit corrosion by the use of passivating pigments. The common type, of coatings used as overlays on a metal surface are discussed herein: conversion coatings (phosphate, chromate, and chromate–free); cementitious linings; glass and porcelain enamels; electroplating; thermal spray coatings; rubber linings; and paints, coatings, and linings.
CONVERSION COATINGS
Conversion coatings can be defined as a material applied to a metal surface that chemically reacts and converts the surface of the metal into a compound that becomes part of the coating. There originally were two principal kinds of conversion coatings: those based on phosphate chemistries and those on chromate chemistries. Because of chromate toxicity, a third type of conversion coating is emerging— those based upon chromate-free chemistries.
Phosphate Conversion Coatings
A phosphate layer M3(PO4)2 · (nH2O) is formed on a metal substrate by exposure to a phosphating bath. M stands for a metal and represents divalent metallic cations of one or more metallic alloying elements in the phosphating bath that deposit onto the metal being phosphated. n is the number of waters of hydration attached to the metal phosphate molecule comprising the phosphate layer. Steel, aluminum, copper, and magnesium, and their alloys are most commonly phosphated. Zinc phosphate is by far the most commonly used phosphate solution and it is applied usually by immersing the work piece in a bath or by spraying in a cabinet or tunnel. For painting purposes, crystalline zinc phosphate layers are formed on steel, zinc galvanized steel, and aluminum. The weight of the phosphate coating ranges from 1 to 7 g/m2 (0.0033 to 0.023 ounces per square foot). The phosphate solution consists of a primary phosphate dissolved in a dilute solution of phosphoric acid. When a metal component is immersed in the solution or sprayed with the solution, the acid attacks the metal surface. The attack to the metal reduces the concentration of the acid in the localized area of the attack at the metal interface. With a less acidic environment, the primary phosphate decomposes into a sparingly soluble secondary phosphate and an insoluble tertiary phosphate, both of which precipitate onto the surface at the areas of lower acid concentration. In the process, they are molecularly bonded to the metal. The resulting phosphate layer consists principally of phosphate crystals, but may have amorphous phases between crystals.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 55
The decomposition of the primary phosphate to the secondary tertiary phosphates release hydrogen ions that restore the acid level in the acid-depleted region. The process continues until the desired thickness of the phosphate layer is attained. This mechanism of coating formation is the same for both zinc and iron phosphatizing processes. Manganese phosphate baths are used for heavier coatings. Most conversion coatings used for painting pretreatments consist of zinc in combination with various amounts of Fe++, Ni++, or Mn++ ions. The major advantage of phosphating pretreatments are an increase in surface area for paint bonding provided by the crystalline phosphate layer and a significant reduction of underfilm corrosion at scribes, discontinuities, and mechanical damage when the phosphate-treated steel is exposed to a corrosive, aqueous environment compared to an unphosphated metal. Most automobile bodies are zinc phosphated prior to painting to increase corrosion resistance and paint adhesion.
Chromate Conversion Coatings
Chromate conversion coatings are formed by chemical or electrochemical treatment of metals or metallic coatings in solutions containing hexavalent chromium (Cr6+). The process results in the formation of an amorphous protective coating comprised of the substrate, complex chromium compounds, and other components in the chromating bath. The most common chromating solution is zinc chromate, and the film formation is similar to that for phosphatizing. The metal being passivated is attacked by the acid (chromic acid and some sulfuric acid), bringing about a decrease in acidity at the metal surface as the acid is consumed during the attack. The hexavalent chromate is reduced to trivalent chromium, which precipitates as a complex chromate gel, bonding itself to the metal. The reduction of the hexavalent chromium results in the formation of hydrogen ions, raising the acidity level at the metal/liquid interface after precipitation of the chromate gel. It is believed that the gel film is formed with chromic hydroxide, providing an insoluble matrix with basic chromic chromate absorbed in it as a soluble component. The gel is soft for the first 24 h or so, after which it hardens and becomes relatively abrasion resistant. Chromate films can be produced on steel, cadmium, copper, and magnesium substrates. Aluminum and aluminum alloys are passivated principally by an acid process consisting of chromate-fluoride and chromate-fluoride-phosphate treatments. Chromate conversion coatings are most frequently applied by immersion or spraying, but can be brushed, roll coated, dipped, squeegeed, or applied in other ways. Chromate conversion coatings are applied primarily to enhance bare or painted corrosion resistance to undercutting and underfilm corrosion, and to improve the adhesion of paint or other organic coatings. Galvanized steel is passivated by immersing or spraying the freshly prepared galvanized strip in a chromating solution for one to three seconds immediately before removing excess chromate by squeegee rolls. The conversion coating is applied to a weight range from 0.01 to 0.02 g of chromium per square meter (1 to 2 mg of chromium per square foot). Chromium passivation of galvanized steel can be classified as clear, iridescent, and colored (yellowish), depending upon its thickness. Corrosion
Q
TYPES OF METAL CORROSION
707
protection provided by these coatings increases with color (thickness).
Chromate-Free Conversion Coatings
Because hexavalent chromium (Cr6+) is a carcinogen and toxin, there has been an attempt to replace chromatecontaining products with safer and more environmentally acceptable alternatives. However, chromate conversion coatings have been successfully used for a long time and the chromating process consists of both cleaning and deoxidizing/desmutting prior to application of the conversion coating. These preparation procedures, prior to application of the chromate conversion coating, have been developed and refined over the years to optimize the performance of the chromate conversion coating. For chromate-free substitutions, the development and optimization of surface preparation and cleaning methods is not as nearly perfected. However, there are technologies under development that may be promising for chromate-free conversion coatings used in various industries. Some of the chromate-free conversion coatings currently under development or evaluation are based on the following technologies: Titanium and zirconium fluorocomplexes—These systems consist of an acidic hexafluoro metal complex (H2ZrF6 or H2TiF6) compound, and commonly a polymer. The added polymers improve corrosion resistance and adhesion of subsequent coats. Titanium and zirconium fluorocomplex conversion coatings are in use on sheet stock for the canning industry and are being evaluated by the automotive industry. Cerium-based conversion coatings—These coatings are being developed for use on aluminum, both for architectural and aerospace applications. The coating generally consists of cerium chloride (CeCl3) or cerium nitrate [Ce(NO3)3] and a silicate (SiO2) seal coat. Cobalt-based conversion coatings—A cobalt nitrate [Co(NO3)2] bath produces layers of cobalt oxides over an aluminum substrate. Cobalt-based conversion coatings are being developed for the marine, automotive, and aerospace industries. Molybdenum-based conversion coatings—These coatings are being developed for use on tin plate and zinc for galvanized surfaces. They have met with little success on aluminum. Hydrotalcite coatings—These consist of clay-like mineral compounds such as hydromagnesite compounds in conjunction with an alkaline lithium salt solution. A system of polycrystalline compounds is formed on aluminum and aluminum alloy substrates that has good corrosion resistance. Subsequently applied organic coatings adhere well to the hydrotalcite coating surface. Siliane coatings—Silanes form a silicate layer on a metal surface; they tend to have good corrosion resistance due to a lack of organic groups, and the ability of surface silanol groups to form siloxane bonds as they age. These are being developed for the automotive industry. Conductive polymer coatings—Conductive polymers, including polyaniline, polypyrroles, and polythiophenebased coatings dispersed in a polymer matrix react with iron on a steel surface and form an iron oxide coating from 0.04 mils to 0.8 mils (1–20 μm thick). These coatings are being developed for ferrous metals.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
708
PAINT AND COATING TESTING MANUAL
Q
HOT-DIP GALVANIZING
Hot-dip galvanizing consists of immersing an object to be protected in a bath of molten zinc. The protective coating so developed consists of zinc-iron alloy of variable composition on the surface of the steel. Closest to the surface, the iron content is high relative to the zinc, and proceeding through the protective layer cross section, the zinc content increases in the alloy until the surface layer is all zinc with no iron. Fig. 10 [16] shows a photomicrograph of the cross section of a hot-dip galvanize coating. At the zinc/steel interface, there is an alloying and diffusion of the zinc into the steel and a gradual transition of iron-zinc intermetallics until a relatively pure zinc layer is reached at the outer layer. The outer zinc layer is relatively soft, but the iron-zinc alloy layers are quite hard and from four to six times more resistant to abrasion than the pure zinc layer. The compositions of the various alloy layers are shown adjacent to the photomicrograph. Galvanizing is applied by either of two methods: batch galvanizing or continuous galvanizing. Batch galvanizing involves cleaning of the steel to be protected, applying a flux to the surface of the zinc bath, and immersing the steel item through the flux into the molten bath of zinc for various periods of time. Thickness of the coating is controlled by the composition of the steel substrate, immersion time, and bath withdrawal rate. (Fast withdrawal rates result in thicker galvanized films.) Batch galvanize thickness is usually between 2 and 4 oz/ft2, or 1.7 to 3.4 mils per side (610 and 220 g/m2 and 43–86 μm in thickness). Continuous galvanizing consists of moving a steel sheet, strip, or wire through a tank of molten zinc in a continuous flow. Continuous galvanize coatings are applied at speeds of over 90 m/min (300 ft/min). Coating weights vary from 90 to 1100 g/m2 (0.29 to 3.6 oz/ft2). Coating thickness equivalents are from 6 to 78 μm; 0.24 to 3.1 mils. The zinc coating may be on only one side of the sheet, of equal weight on both sides of the sheet, or of different weights on each side of the sheet. Most sheet coating lines use a gas-wiping technique in which a stream of air or nitrogen gas is directed against the emerging sheet, blowing off excessive zinc and cooling it. A coated galvanized sheet is often oiled or coated with a chromate conversion coating to inhibit white rusting and superficial corrosion during storage in transit. Application of a wax on galvanized wire serves as a white rust preventative and facilitates handling during subsequent processing. All galvanized coatings on steel protect by acting both as a barrier and providing cathodic protection to the underlying steel. A fresh galvanized surface reacts with oxygen to
15TH EDITION
form zinc oxide (ZnO). In the presence of moisture, zinc oxide converts to zinc hydroxide (ZnOH2) and further reaction with carbon dioxide in the air results in the formation of zinc carbonate (ZnCO3), which is relatively insoluble and provides a gray patina on the surface of weathered galvanize. Cathodic protection is provided to the steel at discontinuities, scratches, and mechanical damage due to the fact that zinc is anodic to steel in most environments.
CEMENTITIOUS LININGS
Linings based upon cement are used as sprayable linings, as centrifugally-cast linings for water pipe interiors, and to add weight to submersed pipelines. For corrosion protection purposes, the high alkalinity (pH of 12) of the cement, due principally to calcium silicates and some aluminum silicates, prevents corrosion of steel. Additionally the great thickness (1/2 to 2 in. or more) and inorganic nature of the lining renders it highly abrasion resistant, and resistant to mechanical damage. Ductile and steel water pipes sometimes have a Portland cement lining centrifugally cast to their interior as part of the manufacturing process. Centrifugal casting provides a more dense cement lining making it more resistant to leeching of alkalinity and more abrasion resistant than sprayed linings. Sprayed cement linings, are more porous, but still provide good corrosion resistance, abrasion resistance, and heat resistance. Spray application is necessary on in situ structures. Often polymers, such acrylic, styrene-butadiene, and polyvinyl acetate latexes, and epoxy emulsions are added to the cement mixture in percentages up to about 20 % to provide better moisture impermeability, acid resistance, tensile strength, and adhesion. Cement is also added to pipeline exteriors as a coating or as a casting to provide weight to keep the pipe from rising when buried or submersed in water.
GLASS AND PORCELAIN ENAMELS
These coatings are based on complex borosilicate glasses that are produced by quenching a molten glassy mixture. Frits are formed by shattering the glass into small granules by mechanical means. The composition of the frits will vary widely depending upon the intended use for the enamel coating. Porcelain enamels are used widely for chemical processing vessels, agricultural storage tanks, piping and pipe components, and as well for water heater tanks, cookware, and as appliance coatings. They can be formulated to have excellent chemical resistance, particularly to alkalis (at room temperature) and many acids. Water resistance is excellent at room temperature and decreases somewhat at elevated temperatures. Weather resistance, even in highly corrosive, polluted, industrial environments, is excellent. Solvent resistance to common organic solvents, greases, and oils is also excellent. High temperature resistance, depending upon formulation, is excellent up to 2,000°F (1,095°C). However, as might be imagined, glass and porcelain enamels do not have good flexibility and will crack when subjected to bending or twisting. Additionally, when impacted, the thicker coatings may fracture and chip.
ELECTROPLATING Fig. 10—Photomicrograph of galvanized steel cross-section, and alloy layer compositions. (Courtesy of GalvInfo Center.)
Electroplating involves the electrodeposition of a metallic coating onto a base metal to improve the corrosion resistance, appearance, or properties of that base metal.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 55
Electroplating is commonly done to improve corrosion resistance, and examples are: chrome-plated steel automobile bumpers; electrogalvanizing or zinc plating of steel strip and sheet metal; zinc and cadmium plating of screws, nut, bolts, and other small hardware items; tin plating of steel strip for food packaging and other container uses. Electroplating consists of the deposition of a layer, usually a thin metal, on an object by passing an electric current, usually DC, through a bath within which the object being plated is immersed. The object being plated is one of the electrodes, usually the cathode. The anode is the metal being deposited, and it dissolves as the plating process proceeds. The plating bath contains dissolved salts of the metals that are being deposited. The bath also contains appropriate acids, bases, and buffering salts to hold the pH to the desired level. Addition agents, such as boric acid, glycine, urea, and gelatin may be added to the bath to regulate plating texture to give, for example, a smooth mirror-like finish. To plate out metals above hydrogen from an aqueous solution, the concentration of hydrogen in the cathodic film must be low. Hence, a basic solution such as that provided by a cyanide bath may be used. To plate out monovalent ions such as silver (Ag+1) will require one electron per ion. Multivalent ions such as gold (Au+1 and Au+3) will require one and three electrons per ion, respectively. In theory, one Faraday of electricity will deposit one equivalent weight of any metal; 107.868 grams of silver, 196.967 grams of Au+1, and 196.967/3 = 65.66 grams of trivalent gold. However, this ideal is almost never realized and the current efficiency is defined as the actual mass of metal deposited/the theoretical mass of metal that would have been deposited using the same amount of electrical current. Metals commonly plated are copper, chromium, nickel, iron, cadmium, zinc, indium, tin, lead, silver, and gold. Brass, bronze, zinc-iron and zinc-nickel alloys are also plated. Each of these metals require a different bath, and different bath operating parameters (temperature, cathode current density, voltage, pH, addition agents, etc.).
THERMAL SPRAY COATINGS
Thermal spray utilizes heat to melt a material (usually a metal, ceramic, organic polymer, or combinations thereof) for corrosion protection. Steel surfaces are by far the most widely used substrate for thermal spray coatings, although any metal or non-metal compatible with the coating may be coated. The heating methods used are flame (combustion), electric arc, and electric plasma spray processes. In all cases, a feed stock, usually consisting of a powder or wire, is heated, atomized, and propelled onto a suitably prepared substrate where, upon impact, it consolidates for form a protective, adherent coating. For purposes of corrosion protection, often metallic and ceramic coatings are sealed with one or more layers of a polymeric topcoat. Sealers are typically thinned coatings including epoxies, polyurethanes, phenolics, and silicones. Thermally sprayed coatings can be readily built to thicknesses from 0.004 to 0.08 in. (0.1 to 2 mm). Thermally sprayed metals for protection of steel substrates are most commonly metallic zinc, aluminum, and zinc-aluminum alloys (85/15 zinc-aluminum). Other metals can be thermally sprayed such as stainless steel, bronze, nickel, tin, and other exotic metals.
Q
TYPES OF METAL CORROSION
709
Ceramics are usually applied over a thin, metallic bond coating. Glass frits, ground to a powder for easier melting, based on borosilicate glasses and other ceramics, can be applied. Polymer coatings, including polyethylene, polypropylene, nylon, polyethylene-tetrafluoroethylene copolymers, as well as thermoplastic acrylics and thermoset epoxies, can be thermally spray applied. These polymers are ground into a powder and fed into the combustion zone by a carrier gas where, upon heating, they melt and as droplets, are transported to a preheated substrate. The molten particles impinge on the substrate, flattening and solidifying into an interlaced network of “splats.” The thickness of the coating is governed by the number of spray passes.
RUBBER LININGS
Rubber, in the form of prefabricated sheets, have been used successfully as a lining material in a variety of immersion environments such as storage tanks, flue gas desulfurization units, chemical scrubbers, pipelines, valves, and pumps. Because of superior impact, abrasion, and chemical resistance, rubber linings in the range of 3–6 mm (1/8–1/4 in.) can often outperform conventional liquidapplied linings. A steel or other metal surface is thoroughly cleaned, usually by solvent cleaning and abrasive blast cleaning. An adhesive system, designed to adhere the rubber to the metal, is then applied. The adhesive may be a single or multicoat system. The rubber sheet is then cut to fit and applied to the surface using rollers and stitchers to press out air pockets and to press the sheet into the adhesive. The applied rubber must then be cured or vulcanized by steam in an autoclave, by heating with exhaust steam, or by a chemical curing agent applied to the rubber surface. The seams where adjacent rubber sheets join are of particular concern, as most rubber lining failures occur at the seams. Rubber sheets used for corrosion protection usually comprise two or more different types of rubbers laminated together to form the finalized sheet. Natural rubbers derived from latex-producing plants (principally in Malaysia, Indonesia, and Thailand) account for nearly 90 % of the global production of raw, natural rubber. Synthetic rubber is made principally from petroleum feedstock. Principal synthetic rubber compounds used for linings are neoprene or chloroprene rubber (CR); butyl rubber (IIR—a blend of isoprene and isobutylene copolymers); chlorobutyl rubber (CIIR—manufactured by the chlorination of IIR); ethylene-propylene rubber (EPDM—a copolymer of ethylene and propylene with the addition of a diene monomer); nitrile rubber (NBR—a copolymerization of acronitrile and butadiene); styrene-butadiene (SBR— formed by copolymerization of styrene and butadiene); and chlorosulfonated polyethylene (CSM—marketed as Hypalon by DuPont). Each of the rubbers, including the various types of natural rubbers, has advantages and disadvantages, along with specific chemical and temperature resistances, and hardness and flexibility. Rubber lining systems are chosen for their resistance to the intended service, and are applied, cured, and inspected for defects prior to placing the rubber-lined structure in service.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
710
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
TABLE 5—Advantages and limitations of principal coating resins. (Courtesy of ASM International.) Resin type
Advantages
Limitations
Comments
Autooxidative cross-linked resins Alkyds
Good resistance to atmospheric weathering and moderate chemical fumes; not resistant to chemical splash and spillage. Long oil alkyds have good penetration but are slow drying: short oil alkyds are fast drying. Temperature resistant to 105°C (225°F)
Not chemically resistant; not suitable for application over alkaline surfaces, such as fresh concrete, or for water immersion
Long oil alkyds make excellent primers for rusted and pitted steel and wooden surfaces. Corrosion resistance is adequate for mild chemical fumes that predominate in many industrial areas. Used as interior and exterior industrial and marine finishes
Epoxy esters
Good weather resistance; chemical resistance better than alkyds and usually sufficient to resist normal atmospheric corrosive attack
Generally the least resistant epoxy resin. Not resistant to strong chemical fumes, splash, or spillage. Temperature resistance: 105°C (225°F) in dry atmospheres. Not suitable for immersion service
A high-quality oil-based coating with good compatibility with most other coating types. Easy to apply. Used widely for atmospheric resistance in chemical environments on structural steel, tank exteriors, etc.
