Forme Farmaceutiche Innovative - Approfondimenti Volume 3

Journal of Colloid and Interface Science 309 (2007) 78–85 www.elsevier.com/locate/jcis

Zinc oxide colloids with controlled size, shape, and structure Mihaela Jitianu 1 , Dan V. Goia ∗ Clarkson University/Center for Advanced Materials Processing, Potsdam, NY 13699, USA Received 31 October 2006; accepted 6 December 2006 Available online 12 December 2006

Abstract Highly dispersed uniform ZnO particles of different sizes and shapes were prepared by slowly adding zinc salt and sodium hydroxide solutions in parallel into aqueous solutions of Arabic gum. Except for the very early stages, the precipitated solids consisted of a well-defined zinc oxide phase. Depending on the experimental conditions, the size of the final polycrystalline particles formed by the aggregation of nanosize entities varied from 100 to 300 nm. The reaction temperature affected both the size of the nanosize precursors and their arrangement in the final particles. At ambient temperature the primary nanoparticles, approximately 10 nm in size, formed spherical aggregates, while at 600 ◦ C they were much larger (44 nm) and combined to form rather uniform hexagonal ZnO prisms. The aspect ratio and the internal structure of the latter could be altered by changing the nature of the zinc salt, the addition rate, and the initial concentration of the reactants. Based on the findings of the study a two-stage mechanism for the formation of uniform polycrystalline particles with well-defined geometric shapes is proposed. © 2006 Published by Elsevier Inc. Keywords: Colloid; Zinc oxide; Double jet precipitation; Aggregation; Primary nanoparticles

1. Introduction In both bulk and colloidal form zinc oxide exhibits unique properties, which have been successfully exploited in many technological applications. The most notable uses of dispersed ZnO are in catalysis [1,2], construction of varistors and gas sensors [3–5], pigments [6], luminescent materials [7,8], and in the pharmaceutical and cosmetic industries [9,10]. Zinc oxide powders have also been employed as model materials for studying the sintering, grain growth, and microstructural development in ceramics [11–13]. Considering the specific needs of these very diverse applications, there has been strong interest in the development of preparation methods enabling to generate highly dispersed zinc oxide with well-controlled properties. Among other possibilities, the thermal decomposition/evaporative and the solution based methods remain the preferred synthetic routes to generate colloidal ZnO. The first approach, which consists in converting zinc precursors directly into crystalline zinc ox* Corresponding author. Fax: +1 315 268 2139.

E-mail address: [email protected] (D.V. Goia). 1 On leave from Institute of Physical Chemistry, Romanian Academy,

Bucharest, Romania. 0021-9797/$ – see front matter © 2006 Published by Elsevier Inc. doi:10.1016/j.jcis.2006.12.020

ide in gas phase at high temperatures, includes spray [14] and plasma [15] pyrolysis, chemical vapor deposition (CVD) [16–19], and physical vapor deposition (PVD) [20]. The solution route relies on the hydrolysis of zinc ions in alkaline conditions, which often yields highly dispersed uniform ZnO particles with well-controlled size and shape [21,22]. Experimentally, this approach can be implemented in different forms, including precipitation in homogeneous solutions [21–27], sol– gel [27–29], microwave [30], hydrothermal [31,32], and microemulsion [33] techniques. The nature of the reactants, their concentrations, the type and the charge of the counterions, the reaction temperature and time, and the presence of dispersants were all found to affect the formation of ZnO particles [23]. Additionally, it was shown that the way the reactant solutions are mixed can also influence the nucleation and growth of the solid phase and, therefore, the properties of the final solids [23,24]. This paper describes a simple precipitation method for preparing highly dispersed uniform zinc oxide colloids at low temperatures and discusses the mechanisms responsible for their formation. It will be shown that the final polycrystalline ZnO particles are formed by the aggregation of nanosize primary particles and their internal structure and shape depend on the size and the spatial arrangement of the constituent subunits.

M. Jitianu, D.V. Goia / Journal of Colloid and Interface Science 309 (2007) 78–85

79

2. Experimental 2.1. Materials High purity (>99.0%) zinc nitrate {Zn(NO3 )2 ·6H2 O}, zinc sulfate {ZnSO4 ·7H2 O}, zinc acetate {Zn(CH3 COO)2 }, and sodium hydroxide pellets (all obtained from Alfa Aesar) were used to obtain stock solutions containing 0.5 mol dm−3 Zn2+ and respectively 1.0 mol dm−3 NaOH. The Daxad 11G (AG, Hampshire Chemical Corp.) and the Arabic gum (Flutarom Inc., North Bergen/NJ) were used as received. 2.2. Preparation methods The precipitations were carried out in a 1000 cm−3 jacketed cylindrical beaker connected to a constant temperature circulating bath. In a typical experiment, equal volumes (100 cm3 ) of zinc salt and sodium hydroxide stock solutions were added in parallel into 150 cm3 of deionized water containing the dissolved dispersing agent. The reacting solutions were delivered at flow rates ranging from 0.28 to 1.1 cm3 min−1 using two metering pumps, the mixing being provided by a 1 diameter three-blade SS propeller spinning at 500 rpm. The majority of the precipitations were carried out at 60 ◦ C using zinc nitrate. In the experiments in which a dispersant was employed, the quantity added was arbitrarily selected to be equal to the amount of ZnO formed assuming the complete precipitation of zinc ions. After the addition of the reactants was completed, the resulting white precipitate was washed repeatedly with distilled water by decantation. In the case of the experiments in which stable dispersions were formed, the solids were separated by ultracentrifugation and then redispersed in DI water. Once the pH of the supernatant was below ∼8.0, the ZnO particles were rinsed twice with alcohol, filtered, and dried at 80 ◦ C for several hours. 2.3. Characterizations The crystalline structure of the dried solids was evaluated by X-ray powder diffraction with a Bruker Advance D8 diffractometer using CuKα (λ = 1.5418 Å) as incident radiation. The 2θ range was scanned with a step of 0.02◦ , the data collection time for each step being 2 s. In addition to identifying the crystalline phases, the XRD data were also used to estimate the size of the constituent crystallites by the Scherrer method. The size and morphology of the ZnO particles, as well as their surface topography, were evaluated by field emission electron microscopy (FE-SEM) with a JEOL-JSM-7400F instrument. The particle size and size distribution of the dispersed ZnO were obtained from the SEM images by averaging minimum 100 particles and by laser diffraction size analysis using a Malvern 2000e instrument. 3. Results The reaction temperature, the zinc salt type, the nature of the dispersing agent, the reactants concentration and the addition rate were varied in this study, as summarized in Table 1.

Fig. 1. Field emission electron micrographs of solids obtained at 60 ◦ C (a) in the absence of dispersant (Sample 1), (b) with Daxad 11G (Sample 2), and (c) with Arabic gum (Sample 3).

The slow addition of zinc nitrate and sodium hydroxide stock solutions at 60 ◦ C in pure DI water (Sample 1) resulted in the formation of nonuniform agglomerated particles, which settled as soon as the agitation was stopped (Fig. 1a). The addition of Daxad 11G in the reactor changed the shape of the particles formed (Fig. 1b) but did not improve the stability of the dispersion (Sample 2). In contrast, in the case of Sample 3, the same amount of Arabic gum (AG) yielded a stable dispersion of uniform particles having an average size of ∼260 nm (Fig. 1c). In all three cases the final solids displayed XRD patterns characteristic for crystalline zinc oxide (Fig. 2). Surprisingly, despite the fact that large ‘crystal-like’ particles were obtained in pure water, the crystallinity of the solids formed in the presence of dispersants was more pronounced as indicated by the sharper

80

M. Jitianu, D.V. Goia / Journal of Colloid and Interface Science 309 (2007) 78–85

Table 1 Experimental conditions and properties of precipitated ZnO particles Sample Zinc salt

Zn2+ conc. (mol cm−1 )

Reactants flow (mL min−1 )

Reaction time (h)

Dispersant

T (◦ C)

Dispersion Particle Particle stabilitya uniformityb size (nm)

Crystallite size (nm)

Particle shape

1 2 3 4 5 6 7 8

0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.75

0.28 0.28 0.28 0.28 1.12 0.28 0.28 0.28

6 6 6 6 1.5 6 6 6

None Daxad AG AG AG AG AG AG

60 60 60 20 60 60 60 60

U U S S S S S S

21 33 44 10 28 39 23 43

Irregular crystals Ellipsoidal Hexagonal prisms Spheroidal Spheroidal Hexagonal prisms Irregular Hexagonal prisms

Zn(NO3 )2 Zn(NO3 )2 Zn(NO3 )2 Zn(NO3 )2 Zn(NO3 )2 ZnSO4 ZnAc2 Zn(NO3 )2

P P VG G G VG P G

Broad Broad ∼260 ∼300 ∼100 ∼270 20–300 ∼300

a Dispersion stability: U—unstable, S—stable. b Particle uniformity: P—poor, G—good, VG—very good.

Fig. 2. X-ray diffraction patterns of solids obtained at 60 ◦ C in the absence of dispersant (Sample 1), with Daxad 11G (Sample 2), and with Arabic gum (Sample 3).

diffraction peaks and the increase in the calculated crystallite size from ∼21 nm in absence of dispersant to ∼33 nm for Daxad 11G and respectively ∼44 nm in the case of Arabic gum. The high degree of dispersion of the solids obtained with Arabic gum was confirmed by the excellent agreement between the mean particle size measured by laser diffraction (Fig. 3) and the average size obtained from the electron micrograph shown in Fig. 1c. Because of the remarkable stability of the dispersions prepared with Arabic gum and the uniformity of the ZnO particles obtained, most of this study was focused on understanding the effect of different experimental conditions on this particular system. When the experiment with Arabic gum (Sample 3) was carried out at 20 ◦ C instead of 60 ◦ C, a very stable dispersion of ZnO was also obtained. While the decrease in the temperature caused only a slight increase in the average particle size to ∼300 nm, the effect on the shape and the internal struc-

Fig. 3. The size distribution of the ZnO particles obtained in Sample 3.

M. Jitianu, D.V. Goia / Journal of Colloid and Interface Science 309 (2007) 78–85

81

Fig. 4. Electron micrographs of ZnO particles obtained at (a) 60 ◦ C (Sample 3) and (b) 20 ◦ C (Sample 4).

ture of the solids was significant. Indeed, in contrast with the ‘hexagonal prism’-like particles obtained at 60 ◦ C (Fig. 4a), the ZnO particles formed at the lower temperature (Sample 4) were spherical/spheroidal aggregates of very small nanoparticles (Fig. 4b). The dramatic impact of the temperature on the crystalline structure was confirmed by the much broader and poorly separated XRD peaks obtained in the case of the solids prepared at 20 ◦ C (Fig. 5a) and the smaller calculated crystallite size (∼10 nm), which was in good agreement with the value estimated based on the high resolution electron micrograph shown in Fig. 5b. Particles with similar shape and internal structure were obtained at the higher temperature (60 ◦ C) as well when the addition rate was increased from 0.28 to 1.1 cm3 min−1 (Fig. 6). They were, however, much smaller in size (∼100 nm) and the constituent primary ZnO nanoparticles were larger (∼28 nm). Changing the nature of the zinc salt in the experimental conditions used for Sample 3 affected both the uniformity and the shape of the ZnO particles. When zinc sulfate was used instead of zinc nitrate (Sample 6), the size of the primary crystallites and final particles as well as their general shape were preserved but their aspect ratio changed due to a lower height of the hexagonal prisms (Fig. 7a). In contrast, with zinc acetate (Sample 7) the zinc oxide particles obtained had irregular shape and a broad size distribution (Fig. 7b), while the crystallite size was considerably smaller (∼23 nm).

Fig. 5. (a) XRD patterns of ZnO powders obtained at 20 ◦ C and (b) high resolution FESEM image showing the fine structure of the polycrystalline aggregated particles.

Fig. 6. Field emission electron micrographs of ZnO particles obtained at 60 ◦ C at an increased addition rate of the reactants (Sample 5).

The increase of the concentration of the zinc solution from 0.5 to 0.75 mol dm−3 and NaOH solution from 0.75 to 1.12 mol dm−3 in otherwise the same conditions (Sample 8) did not change significantly the size of the final ZnO particles but did affect slightly their morphology. The hexagonal prisms obtained in this case had roughly the same diameter as at lower concentration but were more elongated and occasionally ‘branched’ (Figs. 8a and 8b).

82

M. Jitianu, D.V. Goia / Journal of Colloid and Interface Science 309 (2007) 78–85

Fig. 7. Field emission electron micrographs of ZnO particles prepared at 60 ◦ C in the presence of Arabic gum as dispersant with (a) zinc sulfate and (b) zinc acetate.

Fig. 8. FESEM images at two magnifications of zinc oxide particles obtained by reacting 0.75 M solution of zinc nitrate and 1.5 M sodium hydroxide at 60 ◦ C in the presence of Arabic gum (Sample 8).

4. Discussion

The time dependent investigations conducted in this study have revealed that in such ‘double jet’ configuration zinc hydroxide is still formed early in the reaction but it is completely converted later in the process into crystalline ZnO. Indeed, the XRD data in Fig. 9 clearly show that, at both 20 and 60 ◦ C, after the first hour the peaks of both ε-zinc hydroxide (wulfingite) and zinc oxide are present, while after two hours only the latter can be detected. As expected, at the higher temperature the conversion of the hydroxide into ZnO is accelerated as indicated by the more intense peaks of the oxide (Fig. 9b). Interestingly enough, after the complete conversion of the hydroxide into crystalline oxide has occurred, the former could not be detected despite the fact that the Zn2+ and the OH− ions were continuously added at the same rate in the reactor. The instability of the initially formed hydroxide and the direct formation of crystalline ZnO in the late stages of the precipitation are quite surprising findings, particularly for the experiment conducted at room temperature where the dehydration and aging of the hydroxide are assumed to be generally slow. It appears that keeping at all times the concentrations of Zn2+ and OH− ions at or below the solubility limit (as is the situation in the ‘double jet’ process) favors the formation of Zn(OH)+ or a more reactive form of zinc hydroxide, which can be rather easily converted into crystalline ZnO. The fact that the hydroxide did not form at all during the last stages of the precipitation is even more puzzling. One possible explanation could be that the surface of the crystalline zinc oxide formed

Unlike the case of most polyvalent cations, the hydrolysis of zinc ions follows a rather simple path. The reaction of Zn2+ with OH− ions results first in the formation of Zn(OH)+ species followed by the precipitation of zinc hydroxide, which easily redissolves in excess of base to form a soluble complex. Simple calculations based on the values of the solubility product for Zn(OH)2 and the stability constant of Zn(OH)2− 4 species, show that the complete precipitation of the zinc ions (i.e., a residual Zn2+ concentration of less than 10−6 mol dm3 ), can be achieved only if the pH is higher than 8.1 but not above 10.5. The nature of the solids formed, however, depends significantly on how the precipitation process is conducted. If, for example, a strong base is added to a solution of zinc salt, the high concentration of hydroxyl ions at the contact point between the reactants favors the rapid formation of a white gelatinous zinc hydroxide, which through a ‘sol–gel’ transformation changes rapidly into crystalline Zn(OH)2 , most often the orthorhombic wulfingite. The latter is rather stable and can be further dehydrated and converted into ZnO only as a result of a subsequent aging, which can be accelerated at elevated temperatures [34]. In contrast, as previously shown by Matijevi´c and others [23,24], if stoichiometric amounts of zinc salt and base are added in parallel (‘double-jet’) and slowly into the vigorously mixed solvent, crystalline ZnO can be directly precipitated.

M. Jitianu, D.V. Goia / Journal of Colloid and Interface Science 309 (2007) 78–85

83

(a)

Fig. 10. High resolution field emission electron micrograph of solids collected (a) after 1 h and (b) at the end of the 6-h process in the case of the precipitation conducted at 20 ◦ C.

(b) Fig. 9. XRD spectra of solids collected at 1, 2, and 6 h for precipitations conducted at (a) 20 ◦ C and (b) 60 ◦ C.

acts as a template which either accelerates the dehydration of the hydroxide or completely bypasses its formation by favoring an alternative hydrolysis mechanism in which the Zn(OH)− species are converted directly into ZnO. As indicated by both electron microscopy and XRD data, the final uniform ZnO particles consist of aggregates of nanosize primary particles. Such outcome is not at all surprising, this formation mechanism being very often observed in many

inorganic systems such as metals [35], metal sulfides [36], oxides [37], as well as organic ones [38]. A model for the spontaneous aggregation of nanosize precursors, an event triggered most often by the decrease in the thickness of the double layer at high ionic strength or the reduction of the surface charge near the isoelectric point (IP), was described previously in detail for the formation of monodispersed spherical gold [39] and CdS [40]. According to this model, the essential condition for the size selection to occur is that the generation rate of primary particles must decay during the process. In the system investigated here, the aggregation mechanism was particularly obvious when the precipitation was conducted slowly at 20 ◦ C, as confirmed by the electron micrograph image in Figs. 4b and 5b. The time dependent studies showed that the aggregation of the primary nanoparticles occurred very early in the process, as testified by the high resolution electron micrograph of the solids collected after the first hour (Fig. 10a), and that the aggregates continued to grow during the remaining 5 h by the attachment of newly formed primary particles. The attachment appeared to be random during most of the process, except for the end of the precipitation when the precursors showed a tendency to rearrange into organized layered structures (Fig. 10b). While the rearrangement of the constituent nanosize precursors was too slow at 20 ◦ C to affect the shape of the final ZnO particles during the 6 h long reaction, at 60 ◦ C the process was significantly faster. In this case, the ordered structures of the primary particles (this time slightly larger at

84

M. Jitianu, D.V. Goia / Journal of Colloid and Interface Science 309 (2007) 78–85

Fig. 11. FESEM images at two magnifications of ZnO particles collected in the reaction conducted at 60 ◦ C (Sample 3) after (a, b) 1 h, (c, d) 3 h, and (e, f) at the end of the 6-h process.

∼24 nm) were observed after only 1 h (Figs. 11a and 11b) and after 3 h the rearrangement was extensive enough to completely change the shape of a large number of aggregates from spherical to hexagonal prisms (Figs. 11c and 11d). Eventually, by the end of the precipitation all ZnO particles consisted of hexagonal prisms with relatively smooth sides and rough ‘end’ surfaces (Figs. 11e and 11f). The rearrangement of the primary particles was found to be a time dependent process. Indeed, even at 60 ◦ C, when the addition of reactants was completed in only 1.5 h (Sample 5), the final ZnO particles preserved their spherical shape (Fig. 6). The mechanism of aggregation and rearrangement of the primary particles was also affected by the nature of the anions present in the system. The divalent sulfate ions, for example, still favored the formation of hexagonal prisms and changed

only their height-to-diameter ratio (Fig. 7a). The acetate ions, however, prevented the formation of uniform particles with a well-defined morphology, likely because their larger size diminished the effect of the attractive forces between the primary nanoparticles. While it can be easily rationalized why uniform spherical entities are formed by the rapid aggregation of nanosize primary particles, it is quite difficult to understand how such event would yield directly uniform particles with other regular geometric shapes. The findings of this study support a two-step process for the formation of such structures. According to this proposed mechanism, in the first step the aggregation yields larger polycrystalline spheres, as the process is too rapid to allow the spatial rearrangement of the nanosize primary particles. In a much slower second step, the rearrangement of the con-

M. Jitianu, D.V. Goia / Journal of Colloid and Interface Science 309 (2007) 78–85

stituent precursors can eventually proceed leading to particles of other regular geometric shapes. The latter process is likely favored in the case of solids with higher solubility in the dispersion medium and it may or may not involve the growth of the primary particles. The final shape of the particles depends on many parameters including but not limited to the crystal structure of the substance, the nature of the soluble species present (ions, molecules, polymers), mixing, solvent, etc. This two-step mechanism is supported by the findings of other studies, which showed that polycrystalline particles with regular geometric shapes are usually formed as a result of slow precipitation processes [37]. Finally, it must be emphasized that the Arabic gum played a key role in the precipitation process being responsible for the formation of a highly dispersed ZnO particles with controlled size and defined shape despite the high ionic strength of the system. For this reason, the method described may represent a cost effective synthetic route to prepare large quantities of ZnO powders suitable for applications in electronics and sensors. The effectiveness of the dispersant was diminished however at even higher ionic strength where a slight aggregation of the final ZnO particles occurred. Interestingly enough however, the aggregation was somewhat orderly, the individual prisms being stacked mostly along their highest symmetry axis to form elongated hexagonal prisms (Fig. 8b). 5. Summary This study describes a versatile precipitation method for preparing uniform, highly dispersed zinc oxide particles. By adding the zinc salt and base solutions in parallel (‘double jet’), it was possible to generate even at ambient temperatures crystalline ZnO particles without the need for a subsequent heat treatment. It was found that the larger uniform oxide particles were formed as a result of the aggregation of much smaller nanosize primary particles. The time dependent investigations showed that the size, shape, and internal structure of the final particles were dictated by the size and the spatial arrangement of the nanosize subunits, which in turn depended on the nature of the dispersant, the type of zinc salt, the reactants addition rate, and the temperature of the process. The findings of the study indicated that the large ‘hexagonal prism’-like particles were formed by a two-stage mechanism involving first the rapid aggregation of nanosize precursors into larger uniform spheres followed by a slow rearrangement leading to the final shape. Acknowledgment This work was sponsored by Umicore (Hanau, Germany) and by the NSF Grant DMR-0509104.

85

References [1] C.A.K. Gouvea, F. Wypych, S.G. Moraes, N. Duran, N. Nagata, P. PeraltaZamora, Chemosphere 40 (2000) 433. [2] W. Elseviers, H. Verelst, Fuel 78 (1999) 601. [3] A. Degen, M. Kosec, J. Eur. Ceram. Soc. 20 (2000) 667. [4] J. Wrzesinski, D. Frolich, Solid State Commun. 105 (5) (1998) 301. [5] M. Takehana, T. Nishino, K. Sugawara, T. Sugawara, Mater. Sci. Eng. B 41 (1996) 186. [6] F.A. Sigoli, M.R. Davolos, M. Jafelicci Jr., J. Alloys Compd. 262–263 (1997) 292. [7] S. Fujihara, H. Naito, T. Kimura, Thin Solid Films 389 (2001) 227. [8] C. Lorenz, A. Emmerling, J. Fricke, T. Schmidt, M. Hilgendorff, L. Spanhel, G. Muller, J. Non-Cryst. Solids 238 (1998) 1. [9] O. Yamamoto, Int. J. Inorg. Mater. 3 (2001) 643. [10] H. Akiyama, O. Yamasaki, H. Kanzaki, J. Tada, J. Arata, J. Dermatol. Sci. 17 (1998) 67. [11] T. Senda, R.C. Bradt, J. Am. Ceram. Soc. 74 (1991) 1296. [12] S.D. Shin, C.S. Sone, J.H. Han, D.Y. Kim, J. Am. Ceram. Soc. 79 (1996) 565. [13] B.P. Lim, J. Wang, S.C. Ng, C.H. Chew, L.M. Gan, Ceram. Int. 24 (1998) 205. [14] J. De Merchant, M. Cocivera, Chem. Mater. 7 (1995) 1742. [15] X.H. Zhang, S.Y. Xie, Z.Y. Jiang, X. Zhang, Z.Q. Tian, Z.X. Xie, R.B. Huang, L.S. Zheng, J. Phys. Chem. 107 (2003) 10114. [16] A.P. Roth, D.F. Williams, J. Appl. Phys. 52 (1981) 6685. [17] T. Maruyama, J. Shinoya, J. Mater. Sci. Lett. 11 (1992) 170. [18] M. Izaki, T. Omi, Appl. Phys. Lett. 68 (1996) 2439. [19] M. Izaki, T. Omi, J. Electrochem. Soc. 143 (1996) L53. [20] H. Kim, W. Sigmund, J. Nanosci. Nanotechnol. 4 (2004) 275. [21] A. Chittofrati, E. Matijevic, Colloids Surf. 48 (1990) 65. [22] J.E. Rodriguez-Paez, A.C. Caballero, M. Villegas, C. Moure, P. Duran, J.F. Fernandez, J. Eur. Ceram. Soc. 21 (2001) 925. [23] A.P.A. Oliveira, J.F. Hochepied, F. Grillon, M.H. Berger, Chem. Mater. 15 (2003) 3202. [24] Q.P. Zhong, P. Matijevi´c, J. Mater. Chem. 6 (1996) 443. [25] M. Takehana, T. Nishino, K. Sugawara, T. Sugawara, Mater. Sci. Eng. B 41 (1996) 186. [26] T. Trindade, J.D. Dejesus, P. O’Brien, J. Mater. Chem. 4 (1994) 1611. [27] H. Van der Rul, D. Mondelaers, M.K. Van Bael, J. Mullens, J. Sol–Gel Sci. Technol. 39 (2006) 41. [28] J. Zhang, H. Fena, W. Hao, T. Wang, J. Sol–Gel Sci. Technol. 39 (2006) 37. [29] V. Briois, C. Giorgetti, E. Dartyge, F. Baudelet, M. Tokumoto, S. Pulcinelli, C. Santilli, J. Sol–Gel Sci. Technol. 39 (2006) 25. [30] S. Komarneni, M. Bruno, E. Mariani, Mater. Res. Bull. 35 (2000) 1843. [31] C.H. Lu, C.H. Yeh, Ceram. Int. 26 (2000) 351. [32] Z. Zhou, H. Deng, J. Yi, S. Liu, Mater. Res. Bull. 34 (1999) 1563. [33] C.H. Lu, C.H. Yeh, Mater. Lett. 33 (1997) 129. [34] R.A. McBride, J.M. Kelly, D.E. McCormack, J. Mater. Chem. 13 (2003) 1196. [35] D.V. Goia, E. Matijevi´c, Colloids Surf. A 146 (1999) 139. [36] S. Libert, D.V. Goia, E. Matijevi´c, Langmuir 19 (2003) 10673. [37] W.P. Hsu, L. Rönnquist, E. Matijevi´c, Langmuir 4 (1988) 31. [38] G. Pozarnsky, E. Matijevi´c, Colloids Surf. 125 (1997) 593. [39] V. Privman, D.V. Goia, J. Park, E. Matijevi´c, J. Colloid Interface Sci. 213 (1999) 36. [40] S. Libert, V. Gorshkov, V. Privman, D.V. Goia, E. Matijevi´c, Adv. Colloid Interface Sci. 100–102 (2003) 169.

Colloid chemistry Lecture 9: Colloid stability

Theories of the stability of colloidal disperse systems

(1 nm – 10 μm)

Instability of liophobic colloids aggregation:

coalescence:

Colloid stability requires repulsive forces between colliding particles

Main types of (de)stabilization of colloidal dispersions electrostatic potential

steric potential

Coulombic repulsion ĺ ĺ electrostatic stabilization

entropy hindrance ĺ ĺ steric stabilization

van der Waals attractive interactions; attraction potential : VA

x

attraction potential between plate-like particles:

attraction potential between spherical particles:

VA (x)



Ar 12 x

A: Hamaker constant (attraction parameter)

r

0

VA (x) x



A | 2 ȖS 12 ʌ x 2

A: Hamaker constant (attraction parameter)

Hamaker constants of various materials

Hamaker constants of various materials

Combination of Hamaker constants

A132 | A131 A232 A12 | A11 A 22

A132: particle(1)-particle(2) interaction through medium (3) A12: particle(1)-particle(2) interaction in vacuum

A131 | A 313 | A11  A 33  2 A13 |  A11  A 33

2

A132 |  A11  A 33  A 22  A 33

van der Waals attractions in colloidal disperse systems The stability of a colloidal disperse system is strongly dependent on the attractive pair potential, VA, between the dispersed particles. VA is determined by the geometric arrangement, G, of the particles (e.g. lamella-lamella; sphere-sphere; sphere-lamella; etc. interactions, independently of the chemical composition), and the Hamaker constant, A, of the overall system (which depends on the chemical composition of the constituting species, but is independent of the geometrc arrangement). Formally: VA = A × G The Hamaker constant A of the overall system originates from a combination of the individual Hamaker constants, Ai, of the dispersion medium and that of the dispersed particles. It can be derived from the summation of the der Waals dispersion forces between the constituting species (dispersion medium; dispersed particles).

VA, van der Waals interaction energy (kT)

Dependence of the van der Waals pair potential VA between two colliding particles on the overall Hamaker constant ”A” of the system

two spherical particles, R=4μm

surface separation, x (nm), of two interacting particles

DLVO theory: Electrostatic stability Derjagin

Landau

Verwey Overbeek

van der Waals attraction

electrostatic repulsion

Aggregation is hindered by electrostatic repulsion

DLVO theory: the theory of electrostatic stabilization (Derjagin-Landau and Verwey-Overbeek) The electrostatic repulsion depends on the... 1. …Stern potential ȥSt 2. …thickness of the electric double layer, N-1

Electric double layer interaction energies; repulsion energies VR 

VR

64 n kT  Nx e N

between two sheets (lamellae):

VA



A 12S x 2

VA

Ar 12 x

between two spheres:

VR

8 k 2 T 2 Hr D  Nx e e2 z 2

\ ª ze º zkT e  1 « » ze \ « » zkT e  1 »¼ ¬«

2

DLVO theory: the resultant (total) potential is: VT = VA + VR x

repulsion potential

attraction potential (van der Waals forces) VA kT

0

(Coulombic repulsion; aqueous medium, large H)

DLVO theory Derjagin Landau Verwey Overbeek

x

VR kT

0

x

VT kT

total potential 0

VT = VA + VR

x

(VT/kT) t 10: o “colloid stabiliy”

DLVO theory: conditions for colloid stability

Stability

repulsive forces overwhelm attractive forces

Coagulation

attractive forces overwhelm repulsive forces

DLVO theory: conditions for colloid stability

Colloid stability/flocculation/coagulation/ are controlled by the relative magnitudes of the van der Waals and the Coulombic forces

potential

repulsion

0

aqueous Al2O3 suspension at different pHs

total

attraction

particle-particle distance (surface separation)

aqueous As2S3 sol with increasing background electrolyte concentraion (1:1 electrolyte, mM)

Colloid stability/flocculation/coagulation/ are controlled by the relative magnitudes of the van der Waals and the Coulombic forces

stable emulsion

stable suspension

unstable

unstable

(coagulation; creaming and coalescence)

(coagulation / flocculation; sedimentation)

For VT = 0, VR = -VA. In this case, the DLVO theory predicts the ccc ratio for 1:2:3 valence of charge of ions to be 1000:16:1.3. Schulze-Hardy rule: ccc v

1 z6



1 1 1 : : 16 2 6 36



electrolyte

ccc (M) (As2S3 dispersion)

Schulze-Hardy-rule

NaCl

5,1 10-2

1000

KCl

5,0 10-2

1000

MgCl2

7,2 10-4

16

(14)

CaCl2

6,5 10-4

16

(13)

AlCl3

9,3 10-5

1,3

(1,8)

(980)

ccc

interaction energy / 10-19 J

c increases c = 1 mM

23 mM 4 mM

90 mM

360 mM 1500 mM

ccc = 65 mM

r = 0.1Pm; T=298K; A212=10-19J; \St=50mV; z = 1; H 

As2S3 sol in 1:1 electrolyte

ccc

stable Fe(OH)3 sol

The sol undergoes coagulation upon the addition of Al2(SO4)3 solution

Mechanisms of coagulation

perikinetic

differential settling

orthokinetic

perikinetic: collisions by Brownian motion differential settling (polydisperse suspensions) orthokinetic: induced collisions through stirring; shear

The kinetics of coagulation x

r

Fast coagulation (Smoluchowski): each collision leads to aggregation (high electrolyte concentration, Vmin  0, rate constant: kf)

no 1  4 S (2 r  x ) D n o t

n

Slow coagulation (Fuchs): only part of the collisions leads to aggregation (low/intermediate electrolyte concentration, Vmin ~ kT, rate constant: ks)

n

no 1  W 4 S (2 r  x ) D n o t

where W

kf ks

t s,1 / 2 t f,1 / 2

( t12: half life time )

Colloid stability: thermodynamic and kinetic aspects W  1: stability ratio (Fuchs) f

V 1 § V( x ) · d x 2 r ³ exp¨ exp max ¸ 2 | kT 2Nr © kT ¹ x 2r

lec

kf ks

t s,1 / 2 t f,1 / 2

turbidity measurements W

ly tro

tro lyt e

W

c ele

2:2 e

1:1

log W

W

te

fast coagulation region

logW

a  b log c

log c

Summary of main types of particle-particle interactions x

(x > į )

(x < 2į 2į), e.g.:

interactions.

Colloid stability requires repulsive forces between colliding particles

Main types of (de)stabilization of colloidal dispersions electrostatic potential

steric potential

Coulombic repulsion ĺ ĺ electrostatic stabilization

entropy hindrance ĺ ĺ steric stabilization

Conformations of adsorbed polymer chains

Conformations of adsorbed polymer chains

(a) polymer in solution; (b) chemisorbed (end-grafted) copolymer; (c) physisorbed homopolymer; (d) adsorption at low surface coverage with no neares neighbour overlap (‘mushrooms’); (e) adsorption at high coverage (‘brush’); (f) bridging

Typical Polymer Adsorption Isotherms

Amount adsorbed, ī (mg/m2)

the effect of molecular weight 1.4

Langmuir isotherm: ī

ī max

1.2

c K c MW3 MW 3

1 0.8

MW22 MW

0.6

MW1 MW 1

0.4 0.2

MW3 >> MW MW1 MW MW 2> 3 2 > MW 1

0 0

100

200

300

c (mg/L)

400

500

600

Typical thickness of the adsorption layer in case of electrostatic stabilization and steric stabilization hoį [nm]

1:1 electrolyte c [mmol.dm-3]

1/ț [nm]

polymer M [g.mol-1]

0.01

100

1,000,000

60

0.1

30

100,000

20

1

10

10,000

6

10

3

1,000

2

100

1 (thickness of the electric double layer)

-

(end-to-end distance  polymer layer thickness)

Polymers at interfaces

B

C

Conditions for efficient steric stabilization

cpolymer

(1) large ī (point C in the Figure) (2) large į (layer thickness) (3) large 3s (adsorption energy) (4) Ȥ < 0.5 (good solvent for the polymer chain) (5) low c (free polymer concentration) note: (3) may conflict with (4) for homopolymers; this conflict is absent for graft- and block copolymers

bridging flocculation

aggregation for small values of ī (below point B in the Figure)

Steric interaction

r

x/2į

x = 2į

x

Typical potential function of steric stabilization

VT VS steric repulsion

x

~ coil diameter, rg VA van der Waals attraction

Brownian movement/ agitation

electrolytes

polymers

3. Dispersed Systems Introduction. Many food ingredients are completely immiscible and so will form separate phases within the food. However the sizes of these phases can be very small, so to the naked eye the food will appear homogeneous. The techniques and principles of colloid science are suited to dealing with the properties of fine particles and their applications to foods will be explored in this section. A colloidal particle is many times larger than an individual molecule but many times smaller than anything that can be seen with the naked eye. Some examples of colloidal particles in foods are listed in the adjacent table (Table is reproduced from Fennema text). A colloidal particle has at least one length dimension in the range (approximately) of tens of nanometers and tens of micrometers. Although colloidal particles are small, each contains a huge number of molecules. To get an idea of the number of molecules needed to form a colloidal particle: a 10 nm droplet of water would contain about 105 molecules, a 10 µm about 1014 molecules. To stretch the analogy, if a water molecule was a person, a 10 nm droplet would be have as many people as a city the size of Bethlehem (PA), while a 10 µm droplet 10,000 times the population of the planet. Particles must be dispersed in a second phase. (Note that the particles are the dispersed phase and the phase they are dispersed in is the continuous phase). Depending what the phases are (solid, liquid, gas) it is possible to generate different types of colloidal system. Their names and some examples are given in the following table:

Dispersed phase

Continuous phase

Solid

Solid Solid Glass (e.g., frozen food)

Liquid Gas

Solid foam (e.g., whipped candy)

Liquid Sol (e.g., molten chocolate)

Smoke

Emulsion (e.g., cream)

Aerosol (e.g., spray)

Foam (e.g., beer)

Gas

Surface Chemistry. The behavior of large objects is governed by gravity. The behavior of very small objects is governed by thermal motion. Colloidal objects will diffuse randomly in response to thermal energy (Brownian motion) but may also settle out slowly. In addition, for colloidal objects surface properties are important. Interfacial energy depends on the chemical dissimilarity between the continuous and dispersed phases. At a molecular level, like molecules are more attracted to like. Therefore a molecule at the surface will have a net attraction back into the bulk. Any attempt to increase the surface area will force more molecules to the surface and increase the net pull opposing the expansion. The magnitude of the pull depends on the extent of the molecular dissimilarity (i.e., how much more the surface molecules would rather interact with similar molecules). A classic experiment to measure this type of surface force is to measure the force it takes to stretch out a soap bubble. Pulling back the plunger forces more molecules to the airsolution interface and is opposed by a force related to the molecular dissimilarity between soap solution molecules and air. The slope of the surface energy with interfacial area graph is γ – the interfacial tension. dG=γdA where dG is the change in surface excess free energy caused by a change in interfacial area (dA). As in all things there is a drive to minimize Gibbs free energy. For surface energy this can either be done by reducing the surface area (i.e., increasing the size of the droplets by allowing them to merge) or reducing the interfacial tension (i.e., by adsorbing a surfactant to the interface).

Complicating the issue of surface tension is the impact of surface curvature. Surface tension pulls surface molecules towards the center of a droplet, slightly increasing the pressure inside. If a droplet is small (i.e., the surface is highly curved) there are more molecules pulling in per unit volume so the pressure is even higher. The increase in pressure (over atmospheric pressure) is the Laplace pressure (PL) a function of interfacial tension (g) and the radius of curvature:

2γ r Laplace pressure has two important consequences for food colloids: 1. Small fluid (gas or liquid droplets in foams or emulsions) colloids behave as hard spheres. Any attempt to deform them means the curvature changes and the pressure in some parts of the droplet is higher than in others. To equalize the pressure the droplet (or bubble) reverts to its spherical shape. Smaller droplets are obviously more rigid than larger ones because the surface-volume ratio is greater. 2. Small droplets/bubbles are more soluble than large ones because solubility increases with pressure. The solubility differential will drive the diffusion of dispersed phase material from small to large particles through a process known as Ostwald ripening. When the solubility of the dispersed phase in the continuous is intrinsically very low Ostwald ripening is unlikely to be significant (e.g., oil is very insoluble in water so food emulsion have little tendency to Ostwald ripen). On the other hand more soluble dispersed phase is more prone to Ostwald ripening (e.g., carbon dioxide in a foam, ice crystals in ice cream). PL =

Surfaces provide an ideal environment for amphiphilic molecules. An amphiphile has part of its structure water-soluble and part water insoluble. Examples include polymers (e.g., proteins) and small molecules (i.e., surfactants, e.g., soaps, Tween, lecithin). By aligning at an interface they can solubilize their water-soluble parts in the aqueous phase and their hydrophobic parts in the less polar phase (e.g., oil or air). Surface adsorption represents an entropy cost and well as an entropy/enthalpy benefit for the surface-active material. Free surface-active material is able to diffuse freely about the system and is thus has higher entropy than if it were bound. On the other hand bound surface-active material does not have to pay the entropy cost of having hydrophobic portions of their structure in contact with water. The trade-off between cost and benefit means that there needs to be some finite concentration of surface-active material added before there is surface adsorption. The amount of adsorbed surface-active material will increase as more as added until the surface becomes saturated and there is no more available space (i.e., monolayer adsorption). The relationship between total concentration and surface concentration is a surface adsorption isotherm. (Think of this in analogy to a moisture sorption isotherm – a relationship between total moisture, humidity, and bound water). Typical surface saturations are in the order of a few mg per m2 of available surface. More surface area (smaller particles) means there must be more surfactant used to fill it.

Each surface-active molecule that adsorbs at the surface blocks some unfavorable contact between the immiscible phases and goes some way to stabilize the system. If the surface tension in the absence of surfactant is γ0 and in the presence of surface active material is γ then the surface pressure (the extent of surface tension lowering) is π (=γ0-γ). The greater the surface pressure, the lower the interfacial free energy and the more stable the system. The more surface-active material adsorbed, the lower the surface pressure (i.e., lower the surface tension). Because the surface load reaches a plateau at high [surfactant] the surface tension reaches its minimum at a similar concentration (and surface pressure its maximum). [Small molecule surfactants have a lower surface affinity than a polymeric (protein) surfactant so it is necessary to add more to get any surface adsorption. However the surfactant can lower the interfacial tension more than the protein. (Imagine the surfactant molecule “packing” more efficiently at the surface and better block the phases from each other.) An important consequence for foods is that if a mixture of surfactant and protein are used to create a dispersed system, the surfactant will predominate at the interface. Secondly, if surfactant is added to a protein-stabilized dispersed system, it may displace the protein. The competitive adsorption of surfactant in place of protein is important in understanding the role of added small-molecule emulsifiers in ice cream.] Properties of Dispersed Systems. The key parameters governing the properties of all dispersed systems are the type (i.e., which phase is continuous?), particle concentration, and particle size. • The type of emulsion can usually be readily determined – the system “feels” and behaves most like the continuous phase (e.g., mayonnaise disperses relatively easily in water but not in oil because it is an oil-in-water emulsion). • There is always a distribution of sizes in any real dispersed system. The presence of even a few large particles can lead to much more rapid destabilization. Particle size can be reduced (within limits) by increasing the energy used in preparing the dispersion or by adding more surfactant during the dispersion process. • The concentration of the particles (usually given as a volume-fraction φ) can increase from zero to close-packing. Close-packing is a geometric constraint on the number of spheres that can be physically fitted into an available space (e.g., there is a limited number of basketballs that can fit into a room and there will still be some unfilled space). Close packing for monodisperse spheres is about 69% and much higher for polydisperse particles.

Dispersed systems are always more viscous than the pure continuous phase and the viscosity is greater the more dispersed particles present. Viscosity is the intermolecular friction that must be overcome to make a liquid flow. At its simplest, fluid flow is seen as a velocity gradient with one layer (streamline) of fluid flowing relative to another. The boundary layer of fluid adjacent to a stationary surface can only flow at an infinitesimally small speed. The layer beyond that an infinitesimal amount faster and so on. The more friction between the layer, the more energy needed to achieve a given velocity gradient and the higher the viscosity.

When particles are included in the fluid stream they disrupt the streamlines as they force liquid to flow around them. This increases the amount of energy needed to achieve a given velocity gradient and increases the viscosity. Another, more systematic, way of looking at this is to imagine the particles as blocking some streamlines. To achieve the same overall velocity gradient, the velocity gradient over the liquid parts of the system (the unblocked streamlines) must be higher. To achieve the same overall velocity gradient as the particle-free system it costs more energy because a higher velocity gradient in the remaining liquid parts of the system. However you conceptualize the mechanism of viscosity-building by particles the magnitude of the effect can be described by the Einstein equation: η=η0 (1+2.5φ) where η is the viscosity of the dispersion, η0 is the viscosity of the continuous phase (i.e., no particles) and φ is the volume fraction of particles present. The Einstein equation only works quantitatively up to a few percent particles, at higher concentrations the viscosity will increase even more quickly than predicted. Creaming (Sedimentation). The effect of particles on a flowing liquid are the basis of the increased viscosity of dispersed systems. However particles will also diffuse and move in a stationary liquid. Very small particles will be most affected by temperature and diffuse randomly by Brownian motion. Large particles diffuse less and are affected more by gravity so tend to either float or sink depending on their density. Oil is less dense than water so will tend to float to the surface of an emulsion (i.e., creaming). The speed of the droplet as it floats upwards is retarded by the friction it experiences with the continuous phase.

The tendency of a particle to float is the buoyancy – a product of the amount of material trying to float (i.e., particle volume) and the density difference between phases (∆ρ) and the acceleration due to gravity (g):

Fb =

πd 3 g 6

∆ρ

The frictional force opposing a spherical particle as it moves upwards is proportional to its surface area, its speed and the viscosity of the continuous phase (ηc). Ff = 3πη c rv

A droplet will float upwards at ever increasing speeds in response to the buoyancy force until the velocity is sufficient to allow the frictional force to exactly match it. When Fb=Ff the particle will continue to move upwards at its terminal velocity (vs): vs =

d 2 ∆ ρg 18ηc

The Stokes-Einstein equation provides an estimate of the speed of a creaming particle. The lower vs, the more stable the emulsion will be to creaming. Stokes-Einstein is rarely quantitatively applicable but it gives some ideas to reduce creaming rate (increase continuous phase viscosity, reduce particle size, increase oil density). Foams. Foams can be formed either by whipping a gas into a liquid or by bubble nucleation (e.g., from yeast cells or from a supersaturated CO2 solution). Foams are similar in many ways to emulsions but have some distinctive feature worth consideration: • Gasses are more soluble in water than oils so the rate of Ostwald ripening is much more rapid. • Bubbles tend to be much larger than droplets because lower energy levels are used to form them and because very small bubbles tend to disappear quickly by Ostwald ripening. • Creaming is typically much faster in foams as the density difference and the bubble size is much larger. • Large foam bubbles are more capable of deformation than small oil droplets so it is possible to reach very high volume fractions (φ>99%) • Dilute foams (e.g., soda) break down by rapid creaming. Concentrated foams (e.g., meringue) break down by (i) drainage and (ii) film rupture. Particle Aggregation. The mechanisms of emulsion stability considered so far (i.e., creaming – very important especially for large droplets, and Ostwald ripening – rarely important unless the oil has significant water-solubility) require no droplet-droplet contact. Droplet-droplet collisions and interactions are essential for droplet aggregation. There are two types of aggregations: fluid droplets can coalesce when they collide and merge to form one larger droplet. Fluid droplets can also collide, form a semi-permanent link, but maintain their individual identity and do not merge (i.e., flocculation). (Solid droplets can only flocculate). Flocculation usually precedes coalescence and in practice most food emulsions spoil before significant coalescence has occurred.

Flocculated networks of particles are open structures that include a proportion of continuous phase in their structure. The effective volume fraction of a flocculated system is therefore larger than a corresponding non-flocculated system. In fact, most sudden changes in viscosity for a dispersed system depend on the formation or fracture of flocs. A second major consequence of flocculation is because the effective particle size (the hydrodynamic radius) is larger the creaming rate is much faster. Very extensive flocculation allows the formation of an extended particle network spanning the sample. The heavily flocculated fluid will now behave as a gel and will not cream at all because all the particles are interconnected. The rate of droplet aggregation processes depends on the number of droplet collisions and their effectiveness (i.e., how many of the collisions lead to particles “sticking” and forming a floc). Collision kinetics is a second order process (i.e., rate depends on the square of the number of particles present). The second order rate constant can be calculated from the diffusion coefficient of the particles present as: k fast =

4kT 3η

where T is absolute temperature, k is the Boltzmann constant and η is the continuous phase viscosity. The rate predicted by this equation (i.e., Smoulokowski kinetics) is very high. In practice not every collision leads to droplet aggregation and we must reduce this rate by a collision efficiency w (usually>1): k=kfast/w. The collision efficiency is related to an energy barrier preventing the droplets colliding. If the size of the energy barrier (or w) is large, k will tend to zero and the particles will not aggregate. We can imagine the energy barrier as a “force field” that surrounds the droplets. There are some non-covalent attractive forces and some repulsive forces that surround each droplet. Their sum gives the magnitude of the effective force field and the energy barrier that may prevent droplets reacting.

Van der Waals forces are weak transientdipole attractions between all matter. Particles will attract each other by these forces. However, they have a limited range and their effectiveness decreases as 1/separation. The force is conventionally expressed as a pair potential – the free Ar energy cost (or gain) to bring one a particle VA = − 12h from infinite separation to a distance h from a second similar particle. The Van der Waals function is negative because it is attractive (so ∆G2-3 kT) will prevent further droplet approach (some ionic strengths give a small secondary minima that may hold particles at a finite separation in loose flocs). Beyond the repulsive energy barrier the Van der Waals attraction again dominates and droplets that can overcome the barrier will rapidly coalesce in the energy well. Question 1: DVLO Theory. DVLO theory can predict the interaction potential between droplets in an emulsion. The main adjustable parameters in the equation are the surface potential (related to the charge density on the surface of the droplet – increase by either adsorbing more protein or increasing the charge on the each protein, pH-pI) and the ionic strength (=Σcz2, where c is the concentration of ions of charge z. Increase by adding salts). In this exercise, use the Excel spreadsheet on the class website to calculate DVLO potentials between droplets with different surface potential and ionic strength. For surface potentials in the range 0.1-1, calculate the maximum ionic strength that can maintain an energy barrier opposing aggregation. Present your answer as graph of maximum ionic strength (y) against surface potential (x) and one or two sentences explaining what this stability map tells us about the types of emulsion composition we can expect to be stable.

The DVLO approach is in reality an oversimplification. There are many more forces that may be important and a truly successful theory would incorporate all of these. Among the most important are steric repulsion forces. A thick layer of surfactant surrounding a particle can prevent the approach of a second particle and inhibit aggregation. Steric repulsion is very powerful but very short range. They can typically hold droplets at a close separation in flocs without allowing them to merge and coalesce. The two mechanisms of steric repulsion are: (i) Osmotic. As the droplets approach one another the aqueous portion of the surfactants overlap. Locally the concentration of surfactant molecules increases and the concentration of water decreases. This sets up an osmotic pressure gradient between the overlap area and the outside solution. Water will diffuse in according to the Osmotic pressure and force the particles apart. (ii) Mechanical. The physical space taken up by the aqueous portions of the surfactant molecules cannot be taken up by other molecules. When a second droplet approaches the physical presence of one layer prevents the other getting too close. Any tendency for the surface proteins on one droplet to bond with the surface proteins on another droplet will quickly favor aggregation.

Surface Tension

Ishita Patel, M.S.

SURFACE TENSION Surface tension is an effect within the surface layer of a liquid that causes that layer to behave as an elastic sheet. In the bulk of the liquid each molecule is pulled equally in all directions by neighboring liquid molecules, resulting in a net force of zero. At the surface of the liquid, the molecules are pulled inwards by other molecules deeper inside the liquid but they are not attracted as intensely by the molecules in the neighbouring medium. Therefore all of the molecules at the surface are subject to an inward force of molecular attraction which can be balanced only by the resistance of the liquid to compression.

Molecules on the surface of a liquid experience an imbalance of forces The net effect of this situation is the presence of free energy at the surface. The excess energy is called surface free energy and can be quantified as a measurement of energy per unit area. It is also possible to describe this situation as having a line tension or surface tension, which is quantified as a force per unit length measurement. The common units for surface tension are dynes/cm. Surface effects might be expressed in the language of thermodynamics, dG = VdP + γdA + SdT At constant temperature and pressure the Gibbs free energy becomes, dG = γdA Since γ is a positive constant under a given set of conditions we note that, • If dA is positive (surface area increases) then dG is positive • If dA is negative (surface area decreases) then dG is negative This simply means that that decreasing the surface area of a substance is always spontaneous (∆G0). A measure of how spontaneous (or non-spontaneous) is the change in the surface area is precisely the surface tension. The work of surface formation at constant volume and temperature can be expressed as the change in the Helmholtz energy, dA = γdσ where σ is the surface area.

1

Surface Tension

Ishita Patel, M.S.

For example: Compute the work needed to raise a platinum-iridium alloy ring of mean circunference 5.00 cm and to stretch the surface of pure water at 20.00C through a height of 0.02 cm. The surface tension of pure water at 20.00C is 72.8 dynes/cm. Ignore gravitational effects.

Free body diagram of the platinum-iridium alloy ring • The first thing to notice from the free body diagram is that there are two sigmas (σ) because there are two surface areas: one inside and one outside the ring. By using the mean circunference of the ring we make the computed areas nearly equal. dσ(in/out) = 0.02 cm X 5.00 cm = 0.10 cm2 dσ(both) = 2 X 0.10 cm2 = 0.20 cm2 dA = γ X dσ(both) dA = [72.8 dynes/cm] X [0.20 cm2] dA = 14.56 dynes X cm = 14.56 X 10-7 Joules. Compute the weight of the heaviest platinum-iridium alloy ring of mean circunference 5.00 cm that will float in pure water at 20.00C.

Cross section showing the forces acting on the platinum-iridium alloy ring • The sum of the vertical components of F1 and F2 balances the weight W of the ring. F1 + F2 -W = 0 (γL)cosθ + (γL)cosθ -W = 0 W = 2(γL)cosθ • The forces due to the surface tension will balance the largest weight when they point completely vertically and θ = 00. To get an answer in standard units of weight we use γ = 0.0728 N/m. W = 2(0.0728 N/m X 0.05 m)cos00 = 7.28 X 10-3 N • Using the value of 9.8 m/s2 for the acceleration due to gravity, 7.28 X 10-3 N / (9.8 m/s2) = 7.43 X 10-4 kg = 0.743 g 2

Surface Tension

Ishita Patel, M.S.

Du Noüy RING METHOD Historically the Du Noüy Ring method was the first to be developed. In this class, we will be using the Surface Tensiomat Model 21 from Fisher Scientific. The modern design still contains the basic elements of a surface tensiometer that could have been built in the middle ages.

An antique tensiometer and its modern conterpart In the Du Noüy Ring method the liquid is raised until contact with the surface is registered. The sample is then lowered again so that the liquid film produced beneath the ring is stretched. The surface tension would then be given by, γ = F/2L where F is the detachment force, L is the mean circunference of the ring and the factor of two (2) takes care of the inner and outher surface of the ring. In order to understand what is going on, let's look at a graph of Force as a function of ring distance.

Change of force with ring distance

3

Surface Tension

Ishita Patel, M.S.

Suppose you are back in the old days you would have had to dip the ring just under the surface of the liquid of which you want determine the surface tension and level the balance in these conditions. At this point your graph is in the F1 range. Some masses would then be added to the opposite arm until the ring detached from the liquid. At this point your graph is either at the Fmax point or somewhere in the F3 range. Then it becomes clear that you should read the surface tension during a return movement as well and the results should agree. But we are in the modern age now. The lowering and raising of the liquid sample is done by a motor driven mechanisn and the Suface Tensiomat Model 21 can be calibrated to read the surface tension directly in dynes/cm. Still, reproducibility of results within certain tolerance is required.

REQUIRED MATERIALS AND APPARATUS -Surface Tensiomat Model 21 -Destilled Water -Petri Dishes -n-butanol (density 0.8098 g/mL) -n-pentanol (density 0.8110 g/mL) -thermometer

EXPERIMENTAL PROCEDURE

Main parts of the Surface Tensiomat Model 21

4

Surface Tension

Ishita Patel, M.S.

The operation of the Surface Tensiomat Model 21 is described in the respective Instruction Manual. For most laboratory work The following instructions should be enough. Adjust the height of the Movable Table: It is best to obtain petri dishes that have the similar diameter. That way, adding a constant volume of liquid will result in a constant liquid height and the table adjustment will need to be done only once. The liquid level in the dish should be about 1 cm. To move the table by a large distance use the table adjustment (L). To move the table by a small distance (a couple of mm) use the table adjustment (S). Set the Zero Tension Reading: Raise the sample table until the ring is about 2 mm under the liquid. Release the tension of the machine by turning the Tension Knob. Turning the knob counterclockwise increases the tension, turning the knob clockwise decreases the tension. When the Index is aligned with its image in the mirror, the tension should be zero. [1] Adjust the reading in the main dial to zero using the Set Zero Knob. [2] Increase the tension until the film breaks. [3] Release the tension, stopping at the point when the ring returns to the liquid. [4] Readjust the reading in the main dial to zero using the Set Zero Knob. [5] Repeat until reproducibility within a few dynes/cm is obtained. Calibrate the Tensiometer: The readings might be a bit too small or a bit too large. However, we do not want to have to do adjust the tensiometer's mechanism thus we will have to multiply all readings by a calibration constant. For this we use distilled water at 200C, which has a surface tension of 72.8 dynes/cm. Obtain the correction factor: Take at least five (5) measurements and assume that the mean value is a good approximation for γ. The correction factor cf for the subsequent determinations will be, 72.8 dynes/cm --------------- = cf γ

Every future measurement is to be multiplied by this correction factor. Notice that: [1] If γ is smaller than 72.8 dynes/cm, then the correction factor is larger than 1. [2] If γ is larger than 72.8 dynes/cm, then the correction factor is smaller than 1. [3] If γ is exactly 72.8 dynes/cm, then the correction factor is exactly 1 (No correction needed). Prepare a 0.10 M solution of n-butanol, let's call this concentration C. Dilute precisely to 1/2 of the original 0.10 M concentration, let's call this concentration C/2. Repeat as indicated in the following table, Solution # 1 2 3 4 5 6

Concentration C C/2 C/4 C/8 C/16 C/32

5

Surface Tension

Ishita Patel, M.S.

