Journal of Colloid and Interface Science 309 (2007) 78–85www.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 solutionsin parallel into aqueous solutions of Arabic gum. Except for the very early stages, the precipitated solids consisted of a well-defined zinc oxidephase. Depending on the experimental conditions, the size of the final polycrystalline particles formed by the aggregation of nanosize entitiesvaried 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 muchlarger (44 nm) and combined to form rather uniform hexagonal ZnO prisms. The aspect ratio and the internal structure of the latter could bealtered 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 atwo-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 uniqueproperties, which have been successfully exploited in manytechnological applications. The most notable uses of dispersedZnO are in catalysis [1,2], construction of varistors and gas sen-sors [3–5], pigments [6], luminescent materials [7,8], and in thepharmaceutical and cosmetic industries [9,10]. Zinc oxide pow-ders have also been employed as model materials for studyingthe sintering, grain growth, and microstructural development inceramics [11–13]. Considering the specific needs of these verydiverse applications, there has been strong interest in the devel-opment of preparation methods enabling to generate highly dis-persed zinc oxide with well-controlled properties. Among otherpossibilities, the thermal decomposition/evaporative and the so-lution based methods remain the preferred synthetic routes togenerate colloidal ZnO. The first approach, which consists inconverting zinc precursors directly into crystalline zinc ox-

* Corresponding author. Fax: +1 315 268 2139.E-mail address: goidanv@clarkson.edu (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 so-lution route relies on the hydrolysis of zinc ions in alkalineconditions, which often yields highly dispersed uniform ZnOparticles with well-controlled size and shape [21,22]. Experi-mentally, 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 mi-croemulsion [33] techniques. The nature of the reactants, theirconcentrations, the type and the charge of the counterions, thereaction temperature and time, and the presence of dispersantswere all found to affect the formation of ZnO particles [23]. Ad-ditionally, it was shown that the way the reactant solutions aremixed can also influence the nucleation and growth of the solidphase and, therefore, the properties of the final solids [23,24].

This paper describes a simple precipitation method forpreparing highly dispersed uniform zinc oxide colloids at lowtemperatures and discusses the mechanisms responsible fortheir formation. It will be shown that the final polycrystallineZnO particles are formed by the aggregation of nanosize pri-mary particles and their internal structure and shape depend onthe 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)2·6H2O},zinc sulfate {ZnSO4·7H2O}, zinc acetate {Zn(CH3COO)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 (Flu-tarom Inc., North Bergen/NJ) were used as received.

2.2. Preparation methods

The precipitations were carried out in a 1000 cm−3 jacketedcylindrical beaker connected to a constant temperature circu-lating bath. In a typical experiment, equal volumes (100 cm3)of zinc salt and sodium hydroxide stock solutions were addedin parallel into 150 cm3 of deionized water containing the dis-solved dispersing agent. The reacting solutions were deliveredat flow rates ranging from 0.28 to 1.1 cm3 min−1 using twometering pumps, the mixing being provided by a 1′′ diameterthree-blade SS propeller spinning at 500 rpm. The majority ofthe precipitations were carried out at 60 ◦C using zinc nitrate. Inthe experiments in which a dispersant was employed, the quan-tity added was arbitrarily selected to be equal to the amount ofZnO formed assuming the complete precipitation of zinc ions.After the addition of the reactants was completed, the result-ing white precipitate was washed repeatedly with distilled waterby decantation. In the case of the experiments in which stabledispersions were formed, the solids were separated by ultracen-trifugation and then redispersed in DI water. Once the pH ofthe supernatant was below ∼8.0, the ZnO particles were rinsedtwice with alcohol, filtered, and dried at 80 ◦C for several hours.

2.3. Characterizations

The crystalline structure of the dried solids was evaluatedby X-ray powder diffraction with a Bruker Advance D8 dif-fractometer using CuKα (λ = 1.5418 Å) as incident radiation.The 2θ range was scanned with a step of 0.02◦, the data col-lection time for each step being 2 s. In addition to identifyingthe crystalline phases, the XRD data were also used to es-timate the size of the constituent crystallites by the Scherrermethod. The size and morphology of the ZnO particles, as wellas their surface topography, were evaluated by field emissionelectron microscopy (FE-SEM) with a JEOL-JSM-7400F in-strument. The particle size and size distribution of the dispersedZnO were obtained from the SEM images by averaging mini-mum 100 particles and by laser diffraction size analysis using aMalvern 2000e instrument.

3. Results

The reaction temperature, the zinc salt type, the nature ofthe dispersing agent, the reactants concentration and the addi-tion 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 theabsence of dispersant (Sample 1), (b) with Daxad 11G (Sample 2), and (c) withArabic gum (Sample 3).

The slow addition of zinc nitrate and sodium hydroxide stocksolutions at 60 ◦C in pure DI water (Sample 1) resulted in theformation of nonuniform agglomerated particles, which settledas soon as the agitation was stopped (Fig. 1a). The addition ofDaxad 11G in the reactor changed the shape of the particlesformed (Fig. 1b) but did not improve the stability of the disper-sion (Sample 2). In contrast, in the case of Sample 3, the sameamount of Arabic gum (AG) yielded a stable dispersion of uni-form particles having an average size of ∼260 nm (Fig. 1c). Inall three cases the final solids displayed XRD patterns charac-teristic for crystalline zinc oxide (Fig. 2). Surprisingly, despitethe fact that large ‘crystal-like’ particles were obtained in purewater, the crystallinity of the solids formed in the presence ofdispersants 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 1Experimental conditions and properties of precipitated ZnO particles

Sample Zinc salt Zn2+ conc.(mol cm−1)

Reactants flow(mL min−1)

Reactiontime (h)

Dispersant T

(◦C)Dispersionstabilitya

Particleuniformityb

Particlesize (nm)

Crystallitesize (nm)

Particle shape

1 Zn(NO3)2 0.5 0.28 6 None 60 U P Broad 21 Irregular crystals2 Zn(NO3)2 0.5 0.28 6 Daxad 60 U P Broad 33 Ellipsoidal3 Zn(NO3)2 0.5 0.28 6 AG 60 S VG ∼260 44 Hexagonal prisms4 Zn(NO3)2 0.5 0.28 6 AG 20 S G ∼300 10 Spheroidal5 Zn(NO3)2 0.5 1.12 1.5 AG 60 S G ∼100 28 Spheroidal6 ZnSO4 0.5 0.28 6 AG 60 S VG ∼270 39 Hexagonal prisms7 ZnAc2 0.5 0.28 6 AG 60 S P 20–300 23 Irregular8 Zn(NO3)2 0.75 0.28 6 AG 60 S G ∼300 43 Hexagonal prisms

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 crystal-lite size from ∼21 nm in absence of dispersant to ∼33 nm forDaxad 11G and respectively ∼44 nm in the case of Arabic gum.The high degree of dispersion of the solids obtained with Ara-bic gum was confirmed by the excellent agreement between themean particle size measured by laser diffraction (Fig. 3) andthe average size obtained from the electron micrograph shownin Fig. 1c.

Because of the remarkable stability of the dispersions pre-pared with Arabic gum and the uniformity of the ZnO particlesobtained, most of this study was focused on understanding theeffect of different experimental conditions on this particularsystem.

When the experiment with Arabic gum (Sample 3) was car-ried out at 20 ◦C instead of 60 ◦C, a very stable dispersion ofZnO was also obtained. While the decrease in the tempera-ture caused only a slight increase in the average particle sizeto ∼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 nanoparti-cles (Fig. 4b).

The dramatic impact of the temperature on the crystallinestructure was confirmed by the much broader and poorly sep-arated XRD peaks obtained in the case of the solids pre-pared at 20 ◦C (Fig. 5a) and the smaller calculated crystal-lite size (∼10 nm), which was in good agreement with thevalue estimated based on the high resolution electron micro-graph shown in Fig. 5b. Particles with similar shape and inter-nal structure were obtained at the higher temperature (60 ◦C)as well when the addition rate was increased from 0.28 to1.1 cm3 min−1 (Fig. 6). They were, however, much smaller insize (∼100 nm) and the constituent primary ZnO nanoparticleswere larger (∼28 nm).

Changing the nature of the zinc salt in the experimental con-ditions used for Sample 3 affected both the uniformity and theshape of the ZnO particles. When zinc sulfate was used insteadof zinc nitrate (Sample 6), the size of the primary crystallitesand final particles as well as their general shape were preservedbut their aspect ratio changed due to a lower height of thehexagonal prisms (Fig. 7a). In contrast, with zinc acetate (Sam-ple 7) the zinc oxide particles obtained had irregular shape anda broad size distribution (Fig. 7b), while the crystallite size wasconsiderably smaller (∼23 nm).

Fig. 5. (a) XRD patterns of ZnO powders obtained at 20 ◦C and (b) high resolu-tion FESEM image showing the fine structure of the polycrystalline aggregatedparticles.

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

The increase of the concentration of the zinc solutionfrom 0.5 to 0.75 mol dm−3 and NaOH solution from 0.75 to1.12 mol dm−3 in otherwise the same conditions (Sample 8)did not change significantly the size of the final ZnO parti-cles but did affect slightly their morphology. The hexagonalprisms obtained in this case had roughly the same diameter asat lower concentration but were more elongated and occasion-ally ‘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 ◦Cin the presence of Arabic gum as dispersant with (a) zinc sulfate and (b) zincacetate.

4. Discussion

Unlike the case of most polyvalent cations, the hydrolysisof zinc ions follows a rather simple path. The reaction of Zn2+with OH− ions results first in the formation of Zn(OH)+ speciesfollowed by the precipitation of zinc hydroxide, which easilyredissolves in excess of base to form a soluble complex. Sim-ple calculations based on the values of the solubility product forZn(OH)2 and the stability constant of Zn(OH)2−

4 species, showthat the complete precipitation of the zinc ions (i.e., a resid-ual Zn2+ concentration of less than 10−6 mol dm3), can beachieved only if the pH is higher than 8.1 but not above 10.5.The nature of the solids formed, however, depends significantlyon how the precipitation process is conducted. If, for exam-ple, a strong base is added to a solution of zinc salt, the highconcentration of hydroxyl ions at the contact point between thereactants favors the rapid formation of a white gelatinous zinchydroxide, which through a ‘sol–gel’ transformation changesrapidly into crystalline Zn(OH)2, most often the orthorhom-bic wulfingite. The latter is rather stable and can be furtherdehydrated and converted into ZnO only as a result of a sub-sequent aging, which can be accelerated at elevated tempera-tures [34]. In contrast, as previously shown by Matijevic andothers [23,24], if stoichiometric amounts of zinc salt and baseare added in parallel (‘double-jet’) and slowly into the vigor-ously mixed solvent, crystalline ZnO can be directly precipi-tated.

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

The time dependent investigations conducted in this studyhave revealed that in such ‘double jet’ configuration zinc hy-droxide is still formed early in the reaction but it is completelyconverted later in the process into crystalline ZnO. Indeed, theXRD data in Fig. 9 clearly show that, at both 20 and 60 ◦C,after the first hour the peaks of both ε-zinc hydroxide (wulfin-gite) and zinc oxide are present, while after two hours onlythe latter can be detected. As expected, at the higher temper-ature the conversion of the hydroxide into ZnO is acceleratedas indicated by the more intense peaks of the oxide (Fig. 9b).Interestingly enough, after the complete conversion of the hy-droxide into crystalline oxide has occurred, the former couldnot be detected despite the fact that the Zn2+ and the OH− ionswere continuously added at the same rate in the reactor.

The instability of the initially formed hydroxide and thedirect formation of crystalline ZnO in the late stages of theprecipitation are quite surprising findings, particularly for theexperiment conducted at room temperature where the dehydra-tion and aging of the hydroxide are assumed to be generallyslow. It appears that keeping at all times the concentrations ofZn2+ and OH− ions at or below the solubility limit (as is thesituation in the ‘double jet’ process) favors the formation ofZn(OH)+ or a more reactive form of zinc hydroxide, which canbe rather easily converted into crystalline ZnO. The fact thatthe hydroxide did not form at all during the last stages of theprecipitation is even more puzzling. One possible explanationcould be that the surface of the crystalline zinc oxide formed

M. Jitianu, D.V. Goia / Journal of Colloid and Interface Science 309 (2007) 78–85 83

(a)

(b)

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

acts as a template which either accelerates the dehydration ofthe hydroxide or completely bypasses its formation by favor-ing 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 nano-size primary particles. Such outcome is not at all surprising,this formation mechanism being very often observed in many

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 precipitationconducted at 20 ◦C.

inorganic systems such as metals [35], metal sulfides [36], ox-ides [37], as well as organic ones [38]. A model for the spon-taneous aggregation of nanosize precursors, an event triggeredmost often by the decrease in the thickness of the double layerat high ionic strength or the reduction of the surface chargenear the isoelectric point (IP), was described previously in de-tail for the formation of monodispersed spherical gold [39] andCdS [40]. According to this model, the essential condition forthe size selection to occur is that the generation rate of primaryparticles must decay during the process.

In the system investigated here, the aggregation mechanismwas particularly obvious when the precipitation was conductedslowly at 20 ◦C, as confirmed by the electron micrograph imagein Figs. 4b and 5b. The time dependent studies showed that theaggregation of the primary nanoparticles occurred very early inthe process, as testified by the high resolution electron micro-graph of the solids collected after the first hour (Fig. 10a), andthat the aggregates continued to grow during the remaining 5 hby the attachment of newly formed primary particles. The at-tachment appeared to be random during most of the process,except for the end of the precipitation when the precursorsshowed a tendency to rearrange into organized layered struc-tures (Fig. 10b). While the rearrangement of the constituentnanosize precursors was too slow at 20 ◦C to affect the shapeof the final ZnO particles during the 6 h long reaction, at 60 ◦Cthe process was significantly faster. In this case, the orderedstructures 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 theend of the 6-h process.

∼24 nm) were observed after only 1 h (Figs. 11a and 11b) andafter 3 h the rearrangement was extensive enough to completelychange the shape of a large number of aggregates from spheri-cal to hexagonal prisms (Figs. 11c and 11d). Eventually, by theend of the precipitation all ZnO particles consisted of hexagonalprisms with relatively smooth sides and rough ‘end’ surfaces(Figs. 11e and 11f).

The rearrangement of the primary particles was found to bea time dependent process. Indeed, even at 60 ◦C, when the ad-dition 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 pri-mary particles was also affected by the nature of the anionspresent 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 awell-defined morphology, likely because their larger size di-minished the effect of the attractive forces between the primarynanoparticles.

While it can be easily rationalized why uniform sphericalentities are formed by the rapid aggregation of nanosize pri-mary particles, it is quite difficult to understand how such eventwould yield directly uniform particles with other regular geo-metric shapes. The findings of this study support a two-stepprocess for the formation of such structures. According to thisproposed mechanism, in the first step the aggregation yieldslarger polycrystalline spheres, as the process is too rapid to al-low 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 85

stituent precursors can eventually proceed leading to particlesof other regular geometric shapes. The latter process is likelyfavored in the case of solids with higher solubility in the dis-persion medium and it may or may not involve the growth ofthe primary particles. The final shape of the particles dependson many parameters including but not limited to the crystalstructure of the substance, the nature of the soluble speciespresent (ions, molecules, polymers), mixing, solvent, etc. Thistwo-step mechanism is supported by the findings of other stud-ies, which showed that polycrystalline particles with regulargeometric shapes are usually formed as a result of slow pre-cipitation processes [37].