Thermoplastic resins Acrylics
Excellent light and ultraviolet stability, gloss, and color retention. Good chemical resistance and excellent atmospheric weathering resistance. Resistant to chemical fumes and occasional mild chemical splash and spillage. Minimal chalking, little if any darkening on prolonged light exposure.
Thermoplastic and water emulsion acrylics not suitable for any immersion service or any substantial acid or alkaline chemical exposure. Most acrylic coatings are used as topcoats in atmospheric service. Acrylic emulsions have limitations described under “Water emulsion latex.”
Used predominantly where light stability, gloss, and color retention are of primary importance. With cross-linking, greater chemical resistance can be achieved. Crosslinked acrylics are the most common automotive finish. Emulsion acrylics are often used as primers on concrete block and masonry surfaces.
Water emulsion latex
Resistant of water, mild chemical fumes, and weathering. Good alkali resistance. Latexes are compatible with most generic coating types, either as an undercoat or topcoat.
Must be stored above freezing. Does not penetrate chalky surfaces. Exterior weather and chemical resistance not as good as solvent or oil-based coatings. Not suitable for immersion service
Ease of application and cleanup. No toxic solvents. Good concrete and masonry sealers, because breathing film allows passage of water vapor. Used as interior and exterior coatings
Asphalt pitch
Good water resistance and ultraviolet stability. Will not crack or degrade in sunlight. Nontoxic and suitable for exposure to food products. Resistant to mineral salts and alkalies to 30 % concentration
Black color only. Poor resistance to hydrocarbon solvents, oils, fats, and some organic solvents. Does not have the moisture resistance of coal tars. Can embrittle after prolonged exposure to dry environments or temperatures above 150°C (300°F), and can soften and flow at temperatures as low as 40°C (100°F)
Often used as relatively inexpensive coating in atmospheric service, where coal tars cannot be used. Relatively inexpensive. Most common use is as a pavement sealer or roof coating.
Coal tar pitch
Excellent water resistance (greater than all other types of coatings); good resistance to acids, alkalies, and mineral, animal, and vegetable oils
Unless cross-linked with another resin, is thermoplastic and will flow at temperatures of 40°C (100°F) or less. Hardens and embrittles in cold weather. Black color only; will alligator and crack on prolonged sunlight exposure, although still protective
Used as moisture-resistant coatings in immersion and underground service. Widely used as pipeline exterior and interior coatings below grade. Pitch emulsions used as pavement sealers. Relatively inexpensive
Less flexible; requires thorough blast cleaning surface preparation
Increasing use due to high reactivity and low VOC content
Cross-linked thermosetting resins Bisphenol-F epoxies
Lower volatile organic compound (VOC) content than Bis-A epoxies. Better temperature and chemical resistance than Bis-A types. Intermediate temperature and chemical resistance between Bis-A and novolacs
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 55
Q
TYPES OF METAL CORROSION
711
TABLE 5—Advantages and limitations of principal coating resins. (Courtesy of ASM International.) (Continued) Resin type
Advantages
Limitations
Comments
Epoxy novolacs
Greatest chemical resistance and highest temperature resistance of all epoxy resins used for most severe immersion service
Least flexible of epoxy resins. Costly
Used for high-temperature chemical immersion service
Polyamide, cured epoxies
Superior to amine-cured epoxies for water resistance. Excellent adhesion, gloss, hardness impact, and abrasion resistance. More flexible and tougher than amine-curved epoxies. Some formulations can cure on wet surfaces and underwater. Temperature resistance: 105°C (225°F) dry; 65°C (150°F) wet
Cross-linking does not occur below 5°C (40°F). Maximum resistances generally require 7 day cure at 20°C (70°F). Slightly lower chemical resistance than amine-cured epoxies
Easier to apply and topcoat, more flexible, and better moisture resistance than amine-cured epoxies. Excellent adhesion over steel and concrete. A widely used industrial and marine maintenance coating. Some formulations can be applied to wet or underwater surfaces.
Polyurethanes (aromatic or aliphatic)
Aliphatic urethanes are noted for their chemically excellent gloss, color, and ultraviolet light resistance. Properties vary widely, depending on the polyol coreactant. Generally, chemical and moisture resistance are similar to those of polyamide-cured epoxies, and abrasion resistance is usually excellent.
Because of the versatility of the isocyanate reaction, wide diversity exists in specific coating properties. Exposure to the isocyanate should be minimized to avoid sensitivity that may result in an asthmatic, like breathing condition on continued exposure. Carbon dioxide is released on exposure to humidity, which may result in gassing or bubbling of the coating in humid conditions. Aromatic urethanes may darken or yellow on exposure to ultraviolet radiation.
Aliphatic urethanes are widely used as glossy light-fast topcoats on many exterior structures in corrosive environments. They are relatively expensive but extremely durable. The isocyanate can be combined with other generic materials to enhance chemical, moisture, low-temperature, and abrasion resistance.
Amine-cured epoxies
Excellent resistance to alkalies, most organic and inorganic acids, water, and aqueous salt solutions. Solvent resistance and resistance to oxidizing agents are good as long as not continually wetted. Amine adducts have slightly less chemical and moisture resistance.
Harder and less flexible than other epoxies and intolerant of moisture during application. Coating chalks on exposure to ultraviolet light. Strong solvents may lift coatings. Temperature resistance: 105°C (225°F) wet; 90°C (190°F) dry. Will not cure below 5°C (40°F); should be topcoated within 72 h to avoid intercoat determination. Maximum properties require curing time of approximately 7 days.
Good chemical and weather resistance. Best chemical resistance of epoxy family. Excellent adhesion to steel and concrete. Widely used in maintenance coatings and tank linings
Epoxy powder coatings
Good adhesion, chemical and moisture resistance. Allows cathodic protection in shielded areas on pipelines
Color change is difficult due to extensive cleanup of old powder. Like all epoxy resins, chalks on ultraviolet exposure. Must be applied in shop. Hard to field repair
Applied by electrostatic spray/ fluidized bed. Increasing use because of VOC used as pipeline, original equipment manufacturer coatings
Coal tar epoxies
Excellent resistance to saltwater and freshwater immersion. Very good acid and alkali resistance. Solvent resistance is good, although immersion in strong solvents may leach the coal tar.
Embrittles on exposure to cold or ultraviolet light. Cold weather abrasion resistance is poor. Should be topcoated within 48 h to avoid intercoat adhesion problems. Will not cure below 10°C (50°F). Black or dark colors only. Temperature resistance: 105°C (225°F) dry; 65°C (150°F) wet
Good water resistance. Thicknesses to 0.25 mm (10 mils) per coat. Can be applied to bare steel or concrete without a primer. Low cost per unit coverage
Epoxy phenolics
Excellent hot water resistance and acid resistance. More flexible than phenolic. Better alkali resistance than phenolics
Not as alkali resistant as other apoxy resins. Applies in thin films: somewhat brittle
Linings for food and beverage processing and storage tanks
(Continued) Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
712
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
TABLE 5—Advantages and limitations of principal coating resins. (Courtesy of ASM International.) (Continued) Resin type
Advantages
Limitations
Comments
Phenolics
Greatest solvent resistance of all organic coatings described. Excellent resistance to aliphatic and aromatic hydrocarbons, alcohols, esters, ethers, ketones, and chlorinated solvents. Wet temperature resistance to 95°C (200°F). Odorless, tasteless, and nontoxic; suitable for food use
Must be baked at a metal temperature ranging from 175 to 230°C (350 to 450°F). Coating must be applied in a thin film (approximately 0.025 inch, or 1 mil) and partially baked between coats. Multiple thin coats are necessary to allow water from the condensation reaction to be remove. Cured coating is difficult to patch due to extreme solvent resistance. Poor resistance to alkalies and strong oxidants
A brown color results on baking, which can be used to indicate the degree of cross-linking. Widely used as tank lining for alcohol storage and fermentation and other food products. Used for hot water immersion service. Can be modified with epoxies and other resins to enhance water, chemical, and heat resistance
Polyureas
Extremely fast curing, highly elastic thick film coatings with good strength. Abrasion resistance, chemical resistance
Requires specialized spray equipment due to fast curing Not resistant to strong acids.
Concrete coatings for use in secondary containment of chemical tanks. Elastomeric roof coatings
PAINTS, COATINGS, AND LININGS
The application of a pigmented coating to a metal surface to provide corrosion protection, and as well to provide color and appearance is by far the most common means of corrosion protection of steel and most other metals. The reason for this is that coating materials are relatively inexpensive and can be inexpensively and quickly applied as a liquid by brush, roller, or spray to an irregular, complex surface. The applied liquid coating wets and conforms to the shape of a complex surface and, upon drying or curing, form a protective layer, sealing the object from a corrosive environment.
Sometimes the words “paint,” “coating,” and “linings” are used to describe a liquid blend of resins and pigment that will dry or cure into a protective layer on a metal. In general usage, however, “paints” are usually materials that are used primarily for decorative purposes, “coatings” are used for corrosion protection purposes in atmospheric environments, and “linings” are used for corrosion protection purposes in immersion or subsoil environments. As can be seen by these general categories, there is much overlap and, under many circumstances, the terms can be used interchangeably. Coating materials generally consist of a resin (or resin blend), pigments, and solvents or water. Most resins are
TABLE 6—Advantages and limitations of zinc-rich coatings. (Courtesy of ASM International.) Resin type
Advantages
Limitations
Comments
Organic zinc-rich
Galvanic protection afforded by the zinc content, with chemical and moisture resistance similar to that of the organic binder. Should be topcoated in chemical environments with a pH outside the range 5–10. More tolerant of surface preparation and topcoating than inorganic zincrich coatings
Generally have lower service performance than inorganic zinc-rich coatings, but case of application and surface preparation tolerance make them increasingly popular.
Widely used in Europe and the Far East, while inorganic zinc-rich coatings are most common in North America. Organic binder can be closely tailored to topcoats (for example, epoxy topcoats over epoxy-zinc-rich coatings) for a more compatible system. Organic zinc-rich coatings are often used to repair galvanized or inorganic zinc-rich coatings.
Inorganic zinc-rich
Provides excellent long-term protection against pitting in neutral and near-neutral atmospheric, and some immersion, services. Abrasion resistance is excellent, and dry heat resistance exceeds 370°C (700°F). Water-based inorganic silicates are available for confined spaces and compliance with regulations regarding volatile organic compounds.
Inorganic nature necessitates thorough blast-cleaning surface preparation and results in difficulty when topcoating with organic topcoats. Zinc dust is reactive outside the pH range of 5-10, and topcoating is necessary in chemical fume environments. Somewhat difficult to apply; may mudcrack at thicknesses in excess of 0.13 mm (5 mils)
Ethyl silicate zinc-rich coatings require atmospheric moisture to cure and are the most common type. Widely used as a primer on bridges, offshore structures, and steel in the building and chemicalprocessing industries. Used as a weldable preconstruction primer in the automotive and shipbuilding industries. Use eliminates pitting corrosion.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 55
organic polymers, although some, notably the silicones, silanes, and silicates (used principally for inorganic zincrich coatings) are inorganic. Pigments provide color, moisture impermeability, strengthening, and reinforcement of the dried film, and in some cases, inhibition (chromates, molybdates, borates), and sacrificial protection (zinc). Pigments are ground or mixed into the resin blend and either dissolved in the solvent mixture or emulsified in water (sometimes with a solvent coalescing aid). The principal characteristics of a coating are derived from the resin or resin blend. Proper choice of pigments must be made for the intended service, as the pigment must have the appropriate acid, alkali, moisture, temperature, and weatherability resistances for the environment of intended use. In order for any coating system to perform properly in a given environment, the metal must be cleaned of oils, grease, salts, and contaminants that may compromise its performance. Common surface preparation methods include hand and tool scraping, brushing and grinding; abrasive blast cleaning; solvent cleaning; water washing; and high pressure water jetting. Often, more than one layer of coating is used, with primers formulated for adhesion, corrosion inhibition, and sacrificial protection. Topcoats are formulated for environmental resistance, and intermediate coats (when used) provide a means for the topcoat to adhere to the primer and provide additional thickness and resistance to the coating system. Coating systems are usually chosen for their environmental resistance, and the surface preparation and specific coating design are dependent upon the desired environmental service and anticipated service life. Table 5 [17,18] describes the advantages and limitations of principal coating resins used to formulate coatings and linings. Zinc-rich primers, and when untopcoated, zinc-rich coatings, consist of an organic (commonly epoxy, or polyurethane) or inorganic (silicate) binder highly loaded (75 % solids by volume or higher) with zinc dust. The zinc in the zinc-rich is anodic to the underlying steel, and corrodes preferentially to the steel which is the cathode in the galvanic couple. Accordingly, in time, when the zinc dissipates, it can be replaced by reapplication. This is usually much safer and less expensive than steel replacement. Zincrich primers are widely used in aggressive neutral and nearneutral marine industrial and atmospheric environments as they virtually eliminate pitting corrosion. Table 6 lists the advantages and disadvantages of organic and inorganic zinc-rich coatings.
CONCLUSIONS
Metallic corrosion, its mechanisms, and forms have been summarized in the preceding paragraphs, along with some of the more common means to prevent corrosion by use of protective overlayers. A more exhaustive treatment of either subject is beyond the scope of this chapter. As knowledge of corrosion increases as a result of research into corrosion mechanisms and the means to detect and study corrosion, new methods to combat corrosion will be discovered, along with new corrosion protection materials and methods. When one considers the state of the art
Q
TYPES OF METAL CORROSION
713
of corrosion knowledge, and the methods used to deter corrosion 50 years ago, it is evident we have come a long way baby! It is almost unimaginable to think where the science will be in another 50 years. What an exciting future for workers, technicians, scientists, and technologists in the field of corrosion, and corrosion protection!
BIBLIOGRAPHY Ailor, W. H., Handbook of Corrosion Testing and Evaluation, John Wiley and Sons, Inc., New York, 1971. Baboian, R., Electrochemical Techniques for Corrosion Engineers, NACE International, Houston, TX, 1986. Baboian, R., Manual 20 Corrosion Tests and Standards: Application and Interpretation, ASTM International, West Conshohocken, PA, 1995. Beachem, C. D., Hydrogen Damage, ASM, Metals Park, OH,1977. Bertocci, U., and Mansfield, F., “Electrochemical Corrosion Testing,” ASTM STP 727, ASTM International, West Conshohocken, PA, 1979. Bockris, J.O’M. Reddy, A.K.N. and Gamboa-Aldeco, M. E., Modern Electrochemistry, Kluwer Academic/Plenum Publishers, New York, 2000. Buchheit, R. G., Mamidipally, S. B., Schmutz, P., and Guan, H., “Active Corrosion Protection in Chromate and Chromate free Conversion Coatings,” Surface Conversion for Aluminum and Ferrous Alloys, NACE International, Houston, TX, 2007, pp. 67–92. Carpenter, W.W., “Ceramic Coatings and Linings,” ASM Handbook, Vol. 5, ASM International, Metals Park, OH, 1994. “Chemical Conversion Coatings for Coating Aluminum and Aluminum Alloys,” Naval Air Systems Command, Department of the U.S. Navy, 1970. Chen, C., Froning, M., and Verink, E., ‘‘Crevice Corrosion and its Relation to Stress-Corrosion Cracking,” ASTM STP 610, ASTM International, West Conshohocken, PA, 1976. Corrosion Basics, NACE International, Austin, TX, 1984. Edwards, J., Coating and Surface Treatment Systems for Metals, Finishing Publications Ltd./ASM International, Metals Park, OH, 1997. Fontana, M. G. and Greene, N. D., Corrosion Engineering 2nd ed., McGraw Hill, New York, 1986. Francis, R., Galvanic Corrosion: A Practical Guide for Engineers, NACE International, Houston, TX, 2001. Freeman, D. B., Phosphating and Metal Pretreatment, Industrial Press Inc., New York, NY, 1986. Good Painting Practice, Vol. 1, The Society of Protective Coatings, Pittsburgh, PA, 2002; Systems and Specifications, Vol. 2, The Society of Protective Coatings, Pittsburgh, PA, 2002 Hare, C. H., Protective Coatings, Technology Publishing Compan, Pittsburgh, PA, 2004. Haynes, G. S. and Baboian, R., “Laboratory Corrosion Tests and Standards,” ASTM STP 866, ASTM International, West Conshohocken, PA, 1985. “Hot-Dip Galvanizing for Corrosion Protection. A Specifiers Guide,” American Galvanizers Association, 2001. Kayes, A. P., Robinson, M. J., and Impey, S., “Influence of Cleaning and Surface Treatment on Filiform Corrosion of Aluminum Alloys,” J. Corrosion Science and Technology, Vol 2, Paper I, 1999. Kucera, V. and Mattsson, E., Atmospheric Corrosion, John Wiley and Sons, New York, 1982. LaQue, F. L., Corrosion Handbook, John Wiley and Sons, New York, 1948. LaQue, F. L., Marine Corrosion-Causes and Prevention, John Wiley and Sons, Inc., New York, 1975. Marcus, P., Oudar, J., Corrosion Mechanisms in Theory and Practice, Marcel Dekker, Inc, New York, NY, 1995. Morgan, J., Cathodic Protection, 2nd ed., NACE International, Austin, TX, 1993. Munger, C. G., and Vincent, L. D., Corrosion Protection by Protective Coatings, 2nd ed., NACE International, Austin, TX, 1999.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
714
PAINT AND COATING TESTING MANUAL
Q
Peabody, A. W., Control of Pipeline Corrosion, 2nd ed., NACE International, Austin, TX, 2001. Rausch, W, The Phosphating of Metals, Finishing Publications Ltd and ASM International, Metals Park, OH, 1990. Schweitzer, P. A., Corrosion-Resistant Linings and Coatings, Marcel Dekker, New York, NY, 2001. Sedricks, A. J., Corrosion of Stainless Steels, Wiley Interscience, New York, 1996. Senkowski, E. B., “Corrosion Protection with Rubber Linings,” J. Protective Coatings and Linings, November 1998, pp. 17–32. Shreir, L. L., Jarman, R. A., and Burstein, G. T., Metal/ Environmental Reactions, 3rd ed., Butterworth-Heinemann, Oxford, 1994. Staehle, R. W., et al., Stress Corrosion Cracking and Hydrogen Embrittlement of Iron Base Alloys, NACE International, Houston, TX, 1977. Swaraj, P., Surface Coaitngs: Science and Technology, 2nd ed., John Wiley and Sons, New York, 1996. Szlarska-Smialowska, Z., Pitting Corrosion of Metals, NACE International, Houston, TX, 1986. “Thermal Spraying: Practice, Theory and Application,” American Welding Society, Miami, FL, 1985. Tidblad, J., Kucera, V., Mihhailov, A., and Knotkova, D., “Improvement of the ISO Classification System Based upon DoseResponse Functions Describing the Corrosivity of Outdoor Exposures” ASTM STP 1421, ASTM International, West Conshohocken, PA, 2002. Wallinder, C. G., and Leygraf, L. D., “Environmental Effects of Metals Induced by Atmospheric Corrosion” ASTM STP 1421, ASTM International, West Conshohocken, 2002. Wernick, S., Pinner, R., and Sheasby, P. G., The Surface Treatment and Finishing of Aluminum and its Alloys, Vol. 1, 5th ed., ASM, Metals Park, OH, 1987. Zhuang, X. G., “Galvanic Corrosion,” Uhlig’s Corrosion Handbook, R. W Revie, Ed., Wiley, New York, 2000.