Determine the surface tension and temperature of each solution. It is recommended that all measurements be made at nearly the same temperature. Take at least three readings for each concentration. Rinse and dry the ring between measurements. Repeat the procedure with Distilled Water at the following temperatures. 1 2 3 4 5 6

Temperature 0C 10 20 30 40 50 60

DATA PROCESSING -For the n-butanol solutions, plot the surface tension γ in the ordinate and the logarithm of the bulk concentration C in the abscissa. -Determine the slope of the line. -Compute the surface concentration using the Gibbs isotherm, Γ = -[1/RT][dγ/d(ln C)] -Γ has units of mol per surface area in cm2. Convert cm2 to Å2, take the reciprocal of this quantity and use Avogadro's number to obtain the surface area per molecule (σ). -The difference between the surface tension of the solvent and that of the solution, γ0-γ, is the force per centimeter exerted by the adsorbed molecules at the interface. This force per unit length is the surface pressure, π. Find π for each solution. -Plot π as a function of σ. This is the two-dimensional isotherm analogous to the three dimensional P as a function of V graph. -From the graph of π as a function of σ, determine the surface pressure and surface area per mole of adsorbed solute at the point of complete surface coverage (monolayer formation). -For the distilled water trials, plot the surface tension γ in the ordinate and the temperature T in the abscissa. -The general trend is that surface tension decreases with temperature, reaching a value of 0 at the critical temperature Tc. Extrapolate to γ = 0 and determine Tc for water. -Use Eötvös empirical equation to estimate the molar volume of water, γV2/3 = k (Tc - T) where Tc the critical temperature and k = 1.03 erg/0C for water. You could use T = 200C and γ = 72.8 dynes/cm or some other point for which the data is well characterized. -Submit all graphs with your report. Testing your knowledge: In the Gospel of John (John 6:16-21) Jesus is said to have walked over water. Some scholars believe this miracle involved a change in the surface tension of the water in the immediate vicinity of Jesus. Compute the surface tension of this holy region of water. According to the canonical Gospels, Jesus weighted 75 kilograms and wore sandals size 8 (USA). 6

Advanced Drug Delivery Reviews 63 (2011) 456–469

Contents lists available at ScienceDirect

Advanced Drug Delivery Reviews j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / a d d r

Physical and chemical stability of drug nanoparticles☆ Libo Wu, Jian Zhang, Wiwik Watanabe ⁎ MAP Pharmaceuticals, Inc. 2400 Bayshore Parkway, Mountain View, CA 94043, USA

a r t i c l e

i n f o

Article history: Received 12 August 2010 Accepted 2 February 2011 Available online 21 February 2011 Keywords: Drug nanoparticles Nanosuspensions Physical stability Chemical stability Stabilizer

a b s t r a c t As nano-sizing is becoming a more common approach for pharmaceutical product development, researchers are taking advantage of the unique inherent properties of nanoparticles for a wide variety of applications. This article reviews the physical and chemical stability of drug nanoparticles, including their mechanisms and corresponding characterization techniques. A few common strategies to overcome stability issues are also discussed. Published by Elsevier B.V.

Contents 1. 2.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stability of drug nanoparticles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Effect of dosage form on stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. General stability issues related to nanosuspensions . . . . . . . . . . . . . . . . . . . . 2.2.1. Sedimentation or creaming . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.2. Agglomeration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.3. Crystal growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.4. Change of crystalline state . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.5. Stability issues with solidification process of nanosuspensions . . . . . . . . . . . 2.2.6. Chemical stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. Additional stability issues relate to large biomolecules . . . . . . . . . . . . . . . . . . . 3. Characterizing stability of drug nanoparticles and nanoparticle formulations . . . . . . . . . . . . 3.1. Particle size, size distribution and morphology. . . . . . . . . . . . . . . . . . . . . . . 3.2. Sedimentation/creaming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3. Particle surface charge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4. Crystalline state . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5. Chemical stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6. Additional techniques for assessing large biomolecule nanoparticle and formulation stability 4. Recommendations of general strategies for enhancing stability of nanoparticle formulations . . . . 5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

456 457 457 457 457 458 463 463 464 464 464 465 465 466 466 466 466 466 467 467 467

1. Introduction

☆ This review is part of the Advanced Drug Delivery Reviews theme issue on “Nanodrug Particles and Nanoformulations for Drug Delivery”. ⁎ Corresponding author. Tel.: +1 650 386 8193; fax: +1 650 386 3100. E-mail address: [email protected] (W. Watanabe). 0169-409X/$ – see front matter. Published by Elsevier B.V. doi:10.1016/j.addr.2011.02.001

With significant attention focused on nanoscience and nanotechnology in recent years, nanomaterial-based drug delivery has been propelled to the forefront by researchers from both academia and industry [1–3]. Various nano-structured materials were produced and applied to drug delivery such as nanoparticles [4], nanocapsules [5],

L. Wu et al. / Advanced Drug Delivery Reviews 63 (2011) 456–469

nanotubes [6], micelles [7], microemulsions [8] and liposomes [9]. In general, the term “nanoparticles” refers to particles with sizes between 1 and 100 nm. However, submicron particles are also commonly referred as nanoparticles in the field of pharmaceutics and medicine [10–14]. Nanoparticles are categorized as nanocrystals [10], polymeric [15], liposomal [9] and solid lipid nanoparticles (SLN) [16] depending on their composition, function and morphology. Given the extensive available literature reviews on SLN, polymeric and liposomal nanoparticles [4,9,15–18], this article will focus only on nanocrystals (pure drug nanoparticles). The unique nano-scale structure of nanoparticles provides significant increases in surface area to volume ratio which results in notably different behavior, both in-vitro and in-vivo, as compared to the traditional microparticles [10–12]. Consequently, drug nanocrystals have been extensively used in a variety of dosage forms for different purposes [10,11,14,19,20], such as improving the oral bioavailability of poorly water-soluble drugs by utilizing enhanced solubility and dissolution rate of nanoparticles [21–23]. In the field of pulmonary drug delivery, the nanoparticles are able to deliver the drugs into the deep lungs, which is of great importance for systemically absorbed drugs [11,14]. In addition, injection of poorly water-soluble nanosuspension drugs is an emerging and rapidly growing field that has drawn increasing attention due to its benefits in reducing toxicity and increasing drug efficacy through elimination of co-solvent in the formulation [10,20]. Despite the advantages of drug nanocrystals, they present various drawbacks including complex manufacturing [12,24–26], nanotoxicity [27] and stability issues [10,19,20]. Stability is one of the critical aspects in ensuring safety and efficacy of drug products. In intravenously administered nanosuspensions, for example, formation of larger particles (N5 μm) could lead to capillary blockade and embolism [20], and thus drug particle size and size distribution need to be closely monitored during storage. The stability issues of drug nanoparticles could arise during manufacturing, storage and shipping. For instance, the high pressure or temperature produced during manufacturing can cause crystallinity change to the drug particles [12,26,28]. Storage and shipping of the drug products may also bring about a variety of stability problems such as sedimentation, agglomeration and crystal growth [29–31]. Therefore, stability issues associated with drug nanocrystals deserve significant attention during pharmaceutical product development. This article reviews existing literature on drug nanoparticle stability, including theory/mechanisms, methods used to tackle the stability problems and characterization techniques, and provides recommendations to improve the current practices. Since the stability issues related to nanoparticle dry powders are usually trivial, this review will only focus on stability of nanosuspensions (drug nanoparticles dispersed in a liquid medium).

2. Stability of drug nanoparticles 2.1. Effect of dosage form on stability The unique characteristics of drug nanoparticles have enabled their extensive application in various dosage forms including oral, parenteral, ocular, pulmonary, dermal and other specialized delivery systems [10,11,13,20,32]. Although different dosage forms may share some common stability issues, such as sedimentation, particle agglomeration or crystal growth, their effects on drug products are quite different. For instance, particle agglomeration could be a major issue in pulmonary drug delivery since it affects deposition amount/ site, and thus drug efficacy. On the other hand, agglomeration in intravenous formulations can cause blood capillary blockage and obstruct blood flow. Moreover, the selection of stabilizers is also closely related to dispersion medium, dosage form and strictly governed by FDA regulations. To date, there is a wide variety of

457

choices on the approved stabilizers for oral dosage form whereas the excipients allowed for inhalation are very limited [33]. Drug nanoparticles exist in the final drug products either in dry powder or suspension form. Examples of the dry powder form include the dry powder inhaler, lyophilized powder for injection and oral tablets or capsules. Solid dosage forms usually have good storage stability profiles, which is why a common strategy to enhance nanosuspension stability is to transform the suspension into solid form [19,25]. Most of the reported stability concerns arise from nanosuspensions in which the drug nanoparticles are dispersed in a medium with or without stabilizers. In addition, mechanisms involved in the stability of small and large biomolecule formulations are different due to their molecular structure differences. A small molecule drug is defined as a low molecular weight non-polymeric organic compound while large biomolecules refer to large bioactive molecules such as protein/peptide. One of the major issues with protein/peptide stability is to maintain the 3-dimensional molecular conformation, such as secondary and tertiary structure in order to keep their biological activities [34,35], whereas there is no such concern for small organic molecules. 2.2. General stability issues related to nanosuspensions Stability issues associated with nanosuspensions have been widely investigated and can be categorized as physical and chemical stability. The common physical stability issues include sedimentation/creaming, agglomeration, crystal growth and change of crystallinity state. 2.2.1. Sedimentation or creaming Drug particles can either settle down or cream up in the formulation medium depending on their density relative to the medium. The sedimentation rate is described by Stokes' law [36,37] which indicates the important role of particle size, medium viscosity and density difference between medium and dispersed phase in determining the sedimentation rate. Decreasing particle size is the most common strategy used to reduce particle settling. Matching drug particles density with medium or increasing medium viscosity are the other widely used approaches to alleviate sedimentation problems [37,38]. Fig. 1 shows different sedimentation types that occur in suspension formulations. In a deflocculated suspension (Fig. 1a), particles settle independently as small size entities resulting in a slow sedimentation rate. However, densely packed sediment, known as caking [39], is usually formed due to the pressure applied on each individual particle. This sediment is very difficult to be re-dispersed by agitation [36,37,39] and would be detrimental to the drug products performance. In the flocculated suspension (Fig. 1b), the agglomerated particles settle as loose aggregates instead of as individual particles [36,37]. The loose aggregates have a larger size compared to the single particle, and thus higher sedimentation rate. The loose structure of the rapidly settling flocs contains a significant amount of entrapped medium and this structure is preserved in the sediment. The final flocculation volume is therefore relatively large and the flocs can be easily broken and redispersed by simple agitation. K.P. Johnston et al. [40,41] have recently attempted to achieve stable nanosuspensions via a novel design of flocs structure called “open flocs”, as illustrated in Fig. 1c. Thin film freezing was used to produce BSA nanorods with aspect ratio of approximately 24. These BSA nanorods were found to be highly stable when dispersed into hydrofluoroalkane (HFA) propellant, with no apparent sedimentation observed for 1 year. Due to the high aspect ratio of BSA nanorods and relatively strong attractive Van der Waals (VDW) forces at the contact sites between the particles, primary nanorods were locked together rapidly as an open structure upon addition of HFA, inhibiting collapse of the flocs [41]. The lowdensity open flocs structure was then filled with liquid HFA medium, preventing particle settling. Similar results were shown using needle

458

L. Wu et al. / Advanced Drug Delivery Reviews 63 (2011) 456–469

Fig. 1. Sedimentation in (a) deflocculated suspension; (b) flocculated suspension; and (c) open flocs-based suspension.

and plate shaped itraconazole nanoparticles with aspect ratios between 5 and 10 [40]. Although sedimentation is one of the key issues for colloidal suspension, the reported studies examining sedimentation issues in aqueous-based nanosuspensions are very scarce. This could be due to (i) surfactants are generally used in most of the nanosuspensions to inhibit particle agglomeration in the medium, which alleviates the sedimentation issues and (ii) the small nano-sized particles significantly reduce the sedimentation rate. In addition, many of the aqueous nanosuspensions are transformed to dry solid form by spray drying or freeze drying to circumvent the long-term sedimentation issue. Unfortunately, this solidification process cannot be applied to non-aqueous nanosuspensions where sedimentation/creaming is commonly present. An example to illustrate this is metered dose inhaler (MDI) formulations where the nanoparticles are suspended in HFA propellants. Sedimentation or creaming is a key aspect affecting stability of these formulations. Particle engineering to optimize particle surface properties and morphology, e.g. hollow porous particles [42], and introduction of surfactant(s) is generally employed to alleviate the issue.

2.2.2. Agglomeration The large surface area of nanoparticles creates high total surface energy, which is thermodynamically unfavorable. Accordingly, the particles tend to agglomerate to minimize the surface energy. Agglomeration can cause a variety of issues for nanosuspensions including rapid settling/creaming, crystal growth and inconsistent dosing. The most common strategy to tackle this issue is to introduce stabilizers to the formulation. In addition to safety and regulation

considerations, selection of stabilizers is based on their ability to provide wetting to surface of the particles and offer a barrier to prevent nanoparticles from agglomeration [13,19]. There are two main mechanisms through which colloidal suspensions can be stabilized in both aqueous and non-aqueous medium, i.e. electrostatic repulsion and steric stabilization [10,36,37]. These two mechanisms can be achieved by adding ionic and non-ionic stabilizers into the medium, respectively. Stabilization from electrostatic repulsion can be described by the classic Derjaguin–Landau– Verwey–Overbeek (DLVO) theory [43,44]. This theory mainly applies to aqueous suspension while its application in non-aqueous medium is still not well-understood [33]. The DLVO theory assumes that the forces acting on the colloidal particles in a medium include repulsive electrostatic forces and attractive VDW forces. The repulsive forces are originated from the overlapping of electrical double layer (EDL) surrounding the particles in the medium, and thus preventing colloidal agglomeration. The EDL consist of two layers: (i) stern layer composed of counter ions attracted toward the particle surface to maintain electrical neutrality of the system and (ii) Gouy layer which is essentially a diffusion layer of ions (Fig. 2). The total potential energy (VT) of particle–particle interaction is a sum of repulsion potential (VR) generated from electric double layers and attraction potential (VA) from the VDW forces. VA is determined by the Hamaker constant, particle size and inter-particulate distance while VR depends on particle size, distance between the particles, zeta potential, ion concentration and dielectric constant of the medium. VR is extremely sensitive to ion concentration in the medium. As the ion strength is increased in the medium, the thickness of EDL decreases due to screening of the surface charge [36,37]. This causes decrease in VR, increasing the susceptibility of the dispersed particles to form aggregates. Zeta potential (ZP) is electric potential at the shear plane which is the boundary of the surrounding liquid layer attached to the moving particles in the medium. ZP is a key parameter widely used to predict suspension stability. The higher the ZP, the more stable the suspension is. In the case of steric stabilization, amphiphilic non-ionic stabilizers are usually utilized to provide steric stabilization which is dominated by solvation effect. As the non-ionic stabilizers are introduced into nanosuspensions, they are absorbed onto the drug particles through an anchor segment that strongly interacts with the dispersed particles, while the other well-solvated tail segment extends into the bulk medium (Fig. 3). As two colloidal particles approach each other, the stabilizing segments may interpenetrate, squeezing the bulk medium molecules out of the inter-particulate space as illustrated in Fig. 3. This interpenetration is thermodynamically disfavored when a good solvent is used as the bulk medium to stabilize the tail [36]. Accordingly, provided that the stabilizers can be absorbed onto the particle surface through the anchor segment, strong enthalpic interaction (good solvation) between the solvent and the stabilizing segment of the stabilizer is the key factor to achieve steric stabilization and prevent particles from agglomeration in the medium [36,37]. In addition to solvation, the stabilizing moiety needs to be sufficiently long and dense to maintain a steric barrier that is capable of minimizing particle–particle interaction to a level that the VDW attractive forces are less than the repulsive steric forces [43–45]. The main drawback associated with the steric stabilization is the constant need to tailor the anchoring tail according to the particular drug of interest. Due to the lack of fundamental understanding of interaction between the stabilizers and dispersed nanoparticles, current surfactant screening approaches to achieve a successful steric stabilization are mostly empirical and could be very burdensome [45– 49]. In addition, the solvation of the stabilizing segment is susceptible to variation in temperature. Stabilizer concentration could also play a role in causing suspension instability by affecting the absorption affinity of non-ionic stabilizers to drug particles surface. Deng et al.

L. Wu et al. / Advanced Drug Delivery Reviews 63 (2011) 456–469

459

Fig. 2. Illustration of classical DLVO theory. Attractive forces are dominant at very small and large distances, leading to primary and secondary minimum, while repulsive forces are prevailing at intermediate distances and create net repulsion between the dispersed particles, thus preventing particle agglomeration.

[50] used Pluronic® F127 to stabilize paclitaxel nanosuspensions. It was reported that stabilizers had high affinity to nanocrystals surface at concentrations below critical micelle concentration (CMC), and increasing concentrations above CMC caused loss of F127 affinity to the nanocrystals and thus unstable formulation. This was because F127 monomers on the nanocrystals surface started to aggregate with each other to form micelles as the concentration was increased to the CMC level, leading to a lower affinity to the drug crystals. Temperature was also shown to affect the stabilizer affinity to drug crystals. This is expected since the CMC level is dependent on temperature. It is apparent that combination of the two stabilization mechanisms can be very beneficial in achieving a stable colloidal dispersion. In addition, the combination of a non-ionic stabilizer with an ionic stabilizer reduces the self repulsion between the ionic surfactant molecules, leading to closer packing of the stabilizer molecules [10,51].

Besides the steric and electrostatic stabilization mechanisms, some other stabilization mechanisms have also been reported. Makhlof et al. produced indomethacin (IMC) nanocrystals using the emulsion solvent diffusion technique [52]. The nanoparticles were stabilized using various cyclodextrins (CyDs) without adding any surfactants. The stabilizing effect was attributed to the formation of a CyD network in the aqueous medium via intermolecular interaction of CyD molecules. The network-like structure was believed to prevent aggregation and crystal growth of IMC nanoparticles initially produced from the solvent diffusion process. Similar stabilization mechanism was also observed in another study where budesonide microsuspension was stabilized with hydroxypropyl-beta-cyclodextrin in HFA medium [53]. Another approach to enhance suspension stability that has increasingly been utilized is engineering of particle morphology. One breakthrough in this area was the porous particle

Fig. 3. Steric stabilization mechanisms according to Gibbs free energy: ΔG = ΔH−TΔS. A positive ΔG indicates stable suspension while negative ΔG induces particle agglomeration. If the medium is a good solvent for the stabilizing moiety, the adsorbed stabilizing layers on the dispersed particles cannot interpenetrate each other when the particles collide. This reduces the number of configurations available to the adsorbed stabilizing tails, resulting in a negative entropy change and positive ΔG. On the other hand, if the dispersion medium is a poor solvent, the adsorbed layers on the particles may interpenetrate thermodynamically and induces particles agglomeration.

460

L. Wu et al. / Advanced Drug Delivery Reviews 63 (2011) 456–469

concept that was first introduced by Edwards et al. [54]. The porous particles include hollow porous particle [42] and porous nanoparticleaggregate particles (PNAPs) [14]. Unfortunately, most of the work has been focused on microsuspension or polymeric colloidal formulations and has not been applied to pure drug nanoparticles. Table 1 summarizes a few published studies on pharmaceutical nanosuspensions. Due to the vast amount of literature work on the pharmaceutical nanosuspensions, this review will focus only on the studies that provide a more profound enlightenment on the stabilizer selection for nanosuspensions. The summary table shows that most of nanosuspensions were generated in aqueous medium, with only a limited number of nanosuspensions made in non-aqueous environment. The commonly used ionic stabilizers in aqueous medium include sodium dodecyl sulfate (SDS), sodium lauryl sulfate (SLS), lecithin and docusate sodium. The non-ionic surfactants used in aqueous medium are usually selected from Pluronic® surfactants, Tween 80, polyethylene glycol (PEG), polyvinyl alcohol (PVA) polyvinylpyrrolidone (PVP) and cellulose polymers such as hydroxypropyl cellulose (HPC) and hydroxypropyl methylcellulose (HPMC). The stabilizers are not only used to provide short- and long-term storage stability for nanosuspensions, but also to achieve successful formation and stabilization of nanocrystals during particle production. Lee et al. designed and synthesized various amino acid copolymers containing lysine as the hydrophilic segments with alanine, phenylalanine or leucine as hydrophobic moieties [49]. Wet comminution was used to produce naproxen nanosuspensions in presence of HPC and amino acid copolymers. Lysine copolymer with alanine was unable to produce submicron particles while the other copolymers with phenylalanine and leucine were capable of forming the nanoparticles. The size of nanocrystals was proven to be constant over 1 month storage and the crystallinity was also shown to be preserved after the wet comminution process. Furthermore, hydrophobicity of the copolymers was identified as the key factor in achieving the stable nanosuspensions, attributed to strong polymer adsorption onto the hydrophobic drug surfaces. Although this work did not provide an in-depth discussion on how the copolymers interacted with the drug nanoparticles, it illustrated the importance of careful selection of the anchor group (that is attached to the drug surface) in facilitating the production of a stable nanosuspesion. In the subsequent study [45], they attempted to understand the nature of interactions between polymeric stabilizers and drugs with different surface energies. Nanocrystals of seven model drugs with PVP K30 and HPC as stabilizers were generated using wet comminution. It was expected that a close match of surface energy between the stabilizers and drug crystals would promote the absorption of stabilizers onto drug particles, and thus help in reducing the particle size during the wet comminution process. Although surface energy did not seem to correlate well with particle size for HPC stabilized system, some trend was observed for PVP stabilized suspension with only one exception. A further study with seven stabilizers (non-ionic stabilizers: HPC, PVP K30, Pluronic® F127 & F68, PEG and ionic stabilizers: SDS and benzethoinum chloride) and eleven model drugs was conducted by the same group in order to provide more understanding on the stabilization mechanism [48]. Again, the general trend between surface energy and particle size reduction was not observed in this work. PEG was unsuccessful in reducing the particle size of most drug candidates while the other non-ionic stabilizers proved to be effective in reducing the size of five drug candidates that had similar surface energies to the stabilizers. F68 was shown to be the most effective stabilizer (successfully stabilizing nine drug candidates), which could be due to its strong chain adsorption onto the drug crystals through the hydrophobic polypropylene glycol (PPG) units. F127 was found to be less efficient than F68 likely because the short processing time led to inefficient physical adsorption of higher molecular weight F127 to the drug surface. This study demonstrated that a combination of ionic and non-ionic stabilizers is not always beneficial to enhance

stabilization, A few combinations of SDS or benzethoinum chloride with various non-ionic stabilizers resulted in positive stability effects while the others did not. The effects of physicochemical properties of the drugs on the stabilization were also explored in this study. In general, drugs with lower aqueous solubility, higher molecular weight and higher melting point were shown to have higher chance for successful nanosuspension formation. Van Eerdenbrugh et al. conducted an expanded study using 13 stabilizers at 3 different concentrations to stabilize 9 drug compounds [47]. The particles were generated using the wet milling technique. The success rate in producing nanosuspensions using polysaccharide based stabilizers [HPMC, methylcellulose (MC), hydroxyethylcellulose (HEC), HPC, carboxymethylcellulose sodium (NaCMC), alginic acid sodium (NaAlg)] was limited by the high viscosity of these polymeric stabilizer solutions. Increasing concentration of these stabilizers did not appear to be helpful. In contrast, the other stabilizers [PVP K30, PVP K90, PVA, Pluronic® F68, polyvinyl alcohol–polyethylene glycol graft copolymer (K-IR), Tween 80 and D-α-tocopherol polyethylene glycol 1000 succinate (TPGS)] did not encounter the viscosity issue. PVA was ineffective in producing the nanosuspension and the success probability of PVP K30, PVP K90, F68 and K-IR is highly dependent on their concentration. Higher concentrations (25 wt.% and 100 wt.%) increased the stabilizing efficacy significantly. Tween 80 and TGPS were proven to be the most effective stabilizers. Addition of TGPS (at concentrations N25 wt.%) allowed nanosuspension formation for all tested drug compounds. No correlation was observed between drug physicochemical properties (molecular weight, melting point, log p, solubility and density) and nanosuspension formation success rate. It was demonstrated that surface hydrophobicity of the drug candidates was the driving force for nanoparticles agglomeration, thus lowering the success rate of nanosuspension production. Mishra et al. explored nanosuspension stability issues during both production and storage [29]. Hesperetin nanosuspensions were produced using HPH with Pluronic® F68, alkyl polyglycoside (Plantacare 2000) and inulin lauryl carbamate (Inutec SP1), or Tween 80 as stabilizers. It was demonstrated that all stabilizers were suitable for successful production of hesperetin nanosuspensions. The size of nanocrystals was dependent on power density applied in the homogenization process and the hardness of the crystals. The effect of stabilizers on the particle size was negligible. Short-term stability over a period of 30 days was examined in order to evaluate the stabilizer efficiency. ZP was measured as a key parameter to predict the stability. In distilled water, the ZP values of all the nanosuspensions fell between −30 and −50 mV and the values dropped significantly in the original dispersion medium. This can be explained by the fact that adsorbed layers of large molecules shifted the shear plane to a longer distance from the particle surface, thus reducing the measured value of zeta potential (Fig. 4). However, the low ZP value does not point to an unstable suspension in this case, which could be due to the additional presence of steric stabilization mechanism. Both Inutec and Plantacare stabilized nanosuspensions also showed significant reduction of ZP measured from water to dispersion medium, indicating a thick absorbed steric layer and good stability. F68 exhibited only slight decrease in ZP, indicating a relatively thin stabilization layer. The ZP value of Tween 80 was only −13 mv in the dispersion medium, pointing to a potentially problematic stabilization. The study demonstrated that zeta potential measurement is a good predictor for storage stability. Nanosuspensions stabilized by Inutec and Plantacare were stable at all storage conditions (4, 25 and 40 °C) up to 30 days while F68 stabilized nanosuspensions were shown to be less stable. The Tween 80 formulation stability was the poorest. Pardeike et al. [30] conducted a similar study using phospholipase A2 inhibitor PX-18 nanosuspensions produced by HPH with Tween 80 as stabilizer. In this work, ZP of the homogenized nanosuspensions was dropped from − 50 mV to

L. Wu et al. / Advanced Drug Delivery Reviews 63 (2011) 456–469

461

Table 1 Literature summary of pharmaceutical nanosuspensions. Nanoparticles compound

Manufacturing technique

Delivery route

Dispersion medium

Stabilizers

Reference

Oridonin Oridonin Budesonide Buparvaquone Buparvaquone Diclofenac acid Azothromycin Rutin Rutin Tarazepide Omeprazole Amphotericin B Nimodipine Albendazole RMKP 22 Hesperetin

HPH HPH HPH HPH HPH HPH HPH HPH HPH HPH HPH HPH HPH HPH HPH HPH

NA IV Inhalation Inhalation Oral Oral NA Oral Oral NA IV Oral IV Oral NA Dermal

Water Water Water Water Water Water Water Water Water Water Water Water Water Water Water Water

Gao et al. (2007) [55] Gao et al. (2008) [56] Jacobs et al. (2002) [57] Hernadez-Trejo et al.(2005) [58] Jacobs et al. (2002) [59] Lai et al. (2009) [60] Zhang et al. (2007) [61] Mauludin et al. (2009) [62] Mauludin et al. (2009) [63] Jacobs et al. (2000) [64] Moschwitzer, (2004) [65] Kayser et al. (2003) [22] Xiong et al., (2008) [66] Kumar et al. (2008) [23] Peters et al. (1999) [67] Mishra et al. (2009) [29]

Hydrocortisone, prednisolone and dexamethasone Ascorbyl palmitate RMKK99 Nifedipine Undisclosed Hydroxycamptothecin Asulacrine RMKP 22 RMKP 22 PX-18 PX-18 Silybin Tarazepide Omeprazole, albendazole and danazol Fluticasone, budesonide Naproxen Loviride Nine different compounds Zinc Insulin Ethyl Diatrizoate Cinnarizine, itraconazole and phenylbutazone Nine different compounds Beclomethasone dipropionate Rilpivirine Undisclosed Piposulfan, etoposide, camptothecin, paclitaxel Naproxen Seven different compounds Eleven different compounds

HPH

Opthalmic

Water

PVP K25, Brij 78, SDS, Pluronic® F68, lecithin Pluronic® F68, lecithin Lecithin, Span 85, tyloxapol, cetyl alcohol Pluronic® F68 and PVA Pluronic® F68 and lecithin Pluronic® F68 Lecithin, Pluronic® F68, Tween 80 SDS SDS, Tween 80, Pluronic® F68, PVA Tween 80, Pluronic® F68 Pluronic® F68 Tween 80, Pluronic® F68 Pluronic® F68, sodium cholic acid and mannitol SLS, Carbopol, PS 80, hpmc Phospholipon 90 Pluronic® F68, Inutec SP1, Tween 80 and Plantacare 2000 Pluronic® F68

HPH HPH HPH HPH HPH HPH HPH HPH HPH HPH HPH HPH Wet milling

NA NA NA Oral NA IV NA NA NA NA Oral, IV IV Oral

Water Water Water Water Water Water Water Water Water Water Water Water Water

SDS, Tween 80 Potassium oleate, Tween 80 HPMC SLS, HPMC, PVA, Acaciae Gum, Pluronic® F127 Lipoid S75, Pluronic® F68, Solutol® HS 15 Pluronic® F68 Tween 80 Tween 80, Glycerol Tween 80 Tween 80 Lecithin, Poloxamer 188 Pluronic® F68, Tween 80, Glycerol Pluronic® F108, F68

Teeranachaideekul et (2008) [69] Krause et al. (2001) [70] Hecq et al. (2005) [71] Hecq et al. (2006) [72] Zhao et al. (2010) [73] Ganta et al. (2009) [74] Muller et al. (1998) [75] Grau et al. (2000) [76] Pardeike et al. (2010) [30] Wang et al. (2010) [77] Wang et al. (2010) [78] Jacobs et al. (2000) [64] Tanaka et al. (2009) [79]

Wet milling Wet milling Wet milling Wet milling Wet milling Wet milling Wet milling

Inhalation NA NA NA NA NA NA

Water Water Water Water Water Water Water

Tween 80 HPC, arginie hydrochloride Tween 80, Pluronic® F68 13 different stabilizers Pluronic® F68, sodium deoxycholate Poloxamine 908 TPGS 1000

Yang et al. (2008) [80] Ain-Ai et al. (2008) [81] Van Eerdenbrugh et al. (2007) [82] Van Eerdenbrugh et al. (2009) [47] Merisko-Liversidge et al. (2004) [83] Na et al. (1999) [84] Van Eerdenbrugh et al. (2008) [85]

Wet milling Wet milling Wet milling Wet milling Wet milling

NA Inhalation Parenteral NA NA

Water Water Water Water Water

TPGS 1000 PVA Pluronic® F108, TPGS 1000 Plasdone S-630, docusate sodium Tween 80, Span 80, Pluronic® F108, F127

Van Eerdenbrugh et al. (2008) [86] Wiedmann et al. (1997) [87] Baert et al. (2009) [88] Deng et al. (2008) [89] Merisko-Liversidge et al. (1996) [90]

Wet comminution Wet comminution Wet comminution

NA NA NA

Water Water Water

Lee et al. (2005) [49] Choi et al. (2005) [45] Lee et al. (2008) [48]

Vibrational rod milling Vibrational rod milling Precipitation, microfluidization Precipitation, microfluidization Solvent diffusion, melt emulsification Complex precipitation

NA NA NA

Water Water Water

NA

Water

Copolymers of amino acids HPC, PVP HPC, PVP, PEG, SDS, Pluronic® F68, F127, benzethonium chloride PVP K30, sodium deoxycholate PVP, SDS SLS, PVP K30, Pluronic® F68, F127, Tween 80, HPMC PVP, HPMC, SLS

NA

Water

PVA, PVP K25, Pluronic® F68, Tween 80,

Kocbek et al. (2006) [94]

NA

Water

None

Epstein et al. (2007) [95]

Stabilization of nanocrystal (SNC) Antisolvent precipitation & Wet-milling Antisolvent precipitation Antisolvent precipitation Antisolvent precipitation Antisolvent precipitation, Wet milling Antisolvent precipitation

NA

Water

Pluronic® F127

Deng et al. (2010) [50]

NA

Water

PVP K30, SDS, docusate sodium

Lindfors (2007) [96]

Oral NA Inhalation IV, Oral

Water Water Water Water

PVP K15, Pluronic® F127 HPMC, PVP K17 Tween 80 PVP, SDS, Miglyol, docusate sodium

Chen et al. (2009) [97] Douroumis et al. (2007) [98] Tam et al. (2008) [99] Sigfridsson et al. (2007) [100]

NA

Water

HPMC, lipoid S75, PEG-5 soy sterol

Douroumis et al. (2006) [101]

Antisolvent precipitation

NA

Water

None

Zhang et al. (2006) [102]

Dihydroartemisinin Probucol Ibuprofen Hydrocortisone Ibuprofen Alendronate-gallium, alendronate-gadolinium Paclitaxel Felodipine Naproxen Carbamazepine Cyclosporin A Undisclosed β-methasone valerate-17, oxcarbazepine Retinoic acid

Kassem et al. (2007) [68]

Chingunpitak et al.(2008) [91] Pongpeerapat et al. (2008) [92] Verma et al. (2009) [31] Ali et al. (2009) [93]

(continued on next page)

462

L. Wu et al. / Advanced Drug Delivery Reviews 63 (2011) 456–469

Table 1 (continued) Nanoparticles compound

Manufacturing technique

Delivery route

Dispersion medium

Stabilizers

Reference

2-devinyl-2-(1-hexyloxyethyl) pyropheophorbide Nitrendipine

Antisolvent precipitation

NA

Water

None

Baba et al. (2007) [103]

Oral

Water

PVA

Xia et al. (2010) [104]

NA Oral NA NA NA Inhalation

Water Water Water Water Water HFA

Cyclodextrins Tween 80, PVP K30, SDS Tween 80, Oramix CG-110 Tween 80, caprylyl-capryl glucoside, lecithin Lecithin None

Makhlof et al. (2008) [52] Dolenc et al. (2009) [105] Trotta el al.(2003) [106] Trotta el al.(2001) [107] Trotta el al.(2003) [108] Nyambura et al.(2009) [109]

Inhalation Inhalation Inhalation

HFA HFA HFA

None None Citral, cineole

Engstrom et al.(2009) [41] Tam et al.(2010) [40] Nyambura et al.(2009) [110]

Inhalation

HFA

Lecithin, docusate sodium

Dickinson et al. (2001) [111]

NA NA

Acetonitrile Tween 80 Ethyl acetate Methyl-β-cyclodextrin

Precipitation– ultrasonication Indomethancin Emulsion diffusion Celecoxib Emulsion diffusion Griseofulvin Emulsion diffusion Mitotane Emulsion diffusion Griseofulvin Microemulsion diffusion Lysozyme Emulsification/freezedrying Bovine serum albumin Thin film freezing Itraconazole Thin film freezing Insulin Emulsification + freeze-drying Salbutamol sulfate Microemulsion + freeze-drying Salbutamol sulfate HPH Horseradish peroxidase, carbonic Freeze-drying anhydrase, lysozyme, subtilisin carlsberg and α-chymotrypsin Diclofenac Emulsification + freeze drying

Transdermal Isopropyl myristate

around − 20 mV when tested from water to dispersion medium. It is generally believed that ZP of ±20 mV is sufficient to maintain a stable formulation with a combined electrostatic and steric stabilization [30]. The PX-18 nanosuspension was shown to be physically stable (no changes in particle size distribution) for more than half year at the storage condition of 5 and 25 °C. However, physical instability was observed after 1 month storage at a higher storage temperature. This could be due to the decreased dynamic viscosity and enhanced diffusion constant at higher temperature. There is another interesting work by Pongpeerapat et al. investigating probucol/PVP/SDS ternary ground mixture (GM) that was prepared with a vibrational rod mill [92]. The produced primary probutcol nanoparticles were around 20 nm in presence of both SDS and PVP. An interesting phenomenon was observed following the dispersion of the GM into water. For GM stabilized with PVP K17 and SDS, spherical agglomerates of primary nanocrystals were formed immediately in the size of around 90 nm after dispersion of the GM into water. A further agglomeration to around 160 nm in size occurred gradually during the storage stability study. In the case of PVP K12 and SDS, agglomerations of approximately 180 nm were observed after 4 days of storage and then remained stable up to 84 days. This phenomenon is illustrated in Fig. 5. Above critical aggregation concentration, SDS complexes with PVP to form a “necklace” structure in aqueous medium through both electrostatic and hydrophobic interactions. Following dispersion of probucol/PVP K17/SDS into

Sucrose ester

Ahmad et al. (2009) [112] Montalvo et al. (2008) [113]

Piao et al. (2007) [114]

water, PVP K17/SDS “necklace” complex interacted with primary drug nanoparticles, causing immediate agglomeration of the primary nanoparticles into 90 nm aggregates. The 160 nm secondary nanoparticles were formed due to further gradual agglomeration process. The stabilization of probucol nanocrystals was attributed to formation of PVP K17/SDS layered structure on the surface of probucol. For the GM of probucol/PVP K12/SDS, agglomeration of primary drug nanoparticles occurred more rapidly because of the insufficient surface coverage of PVP K12 and SDS on the probucol surface. Stabilization of the nanosuspension was linked to absorption of PVP K12 on the surface of probucol nanocrystals, owing to the absence of layered structure. Despite the proven importance of stabilizers in preventing particle agglomeration, there have been a few studies that generated stable nanosuspensions without stabilizers. Baba et al. prepared 2-devinyl2-(1-hexyloxyethyl)pyropheophorbide (HPPH) nanosuspensions without any stabilizer and reported formulation stability for more than 3 months [103]. The self-stabilization of the nanosuspensions was attributed to a high ZP value (− 40 mv) resulting from the deprotonation of the carboxylic end group of HPPH molecules. A similar self-stabilized nanosuspension was reported in another study in which amorphous all-trans retinoic acid nanoparticles were shown to be stable in aqueous medium up to 6 months. Epstein et al. [95] prepared self-suspended alendronate nanosuspensions by combining the negative charged alendromnic acid with gallium (Ga) or

Fig. 4. Location of shear plane in an electrostatic stabilized system (a) and in a combined steric-electrostatic stabilized system (b). Reprinted from Ref. [30] with permission from ELSEVIER.

L. Wu et al. / Advanced Drug Delivery Reviews 63 (2011) 456–469

463

Fig. 5. Schematic overview of agglomeration/stabilization mechanism of probucol/PVP/SDS ternary ground mixture after dispersion into water. Reprinted from Ref. [92] with permission from ELSEVIER.

gadolinium (Gd) under sonication as complex nanoparticles. The alendronate-Ga nanosuspension was shown to be stable for more than 3 months, while the alendronate-Gd nanosuspension was stable for only 3 days. These stability profiles correlated well with their ZP values (33 mV for Ga complex vs. 21 mV for Gd complex). 2.2.3. Crystal growth Crystal growth in colloidal suspensions is generally known as Ostwald ripening and is responsible for changes in particle size and size distribution. Oswald ripening is originated from particles solubility dependence on their size. Small particles have higher saturation solubility than larger ones according to Ostwald–Freundlich equation [115], creating a drug concentration gradient between the small and large particles. As a consequence, molecules diffuse from the higher concentration surrounding small particles to areas around larger particles with lower drug concentration. This generates supersaturated solution around the large particles, leading to drug crystallization onto the large particles. This diffusion process leaves an unsaturated solution surrounding the small particles, causing dissolution of the drug molecules from the small particles into the bulk medium. This diffusion process continues until all the small particles are dissolved. The Ostwald ripening is essentially a process where large particles grow at the expense of smaller particles [36,37], which subsequently leads to a shift in the particle size and size distribution of the colloidal suspension to a higher range. The diffusion and crystal growth during Ostwald ripening is shown schematically in Fig. 6. A narrow particle size distribution can minimize the saturation solubility difference and drug concentration gradients within the medium, and thus help to inhibit occurrence of the Ostwald ripening [37]. This can perhaps explain why Ostwald ripening is not a major concern for nanosuspensions with uniform particle size [10,20]. Stabilizers may also alleviate Ostwald ripening as long as they do not enhance the drug solubility [116,117]. Being absorbed on the

nanoparticles surface, the stabilizers can reduce the interfacial tension between the solid particles and liquid medium, and thus preventing the Ostwald ripening. Solubility, temperature, and mechanical agitation also affect Ostwald ripening [117]. Verma et al. produced ibuprofen nanosuspensions by microfluidization milling with the aid of various stabilizers (HPMC, Pluronic® F68 & F127, Kollidon 30, SLS) [31]. The particle size obtained with microfluidization showed some correlation with the ibuprofen solubility in aqueous stabilizer solutions. A higher solubility of ibuprofen in the solution of SLS, Tween 80 and Pluronic® F127 resulted in larger particles due to Ostwald ripening that occurred during process. A similar correlation was observed for ibuprofen particles during storage where Ostwald ripening was also believed to be the driving factor for formation of larger particles. Van Eerdenbrugh et al. demonstrated that Ostwald ripening was highly dependent on temperature by exploring TPGS stabilized nanosuspensions for 9 different drug candidates [86]. Following 3 months storage at room temperature, Ostwald ripening occurred in 8 out of 9 nanosuspensions studied. Enhanced Ostwald ripening was observed at 40 °C storage, while lowering temperature to 4 °C slowed down or even stopped Ostwald ripening effects. 2.2.4. Change of crystalline state Crystalline state is one of the most important parameters affecting drug stability, solubility, dissolution and efficacy. The main issue with crystalline state change is the transformation between amorphous and crystalline state. The high energy top-down manufacturing techniques tend to create partially amorphous nanosuspensions and some bottom-up techniques can create completely amorphous particles. The high energy amorphous particles are unstable and inclined to convert to low energy crystalline state over time. This conversion occurs depending on different parameters, such as temperature, dispersion medium, stabilizers and the presence of crystalline particles. Lindfors et al. produced Felodipine amorphous

464

L. Wu et al. / Advanced Drug Delivery Reviews 63 (2011) 456–469

Fig. 6. Schematic illustration of Ostwald ripening.

nanosuspensions via anti-solvent precipitation under sonication [96]. They demonstrated that amorphous nanoparticles were highly unstable in the presence of small amounts of crystalline particles. This was attributed to saturation solubility differences between amorphous and crystalline nanoparticles that initiated a similar diffusion process to Ostwald ripening, leading to a rapid conversion of amorphous nanoparticles to crystalline state. Although most of amorphous particles have been shown to be unstable, a few amorphous nanosuspensions have been demonstrated to be stable over a certain period of time. Amorphous hydrocortisone nanosuspensions, produced through a bottom-up nanoprecipitation technique using microfluidic reactors, was found to remain stable after 3 months storage at room temperature [93]. Amorphous all-trans retinoic acid nanosuspensions, prepared by an anti-solvent precipitation technique, were also shown to be stable over 6 months storage at 4 °C [102]. Manufacturing process might also induce some other type of crystalline transformation. Lai et al. prepared the diclofenac acid (DCF) nanosuspensions by HPH with two different crystalline forms (DCF1 and DCF2) [60]. 5 w/w% Pluronic® F68 was used as a stabilizer. XRD analysis showed that these two crystalline forms belonged to the same polymorph with differences in molecular conformation and crystal size. It was demonstrated that the HPH process caused the partial transformation of DCF2 to DCF1 while no effect on DCF1 was observed. The change in the crystalline structure was attributed to the solubilization of DCF2 during HPH process and its subsequent recrystallization as the DCF1 form.

chemical stability of nanosuspensions is generally superior to that of solutions. Paclitaxel serves as a good example to illustrate this [119]. Fig. 7(a) shows an HPLC diagram of paclitaxel nanosuspensions stabilized with Pluronic® F68 after 4 years of storage at 4–8 °C. No visible degradation product was observed with a recovery of more than 99%. On the other hand, paclitaxel solution with methanol as cosolvent showed clear degradation only after 48 h at room temperature (Fig. 7(b)). The excellent chemical stability of paclitaxel nanosuspensions was attributed to a mechanism similar to oxidized layer on the aluminum surface. Monolayer degradation on the nanocrystals surface was created once they were exposed to water and oxygen, as illustrated in Fig. 7(c). This monolayer could protect the inner part of drug crystals from further degradation, and thus enhance chemical stability of the nanosuspensions. Unlike the physical stability issue that is a common concern for nanosuspensions, chemical stability is drug specific. Each molecule has its particular functional groups and reaction mechanism that affect the stability. For example, chemical functionalities, such as ester and amides, are susceptible to hydrolytic degradation, while amino groups may undergo oxidative degradation [120]. Although chemical stability of nanosuspensions is usually not a major concern, extra attention should be paid to drug molecules with solubility greater than 1 mg/mL or with low concentration in suspension [120]. The common strategy to enhance the chemical stability is to transform the nanosupensions into dry solid dosage form which is much more stable than nanosuspensions [19] or to increase the concentration of the nanosuspensions [120].

2.2.5. Stability issues with solidification process of nanosuspensions When stable nanosuspensions are unattainable, the solid dosage form is the ultimate solution. The most common solidification processes are freeze drying and spray drying [10,19,20,118]. Since most solidified nanoparticle dry powders are usually reconstituted back into nanosuspensions during administration, drug nanocrystal growth or agglomeration during drying process needs to be prevented in order to maintain the nanosizing features such as rapid dissolution following the reconstitution. Adding matrix formers, such as mannitol, sucrose and cellulose, into nanosuspensions prior to drying is the common approach to overcome the stability issues during solidification process [19]. Since several excellent reviews have been published on this topic [19,25,118], the readers are directed to those reviews for more details.

2.3. Additional stability issues relate to large biomolecules

2.2.6. Chemical stability Since drug nanocrystals are usually dispersed in nanosuspensions with a limited solubility, the possibility of chemical reactions is not as substantial as that in solution-based formulations. Consequently,

Large biomolecules discussed in this review are mainly referred to therapeutic protein and peptide. The molecular structure of protein/ peptide is distinctly different and more complicated as compared to that of the small molecules. The structures of large molecules are generally differentiated into four structures, i.e. primary, secondary, tertiary and quaternary structures [34]. These different structures refer to the sequence of the different amino acids, regions where the chains are organized into regular local structures by hydrogen bonding such as alpha helix and beta sheet, the mechanisms on how the protein/peptide chain folds into a 3-dimensional conformation, and the composition of multiple protein/peptide molecules assembly, respectively [32,33,123]. The intact molecular structure of protein/peptide is essential to maintain their therapeutic efficacy [35,121]. Common stability issues associated with protein/peptide include deamidation, oxidation, acylation, unfolding, aggregation and adsorption to surfaces [35,121]. These stability issues are affected by

L. Wu et al. / Advanced Drug Delivery Reviews 63 (2011) 456–469

465

Fig. 7. (a) HPLC diagram of paclitaxel aqueous nanosuspensions stabilized with Pluronic® F68; (b) HPLC diagram of paclitaxel solution (methanol: 10 ml, water: 5 ml, paclitaxel: 20.8 mg); (c) Schematic illustration of the stabilization mechanism of paclitaxel nanosuspensions. Reprinted from Ref. [119] with permission from ELSEVIER.

temperature, solution pH, buffer ion, salt concentration, protein concentration, and added surfactants, with solution formulations being more susceptible to the influence from these factors than the suspension formulations [34,35,121]. Although suspension formulations or solid state of protein/peptide have enhanced stability due to their reduced molecular mobility, other stability issues may arise during particle formation or formulation process. For example, irreversible denaturation and aggregation upon reconstitution were often observed for dehydrated protein through freeze drying or spraydrying [125,126]. To prevent this, supplementary excipients such as bulking agents or surfactants are usually introduced during lyophilization [122]. The vulnerable structure of protein/peptide creates challenges for formulation development. Instead of using “naked” protein, the common strategy to prevent protein/peptide denaturation is to encapsulate the biomolecules with carrier such as liposome [123], SLN [124] or polymeric materials [125,126]. In addition to improving the stability, protein/peptide encapsulation can enhance bioavailability and provide sustained therapeutic release [125–128]. There has been plenty of work reported on encapsulated protein/peptide nanoparticles but very scarce studies on pure protein/peptide nanoparticles. Gomez et al. produced bovine zinc insulin nanoparticles using an electrospray drying technique and reported retained biological activities of the particles [129]. By using HPH, Maschke et al. attempted to micronize insulin in the medium of Myglyol 812 [130]. The stability and bioactivity of the insulin were maintained in spite of the harsh HPH process conditions. Merisko-Liversidge et al. [83] also noticed retained stability and bioactivity of zinc-insulin nanosuspensions that were produced through a wet milling process in presence of Pluronic® F68 and sodium deoxycholate. Nyambura et al. utilized a bottom up technique (combination of emulsification and freeze drying) to generate insulin nanoparticles (80 w/w% insulin with 20 wt.% lactose) [110]. The particles were then dispersed into HFA134a to produce an MDI formulation. The molecular integrity of insulin formulation, measured by HPLC, size exclusion chromatography, circular dichroism and fluorescence spectroscopy, indicated that native structures (primary, secondary and tertiary) were retained after particle formation and formulation process. The presence of surfactant (lecithin) and lyoprotectant (lactose) was believed to be responsible for preservation of the insulin structures. In their follow up work [109], they applied a similar approach to produce composite

nanoparticles of lysozyme and lactose for MDI formulations. The retained biological activity of lysozyme was enhanced with increasing lactose concentration in the particles, and reached maximum (99% retained activity) with 20 w/w% lactose. Nanoprecipitation coupled with freeze drying was used as well in this work to produce spherical nanoparticles containing 80 w/w% lysozyme with fully preserved bioactivity. It was demonstrated that bioactivity of lysozyme nanoparticles remained unchanged when in contact with HFA 134a. Yu et al. compared the effectiveness of spray freezing into liquid (SFL) and spray-freeze drying (SFD) processes in producing bioactive lysozyme particles [131]. Both processes generated highly porous micro-sized aggregates of lysozyme primary nanoparticles in the size of 100–300 nm. SFL process was shown to produce lysozyme with lower aggregation and higher enzyme activity as compared to the SFD process, which was attributed to the shorter exposure time to the air– water interface during the SFL atomization process. 3. Characterizing stability of drug nanoparticles and nanoparticle formulations Selection of characterization techniques for drug nanoparticles stability is dependent on the nature of stability issues and product dosage form. A few commonly used stability characterization techniques are listed in Table 2. 3.1. Particle size, size distribution and morphology Particle size and size distribution are the key parameters used for evaluating the physical stability of nanoparticles. A variety of techniques, including photon correlation spectroscopy (PCS), also known as dynamic light scattering (DLS), laser diffraction (LD) and coulter counter, are commonly used to measure the particle size and size distribution (Table 2). The PCS/DLS is widely used to determine the size and size distribution of small particles suspended in liquid medium. The mean particle size and size distribution indicated as polydispersity index (PDI) are the typical measured parameters of this technique. A PDI value of 0.1 to 0.25 indicates a narrow size distribution while a PDI greater than 0.5 refers to a broad distribution [20]. Unfortunately, this technique is not capable of measuring size of dry powders and its measurement range is too narrow (3 nm to 3 μm) to detect the interference from the microparticles (N3 μm) within the

466

L. Wu et al. / Advanced Drug Delivery Reviews 63 (2011) 456–469

Table 2 Commonly used technique to evaluate the stability of nanoparticles. Measured parameters

Techniques

Remarks

Particle size and size distribution

PCS/DLS

Pros: rapid, non-invasive. Cons: limited measurement range; apply only to liquid suspension. Pros: wide measurement range, rapid, non-invasive, apply to both liquid suspension and dry powder samples. Cons: particles are assumed to be spherical. Pros: precise. Cons: apply only to spherical particles. Pros: evaluate both particle morphology and size, very small quantity of sample required. Cons: challenging to acquire statistical size distribution, usually invasive, time-consuming. Pros: non-invasive, evaluate both particle morphology and size, very small quantity of sample required. Cons: challenging to acquire statistical size distribution, time-consuming. –

Laser diffraction

Coulter counter Particle size and morphology

SEM/TEM AFM

Sedimentation/creaming Particle surface charge/zeta potential Crystallinity state Chemical stability

Visual observation/laser backscattering/ near infrared transmission Laser Doppler electrophoresis XRD/DSC HPLC/FTIR/NMR/MS

– – –

nanosuspensions. Therefore, LD is often used in combination with PCS to circumvent this issue. Laser diffraction has a much wider detection range (20 nm to 2000 μm) and it can be used to evaluate both suspension and dry powder samples. The typical LD characterization parameters are LD50, LD90 and LD99, indicating 50, 90 or 99% of the particles are below the given size, respectively. LD is especially suitable for characterizing parenteral and pulmonary suspensions due to it wide measurement range. LD can detect the presence of microparticles (N5 μm) which are detrimental to parenteral nanosuspensions. However, LD provides only relative size distribution. The Coulter counter, on the other hand, measures the absolute number of particles per volume unit for the different size classes, and is more precise than the LD. Although PCS, LD and coulter counter techniques provide rapid measurement of particle size and size distribution, they do not have the capability in evaluating particle morphology. As direct visualization techniques, Scanning Electron Microscope (SEM), Transmission Electron Microscope (TEM) and Atomic Force Microscope (AFM) are widely used for assessment of particle morphology. However, it is very challenging and time-consuming to measure a significant number of particles to achieve statistical size distribution using these techniques. In addition, they usually require additional sample preparation such as coating that could be invasive to the particles, potentially causing some changes in particle properties. 3.2. Sedimentation/creaming The traditional method to evaluate sedimentation or creaming is by visual observation over a period of time. By measuring the volume of the settled or creamed particle layer relative to the total suspension volume within a specific time, a dimensionless parameter known as sedimentation or flocculation volume can be obtained as a quantitative evaluation of suspension stability. A higher flocculation volume indicates a more stable suspension. The structure of settled/creamed layer can be easily assessed by re-dispersing the suspension, i.e. easily re-dispersed suspension indicates loose flocs while a dense cake is hard to be broken by manual shaking. Other approaches to evaluate sedimentation/creaming include laser backscattering [132] and nearinfrared transmission [133]. 3.3. Particle surface charge Laser Doppler electrophoresis is commonly used to measure ZP. This technique evaluates electrophoretic mobility of suspended particles in the medium. It is a general rule of thumb that an absolute

value of ZP above 60 mV yields excellent stability, while 30, 20 and less than 5 mV generally results in good stability, acceptable shortterm stability and fast particle aggregation, respectively [29]. This rule of thumb is only valid for pure electrostatic stabilization or in combination with low-molecular weight surfactants, and is not valid when high molecular weight stabilizers are present [29].

3.4. Crystalline state The crystallinity of drug nanoparticles is usually assessed by X-Ray Diffraction (XRD) and/or Differential Scanning Calorimetry (DSC). XRD differentiates amorphous and crystalline nanoparticles as well as different polymorphic phases of the particles, while DSC is often used as a supplementary tool to XRD. Crystalline particles usually have a sharp melting peak which is absent in amorphous materials. The melting point can also be utilized to differentiate different polymorphs.

3.5. Chemical stability HPLC is the most common characterization technique used to evaluate chemical stability that provides precise quantitative analysis on the degradation impurities. Mass spectrometry (MS) is often coupled with HPLC to identify the molecular structure of impurities. Some other techniques such as FTIR and NMR can also be used for chemical stability assessment. However, they are not as precise and sensitive as HPLC, and thus not widely used for stability assessment.

3.6. Additional techniques for assessing large biomolecule nanoparticle and formulation stability For large biomolecules, additional characterization tools are generally required depending on the level of molecular structure to be assessed. For instance, size exclusion chromatography and electrophoresis are used to evaluate the primary structure of large biomolecules, circular dichroism is to monitor the secondary and tertiary structures while fluorescence spectroscopy is for tertiary structure [34,134]. In addition, in-vitro bioassays or in-vivo efficacy tests are needed to evaluate biological activities of the large biomolecules. Insulin particles, as an example, have been tested for its bioactivity either by in-vitro chondrocyte culture assays [130] or invivo monitoring of blood glucose level on rats following insulin administration [83].