Finally, it must be emphasized that the Arabic gum playeda key role in the precipitation process being responsible for theformation of a highly dispersed ZnO particles with controlledsize and defined shape despite the high ionic strength of thesystem. For this reason, the method described may represent acost effective synthetic route to prepare large quantities of ZnOpowders suitable for applications in electronics and sensors.The effectiveness of the dispersant was diminished however ateven higher ionic strength where a slight aggregation of the fi-nal ZnO particles occurred. Interestingly enough however, theaggregation was somewhat orderly, the individual prisms beingstacked mostly along their highest symmetry axis to form elon-gated hexagonal prisms (Fig. 8b).

5. Summary

This study describes a versatile precipitation method forpreparing uniform, highly dispersed zinc oxide particles. Byadding the zinc salt and base solutions in parallel (‘double jet’),it was possible to generate even at ambient temperatures crys-talline ZnO particles without the need for a subsequent heattreatment. It was found that the larger uniform oxide particleswere formed as a result of the aggregation of much smallernanosize primary particles. The time dependent investigationsshowed that the size, shape, and internal structure of the finalparticles were dictated by the size and the spatial arrangementof the nanosize subunits, which in turn depended on the natureof the dispersant, the type of zinc salt, the reactants additionrate, and the temperature of the process. The findings of thestudy indicated that the large ‘hexagonal prism’-like particleswere formed by a two-stage mechanism involving first the rapidaggregation of nanosize precursors into larger uniform spheresfollowed 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.

References

[1] C.A.K. Gouvea, F. Wypych, S.G. Moraes, N. Duran, N. Nagata, P. Peralta-Zamora, 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. Span-

hel, 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. Matijevic, 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. Pul-

cinelli, 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. Matijevic, Colloids Surf. A 146 (1999) 139.[36] S. Libert, D.V. Goia, E. Matijevic, Langmuir 19 (2003) 10673.[37] W.P. Hsu, L. Rönnquist, E. Matijevic, Langmuir 4 (1988) 31.[38] G. Pozarnsky, E. Matijevic, Colloids Surf. 125 (1997) 593.[39] V. Privman, D.V. Goia, J. Park, E. Matijevic, J. Colloid Interface Sci. 213

(1999) 36.[40] S. Libert, V. Gorshkov, V. Privman, D.V. Goia, E. Matijevic, 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

electrostatic potential steric potential

entropy hindrancesteric stabilization

Coulombic repulsionelectrostatic stabilization

Main types of (de)stabilization of colloidal dispersions

x r

x0

van der Waals attractive interactions; attraction potential : VA

attraction potential between spherical particles:

A: Hamaker constant(attraction parameter)

x12rA(x)VA

S2A 2x12

A(x)V

attraction potential between plate-like particles:

A: Hamaker constant(attraction parameter)

Hamaker constants of various materials

Hamaker constants of various materials

23311133311313131 AAA2AAAA

33223311132 AAAAA

232131132 A AA

221112 A AA

Combination of Hamaker constants

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

A12: particle(1)-particle(2)interaction in vacuum

The stability of a colloidal disperse systemis 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 theconstituting 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 dispersionforces between the constituting species (dispersion medium; dispersed particles).

van der Waals attractions in colloidal disperse systems

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

VA, v

an d

erW

aals

inte

ract

ion

ener

gy(k

T)

two sphericalparticles, R=4μm

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

DLVO theory:Electrostatic stability

Aggregation is hindered by electrostatic repulsion

van der Waals attraction

electrostatic repulsion

Derjagin Landau Verwey Overbeek

1. …Stern potential St

2. …thickness of the electric double layer, -1

The electrostatic repulsion depends on the...

DLVO theory: the theory of electrostatic stabilization(Derjagin-Landau and Verwey-Overbeek)

xR ekTn64V

2

zkTze

zkTze

x22

r22

R

1e

1eeze

Tk8V

Electric double layer interaction energies;repulsion energies VR

between two sheets (lamellae):

between two spheres:

2A x12AV

x12rAVA

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

x

attraction potential(van der Waals forces)

0 x

DLVO theory

0 x

repulsion potential(Coulombic repulsion;

aqueous medium, large )

0 x

(VT/kT) 10:“colloid stabiliy”

DerjaginLandauVerwey

Overbeek

VR

kT

VA

kT

VT

kTtotal

potential

VT = VA + VR

DLVO theory: conditions for colloid stability

attractive forcesoverwhelm

repulsive forces

repulsive forcesoverwhelm

attractive forces

CoagulationStability

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

aqueous As2S3 sol with increasing background electrolyte concentraion

(1:1 electrolyte, mM)

aqueous Al2O3 suspensionat different pHs

repulsion

total

attraction

pote

ntia

l

particle-particle distance(surface separation)

0

unstable

stable emulsion stable suspension

unstable

(coagulation / flocculation;sedimentation)

(coagulation; creaming and coalescence)

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

Schulze-Hardy rule: 6z1ccc

666 31:

21:

11

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

AlCl3

CaCl2

MgCl2

KCl

NaClelectrolyte

1,3 (1,8)9,3 10-5

16 (13)6,5 10-4

16 (14)7,2 10-4

1000 (980)5,0 10-2

10005,1 10-2

Schulze-Hardy-ruleccc (M) (As2S3 dispersion)

As2S3 sol in 1:1 electrolyte

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

c = 1 mM

23 mM4 mM

90 mM

360 mM1500 mM

ccc = 65 mM

inte

ract

ion

ener

gy/ 1

0-19

J

c increases

ccc

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

ccc

Mechanisms of coagulation

perikinetic orthokinetic

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

differentialsettling

The kinetics of coagulation

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

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

tnD)xr2(41n

no

o

x r

tnD)xr2(4W1n

no

o

2/1

2/1

f,ts,t

skfkWwhere

( t12: half life time )

Colloid stability: thermodynamic and kinetic aspects

W 1: stability ratio (Fuchs)

turbidity measurements

kTV

expr2

1xxd

kT)x(Vexpr2W max

r22

2/1

2/1

f,ts,t

skfkW

2:2 electrolyte

1:1 electrolyte

log c

log

W

clogbalogW

W

fast coagulationregion

Summary of main types of particle-particle interactions

x

(x (x < 2< 2 ), e.g.:), e.g.:

(x (x > > ) )

interactions.interactions.

Colloid stability requires repulsive forces between colliding particles

electrostatic potential steric potential

entropy hindrancesteric stabilization

Coulombic repulsionelectrostatic stabilization

Main types of (de)stabilization of colloidal dispersions

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

MW1

MW2

MW3

MW3 > MW2 > MW1

Typical Polymer Adsorption Isothermsthe effect of molecular weight

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0 100 200 300 400 500 600

c (mg/L)

Am

ount

adso

rbed

,(m

g/m

2 )

MW1

MW2

MW3

MW3 > MW2 > MW1

cKc

maxLangmuir isotherm:

(thicknessof the electric double layer)

(end-to-enddistance polymer

layer thickness)

--110021,000310610,000101

20100,000300.1601,000,0001000.01

ho[nm]

polymerM [g.mol-1]

1/[nm]

1:1 electrolytec [mmol.dm-3]

Typical thickness of the adsorption layer in case ofelectrostatic stabilization and steric stabilization

B

C

Polymers at interfaces

cpolymer

Conditions for efficient steric stabilization(1) large (point C in the Figure)(2) large (layer thickness)(3) large s (adsorption energy)(4) < 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

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

flocculation

bridging

StericSteric interactioninteraction

r

x = 2x = 2

x/2x/2

xx

VT

x

Typical potential function of steric stabilization

VSsteric

repulsion

VAvan der Waals

attraction

~ coil diameter, rg

electrolytes polymers

Brownian movement/Brownian movement/agitationagitation

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: Continuous phase

Solid Liquid Gas

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

Sol (e.g., molten chocolate)

Smoke

Liquid Emulsion (e.g., cream) Aerosol (e.g., spray)

Dis

pers

ed p

hase

Gas Solid foam (e.g., whipped candy)

Foam (e.g., beer)

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 air-solution 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:

rPL

γ2=

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).

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 sthey disrupt the streamlines as they force liquid to flow around them. This increaamount of energy needed to achieve a givevelocity gradient and increases the visAnother, more systematic, way of lookinthis is to imagine the particles as blockisome streamlines. To achieve the same overall velocity gradient, the velocity gradover 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 rem

tream

ses the n

cosity. g at

ng

ient

aining liquid parts of the system.

owever you conceptualize the mechanism of viscosity-building by particles the

η=η0 (1+2.5φ)

here η is the viscosity of the dispersion, η0 is the viscosity of the continuous phase (i.e.,

reaming (Sedimentation). The effect of particles on a flowing liquid are the basis of

he

he tendency of a particle to float is the buoyancy – a product of the amount of material

Hmagnitude of the effect can be described by the Einstein equation:

wno 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. Cthe 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 affectedmore 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). Tspeed of the droplet as it floats upwards is retarded by the friction it experiences with the continuous phase. Ttrying to float (i.e., particle volume) and the density difference between phases (∆ρ) and the acceleration due to gravity (g):

ρπ∆=

6

3gdFb

he frictional force opposing a spherical particle as it moves upwards is proportional to

F

Tits surface area, its speed and the viscosity of the continuous phase (ηc).

f rvcπη3=

droplet will float upwards at ever increasing speeds in response to the buoyancy force A

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):

cs

gdvηρ

18

2∆=

he Stokes-Einstein equation provides an estimate of the speed of a creaming particle.

e ion:

•

• Concentrated foams

article

n.

TThe lower vs, the more stable the emulsion will be to creaming. Stokes-Einstein is rarelyquantitatively applicable but it gives some ideas to reduce creaming rate (increase continuous phase viscosity, reduce particle size, increase oil density).

oams. Foams can be formed either by whipping a gas into a liquid or by bubble Fnucleation (e.g., from yeast cells or from a supersaturated CO2 solution). Foams arsimilar in many ways to emulsions but have some distinctive feature worth considerat

• 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 byOstwald 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. (e.g., meringue) break down by (i) drainage and (ii) film rupture.

Aggregation. The mechanisms of emulsion stability considered so far (i.e., Pcreaming – very important especially for large droplets, and Ostwald ripening – rarelyimportant unless the oil has significant water-solubility) require no droplet-droplet contact. Droplet-droplet collisions and interactions are essential for droplet aggregatioThere 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:

η34kTk fast =

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 transient-dipole 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 energy cost (or gain) to bring one a particle from infinite separation to a distance h from a second similar particle. The Van der Waals function is negative because it is attractive (so ∆G<0 to bring particles closer). While weak, Van der Waals force can attract particles over several nanometers.

where r is the particle radius and h the particle separation and A is the Hamaker constant.

hArVA 12

−=

Electrostatic interactions are repulsive between similarly charged particles. Most particles are charged because the proteins adsorbed to their surface The effective magnitude of the repulsive force depends on the surface charge (i.e., potential). It decays in magnitude exponentially moving away from the charge and the rate of decay is greater in higher ionic strength systems. The formulation commonly used is:

)1ln(2 khE ekrV −+= ψ

where ψ is the surface potential, r is the droplets radius, h is the particle separation, k is a constant and κ is the reciprocal Debye length – a constant whose magnitude increases with ionic strength (concentration of ionic charge in the aqueous phase).

Note the potential is positive – it costs energy to bring one particles closer to a similarly-charged particle.

According the DVLO theory, the net potential between droplets is given by the sum of the Van der Waals attractive forces and electrostatic forces. The sof forces is attractive at someseparations and repulsiothers. It provides the interdroplet pair potential. A pair potential represents an energy surface a particle mustmove across to approach andcollide with a second. If the potential is repulsive at all separations, collisions will be difficult. If the potential is attractive (negative) at all separations the particles will deviate from their trajectories towards one another and collide. If there is an energy barrier (a positive slope) in thepotential it will provide the energy barrier needed to calculate the collision efficiency (w). For example, the Figure shows the effect of increasing ionic strength (and decreasing the range of electrostatic repulsion) on the DVLO potential. At long ranges the potential is attractive and droplets will approach. However, at shorter ranges the electrostatic repulsion will dominate and create a repulsive barrier, that if large enough (>2-3 kT) wiprevent further droplet approach (some ionic strengths give a small secondary minimthat may hold particles at a finite separation in loose flocs). Beyond the repulsive energybarrier the Van der Waals attraction again dominates and droplets that can overcome thebarrier will rapidly coalesce in the energy well.

um

ve at

ll a

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.

1

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 (∆G<0), on the contrary, in order to increase its surface a certain amount of energy is needed, as the process is, per se, non-spontaneous (∆G>0). 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.

Surface Tension Ishita Patel, M.S.

2

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

Surface Tension Ishita Patel, M.S.

3

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

Surface Tension Ishita Patel, M.S.

4

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

Surface Tension Ishita Patel, M.S.

5

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 # Concentration 1 C 2 C/2 3 C/4 4 C/8 5 C/16 6 C/32

Surface Tension Ishita Patel, M.S.

6

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. Temperature 0C 1 10 2 20 3 30 4 40 5 50 6 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).

Surface Tension Ishita Patel, M.S.

Advanced Drug Delivery Reviews 63 (2011) 456–469

Contents lists available at ScienceDirect

Advanced Drug Delivery Reviews

j ourna l homepage: www.e lsev ie r.com/ locate /addr

Physical and chemical stability of drug nanoparticles☆

Libo Wu, Jian Zhang, Wiwik Watanabe ⁎MAP Pharmaceuticals, Inc. 2400 Bayshore Parkway, Mountain View, CA 94043, USA

☆ This review is part of the Advanced Drug Deliv“Nanodrug Particles and Nanoformulations for Drug De⁎ Corresponding author. Tel.: +1 650 386 8193; fax:

E-mail address: wwatanabe@mappharma.com (W. W

0169-409X/$ – see front matter. Published by Elsevierdoi:10.1016/j.addr.2011.02.001

a b s t r a c t

a r t i c l e i n f o

Article history:Received 12 August 2010Accepted 2 February 2011Available online 21 February 2011

Keywords:Drug nanoparticlesNanosuspensionsPhysical stabilityChemical stabilityStabilizer

As nano-sizing is becoming a more common approach for pharmaceutical product development, researchersare taking advantage of the unique inherent properties of nanoparticles for a wide variety of applications. Thisarticle reviews the physical and chemical stability of drug nanoparticles, including their mechanisms andcorresponding characterization techniques. A few common strategies to overcome stability issues are alsodiscussed.

ery Reviews theme issue onlivery”.+1 650 386 3100.atanabe).

B.V.