References [1] Munger, C. G., Corrosion Protection by Protective Coatings, NACE International, Houston, TX, 1984, p. 19.
15TH EDITION
[2] “Corrosion: Fundamentals, Testing, and Protection,” ASM Handbook Volume 13A, ASM International, Metals Park, OH, 2003, p. 189. [3] Skorchelletti, V. V., The Theory of Metal Corrosion, Leningradskoe, Ordellenie, 1973, available from the U.S. Department of Commerce, National Technical Information Service, Springfield, VA, p. 2. [4] Munger, C. G., Corrosion Protection by Protective Coatings, NACE International, Houston, TX, 1984, p. 20. [5] Fontana, M. G., Corrosion Engineering, McGraw-Hill, Inc., New York, 1986, p. 5. [6] KTA-Tator, Inc, Pittsburgh, PA. [7] KTA-Tator, Inc, Pittsburgh, PA. [8] Munger, C. G., Corrosion Protection by Protective Coatings, NACE International, Houston, TX, 1984, p. 25. [9] Munger, C. G., Corrosion Protection by Protective Coatings, NACE International, Houston, TX, 1984, p. 27. [10] Munger, C. G., Corrosion Protection by Protective Coatings, NACE International, Houston, TX, 1984, p. 28. [11] Covino, B. S. and Cramer, S. D., “Introduction to Forms of Corrosion,” ASM Handbook Volume 13A, ASM International, Metals Park, OH, 2003, pp. 187–188. [12] Fontana, M. G., Corrosion Engineering, McGraw-Hill, Inc., New York, 1986, p. 39. [13] Heidersbach, R. H., “Cathodic Protection,” ASM Handbook Volume 13A, ASM International, Metals Park, OH, 2003, p. 855. [14] Heidersbach, R. H., “Cathodic Protection,” ASM Handbook Volume 13A, ASM International, Metals Park, OH, 2003, p. 857. [15] Fontana, M. G., Corrosion Engineering, McGraw-Hill, New York, 1986, p. 104. [16] GalvInfo Center; GalvInfo Note#9 rev 2-2, August 2004; www. galvinfo.com/GINotes/G_Note9.doc. [17] Tator, K. B., “Organic Coatings and Linings,” ASM Handbook Volume 13A, ASM International, Metals Park, OH, 2003, pp. 818–819. [18] Tator, K. B., “Zinc-Rich Coatings,” ASM Handbook Volume 13A, ASM International, Metals Park, OH, 2003, p. 834.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
56
MNL17-EB/Jan. 2012
Accelerated Weathering Valerie S. Sherbondy1 PREFACE
IN PREPARATION OF THIS CHAPTER, THE CONtents of the fifteenth edition were drawn upon. The current edition will review, clarify and update the topics as addressed in the previous edition. New technology and reference materials are acknowledged and included to update the reader with the advances in this industry. Paints and coatings are used both to protect substrates and to provide an aesthetically pleasing appearance. In an outdoor environment, both of these functions can be affected by weathering. The four major factors involved in weathering are solar radiation (sunlight), moisture, oxygen, and heat. Sunlight, especially in the short wavelength/high energy ultraviolet (UV) region, has been proven to lead to discoloration, loss of gloss, scaling, embrittlement, and chalking. Moisture, in the form of rain, dew and humidity can cause blistering, flaking, loss of adhesion and promote the growth of mildew and algae. Heat exposure may cause embrittlement, cracking, peeling, and checking. Oxygen in the atmosphere participates in the oxidation of the surface of the coating, which may eventually lead to oxidation of internal layers, causing embrittlement, softening, cracking, or crazing. The oxygen is often left out of the discussion since it is more constant than the other factors, but this degradation process is a contributing factor to the other three factors of weathering. These elements contribute individually as well as in combination to cause coating failures. Naturally occurring and man-made chemicals in the environment also contribute to coating degradation and could be considered a fifth element of weathering. However, the type and levels of chemicals can vary dramatically, even over short distances. Therefore, they cannot be considered as universal in influencing the degradation process as the four factors mentioned previously. Perhaps as a consequence of this, and also partly due to tradition, chemical resistance testing is usually considered to be separate from artificial weathering. Although the effects of chemicals cannot be ignored, they are discussed elsewhere in the manual. One of the most common chemical exposure tests that is often grouped with accelerated weathering continues to be salt fog or salt spray testing, which is discussed in the corrosion section of this manual. In summary, this chapter will consider only devices that incorporate an UV light source, temperature control or monitoring, and moisture exposure monitoring. Accelerated or artificial weathering involves the use of laboratory equipment (either indoors or outdoors) to simulate the degradation that occurs during actual outdoor 1
exposure [1–3]. Accelerated weathering is the term most often applied to artificial weathering because at least one of the elements of light, heat, and moisture are either longer in duration or more intense than the actual time and conditions encountered in outdoor exposure. This increased exposure causes the coating to weather or degrade more rapidly than when placed in an environment without the intensified weathering factor. However, in an attempt to accelerate the effects of natural weathering, the laboratory conditions may be overly aggressive and thus cause results that are not attained during natural weathering. In some cases, the modes of failure identified by accelerated weathering do not occur in natural conditions, which is at least partially due to the unpredictability of the exposure environment. To evaluate the applicability of the results from any accelerated test, background information about the material being exposed and actual field results of similar materials may be referenced. Internal controls, consisting of materials with known performance characteristics, provide helpful information and can be used to predict expected service life expectations. Artificial weathering devices should be designed to produce test conditions that are controllable and reproducible, so that data are reproducible on a day-to-day basis and comparable on a laboratory-to-laboratory basis. This differs from outdoor or natural weathering where there is natural variation and no control over environmental factors. Due to this variability of conditions, the results of short-term natural weathering often are not reproducible. For example, weathering factors are different for various parts of the world and for different countries and can be broken down to represent various known climates, which are further addressed in the “Natural Weathering” chapter of this book. Some locations receive more sunlight and heat, while others are cloudy and cool. There are also seasonal variations. Even for long-term testing, the data for a specimen placed outside during January would not necessarily concur with data for a duplicate specimen placed outside in July because of initial exposure effects. When considering short term exposure, even if the same location and time of year are used, the natural weathering factors change from year to year. It should be noted that the longer term outdoor weathering does settle into a pattern and exposures of greater than 3 years have shown correlation to other samples exposed over a similar time. The long term outdoor exposure monitoring has indicated that the variations within particular climates vary within a characterized range. In contrast to natural weathering, data produced
Senior Chemist, KTA-Tator, Inc., 115 Technology Dr., Pittsburgh, PA 15275.
715 Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed byASTM International This standard is for EDUCATIONAL USE ONLY. Copyright © 2012 by www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
716
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 1—The electromagnetic spectrum.
under controlled laboratory conditions can be used as comparative data (assuming the same equipment, equally aged lamps, etc.) even if testing was conducted during different times of the year or over several years. In addition to reproducibility, a great advantage of accelerated weathering is that the results are available more quickly than with natural weathering, which is a phenomenon that may take years. The time saved on testing can translate into cost savings when developing new products or when choosing a new coating system for buildings, tanks, or equipment where historic data about coating performance is not available. This acceptance of data should be used cautiously since not all accelerated data is representative of natural weathering degradation processes. Even though there are many advantages to accelerated weathering, it is important to recognize that the data is most often used for comparative purposes only. Although there are correlation tables available for many devices for a variety of materials, the correlation between artificial and natural outdoor results should be validated before using the results as a basis for business decisions. One reason an accepted overall correlation has not been developed is that the data obtained through artificial weathering has been produced by subjecting the sample to unnatural conditions, and the results may or may not duplicate those that occurred during natural weathering. If similar failure modes are experienced in an accelerated weathering device and at an outdoor test site, a direct correlation may be developed for that product. The calculated correlation would be valid only for the tested material under the stated, specific conditions. Note that the correlation is often extended to a variety of similar environments, making further estimates of service life possible, but this correlation is not a universal correlation and should not be represented as such. Although the previous information makes it sound like the only way to get reproducible data is accelerated weathering devices, it should be noted that not all results obtained by accelerated weathering devices are as reproducible as those using and producing the data would like [3]. Studies have concluded that even though the settings may be the same, the maintenance of the device—including light source, filters, and gages—affect the exposure levels and thus the results of the testing. The cleaning and disposal of light filters is often subjective and a major source of variation that changes the amount of light at critical wavelengths. The intensity of the light is also dependent on the changing of, rotation of, and power supplied to the light source. These factors may speed up or slow down the degradation process due to changes in both temperature
and UV light exposure. Since this is known to occur, testing specifications are sometimes written to include a standard material be exposed with the test specimens or to include an actual measurement of the light during the testing. This measurement is discussed later in this chapter. Updates to methods and equipment are decreasing the variability by increasing the requirements for monitoring, ease of monitoring, and required documentation. [See ASTM G151 “Standard Practice for Exposing Non-metallic Materials in Accelerated Test Devices that Use Laboratory Light Sources” for an example of care. The manufacturer of the device should provide exact information.] Even if the light source was not changing with time, each device has regions of more intense exposure. For this reason, most devices have recommended manual rotation procedures for the exposed specimens. These procedures should be performed at regularly spaced intervals to decrease the effects of increased or decreased light intensity on the specimens during the duration of the exposure. These effects would be greater on short-term tests, where the specimens would not be moved through the different exposure zones. Aside from the conditions of the device, the conditions surrounding the device also affect performance. Most devices require a surrounding area that is ventilated and able to maintain ambient conditions, so that temperature changes within the device can be maintained or obtained. For example, a relatively high ambient temperature may reduce the amount of condensation on some devices, while high amounts of ventilation may slow the temperature recovery after water spray in other devices. It should be noted that ambient conditions are not specified by all of the accelerated weathering ASTM standards, although the manufacturers of the testing equipment often recommend general guidelines. Ambient conditions could vary between laboratories, adding an unknown variable that may affect the results.
ELEMENTS OF WEATHERING
There are many component factors that contribute to the weathering of a coating. The general components are light, moisture, heat, and oxygen, which are always present in various amounts. This section concentrates on these general components and on how they are simulated and intensified.
Light
Sunlight is composed of radiation from the visible, UV, and infrared regions of the electromagnetic spectrum
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 56
(Fig. 1) [4]. The most damaging region of sunlight for polymeric materials has been determined to be the UV light region, especially the shorter wavelengths ranging from 400 to 295 nm. The portion of sunlight in the UV region is relatively small, only 5 to 7 %, due to the filtering effects of the atmosphere. The UV region has been divided into three domains: UV-A, UV-B, and UV-C. The UV-A region is 315 to 400 nm, the UV-B region is 280 to 315 nm, and the UV-C region is 100 to 280 nm. Although the high energy/ short wavelength UV-C is the most damaging region, these wavelengths are essentially filtered out by ozone in the upper levels of the atmosphere. Therefore, if light sources produced output only in this region, they would cause abnormal degradation unless the coating was to be used for aerospace applications. Radiation in both the UV-A and the UV-B regions cause degradation of coatings. The energy of the shorter wavelengths present in the UV-B region, ~91 to 102 kcal/ mol (3.8 to 4.3 J/mol), cause more severe and rapid degradation of coatings than the wavelengths in the UV-A region. In the UV-B region, the energy levels are high enough to break carbon-nitrogen, carbon-carbon, nitrogen-hydrogen, carbon-oxygen, and carbon-hydrogen bonds in the polymeric portion of the coating. In the UV-A region, the longer wavelengths do not have sufficient energy, (~71 to 91 kcal/ mol) (3.0 to 3.8 J/mol), to break certain bonds, namely carbon-hydrogen. Thus, it is important to characterize the output of the light source used in the weathering device and determine if the spectral power distribution adequately correlates to the electromagnetic spectrum of sunlight and the actual exposure of the product being evaluated. One type of degradation caused by exposure to solar radiation is classified as erosion. The breaking of chemical bonds first leads to a degradation of the coating surface layers and is manifested by chalking, fading, and loss of gloss. Once the protective outer layer of polymer is lost, pigments are exposed. Without the protective polymeric binder, the pigments can fade and erode, causing a change in color and/or appearance. Although pigments can and do fade even in intact polymer films, this example is provided to show one of the possible effects of solar radiation on the polymer which can lead to a visual performance defect. This is by no means the only mode of degradation, or even the most common. The modes of degradation are variable and are determined by the UV light resistance of the individual components of the coating formulation. Sunlight has been simulated in accelerated weathering devices by filtered and unfiltered mercury arcs, open and enclosed carbon arcs, xenon arcs, and fluorescent lamps, as well as concentrated onto the paint surfaces by reflection of collected sunlight. However, as the results produced by early accelerated testing devices were compared to the results obtained from natural weathering, the light sources were modified to attempt to achieve results more similar to the natural weathering defects. Depending on the exact coating type and the service environment, several different specifications have been developed that indicate which light source should be used. Most of these test specifications require the use of open or “sunshine” carbon arcs, xenon arcs, or fluorescent lamps. These light sources were chosen based on their ability to more closely simulate the degrading UV light range of sunlight or to rapidly produce dramatic changes in the coating. Each
Q
ACCELERATED WEATHERING
717
Fig. 2—Spectra of light produced by an enclosed carbon arc and a sunshine carbon arc. Used with permission of the Atlas Electric Devices Co.
of the light sources provides benefits as will as pitfalls to the testing process. The source chosen will depend on the current research at the time, the light sources currently available, the material being tested, the expected environmental conditions, and historical testing of competitive compounds. Since these factors will vary with each exposure and the advances in the exposure devices, a recommendation of source cannot be made in this chapter and the manufacturers of the devices and current literature should be referenced.
ENCLOSED CARBON ARC
Originally, these lamps were used to test the light fastness of textiles. They were then combined with a water spray to be an industry-wide artificial testing device. Although mercury and carbon arc sources have been in use for over 75 years, when they were used to test paint products, it was discovered that they produce light that accelerates chalking and fading more than cracking and further testing of light sources was initiated. The carbon arc source consists of neutral solid and cored carbon electrodes. The flame portion of the lamp is enclosed in a borosilicate globe. The globe creates a semisealed atmosphere that sustains the arc and filters out UV light below 275 nm. The lamps produce two large spectral peaks that center near 350 and 380 nm, but they deliver essentially no emission below 340 nm, where the light is more severe in its degradation capabilities and more damaging to polymeric materials (Fig. 2). In addition, even the light in the visible region, 400 to 800 nm, was also low compared to sunlight, meaning that any visible color change due to the change of the pigment would require a lengthy exposure. As a result, there was incentive to produce a light source that provided a better simulation of natural sunlight, especially since more durable materials were being developed. In addition, a rigorous maintenance routine was essential for consistent operation. The rods and globe need to be cleaned daily, and the globe checked for deterioration and replacement.
OPEN-FLAME CARBON ARC
Innovation of the open flame carbon arc, or sunshine carbon arc, from the basic carbon arc was a welcomed change
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
718
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 3—Spectrum of light produced by a xenon arc with Boro/Boro filters. Used with permission of the Atlas Electric Devices Co.
since the industry saw the need for a special distribution spectrum closer to that of sunlight. The lamp operates in a free flow of air instead of in a globe. The lamp is composed of copper-coated electrodes and a central core of rare earth. The lamp is surrounded by a stainless-steel filter frame that also acts as an air duct. The frame holds flat, optical, heat-resistant, borosilicate glass panels that filter portions of the lower wavelength light from the spectrum. The spectrum produced by this light source is more similar to that of sunlight in the region between 310 and 370 nm (Fig. 3). However, there was still a spectral imbalance due to large bands between 370 and 450 nm, but the region between 450 and 800 nm was closer to natural sunlight than the enclosed carbon arc lamp shown in Fig. 3. The disadvantage of this light source is the emission between 260 and 310 nm, which includes a portion of the undesirable UV-C region. Although this light source was an improvement, the industry requested the spectral distribution of the light be even closer to that of sunlight. Maintenance issues could lead to variability since the carbon rods need to be replaced daily and the filters need to be cleaned daily and replaced periodically. A good laboratory will perform these functions as a matter of the daily routine; however, fewer laboratories are offering this service. The current use of the filtered carbon arc sources to simulate the effects of outdoor weathering has been widely deprecated by the weathering science community due to the poor simulation of the spectral distribution of the terrestrial solar radiation.
XENON ARC
The adoption of the filtered xenon arc solar radiation sources for accelerated weathering devices was the next improvement in artificial weathering equipment. The xenon lamp consists of a burner tube and a light filter system. There are two types of xenon arc lamps. One type of xenon arc lamp is cooled by water circulated through the lamp housing. The cooling water also filters out a portion of the long-wavelength infrared light. The other type of xenon lamp is air-cooled. Both lamp types produce a spectrum closer to sunlight when filtered and set at the correct irradiance setting. There are several filters and combinations of filters that can be used. The three common filter combinations used for artificial weathering of paint are quartz/
borosilicate, borosilicate/borosilicate, and quartz/quartz. The first combination allows the UV region to extend down to 270 nm (Fig. 4). The borosilicate/borosilicate combination is the most common combination and has a cutoff at 280 nm, which makes the spectrum closer to that of natural sunlight, which cuts off at 295 nm. The quartz/quartz combination produces a spectrum that extends to below 250 nm, i.e., into the undesirable UV-C region, and although used to test coatings, it is the least popular of the three filter combinations. If a more precise match for the solar cutoff and full UV spectrum match is needed, an option of lower temperatures is available with a coated infrared absorbing (CIRA) quartz/soda lime filter combination. With all of these combinations there is an increased spectral output in the 450–500 nm region when compared to the spectral output of natural sunlight. Recently, an additional filter has been introduced to the market [5]. The spectra distribution of this filter used as an inner filter with a quartz or CIRA filter provides a spectral distribution of xenon light that matches the spectral distribution of sunlight more closely than any testing light source to date (Fig. 5). Since the xenon arc source decays as the lamp ages, the irradiance of the light source should be monitored and can be controlled by adjusting the lamp wattage. All modern devices have a monitoring system that adjusts the lamp power to compensate for the decay. The settings for the lamps are referenced in many of the standards and should be referenced depending on the exposure required. Common settings produce a spectrum with approximately the same cutoff wavelength (determined by the filter) as natural sunlight. The settings are chosen to be within the range of natural sunlight. Note that a variance in the wattage settings results in changes of the degradation rate of the coatings.