L. Wu et al. / Advanced Drug Delivery Reviews 63 (2011) 456–469

4. Recommendations of general strategies for enhancing stability of nanoparticle formulations Strategies to address different stability issues are usually tailored according to different aspects, such as therapeutic requirements, dosage form and manufacturing complexity. For example, as the particle size is reduced, the sedimentation rate is decreased so that the particles can stay suspended longer in nanosuspensions. The general wisdom is that the smaller the nanoparticles are, the better. Unfortunately, too small particles are not always desirable, as they may create undesired plasma peaks due to the significant increases in dissolution rate [28]. Moreover, manufacturing complexity may be increased as well when the particles size requirements become too stringent. The use of stabilizers is the most commonly used technique in achieving a stable nanoparticle formulation. However, the stabilizer selection is known to be very challenging. The challenge stems mainly from two aspects: (i) lack of fundamental understanding of interactions within nanosuspensions and (ii) lack of an efficient and high throughput stabilizer screening technique. In the case of aqueous nanosuspensions, it is relatively easy to select stabilizers given that water-based stabilizing moieties such as PEG and PVA are well known. However, selecting the anchor groups that interact strongly with the drug surface can be challenging due to the limited understanding on interactions between nanoparticles and stabilizers in molecular level. For non-aqueous nanosuspensions such as HFA-based MDI delivery system, understanding of solvation in the low-dielectric HFA medium is still in its infancy, which makes stabilizers selection even more challenging. Inefficient screening approaches are another hurdle for stabilizer selection. The current practice for stabilizer screening involves trial production of nanosuspensions with different stabilizers or stabilizer combinations, which could be burdensome and require vast amount of efforts especially with a large number of potential stabilizer candidates. AFM has recently been proven to be a feasible and efficient tool for stabilizer screening. Verma et al. demonstrated the feasibility of using AFM to select stabilizers for Ibuprofen nanosuspensions [135]. The AFM measurements showed that HPMC and HPC had extensive surface absorption on the ibuprofen surface, as opposed to the inadequate surface absorption with PVP and Pluronic® surfactants. These results correlated well with their stabilizing performances in the nanosuspensions. This finding confirmed the significance of AFM in providing a scientific rationale for stabilizer selection and improving understanding of the stabilization mechanisms. Another technique, known as colloidal probe microscopy (CPM) which is derived from AFM, has also been widely used to study interactions between colloidal particles and is expected to be a useful tool for nanosuspension stabilizer screening [136]. Due to the significant challenges associated with stabilizer selection, self-stabilized nanosuspensions with no added stabilizer are highly desirable. This is not only for simplifying the formulation development process but also reducing stabilizer-based toxicity. Unfortunately, the challenges to engineer such self-suspended nanoparticles are tremendous with very few reported studies to date. A couple of approaches that could potentially be used to produce self-stabilized nanosuspensions include the creation of drug nanoparticles with high ZP and controlling morphology or surface properties of drug nanoparticles to minimize inter-particulate forces. 5. Conclusions The stability of drug nanoparticles remains a very challenging issue during pharmaceutical product development. Stability is affected by various factors such as dosage form (nanosuspension vs. dry solid), dispersion medium (aqueous vs. non-aqueous), delivery route (oral, inhalation, IV or other routes), production technique (topdown vs. bottom-up) and nature of drug (small molecules vs. large

467

biomolecules). Despite the significant challenges associated with stabilizer screening, adding a stabilizer or combination of stabilizers is still the most commonly used and preferred approach to enhance the stability of nanosuspensions. Further understanding of particle– particle interactions within nanosuspensions and development of high-throughput stabilizer screening tools are essential to facilitate efficient stabilizer selection. Development of self-stabilized nanosuspensions, although currently seen as very complicated and challenging, is expected to grow with the continuing advancement in the field of particle engineering. References [1] K. Park, Nanotechnology: what it can do for drug delivery, J. Control. Release 120 (1–2) (2007) 1–3. [2] G.A. Hughes, Nanostructure-mediated drug delivery, Nanomed. Nanotechnol. Biol. Med. 1 (1) (2005) 22–30. [3] G.M. Whitesides, Nanoscience, nanotechnology, and chemistry, Small 1 (2) (2005) 172–179. [4] J. Panyam, V. Labhasetwar, Biodegradable nanoparticles for drug and gene delivery to cells and tissue, Adv. Drug Deliv. Rev. 55 (3) (2003) 329–347. [5] P. Couvreur, E. Fattal, P. Legrand, C. Vauthier, Nanocapsule technology: a review, Crit. Rev. Ther. Drug Carrier Syst. 19 (2) (2002) 99–134. [6] A. Bianco, K. Kostarelos, M. Prato, Applications of carbon nanotubes in drug delivery, Curr. Opin. Chem. Biol. 9 (6) (2005) 674–679. [7] G.S. Kwon, T. Okano, Polymeric micelles as new drug carriers, Adv. Drug Deliv. Rev. 21 (2) (1996) 107–116. [8] M.J. Lawrence, G.D. Rees, Microemulsion-based media as novel drug delivery systems, Adv. Drug Deliv. Rev. 45 (1) (2000) 89–121. [9] T.M. Allen, E.H. Moase, Therapeutic opportunities for targeted liposomal drug delivery, Adv. Drug Deliv. Rev. 21 (2) (1996) 117–133. [10] B.E. Rabinow, Nanosuspensions in drug delivery, Nat. Rev. 3 (2004) 785–796. [11] W. Yang, J.I. Peters, R.O. WilliamsIII, Inhaled nanoparticles — a current review, Int. J. Pharm. 356 (2008) 239–247. [12] C.M. Keck, R.H. Müller, Drug nanocrystals of poorly soluble drugs produced by high pressure homogenisation, Eur. J. Pharm. Biopharm. 62 (2006) 3–16. [13] L. Gao, D. Zhang, M. Chen, Drug nanocrystals for the formulation of poorly soluble drugs and its application as a potential drug delivery system, J. Nanopart. Res. 10 (2008) 845–862. [14] J.C. Sung, B.L. Pulliam, D.A. Edwards, Nanoparticles for drug delivery to the lungs, Trends Biotechnol. 25 (12) (2007) 563–570. [15] K.S. Soppimath, T.M. Aminabhavi, A.R. Kulkarni, W.E. Rudzinski, Biodegradable polymeric nanoparticles as drug delivery devices, J. Control. Release 70 (1–2) (2001) 1–20. [16] R.H. Muller, K. Mäder, S. Gohla, Solid lipid nanoparticles (SLN) for controlled drug delivery — a review of the state of the art, Eur. J. Pharm. Biopharm. 50 (1) (2000) 161–177. [17] S.A. Wissing, O. Kayser, R.H. Muller, Solid lipid nanoparticles for parenteral drug delivery, Adv. Drug Deliv. Rev. 56 (9) (2004) 1257–1272. [18] A. Chonn, P.R. Cullis, Recent advances in liposomal drug-delivery systems, Curr. Opin. Biotechnol. 6 (6) (1995) 698–708. [19] B.V. Eerdenbrugh, G.V.D. Mooter, P. Augustijns, Top-down production of drug nanocrystals Nanosuspension stabilization, miniaturization and transformation into solid products, Int. J. Pharm. 364 (2008) 64–75. [20] V.B. Patravale, A.A. Date, R.M. Kulkarni, Nanosuspensions a promising drug delivery strategy, J. Pharm. Pharmacol. 56 (2004) 827–840. [21] J.-I. Jinno, N. Kamada, M. Miyake, K. Yamada, T. Mukai, M. Odomi, H. Toguchi, G.G. Liversidge, K. Higaki, T. Kimura, Effect of particle size reduction on dissolution and oral absorption of a poorly water-soluble drug, cilostazol, in beagle do, J. Control. Release 111 (2006) 56–64. [22] O. Kayser, C. Olbrich, V. Yardley, A.F. Kiderlen, S.L. Croft, Formulation of amphotericin B as nanosuspension for oral administration, Int. J. Pharm. 254 (2003) 73–75. [23] M.P. Kumar, Y.M. Rao, S. Apte, Formultion of nanosuspensions of albendazole for oral administration, Curr. Nanosci. 4 (2008) 53–58. [24] N. Blagden, M.D. Matas, P.T. Gavan, P. York, Crystal engineering of active pharmaceutical ingredients to improve solubility and dissolution rates, Adv. Drug Deliv. Rev. 59 (2007) 617–630. [25] W. Abdelwahed, G. Degobert, S. Stainmesse, H. Fessi, Freeze-drying of nanoparticles formulation, process and storage considerations, Adv. Drug Deliv. Rev. 58 (2006) 1688–1713. [26] J.E. Kipp, The role of solid nanoparticle technology in the parenteral delivery of poorly water-soluble drugs, Int. J. Pharm. 284 (2004) 109–122. [27] H.C. Fischer, W.C.W. Chan, Nanotoxicity: the growing need for in vivo study, Curr. Opin. Biotechnol. 18 (6) (2007) 565–571. [28] M. Radtke, Pure drug nanoparticles for the formulation of poorly soluble drugs, New Drugs 3 (62–68) (2001). [29] P.R. Mishra, L.A. Shaal, R.H. Müller, C.M. Keck, Production and characterization of Hesperetin nanosuspensions for dermal delivery, Int. J. Pharm. 371 (2009) 182–189. [30] J. Pardeike, R.H. Müllera, Nanosuspensions a promising formulation for the new phospholipase A2 inhibitor PX-18, Int. J. Pharm. 391 (2010) 322–329.

468

L. Wu et al. / Advanced Drug Delivery Reviews 63 (2011) 456–469

[31] S. Verma, R. Gokhale, D.J. Burgessa, A comparative study of top-down and bottom-up approaches for the preparation of micro/nanosuspensions, Int. J. Pharm. 380 (2009) 216–222. [32] S.K. Murthy, Nanoparticles in modern medicine state of the art, Int. J. Nanomed. 2 (2) (2007) 129–141. [33] P. Rogueda, Novel hydrofluoroalkane suspension formulations for respiratory drug delivery, Expert Opin. Drug Deliv. 2 (4) (2005) 625–638. [34] A.K. Banga, Structure and analysis of therapeutic peptides and proteins, Therapeutic peptides and proteins: formulation, processing, and delivery systems, second Edition, Taylor & Francis, 2005, pp. 33–66. [35] S. Koppenol, Physical considerations in protein and peptide stability, in: E.J. McNally, J.E. Hastedt (Eds.), Protein formulation and delivery, second edition, Informa Healthcare, 2007, pp. 43–72. [36] C.-J. Kim, Surface chemistry and colloids, Advanced pharmaceutics: physicochemical principles, CRC Press, 2004, pp. 193–256. [37] M.T.H. Nutan, I.K. Reddy, General principles of suspensions, in: A.K. Kulshreshtha, O.N. Singh, G.M. Wall (Eds.), Pharmaceutical suspensions: from formulation development to Manufacturing, Springer, 2009, pp. 39–66. [38] Y. Ali, K. Lehmussaari, Industrial perspective in ocular drug delivery, Adv. Drug Deliv. Rev. 58 (11) (2006) 1258–1268. [39] A.T. Florence, D. Attwood, Emulsions, suspensions and other dispersed systems, Physicochemical principles of pharmacy, 4th Edition, Pharmaceutical Press, 2005, pp. 229–272. [40] J.M. Tam, J.D. Engstrom, D. Ferrer, R.O. Williams III, K.P. Johnston, Templated open flocs of anisotropic particles for pulmonary delivery with pressurized metered dose inhalers J, Pharm. Sci. 99 (7) (2010) 3150–3165. [41] J.D. Engstrom, J.M. Tam, M.A. Miller, R.O. Williams III, K.P. Johnston, Templated open flocs of nanorods for enhanced pulmonary delivery with pressurized metered dose inhalers, Pharm. Res. 26 (1) (2009) 101–117. [42] L.A. Dellamary, T.E. Tarara, D.J. Smith, C.H. Woelk, A. Adractas, M.L. Costello, H. Gill, J.G. Weers, Hollow porous particles in metered dose inhalers, Pharm. Res. 17 (2) (2000) 168–174. [43] B.V. Derjaguin, L. Landau, Theory of the stability of strongly charged lyophobic sols and of the adhesion of strongly charged particles in solutions of electrolytes, Acta Phys. Chim. 14 (1941) 633–662. [44] E.J.W. Verwey, Theory of the stability of lyophobic colloids, J. Phys. Chem. 51 (3) (1947) 631–636. [45] J.-Y. Choi, J.Y. Yoo, H.-S. Kwak, B.U. Nam, J. Lee, Role of polymeric stabilizers for drug nanocrystal dispersions, Curr. Appl Phys. 5 (2005) 472–474. [46] M.J. Choi, S. Briancon, J. Andrieu, S.G. Min, H. Fessi, Effect of freeze-drying process conditions on the stability of nanoparticles, Drying Technol. 22 (1&2) (2004) 335–346. [47] B.V. Eerdenbrugh, J. Vermant, J.A. Martens, L. Froyen, J.V. Humbeeck, P. Auguestijns, G.V.D. Mooter, A screening study of surface stabilization during the production of drug nanocrystals, J. Pharm. Sci. 98 (6) (2009) 2091–2103. [48] J. Lee, J.-Y. Choi, C.H. Park, Characteristics of polymers enabling nanocomminution of water-insoluble drugs, Int. J. Pharm. 355 (2008) 328–336. [49] J. Lee, S.-J. Lee, J.-Y. Choi, J.Y. Yooa, C.-H. Ahn, Amphiphilic amino acid copolymers as stabilizers for the preparation of nanocrystal dispersion, Eur. J. Pharm. Sci. 24 (2005) 441–449. [50] J. Deng, L. Huang, F. Liu, Understanding the structure and stability of paclitaxel nanocrystals, Int. J. Pharm. 390 (2010) 242–249. [51] M.J. Rosen, Dispersion and aggregation of solid in liquid medium by surfactants, Surfactants and Interfacial Phenomena, John Wiley & Sons, 2004, pp. 337–362. [52] A. Makhlof, Y. Miyazaki, Y. Tozuka, H. Takeuchi, Cyclodextrins as stabilizers for the preparation of drug nanocrystals by the emulsion solvent diffusion method, Int. J. Pharm. 357 (2008) 280–285. [53] H. Steckel, S. Wehle, A novel formulation technique for metered dose inhaler (MDI) suspensions, Int. J. Pharm. 284 (1–2) (2004) 75–82. [54] D.A. Edwards, J. Hanes, G. Caponetti, J. Hrkach, A. Ben-Jebria, M.L. Eskew, J. Mintzes, D. Deaver, N. Lotan, R. Langer, Large porous particles for pulmonary drug delivery, Science 276 (5320) (1997) 1868–1872. [55] L. Gao, D. Zhang, M. Chen, T. Zheng, S. Wang, Preparation and characterization of an oridonin nanosuspension for solubility and dissolution velocity enhancement, Drug Dev. Ind. Pharm. 33 (2007) 1332–1339. [56] L. Gao, D. Zhang, M. Chen, C. Duan, W. Dai, L. Jia, W. Zhao, Studies on pharmacokinetics and tissue distribution of oridonin nanosuspensions, Int. J. Pharm. 355 (2008) 321–327. [57] C. Jacobs, R.H. Müller, Production and characterization of a budesonide nanosuspension for pulmonary administration, Pharm. Res. 19 (2) (2002) 189–194. [58] N. Hernández-Trejo, O. Kayser, H. Steckel, R.H. Muller, Characterization of nebulized buparvaquone nanosuspensions — effect of nebulization technology, J. Drug Target. 13 (8–9) (2005) 499–507. [59] R.H. Muller, C. Jacobs, Buparvaquone mucoadhesive nanosuspension: preparation, optimisation and long-term stability, Int. J. Pharm. 237 (2002) 151–161. [60] F. Lai, C. Sinico, G. Ennas, F. Marongiu, G. Marongiu, A.M. Fadd, Diclofenac nanosuspensions influence of preparation procedure and crystal form on drug dissolution behaviour, Int. J. Pharm. 373 (2009) 124–132. [61] D. Zhang, T. Tan, L. Gao, W. Zhao, P. Wang, Preparation of azithromycin nanosuspensions by high pressure homogenization and its physicochemical characteristics studies, Drug Dev. Ind. Pharm. 33 (2007) 569–575. [62] R. Mauludin, R.H. Muller, C.M. Keck, Development of an oral rutin nanocrystal formulation, Int. J. Pharm. 370 (2009) 202–209. [63] R. Mauludin, R.H. Muller, C.M. Keck, Kinetic solubility and dissolution velocity of rutin nanocrystals, Eur. J. Pharm. Sci. 36 (2009) 502–510.

[64] C. Jacobs, O. Kayser, R.H. Muller, Nanosuspensions as a new approach for the formulation for the poorly soluble drug tarazepide, Int. J. Pharm. 196 (2000) 161–164. [65] J. Moschwitzer, G. Achleitner, H. Pomper, R.H. Muller, Development of an intravenously injectable chemically stable aqueous omeprazole formulation using nanosuspension technology, Eur. J. Pharm. Biopharm. 58 (2004) 615–619. [66] R. Xiong, W. Lu, J. Li, P. Wang, R. Xu, T. Chen, Preparation and characterization of intravenously injectable nimodipine nanosuspension, Int. J. Pharm. 350 (2008) 338–343. [67] K. Peters, R.H. Muller, D.Q.M. Craig, An investigation into the distribution of lecithins in nanosuspension systems using low frequency dielectric spectroscopy, Int. J. Pharm. 184 (1999) 53–61. [68] M.A. Kassem, A.A.A. Rahman, M.M. Ghorab, M.B. Ahmeda, R.M. Khalil, Nanosuspension as an ophthalmic delivery system for certain glucocorticoid drugs, Int. J. Pharm. 340 (2007) 126–133. [69] V. Teeranachaideekul, V.B. Junyaprasert, E.B. Souto, R.H. Muller, Development of ascorbyl palmitate nanocrystals applying the nanosuspension technology, Int. J. Pharm. 354 (2008) 227–234. [70] K.P. Krause, R.H. Muller, Production and characterisation of highly concentrated nanosuspensions by high pressure homogenisation, Int. J. Pharm. 214 (2001) 21–24. [71] J. Hecq, M. Deleers, D. Fanara, H. Vranckx, K. Amighi, Preparation and characterization of nanocrystals for solubility and dissolution rate enhancement of nifedipine, Int. J. Pharm. 299 (2005) 167–177. [72] J. Hecq, M. Deleers, D. Fanara, H. Vranckx, P. Boulanger, S.L. Lamer, K. Amighi, Preparation and in vitroin vivo evaluation of nano-sized crystals for dissolution rate enhancement of ucb-35440-3, a highly dosed poorly water-soluble weak base, Eur. J. Pharm. Biopharm. 64 (2006) 360–368. [73] Y.-X. Zhao, H.-Y. Hu, M. Chang, W.-J. Li, Y. Zhao, H.-M. Liu, Preparation and cytotoxic activity of hydroxycamptothecin nanosuspensions, Int. J. Pharm. 392 (2010) 64–71. [74] S. Ganta, J. Paxton, B.C. Baguley, S. Garg, Formulation and pharmacokinetic evaluation of an asulacrine nanocrystalline suspension for intravenous delivery, Int. J. Pharm. 367 (2009) 179–186. [75] R.H. Muller, K. Peters, Nanosuspensions for the formulation of poorly soluble drugs I. Preparation by a size-reduction technique, Int. J. Pharm. 160 (1998) 229–237. [76] M.J. Grau, O. Kayser, R.H. Muller, Nanosuspensions of poorly soluble drugs — reproducibility of small scale production, Int. J. Pharm. 196 (2000) 155–157. [77] Q. Wang, A.Y. Sun, J. Pardeike, R.H. Müller, A. Simonyi, G.Y. Sun, Neuroprotective effects of a nanocrystal formulation of sPLA2 inhibitor PX-18 in cerebral ischemia reperfusion in gerbils, Brain Res. 1285 (2009) 188–195. [78] Y. Wang, D. Zhang, Z. Liu, G. Liu, C. Duan, L. Jia, F. Feng, X. Zhang, Y. Shi, Q. Zhang, In vitro and in vivo evaluation of silybin nanosuspensions for oral and intravenous delivery, Nanotechnology 21 (2010) 1–12. [79] Y. Tanaka, M. Inkyo, R. Yumoto, J. Nagai, M. Takano, S. Nagata, Nanoparticulation of poorly water soluble drugs using a wet-mill process and physicochemical properties of the nanopowders, Chem. Pharm. Bull. 57 (10) (2009) 1050–1057. [80] J.Z. Yang, A.L. Young, P.-C. Chiang, A. Thurston, D.K. Pretzer, Fluticasone and budesonide nanosuspensions for pulmonary delivery preparation, characterization, and pharmacokinetic studies and pharmacokinetic studies, J. Pharm. Sci. 97 (11) (2008) 4869–4878. [81] A. Ain-Ai, P.K. Gupta, Effect of arginine hydrochloride and hydroxypropyl cellulose as stabilizers on the physical stability of high drug loading nanosuspensions of a poorly soluble compound, Int. J. Pharm. 351 (2008) 282–288. [82] B.V. Eerdenbrugh, L. Froyen, J.A. Martens, N. Blaton, P. Augustijns, M. Brewster, G. V.d. Mooter, Characterization of physico-chemical properties and pharmaceutical performance of sucrose co-freeze-dried solid nanoparticulate powders of the anti-HIV agent loviride prepared by media milling, Int. J. Pharm. 338 (2007) 198–206. [83] E. Merisko-Liversidge, S.L. McGurk, G.G. Liversidge, Insulin nanoparticles a novel formulation approach for poorly water soluble zn-Insulin, Pharm. Res. 21 (9) (2004) 1545–1553. [84] G.C. Na, H.J. Stevens, J.B.O. Yuan, N. Rajagopalan, Physical stability of ethyl diatrizoate nanocrystalline suspension in steam sterilization, Pharm. Res. 16 (4) (1999) 569–574. [85] B.V. Eerdenbrugh, L. Froyen, J.V. Humbeeck, J.A. Martens, P. Augustijns, G.V.D. Mooter, Alternative matrix formers for nanosuspension solidification dissolution performance and X-ray microanalysis as an evaluation tool for powder dispersion, Eur. J. Pharm. Sci. 35 (2008) 344–353. [86] B.V. Eerdenbrugh, L. Froyen, J.V. Humbeeck, J.A. Martens, P. Augustijns, G.V.D. Mooter, Drying of crystalline drug nanosuspensions-the importance of surface hydrophobicity on dissolution behavior upon redispersion, Eur. J. Pharm. Sci. 35 (2008) 127–135. [87] T.S. Wiedmann, DeCastro, R.W. Wood, Nebulization of nanocrystalsTM, Pharm. Res. 14 (1) (1997) 112–116. [88] L. Baert, G.v.W. Dries, M. François, A. Wouters, E. Basstanie, K. I., Development of a long-acting injectable formulation with nanoparticles of rilpivirine (TMC278) for HIV treatment, Eur. J. Pharm. Biopharm. 72 (2009) 502–508. [89] Z. Deng, S. Xu, S. Li, Understanding a relaxation behavior in a nanoparticle suspension for drug delivery applications, Int. J. Pharm. 351 (2008) 236–243. [90] E. Merisko-Liversidge, P. Sarpotdar, J. Bruno, S. Hajj, L. Wei, N. Peltier, J. Rake, J.M. Shaw, S. Pug, L. Polin, J. Jons, T. Corbett, E. Cooper, G.G. Liversidge, Formulation and antitumor activity evaluation of nanocrystalline suspensions of poorly soluble anticancer drugs, Pharm. Res. 13 (2) (1996) 272–278.

L. Wu et al. / Advanced Drug Delivery Reviews 63 (2011) 456–469 [91] J. Chingunpitak, S. Puttipipatkhachorn, P. Chavalitshewinkoon-Petmitr, Y. Tozuka, K. Moribe, K. Yamamoto, Formation, physical stability and in vitro antimalarial activity of dihydroartemisinin nanosuspensions obtained by cogrinding method, Drug Dev. Ind. Pharm. 34 (2008) 314–322. [92] A. Pongpeerapat, C. Wanawongthai, Y. Tozuka, K. Moribe, K. Yamamoto, Formation mechanism of colloidal nanoparticles obtained from probucol/PVP/ SDS ternary ground mixture, Int. J. Pharm. 352 (2008) 309–316. [93] H.S.M. Ali, P. York, N. Blagden, Preparation of hydrocortisone nanosuspension through a bottom-up nanoprecipitation technique using microfluidic reactors, Int. J. Pharm. 375 (2009) 107–113. [94] P. Kocbek, S. Baumgartner, J. Kristl, Preparation and evaluation of nanosuspensions for enhancing the dissolution of poorly soluble drugs, Int. J. Pharm. 312 (2006) 179–186. [95] H. Epstein, V. Berger, I. Levi, G. Eisenberg, N. Koroukhov, J. Gao, G. Golomb, Nanosuspensions of alendronate with gallium or gadolinium attenuate neointimal hyperplasia in rats, J. Control. Release 117 (2007) 322–332. [96] L. Lindfors, P. Skantze, U. Skantze, J. Westergren, U. Olsson, Amorphous drug nanosuspensions. 3. Particle dissolution and crystal growth, Langmuir 23 (19) (2007) 9866–9874. [97] X. Chen, M.E. Matteucci, C.Y. Lo, K.P. Johnston, R.O. Williams III, Flocculation of polymer stabilized nanocrystal suspensions to produce redispersible powders, Drug Dev. Ind. Pharm. 35 (2009) 283–296. [98] D. Douroumis, A. Fahr, Stable carbamazepine colloidal systems using the cosolvent technique, Eur. J. Pharm. Sci. 30 (2007) 367–374. [99] J.M. Tam, J.T. Mcconville, R.O. Williams III, K.P. Johnston, Amorphous cyclosporin nanodispersions for enhanced pulmonary deposition and dissolution, J. Pharm. Sci. 97 (11) (2008) 4915–4933. [100] K. Sigfridsson, S. Forsse'n, P. Holla¨nder, U. Skantze, J.d. Verdier, A formulation comparison, using a solution and different nanosuspensions of a poorly soluble compound, Eur. J. Pharm. Biopharm. 67 (2007) 540–547. [101] D. Douroumis, A. Fahr, Nano- and micro-particulate formulations of poorly water-soluble drugs by using a novel optimized technique, Eur. J. Pharm. Biopharm. 63 (2006) 173–175. [102] X. Zhang, Q. Xia, N. Gu, Preparation of all-trans retinoic acid nanosuspensions using a modified precipitation method, Drug Dev. Ind. Pharm. 32 (2006) 857–863. [103] K. Baba, H.E. Pudavar, I. Roy, T.Y. Ohulchanskyy, Y. Chen, R.K. Pandey, P.N. Prasad, New method for delivering a hydrophobic drug for photodynamic therapy using pure nanocrystal form of the drug, Mol. Pharm. 4 (2) (2007) 289–297. [104] D. Xia, P. Quan, H. Piao, H. Piao, S. Sun, Y. Yin, F. Cui, Preparation of stable nitrendipine nanosuspensions using the precipitation–ultrasonication method for enhancement of dissolution and oral bioavailability, Eur. J. Pharm. Sci. 40 (2010) 325–334. [105] A. Dolenc, J. Kristl, S.S. Baumgartner, O.P. Sek, Advantages of celecoxib nanosuspension formulation and transformation into tablets, Int. J. Pharm. 376 (2009) 204–212. [106] M. Trotta, R. Cavalli, D. Chirio, Griseofulvin nanosuspensions from triacetin-inwater emulsions. S.T.P, Pharm. Sci. 13 (6) (2003) 423–426. [107] M. Trotta, M. Gallarate, F. Pattarino, S. Morel, Emulsions containing partially water-miscible solvents for the preparation of drug nanosuspensions, J. Control. Release 76 (2001) 119–128. [108] M. Trotta, M. Gallarate, M.E. Carlotti, S. Morel, Preparation of griseofulvin nanoparticles from water-dilutable microemulsions, Int. J. Pharm. 254 (2003) 235–242. [109] B.K. Nyambura, I. Kellaway, K.M.G. Taylor, The processing of nanoparticles containing protein for suspension in hydrofluoroalkane propellants, Int. J. Pharm. 32 (2009) 140–146. [110] B.K. Nyambura, I. Kellaway, K.M.G. Taylor, Insulin nanoparticles stability and aerosolization from pressurized metered dose inhalers, Int. J. Pharm. 375 (2009) 114–122. [111] P.A. Dickinson, S.W. Howells, J.W. Kellaway, Novel nanoparticles for pulmonary drug administration, J. Drug Target. 9 (4) (2001) 295–302. [112] F.J. Ahmad, R.K. Khar, S. Sultana, A. Bhatnagar, Techniques to develop and characterize nanosized formulation for salbutamol sulfate, J. Mater. Sci. Mater. Med. 20 (2009) 71–76.

469

[113] B.L. Montalvo, Y. Pacheco, B.A. Sosa, D. V'elez, G. S'anchez, K. Griebenow, Formation of spherical protein nanoparticles without impacting protein integrity, Nanotechnology 19 (2008) 1–7. [114] H. Piao, N. Kamiya, A. Hirata, T. Fujii, M. Goto, A novel solid-in-oil nanosuspension for transdermal delivery of diclofenac sodium, Pharm. Res. 25 (4) (2007) 896–901. [115] J.T. Carstensen, Solubility, Advanced pharmaceutical solids, Marcel Dekker, 2001, pp. 27–50. [116] T.F. Tadros, Surfactans in pharmaceutical formulations, Applied Surfactants: Principles and Applications, Wiley-VCH, 2005, pp. 433–502. [117] D.J. McClements, Emulsion stability, Food Emulsions: Principles, Practices, and Techniques, second edition, CRC Press, 2004, pp. 185–234. [118] F. Kesisoglou, S. Panmai, Y. Wu, Nanosizing — oral formulation development and biopharmaceutical evaluation, Adv. Drug Deliv. Rev. 59 (7) (2007) 631–644. [119] R.H. Muller, C.M. Keck, Challenges and solutions for the delivery of biotech drugs — a review of drug nanocrystal technology and lipid nanoparticles, J. Biotechnol. 113 (1–3) (2004) 151–170. [120] S. Garad, J. Wang, Y. Joshi, R. Panicucci, Preclinical development for suspensions, in: A.K. Kulshreshtha, O.N. Singh, G.M. Wall (Eds.), Pharmaceutical Suspensions: From Formulation Development to Manufacturing, Springer, 2009, pp. 39–66. [121] P.M. Bummer, Chemical considerations in protein and peptide stability, in: E.J. McNally, J.E. Hastedt (Eds.), Protein Formulation and Delivery, second edition, Informa Healthcare, 2007, pp. 7–42. [122] L. Chang, D. Shepherd, J. Sun, D. Ouellette, K.L. Grant, X. Tang, M.J. Pikal, Mechanism of protein stabilization by sugars during freeze-drying and storage: native structure preservation, specific interaction, and/or immobilization in a glassy matrix? J. Pharm. Sci. 94 (7) (2005) 1427–1444. [123] V.P. Torchilin, Recent advances with liposomes as pharmaceutical carriers, Nat. Rev. Drug Discov. 4 (2) (2005) 145–160. [124] A.J. Almeida, S. Runge, R.H. Müller, Peptide-loaded solid lipid nanoparticles (SLN): influence of production parameters, Int. J. Pharm. 149 (2) (1997) 255–265. [125] P. Quellec, R. Gref, L. Perrin, E. Dellacherie, F. Sommer, J.M. Verbavatz, M.J. Alonso, Protein encapsulation within polyethylene glycol-coated nanospheres. I. Physicochemical characterization, J. Biomed. Mater. Res. 42 (1) (1998) 45–54. [126] D. Blanco, M.J. Alonso, Protein encapsulation and release from poly(lactide-coglycolide) microspheres: effect of the protein and polymer properties and of the co-encapsulation of surfactants, Eur. J. Pharm. Biopharm. 45 (3) (1998) 285–294. [127] Y. Waeckerle-Men, E.U.-v. Allmen, B. Gander, E. Scandella, E. Schlosser, G. Schmidtke, H.P. Merkle, M. Groettrup, Encapsulation of proteins and peptides into biodegradable poly(d,l-lactide-co-glycolide) microspheres prolongs and enhances antigen presentation by human dendritic cells, Vaccine 24 (11) (2006) 1847–1857. [128] G.E. Hildebrand, J.W. Tack, Microencapsulation of peptides and proteins, Int. J. Pharm. 196 (2) (2000) 173–176. [129] A. Gomez, D. Bingham, L.d. Juan, K. Tang, Production of protein nanoparticles by electrospray drying, J. Aerosol Sci. 29 (5–6) (1998) 561–574. [130] A. Maschke, N. Calı, B. Appel, J. Kiermaier, T. Blunk, A. Gopferich, Micronization of insulin by high pressure homogenization, Pharm. Res. 23 (9) (2006) 2220–2229. [131] Z. Yu, K.P. Johnston, R.O. Williams III, Spray freezing into liquid versus sprayfreeze drying: influence of atomization on protein aggregation and biological activity, Eur. J. Pharm. Sci. 27 (1) (2006) 9–18. [132] K.A. Johnson, Interfacial phenomena and phase behaviour in metered dose inhaler formulations, in: A.J. Hickey (Ed.), Editor Inhalation Aerosols: Physical and Biological Basis for Therapy, second, Informa Healthcare, 2007, pp. 347–372. [133] M. Kuentz, D. Röthlisberger, Rapid assessment of sedimentation stability in dispersions using near infrared transmission measurements during centrifugation and oscillatory rheology, Eur. J. Pharm. Biopharm. 56 (3) (2003) 355–361. [134] M. Baudys, S.W. Kim, Peptide and protein characterization, in: L. Hovgaard, S. Frokjaer, M.v.d. Weert (Eds.), Pharmaceutical formulation development of peptides and proteins, Taylor & Francis, 1999, pp. 41–69. [135] S. Verma, B.D. Huey, D.J. Burgess, Scanning probe microscopy method for nanosuspension stabilizer selection, Langmuir 25 (21) (2009) 12481–12487. [136] L. Wu, S.R.P. da Rocha, Biocompatible and biodegradable copolymer stabilizers for hydrofluoroalkane dispersions: a colloidal probe microscopy investigation, Langmuir 23 (24) (2007) 12104–12110.

Generation of Emulsions by Ultrasonic Cavitation A wide range of intermediate and consumer products, such as cosmetics and skin lotions, pharmaceutical ointments, varnishes, paints and lubricants and fuels are based wholly or in part of emulsions. Hielscher manufactures the world's largest industrial ultrasonic liquid processors for the efficient emulsifying of large volume streams in production plants. In the lab, the emulsification power of ultrasound has been known and applied for long. The video below shows the emulsification of oil (yellow) into water (red) by using a UP400S lab device. Systems consisting of several ultrasonic processors of up to 16,000 watts each, provide the capacity needed to translate this lab application into an efficient production method to obtain finely dispersed emulsions in continuous flow or in a batch - achieving results comparable to that of today's best high-pressure homogenizers available, such as the new orifice valve. In addition to this high efficiency in the continuous emulsification, Hielscher ultrasonic devices require very low maintenance and are very easy to operate and to clean. The ultrasound does actually support the cleaning and rinsing. The ultrasonic power is adjustable and can be adapted to particular products and emulsification requirements. Special flow cell reactors meeting the advanced CIP (clean-in-place) and SIP (sterilize-in-place) requirements are available, too. Emulsions are dispersions of two or more immiscible liquids. Highly intensive ultrasound supplies the power needed to disperse a liquid phase (dispersed phase) in small droplets in a second phase (continuous phase). In the dispersing zone, imploding cavitation bubbles cause intensive shock waves in the surrounding liquid and result in the formation of liquid jets of high liquid velocity. In order to stabilize the newly formed droplets of the disperse phase against coalescence, emulsifiers (surface active substances, surfactants) and stabilizers are added to the emulsion. As coalescence of the droplets after disruption influences the final droplet size distribution, efficiently stabilizing emulsifiers are used to maintain the final droplet size distribution at a level that is equal to the distribution immediately after the droplet disruption in the ultrasonic dispersing zone. Stabilizers actually lead to improved droplet disruption at constant energy density. Studies at oil in water (water phase) and water in oil (oil phase) emulsions have shown the correlation between the energy density and droplet size

Quick Link

Information Request Literature on Emulsifying Information Request your name your email (*required) your phone number your country product or area of interest Click here to request information! Wire & Cable Cleaner UIP1000 - Sonication Set Sonochemistry Twitter, Facebook, Google+

(e.g. Sauter diameter). There is a clear tendency for smaller droplet size at increasing energy density (click at right graphic). At appropriate energy density levels, ultrasound can well achieve a mean droplet sizes below 1 micron (microemulsion).

Bookmark

We would like to see you coming back. To bookmark this page, please press Crtl + D.

Nanometals:

formation and color by Luis M. Liz-Marzán

Metal nanoparticles are very attractive because of their size- and shape-dependent properties. From the plethora of existing procedures for the synthesis of metal nanoparticles, the most widely used wetchemical methods are briefly discussed, which are suitable for production of both spherical and anisometric (rod-like or prismatic) nanoparticles. The optical properties of these nanoparticles are spectacular and, therefore, have promoted a great deal of excitement during the last few decades. The basics of the origin of such optical properties are described and some of the theoretical methods accounting for them are briefly presented. Examples are shown of the color variations arising from changes in the composition, size, and shape of nanoparticles, as well as from the proximity of other metal nanoparticles.

Departamento de QuÌmica FÌsica, Universidade de Vigo, 36200 Vigo, Spain E-mail: [email protected] URL: webs.uvigo.es/coloides/nano

26

February 2004

Nanotechnology, nanoscience, nanostructures, nanoparticles… These are now some of the most widely used terms in materials science literature. But why are nanoscale materials and processes so attractive? From the point of view of the general public, nanotechnology appears to be the fabrication of miniature machines, which will be able to travel through the human body and repair damaged tissues, or supercomputers small enough to fit in a shirt pocket. However, nanostructured materials have potential applications in many more areas, such as biological detection, controlled drug delivery, low-threshold lasers, optical filters, and sensors, among others. In fact, it is relatively easy to find examples of the use of metal nanoparticles (maybe not deliberately) as decorative pigments since the time of the Romans, such as those contained in the glass of the famous Lycurgus Cup (4th century AD). The cup can still be seen at the British Museum1 and possesses the unique feature of changing color depending upon the light in which it is viewed. It appears green when viewed in reflected light, but looks red when a light is shone from inside and is transmitted through the glass. Analysis of the glass reveals that it contains a very small amount of tiny (~70 nm) metal crystals containing Ag and Au in an approximate molar ratio of 14:1. It is the presence of these nanocrystals that gives the Lycurgus Cup its special color display. It was not until 1857, however, that Michael Faraday reported a systematic study of the synthesis and colors of

ISSN:1369 7021 © Elsevier Ltd 2004

REVIEW FEATURE

colloidal gold2. Since that pioneering work, thousands of scientific papers have been published on the synthesis, modification, properties, and assembly of metal nanoparticles, using a wide variety of solvents and other substrates. All this has led not only to reliable procedures for the preparation of metal nanoparticles of basically any desired size and shape, but also to a deep understanding of many of the physico-chemical features that determine the characteristic behavior of these systems. One of the most interesting aspects of metal nanoparticles is that their optical properties depend strongly upon the particle size and shape. Bulk Au looks yellowish in reflected light, but thin Au films look blue in transmission. This characteristic blue color steadily changes to orange, through several tones of purple and red, as the particle size is reduced down to ~3 nm. These effects are the result of changes in the so-called surface plasmon resonance3, the frequency at which conduction electrons oscillate in response to the alternating electric field of incident electromagnetic radiation. However, only metals with free electrons (essentially Au, Ag, Cu, and the alkali metals) possess plasmon resonances in the visible spectrum, which give rise to such intense colors. Elongated nanoparticles (ellipsoids and nanorods) display two distinct plasmon bands related to transverse and longitudinal electron oscillations. The longitudinal oscillation is very sensitive to the aspect ratio of the particles4, so that slight deviations from spherical geometry can lead to impressive color changes. Apart from single particle properties, the environment in which the metal particles are dispersed is also of relevance to the optical properties5. The refractive index of the surrounding medium6, as well as the average distance between neighboring metal nanoparticles7, has been shown to influence the spectral features, as will be described below.

Synthesis of metal nanoparticles In this short review, we do not have the scope to describe all the existing methods for the preparation of metal nanoparticles. This section will be restricted, therefore, to some of the most widely used methods based on chemical reactions in solution (often termed ‘wet chemistry’) that yield metal nanoparticle colloids. Probably the most popular method of preparing Au nanospheres dispersed in water is the reduction of HAuCl4 in a boiling sodium citrate solution8,9. The formation of uniform

Au nanoparticles is revealed by a deep wine red color observed after ~10 minutes10. The average particle diameter can be tuned over quite a wide range (~10-100 nm) by varying the concentration ratio between the Au salt and sodium citrate9. However, for particles larger than 30 nm, deviation from a spherical shape is observed, as well as a larger polydispersity. The same procedure can be used to reduce an Ag salt, but particle size control is very limited. Citrate reduction has also been applied to the production of Pt colloids of much smaller particle sizes (2-4 nm), which can be grown further by hydrogen treatment11,12. Another procedure that has become extremely popular for Au nanoparticle synthesis is the two-phase reduction method developed by Schiffrin and coworkers13,14. Basically, HAuCl4 is dissolved in water and subsequently transported into toluene by means of tetraoctylammonium bromide (TOAB), which acts as a phase transfer agent. The toluene solution is then mixed and thoroughly stirred together with an aqueous solution of sodium borohydride (a strong reductant), in the presence of thioalkanes or aminoalkanes, which readily bind to the Au nanoparticles formed. Depending on the ratio of the Au salt and capping agent (thiol/amine), the particle size can be tuned to between ~1 nm and ~10 nm. Several refinements of the preparative procedure, including the development of analogous methods for the preparation of Ag particles, have been reported15,16. Murray and coworkers have enhanced the method’s popularity by offering an interesting and elegant alternative to the two-phase reduction method, which has opened a new field of preparative chemistry. They explored routes to functionalized monolayer-protected clusters by ligand place exchange reactions17,18. Several examples exist of the reduction of metal salts by organic solvents. Ethanol has been long used for the preparation of metal nanoparticles such as Pt, Pd, Au, or Rh (suitable for catalytic applications) in the presence of a protecting polymer, usually poly(vinyl pyrrolidone) or PVP19,20. Another important example is found in Figlarz’s polyol method, which consists of refluxing a solution of the metal precursor in ethylene glycol or larger polyols21,22. Xia and coworkers recently demonstrated that the polyol method can be applied to the production of Ag nanowires and nanoprisms by reducing AgNO3 with ethylene glycol in the presence of PVP23,24. Ag nanoparticles with high aspect ratios were only grown in the presence of (Pt) seeds formed in situ

February 2004

27

REVIEW FEATURE

prior to the addition of the Ag salt. The dimensions of the Ag nanowires can be controlled by varying the experimental conditions (temperature, seed concentration, ratio of Ag salt and PVP, etc.). On the other hand, Liz-Marzán and coworkers have reported the ability of N,N-dimethylformamide (DMF) to reduce Ag+ ions, so that stable spherical Ag nanoparticles can be synthesized using PVP as a stabilizer25. In addition, SiO2- and TiO2-coated nanoparticles26,27 can be produced by the same method, in the presence of aminopropyltrimethoxysilane and titanium tetrabutoxide, respectively. Interestingly, the shape (and size) of the nanoparticles obtained in this way depends on several parameters, such as Ag salt and stabilizer concentrations, temperature, and reaction time. Specifically, when PVP is used as a protecting agent, spherical nanoparticles form at low AgNO3 concentrations ( b = c), which are defined as

(4) (7) and the parameter r is related to the aspect ratio (r = √1 - (b/a)2). Using these equations, El-Sayed and coworkers46 derived an empirical relationship between the aspect ratio and the wavelength λmax of the longitudinal plasmon resonance:

where h is the film thickness, R is the reflectance at normal incidence, (8) (5)

These expressions predict very well the colors shown in Figs. 1d and 1e. For other geometries, Mie theory has not yet been properly implemented, but different approaches have been devised, such as the discrete dipole approximation, which has been applied to the calculation of the spectra of Ag nanoprisms34. Thin Films When the metal nanoparticle volume fraction is high, the equations above are no longer valid, since dipole-dipole interactions between neighboring nanoparticles are present,

30

February 2004

and α is the absorption coefficient, which can be calculated from ω I m(εav)/cnav. Finally, we define the parameters ζ = 4πnavh/λ and ψ = tan-1(2kav/(nav2 + kav2 - 1). Since eq 7 takes into account the reflection losses, it can be considered as the extinction coefficient of the film. As an example of the absorption and reflection colors of thin films built up via layer-by-layer assembly48 (using a positively charged polyelectrolyte as a molecular glue), Fig. 3 shows 15 nm Au nanoparticles surrounded with inert SiO2 shells of varying thickness7. The absorption colors range from blue (the color of a thin, bulk film) for densely packed metal spheres to light red for well separated particles. The typical

REVIEW FEATURE

Conclusion

Fig. 3 Left: Schematic drawing of a multilayer film formed by layer-by-layer assembly of SiO2-coated Au nanoparticles (1 = glass substrate; 2 = cationic polyelectrolyte; 3 = nanoparticles). Right: Photographs of transmitted (top) and reflected (bottom) colors from Au@SiO2 multilayer thin films with varying silica shell thickness.

Apart from the (linear) optical properties described here, nonlinear optical properties are also of great interest for applications of metal nanoparticles in ultrafast optical switches49. In particular, the third-order susceptibility of metal nanoparticles at wavelengths around the plasmon resonance achieves large values with very fast (0.7 – Very polydisperse. Care should be taken in interpreting results as the sample may not be suitable for the technique (e.g., a sedimenting high size tail may be present)

Non-Negatively constrained Least Squares (NNLS) algorithm • Used for Multimodal size distribution (MSD) – Only positive contributions to the intensityweighted distribution are allowed – Ratio between any two successive diameters is constant – Least squares criterion for judging each criterion is used – Iteration terminates on its own

13

11/15/2012

Correlation function Correlograms Correlograms show the correlation data providing information about the sample The shape of the curve provides clues related to sample quality • • • •

Decay is a function of the particle diffusion coefficient (D) Stokes-Einstein relates D to dH z-average diameter is obtained from an exponential fit Distributions are obtained from multi-exponential fitting algorithms Noisy data can result from • • •

Low count rate Sample instability Vibration or interference from external source

Correlation function Correlograms

14

11/15/2012

Data interpretation Correlograms • • •

Very small particles Medium range polydispersity No large particles/aggregrates present (flat baseline)

Data interpretation Correlograms • •

•

Large particles Medium range polydispersity Presence of large particles/agglomerat es (noisy baseline)

15

11/15/2012

Data interpretation Correlograms • • •

Very large particles High polydispersity Presence of large particles/agglomerat es (noisy baseline)

Upper size limit of DLS • DLS will have an upper limit wrt size and density • When particle motion is not random (sedimentation or creaming), DLS is not the correct technique to use

• Upper limit is set by the onset of sedimentation • Upper size limit is therefore sample dependent • No advantage in suspending particles in a more viscous medium to prevent sedimentation because Brownian motion will be slowed down to the same extent making measurement time longer

16

11/15/2012

Upper size limit of DLS • Need to consider the number of particles in the detection volume – Amount of scattered light from large particles is sufficient to make successful measurements, but … – Number of particles in scattering volume may be too low – Number fluctuations – severe fluctuations of the number of particles in a time step can lead to problems defining the baseline of the correlation function – Increase particle concentration, but not too high or multiple scattering events might arise

Detection volume

Detector

Laser

17

11/15/2012

Lower particle size limit of DLS • Lower size limit depends on

– Sample concentration – Refractive index of sample compared to diluent – Laser power and wavelength – Detector sensitivity – Optical configuration of instrument

Lower limit is typically ~ 2 nm

Sample preparation • Measurements can be made on any sample in which the particles are mobile • Each sample material has an optimal concentration for DLS analysis – Low concentration → not enough scattering – High concentration → multiple scattering events affect particle size

18

11/15/2012

Sample preparation • Upper limit governed by onset of particle/particle interactions – Affects diffusion speed – Affects apparent size

• Multiple scattering events and particle/particle interactions must be considered • Determining the correct particle concentration may require several measurements at different concentrations

Sample preparation An important factor determining the maximum concentration for accurate measurements is the particle size

19

11/15/2012

Sample concentration Small particles • For particle sizes [crit, the size distribution function has two extrema. It can be seen that if t increases both the number and the average size of the micellar aggregates increase. In deriving eq 16-19, the surface area A is taken as the surface area of the micellar core since it is based on the volume of the hydrophobic core. In reality the surface area should account for the surface roughness due to the presence of the head groups. The actual surface area can be calculated by increasing the radius of the micellar core by a value A is calculated using eq 10 where ro is now given by (26) The introduction of this correction makes the algebra leading to the equations for gcrit and [crit more complex. For this case, the value of &it can be computed only by numerical methods. Figure 2 represents the size distribution function for 6 = 3 A and a = 8 X lo4 cal A2/mol for an amphiphile with an octyl hydrocarbon tail. The mole fractions of amphiphiles present both in nonaggregated form and as aggregates for various values of are presented in Table I for the same parameters as in Figure 2. Table I shows that below Ecrit = 27.82, the concentration of micellar aggregates remains extremely small. At = &it, the concentration of micellar aggregates begins to increase even though their contribution to the total amphiphile concentration remains negligible. However as the value of increases beyond &it, the number of aggregates increases significantly and their contribution to the total amphiphile concentration becomes important. The values of cmc normally reported should correspond to the total amphiphile concentration for values of f somewhat higher than Fcrit. The values of Fcrit and the corresponding values of the

2625

Transition Point for Micellar Size Distribution

TABLE I: Variation of the Mole Fraction of the Aggregates a s a Function of the Mole Fraction of the Nonaggregated Amphiphilea NJF~

5 22.14 24.98 27.82 33.50 39.18 44.86

2.09 2.34 2.61 3.14 3.67 4.21

x 10-4 x 10-4 x 10-4 x 10-4 x lV4 X lV4

1.53 x 1.97 X 2.54 X 3.61 x 1.28 x 8.59'

10-1' lo-'' 1r1' 10-9 10-4

= n , = 8, a = 8 X lo4 cal A2/mol, 6 = 3 A. Mole fraction of nonaggregated amphiphile. Mole fraction of the aggregates. See the explanation in the text.

TABLE I I V a r i a t i o n of the Critical Concentration as a Function of Hydrocarbon Chain Lengtha E.373

%

5 cri t

Calcd Ccritb

6 8 10 12 14 16

27 27.8 29.0 29.5 30 31

2.64 2.61 2.58 2.52 2.49 2.45

Q

C

6 = 3

A,

a = 8 X

x 10-3 x 10-4 X lV5 x lom6 X

lV7

Exptl cmcbpc 2.76 2.59 2.44 2.30 2.16 2.04

x 10-3 x 10-4 x 10-5 x 1V6 x 10-7

x 1W8 x 104 cal AZ/mol. In mole fraction units.

Reference 3a.

phiphiles with different hydrocarbon chain lengths. When 6 > 0, the value of &it is determined not only by a but also by n, and 6. However, the value of Fcrit for 6 = 0 is a good first approximation even for 6 > 0. Table I contains an aggregate molar fraction larger than unity because the corresponding value of [ is larger than tmaxwhich for this particular case is equal to 41. (The first constant in eq 25 was taken as 1750 cal/mol.) Of course, an upper bound of fmax can be obtained from the condition

N~

9

Figure 2. Variation in the aggregate size distribution function with nona gregated amphiphile concentration for 6 = 3 A, a = 8 X lo4 cai A /mol, n, = 8.

4

critical concentration are computed for 6 = 3 A, a = 8 X lo4 cal A2/mol and for amphiphiles with various hydrocarbon chain lengths (Table 11).Also shown in the table are the experimental cmc of amphiphiles with hexoxyethylene glycol monoether head groups and various hydrocarbon chain lengths. It should be mentioned that the value of a used here is chosen for illustrative purposes and has not been computed on theoretical grounds for this amphiphile. It can be seen how closely the critical concentration Ccrit, defined here predicts the cmc values normally reported. For 6 = 0 the value of Fcrit is solely determined by the repulsive interaction term a and remains a constant for am-

+

e~,+ F N~ =

2

It has already been mentionedhhat the representation of aggregates as spheres is not essential to the definition of the critical concentration given here. Hoeve and Benson5 suggest a spherical shape for micelles at low aggregation numbers and an oblate spherocylindrical shape at higher aggregation numbers. Tanford3 suggests an ellipsoidal shape for the aggregates. It is possible to incorporate nonspherical shapes of aggregates as well as the variation in shapes accompanying growth in aggregate size in the present treatment. The quantitative results in terms of the size distribution function and critical concentrations will depend on these assumptions. However, the qualitative change in the pattern of the size distribution function from that of a monotonic decreasing function at loa monomer concentration to one exhibiting extrema a t higher monomer concentration is always present. The explicit expressions used here for the various freeenergy terms are of an empirical nature. Of course, it is possible to obtain expressions for the free energy on theoretical grounds following the approaches of Hoeve and Benson5 or Poland and Scheraga4 using the formalism of statistical thermodynamics. The Journal of Physical Chemistry, Vol. 79, No. 24, 1975

Yohji Achiba and Katsumi Kimura

2626

Conclusions A well-defined transition point in the aggregation process emerges as that separating two different types of behavior of the aggregate size distribution function. A critical concentration corresponding to this transition point is defined which is a close lower bound on the cmc values usually reported.

Acknowledgment. This work was supported by NSF.

Referenoes and Notes ( I ) (a) P. Mukerjee, Adv. Colloidhterface Sci., 1, 241-275 (1967); (b) J. M. Corkill and J. F. Qoodman. ib/d., 2, 297-330 (1969). (2) K. Shinoda, T. Nakagawa, B. Tamamushi, and T. Isemura. “Colloidal Sur. factants”, Academic Press, New York, N.Y., 1963. (3) (a) C. Tanford, “The Hydrophobic Effect”, Wiley, New York, N.Y., 1973; (b) C. Tanford, J. Phys. Chem., 78, 2469-2479 (1974). (4) (a) D. C. Poland and H. A. Scheraga, J. Phys. Chem., 80, 2431-2442 (1965); (b) D. C. Poland and H. A. Scheraga, J. Colloid hterface Sci., 21, 273-283 (1966). (5) C. A. J. Hoeve and G. C. Benson, J. Phys. Chem., 61, 1149-1 158 (1957).

Rate Constants of Triplet-State Ionic Photodissociation of Weak Charge-Transfer Complexes Formed between Pyromellitic Dianhydride (PMDA) and Naphthalenes Yohji Achlba and Katsumi Kimura* Physical Chemistry Laboratory, Institute of Applied €/ectricity, Hokkaido University, Sapporo 060, Japan (Received January 27, 1975; Revised Manuscript ReceivedJune 23, 1975) Publication costs assistedby the hstitute of Applied €lectricity, Hokkaido University

The weak charge transfer (CT) complexes of PMDA with naphthalene and several of its derivatives have been excited in CT absorption bands by means of a laser flash technique in solution at room temperature. Transient absorption spectra due to the triplet-triplet transitions of the CT complexes initially appear, then followed by the absorption spectra of the radical anion of PMDA (PMDA-) in the nanosecond region. By a first-order kinetic analysis, it has been indicated that the rise curves of the PMDA- absorption band give rise to approximately the same rate constants as those of the CT triplet decay. The rate constants of such triplet-state anion formations have been determined from the PMDA- rise curves. It has been found that the rate constant increases with the dielectric constant of solvent. In the photolysis of these CT complexes, it has also been suggested that PMDA; is produced via both the lowest excited singlet and triplet states of the CT complexes.

Introduction Spectroscopic evidence of the ionic photodissociation of a ground-state charge transfer (CT) complex may be obtained by analyzing the rise curve of radical ions produced as a result of electron transfer. The first direct, spectroscopic evidence of ionic photodissociation in the excited triplet state was obtained with the PMDA-mesitylene CT complex at low temperature by Potashnik et al.,I who followed the decay curve of CT phosphorescence as well as the rise curve of the optical absorption of the radical anion. So far, several examples have been published on the tripletstate ionic photodissociation of CT complexes.2-6 In the present work, we considered it interesting to determine the rate constants of the CT-triplet ionic photodissociation using a nanosecond laser flash technique. It may also be interesting to study the effects of solvent and electron donor on the rate constant. The reason that PMDA has been used as an electron acceptor is that its CT complexes of naphthalenes can be excited in the CT bands by the second harmonic (347 nm) of the ruby laser and that the resulting radical anion of PMDA (PMDA-) shows a strong absorption band a t about 665 nm, which is well separated from the T-T absorption bands. Naphthalene and its derivatives (1-methyl-, 2-methyl-, 1-chloro-, 2,3-diThe Journal of Physical Chemistry, Vol. 79, No. 24, 1975

methyl-, and 2-hydroxynaphthalene) have been used as electron donors, since their CT complexes with PMDA show considerably strong T-T absorption spectra. Experimental Section 1,2-Dichloroethan’e (DCE) was repeatedly washed with dilute sulfuric acid (lo%), alkaline aqueous solution (lo%), and water, and finally purified by distillation after drying over CaC12. Tetrahydrofuran (THF), dimethoxyethane (DME), and butyronitrile were refluxed over CaHz and distilled. Acetonitrile was refluxed over phosphorus pentoxide and distilled. Naphthalene and 2-hydroxynaphthalene were recrystallized from ligroin, and 2-methyl- and 2,3dimethylnaphthalene from petroleum ether. 1-Methyl-and 1-chloronaphthalene were purified by vacuum distillation and PMDA by vacuum sublimation. Ordinary visible and ultraviolet absorption spectra were measured on a Cary 15 spectrophotometer and phosphorescence spectra on a Hitachi MPF-PA fluorescence spectrophotometer. Laser pulse excitation experiments were carried out with a giant pulse ruby-laser apparatus previously described by Takemura et al.,’ the second harmonic (347 nm) of the ruby laser being generated by an ADP crystal. The laser pulse apparatus was combined with a 50-cm Nar-

07/ 11/ 12

KRÜSS: CM C M easur em ent

You are here: Theory

Measurements

Surface Tension

CMC Measurement

Surface tension Critical micelle concentration (CMC) An important measure for the characterization of surfactants is the critical micelle concentration (CMC). Surfactants consist of a hydrophilic "head" and a hydrophobic "tail". If a surfactant is added to water then it will initially enrich itself at the surface; the hydrophobic tail projects from the surface. Only when the surface has no more room for further surfactant molecules will the surfactant molecules start to form agglomerates inside the liquid; these agglomerates are known as micelles. The surfactant concentration at which micelle formation begins is known as the critical micelle formation concentration (CMC). Micelles are spherical or ellipsoid structures on whose surface the hydrophilic heads of the surfactant molecules are gathered together whereas the hydrophobic tails project inwards. The washing effect of surfactants is based on the fact that hydrophobic substances such as fats or soot can be stored within the micelles.

Standard procedure The critical micelle formation concentration (CMC) can be determined by carrying out surface tension measurements on a series of different surfactant concentrations. Surfactants exhibit a specific surface tension curve as a function of the concentration. Initially the surfactant molecules increasingly enrich themselves at the water surface. During this phase the surface tension decreases linearly with the logarithm of the surfactant concentration. When the CMC is reached, i.e. when the surface is saturated with surfactant molecules, a further increase in surfactant concentration no longer has any appreciable influence on the surface tension.