Published by Elsevier B.V.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4562. Stability of drug nanoparticles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457

2.1. Effect of dosage form on stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4572.2. General stability issues related to nanosuspensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457

2.2.1. Sedimentation or creaming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4572.2.2. Agglomeration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4582.2.3. Crystal growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4632.2.4. Change of crystalline state . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4632.2.5. Stability issues with solidification process of nanosuspensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4642.2.6. Chemical stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 464

2.3. Additional stability issues relate to large biomolecules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4643. Characterizing stability of drug nanoparticles and nanoparticle formulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465

3.1. Particle size, size distribution and morphology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4653.2. Sedimentation/creaming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4663.3. Particle surface charge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4663.4. Crystalline state . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4663.5. Chemical stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4663.6. Additional techniques for assessing large biomolecule nanoparticle and formulation stability . . . . . . . . . . . . . . . . . . . . 466

4. Recommendations of general strategies for enhancing stability of nanoparticle formulations . . . . . . . . . . . . . . . . . . . . . . . . 4675. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467

1. Introduction

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

457L. Wu et al. / Advanced Drug Delivery Reviews 63 (2011) 456–469

nanotubes [6], micelles [7], microemulsions [8] and liposomes [9]. Ingeneral, the term “nanoparticles” refers to particles with sizesbetween 1 and 100 nm. However, submicron particles are alsocommonly referred as nanoparticles in the field of pharmaceuticsand medicine [10–14]. Nanoparticles are categorized as nanocrystals[10], polymeric [15], liposomal [9] and solid lipid nanoparticles (SLN)[16] depending on their composition, function andmorphology. Giventhe extensive available literature reviews on SLN, polymeric andliposomal nanoparticles [4,9,15–18], this article will focus only onnanocrystals (pure drug nanoparticles).

The unique nano-scale structure of nanoparticles provides signif-icant increases in surface area to volume ratio which results in notablydifferent behavior, both in-vitro and in-vivo, as compared to thetraditional microparticles [10–12]. Consequently, drug nanocrystalshave been extensively used in a variety of dosage forms for differentpurposes [10,11,14,19,20], such as improving the oral bioavailabilityof poorly water-soluble drugs by utilizing enhanced solubility anddissolution rate of nanoparticles [21–23]. In the field of pulmonarydrug delivery, the nanoparticles are able to deliver the drugs into thedeep lungs, which is of great importance for systemically absorbeddrugs [11,14]. In addition, injection of poorly water-soluble nanosus-pension drugs is an emerging and rapidly growing field that hasdrawn increasing attention due to its benefits in reducing toxicity andincreasing drug efficacy through elimination of co-solvent in theformulation [10,20].

Despite the advantages of drug nanocrystals, they present variousdrawbacks including complex manufacturing [12,24–26], nanotoxi-city [27] and stability issues [10,19,20]. Stability is one of the criticalaspects in ensuring safety and efficacy of drug products. Inintravenously administered nanosuspensions, for example, formationof larger particles (N5 μm) could lead to capillary blockade andembolism [20], and thus drug particle size and size distribution needto be closely monitored during storage. The stability issues of drugnanoparticles could arise duringmanufacturing, storage and shipping.For instance, the high pressure or temperature produced duringmanufacturing can cause crystallinity change to the drug particles[12,26,28]. Storage and shipping of the drug products may also bringabout a variety of stability problems such as sedimentation,agglomeration and crystal growth [29–31]. Therefore, stability issuesassociatedwith drug nanocrystals deserve significant attention duringpharmaceutical product development. This article reviews existingliterature on drug nanoparticle stability, including theory/mecha-nisms, methods used to tackle the stability problems and character-ization techniques, and provides recommendations to improve thecurrent practices. Since the stability issues related to nanoparticle drypowders are usually trivial, this review will only focus on stability ofnanosuspensions (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 enabledtheir extensive application in various dosage forms including oral,parenteral, ocular, pulmonary, dermal and other specialized deliverysystems [10,11,13,20,32]. Although different dosage forms may sharesome common stability issues, such as sedimentation, particleagglomeration or crystal growth, their effects on drug products arequite different. For instance, particle agglomeration could be a majorissue in pulmonary drug delivery since it affects deposition amount/site, and thus drug efficacy. On the other hand, agglomeration inintravenous formulations can cause blood capillary blockage andobstruct blood flow. Moreover, the selection of stabilizers is alsoclosely related to dispersion medium, dosage form and strictlygoverned by FDA regulations. To date, there is a wide variety of

choices on the approved stabilizers for oral dosage form whereas theexcipients allowed for inhalation are very limited [33].

Drug nanoparticles exist in the final drug products either in drypowder or suspension form. Examples of the dry powder form includethe dry powder inhaler, lyophilized powder for injection and oraltablets or capsules. Solid dosage forms usually have good storagestability profiles, which is why a common strategy to enhancenanosuspension stability is to transform the suspension into solidform [19,25]. Most of the reported stability concerns arise fromnanosuspensions in which the drug nanoparticles are dispersed in amedium with or without stabilizers. In addition, mechanismsinvolved in the stability of small and large biomolecule formulationsare different due to their molecular structure differences. A smallmolecule drug is defined as a low molecular weight non-polymericorganic compound while large biomolecules refer to large bioactivemolecules such as protein/peptide. One of the major issues withprotein/peptide stability is to maintain the 3-dimensional molecularconformation, such as secondary and tertiary structure in order tokeep their biological activities [34,35], whereas there is no suchconcern for small organic molecules.

2.2. General stability issues related to nanosuspensions

Stability issues associated with nanosuspensions have beenwidelyinvestigated and can be categorized as physical and chemical stability.The common physical stability issues include sedimentation/cream-ing, agglomeration, crystal growth and change of crystallinity state.

2.2.1. Sedimentation or creamingDrug particles can either settle down or cream up in the

formulation medium depending on their density relative to themedium. The sedimentation rate is described by Stokes' law [36,37]which indicates the important role of particle size, medium viscosityand density difference between medium and dispersed phase indetermining the sedimentation rate. Decreasing particle size is themost common strategy used to reduce particle settling. Matching drugparticles density with medium or increasing medium viscosity are theother widely used approaches to alleviate sedimentation problems[37,38]. Fig. 1 shows different sedimentation types that occur insuspension formulations.

In a deflocculated suspension (Fig. 1a), particles settle indepen-dently as small size entities resulting in a slow sedimentation rate.However, densely packed sediment, known as caking [39], is usuallyformed due to the pressure applied on each individual particle. Thissediment is very difficult to be re-dispersed by agitation [36,37,39]and would be detrimental to the drug products performance. In theflocculated suspension (Fig. 1b), the agglomerated particles settle asloose aggregates instead of as individual particles [36,37]. The looseaggregates have a larger size compared to the single particle, and thushigher sedimentation rate. The loose structure of the rapidly settlingflocs contains a significant amount of entrapped medium and thisstructure is preserved in the sediment. The final flocculation volume istherefore relatively large and the flocs can be easily broken and re-dispersed by simple agitation. K.P. Johnston et al. [40,41] haverecently attempted to achieve stable nanosuspensions via a noveldesign of flocs structure called “open flocs”, as illustrated in Fig. 1c.Thin film freezing was used to produce BSA nanorods with aspectratio of approximately 24. These BSA nanorods were found to behighly stable when dispersed into hydrofluoroalkane (HFA) propel-lant, with no apparent sedimentation observed for 1 year. Due to thehigh aspect ratio of BSA nanorods and relatively strong attractive Vander Waals (VDW) forces at the contact sites between the particles,primary nanorods were locked together rapidly as an open structureupon addition of HFA, inhibiting collapse of the flocs [41]. The low-density open flocs structure was then filled with liquid HFA medium,preventing particle settling. Similar results were shown using needle

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

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

and plate shaped itraconazole nanoparticles with aspect ratiosbetween 5 and 10 [40].

Although sedimentation is one of the key issues for colloidalsuspension, the reported studies examining sedimentation issues inaqueous-based nanosuspensions are very scarce. This could be due to(i) surfactants are generally used in most of the nanosuspensions toinhibit particle agglomeration in the medium, which alleviates thesedimentation issues and (ii) the small nano-sized particles signifi-cantly reduce the sedimentation rate. In addition, many of theaqueous nanosuspensions are transformed to dry solid form by spraydrying or freeze drying to circumvent the long-term sedimentationissue. Unfortunately, this solidification process cannot be applied tonon-aqueous nanosuspensions where sedimentation/creaming iscommonly present. An example to illustrate this is metered doseinhaler (MDI) formulations where the nanoparticles are suspended inHFA propellants. Sedimentation or creaming is a key aspect affectingstability of these formulations. Particle engineering to optimizeparticle surface properties and morphology, e.g. hollow porousparticles [42], and introduction of surfactant(s) is generally employedto alleviate the issue.

2.2.2. AgglomerationThe large surface area of nanoparticles creates high total surface

energy, which is thermodynamically unfavorable. Accordingly, theparticles tend to agglomerate to minimize the surface energy.Agglomeration can cause a variety of issues for nanosuspensionsincluding rapid settling/creaming, crystal growth and inconsistentdosing. The most common strategy to tackle this issue is to introducestabilizers to the formulation. In addition to safety and regulation

considerations, selection of stabilizers is based on their ability toprovide wetting to surface of the particles and offer a barrier toprevent nanoparticles from agglomeration [13,19].

There are two main mechanisms through which colloidal suspen-sions can be stabilized in both aqueous and non-aqueous medium, i.e.electrostatic repulsion and steric stabilization [10,36,37]. These twomechanisms can be achieved by adding ionic and non-ionic stabilizersinto the medium, respectively. Stabilization from electrostaticrepulsion can be described by the classic Derjaguin–Landau–Verwey–Overbeek (DLVO) theory [43,44]. This theory mainly appliesto aqueous suspension while its application in non-aqueous mediumis still not well-understood [33]. The DLVO theory assumes that theforces acting on the colloidal particles in a medium include repulsiveelectrostatic forces and attractive VDW forces. The repulsive forces areoriginated from the overlapping of electrical double layer (EDL)surrounding the particles in the medium, and thus preventingcolloidal agglomeration. The EDL consist of two layers: (i) sternlayer composed of counter ions attracted toward the particle surfaceto maintain electrical neutrality of the system and (ii) Gouy layerwhich is essentially a diffusion layer of ions (Fig. 2).

The total potential energy (VT) of particle–particle interaction is asum of repulsion potential (VR) generated from electric double layersand attraction potential (VA) from the VDW forces. VA is determinedby the Hamaker constant, particle size and inter-particulate distancewhile VR depends on particle size, distance between the particles, zetapotential, ion concentration and dielectric constant of the medium. VR

is extremely sensitive to ion concentration in the medium. As the ionstrength is increased in the medium, the thickness of EDL decreasesdue to screening of the surface charge [36,37]. This causes decrease inVR, increasing the susceptibility of the dispersed particles to formaggregates. Zeta potential (ZP) is electric potential at the shear planewhich is the boundary of the surrounding liquid layer attached to themoving particles in the medium. ZP is a key parameter widely used topredict suspension stability. The higher the ZP, the more stable thesuspension is.

In the case of steric stabilization, amphiphilic non-ionic stabilizersare usually utilized to provide steric stabilization which is dominatedby solvation effect. As the non-ionic stabilizers are introduced intonanosuspensions, they are absorbed onto the drug particles throughan anchor segment that strongly interacts with the dispersedparticles, while the other well-solvated tail segment extends intothe bulk medium (Fig. 3).

As two colloidal particles approach each other, the stabilizingsegments may interpenetrate, squeezing the bulk medium moleculesout of the inter-particulate space as illustrated in Fig. 3. Thisinterpenetration is thermodynamically disfavored when a goodsolvent is used as the bulk medium to stabilize the tail [36].Accordingly, provided that the stabilizers can be absorbed onto theparticle surface through the anchor segment, strong enthalpicinteraction (good solvation) between the solvent and the stabilizingsegment of the stabilizer is the key factor to achieve stericstabilization and prevent particles from agglomeration in the medium[36,37]. In addition to solvation, the stabilizing moiety needs to besufficiently long and dense to maintain a steric barrier that is capableof minimizing particle–particle interaction to a level that the VDWattractive forces are less than the repulsive steric forces [43–45].

The main drawback associated with the steric stabilization is theconstant need to tailor the anchoring tail according to the particulardrug of interest. Due to the lack of fundamental understanding ofinteraction between the stabilizers and dispersed nanoparticles,current surfactant screening approaches to achieve a successful stericstabilization are mostly empirical and could be very burdensome [45–49]. In addition, the solvation of the stabilizing segment is susceptibleto variation in temperature. Stabilizer concentration could also play arole in causing suspension instability by affecting the absorptionaffinity of non-ionic stabilizers to drug particles surface. Deng et al.

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 areprevailing at intermediate distances and create net repulsion between the dispersed particles, thus preventing particle agglomeration.

459L. Wu et al. / Advanced Drug Delivery Reviews 63 (2011) 456–469

[50] used Pluronic® F127 to stabilize paclitaxel nanosuspensions. Itwas reported that stabilizers had high affinity to nanocrystals surfaceat concentrations below critical micelle concentration (CMC), andincreasing concentrations above CMC caused loss of F127 affinity tothe nanocrystals and thus unstable formulation. This was becauseF127 monomers on the nanocrystals surface started to aggregate witheach other to form micelles as the concentration was increased to theCMC level, leading to a lower affinity to the drug crystals. Temperaturewas also shown to affect the stabilizer affinity to drug crystals. This isexpected since the CMC level is dependent on temperature.

It is apparent that combination of the two stabilizationmechanismscan be very beneficial in achieving a stable colloidal dispersion. Inaddition, the combination of a non-ionic stabilizer with an ionicstabilizer reduces the self repulsion between the ionic surfactantmolecules, leading to closer packing of the stabilizermolecules [10,51].

Fig. 3. Steric stabilization mechanisms according to Gibbs free energy: ΔG=ΔH−TΔS. A posithe medium is a good solvent for the stabilizing moiety, the adsorbed stabilizing layers on treduces the number of configurations available to the adsorbed stabilizing tails, resulting in ais a poor solvent, the adsorbed layers on the particles may interpenetrate thermodynamica

Besides the steric and electrostatic stabilizationmechanisms, someother stabilization mechanisms have also been reported. Makhlof etal. produced indomethacin (IMC) nanocrystals using the emulsionsolvent diffusion technique [52]. The nanoparticles were stabilizedusing various cyclodextrins (CyDs) without adding any surfactants.The stabilizing effect was attributed to the formation of a CyD networkin the aqueous medium via intermolecular interaction of CyDmolecules. The network-like structure was believed to preventaggregation and crystal growth of IMC nanoparticles initiallyproduced from the solvent diffusion process. Similar stabilizationmechanism was also observed in another study where budesonidemicrosuspension was stabilized with hydroxypropyl-beta-cyclodex-trin in HFA medium [53]. Another approach to enhance suspensionstability that has increasingly been utilized is engineering of particlemorphology. One breakthrough in this area was the porous particle

tive ΔG indicates stable suspension while negative ΔG induces particle agglomeration. Ifhe dispersed particles cannot interpenetrate each other when the particles collide. Thisnegative entropy change and positive ΔG. On the other hand, if the dispersion mediumlly 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 porousparticles include hollow porous particle [42] and porous nanoparticle-aggregate particles (PNAPs) [14]. Unfortunately, most of the work hasbeen focused on microsuspension or polymeric colloidal formulationsand has not been applied to pure drug nanoparticles.