FLUORESCENT UV LAMPS
Another development in light sources is the fluorescent UV lamp [6,7]. These lamps were not developed to simulate the entire spectral range of natural sunlight. Rather, they simulate only the damaging UV region found in sunlight. Currently there are three types of fluorescent UV bulbs. The QFS-40 and UVB-313 produce light with a maximum output at 313 nm (Fig. 6). The UVB-313 has a higher
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 56
Q
ACCELERATED WEATHERING
719
Fig. 4—Spectrum of light produced by a xenon arc with Quartz/Boro filters. Used with permission of the Atlas Electric Devices Co.
intensity and thus a greater weathering acceleration rate than the QFS-40. Since the UVB-313 has a higher, more stable output than the QFS-40 and the devices can now be adjusted to lower the irradiance, the QFS-40 bulb is being phased out of production. Both of these lamps have outputs down to 275 nm, which is below the cutoff of natural sunlight, and can lead to anomalous results [8]. The third type of fluorescent lamp is the UVA-340 lamp that produces a spectrum very similar to that of natural sunlight (Fig. 7). The spectrum is made up of wavelengths
in the UV-A region with a small amount of the UV-B region wavelengths. The cutoff matches that of natural sunlight, but provides little exposure to light above 390 nm. Although the results produced by the UVA-340 lamp are closer to that of sunlight, the UVB lamp had been a widely used fluorescent light source because of the rate at which degradation occurred. In some cases the degradation produced by the higher output UVB-313 bulbs was not found to match the actual degradation that occurred in actual application. The bulbs continued to be used
Fig. 5—Right light filter system versus sunlight. Used with permission of the Atlas Electric Devices Co. Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
720
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 6—UVB-lamps versus summer sunlight. Used with permission of the Q-panel Co.
Fig. 7—UVA-340 lamps versus summer sunlight. Used with permission of the Q-panel Co.
since so much of the historical comparative data had been obtained using the UVB-313 bulbs. The change to the UVA-340A bulb has taken time, but these bulbs are currently the most used fluorescent bulb. Since the UV cutoff of the UVA-340 lamp is higher, the degradation process takes a longer period of time than with the UVB lamps, but often more closely matches the degradation seen in the field and is still accelerated when compared to outdoor exposure tests. The degradation of the coatings produced by the fluorescent bulbs may be significantly different from those of natural weathering due to the lack of long wavelength radiation. Beyond the peak output for these lamps, the spectrum is much weaker than the visible light spectrum. Due to the lack of output in the visible light region, coatings of different colors do not vary much in surface temperature, unlike variations produced by natural sunlight. Changes in instrument technology and uses of monitoring devices have been used to increase the irradiance of the fluorescent bulbs. The bulbs can be operated at up to 1.75× the normal level to increase the intensity of the light, without changing the spectral distribution. In this way, the UVA-340 bulbs, set at a higher irradiance, can be used to produce quicker results without introducing an unnatural wavelength as produced by the UVB-313 bulbs. The bulbs used at the higher irradiance level should be monitored to evaluate bulb aging and decay in irradiance output [8].2
spectral range. Several rare earth metals are present and when used together produce a spectrum similar to sunlight. These lamps have to be used with filters, measuring devices to stabilize the irradiance, and electronic power supplies to provide a spectral power distribution that matches solar radiation. These lamps are high efficiency and thus suited for use in large-scale chambers. Several different radiation units are used for the large-scale exposure and can be moved around the testing area to simulate different light exposure conditions, such as sunrise, sunset, and other incident angles of interest. While it has been determined that the metal halide devices can reproduce the relative energy with a defined range, these are not known to be good overall simulators of sunlight. Additionally, the current lamps vary in output from one to another, which means that each lamp must be measured to determine the suitability to the exposure project. The temperature of the environment surrounding the lamps must also be monitored since the spectral distribution will vary with temperature changes.
METAL HALIDE
New technology and testing requirements are pushing manufacturers to perform additional exposure tests on all of the products offered to the consumer. One of the other advances in artificial weathering is the use of metal halide lamps. These do not have a continuous spectrum, but a large number of individual spectrum lines over a wide ASTM has found suitable devices available and used by DSET Laboratories, Inc., Box 1850 Black Canyon Stage 1, Phoenix, AZ 85029 and at Sub-Tropical Testing Service, 8290 S. W. 120th St., P.O. Box 560876, Miami, FL 33156.
2
LAMP STABILITY
Once a light source has been selected, it is assumed that several tests run over a period of months or years can be used to evaluate the relative performance of the paints. However, this assumption is not always true. Essentially all new accelerated weathering devices are now sold with irradiance control. Even with this control, the output of the light source may vary depending on the care and maintenance of the device. All of the devices should be cleaned and the light sources changed as recommended to ensure the most reproducible and even spectral distribution and irradiance. The newer devices have the option of monitoring and adjusting the light output, or irradiance, for all of the different light sources. Many of the methods used for testing now recommend monitoring the irradiance. In the future, these settings may become a required part of the report for the testing results. The monitoring devices take the guesswork out of determining exactly when to clean or change a light source. The monitoring can lead to
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 56
extended service of the light source, increased reproducibility of results and better overall maintenance of the testing devices. In summary, the maintenance of the device according to sound practices and manufacturers recommendations is required for reliable operation and consistent data for comparison purposes. There are several monitoring instruments available to measure the light output including pyranometers, radiometers, spectroradiometers, and light-sensitive materials. However, each of these devices may be used to measure different characteristics of light. Two differentcolored light sources could produce the same response if a pyranometer is used to measure the light. These devices measure the amount of radiant power regardless of the spectral distribution. Even when filtered to restrict the wavelength of light being measured, it was found that the response was not sufficiently sensitive for the UV range. Radiometers have been modified with filters to select areas of the spectral distribution. These are classified as wide band, broad band, or narrow band. The wideband instruments measure the light output over a range of several hundred nanometers. Broad-band instruments function over a range of 20 to 100 nm, while narrow-band instruments measure less than 20 nm. The most commercially successful radiometer measured the total UV light using a wide-band UV filter. However, when these devices were used to measure natural sunlight in comparison to the light sources, it was found that sensitivity to shorter wavelength UV was less than its peak sensitivity to visible light, which could be affected by temperature changes. The currently used radiometer, developed for exterior monitoring, has a narrow band filter and a thermoelectrically cooled detector. This is suited for long-term use and is easily operated in the field or laboratory by relatively inexperienced personnel. For internal use and for the most accurate measurements, spectroradiometers are available. These would not withstand external use and normally require operation by skilled personnel. In contrast to actual light measurements, there are industries, other than the paint industry, that rely on the use of light-sensitive reference materials. The reference material is placed in the cabinet at the same time as the test samples and monitored, usually for color changes, to evaluate the effectiveness of the light source. These materials must be inherently unstable to achieve the desired result. This instability should be considered when choosing a reference material. The reaction of some materials may vary widely, and the sensitivity is often a result of the total environment, so all of the other factors of exposure must remain constant. One of the major problems with using this technique in the paint industry is the short life of these materials relative to the more durable paint systems. Even with all of the variability of light sources, proper care of the instrument can yield consistent results between laboratories. Most of these instruments have been run for many years with less monitoring and fewer quality-control procedures than are in use today, and the data have been acceptable over many years. This is due to the fact that most companies understand that the test results can vary, and to that end often include several competitive materials in the test protocol.
Q
ACCELERATED WEATHERING
Moisture
721
Another important characteristic of weathering is moisture. Moisture is commonly overlooked as a significant factor of paint degradation because it is a common belief that structures are only wet when it rains, when they are splashed, or when they are immersed in fluids. Actual time-of-wetness studies have shown that samples placed outside in several different locations in the United States and Canada were wet approximately 30 % of the time [9]. This averages to approximately 8 h per day. The water in a natural environment is caused by dew (high relative humidity), rain, or melting snow or ice. During contact, the water may be absorbed or pass through the coating and several types of degradation may be initiated. One example is the liquid passing through the coating and interacting or reacting with a water-soluble material resulting in the formation of an osmotic cell. Another process could be initiated if moisture passes through the coating and reacts with the substrate, for example wood; the interfacial bond between coating and substrate may be destroyed or weakened. The process that affects most coatings is the cyclic absorption and desorption of water. The theory behind cyclic moisture exposure is based on the permeation, or absorption, of the liquid into the coating, which may cause certain coatings to swell. During the drying, cycle desorption, evaporation will occur, causing the coating to shrink, resulting in cyclic stressing of the system. Inner layers of the coating material may be at different points in the cycle, relative to the top surface, adding additional stress within the coating system. The cyclic changes can lead to flaking and surface stress cracking, as well as cracking and peeling of the entire coating system. The process by which degradation takes place and how fast it will occur is influenced by the permeability of the coating and the contact time required to initiate water penetration. Additionally, the rate of water and chemical degradation is increased by increased temperature and UV light exposure. The temperature of the coating when the exposure to moisture occurs can change the way the moisture affects the coating. Rain in particular can cause thermal shock and lead to erosion of the paint surface. If the surface of the paint is warm, the rain striking the surface initiates an evaporation process, which quickly cools the surface and may lead to surface degradation. Freeze-thaw cycles and frozen rain or hail can also cause surface degradation to occur more rapidly, but are not addressed in this chapter. For testing purposes, moisture can be simulated by water spray, condensation, fog, or immersion. Depending on the device used, degradation acceleration is possible by increasing the number of wet/dry cycles or increasing the time of exposure. Since accelerated weathering devices run around the clock (and day after day for that matter), it is possible to cycle the specimen through several wet and dry periods and still meet or exceed the 8 h of average wetness found in natural weathering. Another way to accelerate the damage caused by moisture is to increase the time period of exposure to moisture. Prolonged exposure may degrade coating just as much as the stresses caused by wet/dry cycling. Although the time of wetness is a big factor in the initiation and perpetuation of the degradation process, it is important to consider the purity of the water itself. The
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
722
PAINT AND COATING TESTING MANUAL
Q
water used for testing should be clean and meet the specifications outlined in the device manufacturer’s instructions. Introduction of chemical contamination through the water source could greatly change the degradation of the exposed coating and lead to corrosion within the exposure device, thus altering the exposure even more. The contaminant may initiate the degradation prematurely, or enhance color changes. Even small amounts of a contaminant can be of great influence since heat is applied and evaporation is induced. The heat and evaporation may concentrate the contaminant before it can be diluted or washed away by another cycle of moisture exposure. Recommendations for purification and consistency in the water supply include distillation or a combination of deionization and reverse osmosis. The cleanliness of the water supply can also determine the life of the testing equipment, which often contains items susceptible to corrosion.
Temperature
The third factor of weathering is heat or temperature. Testing performed over the years has indicated that degradation of coatings occurs more rapidly at elevated temperatures and temperature variation can lead to expansion and contraction stresses in the coating. These stresses may be magnified by the expansion and contraction of the substrate itself, which can lead to cracking, peeling, checking, or loss of adhesion. Temperature can also accelerate the effects of other weathering factors such as light and moisture. In accelerated weathering, cyclic testing at only slightly elevated temperatures can produce accelerated results. The temperature chosen for testing should be within the expected temperature range of the service environment. Drastic increases in temperature are not necessary to produce noticeable effects and in fact should be avoided. Testing at excessive temperatures can either cause premature or unreasonable failures or even enhanced performance that would not be realized under actual use condition. High temperatures may cause the coating to bake or cure excessively and cause it to become brittle with decreased impact resistance, or it may become more resistant to the environment than would occur if it were only air dried under ambient conditions. To prevent these occurrences, temperatures near those of actual or expected exposure should be used. The temperature should be monitored so the data have meaning relative to other test results.
Oxygen
Changing the degree of oxygen exposure of specimens by introduction of ozone or pure oxygen is possible with a few testing devices, but this modification in technique is not commonly practiced. Oxygen exposure is usually inadvertently changed in artificial weathering devices. The condensation, fog, immersion, or water spray used to create moisture can introduce oxygen to the test environment and the surface of the panels. Even in natural weathering, oxidation of a coating surface usually occurs in the presence of moisture. Oxidation involves breaking bonds within a cured coating. Either primary or secondary bonds may be affected by oxidation. Since the oxidation process is different for different chemical types (acrylic, epoxide, vinyl, etc.) of coatings, the results of oxidation can range from embrittlement to
15TH EDITION
softening, along with crazing, cracking, or discoloration. Oxidation usually begins as a surface phenomenon that breaks down the outer polymeric binder layers. Water can then pass through the film to the inner layers and cause further breakdown (often at an accelerated pace) of binder and additives.
Other Factors
Although light, heat, moisture, and oxygen play important roles in the deterioration process, it should be recognized that there are other factors that affect coating stability. Weather resistance is dependent on the curing or drying process, the substrate being painted, and application methods. These conditions cannot be fully simulated under controlled laboratory conditions since, as with natural weathering, these conditions are seldom consistent. In actual service, most coatings experience environmental factors that often continually change, are not reproducible, or are unforeseeable at the time of application. Examples include acid rain or other transient environmental pollutants. Chemical exposure, particularly in the vicinity of chemical plants or other heavy industrial environments, can also contribute to degradation. Testing for chemical resistance would be relevant when such factors can be identified and is discussed elsewhere in the manual.
ACCELERATED WEATHERING DEVICES Carbon Arc and Xenon Arc
Carbon arc and xenon arc lamp devices are used to expose specimens to UV radiation, elevated temperature, and water spray. The test conditions for paint and coatings are outlined in ASTM Standard Practice for Filtered Open Frame Carbon Arc Exposures of Paint and Related Materials (D822). The basic principles and operating procedures for xenon arc devices are found in ASTM Standard Practice for Operating Xenon-Arc Light Apparatus for Exposure of Nonmetallic Materials (G155). Additional information for these devices can be found in Standard Practice for Operating Open Flame Carbon Arc Light Apparatus for Exposure of Nonmetallic Materials (G152) and ASTM Standard Practice for Operating Enclosed Carbon Arc Light Apparatus for Exposure of Non-metallic Materials (G153). There are several devices manufactured by several companies that use either carbon arc or xenon arc light sources. The light source for these devices should be chosen by the chemical nature of the material to be tested since the spectra of the various light sources are different and produce different weathering results. For both the carbon arc- and the xenon arc-based devices, it is very important to monitor the levels of irradiance at a selected wavelength if the data are to be used for comparison purposes. This is required because there is a progressive decrease in radiation intensity as the lamp ages. This can be overcome by progressively increasing the lamp wattage, thereby minimizing the changes in intensity, and by monitoring the radiation output as described in the corresponding ASTM procedure. The newer devices include the monitor and automatically adjust the wattage to keep the intensity constant. Both carbon arc and xenon arc procedures include four test methods: 1. Continuous exposure to light and intermittent exposure to water spray.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 56
2.
Alternate exposure to light and darkness and intermittent exposure to water spray. 3. Continuous exposure to light without water spray. 4. Alternate exposure to light and darkness without water spray. A typical cycle used for evaluating coatings with these devices is 102 min of light at 145 ± 5°F (63 ± 3°C) and 18 min of light and water spray at 60 to 63±2.5°F (15.5 to 17±1.5°C).
Fluorescent UV/Condensation
With fluorescent UV bulb devices as described in ASTM Standard Practice for Operating Fluorescent Light Apparatus for UV Exposure of Nonmetallic Materials (G154), specimens are cycled between exposure to UV light and condensation in a heated environment. The light source for the QUV3 and UV20004 is composed of eight fluorescent lamps that produce light in the UV range. The light source may be any of the UV fluorescent bulbs produced for accelerated weathering. The particular bulb used will determine the nature of and speed at which the results are acquired. The exterior of the specimen rack is exposed to room temperature, and the inside is exposed to heat and humidity produced by the lights and a heated water bath. The condensation is caused by the temperature differential that exists between the front and back of the mounted specimens. The exposure can be varied by changing the temperature, the length of the light, and/or the condensation segments of the cycle. A few models are also available with a spray option. The spray option can be used to simulate thermal shock or erosion by water. The samples are mounted in brackets, which form the cabinet wall. The panels are stationary and set at an angle so condensate can run off the test surface and be replaced by fresh condensate in a continuous manner. Vents along the bottom of the chamber permit an exchange of ambient air and water vapor to prevent oxygen depletion of the condensate. The specimens are placed approximately 50 mm from the lamps. In instruments with irradiance control, the lamps and samples do not need to be rotated and the lamps are changed only when irradiance drops below the set point. In instruments without the irradiance control, both the lamps and the panels have to be manually rotated at specified intervals to ensure even UV exposure. Various tests can be selected. If no conditions are specified, ASTM G154 suggests 4 h of UV light at 60°C and 4 h of dark condensation at 50°C. Test temperatures of 50, 60, and 70°C are widely used.
Fluorescent UV-Salt Fog
The Mebon Prohesion Cabinet [10], a cyclic corrosion chamber, was originally developed as an alternative to the standard salt fog cabinet, ASTM Standard Practice for ASTM has found suitable devices available from Atlas Electric Devices Co., 4114 Ravenswood Ave., Chicago, IL 60613 and from Quartzlampen GmbH, 6450 Hanau/Main, Germany (domestic distributor is Batson Machinery, Inc., P.O. Box 3978, Greenville, SC 28608).
3
4 ASTM has found suitable devices available from Q-Panel Co., 26200 First St., Cleveland, OH 44145 and from Atlas Electric Devices Co.
Q
ACCELERATED WEATHERING
723
Operating Salt Spray (Fog) Apparatus (Bl17) for conducting corrosion resistance studies. Further studies revealed that when used in conjunction with a device providing a light source, there was use as an accelerated weathering device. The procedure is outlined in ASTM Standard Practice for Cyclic Salt Fog/UV Exposure of Painted Metal, (Alternating Exposures in a Fog/Dry Cabinet and a UV/Condensation Cabinet) (D5894). The cyclic corrosion chamber introduces a spray or fog by means of an external reservoir and a peristaltic pump operating at a flow rate of from 0.5 to 1.5 ml/h. Although a variety of solutions can be used, one consisting of 0.35 % ammonium sulfate and 0.05 % sodium chloride (Harrison’s solution) is recommended for corrosion studies. A series of experiments with this device indicated that, at least for corrosion rate studies, this solution provided more realistic results than the warm, 5 % sodium chloride solution used in salt fog cabinets2 [11,12]. The conclusion that these test results were more realistic was based on an analysis of the corrosion products. This analysis revealed that the amount of sulfur and chloride salts on panels exposed to the particular salt solution was similar to the amount and type found on the panels that had been corroded at an outdoor location. Due to the nature of the instrument, almost any exposure solution can be used to customize the testing for the location. The fog, introduced at ambient temperature, is eliminated by forcing air through the device at 23 to 55°C. Not only does this assist in drying the panels, it also replenishes the oxygen. The wet-dry cycles can be varied from 1 to 10 h. Since the cabinet is not outfitted with a light source, it is necessary to manually move the panels from the cyclic corrosion chamber to a fluorescent UV/condensation device to incorporate the element of light into the test. The standard practice outlined in ASTM D5894 involves weekly rotation between the two instruments.