Determination of the critical micelle formation concentration This means that in order to determine the CMC the two linear sections formed by the measuring points obtained from the series of different concentrations must be determined. The CMC is obtained from the intersection of the straight lines for the linear concentration-dependent section and the concentration-independent section. In the K100 and K12 the CMC is determined by using the CMC Add-In of the LabDesk software. The concentration series is generated automatically with a computer-controlled Dosimat, so that only a surfactant stock solution needs to be made up. The measurements and their evaluation are carried out automatically. www. kr uss. de/ en/ t heor y/ m easur em ent s/ sur f ace- t ension/ cm c- m easur em ent . ht m l

1/ 2

07/ 11/ 12

KRÜSS: CM C M easur em ent

Reverse CMC measurement For reverse CMC measurements not the solvent but the parent solution is first put into the sample vessel and then diluted with the solvent step by step. One case in which the reverse CMC measurement should be chosen is when the concentration of the sample is above the clouding point. Such a solution can’t be dosed homogeneously, but it can be diluted homogeneously by adding the solvent. Another application for reverse CMC measurements is the case that the CMC is expected at a low concentration of the surfactant. With standard CMC measurements the region of interest is also the region where only small amounts of the sample solution are dosed and where therefore the largest error by means of dosing inaccuracy would occur. With reverse CMC measurement low concentrations are reached with large amounts of the solvent and so this error is reduced to a minimum. Another advantage is that the dosimat only gets in contact with the pure solvent and can run in continuous operation. This makes the reverse CMC measurement an ideal method for routine measurements. Since the end volume exceeds the initial volume by many times a cone shaped sample vessel is used.

Extended CMC-Method For the K100 the"Extended CMC" method is available. The software LabDeskTM does not control one but two dosing units. The second unit substracts the amount of liquid previously added by the first one. Thus the accessible concentration range is increased many times over.

Range of conventional CMC method

Range of extended CMC method

KRÜSS GmbH · Borsteler Chaussee 85-99a · 22453 Hamburg · Germany · Tel: +49 (40) 514401-0

www. kr uss. de/ en/ t heor y/ m easur em ent s/ sur f ace- t ension/ cm c- m easur em ent . ht m l

2/ 2

J Pharm Pharmaceut Sci (www.cspscanada.org) 8(2):147-163, 2005

Micellar solubilization of drugs. Carlota O. Rangel-Yagui, Adalberto Pessoa-Jr, and Leoberto Costa Tavares Department of Biochemical and Pharmaceutical Technology /FCF, University of São Paulo, São Paulo, Brazil Received 17 November 2004, Revised 2 February 2005, Accepted 4 March 2005, Published 8 July 2005

Abstract PURPOSE: Micellar solubilization is a powerful alternative for dissolving hydrophobic drugs in aqueous environments. In this work, we provide an insight into this subject. METHODS: A concise review of surfactants and micelles applications in pharmacy was carried out. RESULTS: Initially, a description of surfactants and aqueous micellar systems is presented. Following, an extensive review on micellar drug solubilization, including both the principles involved on this phenomenon and the work already done regarding solubilization of drugs by micelles is presented. The application of micelles in drug delivery, in order to minimize drug degradation and loss, to prevent harmful side effects, and to increase drug bioavailability, is also presented. Special emphasis is given to the more recent use of polymeric micelles. Finally, we briefly discuss the importance of surfactants and micelles as biological systems models as well as its application in micellar catalysis. CONCLUSIONS: As can be seen from the review presented, the use of micelles in pharmacy is an important tool that finds numerous applications. INTRODUCTION Surfactants are known to play a vital role in many processes of interest in both fundamental and applied science. One important property of surfactants is the formation of colloidal-sized clusters in solutions, known as micelles, which have particular significance in pharmacy because of their ability to increase the solubility of sparingly soluble substances in water (1). Micelles are known to have an anisotropic water distribution within their structure. In other words, the water concentration decreases from the surface towards the core of the micelle, with a completely hydrophobic (water-excluded) core. Consequently, the spatial position of a solubilized drug in a micelle will Corresponding Author: Dr. Carlota de Oliveira Rangel Yagui, Av. Prof. Lineu Prestes, 580 – Bloco 16, CEP: 05508-950 – São Paulo, SP. [email protected]

147

depend on its polarity: nonpolar molecules will be solubilized in the micellar core, and substances with intermediate polarity will be distributed along the surfactant molecules in certain intermediate positions. On the other hand, numerous drug delivery and drug targeting systems have been studied in an attempt to minimize drug degradation and loss, to prevent harmful side effects, and to increase drug bioavailability (26). Within this context, the utilization of micelles as drug carriers presents some advantages when compared to other alternatives such as soluble polymers and liposomes. Micellar systems can solubilize poorly soluble drugs and thus increase their bioavailability, they can stay in the body (blood) long enough to provide gradual accumulation in the required area, and their sizes permit them to accumulate in areas with leaky vasculature (7). In general, surfactants play an important role in contemporary pharmaceutical biotechnology, since they are largely utilized in various drug dosage forms to control wetting, stability, bioavailability, among other properties (8). It is important to notice that lyophobic colloids, such as polymers, require certain energy to be applied for their formation, are quite unstable from the thermodynamic point of view, and frequently form large aggregates. Association colloids such as micelles, on the other hand, can form spontaneously under certain conditions (self-assembling systems), and are thermodynamically more stable towards both dissociation and aggregation (9). Therefore, the study of surfactants and their role in pharmacy is of paramount importance, especially with respect to their ability of solubilizing hydrophobic drugs. In this work, we provide a review of micellar solubilization of drugs in surfactant systems, blending it with basic information on surfactants structure and properties, as well as the applications for drug delivery.

J Pharm Pharmaceut Sci (www.cspscanada.org) 8(2):147-163, 2005

Surfactants and Micelles Surfactants are amphiphilic molecules composed of a hydrophilic or polar moiety known as head and a hydrophobic or nonpolar moiety known as tail. The surfactant head can be charged (anionic or cationic), dipolar (zwitterionic), or non-charged (nonionic). Sodium dodecyl sulfate (SDS), dodecyltrimethylammonium bromide (DTAB), n-dodecyl tetra (ethylene oxide) (C12E4) and dioctanoyl phosphatidylcholine (C8lecithin) are typical examples of anionic, cationic, nonionic and zwitterionic surfactants, respectively (Figure 1). The surfactant tail is usually a long chain hydrocarbon residue and less often a halogenated or oxygenated hydrocarbon or siloxane chain (16, 17).

tion occurs and, if so, at what concentration of monomeric surfactant, depends on the balance of the forces promoting micellization and those opposing it (19, 20).

Figure 2: Schematic illustration of the reversible monomer-micelle thermodynamic equilibrium. The black circles represent the surfactant heads (hydrophilic moieties) and the black curved lines represent the surfactant tails (hydrophobic moieties).

Figure 1: Examples of I-anionic (SDS), II-cationic (CTAB), III- nonionic (C12E4) and VI-zwitterionic (C8-lecithin) surfactants.

A surfactant, when present at low concentrations in a system, adsorbs onto surfaces or interfaces significantly changing the surface or interfacial free energy. Surfactants usually act to reduce the interfacial free energy, although there are occasions when they are used to increase it (17). When surfactant molecules are dissolved in water at concentrations above the critical micelle concentration (cmc), they form aggregates known as micelles. In a micelle, the hydrophobic tails flock to the interior in order to minimize their contact with water, and the hydrophilic heads remain on the outer surface in order to maximize their contact with water (see Figure 2) (18,19). The micellization process in water results from a delicate balance of intermolecular forces, including hydrophobic, steric, electrostatic, hydrogen bonding, and van der Waals interactions. The main attractive force results from the hydrophobic effect associated with the nonpolar surfactant tails, and the main opposing repulsive force results from steric interactions and electrostatic interactions between the surfactant polar heads. Whether micelliza-

148

The determination of a surfactant cmc can be made by use of several physical properties, such as surface tension (γ), conductivity (κ) – in case of ionic surfactants, osmotic pressure (π), detergency, etc. When these properties are plotted as a function of surfactant concentration (or its logarithm, in case of surface tension), a sharp break can be observed in the curves obtained evidencing the formation of micelles at that point (16) (Figure 3).

Figure 3: Changes in the physical properties detergency, conductivity (κ), osmotic pressure (π) and surface tension (γ) of an aqueous solution of surfactant as a function of surfactant concentration. The break in the curve of each property corresponds to the Critical Micelle Concentration (cmc).

J Pharm Pharmaceut Sci (www.cspscanada.org) 8(2):147-163, 2005

Another important parameter that characterizes micelles is the aggregation number, Nag, that corresponds to the average number of surfactant monomers in each micelle of a micellar solution. Usually, in a micellar solution the aggregation number is approximately constant for a broad total concentration range (up to about 100 times the cmc), with the number of micelles varying (21). However, at certain conditions micelles can grow with the aggregation number varying with the surfactant concentration. (22).

The interior of the micelle containing the hydrophobic groups presents a radius of approximately the length of the fully extended hydrophobic chain (17). Another important characteristic of micelles is that the aqueous phase penetrates into the micelle beyond the hydrophilic head groups, and the first few methylene groups adjacent to the head are considered in the hydration sphere. Therefore, we can divide the interior region of the micelle in an outer core penetrated by water and in an inner core completely water-excluded (22).

Micelles are labile entities formed by the noncovalent aggregation of individual surfactant monomers. Therefore, they can be spherical, cylindrical, or planar (discs or bilayers). Micelle shape and size can be controlled by changing the surfactant chemical structure as well as by varying solution conditions such as temperature, overall surfactant concentration, surfactant composition (in the case of mixed surfactant systems), ionic strength and pH. In particular, depending on the surfactant type and on the solution conditions, spherical micelles can grow one-dimensionally into cylindrical micelles or two-dimensionally into bilayers or discoidal micelles. Micelle growth is controlled primarily by the surfactant heads, since both one-dimensional and two-dimensional growth require bringing the surfactant heads closer to each other in order to reduce the available area per surfactant molecule at the micelle surface, and hence the curvature of the micelle surface (18, 22).

Based on the geometry of various micellar shapes and the space occupied by the hydrophilic and hydrophobic groups of the surfactants, it is possible to estimate the structure of a micelle (20). Accordingly, the parameter VH/lcao can determine the shape of the micelle, with VH corresponding to the volume of the hydrophobic group in the micellar core, lc is the length of the hydrophobic group in the core and ao the cross-sectional area occupied by the hydrophilic group at the micelle-solution interface. Based on Tanford (19), VH = 27.4 + 26.9n Å, where n is the number of carbon atoms in the chain less one, and lc = 1.5 + 1.265n Å, depending upon the extension of the chain. Therefore, for a fully extended chain, lc = 1.5 + 1.265n Å (Table 1).

For all these micellar structures in aqueous media, the surfactant molecules are oriented with their polar heads towards the water phase and their tail away from it. In ionic micelles, the interfacial region between the micelle and the aqueous phase contains the ionic head groups, the Stern Layer of the electrical double layer related to these groups, approximately half of the counter ions associated with the micelle, and water. The remaining counter ions are contained in the Gouy-Chapman portion of the double layer that extends further into the aqueous phase. The length of the double layer is a function of the ionic strength of the solution and it can be highly compressed in the presence of electrolytes (23). For the nonionic surfactants having a polyethylene oxide (PEO) head group, the structure is essentially the same, except that the counter ions are not present in the outer region, but rather coils of hydrated polyethylene oxide chains. 149

Table 1: Correlation between the parameter VH/lcao and the structure of the micelle.

Micellar Solubilization An important property of micelles that has particular significance in pharmacy is their ability to increase the solubility of sparingly soluble substances in water. In this context, solubilization can be defined as the spontaneous dissolving of a substance by reversible interaction with the micelles of a surfactant in water to form a thermodynamically stable isotropic solution with

J Pharm Pharmaceut Sci (www.cspscanada.org) 8(2):147-163, 2005

reduced thermodynamic activity of the solubilized material (17). If we plot the solubility of a poorly soluble compound as a function of the concentration of surfactant, as shown in Figure 4, usually what happens is that the solubility is very low until the surfactant concentration reaches the cmc. At surfactant concentrations above the cmc the solubility increases linearly with the concentration of surfactant, indicating that solubilization is related to micellization.

(2)

where Stot is the total drug solubility, SW is the water drug solubility, Csurf is the molar concentration of surfactant in solution, and cmc is the critical micelle concentration (25). Since above the cmc the surfactant monomer concentration is approximately equal to the cmc, the term (Csurf – cmc) is approximately equal to the surfactant concentration in the micellar form and, therefore, χ is equal to the ratio of drug concentration in the micelles to the surfactant concentration in the micellar form. On the other hand, the micelle-water partition coefficient is the ratio of drug concentration in the micelle to the drug concentration in water for a particular surfactant concentration, as follows: (3)

Figure 4: Schematic plot of the concentration of a poorly soluble compound as a function of the surfactant concentration in aqueous solution.

From the thermodynamic point of view, the solubilization can be considered as a normal partitioning of the drug between two phases, micelle and aqueous, and the standard free energy of solubilization (ΔGSº) can be represented by the following expression (1):

Combining Equations (2) and (3), we can relate the two solubility descriptors. Accordingly, for a given surfactant concentration: (4) As can be seen, P is related to the water solubility of the compound, in contrary to χ (25). In order to eliminate the dependence of P on the surfactant concentration, a molar micelle-water partition coefficient (PM), corresponding to the partition coefficient when Csurf = 1 M, can be defined as follows:

(1) (5) where R is the universal constant of the gases, T is the absolute temperature, and P is the partition coefficient between the micelle and the aqueous phase. Usually, the solubilization of a molecule by a surfactant can be evaluated based on two descriptors that are the molar solubilization capacity, χ, and the micellewater partition coefficient, P (24). The χ value is defined as the number of moles of the solute (drug) that can be solubilized by one mol of micellar surfactant, and characterizes the ability of the surfactant to solubilize the drug. It can be calculated based on the general equation for micellar solubilization: 150

The lower is the cmc value of a given surfactant, the more stable are the micelles. This is especially important from the pharmacological point of view, since upon dilution with a large volume of the blood, considering intravenous administration, only micelles of surfactants with low cmc value still exist, while micelles from surfactants with high cmc value may dissociate into monomers and their content may precipitate in the blood (26). There are a number of possible loci of solubilization for a drug in a micelle, as represented in Figure 5.

J Pharm Pharmaceut Sci (www.cspscanada.org) 8(2):147-163, 2005

Figure 5: Possible loci of solubilization of drugs in surfactant micelles, depending on the drug hydrophobicity. The black bold lines (⎯) represent the drug at different sites in the micelle. The black circles represent the surfactant heads, the black bold curved lines represent surfactant heads consisting of PEO, and the light black curved lines represent the surfactant tails.

Accordingly, hydrophilic drugs can be adsorbed on the surface of the micelle (1), drugs with intermediate solubility should be located in intermediate positions within the micelle such as between the hydrophilic head groups of PEO micelles (2) and in the palisade layer between the hydrophilic groups and the first few carbon atoms of the hydrophobic group, that is the outer core (3), and completely insoluble hydrophobic drugs may be located in the inner core of the micelle (4) (7,17). The existence of different sites of solubilization in the micelle results from the fact that the physical properties, such as microviscosity, polarity and hydration degree, are not uniform along the micelle (27). Mukerjee and Cardinal (28) studied the microenvironments of benzene, some of its derivatives, Triton X100, and naphthalene when solubilized in micelles at low solubilizate to surfactant ratios and proposed the existence of at least two states (loci) of solubilization with different polarity. According to the authors, the total uptake by micelles could be divided approximately into an “adsorbed” fraction (location at the micelle-water interface) and a “dissolved” fraction (location in the hydrocarbon core). When adsorption takes place the solubility increases beyond the solubility power of the hydrocarbon core. In fact, numerous studies indicate that the solubility of slightly polar substances and aromatic compounds tend to be considerably higher than the solubility of aliphatic compounds presenting similar molar volumes, despite the fact that the later are expected to be more compatible with the aliphatic hydrocarbon core of most micelles (28).

151

The capacity of surfactants in solubilizing drugs depends on numerous factors, such as chemical structure of the surfactant, chemical structure of the drug, temperature, pH, ionic strength, etc (7). Nonionic surfactants usually are better solubilizing agents than ionic surfactants for hydrophobic drugs, because of their lower cmc values. For polar drugs it is more complicate to establish a general relationship between the degree of solubilization and the chemical structure of the surfactant, since solubilization can be in both the inner and the outer regions of the micelle. Krishna and Flanagan (29) observed that, for the antimalarial drug β-Arteether (an endoperoxide containing a sesquiterpene lactone), nonionic surfactants showed much lower solubilization power than ionic surfactants. They suggested that the solubilization of this drug may not only involve incorporation into the micellar interior, but may be substantially due to adsorption at the micelle-water interface. Regarding the influence of structure of the drug, crystalline solids generally show less solubility in micelles than do liquids of similar structure (17). For polar drugs, the depth of penetration into the micelle varies with the structure of the drug. Usually, the less polar the drug (or the weaker its interaction with either the polar head of the surfactant in the micelle or the water molecules at the micelle-water interface) and the longer is the chain length, the smaller its degree of solubilization, reflecting its deeper penetration into the palisade layer (17,23). The extent of solubilization into a particular micelle depends upon the locus of solubilization and therefore the shape of the micelle. As described previously, the shape of the micelle is determined by the value of the parameter VH/lcao and as this parameter increases the micelle becomes more asymmetrical and the volume of the inner core increases relative to that of the outer portion. Therefore, one can expect that the solubilization of drugs in the core will increase with increase in asymmetry, whereas the solubilization of drugs in the outer region will decrease (17). In fact, it was observed that for alkyl sodium sulfates and alkyl trimethylammonium bromides, the solubilization of β-Arteether increases with the increase in the alkyl chain length, due to the larger micellar size (29).

J Pharm Pharmaceut Sci (www.cspscanada.org) 8(2):147-163, 2005

However, Barry and El Eini (30) studying the solubilization of non-polar steroidal drugs in aqueous solutions of long-chain polyoxyethylene nonionic surfactants have observed that the molar solubilizing efficiency of surfactants increased as the length of the PEO chain increased while micellar sizes are known to decrease with the increase in PEO chain length. The authors suggested that, although the inclusion of nonpolar steroids into the micelles decreases as the PEO hydrophilic chain increases, the number of micelles in equimolar amounts of surfactants increases and consequently the total amount of steroid per mole of surfactant is greater, hence the observed increase in solubilizing efficiency with increased hydrophilic chain length when molar concentrations are considered. Ong and Manoukian (31) have studied the solubilization of timobesone acetate, a corticosteroid used in inflammatory therapy, in nonionic surfactants solutions and observed that the solubilization capacity increased with increasing length of the hydrophobic tail of the surfactants. Therefore, timobesone was assumed to be solubilized in the hydrophobic core of the micelles. This observation was also confirmed by the fact that the length of the PEO chain of the surfactants studied did not affect the solubilization capacity, for a given tail length, and thus solubilization should not have occurred in the palisade layer or among the PEO heads. In general, the amount of drug solubilized in a micellar system increases with the increase in temperature. Alkhamis et al. (32) studied the solubilization of the drug gliclazide, a second-generation sulfonylurea used in the treatment of non-insulin dependent diabetes mellitus. The drug solubility was determined as a function of the concentration of different surfactants at 25 and 37oC and, for all the ionic surfactants studied, the solubilization was higher at 37oC than at 25oC. This was attributed to the increase in thermal agitation, which results in an increase in the space available for solubilization in the micelle, in addition to the increase of gliclazide solubility in water at higher temperatures. For the polyoxyethylene nonionic surfactants, the effect of the temperature on the extent of drug solubilization may depend on whether the drug is located inside the hydrophobic core or in the palisade layer. In this same work, the solubility of gliclazide was found 152

to decrease with temperature for the nonionic surfactants studied. Barry and El Eini (30) also observed a significant decrease in the micelle/water molar partition coefficient, PM, obtained for nonpolar steroidal drugs in PEO surfactants solutions when the temperature was increased from 10 to 50oC. The drugs are believed to be located preferentially in the palisade layer, and the increase in temperature causes dehydration of the PEO groups, bringing them closer and consequently reducing the space available for the drugs in this region of the micelle. Nevertheless, the solubility of drugs located preferentially in the inner core of PEO micelles is expected to increase as the temperature is raised, due to micellar growth (23). The ionic strength can influence significantly the solubilization of a drug in micellar solutions, especially in case of ionic surfactants. The addition of small amounts of salts decreases the repulsion between the similarly charged ionic surfactant head groups, thereby decreasing the cmc and increasing the aggregation number and volume of the micelles. The increase in aggregation number favors the solubilization of hydrophobic drugs in the inner core of the micelle. On the other hand, the decrease in mutual repulsion of the ionic head groups causes closer packing of the ionic surfactant molecules in the palisade layer decreasing the volume available for solubilization of polar drugs. The addition of salts to solutions of PEO nonionic surfactants may also increase the extent of solubilization of hydrophobic drugs because of the increase in aggregation number (17). The pH of micellar solutions can also show significant influence on the extent of solubilization of drugs, since it may change the equilibrium between ionized and molecular forms of some drugs. LI et al. (33) studied the solubility of the ionized and un-ionized forms of flavopiridol in polysorbate solutions at different pH values. This drug is a weakly basic (pKa = 5.68) derivative of rohitukine that has been developed for breast cancer treatment. The authors observed that the highest total drug solubility was achieved at pH 4.3 where most of the drug was ionized. More recently, Li and Zhao (34) studied the solubilization of flurbiprofen, a non-steroidal antinflamatory drug used in rheumatoid arthritis, in polysorbate solutions at different pH values. This drug is a weak acid, with a pKa of 4.17. It was observed that the drug solubility increases with the

J Pharm Pharmaceut Sci (www.cspscanada.org) 8(2):147-163, 2005

increase in pH for pH values over the pKa, due to the increase in the ionized form of the drug. The authors have also proposed an equilibrium-based model to characterize drug-surfactant interactions in pH-controlled systems, reflecting both interactions and interdependence among all drug-containing species: unionized drug in water, ionized drug in water, unionized drug in micelles, and ionized drug in micelles. The model proposed yielded reasonably good estimation when compared to experimental data. Regarding ionic surfactants, a particular kind of behavior can be observed for the solubility of drugs at different pH values. Enhanced solubility of a drug may be observed at pH values at which the drug is found mostly ionized, when surfactant and drug are oppositely charged. This behavior is a consequence of the electrostatic interactions between the surfactant molecules and the charged drug that causes a decrease in the repulsive forces between the head groups of the surfactant molecules, contributing to the micellization process and thus decreasing the cmc value. In fact, an early study has demonstrated that the drug chlorpromazine can form mixed monomolecular films with phospholipids such as L-α-dipalmitoyl phosphatidylethanolamine and L-α-dipalmitoyl phosphatidyl-choline (35). More recently, Caetano et al. (36) observed a comicellization phenomenon for the negatively charged surfactant SDS and trifluoperazine, an amphiphilic cationic drug used as antipsychotic and tranquilizer. The authors demonstrated, based on SAXS (Small Angle Xray Scattering) studies, that the presence of the protonated drug mediates the effect that the counter ion has on the SDS micelle, in such a way that the drug is able to promote micellar surface charge screening. Moreover, the electrostatic interaction between the positively charged drug and the negatively charged SDS must cause a decrease in the repulsive forces between the head groups of the surfactant. One interesting approach is to combine micellar solubilization with other properties that may be improved in a drug solution. In this context, recently Palma et al. (37) combined the solubilization properties of a surfactant with the ascorbic acid antioxidant property that protects drugs from degradation by light, heat, dissolved oxygen and other radical producing species, by means of synthesizing an ascorbyl-decanoate surfac153

tant. It was observed that micellar solutions of the surfactant obtained significantly improved the solubility of hydrophobic drugs with respect to pure water, by including these molecules in the hydrophobic micellar core, as well as protected them from degradation. It was also observed that the drug solubilization was more effective for the most hydrophobic drugs (Danthron and Griseofulvin) than for more hydrophilic ones (Phenacetin). A nonionic surfactant that deserves special attention is Cremophor EL (CrEL), which has been used for solubilization of a wide variety of hydrophobic drugs such as anaesthetics, photosensitizers, sedatives, immunosuppressive agents and anticancer drugs. This heterogeneous surfactant is a result of the reaction of castor oil with ethylene oxide, with polyoxyethylene glycerol ricinoleate 35 as the major component identified (38). Formulations containing CrEL have been shown to present important biological side effects, including severe anaphylactic hypersensitivity reactions, hyperlipidaemia, abnormal lipoprotein patterns, aggregation of erythrocytes and peripheral neuropathy (39-44). One of the most recognized applications of CrEL is the pharmaceutical formulation of paclitaxel, a hydrophobic drug active against several murine tumors that has its development suspended for several years due to solubilization problems. Various studies have shown that CrEL influences the pharmacokinetics of many drugs including paclitaxel that presents a nonlinear disposition when formulated with this surfactant (38). Recently, Sparreboom et al. (45) proposed that the effect of CrEL on paclitaxel pharmacokinetics is associated with micellar solubilization, i.e. encapsulation of the drug within CrEL micelles, with the micelles acting as the principal carrier of paclitaxel in the systemic circulation. In paclitaxel I.V. infusions, an exceptionally large amount of CrEL is necessary, resulting in important biological events that can lead to serious acute hypersensitivity reactions and neurological toxicity. Therefore, large variety CrEL-free formulation vehicles for paclitaxel are currently in (pre)clinical development, including liposomes, nanocapsules and microspheres (46).

J Pharm Pharmaceut Sci (www.cspscanada.org) 8(2):147-163, 2005

Polymeric Micelles and Drug Delivery Long-circulating pharmaceuticals and drug carriers represent a growing area of medical and pharmaceutical research. There are several reasons for the search for long-circulating pharmaceuticals and drug carriers, such as: (i) Long-circulating particles may be used to maintain a required level of a pharmaceutical agent in the blood for extended time intervals for better drug availability. Moreover, long-circulating diagnostic agents are of primary importance for blood pool imaging (47). (ii) Long-circulating particles of nanoscopic size can slowly accumulate in pathological sites with affected and leaky vasculature (such as tumors, inflammations, and infarcted areas) and improve or enhance drug delivery in those areas. This phenomenon is usually called enhanced permeability and retention effect, EPR, known also as “passive” targeting or accumulation via an impaired filtration mechanism (48,49). (iii) Prolonged circulation can help to achieve a better targeting effect for specific ligand-modified drugs and drug carriers, since it increases the total quantity of targeted drug/carrier passing through the target, and the number of interactions between the drug and the target (50). As stated before, micellar systems present some advantages when compared to other drug carriers. For example, micelles can be obtained in an easy and reproducible manner in large scale and specific ligands can be attached to their outer surface in order to optimize the controlled releasing and specificity of pharmacological effect (7). Polymeric carriers might lead to precipitation in water, since the drug-polymer interaction can result in conversion of functional water-soluble groups of the drug into more hydrophobic groups. Micelles, on the other hand, offer a core/shell structure and, therefore, stay water-soluble (51). According to Kabanov et al. (52), the ideal self-assembling drug delivery system should spontaneously form from drug molecules, carrier components and targeting moieties; their size should be of around 10 nm in order to enable them to penetrate various tissues and even cells; they should be stable in vivo for a sufficiently 154

long period of time without provoke any biological reactions; should release the drug upon contact with target tissues/cells; and the components of the carrier (surfactant molecules) should be easily removed from the body when the therapeutic function is completed. A very important property of micelles is their size, which is normally around 5 to 100 nm, filling the gap between such drug carriers as individual macromolecules (antibodies, albumin, and dextran) with size below 5 nm, and particles such as liposomes and microcapsules with size of 50 nm and up. The most usual size of a pharmaceutical micelle is between 10 and 80 nm and the optimal cmc value should be in a low millimolar region. In drug delivery, special attention has been given to the so-called polymeric micelles (5,7,53-57). Polymeric micelles are formed from copolymers consisting of both hydrophilic and hydrophobic monomer units, such as PEO and PPO (polypropylene oxide), respectively. These amphiphilic block co-polymers with the length of the hydrophilic block exceeding the length of the hydrophobic block can form spherical micelles in aqueous solution. The micellar core consists of the hydrophobic blocks and the shell region consists of the hydrophilic blocks (53). The PEO coating has been shown to prevent opsonization and subsequent recognition by the macrophages of the reticuloendothelial system (RES), allowing the micelles to circulate longer and deliver drugs more effectively to the desired sites. (58). Another advantage of polymeric micelles refers to the ease of sterilization via filtration and safety for administration (59, 60). Figure 6 presents a schematic representation of the mechanism of polymeric micelles formation. As aforementioned, micelles are subject to extreme dilution upon intravenous injection into humans. However, the slow dissociation of kinetically stable polymeric micelles allows them to retain their integrity and perhaps drug content in blood circulation above or even below the cmc for some time, creating an opportunity to reach the target site before decaying into monomers (51,61). In addition, some polymeric micelles seems to present better solubilization capacity when compared to surfactant micelles due to the higher number of micelles and/or larger cores of the formers (62).

J Pharm Pharmaceut Sci (www.cspscanada.org) 8(2):147-163, 2005

Polymeric micelles made of poly(ethylene oxide)-bpoly(L-amino acid) (PEO-b-PLAA) has been suggested as synthetic analogs of natural carriers presenting a unique ability for chemical modification, since the free functional groups of PLAA blocks constitute sites to attach drugs. In addition, these PLAA blocks are of increasing interest once they may generate biocompatible monomers after hydrolysis and/or enzymatic degradation (61).

Figure 6: Formation of polymeric micelles from different types of amphiphilic block co-polymers (Extracted from Torchillin, 2001).

Besides the solubilization of a drug by physical encapsulation, polymeric micelles can be loaded with hydrophobic molecules that are conjugated or complexed with the polymeric backbone (63). In case of drug conjugates, there should be a cleavage (hydrolysis) of the covalent bond between drug and polymer. Therefore, the release may be dependent on the rate of micellar dissociation, since water diffusion into the hydrophobic micellar core must be restricted, resulting in a sustained drug release (64). Most studies and applications that have been conducted are based on block copolymers of PEO and PPO blocks, commercially known as Pluronics® (65). Studies for the solubilization of drugs such as haloperidol, indomethacin, doxorubicin (DOX), amphotericin B and digoxin have been reported (52, 66-70) and a parenteral formulation of DOX in these polymeric micelles has entered the phase I of clinical trials in Canada. More recently, biodegradable block copolymers with polyester core-forming structures have been developed. For example, micelles of PEO-poly(D,L-lactic acid-cocaprolactone) (PEO-PDLLA) have been used to encapsulate paclitaxel and shown similar in vitro toxicity, fivefold increase in maximum tolerable dose and increased efficacy after intraperitoneal injection in murine P388 leukemia model when compared to the standard formulation with Cremophor EL (71). 155

Yokoyama et col. studied PEO-b-poly(L-aspartic acid)DOX conjugates and, according to the results, the superiority of the block copolymer-drug conjugate over the free drug was a result of the lower toxicity of the former (72-75). Cisplatin (CIS) has also been complexed with PEO-b-poly(Asp), demonstrating increase in cytotoxic concentration against B16 melanoma cells and lower nefrotoxicity (76). In addition, a PEO-bpoly(α-glutamic acid)-CIS complex was investigated and presented greater stability, prolonged circulation in blood stream and improved accumulation in tumor site when compared to the previous complex (77). Despite the several block copolymer-drug conjugates studies, physical encapsulation of drugs within polymeric micelles offers a great alternative, since conjugation of the drug may lead to changes in the biological properties of the drug and consequently difficult the characterization and regulatory approval of the drug. However, physical encapsulation may present low capacity and/or rapid release of the encapsulated drug (51). Other alternative that emerges in the field of polymeric micelles refers to polyion complex micelles. Oppositely charged macromolecules, such as peptides and DNA, can complex with the charges of the side chains of some PLAA blocks resulting in the required amphiphilic character for micellization of the complex and leading to stabilization against digestive enzymes such as nucleases (51). These systems seem to be promising and have been receiving significant attention (7881). Recently, polymeric micelles incorporating CIS were prepared through polymer-metal complex formation between CIS and poly(ethylene glycol)-poly(glutamic acid) block copolymers, and showed remarkably prolonged blood circulation and effective accumulation in

J Pharm Pharmaceut Sci (www.cspscanada.org) 8(2):147-163, 2005

solid tumors (82). Other polyion complex micelles composed of a porphyrin dendrimer and PEG-bpoly(aspartic acid) were evaluated as new photosensitizers for photodynamic therapy in the Lewis lung carcinoma cell line and resulted in reduced dark toxicity of the cationic dendrimer porphyrin, probably due to the biocompatible PEG shell of the micelles (83). In another work, α-lactosyl-PEG-poly(2(dimethylamino)ethyl methacrylate) block copolymer (lactosePEG-PAMA) was synthesized to construct a polyion complex micelle-type gene vector potentially useful for selective transfection of hepatic cells, by spontaneously complexion with plasmid DNA encoding luciferase (pGL3-Luc). The lactose-PEG-PAMA-pDNA micelle revealed enhanced transfection compared to the control polyion complex micelle without the ligand (lactose) at a lower pDNA dose (84). One of the drawbacks of polyion complex micelles is the sensitivity to environment changes such as dilution and ionic strength. To overcome these, polymeric micelles prepared from PEG-poly(α,β,aspartic acid) and the cationic protein trypsin were cross-linked with glutaraldehyde through the Schift base formation, conferring stability to high salt concentrations and increasing the stability of the protein (85). Micelles, Biological Systems and Micellar Catalysis The study of cell membranes and of the roles it plays in living cells contributes significantly to the understanding of cellular function. Membranes have been shown to consist of lipids in association with proteins and glycoproteins (16). The present accepted model of a biomembrane is that the phospholipids are organized in a bilayer structure, resulting in a fluid lipid matrix of varying composition and fluidity. Embedded in this matrix are the integral proteins that are able to undergo lateral and rotational diffusion. A wide variety of lipids is found in biological membranes, with the phospholipids being among the most common (86). Many biological processes occur at membrane surfaces or within their hydrophobic moiety. Owing to the ionic head groups of the lipids, the surface of biological membranes frequently presents a net charge, giving rise to different binding properties of charged and uncharged forms of molecules such as drugs (87-88). In 156

this sense, the relationship between the binding properties of a drug and its active form, as well as its membrane location, deserves attention. Despite the effort aiming at an understanding of drugs mechanism of action at the molecular level, demonstrate by the number of studies on the interaction of drugs with biological membranes, more studies involving model systems are necessary (87,89). Surfactants have a far-ranging use in membrane studies. Because surfactants are amphiphilic molecules, like lipids, some of the same rules governing lipid behavior also apply to the surfactants. Among the membrane models utilized, micellar systems can be considered an interesting alternative to study the interactions of different compounds with membranes because of the relative simplicity of these systems, and therefore have been used with this purpose (13,20). Reactions behavior observed at surfactant interfaces are expected to be more representative of many biological reactions than are reactions studied in dilute aqueous solutions (90). In this sense, micellar catalysis of reactions is important because of the parallel with enzymes behavior. Catalysis by both normal micelles and reversed micelles is possible. In normal micelles in aqueous medium, enhanced reaction of the solubilized substrate generally occurs at the micelle-water interface; in reversed micelles in non-polar medium this reaction occurs deep in the inner core (17). Micellar catalysis in aqueous solution is generally explained in terms of distribution of reactants between water and micelles, with reactions occurring in both media. Therefore, it is possible to treat the rate-surfactant profiles in terms of the concentrations of reactants in the aqueous and micellar pseudo-phases and the rate constants in each pseudo-phase (91). There are different kinetic models to explain micellar catalytic effects in aqueous medium (92-95). In the pseudo-phase kinetic model (92), the kinetics of a nth order reaction is analyzed by considering the partitioning of the reactants between the two pseudo-phases. The reactants (A and B) may be distributed as shown in Figure 7.

J Pharm Pharmaceut Sci (www.cspscanada.org) 8(2):147-163, 2005

In this model, no distinction is made between the various regions of the micelles, although reactions generally occur in the Stern layer, at the micelle/water interface, rather than in the hydrocarbon-like core of the micelle. Nevertheless, the pseudo-phase model explains many features of micellar rate effects and it can be applied, at least qualitatively, to a variety of reactions in colloidal assemblies (96).

Figure 7: Schematic representation of a micellar catalyzed reaction according to the pseudo-phase kinetic model. AW, BW and AM, BM correspond to the concentrations of the reactants in the aqueous (W) and micellar (M) phases; KA and KB are the biding constants of the reactants to the micelles, and kW and kM are the rate constants in the aqueous and micellar paths.

Therefore, a quantitative rate expression for a bimolecular reaction can be given by the following equation: (6)

where kexp is the observed second-order rate constant, PA and PB are the partition coefficients of the reactants A and B, respectively; C is the total surfactant molar concentration minus the cmc; V is the partial molar volume of the surfactant in the micelle and, therefore, CV and (1 – CV) stand for the volume fractions of the micellar and aqueous phases, respectively. The binding constants (KA and KB) are related to the partition coefficient (P), as follows: (7)

For dilute surfactant solutions, where the volume fraction of the micellar phase is small, 1 >> CV and Eq. (6) can be simplified to: (8)

Utilizing Eq. (7), we can rewrite Eq. (8) as follows: (9)

157

Since the binding constant K depends on the extent of hydrophobic bonding between surfactant and substrate, it can be expected that K will increase with increase in the chain length of both the surfactant and the substrate. However, if the hydrophobic group of the substrate is too long, it may be solubilized so deeply in the micelle that access to its reactive site by a reagent in aqueous solution phase is hindered and, therefore, solubilization will inhibit the reaction (17). The charge of surfactant head group also influences the catalytic power of micelles. Thus, catalysis of some nucleophilic aromatic substitution reactions is more pronounced by dicationic micellar surfactants than by cationic micellar hexadecyltrimethylammonium bromide (96). Yu et al. (97) observed that cationic micelles inhibit, anionic micelles accelerate and nonionic micelles show no appreciable effect on metal ion hydrolysis of p-nitrophenyl picolinate. The higher the electron charge of metal ions, the greater these effects are, indicating that electrostatic interactions are the major contribution for this reaction in micellar solution. In another work, it was observed an increase in the oxidation rate of L(+)arabinose by chromic acid with the addition of SDS and Triton X-100 concentrations. On the other hand, the addition of ammonium, lithium and sodium bromides in SDS micelles resulted in rate decrease (98). Plots of rate constant versus surfactant concentration often show a maximum at some surfactant concentration above the CMC. One of the reasons for this is that the number of micelles increases with increase in the surfactant concentration. When the number of micelles exceeds that required to solubilize all of the substrate there is a dilution of the substrate concentration per micelle with further increase in surfactant concentration, leading to a reduction in the rate constant. Moreover, the charge surface of an ionic micelle in aqueous solution may cause not only the concentra-

J Pharm Pharmaceut Sci (www.cspscanada.org) 8(2):147-163, 2005

tion of an oppositely charged reactant at the micellesolution interface, but the adsorption of that reactant on it or even the solubilization into micelles, resulting in a decrease in the reactant activity in the solution phase. Therefore, an increase in surfactant concentration over that required to complete solubilization of the substrate may result in a decrease in the rate constant (17). There is strong evidence to believe that most micellecatalyzed reactions occur on the surface of the ionic micelles, at or near the charged double layer that surrounds the hydrocarbon core. Typically, reactions between very hydrophobic substrates and hydrophilic anions seem to have lower second-order rate constants in the micellar pseudo phase than in water, because the anions are located in the Stern layer at the micelle/ water interface whereas the substrate may be, on average, more deeply in the micelle (96). Micelles also allow co-solubilization of compounds of very different hydrophobic and hydrophilic character and, as a result, chemical reactions can be developed which otherwise would proceed only with difficulty. An example is the formation of o-phthaldehyde adducts from water-insoluble amines of high molecular mass (99). The presence of micelles can also result in the formation of different reaction products. A diazonium salt in an aqueous micellar solution of sodium dodecyl sulfate, for example, yielded the corresponding phenol from reaction with OH- in the bulk phase, but the corresponding hydrocarbon from material solubilized in the micelles (100). One interesting approach refers to functional surfactants, which are surfactants containing a reactive residue, usually at the head group, that can be micellized or co-micellized with a chemically inert, non-functional surfactant. In this sense, micelles functionalized with groups that model the amino acid side chains responsible for enzyme activity are generally impressive catalysts (96). Final Considerations Considering the importance of micellar systems in the pharmaceutical field and the many applications that it presents, our group have been carrying research on this matter, with special attention to the solubilization of 158

drugs in aqueous micellar solutions. We study the solubilization of model drugs, such as the non-steroidal anti-inflammatory ibuprofen, as well as of potential drugs such as p-substituted benzhydrazides compounds in solutions of different surfactants. Recently, we investigated the solubilization of ibuprofen (IBU) in micellar solutions of three surfactants possessing the same hydrocarbon tail but different hydrophilic head groups, namely sodium dodecyl sulphate (SDS), dodecyltrimethylammonium bromide (DTAB), and n-dodecyl octa(ethylene oxide) (C12E8) (101). The results obtained showed that, irrespective of the surfactant type, the solubility of IBU increases linearly with increasing surfactant concentration, because of the association between the drug and the micelles. Nonionic surfactants were shown to provide a combination of good molar solubilization capacity and high micellar concentration, due to their low cmc, resulting in increased solubility of IBU. On should keep in mind that the low toxicity of nonionic surfactants makes them particularly interesting for solubilization and drug delivery purposes. In addition to these studies, Small Angle X-ray Scattering (SAXS) studies on the interaction of ibuprofen with micelles of SDS, DTAB and C12E8 are in progress, aiming at a deeper understanding on the nature of these interactions as well as on the properties of the aggregates obtained. The p-substituted benzhydrazides are hydrophobic compounds synthesized by Taveres et col. that represent potential drugs with anti-staphylococci and antitrypanosome activities. Quantitative Structure-Activity Relationships (QSAR) studies and experimental results have shown a dependency between biological activity and hydrophobicity of these compounds, with the more hydrophobic molecules presenting higher activity (102,103). However, the more hydrophobic the compound, the more difficult the solubilization in the culture media used for activity determination. Therefore, the possibility of micellar solubilization of these molecules should contribute to more precise determinations of biological activity. We are currently carrying on experiments on the solubilization of p-substituted benzhydrazides in aqueous micellar solutions of n-dodecyl octaethylene oxide (C12E8) and n-hexadecyl octaethylene oxide (C16E8), two nonionic surfactants possessing the same hydrophilic head groups but different hydrophobic tails.

J Pharm Pharmaceut Sci (www.cspscanada.org) 8(2):147-163, 2005

ACKNOWLEDGEMENTS Carlota Rangel-Yagui is grateful for the Post-Doctoral fellowship and financial support from FAPESP (Fundação de Amparo à Pesquisa no Estado de São Paulo). We acknowledge CAPES (Coordenação de Aperfeiçoamento de Pessoal de Nível Superior –Brazil) and CNPq (Conselho Nacional de Desenvolvimento Científico e Tecnológico – Brazil) for financial support. REFERENCES [1]

[2]

[3]

[4]

[5]

[6]

[7]

[8] [9] [10]

[11]

[12]

[13]

[14]

[15]

Mall, S., Buckton, G., Rawlins, D.A. Dissolution behaviour of sulphonamides into sodium dodecyl sulphate micelles: A thermodynamic approach. J Pharm Sci, 85(1):75-78, 1996. Allen, T.M., Hansen, C.B., Menenez, D.E.L. Pharmacokinetics of long-circulating liposomes. Adv. Drug Deliv Rev, 16:267–284, 1995. Canto, G.S., Dalmora, S.L., Oliveira, A.G. Piroxicam encapsulated in liposomes: characterization and in vivo evaluation of topical anti-inflammatory effect. Drug Dev Ind Pharm, 25:1235-1239, 1999. Gref, R., Minamitake, Y., Peracchia, M.T., Trubetskoy, V.S., Torchilin, V.P., Langer, R. Biodegradable long-circulating polymeric nanospheres. Science, 263:1600–1603, 1994. Jones, M.C., Leroux, J.C. Polymeric micelles — a new generation of colloidal drug carriers. Eur J Phar Bio Pharm, 48:101–111, 1999. Torchilin, V.P., Trubetskoy, V.S., Whiteman, K.R., Caliceti, P., Ferruti, P., Veronese, F.M. New synthetic amphiphilic polymers for steric protection of liposomes in vivo. J Pharm Sci, 84:1049-1053, 1995. Torchilin, V.P. Structure and design of polymeric surfactant-base drug delivery systems. J Control Rel, 73:137-172, 2001. Martin, A. Physical Pharmacy, 4ed., Williams and Wilkins, Baltimore, USA, pp 396–398, 1993. Hunter, R.J. Introduction to Modern Colloid Science. Oxford University Press, Oxford, 1993. Ford, J.M., Hait, W.N. Pharmacology of drugs that alter multidrug resistance in cancer. Pharmacol Rev, 42(3):155-199, 1990. Chakraborty, H., Banerjee, R., Sarkar, M. Incorporation of NSAIDS in micelles: implication of structural switchover in drug-membrane interaction. Biophys Chem, 104(1):315-325, 2003. Ferreira, H., Lucio, M., De Castro, B., Gameiro, P., Lima, J.L.F.C., Reis, S. Partition and location of nimesulide in EPC liposomes: a spectrophotometric

159

[16] [17] [18] [19]

[20] [21]

[22]

[23]

[24]

[25]

[26]

[27]

and fluorescence study. Anal Bioanal Chem, 377(2):293-298, 2003. Fresta, M., Guccione, S., Beccari, A.R., Furneri, P.M., Puglisi, G. Combining molecular modeling with experimental methodologies: Mechanism of membrane permeation and accumulation of ofloxacin. Bioorg Med Chem, 10(12):3871-3889, 2002. Israelachvili, J.N., Mitchell, D.J., Ninham, B.W. Theory of self-assembly of hydrocarbon amphiphiles into micelles and bilayers. J Chem Soc Far Trans II, 72:1525-1568, 1976. Suwalsky, M., Hernandez, P., Villena, F., Aguilar, F., Sotomayor, C.P. Interaction of the anticancer drug tamoxifen with the human erythrocyte membrane and molecular models. Zeitschrift fur Naturforschung C - J Biosciences, 53(3-4):182-190, 1998. Jones, M.N., Chapman, D. Micelles, monolayers and biomembranes. Wiley-Liss, New York, 1995. Rosen, M.J. Surfactants and interfacial phenomena, 2ed., John Wiley & Sons, New York, 1989. Chevalier, Y., Zemb, T. The structure of micelles and microemulsions. Rep Prog Phys, 53:279-371, 1990. Tanford, C. The hydrophobic effect: Formation of micelles and biological membranes. Wiley, New York, 1980. Israelachvili, J.N. Intermolecular and surface forces, 2ed., Academic Press, London, 1991. Hiemenz, P.C., Rajagopalan, R. Principles of colloid and surface chemistry, 3ed. Marcel Decker, New York, 1997. Puvvada, S., Blankschtein, D. Molecular-thermodynamic approach to predict micellization, phase behavior, and phase separation of micellar solutions. I. Application to nonionic surfactants. J Chem Phys, 92(6):3710-3724. 1990. Florence, A.T., Attwood, D. Physicochemical Principles of Pharmacy, 3rd ed. The MacMillan Press, London, 2003. Attwood, D., Florence, A.T. Surfactant systems: Their chemistry, pharmacy and biology. Chapman and Hall, New York, 1983. Alvarez-Núñez, F.A., Yalkowsky, S.H. Relationship between polysorbate 80 solubilization descriptors and octanol-water partition coefficient of drugs. Int J Pharm, 200:217-222, 2000. Yokoyama, M. Block copolymers as drug carriers. CRC Crit Rev Ther Drug Carrier Syst, 9:213-248, 1992. Dutt, G.B. Rotational diffusion of hydrophobic probes in Brij-35 micelles: Effect of temperature on

J Pharm Pharmaceut Sci (www.cspscanada.org) 8(2):147-163, 2005

[28]

[29]

[30]

[31]

[32]

[33]

[34] [35]

[36]

[37]

[38]

[39]

[40]

micellar internal environment. J Phys Chem B, 107:10546-10551, 2003. Mukerjee, P., Cardinal, J.R. Benzene derivatives and naphthalene solubilized in micelles. Polarity of microenvironment, location and distribution in micelles, and correlation with surface activity in hydrocarbon-water systems. J Phys Chem, 82(14):1620-1627, 1978. Krishna, A.K., Flanagan, D.R. Micellar solubilization of a new antimalarial drug, β-arteether. J Pharm Sci, 78(7):574-576,1989. Barry, B.W., El Eini, D.I.D. Solubilization of hydrocortisone, dexamethasone, testosterone and progesterone by long-chain polyoxyethylene surfactants. J Pharm Pharmac, 28:210-218, 1976. Ong, J.T.H., Manoukian, E. Micellar solubilization of timbesone acetate in aqueous and aqueous propylene glycol solutions of nonionic surfactants. Pharm Res, 3(11):704-708, 1988. Alkhamis, K.A., Allaboun, H., Al-Momani, W.Y. Study of the solubilization of gliclazide by aqueous micellar solutions. J Pharm Sci, 92(4):839-846, 2003. Li, P., Tabib, E., Yalkowsky, S.H. Solubilization of ionized and un-ionized flavopiridol by ethanol and polysorbate 20. J Pharm Sci, 88(5):507-509, 1999. Li, P., Zhao, L. Solubilization of flurbiprofen in pHsurfactant solutions. J Pharm Sci, 92(5):951-956, 2003. Zografi, G., Auslander, D.E. Surface activity of chlorpromazine and chlorpromazine sulfoxide in the presence of insoluble monomolecular films. J Pharm Sci, 54(9):1313-1318, 1965. Caetano, W., Gelamo, E.L., Tabak, M., Itri R. Chlorpromazine and sodium dodecyl sulphate mixed micelles investigated by Small Angle X-Ray Scattering. J Colloid Interf Sci, 248:149-157, 2002. Palma, S., Manzo, R.H., Allemandi, D., Fratoni, L., Lo Nostro, P. Drug solubilization in ascorbyldecanoate micellar solutions. Colloid Surf A, 212:163173, 2003. Gelderblom, H., Verweij, J., Nooter, K., Sparreboom, A. Cremophor EL: the drawbacks and advantages of vehicle selection for drug formulation. Eur J Cancer, 37:1590-1598, 2001. Adams, J.D., Flora, K.P., Goldspiel, B.R., Wilson, J.W., Arbuck, S.G.I. Taxol: a history of pharmaceutical development and current pharmaceutical concerns. J Natl Cancer Inst Monogr, 15:141-147, 1993. Watkins, J., Ward, A.M., Appleyard, T.N. Adverse reactions in intravenous anaesthetic induction agents. Br Med J, 2:1084-1085, 1997.

160

[41]

[42]

[43]

[44]

[45]

[46]

[47]

[48]

[49]

[50]

[51]

[52]

[53]

Huttel, M.S., Olesen, A.S., Stoffersen, E. Complement-mediated reactions to diazepam with cremophor as solvent (steroid MR). Br J Anaesth 52(1):7779, 1980. Kongshaug, M., Cheng, L.S., Moan, J., Rimington, C. Interaction of Cremophor-EL with human plasma. Int J Biochem, 23(4):473-478, 1991. Woodburn, K., Kessel, D. The alteration of plasmalipoproteins by Cremophor EL. J Photoch Photobiol B, 22(3):197-201, 1994. Windebank, A.J., Blexrud, M.D., Degroen, P.C. Potential neurotoxicity of the solvent vehicle for cyclosporine. J Pharmacol Exp Ther, 268(2):1051-1056, 1994. Sparreboom, A., van Zuylen, L., Brouwer, E., Loos, W.J., de Bruijn, P., Gelderblom, B., Pillay, M., Nooter, K., Stoter, G., Verweij, J. Cremophor ELmediated alteration of paclitaxel distribution in human blood: Clinical pharmacokinetic implications. Cancer Res, 59(7):1454-1457, 1999. van Zuylen, L., Verweij, J., Sparreboom, A. Role of formulation vehicles in taxane pharmacology. Inv New Drug, 19(2):125-141, 2001. Torchilin, V.P., Trubetskoy, V.S. Which polymers can make nanoparticulate drug carriers long-circulating? Adv Drug Deliv Rev, 16:141–155, 1995. Palmer, T.N., Caride, V.J., Caldecourt, M.A., Twickler, J., Abdullah, V. The mechanism of liposome accumulation in infarction. Biochim Biophys Acta,797:363–368, 1984. Maeda, H., Wu, J., Sawa, T., Matsumura, Y., Hori, K. Tumor vascular permeability and the EPR effect in macromolecular therapeutics: a review. J Contr Rel, 65:271–284, 2000. Torchilin, V.P. How do polymers prolong circulation time of liposomes? J Liposome Res, 6:99–116, 1996. Lavasanifar, A., Samuel, J., Kwon, G.S. Poly(ethylene oxide)-block-poly(L-amino acid) micelles for drug delivery. Adv Drug Deliv Rev, 54:169-190, 2002. Kabanov, A.V., Batrakova, E.V., Melik-Nubarov, N.S., Fedoseev, N.A., Dorodnich, T.Y., Alakhov, V.Y., Chekhonin, V.P., Nazarova, I.R., Kabanov, V.A. A new class of drug carriers; micelle poly(oxyethylene)-poly(oxypropylene) block copolymers as microcontainers for drug targeting from blood to brain. J Contr Rel, 22:141–158, 1992. Kwon, G.S., Kataoka, K. Block copolymer micelles as long-circulating carriers for delivery of therapeutic and diagnostic agents. Adv Drug Deliv Rev, 16:295309, 1995.

J Pharm Pharmaceut Sci (www.cspscanada.org) 8(2):147-163, 2005

[54]

[55]

[56]

[57]

[58]

[59]

[60] [61]

[62]

[63]

[64]

[65] [66]

Inoue, T., Chen, G.H., Nakamae, K., Hoffman, A.S. An AB block copolymer of oligo(methyl methacrylate) and poly(acrylic acid) for micellar delivery of hydrophobic drugs. J Control Rel 51(2-3):221-229, 1998. Discher, B.M., Won, Y.Y., Ege, D.S., Lee, J.C.M., Bates, F.S., Discher, D.E., Hammer, D.A. Polymersomes: Tough vesicles made from diblock copolymers. Science, 284(5417):1143-1146, 1999. Rapoport, N., Marin, A., Luo, Y., Prestwich, G.D., Muniruzzaman, M. Intracellular uptake and trafficking of pluronic micelles in drug-sensitive and MDR cells: Effect on the intracellular drug localization. J Pharm Sci, 1:157-170, 2002. Soo, P.L., Luo, L., Maysinger, D., Eisenberg, A. Incorporation and release of hydrophobic probes in biocompatible polycaprolactone-block-poly(ethylene oxide) micelles: implications for drug delivery. Langmuir, 18:9996-10004, 2002. Lee, J.H., Lee, H.B., Andrade, J.D. Blood compatibility of polyethylene oxide surfaces. Prog Polym Sci, 20(6):1043-1079, 1995. Kwon, G.S., Kataoka, K. Block copolymer micelles as long circulating drug vehicles. Adv Drug Deliv Rev, 16:295–309, 1995. Kwon, G.S., Okano, T. Polymeric micelles as new drug carriers. Adv Drug Deliv Rev, 21:107–116, 1996. Adams, M.L., Lavasanifar, A., Kwon, G.S. Amphiphilic block copolymers for drug delivery. J Pharm Sci, 92(7):1343-1355, 2003. Gadelle, F., Koros, W.J., Schechter, R.S. Solubilization of aromatic solutes in block copolymers. Macromolecules, 28:4883–4892, 1995. Kataoka, K., Harada, A., Nagasaki, Y. Block copolymer micelles for drug delivery: design,characterization and biological significance. Adv Drug Deliv Rev, 47:113–131,2001. Li, Y., Kwon, G.S. Methotrexate esters of poly(ethylene oxide)-block-poly(2-hydroxyethyl- -aspartamide). Part I: effects of the level of methotrexate conjugation on the stability of micelles and on drug release. Pharm Res, 17:607–611, 2000. Mortensen, K. PEO-related block copolymer surfactants. Coll Surf A, 183-185:277-292, 2001. Kabanov, A.V., Chekhonin, V.P, Alakhov, V.,. Betrakova, E.V, Lebedev, A.S., Mellik-Nubarov, N.S., Arzakov, S.A., Levashov, A.V., Morozov, G.V., Severin, E.S., Kabanov, V.A. The neuroleptic activity of haloperidol increases after its solubilization in surfactant micelles. FEBS Lett, 258:343–345, 1989.