Table 1 summarizes a few published studies on pharmaceuticalnanosuspensions. Due to the vast amount of literature work on thepharmaceutical nanosuspensions, this review will focus only on thestudies that provide a more profound enlightenment on the stabilizerselection for nanosuspensions. The summary table shows that most ofnanosuspensions were generated in aqueous medium, with only alimited number of nanosuspensions made in non-aqueous environ-ment. The commonlyused ionic stabilizers in aqueousmedium includesodium dodecyl sulfate (SDS), sodium lauryl sulfate (SLS), lecithin anddocusate sodium. The non-ionic surfactants used in aqueous mediumare usually selected from Pluronic® surfactants, Tween 80, polyeth-ylene 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-termstorage stability for nanosuspensions, but also to achieve successfulformation and stabilization of nanocrystals during particle produc-tion. Lee et al. designed and synthesized various amino acidcopolymers containing lysine as the hydrophilic segments withalanine, phenylalanine or leucine as hydrophobic moieties [49]. Wetcomminution was used to produce naproxen nanosuspensions inpresence of HPC and amino acid copolymers. Lysine copolymer withalanine was unable to produce submicron particles while the othercopolymers with phenylalanine and leucine were capable of formingthe nanoparticles. The size of nanocrystals was proven to be constantover 1 month storage and the crystallinity was also shown to bepreserved after the wet comminution process. Furthermore, hydro-phobicity of the copolymers was identified as the key factor inachieving the stable nanosuspensions, attributed to strong polymeradsorption onto the hydrophobic drug surfaces. Although this workdid not provide an in-depth discussion on how the copolymersinteractedwith the drug nanoparticles, it illustrated the importance ofcareful selection of the anchor group (that is attached to the drugsurface) in facilitating the production of a stable nanosuspesion. In thesubsequent study [45], they attempted to understand the nature ofinteractions between polymeric stabilizers and drugs with differentsurface energies. Nanocrystals of sevenmodel drugswith PVP K30 andHPC as stabilizers were generated using wet comminution. It wasexpected that a close match of surface energy between the stabilizersand drug crystals would promote the absorption of stabilizers ontodrug particles, and thus help in reducing the particle size during thewet comminution process. Although surface energy did not seem tocorrelate well with particle size for HPC stabilized system, some trendwas 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 andbenzethoinum chloride) and eleven model drugs was conducted bythe same group in order to provide more understanding on thestabilization mechanism [48]. Again, the general trend betweensurface energy and particle size reduction was not observed in thiswork. PEG was unsuccessful in reducing the particle size of most drugcandidates while the other non-ionic stabilizers proved to be effectivein reducing the size of five drug candidates that had similar surfaceenergies to the stabilizers. F68 was shown to be the most effectivestabilizer (successfully stabilizing nine drug candidates), which couldbe due to its strong chain adsorption onto the drug crystals throughthe hydrophobic polypropylene glycol (PPG) units. F127 was found tobe less efficient than F68 likely because the short processing time ledto inefficient physical adsorption of higher molecular weight F127 tothe drug surface. This study demonstrated that a combination of ionicand non-ionic stabilizers is not always beneficial to enhance

stabilization, A few combinations of SDS or benzethoinum chloridewith various non-ionic stabilizers resulted in positive stability effectswhile the others did not. The effects of physicochemical properties ofthe drugs on the stabilization were also explored in this study. Ingeneral, drugs with lower aqueous solubility, highermolecular weightand higher melting point were shown to have higher chance forsuccessful nanosuspension formation.

Van Eerdenbrugh et al. conducted an expanded study using 13stabilizers 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 polysaccharidebased stabilizers [HPMC,methylcellulose (MC), hydroxyethylcellulose(HEC), HPC, carboxymethylcellulose sodium (NaCMC), alginic acidsodium (NaAlg)] was limited by the high viscosity of these polymericstabilizer solutions. Increasing concentration of these stabilizers didnot appear to be helpful. In contrast, the other stabilizers [PVP K30,PVP K90, PVA, Pluronic® F68, polyvinyl alcohol–polyethylene glycolgraft copolymer (K-IR), Tween 80 and D-α-tocopherol polyethyleneglycol 1000 succinate (TPGS)] did not encounter the viscosity issue.PVA was ineffective in producing the nanosuspension and the successprobability of PVP K30, PVP K90, F68 and K-IR is highly dependent ontheir concentration. Higher concentrations (25 wt.% and 100 wt.%)increased the stabilizing efficacy significantly. Tween 80 andTGPS were proven to be the most effective stabilizers. Addition ofTGPS (at concentrations N25 wt.%) allowed nanosuspension forma-tion for all tested drug compounds. No correlation was observedbetween drug physicochemical properties (molecular weight,melting point, log p, solubility and density) and nanosuspensionformation success rate. It was demonstrated that surface hydropho-bicity of the drug candidates was the driving force for nanoparticlesagglomeration, thus lowering the success rate of nanosuspensionproduction.

Mishra et al. explored nanosuspension stability issues during bothproduction and storage [29]. Hesperetin nanosuspensions wereproduced using HPH with Pluronic® F68, alkyl polyglycoside(Plantacare 2000) and inulin lauryl carbamate (Inutec SP1), orTween 80 as stabilizers. It was demonstrated that all stabilizerswere suitable for successful production of hesperetin nanosuspen-sions. The size of nanocrystals was dependent on power densityapplied in the homogenization process and the hardness of thecrystals. The effect of stabilizers on the particle size was negligible.Short-term stability over a period of 30 days was examined in order toevaluate the stabilizer efficiency. ZP was measured as a key parameterto predict the stability. In distilled water, the ZP values of all thenanosuspensions fell between −30 and −50 mV and the valuesdropped significantly in the original dispersion medium. This can beexplained by the fact that adsorbed layers of large molecules shiftedthe shear plane to a longer distance from the particle surface, thusreducing the measured value of zeta potential (Fig. 4). However, thelow ZP value does not point to an unstable suspension in this case,which could be due to the additional presence of steric stabilizationmechanism. Both Inutec and Plantacare stabilized nanosuspensionsalso showed significant reduction of ZP measured from water todispersion medium, indicating a thick absorbed steric layer and goodstability. F68 exhibited only slight decrease in ZP, indicating arelatively thin stabilization layer. The ZP value of Tween 80 wasonly −13 mv in the dispersion medium, pointing to a potentiallyproblematic stabilization. The study demonstrated that zeta potentialmeasurement is a good predictor for storage stability. Nanosuspen-sions stabilized by Inutec and Plantacare were stable at all storageconditions (4, 25 and 40 °C) up to 30 days while F68 stabilizednanosuspensions were shown to be less stable. The Tween 80formulation stability was the poorest. Pardeike et al. [30] conducteda similar study using phospholipase A2 inhibitor PX-18 nanosuspen-sions produced by HPHwith Tween 80 as stabilizer. In this work, ZP ofthe homogenized nanosuspensions was dropped from −50 mV to

Table 1Literature summary of pharmaceutical nanosuspensions.

Nanoparticles compound Manufacturingtechnique

Deliveryroute

Dispersionmedium

Stabilizers Reference

Oridonin HPH NA Water PVP K25, Brij 78, SDS, Pluronic® F68, lecithin Gao et al. (2007) [55]Oridonin HPH IV Water Pluronic® F68, lecithin Gao et al. (2008) [56]Budesonide HPH Inhalation Water Lecithin, Span 85, tyloxapol, cetyl alcohol Jacobs et al. (2002) [57]Buparvaquone HPH Inhalation Water Pluronic® F68 and PVA Hernadez-Trejo et al.(2005) [58]Buparvaquone HPH Oral Water Pluronic® F68 and lecithin Jacobs et al. (2002) [59]Diclofenac acid HPH Oral Water Pluronic® F68 Lai et al. (2009) [60]Azothromycin HPH NA Water Lecithin, Pluronic® F68, Tween 80 Zhang et al. (2007) [61]Rutin HPH Oral Water SDS Mauludin et al. (2009) [62]Rutin HPH Oral Water SDS, Tween 80, Pluronic® F68, PVA Mauludin et al. (2009) [63]Tarazepide HPH NA Water Tween 80, Pluronic® F68 Jacobs et al. (2000) [64]Omeprazole HPH IV Water Pluronic® F68 Moschwitzer, (2004) [65]Amphotericin B HPH Oral Water Tween 80, Pluronic® F68 Kayser et al. (2003) [22]Nimodipine HPH IV Water Pluronic® F68, sodium cholic acid andmannitol Xiong et al., (2008) [66]Albendazole HPH Oral Water SLS, Carbopol, PS 80, hpmc Kumar et al. (2008) [23]RMKP 22 HPH NA Water Phospholipon 90 Peters et al. (1999) [67]Hesperetin HPH Dermal Water Pluronic® F68, Inutec SP1, Tween 80 and

Plantacare 2000Mishra et al. (2009) [29]

Hydrocortisone, prednisoloneand dexamethasone

HPH Opthalmic Water Pluronic® F68 Kassem et al. (2007) [68]

Ascorbyl palmitate HPH NA Water SDS, Tween 80 Teeranachaideekul et (2008) [69]RMKK99 HPH NA Water Potassium oleate, Tween 80 Krause et al. (2001) [70]Nifedipine HPH NA Water HPMC Hecq et al. (2005) [71]Undisclosed HPH Oral Water SLS, HPMC, PVA, Acaciae Gum, Pluronic® F127 Hecq et al. (2006) [72]Hydroxycamptothecin HPH NA Water Lipoid S75, Pluronic® F68, Solutol® HS 15 Zhao et al. (2010) [73]Asulacrine HPH IV Water Pluronic® F68 Ganta et al. (2009) [74]RMKP 22 HPH NA Water Tween 80 Muller et al. (1998) [75]RMKP 22 HPH NA Water Tween 80, Glycerol Grau et al. (2000) [76]PX-18 HPH NA Water Tween 80 Pardeike et al. (2010) [30]PX-18 HPH NA Water Tween 80 Wang et al. (2010) [77]Silybin HPH Oral, IV Water Lecithin, Poloxamer 188 Wang et al. (2010) [78]Tarazepide HPH IV Water Pluronic® F68, Tween 80, Glycerol Jacobs et al. (2000) [64]Omeprazole, albendazoleand danazol

Wet milling Oral Water Pluronic® F108, F68 Tanaka et al. (2009) [79]

Fluticasone, budesonide Wet milling Inhalation Water Tween 80 Yang et al. (2008) [80]Naproxen Wet milling NA Water HPC, arginie hydrochloride Ain-Ai et al. (2008) [81]Loviride Wet milling NA Water Tween 80, Pluronic® F68 Van Eerdenbrugh et al. (2007) [82]Nine different compounds Wet milling NA Water 13 different stabilizers Van Eerdenbrugh et al. (2009) [47]Zinc Insulin Wet milling NA Water Pluronic® F68, sodium deoxycholate Merisko-Liversidge et al. (2004) [83]Ethyl Diatrizoate Wet milling NA Water Poloxamine 908 Na et al. (1999) [84]Cinnarizine, itraconazoleand phenylbutazone

Wet milling NA Water TPGS 1000 Van Eerdenbrugh et al. (2008) [85]

Nine different compounds Wet milling NA Water TPGS 1000 Van Eerdenbrugh et al. (2008) [86]Beclomethasone dipropionate Wet milling Inhalation Water PVA Wiedmann et al. (1997) [87]Rilpivirine Wet milling Parenteral Water Pluronic® F108, TPGS 1000 Baert et al. (2009) [88]Undisclosed Wet milling NA Water Plasdone S-630, docusate sodium Deng et al. (2008) [89]Piposulfan, etoposide,camptothecin, paclitaxel

Wet milling NA Water Tween 80, Span 80, Pluronic® F108, F127 Merisko-Liversidge et al. (1996) [90]

Naproxen Wet comminution NA Water Copolymers of amino acids Lee et al. (2005) [49]Seven different compounds Wet comminution NA Water HPC, PVP Choi et al. (2005) [45]Eleven different compounds Wet comminution NA Water HPC, PVP, PEG, SDS, Pluronic® F68, F127,

benzethonium chlorideLee et al. (2008) [48]

Dihydroartemisinin Vibrational rod milling NA Water PVP K30, sodium deoxycholate Chingunpitak et al.(2008) [91]Probucol Vibrational rod milling NA Water PVP, SDS Pongpeerapat et al. (2008) [92]Ibuprofen Precipitation,

microfluidizationNA Water SLS, PVP K30, Pluronic® F68, F127,

Tween 80, HPMCVerma et al. (2009) [31]

Hydrocortisone Precipitation,microfluidization

NA Water PVP, HPMC, SLS Ali et al. (2009) [93]

Ibuprofen Solvent diffusion, meltemulsification

NA Water PVA, PVP K25, Pluronic® F68, Tween 80, Kocbek et al. (2006) [94]

Alendronate-gallium,alendronate-gadolinium

Complex precipitation NA Water None Epstein et al. (2007) [95]

Paclitaxel Stabilization ofnanocrystal (SNC)

NA Water Pluronic® F127 Deng et al. (2010) [50]

Felodipine Antisolvent precipitation& Wet-milling

NA Water PVP K30, SDS, docusate sodium Lindfors (2007) [96]

Naproxen Antisolvent precipitation Oral Water PVP K15, Pluronic® F127 Chen et al. (2009) [97]Carbamazepine Antisolvent precipitation NA Water HPMC, PVP K17 Douroumis et al. (2007) [98]Cyclosporin A Antisolvent precipitation Inhalation Water Tween 80 Tam et al. (2008) [99]Undisclosed Antisolvent precipitation,

Wet millingIV, Oral Water PVP, SDS, Miglyol, docusate sodium Sigfridsson et al. (2007) [100]

β-methasone valerate-17,oxcarbazepine

Antisolvent precipitation NA Water HPMC, lipoid S75, PEG-5 soy sterol Douroumis et al. (2006) [101]

Retinoic acid Antisolvent precipitation NA Water None Zhang et al. (2006) [102]

(continued on next page)

461L. Wu et al. / Advanced Drug Delivery Reviews 63 (2011) 456–469

Table 1 (continued)

Nanoparticles compound Manufacturingtechnique

Deliveryroute

Dispersionmedium

Stabilizers Reference

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

Antisolvent precipitation NA Water None Baba et al. (2007) [103]

Nitrendipine Precipitation–ultrasonication

Oral Water PVA Xia et al. (2010) [104]

Indomethancin Emulsion diffusion NA Water Cyclodextrins Makhlof et al. (2008) [52]Celecoxib Emulsion diffusion Oral Water Tween 80, PVP K30, SDS Dolenc et al. (2009) [105]Griseofulvin Emulsion diffusion NA Water Tween 80, Oramix CG-110 Trotta el al.(2003) [106]Mitotane Emulsion diffusion NA Water Tween 80, caprylyl-capryl glucoside, lecithin Trotta el al.(2001) [107]Griseofulvin Microemulsion diffusion NA Water Lecithin Trotta el al.(2003) [108]Lysozyme Emulsification/freeze-

dryingInhalation HFA None Nyambura et al.(2009) [109]

Bovine serum albumin Thin film freezing Inhalation HFA None Engstrom et al.(2009) [41]Itraconazole Thin film freezing Inhalation HFA None Tam et al.(2010) [40]Insulin Emulsification+

freeze-dryingInhalation HFA Citral, cineole Nyambura et al.(2009) [110]

Salbutamol sulfate Microemulsion+freeze-drying

Inhalation HFA Lecithin, docusate sodium Dickinson et al. (2001) [111]

Salbutamol sulfate HPH NA Acetonitrile Tween 80 Ahmad et al. (2009) [112]Horseradish peroxidase, carbonicanhydrase, lysozyme, subtilisincarlsberg and α-chymotrypsin

Freeze-drying NA Ethyl acetate Methyl-β-cyclodextrin Montalvo et al. (2008) [113]

Diclofenac Emulsification+freeze drying

Transdermal Isopropylmyristate

Sucrose ester Piao et al. (2007) [114]

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

around −20 mV when tested from water to dispersion medium. It isgenerally believed that ZP of ±20 mV is sufficient to maintain a stableformulation with a combined electrostatic and steric stabilization[30]. The PX-18 nanosuspension was shown to be physically stable(no changes in particle size distribution) formore than half year at thestorage condition of 5 and 25 °C. However, physical instability wasobserved after 1 month storage at a higher storage temperature. Thiscould be due to the decreased dynamic viscosity and enhanceddiffusion constant at higher temperature.

