Scanning Electron Microscopy

Scanning Electron Microscopy

David Muller 2008

Electron Energy Loss Spectrum of SiO2 Most likely energy transfer is ionization of valence electrons 10

8

Incident Beam

Intensity (arb. units)

10

7

Valence Excitations 10

6

Si L edge 10

5

O-K edge 10

4

0 David Muller 2008

100

200 300 400 Energy Loss (eV)

500

600

700

Path of the Electron Beam

BS2

David Muller 2008

SE2

SE1

Kanaya-Okayama Depth Penetration Formula

1.67 0.0276 A E R= ______________ 0.89 (Z ρ)

μm

R= Depth Penetration A= Atomic Weight (g/mole) E= Beam Energy (KV) Z= Atomic number ρ = density (g/cm )2

David Muller 2008

The Affect of Accelerating Voltage 30KV

15KV

Primary Beam 5KV 1KV

.16 μm

.01 μm (100A)

.99 μm

3.1 μm

Depth Penetration in Iron

(predictions from the KO formula) David Muller 2008

.5KV

35 A

Interaction Volume vs Accelerating Voltage

5 kV 15 kV

25 kV

Better control of where SE, BSE and x-rays are produced at lower beam voltages

David Muller 2008

Interaction Volume –Sample Composition (20 kV incident beam in all 3 cases)

Iron

Carbon David Muller 2008

Silver

Pear to apple-shaped

Secondary Electrons

final lens SE3 SE1

BSE

SE2

specimen

David Muller 2008

Electron Interactions (Between Primary Beam and Sample)

• • • • •

SE1- at point of primary interaction SE2- away from initial interaction point SE3- by BSE outside of sample BSE1- at point of primary interaction BSE2- away from initial interaction point

David Muller 2008

Lateral Distribution of SE

SE1

SE2A SE1 > 100KX SE2A- 50KX SE2B< 15KX

David Muller 2008

SE Escape Depth Total Beam Penetration Volume

SE2B

Lateral Distribution of BSE

BS1

BS2A

BS2B

David Muller 2008

BS2A Escape Depth BS2B Escape Depth

One Primary Electron In Can Create Several SEs Out at Low Accelerating Voltages

100 Angstroms

Secondary Electron Yield Coefficient David Muller 2008

SE out δ= PE in

Energy Distribution of Emitted Electrons

SE # of electrons collected

BSE

Auger 0

50 eV

2 kV electron energy

Secondary Electron Yield Coefficient David Muller 2008

SE out δ= PE in

EPE

Secondary Electron Yields Carbon-contaminated

Cleaned in-situ

As-received samples are all coated with a carbon contamination layer Overall scaling factor is from the different backscattering responses of the substrate David Muller 2008

Charge Neutralization Electron Yield δ= # SE out / # Inc e- in Sample charges +ve (increases landing energy Of incident electrons)

1.4

δ

1.2 1 0.8

Sample charges –ve (reduces landing energy Of incident electrons)

0.6 0.4 0.2 0 0 KV

EC 1

EC 2

KV

Incident Beam Voltage

David Muller 2008

Q: If sample charges, does it get brighter?

Voltage Contrast with SE (SE have low energies so are easily deflected by small voltages)

The floating end of the via chain is bright because of trapped negative charge causes secondary electrons to be repelled. The remainder of the chain is neutral, and thus darker.

(http://www.acceleratedanalysis.com/hepvc.html) David Muller 2008

High Vacuum E.T. Secondary Electron Detector

Light guide

Faraday cage (-150 - +300 V)

Phosphorous screen (Al-coated) (10 kV) glass target Scintillator Photomultiplier David Muller 2008

Secondary Electron Detectors

TLD

PMT Internal Lens

E.T. SED

Specimen David Muller 2008

Lens Modes of a Modern SEM Field-Free Operation

Immersion Lens

Large area, lower resolution

Small area, high resolution

David Muller 2008

TLD in BSE Mode

• Within-the-lens detector is part of the final lens SE3

• Bias voltage down to -150V

BSE

SE Specimen

David Muller 2008

Topography Affects Secondary Electron Emission (Angle of Incidence)

David Muller 2008

Scanning Action of the Electron Beam in a 3-D Specimen

David Muller 2008

Location of Detector Leads to Shadowing

+300 V SE-detector

B A

David Muller 2008

C

Increasing the detector bias will wash out the shadows

What Is “Reality” in the SEM ?

David Muller 2008

What Is “Reality” in the SEM ?

David Muller 2008

Previous image turned upside down. We need to know where the detector is to tell bumps from pits!