161

[67]

[68]

[69]

[70]

[71]

[72]

[73]

[74]

[75]

[76]

[77]

Lin, S., Kawashima, Y. Kinetic studies on the stability of indomethacin in alkaline aqueous solutions containing poly(oxyethylene)poly(oxypropylene) surface active block copolymers. Pharm Acta Helv, 60:345–350, 1985. Baratkova, E.V., Dorodnich, T.Y., Klinskii, E.Y., Kliushnenkova, E.N., Shemchukova, O.B., Goncharova, O.N., Arjakov, V.Y., Alakhov, V., Kabanov, A.V. Anthracycline antibiotics noncovalently incorporated into the block copolymer micelles: in vivo evaluation of anticancer activity. Br J Cancer, 74:1545–1552, 1996. Forster, D., Washington, C., Davis, S.S. Toxicity of solubilized and colloidal amphotericin B formulations to human erythrocytes. J Pharm Pharmacol, 40:325–328, 1988. Baratkova, E.V., Miller, D.W., Li, S., Alakhov, V., Kabanov, A.V., Elmquist, W.F. Pluronic P85 enhances the delivery of digoxin to the brain: in vitro and in vivo studies. J Pharmacol Exp Ther, 269:551– 557, 2001. Zhang, X., Burt, H.M., Von Hoff, D., Dexter, D., Mangold, D., Degen, D., Oktaba, M., Hunter, W.L. An investigation of the antitumour activity and biodistribution of polymeric micellar paclitaxel. Cancer Chemother Pharmacol, 40:81–86, 1997. Yokoyama, M., Kwon, G.S., Okano, T., Sakurai, Y., Sato, T., Kataoka, K. Preparation of micelle-forming polymer–drug conjugates. Bioconjug Chem, 3:295– 301, 1992. Kwon, G.S., Yokoyama, M., Okano, T., Sakurai, Y., Kataoka, K. Biodistribution of micelle-forming polymer–drug conjugates. Pharm Res, 10:970–974, 1993. Kataoka, K., Kwon, G.S., Yokoyama, M., Okano, T., Sakurai, Y. Polymeric micelles as novel drug carriers and virus mimicking vehicles. In: J. Kahovec (Ed), Macromolecules, 1992(VSP):267–276, 1993. Yokoyama, M., Okano, T., Sakurai, Y., Ekimoto, H., Shibazaki, C., Kataoka, K. Toxicity and antitumor activity against solid tumors of micelle-forming polymeric anticancer drug and its extremely long circulation in blood. Cancer Res, 51:3229–3236, 1991. Mizumura, Y., Matsumura, Y., Hamaguchi, T., Nishiyama, N., Kataoka, K., Kawaguchi, T., Saito, T., Kakizoe, T. Cisplatin-incorporated polymeric micelles reducing nephrotoxicity, while maintaining antitumor activity. Jpn J Cancer Res, 92:328–336, 2001. Nishiyama, N., Kato, Y., Sugiyama, Y., Kataoka, K. Development of cisplatin-loaded polymeric micelles with a prolonged circulation in the bloodstream and an enhanced accumulation in the solid tumor. Pro-

J Pharm Pharmaceut Sci (www.cspscanada.org) 8(2):147-163, 2005

[78]

[79]

[80]

[81]

[82]

[83]

[84]

[85]

[86] [87]

ceedings of the 20th International Symposium on Controlled Release of Bioactive Compounds, 28:5155–5156, 2001. Kabanov, A.V., Kabanov, V.A. Interpolyelectrolyte and block ionomer complexes for gene delivery: physico-chemical aspects. Adv Drug Deliv Rev, 30:49–60, 1998. Lemieux, P., Vinogradov, S.V., Gebhart, C.L., Guerin, N., Paradis, G., Nguyen, H.K., Ochietti, B., Suzdaltseva, Y.G., Baratkova, E.V., Bronich, T.K., St-Pierre, Y., Alakhov, V.Y., Kabanov, A.V. Block and graft copolymers and nanogel copolymer networks for DNA delivery into cell. J Drug Target, 8: 91–105, 2000. Itaka, K., Yamauchi, K., Harada, A., Nakamura, K., Kawaguchi, H., Kataoka, K. Polyion complex micelles from plasmid DNA and poly(ethylene glycol)-poly(L-lysine) block copolymer as serum tolerable polyplex system: physicochemical properties of micelles relevant to gene transfection efficiency. Biomaterials, 24(24):4495-4506, 2003. Berret, J.F., Vigolo, B., Eng, R., Herve, P., Grillo, I., Yang, L. Electrostatic self-assembly of oppositely charged copolymers and surfactants: A light, neutron, and X-ray scattering study, Macromolecules, 37(13):4922-4930, 2004. Nishiyama, N., Okazaki, S., Cabral, H., Myiamoto, M., Kato, Y., Sugiyama, Y., Nishio, K., Matsumura, Y., Kataoka, K. Novel cisplatin-incorporated polymeric micelles can eradicate solid tumours in mice. Cancer Res, 63(24):8977-8983, 2003. Zhang, G.D., Harada, A., Nishiyama, N., Jiang, D.L., Koyama, H., Aida, T., Kataoka, K. Polyion complex micelles entrapping cationic dendrimer porphyrin: effective photosensitizer for photodynamic therapy of cancer. J Control Rel, 93(2):141-150, 2003. Wakebayashi, D., Nishiyama, N., Yamasaki, K., Itaka, K., Kanayama, N., Harada, A., Nagasaki, Y., Kataoka, K. Lactose-conjugated polyion complex micelles incorporating plasmid DNA as a targetable gene vector system: their preparation and gene transfecting efficiency against cultured HepG2 cells. J Control Rel, 95(3):653-664, 2004. Jaturanpinyo, M., Harada, A., Yuan, X.F., Kataoka, K. Preparation of bionanoreactor based on core-shell structured polyion complex micelles entrapping trypsin in the core cross-linked with glutaraldehyde. Bioconjugate Chem, 15(2):344-348, 2004. Yeagle, P.L. The membranes of cells, 2nd ed. New York: Academic Press, 1993. Caetano, W., Tabak, M. Interaction of chlorpromazine and trifluoperazine with ionic micelles: elec162

[88]

[89]

[90]

[91]

[92]

[93]

[94]

[95]

[96]

[97]

[98]

[99]

tronic absorption spectroscopy studies. Spectrochim. Acta A. 55(12):2513-2528, 1999. Louro, S.R.W., Nascimento, O.R., Tabak, M. Charge-dependent and ph-dependent binding-sites for dibucaine in ionic micelles - a fluorescence study. Biochim Biophys Acta, 1190:319-328, 1994. Yushmanov, V.E., Perussi, J.R., Imasato, A.C., Rugiero, M. Tabak, M. Ionization and binding equilibria of papaverine in ionic micelles studied by H-1-NMR and optical-absorption spectroscopy Biophys Chem, 52:157-163, 1994. Price, S.E., Jappar, D., Lorenzo, P., Saavedra, J.E., Hrabie, J.A., Davies, K.M. Micellar catalysis of nitric oxide dissociation from diazeniumdiolates. Langmuir, 19:2096-2102, 2003. Arikan, B., Tunçay, M. Micellar effects and reactant incorporation in reduction of toluidine blue by ascorbic acid. Dyes and Pigments. 64:1-8, 2005. Martinek, K., Yatsimirskii, A.K. Levashov, A.V., Berezin, I.V. In: Micellization, Solubilization and Microemulsions, Ed: Mittal, K.L., Plenum, New York, vol.2, p.489, 1977. Menger, F.M., Portnoy, C.E. On chemistry of reactions proceeding inside molecular aggregates. J Am Chem Soc, 89(18): 4698-&, 1967. Piszkiewicz, D. Cooperativity in bimolecular micelle-catalyzed reactions - inhibition of catalysis by high-concentrations of detergent. J Am Chem Soc, 99(23):7695-7697, 1977. Dash, A.C., Prusti, J., Pradhan, J., Das, P.K. Micellar effects upon the reactions of complex-ions in solution .4. Kinetics of aquation and base hydrolysis of some cis-(chloro)(amine)bis (ethylenediamine)cobalt(III) complexes in the presence of neutral and anionic surfactants in an aqueous-medium. J Chem Soc Far Trans, 86(3):507-510, 1990. Carreto, M.L., Rubio, S., Pérez-Bendito, D. Organic microheterogeneous systems in kinetic analysis. Self assembled systems – A review. Analyst, 121:33R-44R, 1996. Yu, X.Q., Jiang, B.Y., Cheng, S.Q., Huang, Z., Zeng, X.C. Comparative reactivities of metal cation-catalyzed hydrolysis of p-nitrophenyl picolinate in micellar solutions. J Disp Sci Technol, 24(6):761-765, 2003. Kabir-ud-Din, Morshed, A., Khan, Z. Oxidative degradation of L(+)arabinose by chromium (VI) in absence and presence of sodium dodecyl sulphate and TX-100 micelles. Oxid Commun, 26(1):59-71, 2003. Memon, M.H., Worsfold, P.J. The use of microemulsions in flow-injection analysis - spectrofluorimetric

J Pharm Pharmaceut Sci (www.cspscanada.org) 8(2):147-163, 2005

[100]

[101]

[102]

[103]

determination of primary amines. Anal Chim Acta, 183:179-185, 1986. Abe, M. Suzuki, N., Ogino, K. Replacement reaction of diazonium salt in the presence of sodium dodecylsulfate micelles. J Coll Interface Sci, 93:285-288, 1983. Rangel-Yagui, C. O., Hsu, H.W.L., Pessoa-Jr, A., Tavares, L.C., Micellar solubilization of ibuprofen – The influence of surfactant head on the extent of solubilization. RBCF – Braz J Pharm Sci, 2004. (In Press) Furlanetto, M., Masunari, A., Tavares, L. C. Validation of carbonyl group absorption frequency , Vc=o, as a structural descriptor for QSAR/QSPR studies In: 15th European Symposium on Quantitative Structure-Activity Relationships & Molecular Modeling, 2004, Istanbul. Program & Abstracts , p.142, 2004. Masunari, A., Rezende, P., Tavares, L. C. QSAR studies of nifuroxazide analogs with antimicrobial activity against multidrug-resistant Staphylococcus aureus In: 15th European Symposium on Quantitative Structure-Activity Relationships & Molecular Modeling, 2004, Istanbul. Program & Abstract, p.123, 2004.

163

07/11/12

Lyotropic liquid crystal - Wikipedia, the free encyclopedia

Lyotropic liquid crystal From Wikipedia, the free encyclopedia

A liquid crystalline material is called lyotropic if phases having long-ranged orientational order are induced by the addition of a solvent. Historically the term was used to describe materials composed of amphiphilic molecules. Such molecules comprise a water-loving 'hydrophilic' head-group (which may be ionic or non-ionic) attached to a water-hating 'hydrophobic' group. Typical hydrophobic groups are saturated or unsaturated hydrocarbon chains. Examples of amphiphilic compounds are the salts of fatty acids, phospholipids. Many simple amphiphiles are used as detergents.

Amphiphile Self-Assembly

A highly viscous cubic phase gel made of polysorbate 80, water, and liquid paraffin.

Amphiphilic molecules form aggregates through a self-assembly process that is driven by the 'hydrophobic effect' when they are mixed with a solvent, which is usually water. The aggregates formed by amphiphilic molecules are characterised by structures in which the hydrophilic head-groups shield the hydrophobic chains from contact with water. For most lyotropic systems aggregation occurs only when the concentration of the amphiphile exceeds a critical concentration (known variously as the 'critical micelle concentration' (CMC) or the 'critical aggregation concentration (CAC)'). Micellar solutions are often denoted by the symbol L1. Above the CMC (or CAC) the self-assembled amphiphile aggregates exist as independent entities, in equilibrium with monomeric amphiphiles in solution, and with no long ranged orientational or positional (translational) order. These dispersions are generally referred to as 'micellar solutions', the constituent aggregates being known as 'micelles', and are 'isotropic' phases (i.e. not liquid crystalline). True lyotropic liquid crystalline phases are formed as the concentration of amphiphile in water is increased beyond the point where the micellar aggregates are forced to be disposed regularly in space. For amphiphiles that consist of a single hydrocarbon chain the concentration at which the first liquid crystalline phases are formed is typically in the range 25-30 wt%.

Liquid Crystalline Phases and Composition/Temperature The simplest liquid crystalline phase that is formed by spherical micelles is the 'micellar cubic', denoted by the symbol I1. This is a highly viscous, optically isotropic phase in which the micelles are arranges on a cubic lattice. At higher amphiphile concentrations the micelles fuse to form cylindrical aggregates of indefinite length, and these cylinders are arranged on a long-ranged hexagonal lattice. This lyotropic liquid crystalline phase is known as the 'hexagonal phase', or more specifically the 'normal topology' hexagonal phase and is generally denoted by the symbol HI. At higher concentrations of amphiphile the 'lamellar phase' is formed. This phase is denoted by the symbol Lα. This phase consists of amphiphilic molecules arranged in bilayer sheers separated by layers of water. Each bilayer is a prototype of the arrangement of lipids in cell membranes. For most amphiphiles that consist of a single hydrocarbon chain, one or more phases having complex architectures are formed at concentrations that are intermediate between those required to form a hexagonal phase and those that lead to the formation of a lamellar phase. Often this intermediate phase is a bicontinuous cubic phase.

en.wikipedia.org/wiki/Lyotropic_liquid_crystal

1/3

07/11/12

Lyotropic liquid crystal - Wikipedia, the free encyclopedia

Schematic showing the aggregation of amphiphiles into micelles and then into lyotropic liquid crystalline phases as a function of amphiphile concentration and of temperature. In principle, increasing the amphiphile concentration beyond the point where lamellar phases are formed would lead to the formation of the inverse topology lyotropic phases, namely the inverse cubic phases, the inverse hexagonal phase (HII) and the inverse micellar cubic phase. In practice inverse topology phases are more readily formed by amphiphiles that have at least two hyrocarbon chains attached to a headgroup. The most abundant phospholipids that are found in cell membranes of mammalian cells are examples of amphiphiles that readily form inverse topology lyotropic phases.

References Laughlin R.G. (1996). The Aqueous Phase Behaviour of Surfactants. London: Academic Press. ISBN 0-12437760-2. Fennell Evans D. and Wennerström H. (1999). The Colloidal Domain. New York: Wiley VCH. ISBN 0-47124247-0. Retrieved from "http://en.wikipedia.org/w/index.php?title=Lyotropic_liquid_crystal&oldid=513648868" Categories: Chemical properties Phases of matter Liquid crystals

en.wikipedia.org/wiki/Lyotropic_liquid_crystal

2/3

07/11/12

Lyotropic liquid crystal - Wikipedia, the free encyclopedia

This page was last modified on 20 September 2012 at 06:53. Text is available under the Creative Commons Attribution-ShareAlike License; additional terms may apply. See Terms of Use for details. Wikipedia® is a registered trademark of the Wikimedia Foundation, Inc., a non-profit organization.

en.wikipedia.org/wiki/Lyotropic_liquid_crystal

3/3

MTLNNLNO

qÜÉ=` ÜÉêÉòçî =i ~Ä=J=i ` m=o Éëçì êÅÉë

home

research

LCP resources

publications

members

links

contact

Introduction INTRO LCP Equations LCP Lipids LCP Protocols LCP Tools Crystal Gallery Structures from LCP LCP Publications LCP Links

Figure 1. Temperature-composition phase diagram of monoolein, the most common lipid for LCP crystallization. The phase diagram represents a metastable state at temperatures below 20 C. Re-drawn from Briggs & Caffrey, 1996.

Box1. LCP Properties: Transparent Optically isotropic Viscous and sticky Gel-like Large surface area/volume Single lipid bilayer

Box 2. LCP applications: Drug delivery Sustained release Topical application Uptake of contaminants Food emulsifier Biosensors Separation of biomolecules Chemical synthesis Protein crystallization

ÅÜÉêÉòçî KëÅêáééëKÉÇì LêÉëçì êÅÉëKÜíã

Figure 2. Cartoon pictures of three bicontinuous LCPs w ith different space groups. Two networks of non-intersecting water channels are shown in different colors.

Lipidic cubic phase (LCP) is one of many liquid crystalline phases that form spontaneously upon mixing lipids w ith w ater at proper conditions (Fig 1). Structuraly LCP consists of a single lipid bilayer that follow s an infinite periodic minimal surface (IPMS) dividing the space into tw o non-intersecting netw orks of w ater channels. There are three common types of bicontinuous LCPs w ith different space group symmetry (Fig. 2). LCP possesses many unique properties (Box 1), making it an attractive tool for a number of applications (Box 2). Crystallization of membrane proteins in LCP w as introduced in 1996 by Landau & Rosenbusch (1). This technique has proven to be crucial for elucidating structural mechanisms of action of several microbial rhodopsins, as w ell as provided the first high-resolution details of human G protein-coupled receptors (GPCR) bound to diffusible ligands. Success of using LCP for grow ing highly ordered crystals of challenging human membrane proteins can be attributed to at least tw o factors. LCP provides a more nativelike membrane environment for proteins as opposed to a rather harsh environment associated w ith detergent micelles. Crystals grow n in LCP have type I packing w ith protein molecules making contacts not only through hydrophilic but also through hydrophobic portions resulting in low er solvent content and better crystal ordering (Fig. 3). Click here >> to see the stats on membrane protein structures crystallized in LCP. References 1. Landau, E.M., and J.P. Rosenbusch. (1996) Lipidic cubic phases: a novel concept for the crystallization of m em brane proteins. Proc. Natl. Acad. Sci. U S A 93: 14532-14535. >> Figure 3. Cartoon representation of the in meso crystallization process. Integral membrane protein molecules (blue-green) are initially embedded into the lipid bilayer of the LCP (yellow), which provides a connectivity for the 3D diffusion. Addition of a precipitant induces a crystal nucleation. The crystal (in the center) is attached to the bulk LCP through a multilamellar lipid portal, which feeds the growing crystal with the protein molecules.

NLO

MTLNNLNO

qÜÉ=` ÜÉêÉòçî =i ~Ä=J=i ` m=o Éëçì êÅÉë

>

CONTACT US: The Scripps Research Institute | Cherezov Lab | 10550 North Torrey Pines Road | GAC1200 | La Jolla | CA 92037 | [email protected]

ÅÜÉêÉòçî KëÅêáééëKÉÇì LêÉëçì êÅÉëKÜíã

OLO

pharmacy Just another WordPress.com weblog

Polymer Micelle Structures

July 30, 2009

Posted by desporatist in Polymer Micelles as Drug Carriers. trackback Self-assembled micelles Self-assembled polymer micelles are created from amphiphilic polymers that spontaneously form nanosized aggregates when the individual polymer chains (“unimers”) are directly dissolved in aqueous solution (dissolution method)above a threshold concentration (critical micelle concentration or CMC) and solution temperature (critical micelle temperature or CMT) (Fig. 1). Amphiphilic polymers with very low water solubility can alternatively be dissolved in a volatile organic solvent, then dialyzed against an aqueous buffer (dialysis method).

Fig. 1. Self-assembly of block copolymer micelles. Amphiphilic diblock (hydrophilichydrophobic) or triblock (hydrophilichydrophobichydrophilic) copolymers are most commonly used to prepare selfassembled polymer micelles for drug delivery, although the use of graft copolymers has been reported. For drug delivery purposes, the individual unimers are designed to be biodegradable and/or have a low enough molecular mass (< ~40 kDa) to be eliminated by renal clearance, in order to avoid polymer buildup within the body that can potentially lead to toxicity. The most developed amphiphilic block copolymers assemble into spherical coreshell micelles approximately 10 to 80 nm in diameter, consisting of a hydrophobic core for drug loading and a hydrophilic shell that acts as a physical (“steric”) barrier to both micelle aggregation in solution, and

to protein binding and opsonization during systemic administration (Fig. 2). The most common hydrophilic block used to form the hydrophilic shell is the FDAapproved excipient poly(ethylene glycol) (PEG) or poly(ethylene oxide) (PEO). PEG or PEO consists of the same repeating monomer subunit CH2CH2O, and may have different terminal end groups, depending on the synthesis procedure, e.g. hydroxyl group HO(CH2CH20)nH; methoxy group CH30(CH2CH20)nH, etc. PEG/PEO blocks typically range from 1 to 15 kDa.In addition to its FDA approval, PEG is extremely soluble and has a large excluded volume. This makes it especially suitable for physically interfering with intramicelle interactions and subsequent micelle aggregation. PEG also blocks protein and cell surface interactions, which greatly decreases nanoparticle uptake by the reticuloendothelial system (RES), and consequently increases the plasma half life of the polymer micelle. The degree of steric protection by the hydrophilicshell is a function of both the density and length of the hydrophilic PEG blocks.

Fig. 2. Polymer micelle structures. Unlike the hydrophilic block, which is typically PEG or PEO, different types of hydrophobic blocks have been sufficiently developed as hydrophobic drug loading cores. Examples of diblock copolymers include (a) poly(Lamino acids), (b) biodegradable poly(esters), which includes poly(glycolic acid), poly(D lactic acid), poly(D,Llactic acid), copolymers of lactide/glycolide, and poly(ecaprolactone), (c) phospholipids/long chain fatty acids; and for triblock copolymers, (d) polypropylene oxide (in Pluronics/poloxamers). The choice of hydrophobic block is largely dictated by drug compatibility with the hydrophobic core (when drug is physically loaded, as described later) and the kinetic stability of the micelle. The selfassembly of amphiphilic copolymers is a thermodynamic and, consequently, a reversible process that is entropically driven by the release of ordered water from hydrophobic blocks; it is either stabilized or destabilized by solvent interactions with the hydrophilic shell. As such, the structural potential of amphiphilic copolymer unimers to form micelles is determined by the mass ratio of hydrophilic to hydrophobic blocks, which also affects the subsequent morphology if aggregates are formed. If the mass of the hydrophilic block is too great, the copolymers exist in aqueous solution as unimers, whereas, if the mass of the hydrophobic block is too great, unimer aggregates with nonmicellar morphology are formed. If the mass of the hydrophilic block is similar or slightly greater than the hydrophobic block, then conventional core shell micelles are formed. An important

consideration for drug delivery is the relative thermodynamic (potential for disassembly) and kinetic (rate of disassembly) stability of the polymer micelle complexes, after intravenous injection and subsequent extreme dilution in the vascular compartment. This is because the polymer micelles must be stable enough to avoid burst release of the drug cargo, as in the case of a physically loaded drug, upon systemic administration and remain as nanoparticles long enough to accumulate in sufficient concentrations at the target site. The relative thermodynamic stability of polymer micelles (which is inversely related to the CMC) is primarily controled by the length of the hydrophobic block.An increase in the length of the hydrophobic block alone significantly decreases the CMC of the unimer construct (i.e. increases the thermodynamic stability of the polymer micelle), whereas an increase in the hydrophilic block alone slightly increases the CMC (i.e. decrease the thermodynamic stability of a polymer micelle).Although the CMC indicates the unimer concentration below which polymer micelles will begin to disassemble, the kinetic stability determines the rate at which polymer micelle disassembly occurs. Many diblock copolymer micelles possess good kinetic stability and only slowly dissociate into unimers after extreme dilution. Thus, although polymer micelles are diluted well below typical unimer CMCs (10~6107M) after intravenous injection, their relative kinetic stability might still be suitable for drug delivery. The kinetic stability depends on several factors, including the size of a hydrophobic block, the mass ratio of hydrophilic to hydrophobic blocks, and the physical state of the micelle core. The incorporation of hydrophobic drugs may also further enhance micelle stability. Unimolecular micelles Unimolecular micelles are topologically similar to selfassembled micelles, but consist of single polymer molecules with covalently linked amphiphile chains. For example, copolymers with starlike or dendritic architecture, depending on their structure and composition, can either aggregate into multimolecular micelles,or exist as unimolecular micelles. Dendrimers are widely used as building blocks to prepare unimolecular micelles, because they are highlybranched, have welldefined globular shape and controled surface functionality. For example, unimolecular micelles were prepared by coupling dendritic hypercores of different generations with PEO chains. The dendritic cores can entrap various drug molecules. However, due to the structural limitations involved in the synthesis of dendrimers of higher generation, and relatively compact structure of the dendrimers, the loading capacity of such micelles is limited. Thus, to increase the loading capacity, the dendrimer core can be modified with hydrophobic block, followed by the attachment of the PEO chains. For example, Wang et al. recently synthesized an amphiphilic 16arm star polymer with a polyamidoamine dendrimer core and arms composed of inner lipophilic poly(ecaprolactone) block and outer PEO block. These unimolecular micelles were shown to encapsulate a hydrophobic drug, etoposide, with high loading capacity. Multiarm starlike block copolymers represent another type of unimolecular micelles. Star polymers are generally synthesized by either the armfirst or corefirst methods. In the armfirst method, monofunctional living linear macromolecules are synthesized and then crosslinked either through propagation, using a bifunctional comonomer, or by adding a multifunctional terminating agent to connect precise number of arms to one center. Conversely, in the corefirst method, polymer chains are grown from a multifunctional initiator. One of the first reported examples of unimolecular micelles, suitable for drug delivery, was a threearm star polymer, composed of mucic acid substituted with fatty acids as a lipophilic inner block and PEO as a hydrophilic outer block. These polymers were directly dispersible in aqueous solutions and formed unimolecular micelles. The size and solubilizing capacity of the micelles were varied by changing the ratio of the hydrophilic and lipophilic moieties. In addition, starcopolymers with polyelectrolyte arms can be prepared to develop pHsensitive unimolecular micelles

as drug carriers. Cross-linked micelles The multimolecular micelles structure can be reinforced by the formation of crosslinks between the polymer chains. These resulting crosslinked micelles are, in essence, single molecules of nanoscale size that are stabile upon dilution, shear forces and environmental variations (e.g. changes in pH, ionic strength, solvents etc.). There are several reports on the stabilization of the polymer micelles by crosslinking either within the core domain or throughout the shell layer. In these cases, the crosslinked micelles maintained small size and coreshell morphology, while their dissociation was permanently suppressed. Stable nanospheres from the PEObpolylactide micelles were prepared by using polymerizable group at the core segment. In addition to stabilization, the core polymerized micelles readily solubilized rather large molecules such as paclitaxel, and retained high loading capacity even upon dilution. Formation of interpenetrating network of a temperaturesensitive polymer (polyNisopropylacrilomide) inside the core was also employed for the stabilization of the Pluronic micelles. The resulting micelle structures were stable against dilution, exhibited temperatureresponsive swelling behavior, and showed higher drug loading capacity than regular Pluronic micelles. Recently, a novel type of polymer micelles with crosslinked ionic cores was prepared by using block ionomer complexes as templates. The nanofabrication of these micelles involved condensation of PEObpoly(sodium methacrylate) diblock copolymers by divalent metal cations into spherical micelles of coreshell morphology. The core of the micelle was further chemically crosslinked and cations removed by dialysis. Resulting micelles represent hydrophilic nanospheres of coreshell morphology. The core comprises a network of the crosslinked polyanions and can encapsulate oppositely charged therapeutic and diagnostic agents, while a hydrophilic PEO shell provides for increased solubility. Furthermore, these micelles displayed the pH and ionic strengthresponsive hydrogellike behavior, due to the effect of the crosslinked ionic core. Such behavior is instrumental for the design of drug carriers with controled loading and release characteristics.

Comments» No comments yet — be the first.

Theme: Regulus by Binary Moon. Blog at WordPress.com.

Encyclopedia of Nanoscience and Nanotechnology

www.aspbs.com/enn

Preparation of Vesicles (Liposomes) Peter Walde ETH Zürich, Zürich, Switzerland

CONTENTS 1. What are Vesicles (Liposomes)? 2. Vesicles and the Liquid Crystalline States of Surfactants 3. Methods for Preparing Normal Vesicles 4. Preparation of Reversed Vesicles 5. Characterizations and Applications of Vesicles 6. Concluding Remarks Glossary References

1. WHAT ARE VESICLES (LIPOSOMES)? 1.1. Introduction Vesicles—more precisely “normal vesicles”—are a particular type of polymolecular aggregate (polymolecular assembly) of certain amphipathic molecules, formed in aqueous solution. A vesicle is composed of one or more closed shells (which are usually 4–5 nm thick) that entrap a small volume of the aqueous solution in which the vesicle is formed. Vesicles are often spherical (under osmotically balanced conditions) and can have diameters between about 20 nm and more than 0.1 mm. If an analogous type of aggregate is formed in a water-immiscible, apolar solvent, the aggregate is called a reversed vesicle (see Section 4). In the following, the term “vesicle” or “lipid vesicle” always stands for a normal type of vesicle and not for a reversed vesicle. For the sake of simplicity, a so-called unilamellar vesicle is first considered. It is a closed lamella with an aqueous interior. The lamella is composed of amphipathic molecules, compounds that comprise at least two opposing parts, a hydrophilic part (which is soluble in water) and a hydrophobic part (which is not soluble in water but is soluble in an organic solvent that is not miscible with water, in this context also called “oil”). Amphipathic molecules have a “sympathy” as well as an “antipathy” for water. Because of the mixed affinities within the same chemical structure, amphipathic molecules are also called amphiphiles (meaning “both

ISBN: 1-58883-065-9/$35.00 Copyright © 2004 by American Scientific Publishers All rights of reproduction in any form reserved.

loving,” water as well as oil). They are surfactants, which stands for “surface active agents” and means that they accumulate at the surface of liquids or solids. The accumulation of surfactant molecules on the surface of water (at the waterair interface) leads to a reduction in the surface tension of water, as a result of an alteration of the hydrogen bonds between the interfacial water molecules. The aqueous solution in which vesicles form is present outside of the vesicles as well as inside. Figure 1 is a schematic representation of a unilamellar spherical vesicle formed by the amphiphile POPC (1-palmitoyl-2-oleoyl-snglycero-3-phosphocholine) in water at room temperature. The vesicle drawn in Figure 1 is assumed to have an outer diameter of 100 nm. The bilayer has a thickness of only about 4 nm, which corresponds in a first approximation to about two times the length of an extended POPC molecule. Lipid vesicles in the size range of 0.1
m =100 nm can be visualized by electron microscopy, for example, by the freeze-fracture technique [1, 2], or by cryofixation [2, 3] (see Fig. 2A and B). For vesicles in the micrometer range, light microscopy can be applied [4] (see Fig. 2C). Based on simple geometric considerations, one can calculate the approximate number of lipid molecules present in a particular defined vesicle. In the case of the unilamellar vesicle shown in Figure 1 (outer diameter 100 nm), about 81 × 104 POPC molecules form the shell of one vesicle, and all of these molecules are held together by noncovalent bonds. The single lamella of the giant vesicle shown in Figure 2C (diameter ∼ 60
m) contains about 26 × 1010 POPC molecules. The shell is a molecular bilayer with an arrangement of the POPC molecules in such a way that the hydrophobic (“water-hating”) acyl chains are in the interior of the bilayer and the hydrophilic (“water-loving”) polar head groups are on the two outer sites of the bilayer, in direct contact with either the trapped water inside the vesicle or with the bulk water in which the vesicle is dispersed. Since the bilayer shell in a sphere is necessarily curved, the number of amphiphiles constituting the inner layer is expected to be smaller than the number of amphiphiles present in the outer layer. For the 100-nm vesicle of Figure 1, the calculated number of POPC molecules is 374 × 104 in the inner layer and 436 × 104 in the outer layer, assuming a mean head group area of one POPC molecule of 0.72 nm2 [5]

Encyclopedia of Nanoscience and Nanotechnology Edited by H. S. Nalwa Volume 9: Pages (43–79)

44

Preparation of Vesicles (Liposomes) 5

3 4

2 ~ 4 nm

1

aqueous exterior

aqueous interior

100 nm

POPC O O

cis 9 10

hydrophobic

O

1

O

2

O P O O 3

H

1 O

+

H3 C N CH3 CH3

hydrophilic

Figure 1. Schematic representation of the cross section through a particular unilamellar, spherical vesicle that has an assumed outer diameter of 100 nm and is formed in water from the surfactant POPC (1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine). POPC constituting the single, closed lamellar shell of the vesicle is represented as a filled circle to which two tails are connected. The tails stand for the two hydrophobic (water-insoluble) chains of POPC, and the filled circle symbolizes the hydrophilic (water-soluble) phosphocholine head group. POPC in the vesicle bilayer is present in the fluid-disordered state, above Tm (see Section 2.3). The chemical structure of POPC, which is a naturally occurring glycerophospholipid, is also shown. The glycerol moiety with its stereospecifically numbered carbon atoms is localized inside the dotted rectangle. The important molecular motions above Tm are indicated: (1) conformational transitions in the hydrophobic tails; (2) conformational transitions in the head group; (3) rotational diffusion about the axis perpendicular to the surface of the bilayer; (4) lateral diffusion within the bilayer plane; (5) vertical vibrations, out of the bilayer plane; and (not shown) collective undulations of the membrane. See text for nomenclature details and [392] for a detailed description of the vesicle membrane dynamics.

s

and a bilayer thickness of 3.7 nm [6]. This is certainly a rough estimation, and the real situation in a curved bilayer is always asymmetric with different packing conditions (and mean surfactant head group areas) in the inner and outer layers [7, 8]. Whether vesicles are thermodynamically stable or not (see Sections 2 and 6) depends critically on whether

Figure 2. Electron and light microscopic visualization of vesicles prepared from POPC in water at ∼ 25  C. (A and B) Transmission electron micrographs of a suspension containing LUVs (FAT-VET100 ), prepared by the extrusion technique (see Section 3.8 and Fig. 3). The samples were analyzed by the freeze-fracture technique (A) and by cryo-fixation (B). The length of the bars corresponds to 100 nm. (C) Light micrograph (differential interference contrast mode) of a single GUV prepared by the electroformation method (see Section 3.3). Length of the bar: 10
m. The electron micrographs were taken by M. Müller, E. Wehrli, and N. Berclaz, Service Laboratory for Electron Microscopy, in the Department of Biology at the ETH Zürich. The light micrograph was taken by R. Wick at the Department of Materials Science at the ETH Zürich.

the chains are flexible enough to accommodate these asymmetric constraints [7]. The mean head group area of a surfactant, abbreviated a0 , corresponds to the area that is occupied on average by the polar head group packed within the vesicle’s bilayer as a consequence of the actual molecular size and the interaction between the neighboring surfactant molecules. A simple geometric model that, in addition to a0 , takes into account the overall geometric shape of the surfactant as a critical packing parameter (or “shape factor”), p =v/a0 lc )), has been developed [7–10] and successfully applied in a useful simple theory toward an understanding of molecular selfassembling systems at large [11]; v is the volume of the hydrophobic portion of the molecule and lc is the critical

Preparation of Vesicles (Liposomes)

length of the hydrophobic tails, effectively the maximum extent to which the chains can be stretched out. According to this simple theory, lipid bilayers and vesicles can be prepared if p for a particular amphiphile has a value between 0.5 and 1.0 (conditions under which the surfactants will pack into flexible, curved, or planar bilayers).

1.2. Terminology There are a large number of amphiphiles that form vesicles. The most intensively investigated are certain lipids present in biological membranes, glycerophospholipids, lipids that contain a glycerol-3-phosphate unit. Actually, the geometric structure of vesicles as spherulites—which is the term originally used [12]—containing one (or more) concentric bilayer shell(s) was elaborated for the first time with vesicle preparations made from (mixtures of) naturally occurring glycerophospholipids [12, 13]. This is why it has been proposed to call these aggregates lipid vesicles [14] or liposomes (actually meaning “fat bodies”) [15, 16]. Since the type of aggregate shown in Figures 1 and 2 has not so much to do with a “fat body,” it is in principle more appropriate to use the term lipid vesicle or just vesicle instead of liposome [17]. In any case, all of the terms liposome, lipid vesicle, and vesicle are used here for the same type of polymolecular aggregate. Sometimes, however, one also finds the term synthetic vesicles [18, 19], referring to vesicles formed by synthetic, often charged, nonnatural surfactants. Others use the term vesicle exclusively for a closed unilamellar (not multilamellar) aggregate of amphiphiles [20, 21], such as the one shown schematically in Figure 1. Other names appearing in the literature are niosomes (vesicles prepared from nonionic surfactants) [22, 23], polymer vesicles or polymersomes (vesicles prepared from polymeric surfactants) [24–26], and so on. Table 1 summarizes different terms to describe a particular type of vesicle. Whenever one uses one of these terms, one should specify how it is used and how it is actually defined, to avoid any possible confusion or misunderstanding.

1.3. Nomenclature and Chemical Structures of Vesicle-Forming Glycerophospholipids Although the basic principles of vesicle formation are for all types of vesicle-forming amphiphiles in the end the same—independently of whether they are charged, neutral, or polymeric—the descriptions in Sections 2 and 3 focus on just one particular group of phospholipids, the so-called phosphatidylcholines (PCs). POPC (see Fig. 1) is a particular PC, namely 1-palmitoyl-2-oleoyl-phosphatidylcholine, more precisely, 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (also called 1-palmitoyl-2-oleoyl-sn-glycero-3phosphorylcholine). Since the nomenclature of glycerophospholipids (such as POPC) is often not so well known to those not working with this particular type of biological molecule, a short overview of the main nomenclature rule is given next. The nomenclature of lipids outlined and used here has been proposed by the International Union of Pure and Applied Chemistry–International Union of Biochemistry (IUPAC-IUB) [27, 28].

45 POPC is a chiral phospholipid with one chiral center at that carbon atom that is localized in the middle of the glycerol moiety; POPC belongs to the glycerophospholipids, the quantitatively most important structural group within the class of phospholipids. Glycerophospholipids have a glycerol backbone to which a phosphate group is bound through a phosphoric acid ester linkage to one of the glycerol hydroxyl groups. To designate the configuration of this glycerol derivative, the carbon atoms of the glycerol moiety are numbered stereospecifically (indicated in the chemical name as prefix -sn-). If the glycerol backbone is written in a Fischer projection (see, for example, [29]) in such a way that the three carbon atoms are arranged vertically and the hydroxyl group connected to the central carbon atom is pointing to the left, then the carbon atom on top is C-1, the carbon atom in the middle is C-2 (this is the actual chiral center common to all glycerophospholipids), and the carbon atom at the bottom is C-3. With this convention, the chemical structure of 1-palmitoyl-2-oleoyl-sn-glycero-3phosphocholine is defined. It is the naturally occurring form of the two possible enantiomers. The mirror image of POPC is 2-oleoyl-3-palmitoyl-sn-glycero-1-phosphocholine. In the Cahn-Ingold-Prelog nomenclature system (also called the R/S convention) [29], the configuration of C-2 in POPC is R. (Another shorthand description of POPC sometimes used is PamOleGroP Cho or PamOlePtdCho [27, 28]; Pam stands for palmitoyl, Ole for oleoyl, Gro for glycerol, P for phosphate, Cho for choline, and Ptd for phosphatidyl). All naturally occurring diacylglycerophospholipids that have a zwitterionic phosphocholine head group are also called 3-sn-phosphatidylcholines, or just phosphatidylcholines. In the biochemical and biophysical literature, phosphatidylcholine is also called lecithin (or l--lecithin because of its stereochemical relationship to the naturally occurring l--glycerol-phosphate). Therefore, POPC is a lecithin. The name lecithin implies that egg yolk (lekithos in Greek) contains large amounts of phosphatidylcholines. According to the IUPAC-IUB, the use of the term lecithin is permitted but not recommended [27]. It is indeed better to avoid using this term since it may have another meaning in the food technology literature. The International Lecithin and Phospholipid Society (ILPS) of the American Oil Chemists Society (AOCS) defines lecithin as a mixture of lipids obtained from animal, vegetable, or microbial sources, which includes PCs but also contains a variety of other substances, such as sphingosylphospholipids, triglycerides, fatty acids, and glycolipids [30]. Egg yolk is one of the cheapest commercial sources for the isolation of phosphatidylcholines; the other one is soybeans. Phosphatidylcholines from egg yolk contain a number of chemically different phosphatidylcholines. All of these PCs have, however, the same glycerol backbone and the same polar head group (phosphocholine). They only differ in the acyl chains esterified with the glycerol hydroxyl groups at C-1 and C-2 (see Table 2) [31]. Many studies on the preparation and characterization of lipid vesicles have been carried out with egg yolk PC. On average, in position C-1 in egg yolk PC is often palmitic acid, and in position C-2, oleic acid [31]. POPC is therefore a representative PC molecule for egg yolk PCs. In contrast to egg yolk PCs, however, POPC has a well-defined chemical

46

Preparation of Vesicles (Liposomes)

Table 1. List of some of the terms used to describe a particular type of vesicle. Term Algosome Archaesome Bilosome Catanionic vesicle Cerasome Ethosome Fluorosome Hemosome Immunoliposome Lipid vesicle Liposome Magnetoliposome Marinosome Niosome Novasome Phospholipid vesicle PLARosome Polymer vesicle Polymerized vesicle Polymersome Proliposomes

Proniosomes Reversed vesicle Spherulite Sphingosome Stealth liposome

Synthetic vesicle Toposome Transfersome Ufasome Vesicle

Virosome

Meaning and use of the term in the literature Vesicle prepared on the basis of 1-O-alkylglycerol [485]. Vesicles prepared from archaebacterial, bolaamphiphilic lipids [359, 486]. Vesicle prepared from a particular mixture of non-ionic surfactants (1-monopalmitoyl-glycerol), cholesterol, dihexadecylphosphate (5:4:1 molar ratio), and bile salt (particularly deoxycholate) [487]. Vesicle prepared from a mixture of a cationic and an anionic surfactant [59, 488]. Vesicle with a silicate framework on its surface [366, 489]. Vesicle that contains in the final preparation a considerable amount of ethanol (prepared by a particular method described in Section 3.20) [229–231]. SUV containing a fluorescent dye embedded in its bilayer to monitor the entry of molecules into the bilayer [490–492]. Hemoglobin-containing vesicle [493]. Vesicle as a drug delivery system that contains on the external surface antibodies or antibody fragments as specific recognition sites for the antigen present on the target cells [69, 373, 494]. Vesicle prepared from amphiphilic lipids [31, 69]. Vesicle prepared from amphiphilic lipids [31, 69]. Vesicle containing magnetic nanoparticles (e.g., magnetite Fe3 O4 ) [495–498]. Vesicle based on a natural marine lipid extract composed of phospholipids (PCs and phosphatidylethanolamines) containing a high amount (∼65%) polyunsaturated aycl chains [499]. Vesicle prepared from non-ionic surfactants [23, 500]. In some cases, at room temperature polyhedral niosomes exist, which transform into spherical niosomes upon heating, cholesterol addition, or sonication [501–503]. Oligo- or multilamellar vesicle prepared by a particular technology that involves the addition of vesicle-forming surfactants in the liquid state (at high temperature) to an aqueous solution (Section 3.10) [185]. Vesicle prepared from (amphiphilic) phospholipids [71]. Phospholipid-alkylresocinol liposome: phospholipid vesicle containing resorcinolic lipids or their derivatives [504]. Vesicle prepared from polymeric amphiphiles, such as di- or triblock copolymers [24, 26]. Vesicle prepared from polymerizable amphiphiles that were (partially) polymerized after vesicle formation [33, 352, 353, 505]. Vesicle prepared from polymeric amphiphiles, such as di- or triblock copolymers [24, 26]. A preparation that upon mixing with an aqueous solution results in the formation of vesicles. The preparation contains vesicle-forming amphiphiles and an alcohol (ethanol, glycerol, or propyleneglycol) (see Section 3.19). Dry (ethanol-free) granular preparations of vesicle-forming amphiphiles, which upon hydration lead to vesicle formation, are also called proliposomes (Section 3.19) [506]. A dry, granular product containing mainly (but not exclusively) non-ionic surfactants which, upon the addition of water, disperses to form MLVs [507]. Inverted vesicle formed in a water-immiscible, apolar solvent in the presence of a small amount of water (Section 4) [508]. Onion-like vesicle prepared with spherulite technology, which involves the use of shear forces (Section 3.11). Vesicle prepared on the basis of sphingolipids present in human skin [69, 509]. Sterically stabilized vesicle, achieved through the use of co-amphiphiles that have PEG (poly(ethyleneglycol))-containing hydrophilic head groups [510–512]. Stealth liposomes are not so easily detected and removed by the body’s immune system (they are long-circulating in the blood). The name stems from an analogy to the American “Stealth bomber” aircraft, which is not easily detected by radar. Alternatively to PEG, polysaccharides have also been used [373]. Vesicle prepared from synthetic surfactants (surfactant mixtures) that are not present in biological membranes. The surfactants usually have a single hydrophobic tail [17, 488]. Vesicle that has a surface that is site-selectively (toposelectively) modified in a stable manner at specific and deliberate locations (e.g., through chemical modifications or chemical functionalizations) [513]. Ultradeformable ethanol-containing mixed lipid/detergent vesicle claimed to transfer water-soluble molecules across human skin (Section 3.28) [275, 276]. Vesicle prepared from unsaturated fatty acid/soap mixtures [330]. General term to describe any type of hollow, surfactant-based aggregate composed of one or more shells. In the biological literature, the term vesicle is used for a particular small, membrane-bounded, spherical organelle in the cytoplasm of an eukaryotic cell [97]. Vesicle containing viral proteins and viral membranes, reconstituted from viral envelopes, the shells that surround the virus [69, 514–516].

Note: In this chapter, all of the terms listed in the table are called vesicles (or lipid vesicles), independent on the chemical structure of the amphiphiles (surfactants) constituting the vesicle shell(s).

structure. For more basic studies, POPC may be more suited than the egg yolk PC mixture. For applications, however, the cheaper egg yolk PCs may be advantageous. Although lipid vesicles prepared from egg yolk PCs are similar in many respects to vesicles prepared from

POPC, the properties of POPC vesicles at a particular fixed temperature may be very different from those of the chemically related DPPC (1,2-dipalmitoyl-sn-glycero-3phosphocholine) vesicles, for example. The reason for this is outlined in Section 2.

47

Preparation of Vesicles (Liposomes) Table 2. Main approximate fatty acid content in egg yolk and soybean PCs (see [31]). Relative abundance in phospholipids Egg yolk

Soybeans

at sn-1 (%)

at sn-2 (%)

Total (%)

at sn-1 (%)

∼17

∼34

Fatty acid (trivial name)

Abbreviation

Total (%)

Hexadecanoic acid (palmitic acid) Octadecanoic acid (stearic acid) Cis-9-octadecenoic acid (oleic acid) Cis, cis-9,12-octadecadienoic acid (linoleic acid) Cis, cis-6,9octadecadienoic acid All cis-9,12,15octadecatrienoic acid All cis-5,8,11,14,17eicosapentaenoic acid All cis-4,7,10,13,16,19docosahexaenoic acid

C 16:0

∼35

∼69

∼2

C 18:0

∼14

∼26

∼1

C 18:1c9

∼27

∼5

∼49

18:2c9c12 ∼6

18:2c6c9

at sn-2 (%)

∼8 ∼23

∼30

∼16

∼48

∼24

∼71

∼9

∼4

∼13

∼11

18:3c9c12c15 20:5c5c8c11c14c17

∼4

∼7

22:6c4c7c10c13c16c19

∼13

∼25

Note: The abbreviation 16:0, for example, indicates that the linear fatty acid has 16 carbon atoms without any double bonds; 18:1c9 indicates that the linear fatty acid is composed of 18 carbon atoms with one cis double bond in position 9,10 (starting at position 9), where the carboxy C atom is carbon number 1.

1.4. There Are Not Only Unilamellar Vesicles Vesicles are not only classified by the chemical structure of the molecules constituting the vesicle shell(s) as reported in Table 1, but also according to their size, lamellarity and morphology, and method of preparation (see Table 3). Small unilamellar vesicles (SUVs) have one lamella and diameters of less than about 50 nm. So-called large unilamellar vesicles (LUVs) have one lamella and diameters between about 50 nm and about 500 nm (see Fig. 1 and Fig. 2A and B). Giant vesicles (GVs) can be observed by light microscopy and have diameters of more than about 0.5–1
m (Fig. 2C). Oligolamellar vesicles (OLVs) have a few and multilamellar vesicles (MLVs) have many concentrically arranged lamellae. Multivesicular vesicles (MVVs) contain nonconcentrically arranged vesicles within a larger vesicle. As described in detail in Section 3, the preparation of vesicle suspensions generally involves the use of a particular technique, a particular preparation method. Depending on the technique applied, the vesicle suspensions are characterized by a certain degree of homogeneity, a certain mean size and mean lamellarity, and a certain trapped volume. The trapped volume is the aqueous volume that is encapsulated by the lipid vesicles, expressed as microliters of aqueous solution per micromole of lipid (=liters/mol). A trapped volume of 1
l/
mol means that in a vesicle suspension containing 1
mol lipid, for example, only 1
l of the aqueous solution is trapped by the vesicles. The spherical unilamellar POPC vesicle shown in Figure 1 has a trapped volume of about 3
l/
mol, as calculated based on simple geometric considerations. This means that in a vesicle suspension

prepared from 10 mM POPC (=76 g/liter), only 30
l out of 1 ml is trapped by the vesicles (3 volume %). The total lipid-water interfacial area in this vesicle suspension is 43 × 103 m2 ! Since, in many cases, the vesicle suspensions prepared by one particular technique are not further characterized with respect to mean size, size distribution, and lamellarity, the vesicles are just named according to the method used. Examples include REVs (reversed-phase evaporation vesicles, vesicles prepared by the so-called reversed-phase evaporation method), VETs (vesicles prepared by the so-called extrusion technique), etc.; see some entries in Tables 1 and 3.

2. VESICLES AND THE LIQUID CRYSTALLINE STATES OF SURFACTANTS 2.1. Introduction Since in most cases, lipid vesicles can be considered as dispersions of a liquid crystalline state of a vesicle-forming surfactant, it is useful to give a short introduction to some of the liquid crystalline phases of amphiphilic molecules, particularly focusing on the so-called L·  - and L·  - or L·  phases, which are considered to be the relevant thermodynamic equilibrium states of most glycerophospholipids under the conditions in which vesicle formation is observed. For a recent excellent general review on surfactant liquid crystals, see [32].

48

Preparation of Vesicles (Liposomes)

Table 3. List of some of the abbreviations often used for a particular type of ve sicle. Abbreviation

Meaning of the abbreviation

DRV

MLV

Dehydrated-rehydrated vesicle (or dried-reconstituted vesicle) Ethanol-injected vesicle Vesicle prepared by repeatedly freezing and thawing a MLV suspension Vesicle prepared with a French press Giant unilamellar vesicle Giant vesicle Interdigitation-fusion vesicle Large unilamellar vesicle Large unilamellar Vesicle prepared by the extrusion technique Multilamellar vesicle

MVL

Multivesicular liposome

MVV

Multivesicular vesicle

OLV REV

Oligolamellar vesicle Reversed-phase evaporation vesicle Rapid solvent exchange vesicle Stable plurilamellar vesicle Small (or sonicated) unilamellar vesicle Unilamellar vesicle Unilamellar vesicle Vesicle prepared by the extrusion technique

EIV FATMLV

v FPV GUV GV IFV LUV LUVET

RSE vesicle SPLV SUV ULV UV VET

Characteristics Vesicle prepared by the dehydration-rehydration method (Section 3.7) [31] Vesicle prepared by the ethanol injection method (Section 3.18) Equilibration and homogenization procedure (Section 3.6) [133]

Unilamellar vesicle or OLV prepared with a French press for vesicle size homogenization (Section 3.8) [72] Unilamellar vesicle with a diameter larger than about 500 nm Vesicle with a diameter larger than about 500 nm Vesicle prepared by the interdigitation-fusion method (Section 3.21) [232, 517] Unilamellar vesicle with a diameter between about 50 nm and 500 nm Vesicles prepared by the extrusion technique are usually large and mainly unilamellar (Section 3.8) The vesicle contains several concentrically arranged lamellae osmotically stressed after formation because of an exclusion of solute molecules during their formation [76] A large vesicle that contains internal, nonconcentrically arranged vesicular compartments; also called MVV [211, 212] A large vesicle that contains internal, nonconcentrically arranged vesicular compartments; also called MVL [210] The vesicle contains a few concentrically arranged lamellae Vesicle prepared by the reversed-phase evaporation technique (Section 3.14) Vesicle prepared by the rapid solvent exchange method [238] Similar to MLV but not osmotically stressed after its formation [76] Unilamellar vesicle with a diameter of less than about 50 nm, as typically obtained by sonicating MLVs (Section 3.5) Vesicle with only one lamella Vesicle with only one lamella Vesicles prepared by the extrusion technique are usually large and mainly unilamellar (Section 3.8)

Note: The abbreviations are based on the size and morphology, the lamellarity, or method of preparation.

2.2. Lamellar Phase and “Gel Phase” A liquid crystalline state (also called the “mesophase”) of a substance is a state between a pure crystal (characterized by a high order of rigid molecules) and a pure liquid (characterized by rapid molecular motions of disordered molecules) [32]. There are dozens of different liquid crystalline states, all characterized by a different degree of molecular mobility and order. Liquid crystals can be produced either by heating a particular crystalline solid—called a “thermotropic liquid crystal”—or by dissolution of particular substances in a solvent—called a “lyotropic liquid crystal.” Many surfactants in water form lyotropic liquid crystalline phases, such as the lamellar phase (L·  ), the so-called gel phase (L·  or L·  ), the normal or reversed hexagonal phase (HI or HII ), or one or more of the known cubic phases (Ia3d, Pn3m, Im3m). The type of phase formed depends on the chemical structure of the surfactant used and on the experimental conditions (such as concentration and temperature, or the presence of other compounds). In the L·  -phase (also called the liquid-analogue [33] or liquid-disordered state [34, 35]), the surfactant molecules are arranged in bilayers, frequently extending over large

distances (1
m or more) [32]. The hydrophobic chains are rather disordered, with a lot of gauche conformations in the saturated hydrocarbon parts of the hydrophobic chains, making the bilayers fluid, characterized by fast lateral and rotational diffusions of the surfactant molecules, similar to a liquid. Comparable molecular motions are also present in the liquid-disordered state of vesicles. In the case of SUVs prepared from POPC, for example, the lateral diffusion coefficient seems to be on the order of 3–4 × 10−8 cm2 /s, as determined between 5  C and 35  C [36]. The rotational correlation time is on the order of 10−9 to 2 × 10−8 s [37]. The L·  (or L·  -)-phase of surfactant molecules (also called solid-analogue [33] or solid-ordered state [34, 35]) closely resembles the L·  -phase in the sense that the surfactant molecules are also arranged in bilayers. The viscosity is very high, however. This is a consequence of the rigidity of the individual surfactant molecules which are mostly present with all-trans conformations in the saturated hydrocarbon parts of the hydrophobic chains. The motion of the molecules is rather restricted, similar to the molecules in a crystal. To specify the relative arrangement of the lipid molecules, the gel phase may be abbreviated as L·  or L·  ,

49

Preparation of Vesicles (Liposomes)

depending on whether the alkyl chains are tilted (P·  ) or not titled (L·  ) with respect to the normal of the lipid bilayer. If it is not known whether the chains are tilted, L·  is often used as a general abbreviation. The phase behavior of a number of phosphatidylcholines [38, 39] and a number of other lipids and lipid mixtures [38, 40] has been determined and reviewed.

2.3. Main Phase Transition Temperature Tm of Glycerophospholipids In the case of conventional glycerophospholipids, either DPPC or POPC, the P·  -phase is formed at thermodynamic equilibrium at temperatures at least 5–10 below a lipid specific temperature called the main phase transition temperature (or lamellar chain melting temperature) Tm . Tm is also called the lamellar gel-to-liquid crystalline phase transition temperature and can be determined, for example, as the endothermic peak maximum in heating scans of differential scanning calorimetry (DSC) measurements [41–43]. Above Tm the lipids are in the L·  -phase. Between the L·  - (or P·  -) phase and the L·  -phase, an intermediate gel phase, called the ripple phase (abbreviated as P·  ) is often observed at high water content in the case of PCs [44]. This particular lipid phase takes its name from the fact that in freeze-fracture electron micrographs, a “ripple” structure can be seen if the lipid dispersion is rapidly frozen from the particular temperature interval in which the ripple phase is formed [44–47]. The transition from the “gel phase” to the “ripple phase” is called pretransition. With respect to certain practical aspects in the methods for lipid vesicle preparations described in Section 3, the Tm value of the lipid (or the lipid mixtures) used is important to know. In the case of dilute POPC-water systems, for example, Tm is around −3  C [48, 49]. If the water content is decreased below ∼10 wt%, Tm increases above 0  C, until it reaches a value of 68  C in the anhydrous system [49]. A list of different Tm values for a number of dilute aqueous phosphatidylcholine systems (MLVs) is given in Table 4. For a more detailed list of Tm values, including other glycerophospholipids, see [38, 40, 50]. Please note that in the case of phospholipids with charged head groups, the Tm values depend on the degree of protonation and may depend considerably on the chemical nature of the counter-ions present [51]. Furthermore, measurements carried out with SUVs give values about 4–5 lower than the Tm values obtained from MLVs [43, 52, 53].

3. METHODS FOR PREPARING NORMAL VESICLES 3.1. Introduction The thermodynamic equilibrium state of glycerophospholipids (and many other bilayer-forming amphiphiles) in water (or in a particular aqueous solution) is—probably under most experimental conditions—a stacked bilayer arrangement of the surfactant molecules, either as L·  -phase (above Tm ) or as L·  -, P·  - (or P )-phase (below Tm ) in equilibrium with excess aqueous phase (see Section 2).

Table 4. Main P -L ·  phase transition temperature (Tm ) values of dilute aqueous dispersions of certain common bilayer-forming phosphatidylcholines, data taken from [39] and [518] (for soybean PCs). Phosphatidylcholine DMPC (1,2-dimyristoyl-sn-glycero-3phosphocholine), 14:0/14:0 DPPC (1,2-dipalmitoyl-sn-glycero-3phosphocholine), 16:0/16:0 DSPC (1,2-distearoyl-sn-glycero-3phosphocholine), 16:0/16:0 POPC (1-palmitoyl-2-oleoyl-sn-glycero-3phosphocholine), 16:0/18:1c9 SOPC (1-stearoyl-2-oleoyl-sn-glycero-3phosphocholine), 18:0/18:1c9 DOPC (1,2-dioleoyl-sn-glycero-3phosphocholine), 18:1c9/18:1c9 Egg yolk PCs (see Table 2) Soybean PCs (see Table 2) Hydrogenated soybean PCs

Tm ( C) 23.6 ± 1.5 41.3 ± 1.8 54.5 ± 1.5 −2.5 ± 2.4 6.9 ± 2.9 −18.3 ± 3.6 −5.8 ± 6.5 −15 ± 5 51–52

Note: 16:0/18:1c9, for example, indicates that the linear acyl chain at sn-1 has 16 carbon atoms without any double bonds; the linear acyl chain at sn-2 has 18 carbon atoms with one cis double bond in position 9,10 (see also Table 2).