There is another interesting work by Pongpeerapat et al.investigating probucol/PVP/SDS ternary ground mixture (GM) thatwas prepared with a vibrational rod mill [92]. The produced primaryprobutcol nanoparticles were around 20 nm in presence of both SDSand PVP. An interesting phenomenon was observed following thedispersion of the GM into water. For GM stabilized with PVP K17 andSDS, spherical agglomerates of primary nanocrystals were formedimmediately in the size of around 90 nm after dispersion of the GMintowater. A further agglomeration to around 160 nm in size occurredgradually during the storage stability study. In the case of PVP K12 andSDS, agglomerations of approximately 180 nm were observed after4 days of storage and then remained stable up to 84 days. Thisphenomenon is illustrated in Fig. 5. Above critical aggregationconcentration, SDS complexes with PVP to form a “necklace” structurein aqueous medium through both electrostatic and hydrophobicinteractions. Following dispersion of probucol/PVP K17/SDS into

Fig. 4. Location of shear plane in an electrostatic stabilized system (a) and in a combined stELSEVIER.

water, PVP K17/SDS “necklace” complex interacted with primarydrug nanoparticles, causing immediate agglomeration of the primarynanoparticles into 90 nm aggregates. The 160 nm secondary nano-particles were formed due to further gradual agglomeration process.The stabilization of probucol nanocrystals was attributed to formationof PVP K17/SDS layered structure on the surface of probucol. For theGM of probucol/PVP K12/SDS, agglomeration of primary drugnanoparticles occurred more rapidly because of the insufficientsurface coverage of PVP K12 and SDS on the probucol surface.Stabilization of the nanosuspension was linked to absorption of PVPK12 on the surface of probucol nanocrystals, owing to the absence oflayered structure.

Despite the proven importance of stabilizers in preventing particleagglomeration, there have been a few studies that generated stablenanosuspensions without stabilizers. Baba et al. prepared 2-devinyl-2-(1-hexyloxyethyl)pyropheophorbide (HPPH) nanosuspensionswithout any stabilizer and reported formulation stability for morethan 3 months [103]. The self-stabilization of the nanosuspensionswas attributed to a high ZP value (−40 mv) resulting from thedeprotonation of the carboxylic end group of HPPH molecules. Asimilar self-stabilized nanosuspension was reported in another studyin which amorphous all-trans retinoic acid nanoparticles were shownto be stable in aqueous medium up to 6 months. Epstein et al. [95]prepared self-suspended alendronate nanosuspensions by combiningthe negative charged alendromnic acid with gallium (Ga) or

eric-electrostatic stabilized system (b). Reprinted from Ref. [30] with permission from

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

463L. Wu et al. / Advanced Drug Delivery Reviews 63 (2011) 456–469

gadolinium (Gd) under sonication as complex nanoparticles. Thealendronate-Ga nanosuspension was shown to be stable for morethan 3 months, while the alendronate-Gd nanosuspension was stablefor only 3 days. These stability profiles correlated well with their ZPvalues (33 mV for Ga complex vs. 21 mV for Gd complex).

2.2.3. Crystal growthCrystal growth in colloidal suspensions is generally known as

Ostwald ripening and is responsible for changes in particle size andsize distribution. Oswald ripening is originated from particlessolubility dependence on their size. Small particles have highersaturation solubility than larger ones according to Ostwald–Freun-dlich equation [115], creating a drug concentration gradient betweenthe small and large particles. As a consequence, molecules diffusefrom the higher concentration surrounding small particles to areasaround larger particles with lower drug concentration. This generatessupersaturated solution around the large particles, leading to drugcrystallization onto the large particles. This diffusion process leaves anunsaturated solution surrounding the small particles, causing disso-lution of the drug molecules from the small particles into the bulkmedium. This diffusion process continues until all the small particlesare dissolved. The Ostwald ripening is essentially a process wherelarge particles grow at the expense of smaller particles [36,37], whichsubsequently leads to a shift in the particle size and size distribution ofthe colloidal suspension to a higher range. The diffusion and crystalgrowth during Ostwald ripening is shown schematically in Fig. 6.

A narrow particle size distribution can minimize the saturationsolubility difference and drug concentration gradients within themedium, and thus help to inhibit occurrence of the Ostwald ripening[37]. This can perhaps explain why Ostwald ripening is not a majorconcern for nanosuspensions with uniform particle size [10,20].Stabilizers may also alleviate Ostwald ripening as long as they do notenhance the drug solubility [116,117]. Being absorbed on the

nanoparticles surface, the stabilizers can reduce the interfacial tensionbetween the solid particles and liquid medium, and thus preventingthe Ostwald ripening. Solubility, temperature, and mechanicalagitation also affect Ostwald ripening [117]. Verma et al. producedibuprofen nanosuspensions by microfluidization milling with the aidof various stabilizers (HPMC, Pluronic® F68 & F127, Kollidon 30, SLS)[31]. The particle size obtained with microfluidization showed somecorrelation with the ibuprofen solubility in aqueous stabilizersolutions. A higher solubility of ibuprofen in the solution of SLS,Tween 80 and Pluronic® F127 resulted in larger particles due toOstwald ripening that occurred during process. A similar correlationwas observed for ibuprofen particles during storage where Ostwaldripening was also believed to be the driving factor for formation oflarger particles. Van Eerdenbrugh et al. demonstrated that Ostwaldripening was highly dependent on temperature by exploring TPGSstabilized nanosuspensions for 9 different drug candidates [86].Following 3 months storage at room temperature, Ostwald ripeningoccurred in 8 out of 9 nanosuspensions studied. Enhanced Ostwaldripening was observed at 40 °C storage, while lowering temperatureto 4 °C slowed down or even stopped Ostwald ripening effects.

2.2.4. Change of crystalline stateCrystalline state is one of the most important parameters affecting

drug stability, solubility, dissolution and efficacy. The main issue withcrystalline state change is the transformation between amorphousand crystalline state. The high energy top-down manufacturingtechniques tend to create partially amorphous nanosuspensions andsome bottom-up techniques can create completely amorphousparticles. The high energy amorphous particles are unstable andinclined to convert to low energy crystalline state over time. Thisconversion occurs depending on different parameters, such astemperature, dispersion medium, stabilizers and the presence ofcrystalline particles. Lindfors et al. produced Felodipine amorphous

Fig. 6. Schematic illustration of Ostwald ripening.

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

nanosuspensions via anti-solvent precipitation under sonication [96].They demonstrated that amorphous nanoparticles were highlyunstable in the presence of small amounts of crystalline particles.This was attributed to saturation solubility differences betweenamorphous and crystalline nanoparticles that initiated a similardiffusion process to Ostwald ripening, leading to a rapid conversionof amorphous nanoparticles to crystalline state. Although most ofamorphous particles have been shown to be unstable, a fewamorphous nanosuspensions have been demonstrated to be stableover a certain period of time. Amorphous hydrocortisone nanosus-pensions, produced through a bottom-up nanoprecipitation tech-nique using microfluidic reactors, was found to remain stable after3 months storage at room temperature [93]. Amorphous all-transretinoic acid nanosuspensions, prepared by an anti-solvent precipi-tation technique, were also shown to be stable over 6 months storageat 4 °C [102].

Manufacturing process might also induce some other type ofcrystalline 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 thesame polymorph with differences in molecular conformation andcrystal size. It was demonstrated that the HPH process caused thepartial transformation of DCF2 to DCF1 while no effect on DCF1 wasobserved. The change in the crystalline structure was attributed to thesolubilization of DCF2 during HPH process and its subsequentrecrystallization as the DCF1 form.

2.2.5. Stability issues with solidification process of nanosuspensionsWhen stable nanosuspensions are unattainable, the solid dosage

form is the ultimate solution. Themost common solidification processesare freeze drying and spray drying [10,19,20,118]. Since most solidifiednanoparticle dry powders are usually reconstituted back into nanosus-pensions during administration, drug nanocrystal growth or agglomer-ation during drying process needs to be prevented in order to maintainthe nanosizing features such as rapid dissolution following thereconstitution. Adding matrix formers, such as mannitol, sucrose andcellulose, into nanosuspensions prior to drying is the common approachto overcome the stability issues during solidification process [19]. Sinceseveral excellent reviews have been published on this topic [19,25,118],the readers are directed to those reviews for more details.

2.2.6. Chemical stabilitySince drug nanocrystals are usually dispersed in nanosuspensions

with a limited solubility, the possibility of chemical reactions is not assubstantial as that in solution-based formulations. Consequently,

chemical stability of nanosuspensions is generally superior to that ofsolutions. Paclitaxel serves as a good example to illustrate this [119].Fig. 7(a) shows an HPLC diagram of paclitaxel nanosuspensionsstabilized with Pluronic® F68 after 4 years of storage at 4–8 °C. Novisible degradation product was observed with a recovery of morethan 99%. On the other hand, paclitaxel solution with methanol ascosolvent showed clear degradation only after 48 h at roomtemperature (Fig. 7(b)). The excellent chemical stability of paclitaxelnanosuspensions was attributed to a mechanism similar to oxidizedlayer on the aluminum surface. Monolayer degradation on thenanocrystals surface was created once they were exposed to waterand oxygen, as illustrated in Fig. 7(c). This monolayer could protectthe inner part of drug crystals from further degradation, and thusenhance chemical stability of the nanosuspensions.

Unlike the physical stability issue that is a common concern fornanosuspensions, chemical stability is drug specific. Each moleculehas its particular functional groups and reaction mechanism thataffect the stability. For example, chemical functionalities, such as esterand amides, are susceptible to hydrolytic degradation, while aminogroups may undergo oxidative degradation [120]. Although chemicalstability of nanosuspensions is usually not a major concern, extraattention should be paid to drug molecules with solubility greaterthan 1 mg/mL or with low concentration in suspension [120]. Thecommon strategy to enhance the chemical stability is to transform thenanosupensions into dry solid dosage formwhich is muchmore stablethan nanosuspensions [19] or to increase the concentration of thenanosuspensions [120].

2.3. Additional stability issues relate to large biomolecules

Large biomolecules discussed in this review are mainly referred totherapeutic protein and peptide. The molecular structure of protein/peptide is distinctly different and more complicated as compared tothat of the small molecules. The structures of large molecules aregenerally differentiated into four structures, i.e. primary, secondary,tertiary and quaternary structures [34]. These different structuresrefer to the sequence of the different amino acids, regions where thechains are organized into regular local structures by hydrogenbonding such as alpha helix and beta sheet, the mechanisms onhow the protein/peptide chain folds into a 3-dimensional conforma-tion, and the composition of multiple protein/peptide moleculesassembly, respectively [32,33,123]. The intact molecular structure ofprotein/peptide is essential to maintain their therapeutic efficacy[35,121]. Common stability issues associated with protein/peptideinclude deamidation, oxidation, acylation, unfolding, aggregation andadsorption to surfaces [35,121]. These stability issues are affected by

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.

465L. Wu et al. / Advanced Drug Delivery Reviews 63 (2011) 456–469

temperature, solution pH, buffer ion, salt concentration, proteinconcentration, and added surfactants, with solution formulationsbeing more susceptible to the influence from these factors than thesuspension formulations [34,35,121]. Although suspension formula-tions or solid state of protein/peptide have enhanced stability due totheir reduced molecular mobility, other stability issues may ariseduring particle formation or formulation process. For example,irreversible denaturation and aggregation upon reconstitution wereoften observed for dehydrated protein through freeze drying or spray-drying [125,126]. To prevent this, supplementary excipients such asbulking agents or surfactants are usually introduced during lyophi-lization [122].

The vulnerable structure of protein/peptide creates challenges forformulation development. Instead of using “naked” protein, thecommon strategy to prevent protein/peptide denaturation is toencapsulate the biomolecules with carrier such as liposome [123],SLN [124] or polymeric materials [125,126]. In addition to improvingthe stability, protein/peptide encapsulation can enhance bioavailabil-ity and provide sustained therapeutic release [125–128]. There hasbeen plenty of work reported on encapsulated protein/peptidenanoparticles but very scarce studies on pure protein/peptidenanoparticles. Gomez et al. produced bovine zinc insulin nanoparti-cles using an electrospray drying technique and reported retainedbiological 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 ofthe harsh HPH process conditions. Merisko-Liversidge et al. [83] alsonoticed retained stability and bioactivity of zinc-insulin nanosuspen-sions that were produced through a wetmilling process in presence ofPluronic® F68 and sodium deoxycholate. Nyambura et al. utilized abottom up technique (combination of emulsification and freezedrying) to generate insulin nanoparticles (80 w/w% insulin with20 wt.% lactose) [110]. The particles were then dispersed intoHFA134a to produce an MDI formulation. The molecular integrity ofinsulin formulation, measured by HPLC, size exclusion chromatogra-phy, circular dichroism and fluorescence spectroscopy, indicated thatnative structures (primary, secondary and tertiary) were retainedafter particle formation and formulation process. The presence ofsurfactant (lecithin) and lyoprotectant (lactose) was believed to beresponsible for preservation of the insulin structures. In their followup work [109], they applied a similar approach to produce composite

nanoparticles of lysozyme and lactose for MDI formulations. Theretained biological activity of lysozymewas enhanced with increasinglactose concentration in the particles, and reached maximum (99%retained activity) with 20 w/w% lactose. Nanoprecipitation coupledwith freeze drying was used as well in this work to produce sphericalnanoparticles containing 80 w/w% lysozyme with fully preservedbioactivity. It was demonstrated that bioactivity of lysozymenanoparticles 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 bioactivelysozyme particles [131]. Both processes generated highly porousmicro-sized aggregates of lysozyme primary nanoparticles in the sizeof 100–300 nm. SFL process was shown to produce lysozyme withlower aggregation and higher enzyme activity as compared to the SFDprocess, 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 nanoparticleformulations

Selection of characterization techniques for drug nanoparticlesstability is dependent on the nature of stability issues and productdosage form. A few commonly used stability characterizationtechniques are listed in Table 2.

3.1. Particle size, size distribution and morphology

Particle size and size distribution are the key parameters used forevaluating the physical stability of nanoparticles. A variety oftechniques, including photon correlation spectroscopy (PCS), alsoknown as dynamic light scattering (DLS), laser diffraction (LD) andcoulter counter, are commonly used to measure the particle size andsize distribution (Table 2). The PCS/DLS is widely used to determinethe size and size distribution of small particles suspended in liquidmedium. The mean particle size and size distribution indicated aspolydispersity index (PDI) are the typical measured parameters of thistechnique. A PDI value of 0.1 to 0.25 indicates a narrow sizedistribution while a PDI greater than 0.5 refers to a broad distribution[20]. Unfortunately, this technique is not capable of measuring size ofdry powders and its measurement range is too narrow (3 nm to 3 μm)to detect the interference from the microparticles (N3 μm) within the

Table 2Commonly 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.

Laser diffraction Pros: wide measurement range, rapid, non-invasive, apply to both liquid suspension anddry powder samples.Cons: particles are assumed to be spherical.

Coulter counter Pros: precise.Cons: apply only to spherical particles.

Particle size and morphology SEM/TEM Pros: evaluate both particle morphology and size, very small quantity of sample required.Cons: challenging to acquire statistical size distribution, usually invasive, time-consuming.