Fresnel Reflector
There are three methods for accelerating natural weather exposure: black box, heated black box, and Fresnel reflector. These are described in ASTM Practice for Conducting Black Box and Solar Concentrating Exposures of Coatings (D4141). The first two methods accelerate the weathering by increasing the temperature of the exposed surface and are discussed in the natural weathering section. The Fresnel reflector is the only method that collects and intensifies natural sunlight to accelerate weathering. The basic principles for this method are found in ASTM Practice for Performing Accelerated Outdoor Weathering of Nonmetallic Materials Using Concentrated Natural Sunlight (G90). It is performed in the desert region of Arizona using Sun-10, FRECKLE, EMMA, EMMAQUA, or other similar devices that involve the use of a mirror array [7, 13]. A more detailed description of this device is presented in the chapter that deals with natural weathering. The concentration of sunlight is achieved by collecting sunlight on ten mirrors and focusing the reflected light onto the specimen. The assembly is designed to actually follow the track of the sun as it moves through the sky. The device is equipped with a blower to regulate the surface temperature of the specimen. The maximum surface temperature that can be reached is limited to no more than 10°C above the maximum temperature normally achieved by natural
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
724
PAINT AND COATING TESTING MANUAL
Q
weathering. The test may be performed with or without water spray. The water spray is provided by an oscillating nozzle assembly, which supplies deionized water as a fine, dense mist. The water spray provides a thermal shock effect and is sprayed on the samples for set cycle times. A common cycle is 8 min of water spray per hour. Exact cycles are given in ASTMG90. Since this method uses natural sunlight, the spectrum produced follows that of natural sunlight but at a higher intensity level. The testing can be performed for specific levels of solar exposure or for specific time periods. The quantity of light is measured by a radiometer, and it is expressed as total solar radiant exposure. The preferred method is based on solar exposure since this quantification accounts for the natural seasonal variations of sunlight. This allows, for example, results obtained in January to be directly compared to those obtained in July. Newer technology has permitted better temperature control during cloud cover and night time exposures. Additionally, changing the number of mirrors used at different times of the year can control the irradiance levels throughout the exposure time and lead to results that may be correlated to results obtained in the future [13].
Ultrafast Weathering
In contrast to all of the above methods, this technique of ultrafast weathering does not wait for visual changes to occur on the surface. At the present time, this method is still being evaluated to determine if there is any correlation between the information gathered using electron spin resonance (ESR) spectroscopy to monitor radical formation and natural weathering results [2]. The theory behind ultrafast weathering is based on the assumption that the radicals, which form within the first several hours of the test, will reveal the relative stability of the coating. For this testing, the process of radical formation is induced by UV radiation greater than that of sunlight. The light is filtered to remove unwanted shorter wavelengths and also focused on the sample using a cooled mirror. The sample is placed between the poles of a magnet in the microwave resonator of the spectrometer. The radical formation is plotted as a function of light-exposure time. Each paint will produce a characteristic curve. If a correlation is found to exist, the curves of different paints are then to be compared to determine which exhibits the best UV light stability. This testing could be completed over several hours instead of days, months, or years.
15TH EDITION
BIBLIOGRAPHY Hamburg, H. R., and Morgans, W. M., Hess’s Paint Film Defects and Their Causes and Cure, 3rd ed., Chapman and Hall, London, 1979. Reich, L., and Stivala, S., Elements of Polymer Degradation, McGraw-Hill, New York, 1971. Slusser, J., Kinmonth, R., and Leber, R., Atlas Sun Spots, Vol. 18, Issue 39, 1988. Atlas Electric Devices Company, Weathering Testing Guidebook, Atlas Material Testing Solutions, 2001.
References [1] Kampf, G., Sommer, K., and Zirngiebel, E., “Studies in Accelerated Weathering. Part I. Determination of the Activation Spectrum of Photodegradation in Polymers,” Progress in Organic Coatings, Vol. 19, 1991, pp. 69–67. [2] Sommer, A., Zirngieble, E., Kahl, L., and Schonfelder, M., “Studies in Accelerated Weathering. Part II. Ultrafast Weathering—A New Method for Evaluating the Weather Resistance of Polymers,” Progress in Organic Coatings, Vol. 19, 1991, pp. 79–87. [3] Fischer, R. M., Ketola, W. D., and Morrey, W. P., “Inherent Variability in Accelerated Weathering Devices,” Progress in Organic Coatings, Vol. 19, 1991, pp. 165–179. [4] Brennan, P. J., and Fedor, C., “Sunlight, UV and Accelerated Weathering,” SPE Automotive RETEC, 1987, Technical Bulletin L-822, The Q-Panel Company, 2600 First Street, Cleveland, OH 44145. [5] Q-Panel Technical Bulletin LU-8031, “High Irradiance UV/ Condensation Testers Allow Faster Accelerated Weathering Test Results,” Q-Panel Corporation, 1994. [6] Grossman, G., “Correlation of Weathering,” Journal Coatings Technology, Vol. 49, No. 633, 1977, pp. 78–82. [7] Fischer, R., “Accelerated Test with Fluorescent UV-Condensation,” SAE Technical Paper, No. 84/1022, 1984. [8] Q-Panel Technical Bulletin LU-8160, “A Choice of Lamps for the QUV,” Q-Panel Corporation, 2006. [9] Grossman, D. M., “Know Your Enemy: The Weather,” Journal Vinyl Technology, Vol. 3, No. 1, 1981, pp. 12–19 (also available as a reprint from the Q-Panel Company). [10] Licensed by Mebon Limited, Nottinghamshire, England to Q-Panel Co., 26200 First Street, Cleveland, OH 44145. [11] Harrison, J. B., and Tickle, T. C. K., Journal of Oil and Colour Chemists Association, Vol. 45, 1962, pp. 571–575. [12] Harrison, J. B., Journal of Oil and Colour Chemists” Association, Vol. 62, 1979, pp. 18–25. [13] Atlas SunSpots, Volume 34, Issue 72, page 10–11, “New EMMA/EMMAQUA Suite Revolutionizes Outdoor Testing,” Atlas Material Testing Technology, 2004. [14] Gardner, Gary, “ASTM’s New Coating Test Method Addresses Interactive Effects of Weathering and Corrosion,” Journal of Protective Coatings & Linings, Vol. 15, No. 9, 1998, pp. 50–62. [15] Atlas SunSpots, Volume 38, Issue 81, page 14–15, “Breakthrough Filter Technology,” Atlas Material Testing Technology, 2008.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
57
MNL17-EB/Jan. 2012
Chemical Resistance Latoska N. Price1
THE ABILITY OF A COATING TO RESIST CHEMICAL deterioration or staining is an essential element in its evaluation [1]. Aesthetics play a key role in the decoration of a coated object. Certainly more purchasers of coated products are moved initially by appearance than by price. It has been said that a coating is a complex material that is often a component of a coatings system, which in turn provides added value to a final product.2 What follows is a review of established test procedures with various levels of complexity and equipment sophistication that provide standardized tools for evaluation of potential flaws such as discoloration, softening, swelling, adhesion loss, gloss reduction, and pitting of the paint finish. Some methods include visual standards such as those for blister size and density or for corrosion which offer a common ground for communication or performance properties between the manufacturer and the end user of a coating. Other tests are less definitive, and while they can give insight as to how the paint may function during service, they do not necessarily correlate precisely with real life conditions. Yet, since these tests are to function in lieu of actual field exposure to predict ultimate performance, it is crucial that accelerated or simplified test parameters reproduce both the chemical as well as the physical effects of field exposure.
STAINING
Staining tests provide a thorough method of determining the ability of a coating to resist discoloration from household chemicals, chemical reagents, and other materials common in today’s environment. The tests generally expose the coating surface to a spot of the reagent on the coating surface or by immersion of a coated test panel in the reagent for a specified period with timed checkpoints.
Staining from Household Chemicals
ASTM D1308, Test Method for Effect of Household Chemicals on Clear and Pigmented Organic Finishes, encompasses the evaluation of discoloration, change in gloss, blistering, softening, swelling, loss of adhesion, or other phenomena resulting from a variety of household chemicals such as distilled water, ethyl alcohol, or vinegar. Materials or chemicals suggested as reagents are outlined in Table 1. Other materials can also be used as specified by the customer or seller. The procedure utilizes open and watch glass covered spot tests with the reagent at ambient temperature as well as immersion tests in the reagent.
1 2
Staining in the Transportation Industry
ASTM D1308, as previously mentioned, was the supplement to ASTM D1540, Test Method for Effect of Staining Agents on Organic Finishes Used in the Transportation Industry. The latter document has been withdrawn, but is available from Internet sources. Some materials or chemicals suggested as reagents are listed in Table 2. As with ASTM D1308, discoloration, change in gloss, blistering, softening, swelling, loss of adhesion, or other phenomena are examined after testing. With some reagents, exposure to sunlight or UV radiation for a specified time is required. Elevated temperature is also used to more closely simulate surface conditions in hot, sunny climates. Gasoline resistance tests combine dripping of fuel at ten drops per minute at a 20° angle with a UV lamp trained on the surface at a 90° angle. Adequate ventilation and safe handling of the dispensing and collecting vessels are essential to safety when working with gasoline. The staining potential of solid materials requires close contact of the reagent with the surface and heat exposure before evaluation.
Staining Resistance of Furniture Finishes
Decorative finishes for wood and metal furniture can include substrates such as wood, multi-density fiber board, steel or aluminum which require more from coatings to withstand normal use. Coatings for these substrates can be waterborne or solvent borne and used for such painting techniques as marble and stone replication, the creation of visual and physical textures, painted distressed finishes, glaze effects, one of the various crackle techniques or simply to provide color. One detail these finishes all have in common is that they require a high degree of stain resistance in order to preserve the decorative surface. Staining resistance of furniture finishes is covered as part of ASTM D3023, Standard Practice for Determination of Resistance of Factory-Applied Coatings on Wood Products to Stains and Reagents. This procedure is concerned with materials such as cosmetics, alcohol (ethanol), boiling water, and coffee. Cosmetics contain a number of waxes (e.g., candelilla, carnauba, microcrystalline) and/or oils (e.g., olive, castor) that are particularly staining. Ethanol, the alcohol found in liquors, can do particular damage since ethanol can oxidize in the air to produce acetic acid while coffee, rich in phenol acids (tannins), will wreck havoc on a furniture finish if left in contact with the surface for a period of time. Think of the dreaded coffee ring
Technical Manager Decorative Coatings, BASF Corporation, 205 South James St., Newport, DE 19701. 302-996-2913,
[email protected]. Hegedus, C. R., “A Holistic Perspective of Coatings Technology,” JCT Research, Vol. 1, No. 1, January 2004.
725 Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed byASTM International This standard is for EDUCATIONAL USE ONLY. Copyright © 2012 by www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
726
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
TABLE 1—Staining from Household Chemicals Item
Contains
Found in
Alkali solution
Sodium hydroxide
Oven cleaners
Acid solution
Acetic acid
Vinegars
Soap and detergent solutions
Sodium percarbonate based bleaching systems, antimicrobial agents
Laundry detergents, soap scum removers, degreasers
Lighter fluid
Petroleum distillates
Starter for charcoal briquette fires
Fruit cut
Citric acid, tartaric acid, malic acid
Oranges, apples, grapes
Oils and fats
Simple or mixed glycerol esters of organic fatty acids
Meats, salad dressings, margarines, butter, cooking oils
Condiments
Turmeric, ascorbic acid
Mustard, ketchup, grape jelly
Beverages
Phosphoric acid, food dyes
Colas, wines, Kool-Aid
Lubricating oils and greases
Petroleum distillates
Furniture polish
left by a coffee cup, sans coaster, in direct contact with the furniture surface. Boiling water and hot coffee, prepared by various methods, are poured on a horizontal panel surface and allowed to dry, and the surface is examined for graying, spotting, softening, staining, or other film deterioration. Cosmetics are applied to the coating surface and placed in a 50°C oven overnight and examined for discoloration or film failure. Fifty percent alcohol or 100 proof vodka is trapped on the coating surface using a 25-mm (1-in) square of double acid washed quantitative filter paper to maintain a longer wet contact with the surface of the finish.
SOLVENT/FUEL RESISTANCE
ASTM D2792, Test Method for Solvent and Fuel Resistance of Traffic Paint, relates a method of evaluating the resistance of a coating to solvent and fuel that causes blistering, wrinkling, loss of adhesion, and loss of hardness. The coating is applied to tin panels and air dried for 90 h. Half the panel is immersed in the test liquid, and the vessel is covered for a period of 4 to 18 h as may be specified by the customer. The panels are then removed and examined for defects. The panels are allowed to dry for another 24 h and re-examined for film defects and softening as compared to the unimmersed portion of the control panel. If subtle differences between coatings are important such as comparative research and development efforts, then the panels can be examined more often without drying at intervals such as 1, 2, 4, 6, 24, and 48 h.
Battelle Chemical Resistance Cell
Several advantages over other immersion methods are claimed for this cell [2], which was developed in the course of research sponsored by Steel Shipping Container Institute, Inc. at Battelle Memorial Institute: t Panels may be flat or indented t Edge effects area voided t Simultaneous testing in liquid and vapor t Wider range of temperature The cell consists of a Pyrex glass tube, open at both ends, held horizontally between coated test panels (Fig. 1). A convenient glass tube size is 2 in. (5.08 cm) in diameter and 3 in. (0.762 cm) in length with the ends ground flat.
Gaskets are used between tube and panels to give a liquidtight seal. The frame has screw adjustments for tightening the assembly. A sponge rubber pad behind one panel evens the pressure. A glass ring is used to surround the dimple when an indented panel is under test. The cell is filled through a hole in the middle of the tube. In use, the cell is half filled and stoppered tightly.
Bratt Conductivity Cell for Chemical Resistance
This cell [3], was designed to use conductivity of a film during chemical resistance tests as a measure of its chemical resistance. The cell proper is a 2-oz vial from which the bottom has been removed, in effect becoming a short piece of glass tubing. The cell is formed by the base plate and an additional plate with a hole that fits over the top of the vial and rests on the shoulder.
TABLE 2—Sources of stains Reagent
Examples
Glycol-based antifreeze up to 100 %
Glycol ethers
Acid, alkali, and salt solutions
Roads salts like sodium chloride or calcium chloride
Polish abrasives, creams, and waxes
Carnauba wax
Road oils and tars
Binders for asphalt, road dust control
Gasoline
Petroleum distillates
Water
Minerals
Hydraulic fluids
Glycols or silicones
Alcohol windshield washing solutions
Isopropyl alcohol
Sunscreen
Cinnamates, benzophenone, anthranilates or paraaminobenzoic acid
Insect spray
N,N-diethyl-m toluamide (DEET)
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 57
Fig. 1—Battelle chemical resistance cell. (Courtesy of Battelle Memorial Institute.)
During a test, a potential of 15 V is applied to the cell. The metal substrate serves as the positive electrode. The external resistance is selected to produce a voltage drop of about 14 V across the cell.
Resistance
Gearhart-Ball [4] solvent resistance tests utilize free coating films. The cup test employs a free film fastened over the top of a beaker or dish. Approximately 20 ml of solvent is poured over the film, and the length of time required to puncture the film is noted. This test would be considered a crude screening test. More precise data can be derived from the Distensibility test where a tensile strength strip (see ASTM D2370) is clamped with either a tensile tester jaw or an alligator clip on the upper end, and the bottom end is clamped with an alligator clip with a 12-g weight and immersed in a clear beaker containing the solvent reagent. The beaker is immediately marked with the initial length, and the time required to elongate the strip one inch is recorded.
Solvent Rub Resistance
Although solvent resistance can be evaluated using ASTM D1308, many use or adapt a solvent rubbing technique with a gauze cloth soaked in a solvent (MEK is common) and rubbing with the thumb back and forth in 2-in. (5.08-cm) strokes. This procedure, ASTM D4752, Test Method for Measuring MEK Resistance of Ethyl Silicate (Inorganic) Zinc-Rich Primers by Solvent Rub, is imprecise because the person’s strength and the size of the thumb are variable. Nevertheless, it provides a quick relative test without having to wait for exposure results. The MEK resistance of some two-component ethyl silicate zinc-rich primers has been shown to correlate well with the cure of the primer as determined by diffuse reflectance infrared spectroscopy.
ACID RESISTANCE AND ACID ETCH RESISTANCE
Acid resistance is determined by exposing a coated panel to freshly prepared mortar as well as a hydrochloric acid solution and is fully described in ASTM D3260, Test Method for Acid and Mortar Resistance of Factory-Applied Clear Coatings on Extruded Aluminum Products. The acid resistance test is performed by first sealing the edges of a specially coated panel with a paraffin and beeswax mixture and then immersing the panel in a 10 vole % solution of 31.6 %
Q
CHEMICAL RESISTANCE
727
solution of HCl at ambient temperature for 6 h, followed by rinsing, drying, and examination for blistering, peeling, lifting, crazing, flaking, or discoloration. Acid resistance however, should not be confused with “acid etch” resistance which is a completely different chemical resistance property. The acid etching of an exterior coating results from sulfur dioxide emissions from the atmosphere. In an automotive clearcoat, for example, etching appears as a non-removable water spot. The physical damage of etch is associated with a localized loss of material resulting in visible pitting of the surface. Evaluation of acid etch resistance is typically evaluated through the use of field testing usually in Jacksonville, Florida. While there is no definitive ASTM test for acid etch resistance, several laboratory tests have been developed.3,4 Mortar resistance is performed by applying a fresh mortar patty, prepared to a specified formula, to both sides of the specially coated panel and then placing it in a high relative humidity cabinet for seven days. The mortar is then carefully removed and the panel wiped off with a damp cloth followed by examination as with the acid test.