Upon dispersing in an aqueous solution, vesicles generally form from an amphiphile (or a mixture of amphiphiles) that forms a lamellar L·  -phase at thermodynamic equilibrium. Depending on how the dispersion is actually prepared (in other words, which method or technique is applied), the vesicles formed by the dispersed amphiphiles are either very heterogeneous or rather homogeneous and are mainly small (below about 50 nm), mainly large (between about 50 nm and 500 nm), or mainly very large (above about 500 nm). It all depends on the lipid (or lipid mixture) used, on the aqueous solution, and particularly on the preparation method. In the following, the principles of some of the best known and widely used methods for the preparation of lipid vesicle suspensions—mainly on a laboratory scale of a few milliliters up to about 100 ml—will be described. For each method, a different vesicle preparation with different typical general characteristics is obtained. It is important to point out once more that in most of the cases the resulting vesicle suspension is not at thermodynamic equilibrium, but represents only a metastable, kinetically trapped state. The equilibrium phases are L·  , P , L·  or L·  , as discussed in Section 2. A particular vesicle dispersion prepared is therefore physically (with respect to vesicle size and lamellarity) not indefinitely stable; it may slowly transform into the thermodynamically most stable state (stacked bilayers), as a result of a so-called aging process [54, 55]. This aging may occur either through the fusion of vesicles or because of an exchange of amphiphiles that are not aggregated (free, monomeric surfactant) [55]. This latter process—called Ostwald ripening (in analogy to the corresponding process occurring in emulsion systems)—is expected to be particularly relevant in the case of synthetic short-chain amphiphiles with a high monomer solubility (10−8 M). In the case of DPPC (and probably also POPC), the monomer solubility is on the order of 10−10 M [56], which means that aging through an Ostwald ripening process is less likely in these cases.

50 Although often only metastable, vesicle suspensions may be stable for a prolonged period of time, for example, for days, weeks, or months, provided that the vesicle-forming amphiphiles are chemically stable during this period of time [57, 58]. In the presence of other amphiphilic molecules (cosurfactants), the situation may change, particularly if the cosurfactant tends to form micellar aggregates, characterized by a packing parameter p ∼ 1/3 (relatively large head group cross-sectional area, a0 ). In this case mixed micelles may exist at thermodynamic equilibrium (in equilibrium with cosurfactant monomers), if the micelle-forming surfactant is present to a large extent. Such mixed micellar systems are used as a starting solution in the case of the so-called detergent depletion method described in Section 3.27. Furthermore, there are also known cases where there appear to be thermodynamically stable vesicles (particularly composed of mixtures of surfactants) [59, 60] (see Section 6). The presence of cholesterol (or other sterols)—molecules that are water insoluble and do not form vesicles alone— may also influence the properties of lipid vesicles, depending on the amount of cholesterol present [61, 62]. In the case of DPPC, for example, up to 33 mol% cholesterol, the Tm value of hydrated bilayers changes only slightly [63]. With increasing cholesterol concentration, however, the phase transition temperature is completely eliminated at 50 mol% (1:1 molar ratio of DPPC to cholesterol). The fluidity of the bilayer membrane is thereby changed, resulting in an increase in the fluidity below Tm of the PC and a decrease in fluidity above Tm , a state of the membranes that is intermediate between solid-ordered and liquid-disordered (see Section 2.3). This state is called liquid-ordered [34]. For all of the methods outlined in the following, more detailed descriptions can be found in the original literature cited. Furthermore, most of the generally known methods have already been summarized before—more or less extensively—in review articles or books about lipid vesicles (liposomes) [14, 31, 64–75]. Most of the methods can be roughly divided into two groups: (i) Methods that are based on the simple swelling of initially dried, preorganized lipids and the mechanical dispersion and mechanical manipulation of the dispersed bilayers (Sections 3.2–3.12). (ii) Methods that involve the use of (a) a cosolvent in which the lipids are soluble (Sections 3.14–3.26 and 3.30), (b) an additional non-bilayer-forming “helper amphiphile,” a coamphiphile (Sections 3.27 and 3.30), or (c) certain ions that influence the initial aggregational state of the lipids (Sections 3.13 and 3.29). All three type of molecules control the assembly process of the bilayer-forming amphiphile in a particular way during the vesicle preparation process, and all three types of molecules may at the end be difficult to remove completely from the final vesicle suspensions.

Preparation of Vesicles (Liposomes)

3.2. MLVs, GUVs, or Myelin Figures Formed by Thin Lipid Film Hydration One of the easiest ways to prepare a vesicle suspension is to disperse a dried lipid film in an aqueous solution [13, 76, 77]. The vesicle-forming and swellable [78] amphiphile is first dissolved in an organic solvent in which the amphiphile is soluble (usually chloroform). This solution is then placed inside a round-bottomed flask, and the solvent is completely removed by rotatory evaporation under reduced pressure, followed by high-vacuum drying overnight. The remaining amphiphiles form a dry thin film that is oriented in such a way as to separate hydrophilic and hydrophobic regions from each other [31]. If an aqueous solution is added to this film at a temperature above the main phase transition temperature Tm (see Section 2.3), the head groups of the dry lipids become hydrated and hydrated bilayers form. The hydration and swelling process is usually speeded up by gentle or vigorous shaking (using a vortex mixer), thereby dispersing the bilayers in the aqueous solution, resulting in the formation of mainly MLVs, which are very heterogenous with respect to size and lamellarity. Lipid film thickness and extent of shaking have an influence on the properties of the resulting vesicle suspension. The interlamellar repeat distance in equilibrated, completely hydrated PC bilayers above Tm is around 6.5 nm [79, 80]. This means that in a MLV formed from POPC as example, the aqueous space between two neighboring lipid lamellae is about 2.5 nm thick (taking into account a bilayer thickness of 4 nm). On average, a MLV may be composed of up to 10 bilayers [81]. The formation of closed bilayers (vesicles), in contrast to open bilayers, can be easily understood on the simple basis that interactions between the hydrophobic chains of the amphiphiles and the water molecules—as would be the case in open bilayers—are energetically unfavorable and therefore rather unlikely. The preparation of MLVs by the dispersal of a dried lipid film (also called hand-shaken MLVs [31]) is often a first step in the preparation of more defined vesicle suspensions (see, for example, Sections 3.5 and 3.8). With respect to the equilibration of water-soluble molecules between the bulk aqueous medium and the inner aqueous compartments of MLVs, ionic species may be unevenly distributed [76]. A more even distribution can be achieved by applying freeze/thaw cycles (see Section 3.6). The experimental conditions under which a dried lipid film is hydrated affect the resulting lipid aggregates obtained. In the case of phospholipid mixtures containing 90 mol% PC and 10 mol% of a negatively charged phospholipid (egg yolk phosphatidylglycerol, bovine brain phosphatidylserine, or bovine heart cardiolipin), the dried lipid film prepared inside a test tube can first be prehydrated at 45  C with water-saturated nitrogen gas for 15–25 min. Afterward, an aqueous solution containing water-soluble molecules to be entrapped (e.g., 100 mM KCl and 1 mM CaCl2 ) is gently added, and the tube is sealed under argon and incubated at 37  C for 10–15 h. During this incubation, the lipid film is completely stripped from the glass surface and forms vesicular aggregates as a kind of white, floating precipitate in the aqueous solution. The analysis of this

51

Preparation of Vesicles (Liposomes)

precipitate indicated the presence of many mainly unilamellar giant vesicles (not MLVs) with diameters on the order of tens of micrometers up to more than 300
m [82]. In addition, much smaller vesicles, large multilamellar vesicles as well as myelin figures (see below) and undispersed lipid material, could also be observed [82]. In the case of egg PC alone (no negatively charged phospholipids present), no giant unilamellar vesicles (GUVs) formed under the experimental procedure used [83]. It therefore seems that electrostatic repulsions between the charged head groups facilitate the formation of unilamellar membranes by opposing the intrinsic adhesive force between the membranes [83]. If divalent cations (1–30 mM Ca2+ or Mg2+ ) are present, giant vesicle formation is also observed with zwitterionic phospholipids alone (POPC), with the use of a procedure almost identical to the one just described [84]. Divalent cations seem to promote giant unilamellar vesicle formation in the case of POPC due to a binding of the ions to the free phosphate oxygen of the lipid head group, which is known to alter the mean head group conformation [85] and the fluidity of the lipid bilayer [86], and which makes a zwitterionic PC molecule positively charged overall [87]. In an independent study and with a different experimental procedure, the formation of giant vesicles from PCs (DOPC or soybean PCs) in the presence of Mg2+ ( Tm

solvent exchange at T > Tm

aqueous solution

hydration, mechanical treatment at T > Tm

(i) Figure 4. Simplified schematic representation of the principal pathways for the preparation of (normal) lipid vesicles in the case of “conventional amphiphiles” (surfactants that do not form a true vesicle phase at thermodynamic equilibrium). The pathways involve, as starting state of the amphiphiles, (i) preorganized dry lipids, which are hydrated and (possibly) mechanically manipulated above the Tm of the lipids; (ii) preorganized lipids in w/o emulsions (or w/o mircoemulsions or reversed micelles) or w/o/w emulsions prepared in a volatile solvent that is removed during the vesicle preparation procedure above Tm of the lipids; (iii) preorganized lipids in the presence of micelle-forming detergents (mixed detergent/lipid micelles) existing in dynamic equilibrium with free detergent monomers that are removed during the vesicle preparation procedure above the Tm of the lipids; or (iv) lipids dissolved in a solvent that is miscible with water and is exchanged with water during the vesicle preparation procedure above the Tm of the lipids. Once a vesicle suspension is formed, the mean vesicle size and size distribution can always be altered by mechanical treatments above Tm .

corresponding in a first approximation to a value a bit lower than the detergent’s cmc, the critical concentration for micelle formation determined separately under comparable conditions. The amount of monomeric detergent in the mixed micellar system is important, as it is this nonaggregated amphiphile that is removed from the solution during the detergent removal process, which finally leads to the formation of vesicles [31, 66, 69, 71, 244–247]. The principle of the detergent-depletion method is the following: mixed detergent/lipid vesicles, present in rapid equilibrium with detergent monomers, are put into a dialysis bag or another dialysis device [31] at a temperature above Tm of the lipid used [248]. The dialysis membrane is characterized by a permeability for the monomers, whereas the much larger mixed micelles cannot pass the membrane. Then, at a temperature above Tm [248], the dialysis device is put in contact with a buffer solution in which the mixed micelles were formed. Since the monomers can pass the dialysis membrane, the amount of monomers in the solution inside the dialysis device continuously and slowly decreases, and detergent monomers move from the mixed micellar aggregate into the bulk solution. This process continues until the amount of detergents in the micellar aggregates is so low that mixed micelles can no longer exist, and extended mixed bilayer fragments (sheets) and finally mixed lipid/detergent vesicles form. Extensive dialysis leads to the formation of vesicles that are almost (but not necessarily completely [249]) free from detergent molecules. These vesicle suspensions are often to a large extent unilamellar and have a narrow size distribution. The mean size depends on the experimental conditions, such as type of detergent used, initial lipid and detergent concentrations in the mixed micellar solution, and speed of detergent removal [247, 250–252]. Table 5 lists detergent molecules that are often used for the preparation of vesicles by the detergent-depletion method, together with characteristic size ranges of the mainly unilamellar vesicles formed. In the case of a system containing egg yolk PC and the bile salt sodium taurochenodeoxycholate (which aggregates itself stepwise into a particular type of unconventional, small micelles [253–255]), the mixed micelle-mixed vesicle transformation process—initiated by a rapid dilution process—has been investigated by time-resolved static and dynamic light-scattering measurements [256]. The scattering data analysis indicates that the key kinetic steps during vesicle formation are the rapid appearance of disc-like intermediate micelles, followed by growth of these micelles and closure of the large discs formed into vesicles [256]. In addition to detergent removal through dialysis, gel permeation chromatography [256, 257] (which is based on the partitioning of detergent monomers into the pores of a swollen gel matrix) or so-called Bio Beads [246, 247, 258, 259] (which bind detergent monomers) can also be applied. In the case of saturated PCs like DMPC and DPPC and octyl--d-glucopyranoside, the originally developed detergent-dialysis method has been modified slightly because of the relatively high Tm value of these lipids (see Table 4) [248]. The important modification is a slow dilution step before the actual dialysis procedure [248], resulting in mainly unilamellar vesicles with a mean diameter of 98 nm

61

Preparation of Vesicles (Liposomes) Table 5. Some of the detergents most often used for the preparation of vesicles by the detergent depletion method (Section 3.27) and approximate mean sizes of the vesicles formed in the case of egg yolk PCs.

Detergent

cmc at 25  C (mM)

Method for removing the detergent

Sodium cholate

∼11

Gel permeation chromatography Dialysis

Sodium glycocholate Sodium deoxycholate Sodium chenodeoxycholate n-Octyl--d-glucopyranoside

∼10 ∼4 ∼5 ∼23

Dilution and dialysis Dialysis Dialysis Dialysis Bio Beads

C12 EO8 (n-dodecyl octaethylenglycol monoether)

CHAPS (3-[(3cholamidopropyl) dimethylammonio]1-propane sulfonate)

∼0.08–0.09

∼5–10

Bio Beads

Dialysis

Reported approximate vesicle diameters (nm) 30 60 70 60–80 80–100 50–150 30–100 150 160 180 230 300–500 250 60–90

25–80 120 380

Refs. [257] [244] [252] [245] [251] [519] [143] [252] [252] [245, 524] [525] [246] [526] [247]

[527] [526] [519]

Note: The size of the vesicles may very much depend on the experimental conditions (see text). In particular, the resulting sizes may also depend on the presence of bilayer soluble substances (e.g., cholesterol [245, 246]) or in particular cases (cholate) on the presence of divalent cations (e.g., Ca2+ ) [519]. The approximative cmc values given in the table are taken for the bile salts from [520]; for C12 E8 from [521, 522]; for CHAPS from [522, 523]; and for n-octyl--d-glucopyranoside from [522].

(DMPC) and 94 nm (DPPC) under the corresponding conditions used [248]. It is quite generally possible to first simply dilute the mixed micellar system, followed by dialysis to completely remove the detergent [260]. If the detergent is not completely removed, the vesicle preparation by simple dilution will always contain detergent molecules, even if the dilution is 100- or 200-fold [250]. In a particular case [261], SUVs with a mean diameter of 23 nm were first prepared by sonication from egg yolk PC at a concentration of 20 mM. The detergent sodium deoxycholate was then added to give an aqueous mixture at a ratio of deoxycholate to PC of 1:2 (mol/mol). This mixture contained vesicles that were considerably larger than the SUVs used, because of the uptake of the detergent molecules. Deoxycholate was then removed to about 96–98% first by gel filtration and then almost completely by a second gel filtration. The final preparation—containing less than one deoxycholate molecule per PC molecule—were dispersed unilamellar vesicles with a mean diameter of 100 nm [261]. The same mean vesicle sizes were also obtained if instead of SUVs a dry egg yolk PC film was treated with deoxycholate, followed by bulk sonication and detergent removal [261]. Large-scale production of vesicles by detergent dialysis is possible (e.g., [262]), and commercial devices are available under the trade names Liposomat and Mini Lipoprep. The detergent depletion method is the method of choice for the reconstitution of water-insoluble membraneassociated proteins, which in a first step are extracted from the biological membrane by a mild detergent that does not lead to an irreversible protein denaturation [258, 263]. The

membrane protein-containing mixed detergent/lipid micelles are then converted to membrane protein-containing vesicles by one of the detergent removal techniques described above.

3.28. (Mixed) Vesicles Prepared by Mixing Bilayer-Forming and Micelle-Forming Amphiphiles As mentioned in Section 3.27, mixed lipid/detergent vesicles form transiently during detergent removal from detergent/lipid micelles. Such mixed vesicles can also be prepared by adding to preformed lipid vesicles (prepared by any method) above Tm an appropriate amount of a particular detergent [264, 265] or by simply diluting a mixed detergent/lipid micellar solution [266]. Examples of detergents that have been used include sodium cholate [267], n-octyl--D-glucopyranoside [268], or 3-[(3-cholamidopropyl)dimethylammonio]-1-propane sulfonate (CHAPS) [269]. The size of the resulting mixed lipid/detergent vesicles depends on the experimental conditions, such as the size (and lamellarity) of the initial vesicles, the chemical structure of the lipid, the lipid concentration, the chemical structure of the detergent, and the detergent concentration. Phosphatidylcholines like POPC, SOPC, DMPC, DPPC, or the mixtures extracted from egg yolk or soybeans are the amphiphiles whose mixed lipid/detergent vesicle (bilayer) formation has been studied most extensively. These PCs all have two long hydrophobic acyl chains containing 14 to 18 or 22 carbon atoms (see Tables 2 and 4).

62 PCs with two short acyl chains, containing fewer than 10 carbon atoms, do not form bilayers (vesicles) in dilute aqueous solution, but rather micelles (as long as the concentration is kept above the cmc) [270, 271]. While a very short-chain PC (1,2-dibutanoyl-sn-glycero-3phosphocholine) forms more or less spherical micelles, an increase in the chain length up to eight carbon atoms leads to the formation of extended, elongated micellar structures. All of these short-chain PCs are detergents in the sense of the term used here (Section 3.27) [272], and unilamellar vesicles composed of long-chain PCs and shortchain PCs can easily be prepared, for example, by the addition of a micellar solution of 1,2-diheptanoyl-sn-glycero3-phosphocholine (final concentration 5 mM) to a DPPC dispersion (final concentration 20 mM) in an aqueous solution [273]. The resulting vesicles are rather stable and have mean diameters below as well as above 1
m [274], depending on the experimental conditions, such as the ratio of the short-chain PC to the long-chain PC and cholesterol content [274]. One particular type of mixed lipid/detergent vesicular system (also containing ethanol)—not prepared by detergent dialysis, but by simple detergent and lipid mixing—are the so-called Transfersomes, vesicles that are claimed to be able to transfer water-soluble drugs through the skin to the blood circulation (transdermal drug delivery), particularly through the outermost physical barrier of the skin, the horny layer, called the stratum corneum. Transfersomes seem to be ultradeformable and may squeeze through the stratum corneum pores, which have diameters only one-tenth of the size of the Transfersomes. In this way, encapsulated watersoluble drugs may be transported across the skin [231]. Transfersomes are prepared by mixing an ethanolic solution of soybean PCs with (for example) sodium cholate to yield a suspension typically containing 8.7% (w/w) soybean PCs and 1.3% (w/w) cholate as well as approximately 8.5% (v/v) ethanol [275, 276]. For a reduction in vesicle size to about 500 nm, the suspension was treated by sonication, freezethaw cycles, and processing with a homogenizer. In the case of the particular non-ionic industrial amphiphile Brij 30 (n-dodecyl tetraethylene glycol ether, abbreviated as C12 E4 or C12 EO4 , also called Laureth4), vesicles have been prepared by simple dilution from so-called hydrotrope solutions, which contained in addition to C12 EO4 the micelle-forming surfactant sodium xylene sulfonate [277]. Through a detailed elaboration and analysis of the three-component phase diagram containing C12 EO4 , sodium xylene sulfonate, and water, the conditions under which vesicles form have first been established. Appropriate starting conditions were then chosen, and a corresponding dilution resulted finally in the desired vesicular system, which contained not only the vesicle-forming amphiphile, but also the hydrotrope (as constituent of the aqueous phase) [277].

3.29. Vesicles Prepared from Lipids in Chaotropic Ion Solutions This method is related to the detergent-depletion method described in Section 3.27. Chaotropic substances (e.g., urea, guanidinium hydrochloride, potassium thiocyanate,

Preparation of Vesicles (Liposomes)

trichloroacetate, or trichlorobromate) are known to disturb (break) the structure of water, and lipids are soluble in chaotropic solutions [278]. Solutions of trichloroacetate, for example, dissolve PCs as micellar solution [279], through a binding of the ions to the head groups, which leads to an increase in a0 (see Section 6). If a SUV suspension prepared from phospholipids in buffer solution by probe sonication is mixed with an aqueous solution containing sodium trichloroacetate, a micellar solution is first obtained (at about 1–3 M sodium trichloroacetate in the case of 0.1–0.5 mM phospholipids). This solution is then extensively dialyzed against a buffer solution to remove all of the chaotropic ions, resulting in the formation of uni- or oligolamellar vesicles with diameters between 10 and 20
m, although many small vesicles are also present [278]. If freeze-thaw cycles are used, the amount of chaotropic ions needed to generate giant vesicles is lowered considerably [278].

3.30. GVs Prepared from a W/O Emulsion with the Help of a Detergent This method involves the use of nonphospholipid amphiphiles in a four-step procedure [280]. First, a w/o emulsion is formed by homogenization of an aqueous solution, n-hexane and a mixture of soybean phospholipids and the non-ionic surfactant Span 80 (sorbitan monooleate). Second, n-hexane is removed by rotatory evaporation from the emulsion under reduced pressure, resulting in the formation of a water-in-lipid mixture. Third, this mixture is mixed with an aqueous solution containing a water-soluble detergent (such as sodium dodecyl sulfate, hexadecytrimethylammonium bromide, or Tween 80 (POE (20) sorbitan monooleate, which is a modified sorbitan monooleate that contains on average a total of 20 hydrophilic oxyethylene units). Fourth, the highly water-soluble detergent is removed by dialysis, leading to the formation of micrometer-sized vesicles.

4. PREPARATION OF REVERSED VESICLES There seems to be a kind of symmetry among most selforganized surfactant aggregates with regard to the distribution of the hydrophilic and the hydrophobic parts of the surfactants in the aggregate [281–283]. There are (normal) micelles (L1 -phase) and reversed micelles (L2 -phase), a (normal) hexagonal phase (HI  and a reversed hexagonal phase (HII , and (normal) emulsions (oil-in-water, o/w) and reversed emulsions (water-in-oil, w/o) [235]. In the case of vesicular aggregates, the existence of reversed vesicles has been demonstrated [283], for example, for a system containing the non-ionic surfactant tetraethyleneglycol dodecyl ether (n-dodecyl tetraethyleneglycol monoether, abbreviated R12 EO4 , C12 EO4 , or C12 E4 , the solvent n-dodecane, and water [281, 282]. These reversed vesicles form first upon the preparation of a mixture composed of 1 wt% water and 99 wt% of a n-dodecane solution which contains 2.5 wt% C12 EO4 [281]. After equilibration at 25  C, a two-phase system forms that is composed of a lamellar liquid crystalline

Preparation of Vesicles (Liposomes)

phase that is in equilibrium with an excess n-dodecane phase. When these two phases are mixed by hand-shaking, a heterogeneous dispersion of mainly multilamellar reversed vesicles forms with diameters of the reversed vesicles from less than 1
m to 10–20
m [281]. Strong vortexing leads to a reduction in size below 200 nm [282]. Without water, no reversed vesicles form. A reversed vesicle formed in an oil (in a water-immiscible, apolar solvent) is composed of an oily core and one or several closed “reversed” surfactant bilayer shells. The reversed bilayers are organized in such a way that the hydrated hydrophilic head groups of the amphiphiles are facing toward the center of the reversed bilayer, while the hydrocarbon chains are in contact with the oil [281]. The particular type of reversed vesicles formed from R12 EO4 , water, and n-dodecane seem to be rather unstable and convert back into the thermodynamically stable twophase system within hours or days [281]. Considerably more stable reversed vesicles can be obtained from a mixture of the non-ionic surfactants sucrose monoalkanoate (which is, to at least 95%, a monoester containing 10 wt% tetradecanoyl, 40 wt% hexadecanoyl, and 50 wt% octadecanoyl chains [283]), hexaethyleneglycol hexadecyl ether (abbreviated R16 EO6 , C16 EO6 , or C16 E6 , n-decane, and water [284]. The two surfactants are mixed at a weight ratio of sucrose monoalkanoate to C16 EO6 of 85:15, and vesicle formation is observed at typically 3 wt% surfactant in n-decane with a typical weight ratio of surfactant to water of 6 [283]. Treatment of the vesicle suspension with a probe sonicator at 30  C leads to a reduction in the reversed vesicle size to about 150–300 nm, depending on the experimental conditions, such as the water-to-surfactant ratio and the surfactant concentration [284]. The formation by vortexing of micrometer-sized reversed vesicles (and reversed myelin figures) has also been observed in m-xylene at a surfactant-to-water weight ratio around 0.8 for a number of polyethyleneglycol oleoyl ethers (abbreviated R181 EOx or C181 EOx , with mean values of x varying between 10 and 55) [285]. Furthermore, it seems that some phospholipids can also form reversed vesicles under particular conditions, as reported in the case of a system containing soybean phosphatidylethanolamine and triolein, saturated with water through vapor equilibration [286]. The particles formed—which are most likely reversed vesicles— have diameters in the range of 500 nm to 1.7
m [286].

5. CHARACTERIZATIONS AND APPLICATIONS OF VESICLES 5.1. Characterizations of Vesicles There are a number of typical properties that characterize a vesicle suspension, particularly the mean vesicle size, the vesicle size distribution, the mean number of lamellae per vesicle, and the chemical and physical stability [287]. All of these properties depend on (i) the chemical structure of the amphiphiles (or mixtures of amphiphiles) from which the vesicles are formed; (ii) the solution in which the vesicles are formed (e.g., salt and buffer content in the case of (normal) vesicles and aqueous solutions); and (iii) the method of vesicle preparation (see Section 3). A minimal characterization

63 of a vesicle suspension prepared is always needed. In many cases, a routine size measurement is appropriate, particularly by photon correlation spectroscopy (PCS), also called dynamic light scattering or quasi-elastic light scattering. PCS provides information about the homogeneity of a vesicle suspension, and it is relatively straightforward to find out whether the vesicles in a preparation are relatively monodisperse or whether they are polydisperse. For monodisperse and spherical vesicles, the mean hydrodynamic diameter of the vesicles can be determined relatively easily. In the case of polydisperse vesicle preparations and/or nonspherical vesicles, the interpretation of the scattering curves is more delicate, and it is recommended to apply additional methods as well, such as electron microscopy. A most probable vesicle size can be obtained by cryotransmission electron microscopy (cryo-TEM), although this method is limited to sizes below about 250 nm, and the sizes and morphologies obtained have to be taken with caution [288–290]. The use of different independent methods is always recommended. Table 6 lists a number of methods that are used for characterizing vesicle suspensions. In many cases, fluorescently labeled amphiphiles are used in the vesicle membranes at a concentration of 1–2 mol%. These amphiphiles contain fluorescent groups, and the behavior of these molecules is then taken as a measure for the behavior of the bulk vesicleforming amphiphiles, for example, for the determination of the lateral diffusion coefficient of the amphiphiles. Since the fluorescent groups are sometimes rather bulky, the kinetic constants obtained with these labeled molecules may be different from the actual membrane-forming lipids [291, 292]. The same is true for nitroxide-labeled amphiphiles used for ESR (electron spin resonance) measurements [292] (see Table 6). For the detection of the internal aqueous volume in (normal) vesicles—usually expressed as microliter of trapped aqueous solution per micromole of amphiphile (
l/
mol)— dye molecules are often used, which can be easily quantified spectroscopically. Meaningful results, however, can only be obtained if the dye molecules do not interact with the vesicle membrane (no incorporation inside the membrane, no adsorption to the membrane). In the case of permeability measurements, the use of dye molecules is convenient, and the release of vesicle-trapped molecules into the external medium can easily be quantified. In this case, however, one should always be aware that the permeability measured is the permeability of the particular dye molecules used under the particular experimental conditions. In this respect, NMR methods that are based on the use of paramagnetic shift reagents are better. The external addition of the shift reagents allows a distinction between the particular solute molecules present inside the vesicles and the solute molecules present outside of the vesicles, presuming that the shift reagent does not permeate the membranes. This has first to be proved. The permeability may very much depend on the experimental conditions, such as the presence of certain buffer species [293]. The list in Table 6 is certainly not complete. However, it gives a few hints about the general principles and problems of the vesicle characterization and lists some of the methods and techniques used. Depending on the vesicle preparation, a particular characterization may be more useful than

64

Preparation of Vesicles (Liposomes)

Table 6. Some of the principles and analytical methods used for characterizing the properties of vesicles (see also [66, 287]). Property Mean vesicle size and size distribution

Bilayer thickness

Vesicle shape and morphology

Methods used and comments

Ref.

PCS Analysis of the time dependence of intensity fluctuations in scattered laser light due to the Brownian motion of the vesicles, which is related to the mean hydrodynamic radius (Rh ) of the vesicles. If the vesicle suspension is very polydisperse and/or contains nonspherical vesicles (e.g., caused by an osmotic imbalance between the inside and the outside of the vesicles), the size analysis is rather complex and difficult

[2, 31, 528–531]

SLS Average scattering intensities are measured as a function of the scattering angle and the vesicle concentration, allowing the determination of the mean radius of gyration (Rg ) of the vesicles

[532–534]

FFEM Replicas of fractured vesicles are analyzed. The vesicles are fractured at low temperature (−100  C) under conditions where the water is in an uncharacterized (amorphous), “glassy” solid state. Nonequatorial fracturing leads to replicas that do not represent the “true” vesicle size

[1, 2, 136, 162, 300, 535]

Cryo-TEM The vesicles are directly observed at low temperature (−170  C) after rapid freezing (∼106  C/s)—under conditions where the water is in an amorphous solid state—to represent the state of the vesicles at the temperature from which the sample is cooled. The observed diameters of the vesicles correspond to the “true” vesicle diameters. Only applicable for vesicles with sizes below ∼250–500 nm. Not free of “artefacts” (“artifacts”), i.e., formation of microstructures due to specimen preparation, electron optics, or radiolytic effects [288–290]

[2, 3, 536]

LM Only for GVs

[4, 66, 89, 91, 536]

Size exclusion chromatography Conventional and high-performance liquid chromatography (HPLC) separation based on the principle that the partitioning behavior of vesicles in the pores of a solid column matrix depends on the size of the vesicles

[66, 386, 537]

NMR Size determinations possible from 31 P-NMR spectra of phospholipid vesicles that have sizes below 1
m with the help of a comparison with simulated spectra. For spectra simulation, the dynamics of the vesicles and the phospholipids in the vesicle membranes are considered

[136]

AFM The vesicles have to be deposited on a solid surface, which may lead to vesicle deformations and vesicle aggregations

[538, 539]

SANS Measurements in D2 O that may influence the vesicle size via altered head group interactions. Relative complex analysis involving fitting of experimental data

[540, 541]

X-ray Based on the interaction of X-rays with the electrons of the amphiphiles in the vesicles

[80, 300, 542]

Cryo-TEM

[543]

SANS Measurements in D2 O. Bilayer thickness determinations possible in the gel, in the ripple, and in the lamellar phase of the vesicle-forming surfactants (see Section 2)

[540, 541, 544–546]

LM Only for GVs. Detection of MVVs

[547–552]

FFEM Detection of MVVs Negative staining EM Heavy metal ions have to be added to the vesicle dispersion, which always alters the vesicles. Furthermore, the vesicles are observed in a dry state

[12]

continued

65

Preparation of Vesicles (Liposomes)

Table 6. Continued Property

Lamellarity

Methods used and comments

Ref.

Cryo-TEM Only for vesicles with sizes below ∼250–500 nm

[553]

Cryo-TEM Only for vesicles with sizes below ∼250–500 nm

[3, 293]

Negative Staining EM Addition of heavy metals (osmium tetroxide) needed. The number of lamellae detected may not represent the “true” number of lamellae

[12, 31]

NMR P-NMR in the case of phospholipid vesicles. Externally added membraneimpermeable Mn2+ influences the 31 P-NMR signal of the phospholipids in the outermost monolayers by broadening the resonance beyond detection. The resonances of the inner phospholipids are unaffected. Paramagnetic shift reagents may also be used (e.g., Pr3+ and Eu3+ )

[161, 162, 293, 554]

Fluorescence quenching An appropriate amphiphilic fluorescent probe molecule is used in the vesicle preparation, and the fluorescence of the amphiphiles in the outermost layers is quenched by adding to the vesicles a membrane-impermeable reagent

[555]

SAXS Based on the interaction of X-rays with the electrons of the amphiphiles in the vesicles

[556]

Chemical modification Functional groups present in the hydrophilic head groups of the amphiphiles in the external layer(s) of the vesicle and exposed to the bulk aqueous solution are chemically modified by externally added reagents that do not permeate the membranes and therefore do not react with the amphiphiles that are present in the inner layer(s) of the vesicles with their head groups exposed to the trapped aqueous space. The chemically modified amphiphiles are then quantified

[31]

Two-photon fluorescence microscopy Use of membrane soluble fluorescent probes. Only for GVs

[557]

NMR P-, 13 C-, or 2 H-NMR

[558, 559]

Confocal fluorescence microscopy Distinction between solid-ordered and fluid-disordered domains in vesicle membranes made with two fluorescent probes that have different affinities for the two domains. Only for GVs

[560]

DSC A vesicle sample and an inert reference are heated independently to maintain an identical temperature in each. In endothermic solid-ordered/liquid-ordered phase transitions, heat is required in excess in the vesicle sample over the heat required to maintain the same temperature in the reference

[43, 300, 540]

Fluorescence Fluorescence polarization measurements of a membrane soluble fluorescent probe (e.g., 1,6-diphenylhexatriene, DPH)

[31, 561, 562]

CD Induced circular dichroism below the Tm of an achiral probe (1,6-diphenylhexatriene, DPH) embedded inside the vesicle membrane constituted by chiral amphiphiles. No CD signal above Tm

[563]

NMR H-, 2 H-,

[37, 564–567]

31

Lipid domains in the membranes (rafts)

31

Phase transition temperature (Tm )

Mobility of the amphiphiles in the vesicle membranes

1

13

C-,

31

P-NMR

ESR Use of spin-labeled amphiphiles

[568]

Quasi-elastic neutron scattering

[569, 570] continued

66

Preparation of Vesicles (Liposomes)

Table 6. Continued Property

Methods used and comments

Ref.

Fluorescence Use of fluorescent amphiphiles

[571]

Fluorescence correlation spectroscopy Analysis of the translational diffusion of a fluorescent probe. Only for GVs

[560]

Fluorescence recovery after photobleaching [37] Lateral diffusion of fluorescently labeled amphiphiles determined by first bleaching the lipids with an intense laser pulse, and then analyzing the fluorescence recovery kinetics in the bleached area, due to the diffusion of unbleached molecules into the previously bleached area Molecular conformation of the NMR amphiphiles in the vesicle 13 C- or 2 H-NMR using labeled phospholipids membranes

Membrane fluidity and order

[300, 559, 564, 565, 572]

FT-IR and Raman Determination of the equilibrium conformational characteristics of the amphiphiles

[37, 300, 573]

CD For vesicles composed of chiral amphiphiles

[563]

Fluorescence [37] Use of fluorescent membrane probes. Measurements of static or time-resolved fluorescence anisotropy NMR Use of partially and specifically deuterated amphiphiles

[300, 564]

ESR Use of amphiphiles that have a nitroxide radical. The ESR spectrum of these amphiphiles is sensitive to the motions of the molecules

[37]

Transmembrane lipid diffusion ESR (flip-flop) Use of spin-labeled amphiphiles (nitroxide radical) Fluorescence Use of fluorescently labeled amphiphiles

[292, 574] [292]

Surface charge

Zeta potential measurements, microelectrophoresis Problematic, since the method is derived from classical theories of the double layer that do not include specific ion effects (see Section 6). Limited to particular background salt conditions (usually NaCl)

Internal volume and entrapment (encapsulation) yield

Use of dye molecules [31, 66, 575–577] The dye molecules (often fluorescent) are water-soluble and should not interact with the vesicle membrane. The dye molecules are entrapped during vesicle preparation, and the amount of entrapped dye molecules is determined quantitatively either after separation of the nonentrapped molecules from the vesicles or by addition to the vesicles of a membrane-impermeable reagent, which leads to a complete quenching of the fluorescence of the externally present dyes (e.g., calcein with quencher Co2+ ) [575]. Simple dilution of the externally present dye may also be possible [577] NMR [554] 17 O-NMR of the water oxygen. Addition of the membrane-impermeable paramagnetic shift reagent DyCl3 to the vesicles below Tm leads to a shift in the 17 O resonance corresponding to the external water. The internal water peak remains the same. Above Tm , only one peak is observed because of rapid water equilibration

Membrane permeability

Use of dyes or radioactively labeled molecules [31, 578] Determination of the release of dye molecules entrapped inside the aqueous interior of the vesicles as a function of time under particular storage conditions NMR [579–582] Use of 1 H- or 17 O-NMR and an externally added membrane-impermeable paramagnetic shift reagent (Mn2+ , Pr3+ ) to distinguish between internal and external permeants continued

67

Preparation of Vesicles (Liposomes) Table 6. Continued Property

Chemical stability

Physical stability

Methods used and comments

Ref.

Use of ion-selective electrodes The ions present outside of the vesicles can be detected

[583]

Thin-layer chromatography Analysis of possible degradation product due to hydrolysis and oxidation reactions

[72]

High-performance liquid chromatography Analysis of possible degradation product due to hydrolysis and oxidation reactions

[72]

Turbidity, PCS, FFEM, Cryo-TEM, and others Analysis of possible size changes during storage due to vesicle aggregation and fusion

[66, 69, 584]

Abbreviations: AFM, atomic force microscopy; CD, circular dichroism; cryo-TEM, cryo-transmission electron microscopy (also called cryo-fixation); DSC, differential scanning calorimetry; EM, electron microscopy; ESR, electron spin resonance (also called electron paramagnetic resonance, EPR); FFEM, freeze-fracture electron microscopy; FT-IR, Fourier transformed infrared; LM, light microscopy; NMR, nuclear magnetic resonance; PCS, photon correlation spectroscopy (also called dynamic light scattering (DLS) or quasi-elastic light scattering (QELS)); SANS, small-angle neutron scattering; SAXS, small-angle X-ray scattering; SLS, static (or classical) light scattering.

another, and certain methods may not be applied at all (e.g., 31 P-NMR obviously cannot be used for the determination of the lamellarity of vesicles that are not composed of phosphorous-containing amphiphiles).

5.2. Applications of Vesicles Lipid vesicles are used successfully in many different fields as interesting and versatile submicrometer- or micrometersized compartment systems [69, 294–297]. This wide applicability of vesicles and the broad interest in vesicles can be understood at least on the basis of the following four reasons: (i) Lipid vesicles can be considered membrane or biomimetic systems [298, 299], since the molecular arrangement of conventional vesicle-forming amphiphiles in a vesicle is a (more or less curved) bilayer, like the lipid matrix in biological cell membranes [97, 300] or in the outer coat of certain viruses [97]. (ii) Vesicles prepared from amphilphiles present in biological systems allow applications as biocompatible and biodegradable systems. (iii) Water-soluble as well as certain water-insoluble molecules can be entrapped inside the aqueous or hydrophobic domains of the vesicles, allowing the use of vesicles as carrier systems and nanometer- or micrometer-sized reaction compartments. (iv) Vesicles can be prepared not only from the conventional PC type of bilayer-forming amphiphiles, but also from a large number of different nonphospholipid surfactants (or mixtures of surfactants), allowing the preparation of application-tailored and specifically designed and functionalized systems. Among the more than 30,000 known surface active compounds [301], a large number of surfactants and surfactant mixtures (including many nonphospholipid amphiphiles) have been reported to form vesicles. The basic principles that lead to the formation of vesicles are the same for all, that is, the requirements of (i) an effective packing parameter p (=v/a0 leff ≈ 1 (see Section 1.1), (ii) chain flexibility (T > Tm ), and (iii) sufficiently low amphiphile concentration

(global packing constraints) [7]. Examples of vesicle-forming amphiphiles include • cationic di-n-alkyldimethylammonium ions [302, 303] (such as di-n-dodecyldimethylammonium bromide (DDAB) [304] and other di-n-dodecyldimethylammonium halides [305] and di-n-octadecyldimethylammonium chloride [306–308] or bromide [308, 309] or other counter-ions [310]); • the cationic oleyldimethylaminoxide [311]; • anionic phospholipids (such as egg yolk phosphatidic acid mixtures and ox brain phosphatidylserine mixtures [312, 313]); • anionic di-n-alkylphosphates [314–316] (e.g., di-n-dodecylphosphate [317], di-n-hexadecylphosphate [318– 322], and various di-polyprenylphosphates [323, 324]); • anionic linear and branched monoalkylphosphates (such as different polyprenylphosphates [325], ndodecylphosphate [326], 6-propylnonylphosphate [327], 4-butyloctylphosphate [327], and 2-pentylheptylphosphate [327]); • anionic tridecyl-6-benzene sulfonate in the presence of salt (sodium chloride) [328]; • anionic fatty acid/soap mixtures (e.g., n-octanoic acid/noctanoate [329], n-decanoic acid/n-decanoate [329– 331], oleic acid/oleate [332–334]); • anionic surfactant/alcohol mixtures (such as sodium n-dodecylsulfate/n-dodecanol [329, 335] and sodium n-decanoate/n-decanol [336], or sodium oleate/noctanol [337], which can form a highly viscous phase of densely packed vesicles); • mixtures of cationic and anionic surfactants, so-called catanionic mixtures (such as n-hexadecyltrimethylammonium tosylate/sodium n-dodecylbenzenesulfonate [20, 59], n-hexadecyltrimethylammonium bromide/ sodium n-octylsulfate [338–342], and sodium n-dodecylsulfate and di-n-dodecyldimethylammonium bromide [343, 344]); • ganglioside GM3 [345, 346]; • phosphatidylnucleosides (such as 1-palmitoyl-2-oleoylsn-glycero-3-phosphocytidine [347, 348]); • diblock copolymers [24] (such as the ethylene oxide (EO)/butylene oxide (BO) diblock copolymers EO6 BO11 , EO7 BO12 , EO11 BO11 , EO14 BO10 , and EO19 BO11 [349]);

68

Preparation of Vesicles (Liposomes)

and many more (e.g., [33, 358, 369]). Finally, complex vesicle-based surfactant aggregates can be prepared (e.g., large vesicles composed of one or more types of amphiphiles may contain vesicles of another type of amphiphile), based on principles that include the specific molecular recognition between different types of vesicles initially prepared by one of the established methods described in Section 3 [370–372]. It is rather obvious that one consequence of the fact that vesicle formation is observed from so many different classes of amphiphiles (or mixtures of amphiphiles) is the very broad range of applications in very different fields. Vesicles are applied—or investigated for potential applications—at least in the following areas:

• in research related to the question of the origin and evolution of life, as models for the precursor structures of the first cells [400]; • in research aimed at constructing artificial (or minimal) cells [401, 402], for example, for potential biotechnological applications [402]; • in food technology and nutrition as carrier systems for food additives and ingredients and for the control of certain food processes (e.g., cheese ripening) [403– 405]; • in agrochemistry [406]; • as nanometer- or micrometer-sized bioreactors containing catalytically active enzymes [75, 407–411]; • in nanoparticle technology [190, 299, 412–414], for example, for the preparation of semiconductor particles [412]; • in catalytic processes as simple models for enzymes [415–420] or simple models of other protein functionalities (e.g., as catalysts for the unfolding [421] or folding [387, 422] of proteins); • in biosensor developments [423–425], particularly for the controlled preparation of bilayers adsorbed to solid surfaces [426–429]; • in the extraction of heavy metal ions with the help of functionalized, metal-sorbing vesicles [430]; • as supramolecular, nanostructured polymeric materials (as polymerized vesicles) [431–433]; • in biomineralization [434–436]; • as templates for the synthesis of inorganic mesoporous materials [437, 438] or biomaterials [436] and in the preparation of hollow polymer capsules [356, 439–441]; • as templates for modifying the distribution of reaction products, for example, in reactions that lead to products that are only sparingly soluble in the absence, but soluble in the presence, of vesicles [442, 443]; • as a medium for the preparation of size-defined polymer particles from monomers that are soluble in the vesicle shell [444]; • as supramolecular, self-assembly-based devices [33, 298, 299, 358, 445], for example, for the conversion of light energy into chemical energy (artificial photosynthesis) [446, 447], for signaling and switching [448, 449], for the construction of molecular wires [450], and for a number of different redox processes [418];

• in pharmacology and medicine [69, 73, 296, 297, 373– 378] as parenteral or topical drug delivery systems [297, 379–381], in the treatment of infectious diseases, in anticancer therapy, as gene delivery systems, as immunoadjuvants, and as diagnostics; • in immunoassays [382–385]; • in chromatographic separations using immobilized vesicles [386, 387]; • in cosmetics as formulations for water and nutrient delivery to the skin [388–391]; • in a variety of biophysical investigation of biological membrane components, including the reconstitution and use of membrane-soluble proteins [4, 69, 258, 263, 392]; • in research on membrane-soluble ion channels [393– 399];

and many more (e.g., [295, 299]). One illustration of vesicles loaded with water-soluble molecules is shown in Figure 5. A cryo-transmission electron micrograph of POPC vesicles containing the protein ferritin is shown. The vesicles were prepared in the presence of ferritin, and the nonentrapped protein molecules were separated from the vesicles by size exclusion chromatography after vesicle preparation [451, 452]. Since this particular protein has a dense iron core with a size of about 8 nm, it is visible by electron microscopy, and the actual number of protein molecules per individual vesicle can be directly counted. This type of loaded vesicles can be used, for example, in basic studies on vesicle transformation processes [451, 452]. If the vesicles contain catalytically active proteins (enzymes), they may be used in drug delivery or as small bioreactors [75, 410, 411].

• triblock copolymers (such as the ethylene oxide (EO)/propylene oxide(PO)/ethylene oxide (EO) triblock copolymer EO5 PO68 EO5 (called Pluronic L121) [350] and the polymerizable poly(2-methyloxazoline)/ poly(dimethylsiloxane)/ poly(2-ethyloxazoline) triblock copolymers [351]); • polymerizable amphiphiles [33, 352, 353]; • perfluorated surfactants [354] (such as short-chain perfluorophosphocholines [355] and perfluoroalkyl PCs [356]); • bolaamphiphiles, which are membrane-spanning amphiphiles with two hydrophilic head groups at the two ends of the molecule and actually form vesicles containing not bilayered but monolayered shells [357–360]; • gemini surfactants that contain two hydrophobic tails and two hydrophilic head groups linked together with a short linker [361, 362]; • industrial, not very well defined surfactant/cosurfactant mixtures (such as N -methyl-N -alkanoyl-glucamine/ octanol or oleic acid mixtures [363]); • calixarene-containing [364] or cryptand-containing amphiphiles [365]; • fullerene-containing amphiphiles with two hydrophobic chains and two charged head groups [366]; • double-chain amphiphiles with a polar alkoxysilyl head group, allowing the preparation of a kind of organicinorganic hybrid vesicle, called cerasomes [367]; • triple-chain amphiphiles containing three hydrophobic chains and two charged polar head groups [365, 368];

Preparation of Vesicles (Liposomes)

Figure 5. Cryo-transmission electron micrograph of two unilamellar vesicles that have been prepared from POPC by the reversedevaporation technique (see Section 3.14) loaded with the iron storage protein ferritin, followed by extrusion (REV-VET100 ). Nonentrapped ferritin molecules were removed by size-exclusion chromatography. Each black spot inside the vesicles represents one iron core (the core of an individual protein). The length of the bar corresponds to 100 nm. The electron micrograph was taken by M. Müller and N. Berclaz, Service Laboratory for Electron Microscopy, at the Department of Biology at the ETH Zürich. See [451, 452] for experimental details and for an application of ferritin-containing vesicles in the investigation of vesicle transformation processes.

Conceptually, an interesting principle of vesicle applications is vesicle transformation (at the site of application) to another type of surfactant assembly as a result of temperature changes that lead to changes in the surfactant packing parameter p (see Section 1.1). One particular example is the transformation of a 1-monoolein MLV suspension (L·  phase) into a 1-monoolein bicontinuous cubic phase [9, 453– 456]. There is no doubt that the number of applications will increase with increasing molecular complexity of the vesicular systems. Vesicles—and other surfactant assemblies (e.g., hexagonal and cubic phases)—will be applied more and more in all fields related to nanoscience and nanotechnology. The title of a recent publication in the field of vesicle drug delivery is an example of one of the directions in which vesicle research and application may go: “Biotinylated Stealth Magnetoliposomes” [457], a particular vesicle preparation that combines steric stabilization (Stealth) with molecular recognition (biotin) and magnetic nanoparticle properties. In other words, it is likely that the complexity of the vesicles prepared and investigated will increase to make them functional, possibly by combining the principles of supramolecular chemistry and surfactant self-assembly, to prepare nanometer- or micrometer-sized synthetic systems that may carry out some of the functions biological systems do rather efficiently [458, 459].

6. CONCLUDING REMARKS There are many different methodologies that have been described in the literature for the preparation of lipid vesicles. Some (but not all), and certainly the most prominent and well known, are mentioned in this review, focusing on normal lipid vesicles (Section 3), although reversed vesicles are briefly mentioned as well (Section 4).

69 As outlined in Sections 1 and 2, the use of a number of different terms in the field of vesicles (and surfactant assemblies at large) is often rather confusing, and confusion even exists over the general use of the term “surfactant.” It is worth pointing out that a reduction in the surface tension of water by surfactants can be achieved not only with classical micelle-forming, single-chain amphiphiles (such as sodium n-dodecyl sulfate, SDS) [460], but also with typical vesicle-forming double-chain phospholipids (such as SOPC or DPPC). Certainly, the kinetics and extent of adsorption of a particular surfactant at the water-air interface very much depend on the precise experimental conditions [461]. In the case of DPPC, for example, the equilibrium surface tension and rate of monolayer formation at the water-air interface depend on the temperature [462]. Below Tm , in the solidordered state of the molecules (see Section 2), the decrease in the surface tension of water by DPPC is considerably lower than that above Tm [462]. However, there is no doubt that glycerophospholipids like SOPC, POPC, or DPPC are surfactants in the sense of the definition of this term given in Section 1.1. Since in dilute aqueous solution many of the bilayerforming surfactants known (at least the most studied phospholipids) at thermodynamic equilibrium do not form a true vesicle phase (with a defined vesicle size, size distribution, and lamellarity), but under the appropriate conditions rather form a lamellar phase (L·  , L·  , L·  or P ) with extended stacked bilayers, the preparation of vesicle dispersions necessarily involves the use of a particular method, a particular technology. Depending on the experimental conditions, as mentioned in Section 3, depending on the chemical structure of the amphiphile or mixtures of amphiphiles used, the amphiphile concentration, the substances that may be encapsulated in the vesicle’s aqueous interior or in the vesicle membrane, etc., there may be one particular method that is more advantageous in comparison with others. Some of the methods described in Section 3 involve the use of organic solvents. This may be a problem for certain applications. Some methods can be scaled up to a bulk preparation [287], some methods involve mechanical treatments that may cause an inactivation of sensitive molecules one may like to entrap [75]; some methods are cheaper than others; some methods lead to very small vesicles (SUVs); other methods lead mainly to MLVs or LUVs or GUVs, or even MVVs; some methods are fast; some methods are relatively time consuming, etc. There is no “best method.” The choice of a certain method that may be useful very much depends on the particular problem one is trying to solve [67]. There are, at least in the case of the conventional type of phospholipids (or in the case of lipid mixtures containing phospholipids, particularly PC), a few general findings that are often valid; but one should be aware of the existence of possible exceptions: • The hydration of a dried thin lipid film above the Tm of the lipids usually leads to the formation of MLVs if the film is dispersed vigorously (vortexing or hand-shaking) (see Section 3.2). • The hydration of a dried thin lipid film above the Tm of the lipids usually leads to the formation of GUVs if

70

•

•

•

• •

•

•

Preparation of Vesicles (Liposomes)

the film is undisturbed while being hydrated slowly (see Section 3.2). The electroformation method can be used for the preparation of GUVs with diameters between about 10 and 50
m (or more) in the case of certain lipids (and lipid mixtures) and buffers with low ionic strength (Section 3.3). Relatively homogeneous preparations of SUVs with diameters around 50 nm or below can often be obtained by probe sonication (Section 3.5), although degradation of the surfactants and consequences of the possibly present metal particles (which may be a source of nucleation for vesicle transformation processes) have to be taken into account. Rather homogeneous preparations of LUVs with diameters around 50 or 100 nm can be obtained by the extrusion technique as FAT-VET50 or FAT-VET100 (Section 3.8, Fig. 3). The dehydration-rehydration method usually results in high encapsulation yields (Section 3.7). Freeze-thaw cycles may lead to vesicle size homogenization and solute equilibration between the external bulk solution and the trapped aqueous solution (Section 3.6). The detergent-depletion method often results in rather uniform vesicles with sizes below 100 nm, although the possibility of an incomplete removal of the detergent should be considered (Section 3.27, Table 5). The use of volatile cosolvents (oils) during vesicle preparation is often based on the principle that either w/o or w/o/w emulsions are formed from which the solvent is removed, limiting the solvents to all those that have a boiling point considerably lower than the boiling point of water (Sections 3.14–3.17, 3.24, and 3.25).

The physicochemical properties (such as chemical and physical stability, membrane permeability) of the vesicles formed very much depend on the surfactant (or surfactant mixtures) used. Depending on the surfactant, the characteristic properties may be very different from those of the conventional PC type of vesicles. Furthermore, the whole equilibrium phase behavior may be different, and cases are known where even true vesicle phases seem to form “spontaneously” [7, 463–465] and seem to exist at thermodynamic equilibrium [7, 59, 60, 463]. In these cases, vesicles of a certain size and lamellarity just form by mixing, independent of the method of preparation. The stability of the vesicles is understood in the case of the so-called catanionic vesicles (which are composed of a mixture of positively and negatively charged amphiphiles) on the basis of a most likely uneven distribution of the amphiphiles in the two bilayers. This allows the required differences in curvature and molecular packing in the two halves of a bilayer [60, 466], similar to the case of lipid vesicles prepared from mixtures of long-chain and short-chain PCs (Section 3.28) [273, 274]. With respect to the spontaneity in the formation of this type of vesicles, it has been argued, however, that shear forces present during the preparation (mixing of solutions) play an important role in vesicle formation [467]. An interesting case of unilamellar vesicles as thermodynamic equilibrium state has been described in the case

of certain ionized phospholipids (e.g., 1,2-dimyristoyl-snglycero-3-phosphoglycerol, DMPG) [468–472]. Under the experimental conditions used, unilamellar DMPG vesicles apparently only form at the critical temperature (T ∗ ) of 31.6  C [471], which is different from Tm . Above T ∗ MLVs are formed; below T ∗ the phospholipids arrange into a sponge phase [471]. Further investigations are needed to fully understand this critical phenomenon and to clarify whether this unilamellar vesicle formation is a particular case or whether it can be more generally observed. With the exception of a few cases—such as the detergent depletion method (Section 3.27) (e.g., [256])—the general mechanism of vesicle formation is not yet completely understood in its details, although general principles have been elaborated [7, 69, 473–477]. From a more practical point of view, and by looking at all of the methods described in Section 3, it is evident that the vesicles often form from a preorganized state of the lipids. This preorganization may be (i) Lamellar sheets present in a dry film deposited on a solid surface (Sections 3.2 and 3.3) or in the ethanolic pro-liposome state (Section 3.19). (ii) W/o or w/o/w emulsion droplets (Sections 3.14, 3.15, and 3.30). (iii) Mixed micelles in the case of the detergent-depletion method (Section 3.27) or micelles in the case of chaotropic ion solutions (Section 3.29). In a few cases, the vesicle formation is triggered as a result of a direct contact of a nonorganized state of the lipids with an aqueous environment, as in the case of the Novasome technology (Section 3.10), the ethanol injection method (Section 3.18), or the ether injection method (Section 3.24). A simple summarizing schematic representation of some of the different pathways for the formation of (normal) lipid vesicles is given in Figure 4. It is expected that more methods for the preparation of vesicles will be developed, although the general pathways may remain the same. The range of pre-organized starting systems from which vesicles can be formed may be expanded, and more will probably soon be understood for the vesicle formation at large. (Normal) vesicles are just a particular state of aggregation of surfactants (or surfactant mixtures) in an aqueous solution that are topologically closed with an internal aqueous space. Vesicles can be unilamellar or multilamellar, and the general principles of surfactant assembly have been outlined [7–11]. As mentioned above, vesicles sometimes appear to be thermodynamically stable, sometimes not. Like all self-assembled amphiphile aggregates—micelles, microemulsions, cubic phases, even biological membranes, etc.—the formation of vesicles depends on only a few things: (i) local curvature, (ii) global packing constraints (including interaggregate interactions), and (iii) flexibility of the hydrophobic chain(s). All of the methods described essentially depend on satisfying these criteria. Physicochemical conditions of inside and outside of vesicles often differ greatly. This is a consequence of the closed topology. This fact, together with adverse packing conditions, can often result in a stable state of so-called supra-selfassembly (e.g., surfactant micelles existing inside surfactant vesicles) [478–480].