Why Edges Appear Brighter

David Muller 2008

Edge Effect at Lower Voltage

David Muller 2008

Edge Effects on a Sphere A

2nd e- Intensity

250 200 150 100

B 50 0

David Muller 2008

200 400 600 800 Distance (microns)

1000

Example of Sample charging in a Secondary Electron Image

Charging is worse On this face As more secondaries escape

David Muller 2008

A Line Profile on a Semi-conductor Line

One Micron SiO2 in Si

The E-Beam line profile of the specimen

David Muller 2008

Where do you measure “One Micron” ?

A Line Profile on a Semi-conductor Line .99 μm One Micron SiO2 in Si

1KV

A 1 KV bean has minimal beam penetration and can give an image that is closer to ‘reality’.

David Muller 2008

A Line Profile on a Semi-conductor Line .74uM One Micron SiO2 in Si

5KV

A 5KV beam penetrates deep into the specimen which gives the appearance of the peaks being closer together

David Muller 2008

Line Profiles on the Same Sample Can Change with Accelerating Voltages .92 μm

.99 μm

1KV

2KV .74 μm

.85 μm

3KV

5KV This was a 1 μm line

David Muller 2008

Secondary Electrons

final lens SE3 SE1

BSE

SE2

specimen

David Muller 2008

Angular Distribution of BSE

• Normal angle of incidence

• Greater angle of incidence

David Muller 2008

Angular Distribution of BSE

• Contrary to SE images, BSE images can have dark edges David Muller 2008

Electron Emission Coefficient Vs. Atomic Number at 20 KV

1 Total • Electron Emission Coefficient

BSE .2

SE 20

• Atomic Number David Muller 2008

80

SE Electron Emission Coefficient Vs Atomic Number at Various KV 2KV

1

5KV

• Secondary Electron Emission Coefficient .2

10KV 15KV 20KV 20 • Atomic Number

David Muller 2008

80

SE Emission Coefficient Vs. KV at Various Atomic Numbers

1 • Secondary Electron Emission Coefficient

AU AL

.2

C 5

15

• Accelerating Voltage in KV David Muller 2008

25

Z Dependence of BSE

David Muller 2008

From “Scanning Electron Microscopy and X-Ray Microanalysis”, Goldstein et al, 3rd ed. Chap 3

Tilt Dependence of BSE

David Muller 2008

From “Scanning Electron Microscopy and X-Ray Microanalysis”, Goldstein et al, 3rd ed. Chap 3

Reverse Biased S.E.D. Repulses Secondary Electrons

-150 V SE-detector

B A

David Muller 2008

C

Backscatter Electrons Ignore the Bias

-150 V SE-detector BSE B A

David Muller 2008

C

A Solid State BSD Can Image Two Ways • Elemental backscatter images are acquired by adding detectors A+B.

• Topographical backscatter electron images can be acquired by subtracting b from a (AB)

David Muller 2008

Solid State BSD

From http://www.jeol.com/sem_gde/bkscat.html

David Muller 2008

Grains in a Polished Fe-Si Alloy imaged by Different SEM methods -ve Biased E-T Noisy Backscattered Signal

Backscattered A+B “Composition” Signal

David Muller 2008

-ve Biased E-T Secondary Electron Signal

Backscattered A-B “Topographic” Signal

Electron Channeling in a Crystal

Electron wave fields within a crystal for incident electron directions close to the Bragg angle qB. The vertical lines are the position of the Bragg reflecting atomic planes. From H. Niedrig, “Electron backscattering from thin films”, Journal of Applied Physics -- April 1982 -- Volume 53, Issue 4, pp. R15-R49

David Muller 2008

Electron Channeling in a Crystal

Electron Backscatter Diffraction Pattern of Germanium. Right –automatic indexing software matches the high symmetry zone axes and spacing between them to identify the crystal type and orientation. (University of Queensland, http://www.uq.edu.au/nanoworld/xl30_anl.html) David Muller 2008

Electron Backscattering Diffraction Patterns (EBSD or EBDP) for Orientational Imaging Orientation Imaging Map (color shows grain orientation)

Boundary – Color shows angle of grain boundary

David Muller 2008

(From http://www.edax.com/technology/EBSD/OIM/intro6.html)

Sample Prep for EBSD Damage layer must be much less than the range of the electrons 3 micron diamond polish

1 micron alpha alumina

David Muller 2008

No pattern visible

Pattern Image Quality (IQ) = 25

10 minutes colloidal silica

30 minutes colloidal silica

http://www.edax.com/TSL/support/EBSD_Sample_Prep.html

Pattern IQ = 177

Pattern IQ = 224