ALKALI AND DETERGENT RESISTANCE
ASTM D1308, Test Method for Effect of Household Chemicals on Clear and Pigmented Organic Finishes, is also recommended for evaluating alkali and detergent resistance. It is a simple common sense technique that can be used for many materials or chemicals that can stain or discolor a coating. The choice of testing materials should be related to the coatings end use. As previously mentioned, ASTM D1308 describes techniques for either immersing the coated substrate in the reagent, putting a small amount of the material on the coated surface and covering with a small watch glass, or just leaving the material on the surface uncovered. The coating is then checked for staining after a period of time or at 1, 2, 4, 6, 24, and 48 h intervals. For detergent resistance of appliance finishes, a solution of a specified detergent containing a high percentage of sodium phosphate is applied and the temperature maintained at 165°F (73.8°C) for the duration of exposure, usually 250 to 500 h. The test panel is submerged at least six inches into the solution. This is a more severe test than a spot test, but it is more representative of actual service conditions. Examination is done after rinsing and blotting the panel and looking for any manifestation of coating failure. ASTM D154, Standard Guide for Testing Varnishes, and related but withdrawn ASTM D1647, Test Method for Resistance of Dried Films of Varnishes to Water and Alkali, describes an alkali resistance immersion test using coated test tubes. The use of test tubes coated by dipping into the coating prevents the reagent from creeping under the edge of a flow-on film. As many as 20 tubes are prepared for this test, allowing examination after 1 to 8, 16, and 24 h. The exposed specimens are rinsed with water and allowed to dry for 30 min before examining for whitening, blistering, or removal of the film. Holubka, J.W., Schmitz, P.J., Xu, L. “Acid Etch Resistance of Automotive Clearcoats I, Laboratory Test Method Development,” Journal of Coatings Technology, Vol. 72, No. 901, February 2000, pp. 77–82. 4 Holubka, J.W., Schmitz, P.J., Xu, L. “Acid Etch Resistance of Automotive Clearcoats II, Comparison of Degradation Chemistry in Laboratory and Field Testing,” Journal of Coating Technology, Vol. 72, No. 902, March 2000, pp. 53–61. 3
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
728
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
WATER AND MOISTURE RESISTANCE
Resistance to water/moisture is potentially affected by many variables such as water source (deionized, tap, pH, saltwater, fresh), exposure format (water vapor or liquid), immersion conditions (static or stirred, stagnate or refreshed), application schedule (cyclic or continuous), temperature (heat/ drying; freeze/thaw). A number of exposure chambers are commercially available including salt fog cabinets, humidity cabinets, and cyclic chambers, that test moisture resistance as well as other resistance criteria, especially weathering. For resistance to continuously wet conditions, a simple practical test is to partially immerse a coated sample in a glass beaker containing water. The water is maintained at 100°F (37.7°C) for an extended period of time, and panels are periodically checked for discoloration, whitening, or blistering of the film. The test results are compared to a specification or standard sample run concurrently. The procedure for this test is ASTM D870, Standard Practice for Testing Water Resistance of Coatings using Water Immersion.
SALT FOG TEST
Salt fog resistance is important for marine, automobile, and aircraft coatings and any other exterior coating exposed to salt spray by being near the ocean or exposed to salted road conditions. The severe corrosion caused by salt is well known. Salt Fog testing (ASTM B117) is popular because it is simple to run and powerfully accelerates corrosion. However, reproducibility is highly dependent on strict process control of the equipment operating parameters. The use of replicate panels plus positive and negative controls is likewise critical. Caution is required when comparing salt fog resistance among various coating/substrate combinations in order to predict service life. Certain substances are notorious for poor resistance to salt fog, yet have an excellent record of field exposure corrosion resistance. Additional factors such as sample configuration (especially the presence of moisture traps such as hem flanges or coach joints) also strongly influence corrosion. Familiarity with ASTM G1 Practice for Preparing, Cleaning, and Evaluating Corrosion Test Specimens and G16 Guide for Applying Statistics to Analysis of Corrosion Data will also be beneficial. Communication of corrosion results will be facilitated by reference to visual standards such as those for blister size and density (ASTM D714 Standard Test Method for Evaluating Degree of Blistering of Paints) and for corrosion (ASTM D610 Standard Test Method for Evaluating Degree of Rusting on Painted Steel Surfaces). The test requires a salt fog cabinet and coated panels. The coating is scored to the bare substrate with an X shape (Fig. 2). The edges are sealed with a weatherproof tape, and the panel is placed in the cabinet for a specified period of time. The metal panels are exposed to the settling fog of an atomized neutral (pH 6.5 to 7.2) sodium chloride solution consisting of five parts by weight sodium chloride and 95 parts distilled or deionized water. The sample is then periodically checked to see if the rusted exposed metal has propagated under the coating causing coating failure (Fig. 3). As with water
Fig. 2—Apparatus for alkali resistance test.
resistance, the results are compared to a standard or a specification. A typical salt spray cabinet as shown in Fig. 4 incorporates a basic chamber, an air saturator tower, a salt solution reservoir, atomizing nozzles, specimen supports, a heater, and controls for maintaining specified temperature. Such chambers are available commercially from several suppliers. The testing procedure, ASTM B117, Practice for Operating Salt Spray (Fog) Apparatus, describes this method in more detail.
Fig. 3—Q panel is scored with an “x” to expose bare substrate.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 57
Q
CHEMICAL RESISTANCE
729
Fig. 4—Diagram of salt-spray (fog) cabinet.
HUMIDITY EXPOSURE
Humidity exposure tests are widely used to test paint film integrity. Exposure to humidity can cause deleterious effects to a coating which can manifest as blisters, or delamination. In ASTM D2247 Practice for Testing Water Resistance of Coatings in 100 % Relative Humidity, the panel face is exposed to 100 % relative humidity at 38°C while the back of the panel is exposed to room temperature conditions. This allows for water to continually condense on the panel surface. ASTM D1735-08 Standard Practice for Testing Water Resistance of Coatings Using Water Fog Apparatus covers the basic principles and operating procedures for testing water resistance of coatings in an apparatus similar to that used for salt spray testing. Alternative practices for testing the water resistance of coatings include ASTM D870 Practice for Testing Water Resistance of Coatings Using Water Immersion and ASTM D4585 Practice for Testing Water Resistance of Coatings Using Controlled Condensation.
PHOTOCHEMICAL WEATHERING
Virtually all exterior field service on earth includes exposure to UV radiation and photochemically induced changes are a key factor in reproducing changes in chemical resistance during use. ASTM D4141, Conducting Black Box and Solar Concentrating Exposures of Coatings, describes several procedures for use in evaluating the degradation of
coatings as a result of outdoor exposure. Environmental factors such as acid rain or alkaline bird droppings, as well as repeated heating and cooling, also contribute to a loss of coating film integrity which can result in loss of adhesion and a decrease in gloss. For these reasons, accelerated weather testing now frequently includes cyclic exposure to elevated temperatures and moisture as well as UV radiation. ASTM G155 Operating Xenon Arc Light Apparatus for Exposure of Non-Metallic Materials describes a frequently used artificial weathering method.
Cyclic Testing
The recognition that a warm, continuously wet environment does not accurately simulate the field exposure condition for many coatings/substrate systems, has led to implementation of cyclic test methods. The objective is to shorten test length while preserving the chemical degradation that occurs in the actual field service as well as replicating damage to coating and substrate. Research has documented the transitions from wet to dry more accurately replicated forms of corrosion such as pitting observed in the field. Similarly, cycling above and below the Tg of the coating also accelerates degradation. ASTM D6675 Practice for Salt-Accelerated Outdoor Cosmetic Corrosion Testing of Organic Coatings on Automotive Sheet Steel, GM 9504P General Motors Accelerated Corrosion Test, JASO M610 Cosmetic Corrosion Test
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
730
PAINT AND COATING TESTING MANUAL
Q
Method for Automotive Parts, SAE J1563 Guideline for Laboratory Cyclic Corrosion Test Procedures for Painted Automotive Parts, and SAE J2334 Cosmetic Corrosion Lab Test describes practice and schedules for various moisture, heat and electrolyte exposure combinations for automotive products. Other cyclic methods are oriented to the specifics of other application and expected field conditions. Apparatus for cyclic test can be fully or partially automated. Equipment may or may not include exposure to radiation. In partially automated systems, samples may manually move between different test chambers such as in D5894 Practice for Cyclic Salt Fog/UV Exposure of Painted Metal, Alternating Exposures in a Fog/Dry Cabinet and a UV/ Condensation Cabinet.
15TH EDITION
References [1] Lambourne, R., Ed., “Paint and Surface Coatings: Theory and Practice,” John Wiley and Sons, New York, 1987, pp. 664–671. [2] Nowacki, L. J., “Protective Linings for Steel Shipping Containers,” Corrosion (Houston), Vol. 14, 1958, p. 100. [3] Hough, R. W., Chairman, “The Bratt Conductivity Cell for Measuring Chemical Resistance,” Official Digest, Federation of Paint and Varnish Production Clubs, ODFPA, Vol. 31, 1959, p. 1460. [4] Gearhart, W. M., and Ball, F. M., “Half-Second Cellulose Acetate-Butyrate: IV,” Official Digest, Federation of Societies for Coatings Technology, Vol. 31, 1959, p. 1460.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
58
MNL17-EB/Jan. 2012
Water-Resistance Testing of Coatings John Fletcher1 and Joseph Walker2 PREFACE
IN PREPARATION OF THIS CHAPTER, THE CONTENTS of the fourteenth edition were drawn upon. The authors acknowledge the author of the fourteenth edition, Wayne Ellis (Deceased). The current edition will review and update the topics as addressed by the previous author, introduce new technology that has been developed, and include upto-date references. Other chapters in this manual cover the influence of water on the formulation and durability of organic coatings. Recent ASTM publications about the moisture effects in building materials are listed in the Bibliography at the end of this chapter. These publications describe moisture problems and solutions involving coatings and their uses. Although this chapter is intended to describe only the principal testing and evaluation of water resistance, some general remarks may be helpful in understanding the test conditions.
EFFECTS ON COATINGS OF EXPOSURE TO WATER AND WATER VAPOR
The adhesion of coatings to substrates is strongly influenced by the absorption of water and by the permeability of the coating to water vapor. The mechanism of this influence proceeds as follows [1]: 1. Absorption of water molecules in the coating film. 2. Inclusion of water in the interface between film and substrate. 3. Blister formation. 4. Corrosion/erosion of the substrate. 5. Flaking or peeling of the film. Although there is no fixed relationship between water absorption and water vapor permeability, generally the higher the water absorption the more permeable the film is to water vapor. Normally the permeability measurements are made on freshly applied films. It should be noted that with progressive aging and weathering, films become more cross-linked, and in the case of water-sensitive binders, the water-soluble additives are washed out by exposure to rain and dew. Hence such films will show decreasing permeability with time. Water resistance is defined by ASTM as “measured ability to retard both penetration and wetting by water in liquid form” [2]. It is a required property for coatings in almost all cases. Water resistance generally is measured on specimens of coatings applied to nominally impermeable substrates such as metal, wood, or masonry. (Water vapor permeability, on the other hand, is evaluated on free films.) Water 1 2
resistance may be evaluated as part of other test regimes, such as exposure testing, water repellency, corrosion, chemical resistance, salt-fog resistance, and cycling tests such as humidity-cycling and light-and-water-exposure testing. Water repellency is defined as the property of some materials which makes water hardly stick to them (hydrophobic). Chapter 67 of the Gardner Sward Coating Handbook deals with water repellent coatings. Water-repellent coatings appear in some ASTM test methods for coatings on concrete and wood. D6489, Standard Test Method for Determining the Water Absorption of Hardened Concrete Treated with a Water Repellent Coating, describes a procedure for the determination of the water absorption by a core of concrete taken from a surface treated with a water repellent. The specimen is dried to a constant weight and the portions of the specimen not treated with the water repellent are sealed with an impervious sealing material. The specimens are weighed and immersed in water. The specimens are removed from the water, weighed, and percent water absorption is calculated. Water repellent coatings on wood are described in D4446 and D5401. D4446, Standard Test Method for AntiSwelling Effectiveness of Water-Repellent Formulations and Differential Swelling of Untreated Wood When Exposed to Liquid Water Environments, describes measurement of the swelling after immersion. Wood samples in the form of elongated slats that represent the timber species or product/ treatment combination to be evaluated are exposed in soak containers. The elongated slats are immersed in the waterrepellent formulation, conditioned with appropriate weighing, and then subjected to immersion in distilled water for a prescribed period. The untreated slats omit the immersion in the water-repellent formulation. The swelling resulting from immersion for the selected time period is determined by reading a dial gage calibrated in increments of 0.025 mm (0.001 in.). A water repellent efficiency of 60 % is required to pass this test. D5401, Standard Test Method for Evaluating Clear Water Repellent Coatings on Wood, evaluates the effectiveness of clear water repellent coatings on wood before or after exterior exposure. Five ponderosa pine specimens are treated with the clear water repellent under test and allowed to dry for seven days. Five untreated specimens serve as controls. The treated and untreated specimens are each weighed and then allowed to float in water for 30 min. The specimens are removed, the excess water is wiped off, and each is reweighed. The effectiveness of the water repellent coating is then calculated.
Technical Support Manager Elcometer Instruments Ltd, UK. VP Sales & Marketing Elcometer Inc. Rochester Hills. Michigan.
731 Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed byASTM International This standard is for EDUCATIONAL USE ONLY. Copyright © 2012 by www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
732
PAINT AND COATING TESTING MANUAL
Q
PERMEABILITY CUPS
ASTM E96, [3] Standard Test Methods for Water Vapor Transmission of Materials and D1653 [4], Standard Test Methods for Water Vapor Transmission of Organic Coating Films, both include wet cup methods using the Payne Cup (Fig. 1). A free film of paint is created on a surface that can be peeled off the film when it is cured, for example, silicone paper is suitable for some types of paint. This film is used to cover the aluminum cup, which contains a quantity of water. The film is tightly clamped with a sealing gasket to the lip of the cup and the cup together with the film and the water are carefully weighed. The cup is left for the period of the test so that the water evaporates and the cup and its contents are weighed and the weight loss is noted. The permeability of the film can be calculated from the weight loss, the area of the cup and the duration of the test. There is a useful paper describing problems with the wet cup method for films with moderate to high water vapor permeability and how to improve results [5]. There are instruments on the market that measure the water vapor transmission of films more easily and rapidly than the methods described in ASTM E96 and D1653. However, it appears that no side-by-side tests have been run comparing such instruments to the wet cup methods for precision and accuracy. It is probable that instrumental methods offer considerable advantages where water vapor transmission measurements must be carried out often and/or in large numbers, but where only occasional measurements are needed, Payne Cups are adequate and economic.
TRADITIONAL TEST METHODS
Spot tests and immersion tests of coatings applied to substrates traditionally have been used as “quick and dirty” techniques to compare specimens. Criteria for evaluation include softening, blistering, solvation, color change, loss of adhesion, and rusting or other deterioration of the substrate. These observations should lead to further and more comprehensive testing related to intended conditions of product use. Such testing may include immersion, exposure to controlled condensation, 100 % relative humidity, or water fog to evaluate moisture-blistering resistance. Even more intensive testing may involve washability and scrub resistance.
15TH EDITION
composition and surface preparation, specimen preparation, and the number of specimens should be agreed upon between involved parties prior to testing. Applicable methods for the preparation of test panels are given in ASTM Methods D609 [6], D1734 [7], and Practice D1730 [8]. Test Methods D823 [9] cover application techniques for the production of uniform films.
EVALUATION
ASTM D1654, Standard Test Method for Evaluation of Painted or Coated Specimens Subjected to Corrosive Environments describes the evaluation in terms of the blistering associated with corrosion, the loss of adhesion at scribe marks on the specimen, softening of the film or other film failure.
IMMERSION TESTING
ASTM Practice for Testing Water Resistance of Coatings Using Water Immersion (D870) describes basic principles and operating procedures for testing water resistance of coatings by the partial or complete immersion of coated specimens in distilled or demineralized water at ambient or elevated temperatures. Coated specimens are partly or wholly immersed in water in a container that is resistant to corrosion (Fig. 2). The exposure conditions are varied by selecting the temperature of the water and the duration of the test. Failure may be caused by a number of factors, including a deficiency in the coating itself, contamination of the substrate, or inadequate surface preparation. Any effects such as color change, blistering, loss of adhesion, softening, or embrittlement are observed and reported. These test results typically are a pass or fail determination, but the degree of failure also may be measured.
SPECIMEN PREPARATION
Careful preparation of coated specimens is essential to assure a proper and meaningful relationship to field exposure and to avoid false test results. The substrate
Fig. 1—Payne Permeability Cup. (Reprinted by permission of Elcometer Inc.)
Fig. 2—Ford Bath Immersion Test. (Reprinted by permission of Elcometer Inc.)
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 58
Q
WATER-RESISTANCE TESTING OF COATINGS
733
Fig. 3—Typical Salt-spray Cabinet. (Reprinted by permission of Elcometer Inc.)
WATER FOG TESTING
ASTM Practice for Testing Water Resistance of Coatings Using Water Fog Apparatus (D1735) covers the basic principles and operating procedures for testing water resistance of coatings in an apparatus similar to that used for salt spray testing [10]. The apparatus required (Fig. 3) consists of a fog chamber, a water reservoir, a supply of suitably conditioned compressed air, one or more atomizing nozzles, specimen supports, provision for heating the chamber, and necessary means of control. Coated specimens are placed in an enclosed chamber where a water fog surrounds them. The temperature of the chamber is usually maintained at 100°F (38°C). The exposure condition is varied by selecting the duration of the test. Failure in water fog tests may be caused by a number of factors, including a deficiency in the coating itself, contamination of the substrate, or inadequate surface preparation. Any effects such as softening (measured by pencil hardness), color change, gloss change, blisters, loss of adhesion, embrittlement, and rusting or corrosion of substrate are observed and reported.
100 % RELATIVE HUMIDITY TESTING
Practice for Testing Water Resistance of Coatings in 100 % Relative Humidity (D2247) covers the basic principles and operating procedures for testing water resistance of coatings by exposing coated specimens in an atmosphere maintained at 100 % relative humidity so that condensation forms on the specimens. The apparatus (Fig. 4) consists of a test chamber, a heated water tank, and suitable controls. Heated water vapor is generated at the bottom of the chamber, causing saturation of the air immediately above the water tank. As the saturated mixture rises, it cools below the dew point temperature, causing condensation on the specimens suspended above. Condensation is uncontrolled. (For testing at 100 % RH with controlled condensation, see ASTM D4585, described below.) Effects of exposure at 100 % relative humidity may be color change, blistering, loss of adhesion, softening, or
Fig. 4—Humidity Cabinet. (Reprinted by permission of Elcometer Inc.)
embrittlement. Any such effects are observed and reported. They may be caused by a number of factors including a deficiency in the coating itself, contamination of the substrate, or inadequate surface preparation.
CONTROLLED CONDENSATION TESTING
Practice for Testing Water Resistance of Coatings Using Controlled Condensation (D4585) differs from D2247 in that the coated specimens are mounted with the uncoated face exposed to room temperature air. The apparatus (Fig. 5) consists of a test chamber in which the specimens form the roof of a chamber that is fitted with suitable water supply and controls. Water vapor is generated by heating a pan of water at the bottom of the test chamber. The specimens form the roof or walls of the test chamber so that the reverse sides of the specimens are exposed to the cooling effects of room temperature air. The resulting heat transfer causes water
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
734
PAINT AND COATING TESTING MANUAL
Q
15TH EDITION
Fig. 5—Controlled Condensation Apparatus. (Reprinted by permission of Elcometer Inc.)
vapor to condense on the coated specimens as liquid water saturated with air. The temperature and amount of condensate forming on the specimens are controlled by the test temperature and room temperature. The specimens are inclined so that the condensate runs off the test surface by gravity and is replaced by fresh condensate in a continuous process during the condensate cycle. Exposure conditions are varied by selecting the temperature of the test, the duration of the test, and periodic drying of the specimens. Failure may be caused by a number of factors including a deficiency in the coating itself, contamination of the substrate, or inadequate surface preparation. Any effects such as color change, blistering, loss of adhesion, softening, or embrittlement are observed and reported.