71

Preparation of Vesicles (Liposomes)

The basic principles that underlie self-assembly of vesicles are quite general and are firmly based in thermodynamics and statistical mechanics [7, 9, 10, 481]. The requirements for “designing” a vesicle are conceptually simple: a packing parameter close to unity (which means effectively a doublechain surfactant or mixed single-chain surfactants), flexible hydrophobic chains (a temperature above Tm ), and control of inter- and intraaggregate interactions. Local and global packings are the key principles to consider. Vesicles, however, are a very small part of a much larger class of self-assembled surfactant aggregates that include cubic phases, which are usually bicontinuous structures of zero average curvature. “Bicontinuous” means that both the aqueous and “oily” parts of the structure are continuously connected over the whole system. Cubic phases or bicontinuous structures in general are ubiquitous in biology for directing biochemical traffic [11]. Likewise, hexagonal phases and microtubules are close to lamellar (and vesicular) phases in a phase diagram [11]. Although these things are known and are even beginning to be understood, quantitative predictions remain a problem. This can be traced to the fact that the underlying theory of molecular forces that underpins physical chemistry and colloid and surface science is flawed [482, 483]. Previous theories cannot deal with specific ion effects (so-called Hofmeister series), dissolved gas, and other solutes that change the water structure. There is rapid development in this area at the moment, which is likely to provide predictability in vesicle design [482–484], which is certainly what one is aiming for.

GLOSSARY Amphiphile A molecule that comprises at least two opposing parts, a solvophilic (for example “hydrophilic”, meaning water-loving) and a solvophobic (for example “hydrophobic”, meaning water-hating). Amphiphiles are surfactants. Detergent A surfactant that in dilute aqueous solution forms micelles, spherical or non-spherical aggregates that contain in the interior of the aggregate the hydrophobic part of the surfactant and on the surface the hydrophilic part of the surfactant. Glycerophospholipid A particular phospholipid that contains a glycerol backbone to which a phosphate group is bound. Lamellar phase, L A particular liquid crystalline equilibrium state of surfactant molecules, also called liquiddisordered state. In an aqueous environment, the surfactant molecules are arranged in layers in which the hydrophobic part of a surfactant is in the interior of the layer and the hydrophilic part on the two surfaces of the layer, exposed to the aqueous environment. Liposome Vesicle prepared from amphiphilic lipids. Phase transition temperature, Tm Characteristic temperature (also called lamellar chain melting temperature) of surfactants that form a lamellar phase. Above the phase transition temperature, the surfactant molecules are in a liquid-disordered state, below in a solid-ordered state.

Phosphatidylcholine, PC A particular glycerophospholipid that contains a choline group in the hydrophilic part, bound to the phosphate. Phospholipid A surfactant that is present in some of the biological membranes and contains at least one phosphate group. Reversed vesicle Inverted vesicle formed in a waterimmiscible, apolar solvent in the presence of a small amount of water. Surfactant A molecule that is surface active, meaning that it accumulates at the surface of liquids or solids. Surfactants are amphiphiles. Vesicle General term to describe any type of hollow, surfactant-based aggregate composed of one or more shells. In the biological literature, the term vesicle is used for a particular small, membrane-bounded, spherical organelle in the cytoplasm of an eukaryotic cell.

ACKNOWLEDGMENTS The author thanks Barry W. Ninham (Department of Applied Mathematics, Australian National University, Canberra, Australia) for extensive discussions on the principles of surfactant self-assembly and for critically reading the manuscript. Manuscript reading and literature advice by Pasquale Stano (Department of Materials Science, ETH, Zürich, Switzerland) are also acknowledged. Furthermore, discussions with Saša Svetina (Institute of Biophysics, University of Ljubljana, Slovenia), Martien A. Cohen Stuart, Frans A. M. Leermakers, and Mireille M. A. E. Claessens (all from the Laboratory of Physical Chemistry and Colloid Science, Wageningen University, Wageningen, The Netherlands) are appreciated.

REFERENCES 1. H. W. Meyer and W. Richter, Micron 32, 615 (2001). 2. S. U. Egelhaaf, E. Wehrli, M. Müller, M. Adrian, and P. Schurtenberger, J. Microsc. 184, 214 (1996). 3. M. Almgren, K. Edwards, and G. Karlsson, Colloids Surf. A: Physicochem. Eng. Asp. 174, 3 (2000). 4. P. L. Luisi and P. Walde, Eds., Giant Vesicles, Perspectives in Supramolecular Chemistry, Vol. 6. Wiley, Chichester, 2000. 5. B. A. Cornell, J. Middlehurst, and F. Separovic, Biochim. Biophys. Acta 598, 405 (1980). 6. Y. Tahara and Y. Fujiyoshi, Micron 25, 141 (1994). 7. D. J. Mitchell and B. W. Ninham, J. Chem. Soc., Faraday Trans. 2 77, 601 (1981). 8. B. W. Ninham and D. F. Evans, Faraday Discuss. Chem. Soc. 81, 1 (1986). 9. J. N. Israelachvili, D. J. Mitchell, and B. W. Ninham, Biochim. Biophys. Acta 470, 185 (1977). 10. J. N. Israelachvili, “Intermolecular and Surface Forces,” 2nd ed. Academic Press, London, 1992. 11. S. Hyde, S. Andersson, K. Larsson, Z. Blum, T. Landh, S. Lidin, and B. W. Ninham, “The Language of Shape.” Elsevier, Amsterdam, 1997. 12. A. D. Bangham and R. W. Horn, J. Mol. Biol. 8, 660 (1964). 13. A. D. Bangham, M. M. Standish, and J. C. Watkins, J. Mol. Biol. 13, 238 (1965). 14. F. Szoka, Jr. and D. Papahadjopoulos, Annu. Rev. Biophys. Bioeng. 9, 467 (1980).

72 15. G. Sessa and G. Weissmann, J. Lipid Res. 9, 310 (1968). 16. S. C. Kinsky, J. Haxby, C. B. Kinsky, R. A. Demel, and L. L. M. van Deenen, Biochim. Biophys. Acta 152, 174 (1968). 17. D. D. Lasic, in “Handbook of Nonmedical Applications of Liposomes” (D. D. Lasic and Y. Barenholz, Eds.), Vol. I, p. 1. CRC Press, Boca Raton, FL, 1996. 18. R. T. Hamilton and E. W. Kaler, J. Colloid Interface Sci. 116, 248 (1987). 19. E. M. A. Pareira, D. F. S. Petri, and A. M. Carmona-Ribeiro, J. Phys. Chem. B 106, 8762 (2002). 20. E. W. Kaler, K. L. Herrington, A. K. Murthy, and J. A. N. Zasadzinski, J. Phys. Chem. 96, 6698 (1992). 21. R. G. Laughlin, Colloids Surf. A: Physicochem. Eng. Asp. 128, 27 (1997). 22. A. T. Florence, in “Liposome Technology” (G. Gregoriadis, Ed.), Vol. 1, p. 157. CRC Press, Boca Raton, FL, 1984. 23. I. F. Uchegbu and A. T. Florence, Adv. Colloid Interface Sci. 58, 1 (1995). 24. B. M. Discher, Y.-Y. Won, D. S. Ege, J. C.-M. Lee, F. S. Bates, D. E. Discher, and D. A. Hammer, Science 284, 1143 (1999). 25. L. Luo and A. Eisenberg, J. Am. Chem. Soc. 123, 1012 (2001). 26. D. E. Discher and A. Eisenberg, Science 297, 967 (2002). 27. IUPAC-IUB Commission on Biochemical Nomenclature, Eur. J. Biochem. 2, 127 (1967); Eur. J. Biochem. 79, 11 (1977); Biochem. J. 171, 21 (1978). 28. J. R. Silvius, in “Phospholipids Handbook” (G. Cevc, Ed.), p. 1. Dekker, New York, 1993. 29. A. Streitwieser, C. H. Heathcook, and E. M. Kosower, “Introduction to Organic Chemistry,” 4th ed. Macmillian, New York, 1992. 30. A. Wendel, in “Kirk-Othmer Concise Encyclopedia of Chemical Technology,” Vol. 15, 4th ed., p. 192. Wiley, New York, 1995. 31. R. R. C. New, Ed., “Liposomes: A Practical Approach.” Oxford University Press, New York, 1990. 32. S. Hassan, W. Rowe, and G. J. T. Tiddy, in “Handbook of Applied Surface and Colloid Chemistry” (K. Holmberg, Ed.), Vol. I, p. 466. Wiley, Chichester, 2002. 33. H. Ringsdorf, B. Schlarb, and J. Venzmer, Angew. Chem. Int. Ed. 27, 113 (1988). 34. C. Trandum, P. Westh, K. Jørgensen, and O. G. Mouritsen, Biophys. J. 78, 2486 (2000). 35. R. M. R. S. Mesquita, E. Melo, T. E. Thompson, and W. L. C. Vaz, Biophys. J. 78, 3019 (2000). 36. B.-S. Lee, S. A. Mabry, A. Jonas, and J. Jonas, Chem. Phys. Lipids 78, 103 (1995). 37. A. Blume, in “Phospholipids Handbook” (G. Cevc, Ed.), p. 455. Dekker, New York, 1993. 38. The Lipid Data Bank at http://www.ldb.chemistry.ohio-state.edu/ (accessed October 2002). 39. R. Koynova and M. Caffrey, Biochim. Biophys. Acta 1376, 91 (1998). 40. R. Koynova and M. Caffrey, Chem. Phys. Lipids 115, 107 (2002). 41. S. Marbey-Gaud, in “Liposomes: From Physical Structure to Therapeutic Applications” (C. G. Knight, Ed.), p. 105. Elsevier/NorthHolland Biomedical Press, Amsterdam, 1981. 42. R. N. A. H. Lewis and R. N. McElhaney, in “The Structure of Biological Membranes” (P. Yeagle, Ed.), p. 73. CRC Press, Boca Raton, FL, 1992. 43. R. L. Biltonen and D. Lichtenberg, Chem. Phys. Lipids 64, 129 (1993). 44. J. Katsaras, S. Tristam-Nagle, Y. Liu, R. L. Headrick, E. Fontes, P. C. Mason, and J. F. Nagle, Phys. Rev. E. 61, 5668 (2000). 45. A. J. Verkleij and J. de Gier, in “Liposomes: From Physical Structure to Therapeutic Applications” (C. G. Knight, Ed.), p. 83. Elsevier/North-Holland Biomedical Press, Amsterdam, 1981. 46. W.-J. Sun, S. Tristram-Nagle, R. M. Suter, and J. F. Nagle, Proc. Natl. Acad. Sci. USA 93, 7008 (1996).

Preparation of Vesicles (Liposomes) 47. K. Sengupta, V. A. Raghunathan, and J. Katsaras, Europhys. Lett. 49, 722 (2000). 48. W. Curatolo, B. Sears, and L. J. Neuringer, Biochim. Biophys. Acta 817, 261 (1985). 49. D. V. Lynch and P. L. Steponkus, Biochim. Biophys. Acta 984, 267 (1989). 50. G. Cevc, in “Phospholipids Handbook” (G. Cevc, Ed.), p. 939. Dekker, New York, 1993. 51. G. Cevc and D. Marsh, “Phospholipid Bilayers: Physical Principles and Models.” Wiley, New York, 1987. 52. P. Machy and L. Leserman, “Liposomes in Cell Biology and Pharmacology.” John Libbey Eurotext Ltd./INSERM, London, 1987. 53. D. Lichtenberg, E. Freire, C. F. Schmidt, Y. Barenholz, P. L. Felgner, and T. E. Thompson, Biochemistry 20, 3462 (1981). 54. H. Madani and E. W. Kaler, Langmuir 6, 125 (1990). 55. U. Olsson and H. Wennerström, J. Phys. Chem B 106, 5135 (2002). 56. R. Smith and C. Tanford, J. Mol. Biol. 67, 75 (1972). 57. M. Grit and D. J. A. Crommelin, Chem. Phys. Lipids 62, 113 (1992). 58. N. J. Zuidam, H. K. M. E. Gouw, Y. Barenholz, and D. J. A. Crommelin, Biochim. Biophys. Acta 1240, 101 (1995). 59. E. W. Kaler, A. K. Murthy, B. E. Rodriguez, and J. A. Zasadzinski, Science 245, 1371 (1989). 60. S. A. Safran, P. Pincus, and D. Andelman, Science 248, 354 (1990). 61. Y. Barenholz, Prog. Lipid Res. 41, 1 (2002). 62. H. Ohvo-Rekilä, B. Ramstedt, P. Leppimäki, and J. P. Slotte, Prog. Lipid Res. 41, 66 (2002). 63. P. R. Cullis, D. B. Fenske, and M. J. Hope, in “Biochemistry of Lipids, Lipoproteins and Membranes” (D. E. Vance and J. E. Vance, Eds.), p. 1. Elsevier, Amsterdam, 1996. 64. A. D. Bangham, M. W. Hill, and N. G. A. Miller, Methods Membr. Biol. 1, 1 (1974). 65. M. J. Hope, M. B. Bally, L. D. Mayer, A. S. Janoff, and P. R. Cullis, Chem. Phys. Lipids 40, 89 (1986). 66. D. Lichtenberg and Y. Barenholz, Methods Biochem. Anal. 33, 337 (1988). 67. M. C. Woodle and D. Papahdjopoulos, Methods Enzymol. 171, 193 (1989). 68. U. K. Nässander, G. Storm, P. A. M. Peeters, and D. J. A. Crommelin, in “Biodegradable Polymers as Drug Delivery Systems” (M. Chasin and R. Langer, Eds.), p. 261. Dekker, New York, 1990. 69. D. D. Lasic, “Liposomes: from Physics to Applications.” Elsevier, Amsterdam, 1993. 70. G. V. Betagri, S. A. Jenkins, and D. L. Parsons, “Liposome Drug Delivery Systems.” Technomic Publishing, Lancaster, PA, 1993. 71. H. Hauser, in “Phospholipids Handbook” (G. Cevc, Ed.), p. 603. Dekker, New York, 1993. 72. D. J. A. Crommelin and H. Schreier, in “Colloidal Drug Delivery Systems” (J. Kreuter, Ed.), p. 73. Dekker, New York, 1994. 73. Y. Barenholz and D. J. A. Crommelin, in “Encyclopedia of Pharmaceutical Technology” (J. Swarbrick and J. C. Boylan, Eds.), Vol. 9, p. 1. Dekker, New York, 1994. 74. S. Vemuri and C. T. Rhodes, Pharm. Acta Helv. 70, 95 (1995). 75. P. Walde and S. Ichikawa, Biomol. Eng. 18, 143 (2001). 76. S. M. Gruner, R. P. Lenk, A. S. Janoff, and M. J. Ostro, Biochemistry 24, 2833 (1985). 77. W. R. Perkins, S. R. Minchey, M. J. Ostro, T. F. Taraschi, and A. S. Janoff, Biochim. Biophys. Acta 943, 103 (1988). 78. R. W. Corkery and S. T. Hyde, Langmuir 12, 5528 (1996). 79. R. P. Rand and V. A. Parsegian, Biochim. Biophys. Acta 988, 351 (1989). 80. J. F. Nagle and S. Tristram-Nagle, Biochim. Biophys. Acta 1469, 159 (2000). 81. D. Lichtenberg and T. Markello, J. Pharm. Sci. 73, 122 (1984). 82. K. Akashi, H. Miyata, H. Itoh, and K. Kinosita, Jr., Biophys. J. 71, 3242 (1996). 83. E. Evans and D. Needham, J. Phys. Chem. 91, 4219 (1987).

Preparation of Vesicles (Liposomes) 84. K. Akashi, H. Miyata, H. Itoh, and K. Kinosita, Jr., Biophys. J. 74, 2973 (1998). 85. M. F. Brown and J. Seelig, Nature 269, 721 (1977). 86. R. Kataoka, S. Aruga, S. Mitaku, K. Kinosita, Jr., and A. Ikegami, Biophys. Chem. 21, 277 (1985). 87. N. Magome, T. Takemura, and K. Yoshikawa, Chem. Lett. 205 (1997). 88. M. Alfridsson, B. Ninham, and S. Wall, Langmuir 16, 10087 (2000). 89. R.-M. Servuss, Z. Naturforsch. 43c, 938 (1988). 90. J. P. Reeves and R. M. Dowben, J. Cell Physiol. 73, 49 (1969). 91. P. Mueller, T. F. Chien, and B. Rudy, Biophys. J. 44, 375 (1983). 92. D. Needham and E. Evans, Biochemistry 27, 8261 (1988). 93. D. Needham and D. Zhelev in “Giant Vesicles,” Perspectives in Supramolecular Chemistry (P. L. Luisi and P. Walde, Eds.), Vol. 6, p. 103. Wiley, Chichester, 2000. 94. D. Needham and D. V. Zhelev, in “Vesicles,” Surfactant Science Series (M. Rosoff, Ed.), Vol. 62, p. 373. Dekker, New York, 1996. 95. R. Virchow, Arch. Pathol. Anat. Physiol. Klin. Med. 6, 562 (1854). 96. O. Lehmann, “Die neue Welt der flüssigen Kristalle.” Akad. Verlagsgesellschaft, Leipzig, 1911. 97. B. Alberts, A. Johnson, J. Lewis, M. Raff, K. Roberts, and P. Walter, “Molecular Biology of the Cell,” 4th ed. Garland Science, New York, 2002. 98. I. Sakurai and Y. Kawamura, Biochim. Biophys. Aca 777, 347 (1984). 99. I. Sakurai, T. Suzuki, and S. Sakurai, Biochim. Biophys. Acta 985, 101 (1989). 100. D. D. Lasic, J. Kidric, and S. Zagorc, Biochim. Biophys. Acta 896, 117 (1987). 101. D. D. Lasic, J. Colloid Interface Sci. 124, 428 (1988). 102. M. I. Angelova and D. D. Dimitrov, Faraday Discuss. Chem. Soc. 81, 303 (1986). 103. M. I. Angelova, in “Giant Vesicles,” Perspectives in Supramolecular Chemistry (P. L. Luisi and P. Walde, Eds.), Vol. 6, p. 27. Wiley, Chichester, 2000. 104. P. Bucher, A. Fischer, P. L. Luisi, T. Oberholzer, and P. Walde, Langmuir 14, 2712 (1998). 105. R. Wick, M. I. Angelova, P. Walde, and P. L. Luisi, Chem. Biol. 3, 105 (1996). 106. J. M. Holopainen, M. I. Angelova, and P. K. J. Kinnunen, Biophys. J. 78, 830 (2000). 107. M. I. Angelova, S. Soléau, P. Méléard, J.-F. Faucon, and P. Bothorel, Prog. Colloid Polym. Sci. 89, 127 (1992). 108. C. Dietrich, M. I. Angelova, and B. Pouligny, J. Phys. II France 7, 1651 (1997). 109. F. M. Menger and M. I. Angelova, Acc. Chem. Res. 31, 789 (1998). 110. P. Méléard, C. Gerbeaud, T. Pott, and M. D. Mitov in “Giant Vesicles,” Perspectives in Supramolecular Chemistry (P. L. Luisi and P. Walde, Eds.), Vol. 6, p. 185. Wiley, Chichester, 2000. 111. B. Bozic, S. Svetina, B. Zekš, and R. E. Waugh, Biophys. J. 61, 963 (1992). 112. L. Mathivet, S. Cribier, and P. F. Devaux, Biophys. J. 70, 1112 (1996). 113. V. Kralj-Iglic, G. Gomišcek, J. Majhenc, V. Arrigler, and S. Svetina, Colloids Surf. A: Physicochem. Eng. Aspects 181, 315 (2001). 114. L. A. Bagatolli, T. Parasassi, and E. Gratton, Chem. Phys. Lipids 105, 135 (2000). 115. H. Kikuchi, H. Yamauchi, and S. Hirota, Chem. Pharm. Bull. 39, 1522 (1991). 116. C. Huang and J. T. Mason, Proc. Natl. Acad. Sci. USA 75, 308 (1978). 117. R. Lawaczek, M. Kianosho, J. L. Giraudet, and S. I. Chan, Nature 256, 584 (1975). 118. R. Lawaczek, M. Kianosho, and S. I. Chan, Biochim. Biophys. Acta 443, 313 (1976). 119. C. Huang, Biochemistry 8, 344 (1969).

73 120. O. Bakouche, D. Gerlier, J.-M. Letoffe, and P. Claudy, Biophys. J. 50, 1 (1986). 121. M. P. Sheetz and S. I. Chan, Biochemistry 11, 4573 (1972). 122. D. Paphadjopoulos and J. C. Watkins, Biochim. Biophys. Acta 135, 639 (1967). 123. S. M. Johnson, Biochim. Biophys. Acta 307, 27 (1973). 124. Y. Barenholz, D. Gibbes, B. J. Litman, J. Goll, T. E. Thompson, and F. D. Carlson, Biochemistry 16, 2806 (1977). 125. H. Hauser, Biochem. Biophys. Res. Commun. 45, 1049 (1971). 126. S. Mitaku and T. Date, Biochim. Biophys. Acta 688, 411 (1982). 127. R. A. Klein, Biochim. Biophys. Acta 210, 486 (1970). 128. R. Qiu and R. C. MacDonald, Biochim. Biophys. Acta 1191, 343 (1994). 129. W. H. Pitcher III and W. H. Huestis, Biochem. Biophys. Res. Commun. 296, 1352 (2002). 130. M. Wong, F. H. Anthony, T. W. Tillack, and T. E. Thompson, Biochemistry 21, 4126 (1982). 131. S. E. Schullery, C. F. Schmidt, P. Felgner, T. W. Tillack, and T. E. Thompson, Biochemistry 19, 3919 (1980). 132. A. L. Larrabee, Biochemistry 18, 3321 (1979). 133. M. J. Hope, R. Nayar, L. D. Mayer, and P. R. Cullis, in “Liposome Technology” (G. Gregoriadis, Ed.), Vol. 1, 2nd ed., p. 123. CRC Press, Boca Raton, FL, 1993. 134. U. Pick, Arch. Biochem. Biophys. 212, 186 (1981). 135. K. Anzai, M. Yoshida, and Y. Kirino, Biochim. Biophys. Acta 1021, 21 (1990). 136. M. Traïkia, D. E. Warschawski, M. Recouvreur, J. Cartaud, and P. F. Devaux, Eur. Biophys. J. 29, 184 (2000). 137. R. C. MacDonald, F. D. Jones, and R. Qui, Biochim. Biophys Acta 1191, 362 (1994). 138. R. C. MacDonald and R. I. MacDonald, in “Liposome Technology” (G. Gregoriadis, Ed.), Vol. 1, 2nd ed., p. 209. CRC Press, Boca Raton, FL, 1993. 139. N. F. Bunkin, O. A. Kiseleva, A. V. Lobeyev, T. G. Movchan, B. W. Ninham, and O. I. Vinogradova, Langmuir 13, 3024 (1997). 140. N. Oku and R. C. MacDonald, Biochemistry 22, 855 (1983). 141. M. Kasahara and P. C. Hinckle, J. Biol. Chem. 252, 7384 (1977). 142. L. D. Mayer, M. J. Hope, P. R. Cullis, and A. S. Janoff, Biochim. Biophys. Acta 817, 193 (1985). 143. C. J. Chapman, W. L. Erdahl, R. W. Talor, and D. R. Pfeiffer, Chem. Phys. Lipids 55, 73 (1990). 144. C. J. Chapman, W. E. Erdahl, R. W. Taylor, and D. R. Pfeiffer, Chem. Phys. Lipids 60, 201 (1991). 145. P. A. Monnard, N. Berclaz, K. Conde-Frieboes, and T. Oberholzer, Langmuir 15, 7504 (1999). 146. K. Higashi, S. Suzuki, H. Fujii, and Y. Kirino, J. Biochem. 101, 433 (1987). 147. D. W. Deamer and G. L. Barchfeld, J. Mol. Evol. 18, 203 (1982). 148. C. Kirby and G. Gregoriadis, BioTechnology 2, 979 (1984). 149. R. L. Shew and D. W. Deamer, Biochim. Biophys. Acta 816, 1 (1985). 150. T. D. Madden and N. Boman, in “Liposomes: Rational Design” (A. S. Janoff, Ed.), p. 261. Dekker, New York, 1999. 151. P. R. Harrigan, T. D. Madden, and P. R. Cullis, Chem. Phys. Lipids 52, 139 (1990). 152. T. Suzuki, H. Komatsu, and K. Miyajima, Biochim. Biophys. Acta 1278, 176 (1996). 153. E. C. A. van Winden, W. Zhang, and D. J. A. Crommelin, Pharm. Res. 14, 1151 (1997). 154. K. Miyajima, Adv. Drug Delivery Rev. 24, 151 (1997). 155. W. Zhang, A. C. A. van Winden, J. A. Bouwstra, and D. J. A. Crommelin, Cryobiology 35, 277 (1997). 156. E. C. A. van Winden and D. J. A. Crommelin, J. Controlled Release 58, 69 (1999). 157. J. H. Cowe and L. M. Cowe, in “Liposome Technology” (G. Gregoriadis, Ed.), Vol. 1, 2nd ed., p. 229. CRC Press, Boca Raton, FL, 1993.

74 158. J. D. Castile and K. M. G. Taylor, Int. J. Pharm. 188, 87 (1999). 159. K. L. Koster, M. S. Webb, G. Bryant, and D. V. Lynch, Biochim. Biophys. Acta 1193, 143 (1994). 160. B. Zadi and G. Gregoriadis, J. Liposome Res. 10, 73 (2000). 161. M. J. Hope, M. B. Bally, G. Webb, and P. R. Cullis, Biochim. Biophys. Acta 812, 55 (1985). 162. L. D. Mayer, M. J. Hope, and P. R. Cullis, Biochim. Biophys. Acta 858, 161 (1986). 163. M. J. Hope, R. Nayar, L. D. Mayer, and P. R. Cullis, in “Liposome Technology” (G. Gregoriadis, Ed.), Vol. 1, 2nd ed., p. 123. CRC Press, Boca Raton, FL, 1993. 164. A. M. Brendzel and I. F. Miller, Biochim. Biophys. Acta 596, 129 (1980). 165. R. Nayar, M. J. Hope, and P. R. Cullis, Biochim. Biophys. Acta 986, 200 (1989). 166. H, Jousma, H. Talsma, F. Spies, J. G. H. Joosten, H. E. Junginger, and D. J. A. Crommelin, Int. J. Pharm. 35, 263 (1987). 167. V. Dorovska-Taran, R. Wick, and P. Walde, Anal. Biochem. 240, 37 (1996). 168. F. Olson, C. A. Hunt, F. C. Szoka, W. J. Vail, and D. Papahadjopoulos, Biochim. Biophys. Acta 557, 9 (1979). 169. F. Szoka, F. Olson, T. Heath, W. Vail, E. Mayhew, and D. Papahadjopoulos, Biochim. Biophys. Acta 601, 559 (1980). 170. S. G. Clerc and T. E. Thompson, Biophys. J. 67, 475 (1994). 171. D. G. Hunter and B. J. Frisken, Biophys. J. 74, 2996 (1998). 172. B. J. Frisken, C. Asman, and P. J. Patty, Langmuir 16, 928 (2000). 173. N. Berger, A. Sachse, J. Bender, R. Schubert, and M. Brandl, Int. J. Pharm. 223, 55 (2001). 174. R. C. MacDonald, R. I. MacDonald, B. P. M. Menco, K. Takeshita, N. K. Subbarao, and L.-R. Hu, Biochim. Biophys. Acta 1061, 297 (1991). 175. T. Schneider, A. Sachse, G. Rössling, and M. Brandl, Drug Dev. Ind. Pharm. 20, 2787 (1994). 176. Y. Barenholz, S. Amselem, and D. Lichtenberg, FEBS Lett. 99, 210 (1979). 177. R. L. Hamilton, J. Goerke, L. S. S. Guo, M. C. Williams, and R. J. Havel. J. Lipid Res. 21, 981 (1980). 178. E. Mayhew, R. Lazo, W. J. Vail, and A. M. Green, Biochim. Biophys. Acta 775, 169 (1984). 179. H. Talsma, A. Y. Özer, L. van Bloois, and D. J. A. Crommelin, Drug Dev. Ind. Pharm. 15, 197 (1989). 180. G. Gregoriadis, H. da Silva, and A. T. Florence, Int. J. Pharm. 65, 235 (1990). 181. J.-C. Vuillemard, J. Microencapsulation 8, 547 (1991). 182. M. M. Brandl, D. Bachmann, M. Drechsler, and K. H. Bauer, in “Liposome Technology” (G. Gregoriadis, Ed.), Vol. 1, 2nd ed., p. 49. CRC Press, Boca Raton, FL, 1993. 183. R. Barnadas-Rodríguez and M. Sabés, Int. J. Pharm. 213, 175 (2001). 184. F. L. Sorgi and L. Huang, Int. J. Pharm. 144, 131 (1996). 185. D. F. H. Wallach and J. R. Philippot, in “Liposome Technology” (G. Gregoriadis, Ed.), Vol. 1, 2nd ed., p. 141. CRC Press, Boca Raton, FL, 1993. 186. O. Diat and D. Roux, J. Phys. II France 3, 9 (1993). 187. F. Gauffre and D. Roux, Langmuir 15, 3070 (1999). 188. T. Gulik-Krzywicki, J. C. Dedieu, D. Roux, C. Degert, and R. Laversanne, Langmuir 12, 4668 (1996). 189. O. Diat, D. Roux, and F. Nallet, J. Phys. II France 3, 1427 (1993). 190. D.-W. Kim, S.-G. Oh, S.-C. Yi, S.-Y. Bae, and S.-K. Moon, Chem. Mater. 12, 996 (2000). 191. J. I. Escalante, M. Gradzielski, H. Hoffmann, and K. Mortensen, Langmuir 16, 8653 (2000). 192. T. D. Le, U. Olson, and K. Mortensen, Phys. Chem. Chem. Phys. 3, 1310 (2001). 193. F. Gauffre and D. Roux, Langmuir 15, 3070 (1999). 194. A. Bernheim-Grosswasser, S. Ugazio, F. Gauffre, O. Viratelle, P. Mahy, and D. Roux, J. Chem. Phys. 112, 3424 (2000).

Preparation of Vesicles (Liposomes) 195. N. Mignet, A. Brun, C. Degert, B. Delord, D. Roux, C. Hélène, R. Laversanne, and J.-C. François, Nucleic Acids Res. 28, 3134 (2000). 196. P. André, B. W. Ninham, and M. P. Pileni, New J. Chem. 25, 563 (2001). 197. H. Talsma, M. J. van Steenbergen, J. C. H. Borchert, and D. J. A. Crommelin, J. Pharm. Sci. 83, 276 (1994). 198. S. Gould-Fogerite and R. J. Mannino, in “Liposome Technology” (G. Gregoriadis, Ed.), Vol. 1, 2nd ed., p. 67. CRC Press, Boca Raton, FL, 1993. 199. D. Papahadjopoulos, W. J. Vail, K. Jacobson, and G. Poste, Biochim. Biophys. Acta 394, 483 (1975). 200. F. Szoka, Jr. and D. Papahadjopoulos, Proc. Natl. Acad. Sci. USA 75, 4194 (1978). 201. N. Düzgünes, J. Wilschut, K. Hong, R. Fraley, C. Perry, D. S. Friend, T. L. James, and D. Papahadjopoulos, Biochim. Biophys. Acta 732, 289 (1983). 202. C. Pidgeon, A. H. Hunt, and K. Dittrich, Pharm. Res. 3, 23 (1986). 203. C. Pidgeon, S. McNeely, T. Schmidt, and J. E. Johnson, Biochemistry 26, 17 (1987). 204. C. Pidgeon, in “Liposome Technology” (G. Gregoriadis, Ed.), Vol. 1, 2nd ed., p. 99. CRC Press, Boca Raton, FL, 1993. 205. K. Otake, T. Imura, H. Sakai, and M. Abe, Langmuir 17, 3898 (2001). 206. F. Ishii, A. Takamura, and H. Ogata, J. Dispersion Sci. Technol. 9, 1 (1988). 207. F. Ishii, in “Liposome Technology” (G. Gregoriadis, Ed.), Vol. 1, 2nd ed., p. 111. CRC Press, Boca Raton, FL, 1993. 208. S. Kim, R. E. Jacobs, and S. H. White, Biochim. Biophys. Acta 812, 793 (1985). 209. S. Kim and G. M. Martin, Biochim. Biophys. Acta 646, 1 (1981). 210. S. Kim, M. S. Turker, E. Y. Chi, S. Sela, and G. M. Martin, Biochim. Biophys. Acta 782, 339 (1983). 211. S. Mantripragada, Prog. Lipid Res. 41, 392 (2002). 212. M. S. Spector, A. Zasadzinksi, and M. B. Sankaram, Langmuir 12, 4704 (1996). 213. A. Moscho, O. Orwar, D. T. Chiu, B. P. Modi, and R. N. Zare, Proc. Natl. Acad. Sci. USA 93, 11443 (1996). 214. D. T. Chiu, C. F. Wilson, F. Ryttsén, A. Strömberg, C. Farre, A. Karlsson, S. Nordholm, A. Gaggar, B. P. Modi, A. Moscho, R. A. Garza-López, O. Orwar, and R. N. Zare, Science 283, 1892 (1999). 215. S. Batzri and E. D. Korn, Biochim. Biophys. Acta 298, 1015 (1973). 216. J. M. H. Kremer, M. W. J. v. d. Esker, C. Pathmamanoharan, and P. H. Wiersema, Biochemistry 16, 3932 (1977). 217. A. S. Domazou and P. L. Luisi, J. Liposome Res. 12, 205 (2002). 218. M. Pons, M. Foradada, and J. Estelrich, Int. J. Pharm. 95, 51 (1993). 219. A. Wagner, K. Vorauer-Uhl, G. Kreismayr, and H. Katinger, J. Liposome Res. 12, 271 (2002). 220. A. Wagner, K. Vorauer-Uhl, G. Kreismayr, and H. Katinger, J. Liposome Res. 12, 259 (2002). 221. R. W. Kriftner, in “Liposome Dermatics” (O. Braun-Falco, H. C. Korting, and H. I. Maibach, Eds.), p. 203. Springer-Verlag, Berlin, 1992. 222. R. Naeff, Adv. Drug Delivery Rev. 18, 343 (1996). 223. Y. Maitani, H. Soeda, W. Junping, and K. Takayama, J. Liposome Res. 11, 115 (2001). 224. W. P. Williams, S. Perrett, M. Golding, J.-P. Arnaud, and M. Michez, in “Phospholipids: Characterization, Metabolism, and Novel Biological Application” (G. Cevc and F. Paltauf, Eds.), p. 181. AOCS Press, Champaign, IL, 1995. 225. S. Perrett, M. Golding, and W. P. Williams, J. Pharm. Pharmacol. 43, 154 (1991). 226. P. Williams, SÖFW J. 118, 377 (1992). 227. U. Isele, P. van Hoogevest, R. Hilfiker, H.-G. Capraro, K. Schieweck, and H. Leuenberger, J. Pharm. Sci. 83, 1608 (1994).

Preparation of Vesicles (Liposomes) 228. N. I. Payne, R. Timmins, V. Ambrose, M. D. Ward, and F. Ridgway, J. Pharm. Sci. 75, 325 (1986). 229. E. Touitou, N. Dayan, L. Bergelson, B. Godin, and M. Eliaz, J. Controlled Rel. 65, 403 (2000). 230. N. Dayan and E. Touitou, Biomaterials 21, 1879 (2000). 231. B. W. Barry, Eur. J. Pharm. Sci. 14, 101 (2001). 232. P. L. Ahl, L. Chen, W. R. Perkins, S. R. Minchey, L. T. Boni, T. F. Taraschi, and A. S. Janoff, Biochim. Biophys. Acta 1195, 237 (1994). 233. F. Ishii, A. Takamura, and Y. Ishigami, Langmuir 11, 483 (1995). 234. F. M. Menger and B. M. Sykes, Langmuir 14, 4131 (1998). 235. D. F. Evans and H. Wennerstöm, “The Colloidal Domain: Where Physics, Chemistry, Biology and Technology Meet,” 2nd ed. WileyVCH, New York, 1999. 236. L. Frederiksen, K. Anton, P. van Hoogevest, H. R. Keller, and H. Leuenberger, J. Pharm. Sci. 86, 921 (1997). 237. D. W. Deamer, Ann. N.Y. Acad. Sci. 308, 250 (1978). 238. J. T. Buboltz and G. W. Feigenson, Biochim. Biophys. Acta 1417, 232 (1999). 239. L. Zhang, J. Hu, and Z. Lu, J. Colloid Interface Sci. 190, 76 (1997). 240. A. Helenius and K. Simons, Biochim. Biophys. Acta 415, 29 (1975). 241. D. Lichtenberg, R. J. Robson, and E. A. Dennis, Biochim. Biophys. Acta 737, 285 (1983). 242. D. Lichtenberg, Biochim. Biophys. Acta 821, 470 (1985). 243. J. Lasch, Biochim. Biophys. Acta 1241, 269 (1995). 244. M. H. W. Milsmann, R. A. Schwendener, and H.-G. Weder, Biochim. Biophys. Acta 512, 147 (1978). 245. O. Zumbühl and H.-G. Weder, Biochim. Biophys. Acta 640, 252 (1981). 246. J. Philippot, S. Mustafschiev, and J. P. Liautard, Biochim. Biophys. Acta 734, 137 (1983). 247. M. Ueno, C. Tanford, and J. A. Reynolds, Biochemistry 23, 3070 (1984). 248. R. A. Parente and B. R. Lentz, Biochemistry 23, 2353 (1984). 249. T. M. Allen, A. Y. Romans, H. Kercret, and J. P. Segrest, Biochim. Biophys. Acta 601, 328 (1980). 250. M. Rotenberg and D. Lichtenberg, J. Colloid Interface Sci. 144, 591 (1991). 251. J. S. Tauskela, M. Akler, and M. Thompson, Anal. Biochem. 201, 282 (1992). 252. M. Wacker and R. Schubert, Int. J. Pharm. 162, 171 (1998). 253. N. Funasaki, S. Hada, and S. Neya, J. Phys. Chem. B 103, 169 (1999). 254. N. Funasaki, M. Nomura, S. Ishikawa, and S. Neya, J. Phys. Chem. B 104, 7745 (2000). 255. Y. Li, J. F. Holzwarth, and C. Bohne, Langmuir 16, 2038 (2000). 256. J. Leng, S. U. Egelhaaf, and M. E. Cates, Europhys. Lett. 59, 311 (2002). 257. J. Brunner, P. Skrabal, and H. Hauser, Biochim. Biophys. Acta 455, 322 (1976). 258. J.-L. Rigaud, D. Levy, G. Mosser, and O. Lambert, Eur. Biophys. J. 27, 305 (1998). 259. J. R. Philippot and J. P. Liautard, in “Liposome Technology” (G. Gregoriadis, Ed.), Vol. 1, 2nd ed., p. 81. CRC Press, Boca Raton, 1993. 260. P. Schurtenberger, N. Mazer, S. Waldvogel, and W. Känzig, Biochim. Biophys. Acta 775, 111 (1984). 261. H. G. Enoch and P. Strittmatter, Proc. Natl. Acad. Sci. USA 76, 145 (1979). 262. R. Peschka, T. Purmann, and R. Schubert, Int. J. Pharm. 162, 177 (1998). 263. J.-L. Rigaud, B. Pitard, and D. Levy, Biochim. Biophys. Acta 1231, 223 (1995). 264. D. Lichtenberg, E. Opatowski, and M. M. Kozlov, Biochim. Biophys. Acta 1508, 1 (2000). 265. M. Almgren, Biochim. Biophys. Acta 1508, 146 (2000).

75 266. M. Ollivon, S. Lesieur, C. Grabielle-Madelmont, and M. Paternostre, Biochim. Biophys. Acta 1508, 34 (2000). 267. S. Almog, T. Kushnir, S. Nir, and D. Lichtenberg, Biochemistry 25, 2597 (1986). 268. M. Paternostre, O. Meyer, C. Grabielle-Madelmont, S. Lesieur, M. Ghanam, and M. Ollivon, Biophys. J. 69, 2476 (1995). 269. C. Sun and M. Ueno, Colloid Polym. Sci. 278, 855 (2000). 270. M. N. Jones and D. Chapman, “Micelles, Monolayers, and Biomembranes.” Wiley-Liss, New York, 1995. 271. P. Walde, E. Blöchliger, and K. Morigaki, Langmuir 15, 2346 (1999). 272. H. Hauser, Biochim. Biophys. Acta 1508, 164 (2000). 273. N. E. Gabriel and M. F. Roberts, Biochemistry 23, 4011 (1984). 274. N. E. Gabriel and M. F. Roberts, Biochemistry 25, 2812 (1986). 275. G. Cevc, A. Schätzlein, and G. Blume, J. Controlled Release 36, 3 (1995). 276. G. Cevc, D. Gebauer, J. Stieber, A. Schätzlein, and G. Blume, Biochim. Biophys. Acta 1368, 201 (1998). 277. S. E. Friberg, H. Yang, L. Fei, S. Sadasivan, D. H. Rasmussen, and P. A. Aikens, J. Dispersion Sci. Technol. 19, 19 (1998). 278. N. Oku and R. C. MacDonald, Biochim. Biophys. Acta 734, 54 (1983). 279. N. Oku and R. C. MacDonald, J. Biol. Chem. 258, 8733 (1983). 280. S. Matsumoto, M. Kohda, and S.-I. Murata, J. Colloid Interface Sci. 62, 149 (1977). 281. H. Kunieda, K. Nakamura, and D. F. Evans, J. Am. Chem. Soc. 113, 1051 (1991). 282. H. Kunieda, K. Nakamura, H. T. Davis, and D. F. Evans, Langmuir 7, 1915 (1991). 283. H. Kunieda, N. Kanai, A. Uemoto, and I. Tobita, Langmuir 10, 4006 (1994). 284. U. Olsson, K. Nakamura, H. Kunieda, and R. Strey, Langmuir 12, 3045 (1996). 285. H. Kunieda, K. Shigeta, and M. Suzuki, Langmuir 15, 3118 (1999). 286. H. Mays, M. Almgren, A. Dedinaite, and P. M. Claesson, Langmuir 15, 8072 (1999). 287. Y. Barenholz, in “Medical Applications of Liposomes” (D. D. Lasic and D. Papahadjopoulos, Eds.), p. 545. Elsevier, Amsterdam, 1998. 288. J. R. Bellare, H. T. Davis, L. E. Scriven, and Y. Talmon, J. Electron Microsc. Tech. 10, 87 (1988). 289. B. Klösgen and W. Helfrich, Eur. Biophys. J. 22, 329 (1993). 290. Y. Talmon, Ber. Bunsenges. Phys. Chem. 100, 364 (1996). 291. O. Maier, V. Oberle, and D. Hoeckstra, Chem. Phys. Lipids 116, 3 (2002). 292. P. F. Devaux, P. Fellmann, and P. Hervé, Chem. Phys. Lipids 116, 115 (2002). 293. M. Fröhlich, V. Brecht, and R. Peschka-Süss, Chem. Phys. Lipids 109, 103 (2001). 294. D. D. Lasic, in “Handbook of Biological Physics” (R. Lipowsky and E. Sackmann, Eds.), Vol. 1, p. 491. Elsevier, Amsterdam, 1995. 295. D. D. Lasic and Y. Barenholz, Eds., “Handbook of Nonmedical Applications of Liposomes,” Vols. 1–4. CRC Press, Boca Raton, FL, 1996. 296. D. D. Lasic and D. Papahadjopoulos, Eds., “Medical Applications of Liposomes.” Elsevier, Amsterdam, 1998. 297. A. S. Janoff, Ed., “Liposomes: Rational Design.” Dekker, New York, 1999. 298. J. H. Fendler, “Membrane Mimetic Chemistry.” Wiley, New York, 1982. 299. J. H. Fendler, Chem. Rev. 87, 877 (1987). 300. R. B. Gennis, “Biomembranes: Molecular Structure and Function.” Springer-Verlag, New York, 1989. 301. I. Ash, “Handbook of Industrial Surfactants.” Synapse Information Resources, New York, 2000. 302. Y. Okahata, R. Ando, and T. Kunitake, Ber. Bunsenges. Phys. Chem. 85, 789 (1981).

76 303. M. J. Blandamer, B. Briggs, P. M. Cullis, S. D. Kirby, and J. B. F. N. Engberts, J. Chem. Soc. Faraday Trans. 93, 453 (1997). 304. T. Kunitake and Y. Okahata, J. Am. Chem. Soc. 99, 3860 (1977). 305. B. W. Ninham, D. F. Evans, and G. J. Wei, J. Phys. Chem. 87, 5020 (1983). 306. A. M. Carmona-Ribeiro and B. R. Midmore, J. Phys. Chem. 96, 3542 (1992). 307. R. G. Laughlin, R. L. Munyon, J. L. Burns, T. W. Coffindaffer, and Y. Talmon, J. Phys. Chem. 96, 374 (1992). 308. E. Feitosa, P. C. A. Barreleiro, and G. Olofsson, Chem. Phys. Lipids 105, 201 (2000). 309. E. Feitosa and W. Brown, Langmuir 13, 4810 (1997). 310. D. B. Nascimento, R. Rapuano, M. M. Lessa, and A. M. CarmonaRibeiro, Langmuir 14, 7387 (1998). 311. H. Kawasaki, M. Souda, S. Tanaka, N. Nemoto, G. Karlsson, M. Almgren, and H. Maeda, J. Phys. Chem. B 106, 1524 (2002). 312. H. Hauser, H. H. Mantsch, and H. L. Casal, Biochemistry 29, 2321 (1990). 313. N. K. Subbarao, R. I. MacDonald, K. Takeshita, and R. C. MacDonald, Biochim. Biophys. Acta 1063, 147 (1991). 314. T. A. A. Fonteijn, D. Hoeckstra, and J. B. F. N. Engberts, Langmuir 8, 2437 (1992). 315. M. J. Blandamer, B. Briggs, P. M. Cullis, J. B. F. N. Engberts, A. Wagenaar, E. Smits, D. Hoeckstra, and A. Kacperska, Langmuir 10, 3507 (1994). 316. E. Smits, M. J. Blandamer, B. Briggs, P. M. Cullis, and J. B. F. N. Engberts, Recl. Trav. Chim. Pays-Bas 115, 37 (1996). 317. L. A. M. Rupert, J. F. L. van Breemen, D. Hoeckstra, and J. B. F. N. Engberts, J. Phys. Chem. 92, 4416 (1988). 318. Y.-M. Tricot, D. N. Furlong, and W. H. F. Sasse, Aust. J. Chem. 37, 1147 (1984). 319. A. M. Carmona-Ribeiro and S. Hix, J. Phys. Chem. 95, 1812 (1991). 320. A. M. Carmona-Ribeiro, C. E. Castuma, A. Sesso, and S. Schreier, J. Phys. Chem. 95, 5361 (1991). 321. L. Hammarström, I. Velikian, G. Karlsson, and K. Edwards, Langmuir 11, 408 (1995). 322. R. Humphry-Baker, D. H. Thompson, Y. Lei, M. J. Hope, and J. K. Hurst, Langmuir 7, 2592 (1991). 323. N. Plobeck, S. Eifler, A. Brisson, Y. Nakatani, and G. Ourisson, Tetrahedron Lett. 33, 5249 (1992). 324. V. Birault, G. Pozzi, N. Plobeck, S. Eifler, M. Schmutz, T. Palanché, J. Raya, A. Brisson, Y. Nakatani, and G. Ourisson, Chem. Eur. J. 2, 789 (1996). 325. G. Pozzi, V. Birault, B. Werner, O. Dannenmuller, Y. Nakatani, G. Ourisson, and S. Terekawa, Angew. Chem. Int. Ed. Engl. 35, 177 (1996). 326. P. Walde, M. Wessicken, U. Rädler, N. Berclaz, K. CondeFrieboes, and P. L. Luisi, J. Phys. Chem. B 101, 7390 (1997). 327. B. J. Ravoo and J. B. F. N. Engberts, Langmuir 10, 1735 (1994). 328. U. Brinkmann, E. Neumann, and B. H. Robinson, J. Chem. Soc., Faraday Trans. 94, 1281 (1998). 329. W. R. Hargreaves and D. W. Deamer, Biochemistry 17, 3759 (1978). 330. D. P. Cistola, J. A. Hamilton, D. Jackson, and D. M. Small, Biochemistry 27, 1881 (1988). 331. K. Morigaki, P. Walde, M. Misran, and B. H. Robinson, Colloids Surf. A: Physicochem. Eng. Aspects, 213, 37 (2003). 332. J. M. Gebicki and M. Hicks, Nature 243, 232 (1973). 333. T. Hains, Proc. Natl. Acad. Sci. USA 80, 160 (1983). 334. H. Fukuda, A. Goto, H. Yoshioka, R. Goto, K. Morigaki, and P. Walde, Langmuir 17, 4223 (2001). 335. M. Bergström and J. C. Eriksson, Langmuir 12, 624 (1996). 336. C. L. Apel, D. W. Deamer, and M. N. Mautner, Biochim. Biophys. Acta 1559, 1 (2002). 337. M. Gradzielski, M. Bergmeier, M. Müller, and H. Hoffmann, J. Phys. Chem. B 101, 1719 (1997). 338. L. L. Brasher and E. W. Kaler, Langmuir 12, 6270 (1996).

Preparation of Vesicles (Liposomes) 339. M. T. Yatcilla, K. L. Herrington, L. L. Brasher, E. W. Kaler, S. Chiruvolu, and J. A. Zasadzinski, J. Phys. Chem. 100, 5874 (1996). 340. C. Caillet, M. Hebrant, and C. Tondre, Langmuir 16, 9099 (2000). 341. A. Shioi and T. A. Hatton, Langmuir 18, 7341 (2002). 342. Y. Xia, I. Goldmints, P. W. Johnson, T. A. Hatton, and A. Bose, Langmuir 18, 3822 (2002). 343. E. F. Marques, O. Regev, A. Khan, M. G. Miguel, and B. Lindman, J. Phys. Chem. B 102, 6746 (1998). 344. E. F. Marques, O. Regev, A. Khan, M. G. Miguel, and B. Lindman, J. Phys. Chem. B 103, 8353 (1999). 345. L. Cantù, M. Corti, E. Del Favero, and A. Raudino, J. Phys. II France 4, 1585 (1994). 346. L. Cantù, M. Corti, E. Del Favero, M. Dubois, and Th. N. Zemb, J. Phys. Chem. B 102, 5737 (1998). 347. S. Bonaccio, P. Walde, and P. L. Luisi, J. Phys. Chem. 98, 6661 (1994). 348. S. Bonaccio, D. Capitani, A. L. Segre, P. Walde, and P. L. Luisi, Langmuir 13, 1952 (1997). 349. J. K. Harris, G. D. Rose, and M. L. Bruening, Langmuir 18, 5337 (2002). 350. K. Schillén, K. Bryskhe, and Y. S. Mel’nikova, Macromolecules 32, 6885 (1999). 351. C. Nardin, T. Hirt, J. Leukel, and W. Meier, Langmuir 16, 1035 (2000). 352. S. L. Regen, in “Liposomes: From Biophysics to Therapeutics” (M. J. Ostro, Ed.), p. 73. Dekker, New York, 1987. 353. B. A. Armitage, D. E. Bennett, H. G. Lamparski, and D. F. O’Brien, Adv. Polym. Sci. 126, 53 (1996). 354. J. R. Riess, F. Frézard, J. Greiner, M. P. Krafft, C. Santaella, P. Vierling, and L. Zarif, in “Handbook of Nonmedical Applications of Liposomes” (D. D. Lasic and Y. Barenholz, Eds.), Vol. 3, p. 97. CRC Press, Boca Raton, FL, 1996. 355. M. P. Krafft, F. Giulieri, and J. G. Riess, Angew. Chem. Int. Ed. Engl. 32, 741 (1993). 356. M. P. Krafft, L. Schieldknecht, P. Marie, F. Giulieri, M. Schmutz, N. Poulain, and E. Nakache, Langmuir 17, 2872 (2001). 357. P. Lo Nostro and G. Gabrielli, Colloids Surf. 44, 119 (1990). 358. J.-H. Fuhrhop and J. Köning, “Membrane and Molecular Assemblies: The Synkinetic Approach.” Royal Society of Chemistry, Cambridge, 1994. 359. G. B. Patel, B. J. Agnew, L. Deschatelets, L. P. Fleming, and G. D. Sprott, Int. J. Pharm. 194, 39 (2000). 360. A. Gliozzi, A. Relini, and P. L.-G. Chong, J. Membr. Sci. 206, 131 (2002). 361. N. A. J. M. Sommerdijk, T. H. L. Hoeks, M. Synak, M. C. Feiters, R. J. M. Nolte, and B. Zwanenburg, J. Am. Chem. Soc. 119, 4338 (1997). 362. M. Bergsma, M. L. Fielden, and J. B. F. N. Engberts, J. Colloid Interface Sci. 243, 491 (2001). 363. H. Hoffmann, C. Thunig, and D. Miller, Colloids Surf. A: Physicochem. Eng. Aspects 210, 147 (2002). 364. Y. Tanaka, M. Miyachi, and Y. Kobuke, Angew. Chem. Int. Ed. 38, 504 (1999). 365. P. Gosh, S. Sengupta, and P. K. Bharadwaj, Langmuir 14, 5712 (1998). 366. M. Sano, K. Oishi, T. Ishii, and S. Shinkai, Langmuir 15, 3773 (2000). 367. K. Katagiri, K. Ariga, and J. Kikuchi, Chem. Lett. 662 (1999). 368. Y. Sumida, A. Masuyama, H. Maekawa, M. Takasu, T. Kida, Y. Nakatsuji, I. Ikeda, and M. Nojima, Chem. Commun. 2385 (1998). 369. T. Kunitake, Angew. Chem. Int. Ed. 31, 709 (1992). 370. S. Chiruvolu, S. Walker, J. Israelachvili, F.-J. Schmitt, D. Leckband, and J. A. Zasadzinski, Science 264, 1753 (1994). 371. J. A. Zasadzinski, E. Kisak, and C. Evans, Curr. Opin. Colloid Interface Sci. 6, 85 (2001).

Preparation of Vesicles (Liposomes) 372. C. M. Paleos, Z. Sideratou, and D. Tsiourvas, ChemBioChem. 2, 305 (2001). 373. T. Sato and J. Sunamoto, Prog. Lipid Res. 31, 345 (1992). 374. A. Sharma and U. S. Sharma, Int. J. Pharm. 154, 123 (1997). 375. D. D. Lasic, Trends Biotechnol. 16, 307 (1998). 376. A. Bloom, in “Liposomes: Rational Design” (A. S. Janoff, Ed.), p. 397. Dekker, New York, 1999. 377. Y. Barenholz, Curr. Opin. Colloid Interface Sci. 6, 66 (2001). 378. R. Banerjee, J. Biomater. Appl. 16, 3 (2001). 379. K. Egbaria and N. Weiner, Cosmetics Toiletries 106, 79 (1991). 380. H. Schreier and J. Bouwstra, J. Controlled Release 30, 1 (1994). 381. T. Lian and R. J. Y. Ho, J. Pharm. Sci. 90, 667 (2001). 382. A. K. Singh and R. G. Carbonell, in “Handbook of Nonmedical Applications of Liposomes” (D. D. Lasic and Y. Barenholz, Eds.), Vol. 4, p. 209. CRC Press, Boca Raton, FL, 1996. 383. H. A. H. Rongen, A. Bult, and W. P. van Bennekom, J. Immunol. Methods 204, 105 (1997). 384. T. Wink, S. J. van Zuilen, A. Bult, and W. P. van Bennekom, Anal. Chem. 70, 827 (1998). 385. J.-A. A. Ho and R. D. Wauchope, Anal. Chem. 74, 1493 (2002). 386. P. Lundahl, C.-M. Zeng, C. Lagerquist Hägglund, I. Gottschalk, and E. Greijer, J. Chromatogr. B 720, 103 (1999). 387. M. Yoshimoto, R. Kuboi, Q. Yang, and J. Miyake, J. Chromatogr. B 712, 59 (1998). 388. O. Braun-Falco, H. C. Korting, and H. I. Maibach, Eds., “Liposome Dermatics.” Springer-Verlag, Berlin, 1992. 389. R. M. Handjani-Vila, A. Ribier, and G. Vanlerberghe, in “Liposome Technology” (G. Gregoriadis, Ed.), Vol. 2, 2nd ed., p. 201. CRC Press, Boca Raton, FL, 1993. 390. G. Vanlerberghe, in “Handbook of Nonmedical Applications of Liposomes” (D. D. Lasic and Y. Barenholz, Eds.), Vol. 4, p. 77. CRC Press, Boca Raton, FL, 1996. 391. T. E. Redelmeier and N. Kitson, in “Liposomes: Rational Design” (A. S. Janoff, Ed.), p. 283. Dekker, New York, 1999. 392. E. Sackmann in “Handbook of Biological Physics” (R. Lipowsky and E. Sackmann, Eds.), Vol. 1, p. 213. Elsevier, Amsterdam, 1995. 393. M. J. Pregel, L. Jullien, and J.-M. Lehn, Angew. Chem. Int. Ed. 31, 1637 (1992). 394. M. R. Ghadiri, J. R. Granja, and L. K. Buehler, Nature 369, 301 (1994). 395. W. L. Erdahl, C. J. Chapman, R. W. Taylor, and D. R. Pfeiffer, Biophys. J. 66, 1678 (1994). 396. T. D. Clark, L. K. Buehler, and M. R. Ghadiri, J. Am. Chem. Soc. 120, 651 (1998). 397. M. Fritz, P. Walde, and D. Seebach, Macromolecules 32, 574 (1999). 398. P. H. Schlesinger, R. Ferdani, J. Liu, J. Pajewska, R. Pajewski, M. Saito, H. Shabany, and G. W. Gokel, J. Am. Chem. Soc. 124, 1848 (2002). 399. V. Sidorov, F. W. Kotch, G. Abdrakhmanova, R. Mizani, J. C. Fettinger, and J. T. Davis, J. Am. Chem. Soc. 124, 2267 (2002). 400. P. L. Luisi, P. Walde, and T. Oberholzer, Curr. Opin. Colloid Interface Sci. 4, 33 (1999). 401. J. W. Szostak, D. P. Bartel, and P. L. Luisi, Nature 409, 387 (2001). 402. A. Pohorille and D. Deamer, Trends Biotechnol. 20, 123 (2002). 403. G. A. Reineccius, in “Encapsulation and Controlled Release of Food Ingredients,” ACS Symposium Series 590 (S. J. Risch and G. A. Reineccius, Eds.), p. 113. Am. Chem. Soc., Washington, DC, 1995. 404. P. Dufour, E. Laloy, J.-C. Vuillemard, and R. E. Simard, in “Handbook of Nonmedical Applications of Liposomes” (D. D. Lasic and Y. Barenholz, Eds.), Vol. 4, p. 153. CRC Press, Boca Raton, FL, 1996. 405. B. C. Keller, Trends Food Sci. Technol. 12, 25 (2001). 406. P. J. Quinn and S. F. Perrett, in “Handbook of Nonmedical Applications of Liposomes” (D. D. Lasic and Y. Barenholz, Eds.), Vol. 4, p. 127. CRC Press, Boca Raton, FL, 1996.