AFM Pros: non-invasive, evaluate both particle morphology and size, very small quantity ofsample required.Cons: challenging to acquire statistical size distribution, time-consuming.

Sedimentation/creaming Visual observation/laser backscattering/near infrared transmission

–

Particle surface charge/zeta potential Laser Doppler electrophoresis –

Crystallinity state XRD/DSC –

Chemical stability HPLC/FTIR/NMR/MS –

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

nanosuspensions. Therefore, LD is often used in combination with PCSto circumvent this issue. Laser diffraction has a much wider detectionrange (20 nm to 2000 μm) and it can be used to evaluate bothsuspension and dry powder samples. The typical LD characterizationparameters are LD50, LD90 and LD99, indicating 50, 90 or 99% of theparticles are below the given size, respectively. LD is especiallysuitable for characterizing parenteral and pulmonary suspensions dueto it wide measurement range. LD can detect the presence ofmicroparticles (N5 μm) which are detrimental to parenteral nanosus-pensions. However, LD provides only relative size distribution. TheCoulter counter, on the other hand, measures the absolute number ofparticles per volume unit for the different size classes, and is moreprecise than the LD.

Although PCS, LD and coulter counter techniques provide rapidmeasurement of particle size and size distribution, they do not havethe capability in evaluating particle morphology. As direct visualiza-tion techniques, Scanning Electron Microscope (SEM), TransmissionElectron Microscope (TEM) and Atomic Force Microscope (AFM) arewidely used for assessment of particle morphology. However, it isvery challenging and time-consuming to measure a significantnumber of particles to achieve statistical size distribution usingthese techniques. In addition, they usually require additional samplepreparation 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 isby visual observation over a period of time. By measuring the volumeof the settled or creamed particle layer relative to the total suspensionvolume within a specific time, a dimensionless parameter known assedimentation or flocculation volume can be obtained as a quantita-tive evaluation of suspension stability. A higher flocculation volumeindicates a more stable suspension. The structure of settled/creamedlayer can be easily assessed by re-dispersing the suspension, i.e. easilyre-dispersed suspension indicates loose flocs while a dense cake ishard to be broken by manual shaking. Other approaches to evaluatesedimentation/creaming include laser backscattering [132] and near-infrared transmission [133].

3.3. Particle surface charge

Laser Doppler electrophoresis is commonly used to measure ZP.This technique evaluates electrophoretic mobility of suspendedparticles 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 andless than 5 mV generally results in good stability, acceptable short-term stability and fast particle aggregation, respectively [29]. This ruleof thumb is only valid for pure electrostatic stabilization or incombination with low-molecular weight surfactants, and is not validwhen high molecular weight stabilizers are present [29].

3.4. Crystalline state

The crystallinity of drug nanoparticles is usually assessed by X-RayDiffraction (XRD) and/or Differential Scanning Calorimetry (DSC).XRD differentiates amorphous and crystalline nanoparticles as well asdifferent polymorphic phases of the particles, while DSC is often usedas a supplementary tool to XRD. Crystalline particles usually have asharp melting peak which is absent in amorphous materials. Themelting point can also be utilized to differentiate differentpolymorphs.

3.5. Chemical stability

HPLC is the most common characterization technique used toevaluate chemical stability that provides precise quantitative analysison the degradation impurities. Mass spectrometry (MS) is oftencoupled with HPLC to identify the molecular structure of impurities.Some other techniques such as FTIR and NMR can also be used forchemical stability assessment. However, they are not as precise andsensitive as HPLC, and thus not widely used for stability assessment.

3.6. Additional techniques for assessing large biomolecule nanoparticleand formulation stability

For large biomolecules, additional characterization tools aregenerally required depending on the level of molecular structure tobe assessed. For instance, size exclusion chromatography andelectrophoresis are used to evaluate the primary structure of largebiomolecules, circular dichroism is to monitor the secondary andtertiary structures while fluorescence spectroscopy is for tertiarystructure [34,134]. In addition, in-vitro bioassays or in-vivo efficacytests are needed to evaluate biological activities of the largebiomolecules. Insulin particles, as an example, have been tested forits bioactivity either by in-vitro chondrocyte culture assays [130] or in-vivo monitoring of blood glucose level on rats following insulinadministration [83].

467L. Wu et al. / Advanced Drug Delivery Reviews 63 (2011) 456–469

4. Recommendations of general strategies for enhancing stabilityof nanoparticle formulations

Strategies to address different stability issues are usually tailoredaccording to different aspects, such as therapeutic requirements,dosage form and manufacturing complexity. For example, as theparticle size is reduced, the sedimentation rate is decreased so that theparticles can stay suspended longer in nanosuspensions. The generalwisdom is that the smaller the nanoparticles are, the better.Unfortunately, too small particles are not always desirable, as theymay create undesired plasma peaks due to the significant increases indissolution rate [28]. Moreover, manufacturing complexity may beincreased as well when the particles size requirements become toostringent.

The use of stabilizers is the most commonly used technique inachieving a stable nanoparticle formulation. However, the stabilizerselection is known to be very challenging. The challenge stemsmainlyfrom two aspects: (i) lack of fundamental understanding of interac-tions within nanosuspensions and (ii) lack of an efficient and highthroughput stabilizer screening technique. In the case of aqueousnanosuspensions, it is relatively easy to select stabilizers given thatwater-based stabilizingmoieties such as PEG and PVA arewell known.However, selecting the anchor groups that interact strongly with thedrug surface can be challenging due to the limited understanding oninteractions between nanoparticles and stabilizers in molecular level.For non-aqueous nanosuspensions such as HFA-based MDI deliverysystem, understanding of solvation in the low-dielectric HFA mediumis still in its infancy, which makes stabilizers selection even morechallenging. Inefficient screening approaches are another hurdle forstabilizer selection. The current practice for stabilizer screeninginvolves trial production of nanosuspensions with different stabilizersor stabilizer combinations, which could be burdensome and requirevast amount of efforts especially with a large number of potentialstabilizer candidates. AFM has recently been proven to be a feasibleand efficient tool for stabilizer screening. Verma et al. demonstratedthe feasibility of using AFM to select stabilizers for Ibuprofennanosuspensions [135]. The AFM measurements showed that HPMCand HPC had extensive surface absorption on the ibuprofen surface, asopposed to the inadequate surface absorptionwith PVP and Pluronic®surfactants. These results correlated well with their stabilizingperformances in the nanosuspensions. This finding confirmed thesignificance of AFM in providing a scientific rationale for stabilizerselection and improving understanding of the stabilization mecha-nisms. Another technique, known as colloidal probe microscopy(CPM) which is derived from AFM, has also been widely used to studyinteractions between colloidal particles and is expected to be a usefultool for nanosuspension stabilizer screening [136].

Due to the significant challenges associated with stabilizerselection, self-stabilized nanosuspensions with no added stabilizerare highly desirable. This is not only for simplifying the formulationdevelopment process but also reducing stabilizer-based toxicity.Unfortunately, the challenges to engineer such self-suspendednanoparticles are tremendous with very few reported studies todate. A couple of approaches that could potentially be used to produceself-stabilized nanosuspensions include the creation of drug nano-particles with high ZP and controlling morphology or surfaceproperties of drug nanoparticles to minimize inter-particulate forces.

5. Conclusions

The stability of drug nanoparticles remains a very challengingissue during pharmaceutical product development. Stability isaffected by various factors such as dosage form (nanosuspension vs.dry solid), dispersion medium (aqueous vs. non-aqueous), deliveryroute (oral, inhalation, IV or other routes), production technique (top-down vs. bottom-up) and nature of drug (small molecules vs. large

biomolecules). Despite the significant challenges associated withstabilizer screening, adding a stabilizer or combination of stabilizers isstill the most commonly used and preferred approach to enhance thestability of nanosuspensions. Further understanding of particle–particle interactions within nanosuspensions and development ofhigh-throughput stabilizer screening tools are essential to facilitateefficient stabilizer selection. Development of self-stabilized nanosus-pensions, although currently seen as very complicated and challeng-ing, is expected to grow with the continuing advancement in the fieldof 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 genedelivery 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 drugdelivery, 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 deliverysystems, Adv. Drug Deliv. Rev. 45 (1) (2000) 89–121.

[9] T.M. Allen, E.H. Moase, Therapeutic opportunities for targeted liposomal drugdelivery, 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, Biodegradablepolymeric 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 controlleddrug 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 drugdelivery, 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 drugnanocrystals Nanosuspension stabilization, miniaturization and transformationinto solid products, Int. J. Pharm. 364 (2008) 64–75.

[20] V.B. Patravale, A.A. Date, R.M. Kulkarni, Nanosuspensions a promising drugdelivery 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 dissolutionand 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 ofamphotericin 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 fororal administration, Curr. Nanosci. 4 (2008) 53–58.

[24] N. Blagden, M.D. Matas, P.T. Gavan, P. York, Crystal engineering of activepharmaceutical 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 ofnanoparticles formulation, process and storage considerations, Adv. DrugDeliv. Rev. 58 (2006) 1688–1713.

[26] J.E. Kipp, The role of solid nanoparticle technology in the parenteral delivery ofpoorly 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 ofHesperetin nanosuspensions for dermal delivery, Int. J. Pharm. 371 (2009)182–189.

[30] J. Pardeike, R.H. Müllera, Nanosuspensions a promising formulation for the newphospholipase 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 andbottom-up approaches for the preparation of micro/nanosuspensions, Int. J.Pharm. 380 (2009) 216–222.

[32] S.K. Murthy, Nanoparticles inmodernmedicine state of the art, Int. J. Nanomed. 2(2) (2007) 129–141.

[33] P. Rogueda, Novel hydrofluoroalkane suspension formulations for respiratorydrug 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 deliverysystems, 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: physico-chemical 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 formulationdevelopment to Manufacturing, Springer, 2009, pp. 39–66.

[38] Y. Ali, K. Lehmussaari, Industrial perspective in ocular drug delivery, Adv. DrugDeliv. 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, Templatedopen flocs of anisotropic particles for pulmonary delivery with pressurizedmetered 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, Templatedopen flocs of nanorods for enhanced pulmonary delivery with pressurizedmetered 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 lyophobicsols 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 fordrug 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 processconditions 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 duringthe 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 nano-comminution 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 copolymersas 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 paclitaxelnanocrystals, 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 forthe 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 pulmonarydrug delivery, Science 276 (5320) (1997) 1868–1872.

[55] L. Gao, D. Zhang, M. Chen, T. Zheng, S. Wang, Preparation and characterization ofan oridonin nanosuspension for solubility and dissolution velocity enhance-ment, Drug Dev. Ind. Pharm. 33 (2007) 1332–1339.

[56] L. Gao, D. Zhang, M. Chen, C. Duan, W. Dai, L. Jia, W. Zhao, Studies onpharmacokinetics and tissue distribution of oridonin nanosuspensions, Int. J.Pharm. 355 (2008) 321–327.

[57] C. Jacobs, R.H. Müller, Production and characterization of a budesonidenanosuspension for pulmonary administration, Pharm. Res. 19 (2) (2002)189–194.

[58] N. Hernández-Trejo, O. Kayser, H. Steckel, R.H. Muller, Characterization ofnebulized buparvaquone nanosuspensions — effect of nebulization technology,J. Drug Target. 13 (8–9) (2005) 499–507.

[59] R.H. Muller, C. Jacobs, Buparvaquone mucoadhesive nanosuspension: prepara-tion, 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, Diclofenacnanosuspensions influence of preparation procedure and crystal form on drugdissolution behaviour, Int. J. Pharm. 373 (2009) 124–132.

[61] D. Zhang, T. Tan, L. Gao, W. Zhao, P. Wang, Preparation of azithromycinnanosuspensions by high pressure homogenization and its physicochemicalcharacteristics studies, Drug Dev. Ind. Pharm. 33 (2007) 569–575.

[62] R. Mauludin, R.H. Muller, C.M. Keck, Development of an oral rutin nanocrystalformulation, Int. J. Pharm. 370 (2009) 202–209.

[63] R. Mauludin, R.H. Muller, C.M. Keck, Kinetic solubility and dissolution velocity ofrutin nanocrystals, Eur. J. Pharm. Sci. 36 (2009) 502–510.

[64] C. Jacobs, O. Kayser, R.H. Muller, Nanosuspensions as a new approach for theformulation 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 anintravenously injectable chemically stable aqueous omeprazole formulationusing 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 ofintravenously 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 oflecithins 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 glucocorticoiddrugs, Int. J. Pharm. 340 (2007) 126–133.

[69] V. Teeranachaideekul, V.B. Junyaprasert, E.B. Souto, R.H. Muller, Development ofascorbyl 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 concentratednanosuspensions by high pressure homogenisation, Int. J. Pharm. 214 (2001)21–24.

[71] J. Hecq, M. Deleers, D. Fanara, H. Vranckx, K. Amighi, Preparation andcharacterization of nanocrystals for solubility and dissolution rate enhancementof 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 dissolutionrate enhancement of ucb-35440-3, a highly dosed poorly water-soluble weakbase, 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 andcytotoxic activity of hydroxycamptothecin nanosuspensions, Int. J. Pharm. 392(2010) 64–71.

[74] S. Ganta, J. Paxton, B.C. Baguley, S. Garg, Formulation and pharmacokineticevaluation 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 solubledrugs 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, Neuroprotectiveeffects of a nanocrystal formulation of sPLA2 inhibitor PX-18 in cerebral ischemiareperfusion 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 andintravenous delivery, Nanotechnology 21 (2010) 1–12.

[79] Y. Tanaka, M. Inkyo, R. Yumoto, J. Nagai, M. Takano, S. Nagata, Nanoparticulationof poorly water soluble drugs using a wet-mill process and physicochemicalproperties 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 andbudesonide nanosuspensions for pulmonary delivery preparation, characteriza-tion, 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 hydroxypropylcellulose as stabilizers on the physical stability of high drug loadingnanosuspensions 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 pharmaceu-tical performance of sucrose co-freeze-dried solid nanoparticulate powders ofthe 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 novelformulation 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 ethyldiatrizoate 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 dissolutionperformance and X-ray microanalysis as an evaluation tool for powderdispersion, 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 surfacehydrophobicity 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 ofa 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 nanoparticlesuspension 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, Formulationand antitumor activity evaluation of nanocrystalline suspensions of poorlysoluble anticancer drugs, Pharm. Res. 13 (2) (1996) 272–278.

469L. 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 vitroantimalarial activity of dihydroartemisinin nanosuspensions obtained by co-grinding 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 nanosuspensionthrough 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 nanosuspen-sions 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 neointi-mal hyperplasia in rats, J. Control. Release 117 (2007) 322–332.

[96] L. Lindfors, P. Skantze, U. Skantze, J. Westergren, U. Olsson, Amorphous drugnanosuspensions. 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 ofpolymer 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 thecosolvent technique, Eur. J. Pharm. Sci. 30 (2007) 367–374.

[99] J.M. Tam, J.T. Mcconville, R.O. Williams III, K.P. Johnston, Amorphous cyclosporinnanodispersions 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 formulationcomparison, using a solution and different nanosuspensions of a poorly solublecompound, Eur. J. Pharm. Biopharm. 67 (2007) 540–547.

[101] D. Douroumis, A. Fahr, Nano- and micro-particulate formulations of poorlywater-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 nanosuspensionsusing 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 usingpure 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 stablenitrendipine nanosuspensions using the precipitation–ultrasonication methodfor 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 celecoxibnanosuspension formulation and transformation into tablets, Int. J. Pharm. 376(2009) 204–212.