CYCLE TESTING
Test Method for Finishes on Primed Metallic Substrates for Humidity-Thermal Cycle Cracking (D2246) covers an accelerated means for determining the tendency of an organic coating to fail by cracking when exposed to humidity-thermal cycling. Although this method was withdrawn in 1992 without replacement it describes evaluation alternate exposure of prepared specimens in a cabinet maintained at 100°F (38°C) and 100 % relative humidity (with continuous condensation on the specimens), then in a cold box at –10°F (–23°C), allowing only a maximum of 30 s for transfer. Specimens are rated (using a grid overlay) by counting the number of grid squares within which one or more cracks is visible. Evaluation before and after exposure may be done for checking (D660 [11]), for cracking (D661 [12]), for erosion (D662 [13]), and sometimes for water beading (D2921 [14]). ASTM test method for Humid-Dry Cycling for Coatings on Wood and Wood Products (D3459) was withdrawn without replacement in 2007 but describes a procedure for evaluation of coatings designed for use on interior wood and wood products by exposure alternately to low and high humidity at an elevated temperature. Test panels having a minimum area of 12 × 12 in. (300 × 300 mm) are exposed in 48-h cycles in exposure chambers maintained at 97 ± 2 % relative humidity and 122 ± 3.5°F (50 ± 2°C); the other chamber maintained at 50 ± 5 % relative humidity and 73.5 ± 3.5°F (23 ± 2°C). At each change of conditions during cycling, the panels are inspected under strong light for possible damage or change, which may be in the base material or in the coating. The 1998 version of D3459 is still available from ASTM.
Testing for the effect of water exposure combined with other exposures is not described in this chapter. Such testing includes light-and-water exposure apparatus using carbon-arc ultraviolet [15], Xenon-arc [16], and fluorescent UV [17]. It should be noted that UV or other weathering degradation (accelerated or natural) may impact coating water resistance characteristics. This is especially critical for substances that tend to be subject to dimensional instability (such as wood) or for coatings that are not opaque where the substrate can be affected by UV light.
ACCELERATED MECHANICAL EXPOSURE
In recent years efforts have been made to simulate real life conditions such as the effects of repeated automated washing of automotive paints. Such tests use a water, abrasive, detergent mix with rotating brushes to simulate the damage caused by worn, contaminated brushes to be found in automated car washes in commercial use. The Car Wash Simulator (Fig. 6) is an example of such a test, utilizing a rotary brush, which can be set up using the same filaments that are used on commercial car washes.
WASHABILITY TESTING
Combination abrasion and scrub testing (D2486, D4213 and D3450) can be undertaken in wet conditions using individually controlled dosing units to introduce water between the abrasive brush or sponge to test paint on test panels or small components (Fig. 7).
Fig. 6—Car Wash Simulator. (Reprinted by permission of Elcometer Inc.)
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 58
Q
WATER-RESISTANCE TESTING OF COATINGS
[5] [6]
[7] [8]
[9]
[10] Fig. 7—Abrasion Scrubbing & Washability Tester. (Reprinted by permission of Elcometer Inc.)
[11]
BIBLIOGRAPHY
[12]
Lieff, M., and Trechsel, H. R., Eds., Moisture Migration in Buildings, ASTM STP 779, 1982. Schwartz, T. A., Ed., Water in Exterior Building Walls: Problems and Solutions, ASTM STP 1107, 1991. Trechsel, H. R., and Bomberg, M., Eds., Water Vapor Transmission Through Building Walls and Systems: Mechanisms and Measurement, ASTM STP 1039, 1989.
References [1] Schmid, E. V., Exterior Durability of Organic Coatings, Redhill, Surrey, FMJ International Publications, Redhill, Surrey, UK, 1988, pp. 81–90. [2] D996-10q, “Terminology of Packaging and Distribution Environment,” Annual Book of ASTM Standards, Vol. 15.10, ASTM International, West Conshohocken, PA. [3] E96/E96M 10, “Standard Test Methods for Water Transmission,” Annual Book of ASTM Standards, Vol. 04.06, ASTM International, West Conshohocken, PA. [4] D1653-03, “Standard Test Methods for Water Vapor Transmission of Organic Coating Films,” Annual Book of ASTM
[13] [14]
[15]
[16]
[17]
735
Standards, Vol. 06.01, ASTM International, West Conshohocken, PA, 2009. Hu, Y., Topolkaraev, V., Hiltner, A., and Baer, E., “Measurement of Water Vapor Transmission Rate in High Permeability Films,” J. Appl. Polym. Sci., Vol. 81, 2001, pp. 1624–1633. D609-00, “Practice for Preparation of Cold-Rolled Steel Panels for Testing Paint, Varnish, Conversion Coatings, and Related Coating Products,” Annual Book of ASTM Standards, Vol. 06.01, ASTM International, West Conshohocken, PA, 2006. D1734-93, “Standard Practices for Making Cementitious Panels for Testing Coatings,” Annual Book of ASTM Standards, Vol. 06.02, ASTM International, West Conshohocken, PA, 2007. D1730-09, “Practices for Preparation of Aluminum and Aluminum-Alloy Surfaces for Painting,” Annual Book of ASTM Standards, Vol. 02.05, ASTM International, West Conshohocken, PA. D823-95, “Standard Practices for Producing Films of Uniform Thickness of Paint Varnish & Related Products on Test Panels,” Annual Book of ASTM Standards, Vol. 06.01, ASTM International, West Conshohocken, PA, 2007. B117-09, “Standard Practice for Operating Salt Spray (Fog) Apparatus,” Annual Book of ASTM Standards, Vol. 03.02, ASTM International, West Conshohocken, PA. D660-93, “Method for Evaluating Degree of Checking of Exterior Paints,” Annual Book of ASTM Standards, Vol. 06.01, ASTM International, West Conshohocken, PA, 2005. D661-93, “Method for Evaluating Degree of Cracking of Exterior Paints,” Annual Book of ASTM Standards, Vol. 6.01, ASTM International, West Conshohocken, PA, 2005. D662-93, “Method for Evaluating Erosion of Exterior Paints,” Annual Book of ASTM Standards, Vol. 06.01, ASTM International, West Conshohocken, PA, 2005. D2921-98, “Method for Qualitative Tests for the Presence of Water Repellents and Preservatives in Wood Products,” Annual Book of ASTM Standards, Vol. 06.02, ASTM International, West Conshohocken, PA, 2005. 5031-01, “Practice for Testing Paints, Varnishes, Laquers, and Related Products Using Enclosed Carbon-Arc Light and Water Exposure Apparatus,” Annual Book of ASTM Standards, Vol. 06.01, ASTM International, West Conshohocken, PA, 2006. G155-05, “Practice for Operating Xenon-Arc Light Apparatus for Exposure of Non-Metallic Materials,” Annual Book of ASTM Standards, Vol. 06.01, ASTM International, West Conshohocken, PA. D4587, “Practice for Conducting Tests on Paints and Related Coatings and Materials Using a Fluorescent UV-Condensation Light- and Water-Exposure Apparatus,” Annual Book of ASTM Standards, ASTM International, West Conshohocken, PA.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
Part 13: Specific Product Testing
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
59
MNL17-EB/Jan. 2012
Aerospace and Aircraft Coatings Charles R. Hegedus,1 Stephen J. Spadafora,2 Anthony T. Eng,3 David F. Pulley,4 and Donald J. Hirst5 ORGANIC COATINGS ARE PRIMARILY APPLIED TO aircraft for environmental protection and appearance. Reference [1] concludes, “The rate controlling parameter for the corrosion of aircraft alloys, excluding the mechanical damage factor, is the degradation time of the protective coating system.” This clearly indicates the importance of the coating system’s durability and its ability to control corrosion and erosion. Relative to appearance, commercial aircraft benefit from the aesthetic characteristics of the coating system, while military aircraft rely on camouflage properties to minimize enemy detection and tracking during mission operations. To meet operational requirements, aircraft coating systems traditionally consist of a primer and a topcoat. Primers inhibit corrosion of the substrate and enhance adhesion of subsequent topcoats, while topcoats are applied for appearance and to enhance overall durability of the coating system. Self-priming topcoats, which perform as both primer and topcoat in a single coating, have recently been introduced [2,3]. In addition, specialty coatings are strategically applied to perform various functions such as protection against rain erosion, chafing, immersion in fuel, and high temperature. References [4–6] provide more detail on the formulation and properties of aircraft coatings. A number of factors affect the performance of aircraft coatings, including the substrate material, the aircraft’s operational environment, and flight conditions. Aircraft structures and skins are manufactured from numerous metallic alloys and polymeric composites with a variety of pre-paint treatments, thus complicating the adhesion and corrosion inhibition characteristics of the coating system. Environmental conditions also can vary dramatically (arctic, tropical, marine, industrial, desert, etc.). Skin temperatures during flight can range from −54 to 177°C (−65 to 350°F) while ground conditions may be relatively benign or highly corrosive. Aircraft type and mission also play important roles in coating system performance. A commuter aircraft that hops from island to island in the tropics sees frequent pressurization and depressurization along with high temperature, humidity, and salt water. In contrast, a military tactical aircraft may fly far fewer hours but will experience extreme structural loads during flight conditions. These flight conditions place environmental
and mechanical stresses on the aircraft coating system. Therefore, selection of appropriate test and evaluation procedures is an essential component for determining acceptable coatings for aircraft application [7].
VISCOSITY
The viscosity of aircraft coating components, component mixtures, and raw materials is valuable to the formulator, manufacturer, and applicator for assessment of rheological characteristics. These characteristics affect paint application properties such as atomization, leveling, sagging, and brushability.
Cup Methods (Cup Viscometers)
In aircraft coating specifications and at application sites, Zahn and Ford cup viscometers are used for admixed paints because they are inexpensive, easy to use and maintain, and produce practical quantitative data. Although the stand-mounted, standard Ford cup is more accurate due to its stability, deeper capillary orifice, and larger volume, the dip-type Zahn cup is preferred since it is easiest to use and maintain. Zahn and Ford viscometers are described in ASTM Test Method for Viscosity by Dip-Type Viscosity Cups (D4212) and ASTM Test Method for Viscosity by Ford Viscosity Cups (D1200), respectively.
Brookfield and Stormer Methods
Brookfield and Stormer viscometers are rotation-type viscometers described in ASTM Test Method for Rheological Properties of Non-Newtonian Materials by Rotational (Brookfield) Viscometer (D2196) and ASTM Test Method for Consistency of Paints Measuring Kerbs Unit (KU) Viscosity Using a Stormer-Type Viscometer (D562), respectively. The Brookfield viscometer, which measures absolute viscosities in centipoises (cP) using a rotating spindle, is particularly effective in determining viscosities of non-Newtonian fluids due to its ability to measure shear stress (that is, torque of rotating spindle) at various speeds and shear rates. Since the cup viscometers described above offer relative simplicity and ease of use, the Brookfield viscometer is not used at application sites. However, it is frequently used as a research tool to characterize the viscosities of polymeric resin materials and dispersions.
Research Associate, Air Products and Chemicals, 7201 Hamilton Blvd., Allentown, PA 18195–1501. NAVAIR Fellow/Head, Materials Engineering Division, Naval Air Systems Command, Patuxent River, MD 20670. 3 Materials Engineer, Naval Surface Warfare Center, Carderock Division, Philadelphia, PA 19112. 4 Chemical Engineer (Retired), Formally with Materials Engineering Division, Naval Air Systems Command, Patuxent River, MD 20670. 5 Materials Engineering Technician (Retired), Formally with Materials Engineering Division, Naval Air Systems Command, Patuxent River, MD 20670. 1 2
739 Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed byASTM International This standard is for EDUCATIONAL USE ONLY. Copyright © 2012 by www.astm.org University of Virginia pursuant to License Agreement. No further reproductions authorized.
740
PAINT AND COATING TESTING MANUAL
Q
The Stormer or Krebs-Stormer viscometer uses a rotating paddle to measure relative viscosity expressed in Krebs units. It is designed to provide controlled, uniform, and relative data based on a paddle-stirring motion. This type of viscometer is rarely used in the laboratory or required in current specifications due to the ease of use and maintenance of cup methods and to the accuracy of the Brookfield method.
DENSITY
Wet coating density measurements provide a check on the theoretical density value and on the uniformity of the manufactured product. Determination of density by any convenient or suitable method in this Manual is acceptable; however, a weight-per-gallon cup is normally used due to the ease of use of this instrument. Typically ASTM D1475, Density of Liquid Coatings, Inks and Related Materials is followed.
FINENESS OF GRIND AND COARSE PARTICLES
Fineness of grind and presence of coarse particles are determined to assess the quality and uniformity of the pigment dispersion and coating finish. In order to produce a high gloss coating of good appearance, a paint should be free of coarse particles. However, the extremely low gloss requirements of some aircraft camouflage paints require relatively large particle sizes. Fineness of grind is determined by ASTM Test Method for Fineness of Dispersion of PigmentVehicle Systems by Hegman-Type Gage (D1210), commonly referred to as the Hegman scale. The coarse particle content is determined by the weight retained on a 325-mesh sieve as specified in ASTM Test Methods for Coarse Particles in Pigments (D185).
SOLID AND VOLATILE CONCENTRATION/ CONTENT General
Several analysis techniques are used to determine the total solids, pigment concentration, and volatile concentration of aircraft coatings. This information can be used as a check on the coating composition when compared to the theoretical value as determined from the formulation. It can also be used to determine the quality of an as-received product and its potential surface coverage per volume of paint. In addition, restrictions on the volatile organic compounds (VOC) content of coatings increase the importance of determining the volatile concentration, and methods to determine this value are continuously being developed and refined. The following methods are currently used to determine these compositional properties for aircraft coatings.
Total Solids Content
The total solids content of a coating, often referred to as its nonvolatile content, is a measure of the combined polymer and pigment content in a paint. It is typically represented as the weight fraction or percentage of these “solid” components relative to the as-received “wet” coating. For aircraft coatings, this is determined by subtracting the volatile fraction of the coating from the total to determine the nonvolatile content. The method specified in ASTM Test Method for Volatile Content of Coatings (D2369) is used. A method that provides a volumetric assessment is ASTM Test Method for Volume Nonvolatile Matter in Clear or Pigmented Coatings (D2697).
15TH EDITION
Pigment Concentration
Three methods are available to determine the pigment weight concentration within a coating: ASTM Test Method for Determination of the Pigment Content of Solvent-Reducible Paints by High Speed Centrifuging (D2698); ASTM Practice for Separation of Vehicle from Solvent-Reducible Paints (D2372); and Federal Standard 141 Method 4021, Pigment Content (Ordinary Centrifuge). These methods use the fact that pigment particles, generally being more dense than the vehicle, will settle under centrifugal force. One distinction between these methods is the variation in rinsing solvent(s) used to separate the polymer from the pigment. Although each method utilizes a different solvent blend, it is common to deviate from the specified method by using solvents which are appropriate for the specific paint under analysis. All three methods result in a quantitative determination of pigment weight concentration in the paint; however, the latter two methods lend themselves to chemical analysis of the pigment sample after this determination.
Volatile Concentration
ASTM D2369 is a standard experimental method to determine the total volatile content of a coating. In contrast, ASTM Practice for Determining Volatile Organic Compound (VOC) Content of Paints and Related Coatings (D3960), offers a method of calculating the VOC using the nonvolatile content, the water content (if any), and the density of the coating. These latter values are predetermined using referenced ASTM methods, and they are subsequently used in the calculations to determine VOC of the paint. It should be noted that in solvent-borne coatings the VOC is the volatile content; however, water-borne coatings obviously have a nonorganic volatile component which must be taken into account when performing these calculations. The VOC is typically recorded in units of grams of organic volatiles per liter of paint with pounds per gallon used as an alternate.
Chemical Analysis
A variety of methods are used to analyze the chemical composition of aircraft coatings and their components. The specific technique that is selected is determined by the material being analyzed and the level of quantitative or qualitative data required. Some of the more common techniques to analyze polymer and solvent systems are gas chromatography, infrared and ultraviolet spectroscopy, and nuclear magnetic resonance (NMR). Atomic absorption and x-ray spectroscopy are common techniques for determining the chemistry of inorganic pigments. Table 1 provides some of the common ASTM methods used to analyze aircraft coatings.
TABLE 1—Analytical methods for aircraft paint components Compound
ASTM Method
Ketones
D2804
Isocyanate
D3432
Water
D4017
Lead, cadmium, and cobalt
D3335
Chromium
D3718
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 59
Q
AEROSPACE AND AIRCRAFT COATINGS
741
Because of growing concerns over potentially toxic materials, restrictions are being placed on the lead and chromium content of many aircraft paints. These restrictions are in direct conflict with corrosion control requirements since chromate salts, such as strontium chromate, barium chromate, and zinc chromate, have been shown to be excellent corrosion inhibitors for many metals. Therefore, many specifications require analysis for these pigments to ensure either their presence or their absence. Specific methods to determine lead and chromate content are listed in Table 1.
Another technique for predicting storage stability of coatings is by evaluating the settling properties of pigments under accelerated conditions. Aircraft coating specifications rarely contain this type of evaluation; however, centrifugation is often used as a research tool to evaluate the tendency of various pigments to settle and compact in specific vehicles. The centrifugal force and the duration are selected on a case by case basis to ascertain differences between systems.
STORAGE STABILITY
The flash point of a coating is the minimum fluid temperature at which the solvent vapors are ignited by a spark or flame. It can be predicted roughly as the weighted average of the individual flash points for each of the solvents in a coating formulation. The closed-cup technique is generally preferred since container breakage during shipment and storage often leads to flammable vapors being trapped in a confined space. Naturally, these flash points tend to be lower than the corresponding open-cup values.
General
The effects of long-term storage on aircraft coating performance are a major concern. Long durations of time and extreme temperatures can have drastic effects on the chemical and physical nature of paints, causing them to have different properties than when they were originally manufactured. Since many aircraft manufacturing, rework, and maintenance activities have paint storage facilities that have only moderate environmental controls, determining the effects of these conditions is necessary. Of primary importance are chemical stability and pigment dispersion. These properties are assessed in the laboratory following long-term and accelerated storage conditions.