77 407. G. Mossa, M. C. Annesini, and L. Di Marzio, in “Handbook of Nonmedical Applications of Liposomes” (D. D. Lasic and Y. Barenholz, Eds.), Vol. 1, p. 273. CRC Press, Boca Raton, FL, 1996. 408. F. Cioci and R. Lavecchia, in “Handbook of Nonmedical Applications of Liposomes” (D. D. Lasic and Y. Barenholz, Eds.), Vol. 1, p. 287. CRC Press, Boca Raton, FL, 1996. 409. M. Blocher, P. Walde, and I. J. Dunn, Biotechnol. Bioeng. 62, 36 (1999). 410. M. Treyer, P. Walde, and T. Oberholzer, Langmuir 18, 1043 (2002). 411. M. Yoshimoto, S. Wang, K. Fukunaga, P. Walde, R. Kuboi, and K. Nakao, Biotechnol. Bioeng., 81, 695 (2003). 412. B. A. Korgel and H. G. Monbouquette, J. Phys. Chem. B 100, 346 (1996). 413. M. T. Kennedy, B. A. Korgel, H. G. Monbouquette, and J. A. Zasadzinski, Chem. Mater. 10, 2116 (1998). 414. B. A. Korgel and H. G. Monbouquette, Langmuir 16, 3588 (2000). 415. T. Kunitake and S. Shinkai, Adv. Phys. Org. Chem. 17, 435 (1980). 416. I. M. Cuccovia and H. Chaimovich, in “Handbook of Nonmedical Applications of Liposomes” (D. D. Lasic and Y. Barenholz, Eds.), Vol. 3, p. 239. CRC Press, Boca Raton, FL, 1996. 417. P. Scrimin, P. Tecilla, and U. Tonellato, J. Am. Chem. Soc. 114, 5086 (1992). 418. P. Scrimin, in “Supramolecular Control of Structure and Reactivity,” Perspectives in Supramolecular Chemistry (A. D. Hamilton, Ed.), Vol. 3, p. 101. Wiley, Chichester, 1996. 419. J. Pérez-Juste, F. Hollfelder, A. J. Kirby, and J. B. F. N. Engberts, Org. Lett. 2, 127 (2000). 420. T. Rispens and J. B. F. N. Engberts, Org. Lett. 3, 941 (2001). 421. I. Shin, T. Silman, C. Bon, and L. Weiner, Biochemistry 37, 4310 (1998). 422. R. Kuboi, M. Yoshimoto, P. Walde, and P. L. Luisi, Biotechnol. Prog. 13, 828 (1997). 423. S. Okada, S. Peng, W. Spevak, and D. Charych, Acc. Chem. Res. 31, 229 (1998). 424. R. Jelinek and S. Kolusheva, Biotechnol. Adv. 19, 109 (2001). 425. J. Song, Q. Cheng, S. Zhu, and R. C. Stevens, Biomed. Microdev. 4, 213 (2002). 426. J. T. Groves, N. Ulman, P. S. Cremer, and S. G. Boxer, Langmuir 14, 3347 (1998). 427. H. Mueller, H.-J. Butt, and E. Bamberg, J. Phys. Chem. B 104, 4552 (2000). 428. C. A. Keller, K. Glasmästar, V. P. Zhdanov, and B. Kasemo, Phys. Rev. Lett. 84, 5443 (2000). 429. E.-K. Sinner and W. Knoll, Curr. Opin. Chem. Biol. 5, 705 (2001). 430. I. Stanish and H. G. Monbouquette, J. Membr. Sci. 192, 99 (2001). 431. S. A. Miller, J. H. Ding, and D. L. Gin, Curr. Opin. Colloid Interface Sci. 4, 338 (1999). 432. S. Liu and D. F. O’Brian, J. Am. Chem. Soc. 124, 6037 (2002). 433. A. Mueller and D. F. O’Brian, Chem. Rev. 102, 727 (2002). 434. S. Mann, J. P. Hannington, and R. J. P. Williams, Nature 324, 565 (1991). 435. E. D. Eanes, in “Handbook of Nonmedical Applications of Liposomes” (D. D. Lasic and Y. Barenholz, Eds.), Vol. 2, p. 349. CRC Press, Boca Raton, 1996. 436. J. H. Collier and P. B. Messersmith, Annu. Rev. Mater. Res. 31, 237 (2001). 437. S. Mann and G. A. Ozin, Nature 382, 313 (1996). 438. G. J. de, A. A. Soler-Illia, C. Sanchez, B. Lebeau, and J. Patarin, Chem. Rev. 102, 4093 (2002). 439. J. D. Morgan, C. A. Johnson, and E. W. Kaler, Langmuir 13, 6447 (1997). 440. J. Hotz and W. Meier, Langmuir 14, 1031 (1998). 441. F. Caruso, Chem. Eur. J. 6, 413 (2000). 442. M. Blocher, D. J. Liu, P. Walde, and P. L. Luisi, Macromolecules 32, 7332 (1999).

78 443. M. Blocher, D. J. Liu, and P. L. Luisi, Macromolecules 33, 5787 (2000). 444. M. Jung, B. H. Robinson, D. C. Steytler, A. L. German, and R. K. Heenan, Langmuir 18, 2873 (2002). 445. J. H. Fuhrhop and J. Mathieu, Angew. Chem. Int. Ed. 23, 100 (1984). 446. G. Steinberg-Yfrach, J.-L. Rigaud, E. N. Durantini, A. L. Moore, D. Gust, and T. A. Moore, Nature 392, 479 (1998). 447. D. K. Lee and Y. S. Kang, Mol. Cryst. Liq. Cryst. 377, 261 (2002). 448. K. Fukuda, Y. Sasaki, K. Ariga, and J. Kikuchi, J. Mol. Catal. B: Enzym. 11, 971 (2001). 449. J. Kikuchi, K. Ariga, Y. Sasaki, and K. Ikeda, J. Mol. Catal. B: Enzym. 11, 977 (2001). 450. T. S. Arrhenius, M. Blanchard-Desce, M. Dvolaitzky, J. M. Lehn, and J. Malthete, Proc. Natl. Acad. Sci. USA 83, 5355 (1986). 451. N. Berclaz, M. Müller, P. Walde, and P. L. Luisi, J. Phys. Chem. B 105, 1056 (2001). 452. N. Berclaz, E. Blöchliger, M. Müller, and P. L. Luisi, J. Phys. Chem. B 105, 1065 (2001). 453. S. Engström, L. Lindahl, R. Wallin, and J. Engblom, Int. J. Pharm. 86, 137 (1992). 454. K. Larsson, J. Phys. Chem. 93, 7304 (1989). 455. J. C. Shah, Y. Sadhale, and D. M. Chilukuri, Adv. Drug. Del. Rev. 47, 229 (2001). 456. M. Nakano, A. Sugita, H. Matsuoka, and T. Handa, Langmuir 17, 3917 (2001). 457. M. Hodenius, M. De Cuyper, L. Desender, D. Müller-Schulte, A. Steigel, and H. Lueken, Chem. Phys. Lipids 120, 75 (2002). 458. J. M. Lehn and P. Ball, in “The New Chemistry” (N. Hall, Ed.), p. 300. Cambridge University Press, Cambridge, 2000. 459. J.-M. Lehn, Proc. Natl. Acad. Sci. USA 99, 4763 (2002). 460. P. Durbut, in “Handbook of Detergents: Part A: Properties” (G. Broze, Ed.), p. 47. Dekker, New York, 1999. 461. A. Pinazo, M. R. Infante, S. Y. Park, and E. I. Franses, Colloids Surf. B: Biointerfaces 8, 1 (1996). 462. S. Lee, D. H. Kim, and D. Needham, Langmuir 17, 5544 (2001). 463. H. T. Jung, B. Coldren, J. A. Zasadzinski, D. J. Iampietro, and E. W. Kaler, Proc. Natl. Acad. Sci. USA 98, 1353 (2001). 464. D. D. Lasic, J. Liposome Res. 9, 43 (1999). 465. D. D. Lasic, R. Joannic, B. C. Keller, P. M. Frederik, and L. Auvray, Adv. Colloid Interface Sci. 89–90, 337 (2001). 466. S. A. Safran, P. A. Pincus, D. Andelman, and F. C. MacKintosh, Phys. Rev. A 43, 1071 (1991). 467. K. Horbaschek, H. Hoffmann, and J. Hao, J. Phys. Chem. B 104, 2781 (2000). 468. N. L. Gershfeld, W. F. Stevens, Jr., and R. J. Nossal, Faraday Discuss. Chem. Soc. 81, 19 (1986). 469. N. L. Gershfeld, Biochim. Biophys. Acta 988, 335 (1989). 470. N. L. Gershfeld and L. Ginsberg, J. Membr. Biol. 156, 279 (1997). 471. M. Koshinuma, K. Tajima, A. Nakamura, and N. L. Gershfeld, Langmuir 15, 3430 (1999). 472. K. Tajima, M. Koshinuma, A. Nakamura, and N. L. Gershfeld, Langmuir 16, 2576 (2000). 473. D. D. Lasic, Biochem. J. 256, 1 (1988). 474. T. E. Thompson, Hepatology 12, 51S (1990). 475. D. D. Lasic and F. J. Martin, J. Membr. Sci. 50, 215 (1990). 476. W. Helfrich, Prog. Colloid Polym. Sci. 95, 7 (1994). 477. D. D. Lasic, J. Liposome Res. 5, 431 (1995). 478. B. W. Ninham, Y. Talmon, and D. F. Evans, Science 221, 1047 (1983). 479. E. Z. Radlinska, B. W. Ninham, J.-P. Dalbiez, and T. N. Zemb, Colloids Surf. 46, 213 (1990). 480. P. André, A. Filankembo, I. Lisiecki, C. Petit, T. Gulik-Krzywicki, B. W. Ninham, and M. P. Pileni, Adv. Mater. 12, 119 (2000). 481. C. Tanford, “The Hydrophobic Effect.” Wiley, New York, 1980. 482. M. Boström, D. R. M. Williams, and B. W. Ninham, Phys. Rev. Lett. 87, 168103 (2001).

Preparation of Vesicles (Liposomes) 483. B. W. Ninham, Prog. Colloid Polym. Sci. 120, 1 (2002). 484. M. Boström, D. R. M. Williams, and B. W. Ninham, Langmuir 18, 8609 (2002). 485. D. Gopinath, D. Ravi, B. R. Rao, S. S. Apte, and D. Rambhau, Int. J. Pharm. 246, 187 (2002). 486. G. B. Patel and G. D. Sprott, Crit. Rev. Biotechnol. 19, 317 (1999). 487. M. Conacher, J. Alexander, and J. M. Brewer, Vaccine 19, 2965 (2001). 488. C. Tondre and C. Caillet, Adv. Colloid Interface Sci. 93, 115 (2001). 489. K. Katagiri, R. Hamasaki, K. Ariga, and J. Kikuchi, Langmuir 18, 6709 (2002). 490. D. L. Melchior and A. Makriyannis, Biotechnol. Tech. 11, 885 (1997). 491. D. L. Melchior, J. Pharm. Sci. 91, 1075 (2002). 492. M. Fix and D. L. Melchior, FEBS Lett. 516, 109 (2002). 493. Y. C. Wang, C. J. Lee, W. K. Chen, C. I. Huang, W. F. Chen, G. J. Chen, and S. Z. Lin, Artif. Cell Blood Sub. 24, 35 (1996). 494. E. Mastrobattista, G. A. Koning, and G. Storm, Adv. Drug Delivery Rev. 40, 103 (1999). 495. M. De Cuyper and M. Joniau, Eur. Biophys. J. 15, 311 (1988). 496. M. De Cuyper and M. Joniau, Langmuir 7, 647 (1991). 497. M. De Cuyper, in “Handbook of Nonmedical Applications of Liposomes” (D. D. Lasic and Y. Barenholz, Eds.), Vol. 3, p. 325. CRC Press, Boca Raton, FL, 1996. 498. C. Sangregorio, J. K. Wiemann, C. J. O’Connor, and Z. Rosenzweig, J. Appl. Phys. 85, 5699 (1999). 499. N. Moussaoui, N. Cansell, and A. Denizot, Int. J. Pharm. 242, 361 (2002). 500. I. F. Uchegbu and S. P. Vyas, Int. J. Pharm. 172, 33 (1998). 501. P. Arunothayanun, I. F. Uchegbu, D. Q. M. Craig, J. A. Turton, and A. T. Florence, Int. J. Pharm. 183, 57 (1999). 502. P. Arunothayanun, I. F. Uchegbu, and A. T. Florence, J. Pharm. Pharmacol. 51, 651 (1999). 503. P. Arunothayanun, M.-S. Bernard, D. Q. M. Craig, I. F. Uchegbu, and A. T. Florence, Int. J. Pharm. 201, 7 (2000). 504. A. Kozubek, J. Gubernator, E. Przeworska, and M. Stasiuk, Acta Biochim. Pol. 47, 639 (2000). 505. H. M. Chen, V. Torchilin, and R. Langer, J. Controlled Release 42, 263 (1996). 506. K.-H. Song, S.-J. Chung, and C.-K. Shim, J. Controlled Release 84, 27 (2002). 507. C. Hu and D. G. Rhodes, Int. J. Pharm. 206, 110 (2000). 508. H. Kunieda and V. Rajagopalan, in “Vesicles,” Surfactant Science Series (M. Rosoff, Ed.), Vol. 62, p. 79. Dekker, New York, 1996. 509. R. Brunke, Manufacturing Chem. 61, 36 (1990). 510. T. M. Allen, Trends Pharm. Sci. 15, 215 (1994). 511. D. D. Lasic and D. Needham, Chem. Rev. 95, 2601 (1995). 512. D. D. Lasic and F. J. Martin, Eds., “Stealth Liposomes.” CRC Press, Boca Raton, FL, 1995. 513. J.-C. Bradley, M.-A. Guedeau-Boudeville, G. Jandeau, and J.-M. Lehn, Langmuir 13, 2457 (1997). 514. R. Glück and A. Wegmann, in “Medical Applications of Liposomes” (D. D. Lasic and D. Papahadjopoulos, Eds.), p. 75. Elsevier, Amsterdam, 1998. 515. T. Daemen, L. Burgener, A. Huckriede, and J. Wilschut, J. Liposome Res. 10, 329 (2000). 516. I. P. Hunziker, R. Zurbriggen, R. Glueck, O. B. Engler, J. Reichen, W. J. Dai, W. J. Pichler, and A. Cerny, Mol. Immunol. 38, 475 (2001). 517. L. T. Boni, S. R. Minchey, W. R. Perkins, P. L. Ahl, J. L. Slater, M. W. Tate, S. M. Gruner, and A. S. Janoff, Biochim. Biophys. Acta 1146, 247 (1993).

Preparation of Vesicles (Liposomes) 518. G. Cevc, in “Phospholipids Handbook” (G. Cevc, Ed.), p. 1. Dekker, New York, 1993. 519. C. Sun and M. Ueno, Colloid Polym. Sci. 278, 855 (2000). 520. T. Nakashima, T. Anno, H. Kanda, Y. Sato, T. Kuroi, H. Fujii, S. Nagadome, and G. Sugihara, Colloids Surf., B: Biointerfaces 24, 103 (2002). 521. K. Esumi, M. Miyazaki, T. Arai, and Y. Koide, Colloids Surf., A: Physicochem. Eng. Aspects 135, 117 (1998). 522. H. Heerklotz and J. Seelig, Biophys. J. 78, 2435 (2000). 523. C. E. Giacomelli, A. W. P. Vermeer, and W. Norde, Langmuir 16, 4853 (2000). 524. R. A. Schwendener, M. Asanger, and H. G. Weder, Biochem. Biophys. Res. Commun. 100, 1055 (1981). 525. L. T. Mimms, G. Zampighi, Y. Nozaki, C. Tanford, and J. A. Reynolds, Biochemistry 20, 833 (1981). 526. Y. Nozaki, D. D. Lasic, C. Tanford, and J. A. Reynolds, Science 217, 366 (1982). 527. D. Levy, A. Gulik, M. Seigneuret, and J.-L. Rigaud, Biochemistry 29, 9480 (1990). 528. N. Ostrowsky, Chem. Phys. Lipids 64, 45 (1993). 529. S. Kölchens, V. Ramaswami, J. Birgenheier, L. Nett, and D. F. O’Brien, Chem. Phys. Lipids 65, 1 (1993). 530. A. J. Jin, D. Huster, and K. Gawrisch, Eur. Biophys. J. 28, 187 (1999). 531. J. Pencer, G. F. White, and F. R. Hallett, Biophys. J. 81, 2716 (2001). 532. J. H. van Zanten and H. G. Monbouquette, J. Colloid Interface Sci. 146, 330 (1991). 533. J. H. van Zanten and H. G. Monbouquette, J. Colloid Interface Sci. 165, 512 (1994). 534. O. Stauch, R. Schubert, G. Savin, and W. Burchard, Biomacromolecules 3, 565 (2002). 535. J. P. P. Starink and A. J. Verkleij, J. Microsc. 167, 181 (1992). 536. D. D. Miller, J. R. Bellare, D. F. Evans, Y. Talmon, and B. W. Ninham, J. Phys. Chem. 91, 674 (1987). 537. S. Lesieur, C. Grabielle-Madelmont, M. Paternostre, and M. Ollivon, Chem. Phys. Lipids 64, 57 (1993). 538. D. Kardassi, D. Tsiourvas, and C. M. Paleos, J. Colloid Interface Sci. 186, 203 (1997). 539. O. Teschke and E. F. de Souza, Langmuir 18, 6513 (2002). 540. P. C. Mason, B. D. Gaulin, R. M. Epand, G. D. Wignall, and J. S. Lin, Phys. Rev. E 59, 3361 (1999). 541. M. A. Kiselev, P. Lesieur, A. M. Kisselev, D. Lombardo, M. Killany, and S. Lesieur, J. Alloys Compd. 328, 71 (2001). 542. A. Tardieu, V. Luzzati, and F. C. Reman, J. Mol. Biol. 75, 711 (1973). 543. Y. Tahara and Y. Fujiyoshi, Micron 25, 1412 (1994). 544. J. Pencer and F. R. Hallett, Phys. Rev. E 61, 3003 (2000). 545. P. Balgavy, M. Dubnicková, N. Kucerka, M. A. Kiselev, S. P. Yaradaikin, and D. Uhriková, Biochim. Biophys. Acta 1512, 40 (2001). 546. G. Ma, D. J. Barlow, M. J. Lawrence, R. K. Heenan, and P. Timmis, J. Phys. Chem B 104, 9081 (2000). 547. B. Kachar, D. F. Evans, and B. W. Ninham, J. Colloid Interface Sci. 100, 287 (1984). 548. J. Käs and E. Sackmann, Biophys. J. 60, 825 (1991). 549. E. Farge and P. F. Devaux, Biophys. J. 61, 347 (1992). 550. R. Lipowsky, Curr. Opin. Struct. Biol. 5, 531 (1995). 551. W. Wintz, H.-G. Döbereiner, and U. Seifert, Europhys. Lett. 33, 403 (1996).

79 552. F. M. Menger, S. J. Lee, and J. S. Keiper, Langmuir 12, 4479 (1996). 553. B. L.-S. Mui, H.-G. Döbereiner, T. D. Madden, and P. R. Cullis, Biophys. J. 69, 930 (1995). 554. L. Vander Elst, C. Piérart, S. L. Fossheim, J.-C. Raux, A. Roch, and R. N. Muller, Supramol. Chem. 14, 411 (2002). 555. J. C. McIntyre and R. G. Sleight, Biochemistry 30, 11819 (1991). 556. J. A. Bouwstra, G. S. Gooris, W. Bras, and H. Talsma, Chem. Phys. Lipids 64, 83 (1993). 557. L. A. Bagatolli and E. Gratton, Biophys. J. 78, 290 (2000). 558. K. Arnold, A. Lösche, and K. Garwrisch, Biochim. Biophys. Acta 645, 143 (1981). 559. A. Blume, R. J. Wittebort, S. K. Das Gupta, and R. G. Griffin, Biochemistry 21, 6243 (1982). 560. J. Korlach, P. Schwille, W. W. Webb, and G. W. Feigenson, Proc. Natl. Acad. Sci. USA 96, 8461 (1999). 561. V. Borenstain and Y. Barenholz, Chem. Phys. Lipids 64, 117 (1993). 562. M. Shinitzky and Y. Barenholz, Biochim. Biophys. Acta 515, 367 (1978). 563. P. Walde and E. Blöchliger, Langmuir 13, 1668 (1997). 564. J. Seelig and A. Seelig, Q. Rev. Biophys. 13, 19 (1980). 565. D. B. Fenske, Chem. Phys. Lipids 64, 143 (1993). 566. B.-S. Lee, S. A. Mabry, A. Jonas, and J. Jonas, Chem. Phys. Lipids 78, 103 (1995). 567. A. A. Nezorov and M. F. Brown, J. Chem. Phys. 107, 10288 (1997). 568. P. F. Devaux and H. M. McConnell, J. Am. Chem. Soc. 94, 4475 (1972). 569. J. Tabony and B. Perly, Biochim. Biophys. Acta 1063, 67 (1990). 570. S. König, W. Pfeiffer, T. Bayerl, D. Richter, and E. Sackmann, J. Phys. II France 2, 1589 (1992). 571. B. R. Lentz, Chem. Phys. Lipids 64, 99 (1993). 572. H. Hauser, W. Guyer, I. Pascher, P. Skrabal, and S. Sundell, Biochemistry 19, 366 (1980). 573. R. G. Snyder, G. L. Liang, H. L. Strauss, and R. Mendelsohn, Biophys. J. 71, 3186 (1996). 574. R. D. Kornberg and H. M. McConnell, Biochemistry 10, 1111 (1971). 575. N. Oku, D. A. Kendall, and R. C. MacDonald, Biochim. Biophys. Acta 691, 332 (1982). 576. W. R. Perkins, S. R. Minchey, P. L. Ahl, and A. S. Janoff, Chem. Phys. Lipids 64, 197 (1993). 577. K. Kawakami, Y. Nishihara, and K. Hirano, Anal. Biochem. 269, 139 (1999). 578. J.-H. Fuhrhop, C. Endisch, A. Schulz, J. Turner, and M. Eaton, Langmuir 15, 3707 (1999). 579. D. Huster, A. J. Jin, K. Arnold, and K. Gawrisch, Biophys. J. 73, 855 (1997). 580. R. G. Males, J. C. Nelson, P. S. Phillips, W. R. Cullen, and F. G. Herring, Biophys. Chem. 70, 75 (1998). 581. T.-X. Xiang and B. D. Anderson, Biophys. J. 75, 2658 (1998). 582. R. G. Males, J. C. Nelson, P. S. Phillips, W. R. Cullen, and F. G. Herring, Biophys. Chem. 70, 75 (1998). 583. S. Paula, A. G. Volkov, A. N. Van Hoeck, and D. W. Deamer, Biophys. J. 70, 339 (1996). 584. K. Matsuzaki, O. Murase, K. Sugishita, S. Yoneyama, K. Akada, M. Ueha, A. Nakamura, and S. Kobayashi, Biochim. Biophys. Acta 1467, 219 (2000).

1. Introduction Liposomes are spherical, selfclosed vesicles of colloidal dimensions, in which (phospho)lipid bilayer sequesters part of the solvent, in which they freely float, into their interior [1]. In the case of one bilayer encapsulating the aqueous core one speaks either of small or large unilamellar vesicles while in the case of many concentric bilayers one defines large multilamellar vesicles [2]. Due to their structure, chemical composition and colloidal size, all of which can be well controlled by preparation methods, liposomes exhibit several properties which may be useful in various applications. The most important properties are colloidal size, i.e. rather uniform particle size distributions in the range from 20 nm to 10 μm, and special membrane and surface characteristics. They include bilayer phase behavior, its mechanical properties and permeability, charge density, presence of surface bound or grafted polymers, or attachment of special ligands, respectively. Additionally, due to their amphiphilic character, liposomes are a powerful solubilizing system for a wide range of compounds. In addition to these physico-chemical properties, liposomes exhibit many special biological characteristics, including (specific) interactions with biological membranes and various cells [3]. These properties point to several possible applications with liposomes as the solubilizers for difficult-to-dissolve substances, dispersants, sustained release systems, delivery systems for the encapsulated substances, stabilizers, protective agents, microencapsulation systems and microreactors being the most obvious ones. Liposomes can be made entirely from naturally occurring substances and are therefore nontoxic, biodegradable and non immunogenic. In addition to these applications which had significant impact in several industries, the properties of liposomes offer a very useful model system in many fundamental studies from topology, membrane biophysics, photophysics and photochemistry, colloid interactions, cell function, signal transduction, and many others [3–5]. The industrial applications include liposomes as drug delivery vehicles in medicine, adjuvants in vaccination, signal enhancers/carriers in medical diagnostics and analytical biochemistry, solubilizers for various ingredients as well as support matrix for various ingredients and penetration enhancer in cosmetics. 2. Applications of liposomes in basic sciences Lipid membranes are two dimensional surfaces floating in three dimensional space. In the simplest models, they can be characterised only by their flexibility which is related to their bending elasticity. A number of new theoretical concepts were developed to understand their conformational behaviour [4]. On the other hand they 493

494

D.D. Lasic

Table 1 Applications of liposomes in the sciences. Discipline

Application

Mathematics

Topology of two-dimensional surfaces in three-dimensional space governed only by bilayer elasticity Aggregation behaviour, fractals, soft and high-strength materials Permeability, phase transitions in two-dimensions, photophysics Colloid behaviour in a system of well-defined physical characteristics, interand intra-aggregate forces, DLVO Photochemistry, artificial photosynthesis, catalysis, microcompartmentalization Reconstitution of membrane proteins into artificial membranes Model biological membranes, cell function, fusion, recognition Studies of drug action Drug-delivery and medical diagnostics, gene therapy

Physics Biophysics Physical Chemistry Chemistry Biochemistry Biology Pharmaceutics Medicine

Table 2 Liposomes in the pharmaceutical industry. Liposome Utility Solubilization Site-Avoidance

Current Applications

Amphotericin B, minoxidil Amphotericin B – reduced nephrotoxicity, doxorubicin – decreased cardiotoxicity Sustained-Release Systemic antineoplastic drugs, hormones, corticosteroids, drug depot in the lungs Drug protection Cytosine arabinoside, interleukins RES Targeting Immunomodulators, vaccines, antimalarials, macophage-located diseases Specific Targeting Cells bearing specific antigens Extravasation Leaky vasculature of tumours, inflammations, infections Accumulation Prostaglandins Enhanced Penetration Topical vehicles Drug Depot Lungs, sub-cutaneous, intra-muscular, ocular

Disease States Treated Fungal infections Fungal infections, cancer Cancer, biotherapeutics Cancer, etc. Cancer, MAI, tropical parasites Wide therapeutic applicability Cancer, bacterial infections Cardiovascular diseases Dermatology Wide therapeutic applicability

can be used as a model in order to understand the topology, shape fluctuations, phase behaviour, permeability, fission and fusion of biological membranes. Their aggregation leads to fractal clusters. In addition they can serve as a model to study vesiculation, including vesicle shedding and endo- and exo-cytosis, of living cells (table 1). Despite their widespread application, the mechanism of liposome formation is not yet well understood. The equilibrium calculations of the shapes of giant unilamellar vesicles [7, 8] and their observations (fig. 1A) [9, 10], however, offer a qualitative guidance in the modeling of structural transformations in the various vesiculation processes. Figure 1B shows that similar shapes occur also in multilamellar aggregates and it is reasonable to assume that the gradient of hydration across the stack of concentric lamellae causes also the gradient of surface areas of polar heads in the consecutive monolayers because the area of polar head is proportional to hydration. As a result of this imbalance the curvature is induced. This indicates that a similar

498

D.D. Lasic

3. Applications of liposomes in medicine Applications of liposomes in pharmacology and medicine can be divided into therapeutic and diagnostic applications of liposomes containing drugs or various markers, and their use as a model, tool, or reagent in the basic studies of cell interactions, recognition processes, and of the mode of action of certain substances [3]. Unfortunately many drugs have a very narrow therapeutic window, meaning that the therapeutic concentration is not much lower than the toxic one. In several cases the toxicity can be reduced or the efficacy enhanced by the use of an appropriate drug carrier which changes the temporal and spatial distribution of the drug, i.e. its pharmacokinetics and biodistribution. The benefits and limitations of liposome drug carriers critically depend on the interaction of liposomes with cells and their fate in vivo after administration. In vitro and in vivo studies of the interactions with cells have shown that the predominant interaction of liposomes with cells is either simple adsorption or subsequent endocytosis. Fusion with cell membranes is much more rare. The fourth possible interaction is exchange of bilayer constituents, such as lipids, cholesterol, and membrane bound molecules with components of cell membranes. These interactions, schematically shown in fig. 3, determine also the fate of liposomes in vivo. The body protects itself with a complex defense system. Upon entering into the body, larger objects cause thrombus formation and eventually their surface is passivated by coating with biomacromolecules while smaller particles, including microbes, bacteria, and colloids are eaten up by the cells of the immune system. This response of the immune system has triggered substantial efforts in the development of biocompatible and nonrecognizable surfaces and has also, on the other hand, narrowed the spectrum of applications of microparticulate drug carriers only to targeting of the very same cells of the immune system.

Fig. 3. Schematic presentation of liposome interactions with cells. Endocytosis is shown in the upper left part of the cell. In addition, clockwise fusion, lipid exchange and adsorption (of leaky vesicle) are shown. From ref. [3], with permission.

Applications of liposomes

499

Although they are composed from natural substances liposomes are no exception. They are rapidly cleared from the circulation by the macrophages which are located mainly in the liver, spleen, and bone marrow. 3.1. Modes of liposome action Liposomes as drug delivery systems can offer several advantages over conventional dosage forms especially for parenteral (i.e. local or systemic injection or infusion), topical, and pulmonary route of administration. The preceding discussion shows that liposomes exhibit different biodistribution and pharmacokinetics than free drug molecules. In several cases this can be used to improve the therapeutic efficacy of the encapsulated drug molecules. The limitations can be reduced bioavailability of the drug, saturation of the cells of the immune system with lipids and potentially increased toxicity of some drugs due to their increased interactions with particular cells. The benefits of drug loaden liposomes, which can be applied as (colloidal) solution, aerosol, or in (semi) solid forms, such as creams and gels, can be summarized into seven categories: (i) Improved solubility of lipophilic and amphiphilic drugs. Examples include Porphyrins, Amphotericin B, Minoxidil, some peptides, and anthracyclines, respectively; furthermore, in some cases hydrophilic drugs, such as anticancer agent Doxorubicin or Acyclovir can be encapsulateded in the liposome interior at concentrations several fold above their aqueous solubility. This is possible due to precipitation of the drug or gel formation inside the liposome with appropriate substances encapsulated [17]; (ii) Passive targeting to the cells of the immune system, especially cells of the mononuclear phagocytic system (in older literature reticuloendothelial system). Examples are antimonials, Amphotericin B, porphyrins and also vaccines, immunomodulators or (immuno)supressors; (iii) Sustained release system of systemically or locally administered liposomes. Examples are doxorubicin, cytosine arabinose, cortisones, biological proteins or peptides such as vasopressin; (iv) Site-avoidance mechanism: liposomes do not dispose in certain organs, such as heart, kidneys, brain, and nervous system and this reduces cardio-, nephro-, and neuro-toxicity. Typical examples are reduced nephrotoxicity of Amphotericin B, and reduced cardiotoxicity of Doxorubicin liposomes; (v) Site specific targeting: in certain cases liposomes with surface attached ligands can bind to target cells (‘key and lock’ mechanism), or can be delivered into the target tissue by local anatomical conditions such as leaky and badly formed blood vessels, their basal lamina, and capillaries. Examples include anticancer, antiinfection and antiinflammatory drugs; (vi) Improved transfer of hydrophilic, charged molecules such as chelators, antibiotics, plasmids, and genes into cells; and (vii) Improved penetration into tissues, especially in the case of dermally applied liposomal dosage forms. Examples include anaesthetics, corticosteroids, and insulin.

Applications of liposomes

505

Fig. 5. Comparison of blood clearance of various liposome formulations in rats following intravenous administration of 5 μM lipid/kg. Liposomes contained 39.5 mol% of egg phosphatidylcholine (PC), 33 mol% cholesterol (Chol) and 7.5 mol% of either egg phosphatidylglycerol (PG) or polyethylene (1900 Da) coupled to distearoyl phosphatidylethanolamine (1900 PEG-DSPE). Clearance of the free label, Galium desferal is also shown. Courtesy M.C. Woodle and M. Newman.

3.3.1. Sterically stabilised liposomes The fate of liposomes, i.e. their rapid clearance from the body, was realized rather early. First attempts to alter their biodistribution by either surface ligands or membrane composition were undertaken in the late 70’s. The results showed that liposome disposition can be altered, but predominantly within the mononuclear phagocytic system including the intrahepatic uptake itself. Blood circulation times were prolonged but the first substantial improvements were achieved by the incorporation of ganglioside GM1 or phosphatidylinositol at 5–10 mol% into the bilayer [33, 34]. The best results were obtained by substituting these two lipids with synthetic polymer containing lipids. The longest circulation times were achieved when polyethylene glycol covalently bound to the phospholipid was used. It seems that intermediate molecular weights, from 1500 to 5000 Da are the optimum [35]. Figure 5 shows blood clearance profiles of several different formulations. It was suggested [36] that the presence of a steric barrier reduces adhesion and adsorption (or at least adsorption with a conformational change) of blood components, such as immunoglobulins, complement proteins, fibronection and similar molecules, which mark foreign particles for subsequent macrophage uptake as schematically shown in fig. 6. The origin of steric stabilisation is well documented although not well understood. Recently it was shown that the Alexander-de-Gennes model of polymers at interfaces [37] can qualitatively explain the stability of liposomes in biological systems [35]. The model can explain minimal polymer concentration above the surface of the bilayer at which polymer forms the so-called brush conformation and which acts as

506

D.D. Lasic

Fig. 6. A schematic presentation of the proposed mechanism of the stabilisation of liposomes in biological environments. The presence of the polymer mushroom or brush [68] reduces the adsorption and adhesion rate of immunoglobulins and other antibodies (A) and plasma (lipo) proteins (P) which either mark liposomes for the subsequent macrophage uptake or deplete lipids which can cause liposome disintegration.

a steric shield. Small angle X-ray scattering measurements of force-distance profiles of polyethylene glycol grafted liposomes have shown enhanced bilayer repulsion [38, 39] in agreement with the hypothesis that reduced surface adhesion and adsorption stabilises liposomes. Recent theoretical work also explained the experimentally wellknown fact that increased concentrations of longer chains start to reverse the effect at particular polymer density. This is due to the so-called collapse of the brush which occurs at certain polymer density and results in polymer selfaggregation [40], a wellknown fact from the experimental polymer science. Longer chains can also exhibit increased attractive and bridging forces with macromolecules [41]. Of course, the in vivo and in vitro stability are not necessarily correlated and, for example, in vitro very stable formulations, such as highly charged ones, or the ones with charged brush, are cleared in vivo very rapidly. Another factor which may differ between the two tests is the role of chain flexibility on the interactions with particles and proteins. It is possible that the decreased mobility of chains in the denser brush regimes, when the chain motion correlation times may approach times required for protein binding, can account for the weak physisorption of proteins. 3.3.2. Medical applications of stealth liposomes Sterically stabilised vesicles can act either as long circulating microreservoirs or tumour (or site of inflammation and infection) targetting vehicles. The former applications requires larger liposomes (∼ 0.2 μm) while the latter one is due to the ability of small vesicles to leave the blood circulation. The prolonged presence of small particulates in blood results in effective extravasation in regions with porous, damaged, or badly formed blood vessels which often characterise tumours or their

Applications of liposomes

507

Fig. 7 Blood vessels (vasculature) of (A) normal and (B) tumour tissue as viewed by rhodaminephosphatidylethanolamine labelled long circulating liposomes approximately one hour after injection. The difference between the healthy and tumour tissue, which shows accumulation of fluorescent liposomes in extravascular sites can be clearly observed. Plates C and D show accumulation of Doxorubicin loaded Stealth liposomes in mammary adenocarcinoma tumour one hour and one day, respectively, after tail vein bolus injection in female Fischer rat skin flap window chamber model. In this study fluorescence of the encapsulated drug Doxorubicin was used as a marker. The intensity is much lower, however, due to the self quenching effect of highly concentrated drug inside the liposomes. Due to this high concentration, which is 3–5 fold above its aqueous solubility the drug is precipitated or gelled with counterions in the vesicle interior. Because these liposomes practically do not leak their contents the blood vessels can be barely seen either immediately after injection or one hour later (C). Few focal points of accumulation which are outside blood circulation (as can be easily verified by the bright field light microscopy of the same area) show that after one hour there is already some extravasation (C). The increase of fluorescence intensity at 24 hours, however, indicates not only high accumulation of the drug but also the fact that it is being released from the liposomes which reduces or eliminates selfquenching (D). (N. Wu, M. Dewhirst, D.D. Lasic, D. Needham, unpublished data. For details see ref. [39].)

vicinity. While normal molecules and macromolecules quickly come to equilibrium large doses of liposomes can accumulate due to their adhesion or immobilization. (In analogy with biocompatible surfaces we can speculate that PEG chains effectively reduce the adsorption of proteins while for the prevention of cell adhesion much longer chains would be required [42]. At present, it is still not known if such long chains can be effectively incorporated into liposomes.) This allows larger doses of liposome loaden drugs to be delivered to malignant tissues. For instance more then 10% of the injected dose of stealth liposome encapsulated Doxorubicin was found in tumours [43] as opposed to around 1% when free drug was administered. Figure 7 shows extravasation of Doxorubicin encapsulated in Stealth liposomes in

508

D.D. Lasic

Applications of liposomes

509

the dorsal flap window rat model, i.e. an animal model which allows the viewing in the fluorescence microscope of the biodistribution of fluorescent molecules or labelled liposomes in vivo, in this case dorsal tissue of rat clipped between special microscopic slides [39]. Extensive liposome localization in the tumours was observed. Healthy tissue did not accumulate any signal which was due to Doxorubicin fluorescence [39]. Efficacy studies in various mice tumour models, such as implanted solid C26 carcinoma and inoculated mammary carcinoma, have shown dramatic improvements [44–48]. Solid C26 colon tumour is practically resistant to free drug, conventional Epirubicin (a very similar drug to Doxorubucin) liposomes, and mixtures of free drug and empty stealth liposomes. Stealth Epirubicin and Doxorubicin liposomes resulted, however, in complete remissions of tumours in the early treatment schedule and substantial reduction of tumour size in the delayed treatment regime (fig. 8) [45]. Similar improvement in therapy was observed also in mammary carcinoma (fig. 9) [46]. These formulations were substantially more effective not only in curing mice with recent implants from various tumours but also in reducing the incidence of metastases originating from these intra mammary implants. Similarly, several fold increased drug accumulation was observed also in sites of infections which are also characterized by the enhanced vascular permeability. For instance, in mice with infected lungs 10 fold more antibiotic drug accumulated in the infected lung as compared to the noninfected one [49]. Sterically stabilised liposomes may act also as a sustained drug release system either as a long circulating microreservoir or localised drug depot. The first example

Fig. 9. Tumour size as a function of time for various treatments. Mammary carcinima MC2 were implanted into syngeneic female mice and animals were treated at days 3, 10, 17 (A) or 10, 17, 24 (B) after implantation with saline, free doxorubicin (Dox) and doxorubicin encapsulated in Stealth liposomes (S-Dox) at 2 concentrations. Each point is the average of 20 tumours. (From ref. [46], with permission)

510

D.D. Lasic

Fig. 8. Effect of various formulations of Epirubicin on the growth of C26 tumour. Ten mice in each group were injected with one million C26 cells and treatments began and continued on days 3, 10, 17 (left) or 10, 17, 24 (right) after inoculation. (A): saline control, (B): free Epi at 6 mg/ml, (C): Epirubicin in stealth liposomes (S-Epi) at 6 mg/ml and, (D): at 9 mg/ml (which resulted in no observable tumour at all). (E): mixture of free drug and empty liposomes. (F): the average values of ten animals in each experiment from A to E. From ref. [45], with permission. Practically identical results were also observed with Doxorubicin (see ref. [47]).

Applications of liposomes

511

is provided by improved therapeutic efficacy of cytosine arabinose in the treatment of lymphoma [48] while the subcutaneous/intramuscular sustained release system was demonstrated by the action of polypeptide vasopressin [50]. Its action was prolonged up to a month as compared to few days for a free drug and a week for the peptide encapsulated in conventional liposomes. It is important to note that these concepts are becoming more and more important with the introduction of genetically engineered polypeptides and proteins which are hampered by the rapid blood clearance, degradation and/or deactivation in the body. The altered biodistribution of stealth liposomes, in addition to the accumulation at the sites characterised with porous blood capillaries, such as in tumours, inflammations, and infections, may benefit several other applications. In the intact vasculature the distribution of stealth liposomes is shifted from the liver, spleen, and bone marrow towards skin. This opens the opportunity to deliver antivirals and dermatological agents to these sites. On the other hand, and while it was shown that the administration of empty stealth liposomes is well tolerated [5], it requires careful toxicology and tolerability studies when liposomes loaden with potent drugs are used. 3.3.3. Applications of Stealth liposomes in man The encouraging results of Doxorubicin encapsulated in Stealth liposomes in preclinical studies were observed also in clinical trials in humans. Blood circulation times around 45 hours were found [51] and at reduced toxicity very good response in AIDS patients with Kaposi sarcoma was observed [52, 53]. Long circulation times significantly, i.e. 200-fold, increased the area under curve of drug concentration vs time and accumulation in various tumours was proportional to the liposome circulation times [51]. The drug remained encapsulated in circulating liposomes up to one week after injection while at tumour sites drug metabolites were found indicating that it had been released from liposomes. The concentration of the drug in tumours was 4–10 times greater than in control group which was treated with free drug [51]. The same selective targetting was observed also in patients with Kaposi sarcoma. Practically all patients showed considerable decrease in modularity of skin lesions while total flattening was observed in 25% of the cases [52]. The high efficacy was due to the approximately ten fold higher drug concentration in lesions as compared to the administration of free drug (table 3) [53]. In conclusion, it seems that stealth liposomes loaded with anticancer drugs will achieve substantial improvements in the treatments of various tumours. In addition, it Table 3 Localization of Doxorubicin in Kaposi Sarcoma lesions after intravenous injection of free drug and drug encapsulated in Stealth liposomes (from ref. [53]). dose [mg/m2 ] 10 20 40

Doxorubicin concentration μg/g tissue free drug in stealth liposome 0.18 0.31 0.72

2.06 1.61 7.11

Selectivity Index 11.4 5.2 9.9

Applications of liposomes

513

5. Application of liposomes in cosmetics The same properties of liposomes can be utilized also in the delivery of ingredients in cosmetics. In addition, liposomes as a carrier itself offers advantages because lipids are well hydrated and can reduce the dryness of the skin which is a primary cause for its ageing. Also, liposomes can act as a supply which acts to replenish lipids and, importantly, linolenic acid. In general the rules for topical drug applications and delivery of other compounds are less stringent than the ones for parenteral administration and several hundred cosmetic products are commercially available since Capture (C. Dior) and Niosomes (L’Or´eal) were introduced in 1987. They range from simple liposome pastes which are used as a replacement for creams, gels, and ointments for do-it-yourself cosmetical products to formulations containing various extracts, moisturizers, antibiotics, and to complex products containing recombinant proteins for wound or sunburn healing. Most of the products are anti-ageing skin creams. Unrinsable sunscreens, long lasting perfumes, hair conditioners, aftershaves and similar products, are also gaining large fractions of the market. Liposomal skin creams already share more than 10% of the over $10 billion market. Table 4 shows some of these products. As in the case of topical delivery in medical applications, the workers in the field do not agree on the mechanism of action. While some claim enhanced permeability into the skin the others claim mostly that liposomes are a noninteractive, skin-nonirritating, water based matrix (without alcohols, detergents, oils and other non-natural solubilizers) for the active ingredients. In addition to the natural lipids, either phospholipids or ‘skin lipids’, which contain mostly sphingolipids, ceramides, oleic acid, and cholesterol sulphate, liposomes made from synthetic lipids are also being used. They include mostly liposomes made from nonionic surfactant lipids, which can be chemically more stable. Some of these Table 4 Some liposomal cosmetic formulations currently on the market. According to the manufacturers, liposomes may deliver moisture and a novel supply of lipid molecules to skin tissue in a superior fashion to other formulations. In addition they can entrap a variety of active molecules and can therefore be utilized for skin creams, anti-aging creams, after shave, lipstic, sun screen and make-up. Product

Manufacturer

Liposomes and key ingredients

Capture Efect du Soleil Niosomes Nactosomes Formule Liposome Gel Future Perfect Skin Gel

Cristian Dior L’Or´eal Lancome (L’Or´eal) Lancome (L’Or´eal Payot (Ferdinand Muehlens) Estee Lauder

Symphatic 2000 Natipide II Flawless finish Inovita Eye Perfector Aquasome LA

Biopharm GmbH Nattermann PL Elizabeth Arden Pharm/Apotheke Avon Nikko Chemical Co.

Liposomes in gel with ingredients Tanning agents in liposomes Glyceropolyether with moisturizers Vitamins Thymoxin, hyaluronic acid TMF, vitamins E, A palmitate, cerebroside ceramide, phospholipid Thymus extract, vitamin A palmitate Liposomal gel for do-it-yourself cosmetics Liquid make-up Thymus extract, hyaluronic acid, vitamin E Soothing cream to reduce eye irritation Liposomes with humectant

The FASEB Journal • Milestone

From “Banghasomes” to liposomes: A memoir of Alec Bangham, 1921–2010 David W. Deamer1 Department of Biomolecular Engineering, University of California Santa Cruz, Santa Cruz, California, USA.

Alec Bangham (1921-2010)

“It was the odd pattern of a well-drawn drop of blood that initiated my curiosity and eased my career away from morbid anatomy to that of the physical chemistry of cell surfaces (1).” A few individuals change the course of a scientific life if one is fortunate enough to work with them. Alec 1308

Bangham was such a person, and I feel privileged to have known him as a friend and colleague. On March 9,

1

Correspondence: E-mail: [email protected] doi: 10.1096/fj.10-0503 0892-6638/10/0024-1308 © FASEB

2010, Alec died peacefully at his home in Great Shelford, England, after a brief illness. Alec Bangham is best known for his seminal research on what he called “multilamellar smectic mesophases,” sometimes less seriously referred to as “banghasomes.” Gerald Weissmann, current editor of The FASEB Journal, was one of the first visitors to the Bangham laboratory and decided that a more descriptive term was needed. He proposed “liposomes,” defined as microscopic vesicles composed of one or more lipid bilayers (2). The name stuck, and over the next 30 years liposomes grew into a minor industry, with the word incorporated into a number of book titles and a scientific journal, two successful companies, a drug delivery agent, a treatment for infantile respiratory distress syndrome, and even cosmetic formulations by Dior and Lancoˆme. By March 9, 2010 PubMed had listed 35,604 articles on liposomes. The liposome story began with a paper in 1964, published in the Journal of Molecular Biology, in which Bangham and Horne showed electron microscopic images of multilamellar phospholipid vesicles (3). By 1965, Bangham and his co-workers had done the crucial experiments and in twin publications reported that the lipid bilayers of the vesicles could maintain concentration gradients of ions such as potassium and sodium. Moreover, when natural or synthetic detergents perturbed the bilayer, the gradient was disrupted (4, 5). This evidence, along with the planar bilayer models being developed at the same time, established that lipid bilayers are the primary permeability barrier of all cell membranes. It was the membrane equivalent of finding the double helix structure of DNA, another Cambridge discovery in the life sciences. In 1975, I decided to spend a sabbatical in Alec’s laboratory at Babraham, at the time little more than a village a few miles south of Cambridge. I knew that several years earlier, Alec had given a talk at Bristol University with the title “Membranes Came First!” in which he proposed that something like liposomes must have been available to house the first forms of cellular life. Inspired by this idea, I returned to my home campus and soon found that phospholipids could be synthesized in simulated prebiotic conditions (6) and that membrane-forming amphiphilic molecules were present in the Murchison meteorite (7). Furthermore, simple cell-like structures could be produced by encapsulating functional nucleic acid polymerases in lipid vesicles (8). With Jack Szostak, I am editing a book to be published in 2010 by Cold Spring Harbor Press in which expert authors give their perspectives on the origins of life, several of whom echo Alec’s prescient assertion that membranes came first. Throughout his time at Babraham, Alec’s laboratory consisted of a set of connected barrack structures left over from the 1940s. There was nothing grand about it, nor was there any pretentiousness about Alec himself. Alec was not ambitious in the usual sense, so he never fit very well into the British scientific establishment. His research was curiosity-driven, and he boldly followed MILESTONE: A MEMOIR OF ALEC BANGHAM, 1921–2010

his ideas wherever they led. At first, liposomes seemed to be just a laboratory curiosity, but the excitement generated by Alec’s discovery finally penetrated the establishment and he was named a Fellow of the Royal Society in 1977. Alec probably never intended to be a research scientist. Instead, he trained as a clinician, and as a young medic he was posted to Israel where the British forces were being attacked in the chaotic struggle from which the nation of Israel emerged. He returned for a short visit in 1998 while I was working at the Weizmann Institute, and we drove around the countryside near Rehovot that he remembered clearly from his experiences as a young physician half a century earlier. One of the sites had become a geriatric hospital, and Alec pointed to a tree under which he had buried a pile of medical equipment when the British were forced to abandon Palestine. In 1952, Alec accepted a research position at the Institute of Animal Physiology at Babraham, where he spent the rest of his scientific career. By this time he had met Rosalind, who had her own practice as a physician. They were soon wed, and settled in nearby Great Shelford to raise their family of four children, Andrew, Janet, Oliver, and Daniel. Andrew now serves on the faculty of the University of East Anglia, specializing in computational biology; Oliver runs a private consulting firm for industrial management in London; Janet is a well-known water color artist in Cambridge; and Daniel has earned an international reputation for his superb hand-crafted woodwind instruments. Alec’s first application of liposomes as a model membrane system was to test a hypothesis related to the action of general anesthetics, which proposed that anesthetics partitioned into the lipid bilayer moiety and in some way inhibited nervous function. In a collaborative study with Sheena Johnson and Keith Miller (9), Alec demonstrated that liposomes exposed to general anesthetics became significantly more permeable to ionic solutes. It had previously been shown that anesthesia could be reversed by hyperbaric pressure, but the reason was not at all clear. Alec and his co-authors demonstrated that pressure could also reverse the permeabilizing effect of anesthetics on liposome membranes. This suggested that anesthetic action could be understood in thermodynamic terms, in which pressure, temperature, and ionic permeability all must be taken into account. Keith Miller, now in the Department of Anesthesia, Massachusetts General Hospital, later commented: “To a field whose most powerful model nearly seven decades ago had been a jar of olive oil, the liposome’s arrival was a liberating force (10).” Alec recently wrote an account of this research for The FASEB Journal (11). Alec also applied liposomes to other clinical questions, one of which involved infantile respiratory distress syndrome. It is known that a monomolecular layer of a phospholipid called dipalmitoylphosphatidylcholine (DPPC) coats the inner surface of lung alveoli. At birth, the decreased surface tension produced by the 1309

monolayer is essential for expansion of the lung with the first breath of the infant. However, particularly in premature newborns, there is sometimes an insufficient amount of DPPC so that their lungs are unable to expand. Alec had the idea that a mixture of two phospholipids he called “artificial lung expanding compound” (note the acronym) could be delivered to the airways to supply the monolayer (12). To quote the concluding sentence of a later paper describing a clinical trial of the preparation: “Artificial surfactant (ALEC) given to very premature babies at birth significantly reduces their mortality and the respiratory support needed and should prove a valuable addition to treatment (13).” Alec’s creative instincts were not limited to his research, but were also on display when he gave talks about liposomes. I once invited him to give a seminar at UC Davis, where I was a member of the biology faculty. Alec asked if he could borrow my Teflon Langmuir trough, some water saturated with diethyl ether, a sample of phospholipid, a graduate student, and a fire extinguisher. The lecture hall was packed, and Alec began by introducing the barrier properties of bilayer membranes, pointing out that even a lipid monolayer could inhibit molecular diffusion across an interface. He poured the ether-water into the trough, stationed the graduate student and fire extinguisher nearby, and held a match to the trough which erupted in flaming ether! He then dipped a glass rod into the phospholipid sample and touched it to the water surface. In a few seconds, as the monolayer spread over the trough, the flames were magically extinguished. Alec cleared the monolayer and lit the ether again. “Any surfaceactive substance works,” he said, then stuck his finger in his ear to get a bit of earwax with which he again extinguished the flaming ether. Memorable! Last year Alec wrote to several of his friends in great excitement. He had been pondering another utterly novel idea: that mixtures of volatile materials he called bouquets could affect the immune response by altering the surface charge on membranes. Its editor recommended that he submit the manuscript to The FASEB Journal, and Alec’s last sole author paper was published in 2009 (1). I know that this publication gave him immense pleasure, a remarkable achievement for an 88-year-old scientist and something we can all hope to emulate.

I was able to visit Alec late in 2009. Rosalind had died a few months earlier, and this loss clearly affected him deeply. However, he still managed to prepare a pot of his favorite meringue biscuits from egg whites (see the photograph) and ordered in a curry for lunch. We spent a couple of hours discussing surface charges on cell membranes and how bouquets of volatile compounds might be able to hide the cells from the immune response. Once again, Alec has given me something to think about. I will miss him.

REFERENCES 1. 2. 3.

4. 5. 6. 7. 8. 9.

10. 11. 12.

13.

Bangham, A. (2009) The physical chemistry of self/non-self: jigsaws, transplants and fetuses. FASEB J. 23, 3644 –3646 Sessa, G., and Weissmann, G. (1968) Phospholipid spherules (liposomes) as a model for biological membranes. J. Lipid Res. 9, 310 –318 Bangham, A. D., and Horne, R.W. (1964) Negative staining of phospholipids and their structural modification by surface active agents as observed in the electron microscope. J. Mol. Biol 8, 660 – 668 Bangham, A. D., Standish, M. M., and Watkins, J. C. (1965) Diffusion of univalent ions across the lamellae of swollen phospholipids. J. Mol. Biol. 13, 238 –252 Bangham, A. D., Standish, M. M., and Weissmann, G. (1965) The action of steroids and streptolysin S on the permeability of phospholipid structures to cations. J. Mol. Biol. 13, 253–259 Hargreaves, W.W., Mulvihill, S. J., and Deamer, D.W. (1977) Synthesis of phospholipids and membranes in prebiotic conditions. Nature 266, 78 – 80 Deamer, D.W. (1985) Boundary structures are formed by organic compounds of the Murchison carbonaceous chondrite. Nature 317, 792–794 Chakrabarti, A., Joyce, G. F., Breaker, R. R., and Deamer, D. W. (1994) RNA synthesis by a liposome-encapsulated polymerase. J. Mol. Evol. 39, 555–559 Johnson, S. M., Miller, K. W., and Bangham, A. D. (1973) The opposing effects of pressure and general anaesthetis on the cation permeability of liposomes of varying lipid composition. Biochim. Biophys. Acta. 307, 42–57 Miller, K. W. (1983) Anaesthetized Liposomes. In: Liposome Letters, pp. 251–9. (A. D. Bangham, ed.) Academic Press, London Bangham, A. D. (2005) Liposomes and the physico-chemical basis of unconsciousness. FASEB J. 19, 1766 –1768 Bangham, A. D., Miller, N. G. A., Davies, R. J., Greenough, A., and Morley, C. J. (1984) Introductory remarks about artificial lung expanding compounds (ALEC). Colloids and Surfaces Physical chemistry of colloids and interfaces: Biotechnologies and drug research. 10, 337– 341 Ten Centre Study Group (1987) Ten Centre trial of artificial surfactant (artificial lung expanding compound) in very premature babies. Brit. Med. J. (Clinical Research Ed.) 294, 991–996.

The opinions expressed in editorials, essays, letters to the editor, and other articles comprising the Up Front section are those of the authors and do not necessarily reflect the opinions of FASEB or its constituent societies. The FASEB Journal welcomes all points of view and many voices. We look forward to hearing these in the form of op-ed pieces and/or letters from its readers addressed to [email protected]. 1310

Vol. 24

May 2010

The FASEB Journal 䡠 www.fasebj.org

DEAMER