[106] M. Trotta, R. Cavalli, D. Chirio, Griseofulvin nanosuspensions from triacetin-in-water emulsions. S.T.P, Pharm. Sci. 13 (6) (2003) 423–426.

[107] M. Trotta, M. Gallarate, F. Pattarino, S. Morel, Emulsions containing partiallywater-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 griseofulvinnanoparticles from water-dilutable microemulsions, Int. J. Pharm. 254 (2003)235–242.

[109] B.K. Nyambura, I. Kellaway, K.M.G. Taylor, The processing of nanoparticlescontaining 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 andaerosolization 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 pulmonarydrug administration, J. Drug Target. 9 (4) (2001) 295–302.

[112] F.J. Ahmad, R.K. Khar, S. Sultana, A. Bhatnagar, Techniques to develop andcharacterize nanosized formulation for salbutamol sulfate, J. Mater. Sci. Mater.Med. 20 (2009) 71–76.

[113] B.L. Montalvo, Y. Pacheco, B.A. Sosa, D. V'elez, G. S'anchez, K. Griebenow,Formation of spherical protein nanoparticles without impacting proteinintegrity, Nanotechnology 19 (2008) 1–7.

[114] H. Piao, N. Kamiya, A. Hirata, T. Fujii, M. Goto, A novel solid-in-oil nanosuspen-sion 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, andTechniques, second edition, CRC Press, 2004, pp. 185–234.

[118] F. Kesisoglou, S. Panmai, Y. Wu, Nanosizing— oral formulation development andbiopharmaceutical 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 aglassy 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-co-glycolide) microspheres: effect of the protein and polymer properties and of theco-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 biodegrad-able poly(d,l-lactide-co-glycolide) microspheres prolongs and enhances antigenpresentation 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 byelectrospray drying, J. Aerosol Sci. 29 (5–6) (1998) 561–574.

[130] A. Maschke, N. Calı, B. Appel, J. Kiermaier, T. Blunk, A. Gopferich, Micronization ofinsulin 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 spray-freeze drying: influence of atomization on protein aggregation and biologicalactivity, Eur. J. Pharm. Sci. 27 (1) (2006) 9–18.

[132] K.A. Johnson, Interfacial phenomena and phase behaviour in metered doseinhaler formulations, in: A.J. Hickey (Ed.), Editor Inhalation Aerosols: Physicaland Biological Basis for Therapy, second, Informa Healthcare, 2007, pp. 347–372.

[133] M. Kuentz, D. Röthlisberger, Rapid assessment of sedimentation stability indispersions using near infrared transmission measurements during centrifuga-tion 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 ofpeptides and proteins, Taylor & Francis, 1999, pp. 41–69.

[135] S. Verma, B.D. Huey, D.J. Burgess, Scanning probe microscopy method fornanosuspension stabilizer selection, Langmuir 25 (21) (2009) 12481–12487.

[136] L. Wu, S.R.P. da Rocha, Biocompatible and biodegradable copolymer stabilizersfor 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 ascosmetics and skin lotions, pharmaceutical ointments, varnishes,paints and lubricants and fuels are based wholly or in part ofemulsions. Hielscher manufactures the world's largest industrialultrasonic liquid processors for the efficient emulsifying of largevolume streams in production plants.In the lab, the emulsification power of ultrasound has been known andapplied 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 wattseach, provide the capacity needed to translate this lab application into anefficient production method to obtain finely dispersed emulsions incontinuous flow or in a batch - achieving results comparable to that oftoday's best high-pressure homogenizers available, such as the neworifice valve. In addition to this high efficiency in the continuousemulsification, Hielscher ultrasonic devices require very low maintenanceand are very easy to operate and to clean. The ultrasound does actuallysupport the cleaning and rinsing. The ultrasonic power is adjustable andcan be adapted to particular products and emulsification requirements.Special flow cell reactors meeting the advanced CIP (clean-in-place) andSIP (sterilize-in-place) requirements are available, too.

Emulsions are dispersions of two or moreimmiscible liquids. Highly intensive ultrasoundsupplies the power needed to disperse a liquidphase (dispersed phase) in small droplets in asecond phase (continuous phase). In thedispersing zone, imploding cavitation bubbles

cause intensive shock waves in the surrounding liquid and result in theformation of liquid jets of high liquid velocity. In order to stabilize the newlyformed droplets of the disperse phase against coalescence, emulsifiers(surface active substances, surfactants) and stabilizers are added to theemulsion. As coalescence of the droplets after disruption influences thefinal droplet size distribution, efficiently stabilizing emulsifiers are used tomaintain the final droplet size distribution at a level that is equal to thedistribution immediately after the droplet disruption in the ultrasonicdispersing zone. Stabilizers actually lead to improved droplet disruption atconstant energy density.

Studies at oil in water (water phase) and water in oil (oil phase) emulsionshave shown the correlation between the energy density and droplet size

(e.g. Sauter diameter). There is a cleartendency for smaller droplet size atincreasing energy density (click at rightgraphic). At appropriate energy densitylevels, ultrasound can well achieve a meandroplet sizes below 1 micron(microemulsion).

Quick Link

Information RequestLiterature on Emulsifying

Information Requestyour 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+

BookmarkWe would like to see you comingback. To bookmark this page,please press Crtl + D.

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

by Luis M. Liz-Marzán

Nanometals:formation and color

Departamento de QuÌmica FÌsica,

Universidade de Vigo,

36200 Vigo, Spain

E-mail: lmarzan@uvigo.es

URL: webs.uvigo.es/coloides/nano

February 200426 ISSN:1369 7021 © Elsevier Ltd 2004

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 wet-

chemical 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.

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 nanoparticlesIn 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

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 aminopropyl-

trimethoxysilane 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 (<1 mM)25, while increasing Ag concentration

(up to 0.02 M) largely favors the formation of anisotropic

particles28,29. Again, the concentration of PVP and the

reaction temperature strongly influence the shape of the

final particles.

As mentioned above, deviations from spherical geometry

strongly affect the optical properties of metal nanoparticles.

For this reason, methods for the synthesis of anisometric

nanoparticles in solution (nanorods, nanowires, nanodisks,

nanoprisms, etc.) are continuously being reported, in

particular for Au30-33 and Ag34-36. Some of these are

described in the previous paragraph, but others deserve a

brief description as well.

The first method for synthesis of Au nanorods in solution

(electrochemical growth within inorganic templates37,38 had

been successful before) was based on the electrochemical

reduction of HAuCl4 in the presence of ‘shape-inducing’

cationic surfactants and other additives. These had been

found empirically to favor rod formation and act as both the

supporting electrolyte and stabilizer for the resulting

cylindrical Au nanoparticles39. In the electrochemical

method, the micellar system consists of two cationic

surfactants: cetyltrimethylammonium bromide (CTAB) and

the much more hydrophobic tetradecylammonium bromide

(TDAB) or TOAB. The ratio between the surfactants controls

the average aspect ratio of the Au nanorods. The synthesis

works on a small scale and, although it is very difficult to

carry out on a large scale, represents a landmark in terms of

shape control.

Subsequently, Murphy and coworkers discovered reduction

conditions that enable the entire synthesis of Au and Ag

nanorods to be carried out directly in solution30,40, with

excellent aspect ratio control. The method uses preformed

Ag or Au seeds on which additional metal is grown in solution

by means of a mild reducing agent (ascorbic acid), in the

REVIEW FEATURE

February 200428

Fig. 1 Left: Transmission electron micrographs of Au nanospheres and nanorods (a,b) and Ag nanoprisms (c, mostly truncated triangles) formed using citrate reduction, seeded growth, and

DMF reduction, respectively. Right: Photographs of colloidal dispersions of AuAg alloy nanoparticles with increasing Au concentration (d), Au nanorods of increasing aspect ratio (e), and

Ag nanoprisms with increasing lateral size (f).

REVIEW FEATURE

presence of CTAB to promote nanorod formation. The

addition of different volumes of the seed solution produces

Au nanorods with different aspect ratios. Recently, this

method has been improved by El-Sayed and coworkers31,

resulting in a spectacular increase in the rod yield. Another

modification of the method is based on a photochemical

process41 but, in contrast to the seed-mediated mechanism,

the addition of a preformed Au seed is not needed. However,

AgNO3 is added to the growth solution containing CTAB and

the Au salt before irradiating with ultraviolet light.

The controlled synthesis and optical characterization of

metal nanoprisms was pioneered by Jin et al.34, who

converted citrate-stabilized Ag nanospheres into (truncated)

triangular nanoprisms by irradiation with a fluorescent lamp

in the presence of bis(p-sulfonatophenyl) phenylphosphine.

The optical spectra of the nanoprisms display bands for

in-plane dipole resonance, as well as for in-plane and

out-of-plane quadrupole resonances, while the out-of-plane

dipole is only reflected as a small shoulder.

Truncated Ag triangles have also been obtained by Chen

and Carroll35 using a procedure similar to the seeded growth

method described for Au nanorod formation. In this case, the

reduction was performed using ascorbic acid to reduce Ag

ions on Ag seeds in a basic solution of concentrated CTAB.

Malikova et al.32 synthesized Au nanoprisms in aqueous

solution by reduction of neutralized HAuCl4 with salicylic

acid at 80°C. Although the yield of nanoprisms was not

outstanding, the formation of a thin film shows that strong

optical coupling occurs when the nanoprisms are close

enough to each other.

The mechanisms involved in the synthesis of spherical

metal nanoparticles are well understood in general, but those

leading to preferential growth in one particular direction are

still the subject of debate. Evidence for an aggregative

mechanism in some cases has been presented28,42, but it is

clear that such a mechanism does not apply in others. Further

work in this direction is still needed.

Examples of metal nanoparticles with various shapes and

sizes, together with dispersions of varying colors arising from

different effects, are shown in Fig. 1.

Optical propertiesThe optical properties of small metal nanoparticles are

dominated by the collective oscillation of conduction

electrons resulting from the interaction with electromagnetic

radiation. These properties are mainly observed in Au, Ag, and

Cu, because of the presence of free conduction electrons. The

electric field of the incoming radiation induces the formation

of a dipole in the nanoparticle. A restoring force in the

nanoparticle tries to compensate for this, resulting in a

unique resonance wavelength (Fig. 2, top)43.

The oscillation wavelength depends on a number of

factors, among which particle size and shape, as well as the

nature of the surrounding medium, are the most important5.

For nonspherical particles, such as rods, the resonance

wavelength depends on the orientation of the electric field.

Therefore, two oscillations, transverse and longitudinal, are

possible (Fig. 2, bottom). In addition, when nanoparticles are

sufficiently close together, interactions between neighboring

particles arise, so that the models for isolated particles do

not hold. The properties of dilute dispersions will be briefly

discussed first, while a simple effective medium theory

(Maxwell-Garnett) will be described next to account for the

behavior of more concentrated systems, such as close-packed

thin films.

DDiilluuttee DDiissppeerrssiioonnss

The optical properties of dispersions of spherical particles

with a radius R can be predicted by Mie theory44, through

expressions for the extinction cross section Cext. For very

small particles with a frequency dependent, complex

dielectric function, ε = ε’ + iε’’, embedded in a medium of

dielectric constant εm, this can be expressed as:

(1)

February 2004 29

Fig. 2 (Top) Schematic drawing of the interaction of an electromagnetic radiation with a

metal nanosphere. A dipole is induced, which oscillates in phase with the electric field of

the incoming light. (Bottom) Transverse and longitudinal oscillation of electrons in a

metal nanorod.

The origin of the strong color changes displayed by small

particles lies in the denominator of eq 1, which predicts the

existence of an absorption peak when

ε’ = -2εm (2)

In a small metal particle, the dipole created by the electric

field of light induces a surface polarization charge, which

effectively acts as a restoring force for the free electrons. The

net result is that, when condition (2) is fulfilled, the long

wavelength absorption by the bulk metal is condensed into a

single surface plasmon band.

To calculate the spectra of elongated particles (nanorods),

the orientation with respect to the oscillating electric field

must be taken into account. The corresponding expression

was derived by Gans45:

(3)

where Pj represents the depolarization factors for the

nanorod axes (a > b = c), which are defined as

(4)

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:

(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.

TThhiinn FFiillmmss

When the metal nanoparticle volume fraction is high, the

equations above are no longer valid, since dipole-dipole

interactions between neighboring nanoparticles are present,

i.e. the oscillating dipoles of neighboring particles influence

the frequency of a central particle. Under these conditions,

effective medium theories are the simplest way to describe

the optical response of the system. Such theories provide us

with expressions to calculate the effective dielectric constant

of the composite material, which can then be used to

determine the corresponding absorption and reflection

coefficients.

Among the various effective medium theories available, it

has been found7 that the one devised by Maxwell-Garnett47

is suitable to describe these dipole-dipole interactions.

Through the average polarization of the nanoparticles and the

surrounding medium, the average dielectric function,

εav = (nav + ikav)2, is calculated as

(6)

where φ is the metal volume fraction, εm is the dielectric

function of the surrounding medium, and ε is the complex

dielectric function of the nanoparticles. The transmittance, T,

of radiation with a frequency ω through the film can be then

calculated:

(7)

where h is the film thickness, R is the reflectance at normal

incidence,

(8)

and α is the absorption coefficient, which can be calculated

from ω Im(εav)/cnav. Finally, we define the parameters

ζ = 4πnavh/λ and ψ = tan-1(2kav/(nav2 + kav

2 - 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

February 200430

REVIEW FEATURE

golden reflection is gradually lost as the particles are

separated from each other.

ConclusionApart 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 (<1 ps)

response times. This is a consequence of the relaxation of the

nonequilibrium electron distribution generated in the metal

nanoparticles50,51. Theoretical and experimental studies

taking into account the nature of the surrounding medium,

interparticle interactions, and shape effects are still in their

infancy and need much further attention. MT

AcknowledgmentsI wish to acknowledge Jorge Pérez-Juste, Isabel Pastoriza-Santos, and Benito Rodríguez-

González for providing unpublished figures, and Paul Mulvaney for having injected me with

metal nanoparticle fever.