Long-Term Evaluation
Methods typically used for determination of stability of aircraft coatings are specified in ASTM Test Method for Package Stability of Paint (D1849), and Federal Test Method Standard (FTMS) 141 method 3022.1, Storage Stability (Filled Container). In these methods, the packaged coating is allowed to sit undisturbed at ambient conditions for an appropriate period of time. (One year is typical for aircraft coatings.) At that time, the coating is reevaluated to compare its physical and optical properties with those originally found for the as-manufactured material. Other methods may be used to evaluate specific aspects of the material at that time, such as FTMS 141 method 3011.2, Condition in Container; method 3021.1, Skinning; or method 4208, Evaluating Degree of Settling of Paint. In most cases, it is essential that (1) the paint be free from skinning, (2) the pigment has not reagglomerated or formed a compacted cake at the bottom of the container and it can be easily redispersed to form a consistent mixture, and (3) the applied coating has properties similar to when it was manufactured.
Accelerated Conditions
To speed the evaluation of a coating’s behavior under storage conditions, methods have been devised to accelerate this behavior by subjecting the coating to extreme conditions. One common example is specified in FTMS 141 method 3019.1, Storage Stability at Thermal Extremes, which subjects the coating to 49°C (120°F) or −12°C (10°F) for 168 h, depending on the type of storage suspected. Methods involving cyclic exposure to high and low temperatures have also been used. With the development of high-performance water-borne coatings for aerospace applications, one major concern is the freeze-thaw stability of these coatings. ASTM Test Method for Freeze-Thaw Resistance of WaterBorne Coatings (D2243) is used to evaluate coating consistency and performance after 17 hat –18°C (0°F).
FLASH POINT General
Pensky–Martens
The Pensky-Martens test is a closed-cup method that can be conducted on an admixed coating or on one of the separate components in the liquid (uncured) state. The material may be refrigerated to bring it to a temperature below the expected flash point. It is then placed in a closed metal cup, heated slowly while stirring, and periodically exposed to a pilot flame through a shutter mechanism. A thermometer immersed in the fluid measures the coating temperature. The flash point is the minimum temperature at which the solvent vapors ignite, yielding a large flame that propagates over the surface of the fluid. The procedure is covered in ASTM Test Methods for Flash Point by Pensky–Martens Closed Cup Tester (D93), Method A for clear coatings and Method B for pigmented coatings.
Setaflash
The Setaflash test uses an enclosed apparatus with automatic controls to determine the flash point. It is easier to operate than the Pensky-Martens tester and utilizes an electric heater for efficient heat transfer and greater accuracy. ASTM Test Methods for Flash Point of Liquids by Small Scale Closed-Cup Apparatus (D3278), Method B describes the test procedure.
Tag
The Tag tester is similar to Pensky–Martens, except that the cup is placed in a bath containing a mixture of water and ethylene glycol. Only coatings with no suspended solids (such as, fillers) can be evaluated. ASTM Test Method for Flash Point by Tag Closed Cup Tester (D56) is the applicable test method.
POT LIFE
Pot life is the length of time in which the flow properties (such as, viscosity) of catalyzed paints will not change within an acceptable range for application. Acceptable coating conditions can vary from no change up to gellation. Pot life requirements in the aerospace industry tend to be controlled by production limitations. Since a normal production work shift is 8 h, many paints have been required to
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
742
PAINT AND COATING TESTING MANUAL
Q
have a pot life spanning this period. More recently, restrictions on the volatile organic content compounds (VOC) of paints have resulted in high solids coatings that tend to inherently possess higher viscosity and shorter pot life (2–6h). Pot life of aircraft coatings is usually determined by measuring viscosity as a function of time after mixing the paint for application. Substantial increases in viscosity are an indication that the pot life is nearly expended. It should be noted that the issue of pot life can be circumvented via the use of plural-component spray equipment, which is becoming more common in the production painting of aircraft. This equipment stores the various components of a multi-component paint separately and then mixes those components at the desired ratio immediately before or at the spray gun. This equipment not only negates pot life as a factor but also minimizes the amount of catalyzed multi-component paint that is wasted due to unused material. The only drawback is that the equipment is most effectively and efficiently used when large volumes of paint (5 gal or more) are used as opposed to small quantities (1 gal or less), which are typical of touchup maintenance operations.
DRYING TIME
ASTM Test Method for Drying, Curing, or Film Formation of Organic Coatings at Room Temperature (D1640) defines drying times of paint films. As many as eight stages of the paint drying process are recognized by coatings authorities. Determination of a particular set of drying properties is important for aerospace coatings due to the strict time and processing constraints placed on production and maintenance painting facilities. No particular drying time parameters are universal throughout the aircraft industry. Each paint facility has acceptable limits for these drying properties depending on their function, schedule, and climatic conditions. For example, set-to-touch time may be more important to a small parts shop where components are handled shortly after painting. In contrast, dry-hard times may be paramount at a production facility for painting entire aircraft that must be flown shortly after painting.
Set-to-Touch
A film is set-to-touch when it clings weakly to the finger under gentle pressure but none of the film transfers to the finger. This property indicates that the painted piece can be handled gently, but excessive contact will diminish the quality of the coating. This property may be considered important to shops painting small aircraft components which must be moved from the application area.
Tack-Free
Basically, tack is the tenacity of the film to cling to foreign objects. This component of drying is not considered of major importance in production painting; however, it may be used to ascertain the overall drying characteristics of a coating.
Dry-to-Recoat
This stage of drying is considered to be the most important for painting of aircraft because it is one of the major factors controlling the production rate. Dry-to-recoat is the time at which a second coat, or specified overcoat (such as, topcoat), can be applied without developing irregularities in the coating system, such as lifting, blistering, or loss
15TH EDITION
of adhesion. If overcoated prior to the recoat time, these defects can be caused by a number of factors, one of which is the trapping of solvent in the original coat.
Dry-Hard
The most common technique is to squeeze or pinch the coated surface with the thumb and forefinger with maximum pressure. The dry-hard time is when this procedure can be performed without leaving a permanent mark on the coating surface. This drying stage is also important in production painting of aircraft since painting is usually the last stage of maintenance and the aircraft can be flown after the coating is dry-hard.
FILM THICKNESS
Aircraft coatings are developed and manufactured to be applied within acceptable thickness limits. Coatings which are applied outside of these limits will not exhibit their optimum performance properties. For example, thin coatings may not have the required tensile and shear strength, making them susceptible to cracking and chipping, especially around fastener heads and at panel seams. In contrast, coatings which are too thick may lack flexibility and can exhibit excessive internal stresses. Another important consideration is the excessive weight that coatings add to an aircraft, causing additional fuel consumption. Therefore, coating thickness is diligently controlled and measurements are performed in both the wet and dry states. Wet coating films are checked for thickness during aircraft application processes using various wet film thickness gages according to ASTM Test Method for Measurement of Wet Film Thickness of Organic Coatings (D1212). A tooth gage is commonly used in the aerospace industry since it is simple to use, inexpensive, and can be used to obtain good approximations of dry film thickness. These approximations are made by multiplying the wet film thickness by the coating volume percent solids. The thickness of cured aircraft coatings can be determined either destructively or nondestructively. Destructive methods involve the chipping or cutting of dry films from a substrate and subsequent measurement of the coating thickness with a gage. This approach is typically not taken for aircraft coatings since the coating-substrate bond must be destroyed. Nondestructive techniques include micrometer, eddy current, magnetic induction, and a magnetoresistor/ thermistor system. These are the most desirable methods of dry film thickness determination used in the aerospace industry, both in the field and in the laboratory due to their accuracy and ease of use. One problem encountered is that aircraft are constructed of a number of types of materials. Therefore, the method used to determine coating thickness may depend on the substrate material, especially if a nondestructive method is used. One method for film thickness testing is ASTM D7091, Practice for Nondestructive Measurement of Dry Film Thickness of Nonmagnetic Coatings Applied to Ferrous Metals and Nonmagnetic, Nonconductive Coatings Applied to Non-ferrous Metals.
OPTICAL PROPERTIES General
The appearance of an applied coating is determined by a number of optical properties (color, gloss, opacity, etc.).
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 59
In the commercial aerospace industry, where public opinion is a consideration, these optical properties can strongly affect the initial impression of the aircraft and are considered to be important with respect to customer recognition. The optical properties of coating systems for military aircraft affect the detectability of these aircraft and thus effect their survivability in potentially threatening scenarios. In both the commercial and military sectors, optical properties are specified with strict tolerances and are closely monitored with instrumentation as well as with the naked eye.
Color
The colors used on commercial aircraft tend to be strong and vibrant for recognition and attraction. The colors of military aircraft coatings are empirically selected in order to achieve either theater-specific (desert, forest, arctic, etc.) or multitheater (world-wide) camouflage paint scheme requirements. In either case, paint color is considered to be important. In addition, color change is often used to assess the effects of environmental exposure (weathering, immersion in fluids, etc.) on aircraft coatings. Means of characterizing aircraft coating colors range from qualitative visual assessment to quantitative measurements with instruments, the latter providing the obvious advantage of consistent, numerical results. For the latter, ASTM Standard Practice for Calculation of Color Tolerances and Color Differences from Instrumentally Measured Color Coordinates (D2244) is followed, resulting in a color description in either the XYZ or Lab color coordinates. Color differences, as qualified by a delta E value, are determined by using these color coordinates as described elsewhere in this manual to determine delta E. One additional note must be made with respect to color measurement of aircraft coatings. Colorimeters and spectrophotometers can both be used to quantitatively measure color. Colorimeters use a specified light source with defined radiation wavelengths and intensities, typically resulting in a direct output of XYZ and/or Lab color coordinates. Spectrophotometers measure light reflected from the coating surface over an entire wavelength region. (In the case of color characterization, this is the visible region of the spectrum.) Of the two, colorimeters are less expensive and easier to use. However, colorimeters do not account for metamerism. Metamerism is a phenomenon that occurs when an object appears to be different colors under different light sources. This effect is observed with some coatings for military aircraft, which use unconventional pigments to obtain camouflage properties. In order to confirm that two colors are the same, their total reflectance over the entire visible spectrum must match. These data are obtained using a spectrophotometer. In addition, spectrophotometers are frequently used for military aircraft coatings to determine their reflectance characteristics outside of the visible region, most commonly the infrared and ultraviolet regions.
Opacity (Contrast Ratio)
Opacity or hiding power is the ability of a coating to mask the underlying substrate. Opacity of aircraft coatings is usually quantified by its contrast ratio according to ASTM Test Method for Hiding Power of Paints by Reflectometry (D2805). In this method, the coating is applied over white and black substrates. (Contrast ratio or Leneta charts are
Q
AEROSPACE AND AIRCRAFT COATINGS
743
typically used.) A colorimeter is used to measure the coating’s luminous reflectance (Y) on the black and white surfaces. Contrast ratio is then determined by: Contrast ratio = Y(black)/Y(white) Contrast ratio is a function of coating dry film thickness. The method must define the coating thickness at which the contrast ratio is determined; 2.0 mils is common for aircraft topcoats. Applying a coating to a precise thickness is difficult and may require several attempts. An alternative approach is to determine the contrast ratio at several coating thicknesses and subsequently fit the data to a quadratic equation. This equation can then be used to determine the contrast ratio at the desired coating thickness.
Reflectance and Gloss
Incident light which is reflected from a surface has two components, specular and diffuse. Specular reflectance or gloss represents light that has been reflected off a surface when the angle of incidence equals the angle of viewing. Diffuse reflectance is light reflected in all directions other than specular. Specular reflectance data are usually obtained via glossmeters whereas diffuse reflectance data are obtained by the use of goniophotometers. Gloss of aircraft coatings is measured by instruments that have been standardized to either 60° or 85° incident angles according to ASTM Test Method for Specular Gloss (D523). Commercial aircraft coatings have high gloss and are usually characterized at 60°. In contrast, many military aircraft are painted with low gloss coatings for camouflage purposes. In addition to gloss analysis at 60°, military coatings are also analyzed at 85° in order to minimize glare at grazing angles. Directional reflectance represents the optical data that are obtained from analyzing reflected diffuse light. This type of data is normally obtained via goniophotometric equipment. Several methods describe procedures for these measurements, including ASTM Test Method for Color and Color-Difference Measurement by Tristimulus Colorimetry (E1347) and Practice for Goniophotometry of Objects and Materials (E167). The use of diffuse reflectance of aircraft coatings is typically limited to those for tactical military functions that require camouflage properties.
ADHESION General
For aircraft coatings to provide maximum protection against degradation, they must firmly adhere to their substrate. Because of the complex nature of adhesion, various techniques have been devised to determine the adhesive characteristics of coatings. Since a coating’s adhesion to its substrate can be significantly affected by environmental conditions, these tests are often performed after exposure to accelerated conditions such as immersion in water at elevated temperature. Common adhesion tests used to evaluate aircraft coatings are tape and scrape tests. Other techniques, such as mechanical peel, tensile, and shear tests, are used less frequently.
Tape Tests
Adhesion tape tests are described in ASTM Test Methods for Measuring Adhesion by Tape Test (D3359). These are
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
744
PAINT AND COATING TESTING MANUAL
Q
the easiest and most versatile of the adhesion tests because they can be conducted in both laboratory and field environments. Tape tests are performed by cutting through the applied coating, into the substrate. A variety of scribe patterns can be used. One common pattern (described in the “A” method of D3359), which is used to evaluate aircraft coatings, is two scribes forming an “X.” An alternative pattern (described in the “B” method of D3359), is formed by scribing eleven parallel lines through the coating, followed by a second set of eleven lines which are perpendicular to the first set. These scribe lines create a matrix of 100 squares. Adhesive tape is firmly applied over the scribe area and is subsequently removed with a quick upward motion. Coating adhesion is then characterized using standard ratings for the amount of coating removed according to the ASTM method. As mentioned above, the coated specimens can be subjected to a specific environment prior to testing adhesion characteristics. This practice is most commonly performed with the tape tests. A typical exposure for aerospace coatings is a 24 h immersion in water at 21°C (70°F). More severe exposures are becoming common as coating technology advances. One example is immersion in water at 65°C (150°F) for seven days. Immediately upon removal from the water, the coating is scribed and tested.
Scrape Adhesion Test
The scrape adhesion test is described in ASTM Test Method for Adhesion of Organic Coatings by Scrape Adhesion (D2197). Scrape adhesion is used to characterize the adhesive and shear strength properties at the primer-substrate interface. The instrument used for this test is the Gardner Labs Scrape Adhesion Test Apparatus (Model SG-1605). The specimen for this test has an area which is uncoated with the substrate exposed. The test is performed by guiding a weighted stylus at a 45° angle to the specimen along the exposed substrate into the coating system. The scrape adhesion value is recorded as the heaviest weight used without shearing the coating from the substrate. Typical scrape adhesion values for aerospace primers fall into the 3 to 5 kg range. The scrape adhesion test is also used to determine the intercoat adhesion between a topcoat and a primer. When testing this property, a specimen with a section of primer exposed is used and the stylus is guided across the primer into the topcoat. Typical scrape adhesion values for aerospace topcoats also fall into the range of 3 to 5 kg. This test can also be performed after exposure to severe environmental conditions.
Peel and Tensile Tests
Peel test methods are rarely used for aircraft coatings. Instead, a modified version of an adhesive peel test is occasionally used. In this test, two thin metal strips are bonded together using the coating as the adhesive. One end of the specimen is left unbonded, and the metal strips at this end are separated to form a “T.” They are subsequently pulled apart in a tensile machine, and the force needed to pull the two strips apart is recorded as the adhesive peel strength of the coating. The type of failure (adhesive or cohesive) is also recorded. An adhesive failure indicates that the coating strength exceeds that of the coating-substrate adhesive strength.
15TH EDITION
The most common tensile adhesion test used on aerospace coatings is a tensile pull-off test, also referred to as a button test. In this test, a flat head paten is bonded to the surface of the coating with an adhesive. A tensile force is then applied to the paten perpendicular to the coating surface until the paten is removed. The location at which the paten is removed from the surface must be carefully examined. In many cases, the adhesive which bonds the paten to the coating will fail or the coating may fail cohesively. In these cases, it can only be stated that the adhesive strength of the coating exceeds the tensile strength recorded. Only if the coating fails adhesively is the recorded tensile strength that of the coating-substrate adhesion.
FLEXIBILITY General
Flexibility is an important property for aerospace coatings, particularly at low temperatures (−51°C), which is common for aircraft cruising at high altitudes. Cracking of coatings at skin joints and around fastener heads can lead to corrosion of exposed areas. Corrosion inhibiting epoxy primers used on aircraft tend to be brittle and exhibit poor flexibility, whereas the urethane topcoats are generally more flexible, especially at low temperatures. Sealant materials and elastomeric primers occasionally are used to improve the overall flexibility of the paint system.
Mandrel Bend
The flexibility of high-performance coatings is commonly characterized by the mandrel bend test method outlined in ASTM Test Methods for Mandrel Bend Test of Attached Organic Coatings (D522). This test is normally conducted at low temperatures such as −51°C and is performed by bending a painted specimen 180° around mandrels of various diameters with the coating aligned away from the mandrel. After allowing the specimen to return to room temperature, the coating is examined for cracking along the bend. The smallest (most severe) mandrel diameter that the coating can withstand without cracking is recorded. One standard military specification requirement for this test on low gloss coatings is a 2-in. mandrel bend while more flexible coatings can withstand mandrel bends of 1/8 to 1/4 in. at low temperatures without cracking.
Impact Tests
The most universal test for measuring impact flexibility of aerospace coatings is Method 6226 (G.E. Impact) of Federal Test Method Standard 141B. The test apparatus consists of a solid steel cylinder weighing 1.69 kg (3.7 lb), which has spherical knobs protruding from the end. These knobs are designed such that the coating system is subjected to elongations of 0.5, 1, 2, 5, 10, 20, 40, and 60 %. The impact is accomplished by allowing the steel cylinder to fall freely from a height of 1.05 m (42 in.) through a hollow cylinder guide, striking the reverse uncoated side of the specimen. The imprints formed from the knobs are examined, and the impact elongation is recorded as the highest deformation without cracking of the coating. A standard requirement for this test on aircraft topcoats is 40 % elongation. Another commonly used test is the Gardner impact test specified in ASTM Test Method for Resistance of Organic Coatings to the Effects of Rapid Deformation (Impact)
Copyright by ASTM Int'l (all rights reserved); Tue Apr 22 03:46:21 EDT 2014 Downloaded/printed by This standard is for EDUCATIONAL USE ONLY. University of Virginia pursuant to License Agreement. No further reproductions authorized.
CHAPTER 59
(D2794). This test is performed in a similar fashion to the G.E. impact test described above; however, the weighted cylinder has a rounded end and is dropped from various heights. The weight of the impact cylinder and the highest height that causes no cracking or disbondment of the coating are used to calculate an impact strength, usually in inch-pounds. This test can be performed directly on the coated side of the panel or, like the GE impact test, on the reverse, uncoated side.
Tensile (Elongation) and Fatigue Tests
For aerospace coatings, free films of the coating generally 5 mils thick or greater are prepared and tested for tensile characteristics. Epoxy primers have relatively high tensile strengths (>2500 psi) but very poor elongations (