February 2004 31

REFERENCES

1. www.thebritishmuseum.ac.uk/science/text/lycurgus/sr-lycurgus-p1-t.html

2. Faraday, M., Philos. Trans. Royal Soc. London (1857) 114477, 145

3. Kreibig, U., and Vollmer, M., Optical Properties of Metal Clusters, Springer-

Verlag, Berlin (1996)

4. Link, S., and El-Sayed, M. A., J. Phys. Chem. B (1999) 110033, 8410

5. Mulvaney, P., Langmuir (1996) 1122, 788

6. Underwood, S., and Mulvaney, P., Langmuir (1994) 1100, 3427

7. Ung, T., et al., J. Phys. Chem. B (2001) 110055, 3441

8. Turkevich, J., et al., Discuss. Faraday Soc. (1951) 1111, 55

9. Frens, G., Nature Physical Science (1973) 224411, 20

10. A movie showing the procedure for synthesis of colloidal gold can be found at

www.mrsec.wisc.edu/edetc/cineplex/gold/index.html

11. Turkevich, J., et al., J. Phys. Chem. (1986) 9900, 4765

12. Furlong, D. N., et al., J. Chem. Soc., Faraday Trans. 1 (1984) 8800, 571

13. Brust, M., et al., Chem. Comm. (1994), 801

14. Brust, M., et al., Chem. Comm. (1995), 1655

15. Heath, J. R., et al., J. Phys. Chem. B (1997) 110011, 189

16. Korgel, B. A., et al., J. Phys. Chem. B (1998) 110022, 8379

17. Hostetler, M. J., et al., J. Am. Chem. Soc. (1996) 111188, 4212

18. Templeton, A. C., et al., J. Am. Chem. Soc. (1999) 112211, 7081

19. Hirai, H., et al., J. Macromol. Sci., Chem. (1979) 1133, 727

20. Wang, Y., and Toshima, N., J. Phys. Chem. B (1997) 110011, 5301

21. Ducamp-Sanguesa, C., et al., J. Solid State Chem. (1992) 110000, 272

22. Silvert, P.-Y., et al., J. Mater. Chem. (1996) 66, 573

23. Sun, Y., et al., Nano Lett. (2002) 22, 165

24. Sun, Y., et al., Chem. Mater. (2002) 1144, 4736

25. Pastoriza-Santos, I., and Liz-Marzán, L. M., Langmuir (2002) 1188, 2888

26. Pastoriza-Santos, I., and Liz-Marzán, L. M., Langmuir (1999) 1155, 948

27. Pastoriza-Santos, I., et al., Langmuir (2000) 1166, 2731

28. Giersig, M., et al., J. Mater. Chem., in press (DOI:10.1039/B311454F)

29. Pastoriza-Santos, I., and Liz-Marzán, L. M., Nano Lett. (2002) 22, 903

30. Jana, N. R., et al., J. Phys. Chem. B (2001) 110055, 4065

31. Nikoobakht, B., and El-Sayed, M. A., Chem. Mater. (2003) 1155, 1957

32. Malikova, N., et al., Langmuir (2002) 1188, 3694

33. Sanyal, A., and Sastry, M., Chem. Comm. (2003) 1111, 1236

34. Jin, R., et al., Science (2001) 229944, 1901

35. Chen, S., and Carroll, D. L., Nano Lett. (2002) 22, 1003

36. Maillard, M., et al., J. Phys. Chem. B (2003) 110077, 2466

37. Martin, C. R., Science (1994) 226666, 1961

38. Schonenberger, C., et al., J. Phys. Chem. B (1997) 110011, 5497

39. Yu, Y.-Y., et al., J. Phys. Chem. B (1997) 110011, 6661

40. Murphy, C. J., and Jana, N. R., Adv. Mater. (2002) 1144 (1), 80

41. Kim, F., et al., J. Am. Chem. Soc. (2002) 112244, 14316

42. Viau, G., et al., Chem. Comm. (2003) 1177, 2216

43. Henglein, A., J. Phys. Chem. (1993) 9977, 5457

44. Bohren, C. F., and Huffman, D. F., Absorption and Scattering of Light by Small

Particles, Wiley, New York, (1983)

45. Gans, R., Ann. Phys. (1915) 4477, 270

46. Link, S., et al., J. Phys. Chem. B (1999) 110033, 3073

47. Maxwell-Garnett, J. C., Philos. Trans. R. Soc. London (1904) 220033, 385

48. Kotov, N. A., et al., J. Phys. Chem. (1995) 9999, 13065

49. Hache, F., et al., J. Opt. Soc. Am. B (1986) 33 (12), 1647

50. Tokizaki, T., et al., Appl. Phys. Lett. (1994) 6655, 941

51. Selvan, S. T., et al., J. Phys. Chem. B (2002) 110066, 10157

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.

Quantum Dots:Science and Applications

James McDanielPhysics 3500

Nanochemistry

Why I chose this topic?

• Useful Now: I chose to research and present on the science and applications of Quantum Dots because of the many interesting and important applications currently in use by this nano-technology.

• The Future is Bright: The usefulness and application of Quantum Dot technology continues to expand and research is striving to bring their benefits to more and more technologically applied fields.

What are Quantum Dots?

• Quantum dots are semi-conductors that are on the nanometer scale.

• Obey quantum mechanical principle of quantum confinement.

• Exhibit energy band gap that determines required wavelength of radiation absorption and emission spectra.

• Requisite absorption and resultant emission wavelengths dependent on dot size.

Fig. 1. Schematic plot of the single particle energy band gap. The upper parabolic band is the conduction band, the lower the valence.

Confinement - Infinite Square Well Potential

Fig. 2. Quantized energy levels of a particle in a box.

Quantum Dots Description…

• The emission and absorption spectra corresponding to the energy band gap of the quantum dot is governed by quantum confinement principles in an infinite square well potential.

• The energy band gap increases with a decrease in size of the quantum dot.

Quantum Confinement

Fig. 3. As the energy well, or the particle, shrinks the gap in energy levels increases.

Fig. 4. The energy band gap associated with semi-conducting materials. In order to produce electric current electrons must exist in the conduction band.

Medical Imaging and Disease Detection

• Can be set to any arbitrary emission spectra to allow labeling and observation of detailed biological processes.

• Quantum Dots can be useful tool for monitoring cancerous cells and providing a means to better understand its evolution.

• In the future, Qdots could also be armed with tumor-fighting toxic therapies to provide the diagnosis and treatment of cancer.

• Qdots are much more resistant to degradation than other optical imaging probes such as organic dyes, allowing them to track cellprocesses for longer periods of time.

• Quantum dots offer a wide broadband absorption spectrum while maintaining a distinct, static emission wavelength.

Figure 5. Solutions of quantum dots of varying size. Note the variation incolor of each solution illustrating the particle size dependence of the optical absorption for each sample. Note that the smaller particles are in the blue solution (absorbs blue), and that the larger ones are in the red (absorbs red).

Quantum Dot LEDs

• Used to produce inexpensive, industrial quality white light.• Marked improvement over traditional LED–phosphor

integration by dot’s ability to absorb and emit at any desired wavelength.

• Produce white light by intermixing red, green, and blue emitting dots homogenously within the phosphor difficult to accomplish with the traditional LED-phosphor set up.

Solar Cells and Photovoltaics

• Traditional solar cells are made of semi-conductors and expensive to produce. Theoretical upper limit is 33% efficiency for conversion of sunlight to electricity for these cells.

• Utilizing quantum dots allows realization of third-generation solar cells at ~60% efficiency in electricity production while being $100 or less per square meter of paneling necessary.

• Effective due to quantum dots’ ability to preferentially absorb and emit radiation that results in optimal generation of electric current and voltage.

Other Future Quantum Dot Applications…

• Anti-counterfeiting capabilities: inject dots into liquid mixtures, fabrics, polymer matrices, etc. Ability to specifically control absorption and emission spectra to produce unique validation signatures. Almost impossible to mimic with traditional semi-conductors.

• Counter-espionage / Defense applications: Integrate quantum dots into dust that tracks enemies. Protection against friendly-fire events.

• Research continues. The possibilities seem endless…

The Silverson wayFor over 60 years Silverson has

specialized in the manufacture

of quality high shear mixers for

processing and manufacturing

industries worldwide.

With customers in over

150 countries, and serving

industries as diverse as food,

pharmaceuticals, cosmetics,

luboils and petrochemicals,

Silverson has become the

world leader in the field of

high shear mixing. Time after

time, companies specify

Silverson mixers as the

“standard” equipment for their

manufacturing process.

The key to this success is based

on the professionalism and

commitment Silverson shows

to each of its customer’s needs.

Whether supplying machines

from its standard range of

mixers or designing equipment

specifically to meet an individual

customer’s requirements, quality

is guaranteed.

With a customer base that

includes many of the world’s

largest companies, Silverson is

constantly at the forefront of new

technologies. Developing and

applying new high shear mixing

techniques to meet these needs,

Silverson has the experience,

knowledge and commitment to

both quality and service to solve

today’s mixing needs and those

of the future.

A truly international company,

Silverson is represented by

a network of associated

companies, distributors and

agents in over 50 countries,

serving North America, Europe,

Asia, Australasia, South America

and Africa.

IntroductionThe first name in high shear mixers

www.silverson.com3

The Silverson advantageSpeedThe exceptionally rapid Silverson

mixing action substantially

reduces process times compared

with conventional agitators and

mixers, and can reduce mixing

times by up to 90%.

VersatilityThe advantage of the Silverson

approach to mixing is that any

one machine can perform the

duties that in the past may have

required several different pieces

of process equipment. This

unrivalled versatility allows any

machine to perform the widest

range of mixing applications:

Blending

A homogeneous product is

rapidly produced when blending

liquids of similar or greatly

varying viscosities, eliminating

problems such as stratification.

Emulsifying and homogenizing

Emulsions (typically in the range

of 0.5 to 5 microns) can be

easily achieved.

Disintegration

All Silverson rotor/stator mixers

can disintegrate matter of

animal, vegetable, mineral

or synthetic origin in a

single operation.

Particle size reduction

Uniformly mill both solid and

semi-solid materials into either

solution or fine suspension.

Gelling and solubilizing

The high shear action of the

Silverson rotor/stator can rapidly

disperse gums, alginates,

C.M.C., carbopols, etc.,

resulting in an agglomerate-free

solution within minutes.

AdvantageThe first name in high shear mixers

How the Silverson worksThe advantages of Silverson’s

high shear rotor/stator mixer

over simple conventional stirrers

or agitators stem from the

multistage mixing/shearing

action as materials are drawn

through the specially designed

Silverson workhead - the heart

of every machine.

Stage 1The high-speed rotation of the

rotor blades within the precision

machined mixing workhead

exerts a powerful suction,

drawing liquid and solid materials

upwards from the bottom of

the vessel and into the center

of the workhead.

Stage 2Centrifugal force then drives

materials towards the periphery

of the workhead where they

are subjected to a milling action

in the precision machined

clearance between the ends of

the rotor blades and the inner

wall of the stator.

Stage 3This is followed by intense

hydraulic shear as the materials

are forced, at high velocity, out

through the perforations in the

stator and circulated into the

main body of the mix.

Stage 4The materials expelled from the

head are projected radially at

high speed towards the sides of

the mixing vessel. At the same

time, fresh material is continually

drawn into the workhead

maintaining the mixing cycle.

The effect of the horizontal

(radial) expulsion and suction into

the head is to set up a circulation

pattern that minimizes aeration

caused by the disturbance of the

liquid’s surface.

Interchangeable heads and screensA comprehensive range of

workheads and screens is

available for all Silverson rotor/

stator mixers.These easily

interchangeable workheads

offer great versatility by allowing

any machine to be adapted to

perform a wide range of mixing

operations including emulsifying,

homogenizing, disintegrating,

dissolving, dispersing, blending,

particle size reduction and

de-agglomerating. Changing

from one head or screen to

another is quick and simple.

www.silverson.com5

General purpose disintegrating headThis is the most versatile

of all the heads, giving an

exceptionally vigorous mixing

action. Ideal for general mixing

applications, its uses also

include the disintegration of

solids and the preparation of gels

and thickeners, suspensions,

solutions and slurries.

Slotted disintegrating headFor the disintegration of fibrous

materials such as animal and

vegetable tissue, as well as the

disintegration and solubilization

of “elastic” materials such as

rubbers and polymers.

Square hole high shear screen™

Provides exceptionally high

shear rates ideal for the rapid

size reduction of soluble

and insoluble granular solids.

It is also suitable for the

preparation of emulsions and

fine colloidal suspensions.

Standard emulsor head and emulsor screenSuitable for liquid/liquid

preparations and especially

useful for all emulsions. Emulsor

screens are available in fine,

medium or coarse perforations.

www.silverson.com21

How the Ultramix worksStage 1As the mixer rotates at high

speed, the propulsion grooves

on the outside of the mixing

head project the surrounding

product outwards, creating an

area of low pressure outside the

head. This draws product from

within the mixing head through

the radial slots to the outside.

Stage 2The material is subjected to a

shearing action as it passes

through the slots and is

projected back into the body

of the mix with great force.

Fresh material is simultaneously

sucked in through the holes in

the top and bottom of the mixing

head, setting up a vigorous flow

pattern throughout the vessel.

Mixing StationThe Ultramix is ideal for mixing

in problematic containers such

as IBCs and Tote Bins - the

highly efficient in-tank movement

provided by the small diameter

workhead overcomes the

problems often encountered

with difficult vessel geometry.

Silverson offers a “Mixing

Station” for this type of

application equipped with a

range of features and options:

- Electric raise and lower

mechanism.

- Guarded shaft.

- Vessel proximity sensor and

height interlock.

- Optional safety interlocked

cage.

- A special hopper allowing

powder to be fed directly into

the container can be offered

for some applications.

www.aspbs.com/enn

Encyclopedia ofNanoscience andNanotechnology

Pharmaceutical Nanotechnology

Paul A. McCarron, Maurice Hall

The Queen’s University of Belfast, Belfast, Ireland

CONTENTS

1. Introduction

2. Definition of Nanoparticles

3. Preparation

4. Characterization

5. Drug Loading

6. Incorporation of Water-Soluble Drugs

7. Incorporation of 5-Fluorouracilinto Nanoparticles

8. Drug Release

9. Conclusions

Glossary

References

1. INTRODUCTIONThe dream of delivering a pharmaceutically active moleculeto a specific site in the body has been a long-held aspirationwith beginnings that may be traced back to Paul Erhlich,who in the early 20th century coined the phrase “magic bul-let” to describe such an entity [1]. Today, extensive pharma-ceutical research has led to the development of drug deliverysystems and strategies, which go some way to fulfilling thisidea, but few which could be described as “magic bullets.”Side-effects and toxicities still afflict these approaches and,hence, Erhlich’s visionary thinking has not yet been fullyrealized. This is especially relevant in tumor chemotherapy,where selective delivery to neoplastic cells in comparison tosurrounding normal cells is an important principle [2]. Ascan be seen in Figure 1a, the challenge faced in site-specificdelivery of drugs is immense due to the numerous obstaclesbarricading the drug along its desired route. Cellular struc-tures and indeed the very components of the cell itself willeither prevent or act in some selective manner to hinder tothe migration of drug from its point of administration to theintended destination site.

On moving forward into the 21st century, it is apparentthat modern medicine still faces many challenges. Nano-technology is indeed one area that may offer scientificadvances in the coming years, which could lead to significantprogress in the improvement of therapeutic outcomes. Inparticular, the development of nanoparticulate drug deliv-ery systems may enhance the probability of getting a drug toits target site [3]. Instead of relying on the physicochemicalproperties of the drug to dictate its biodistribution, the drugis incorporated as a payload into a particle resulting in a dif-ferent transit mechanism for the drug after administration[4]. This can be enhanced further by virtue of the flexiblenature of the nanoparticle scaffold, on to which subsectionsmay be chemically bolted, producing a tailor-made and mul-tifaceted device. This can be represented in Figure 1b, whichgives the blueprint of an idealized nanoparticulate deliverysystem that must make its way to the target cell.

The particle in Figure 1b has several properties which areincorporated onto the particle, mostly by covalent bondingto surface groups. A targeting system, such as a monoclonalantibody, will recognize binding sites that are unique to thetarget cell and allow the particle to dock onto the exposedsurface [5]. This has been the focus of much current researchinto developing strategies for targeting nanoparticles to thesite of drug action [6–8]. A fusion protein will instigatethe process of merging with the target cell, thereby bringingthe particle into the cytoplasm [9]. As polymeric nano-particles are recognized as foreign by the body’s immune sys-tem, they are removed quite effectively by phagocytosis onexposure to the endoreticular system [10]. This will preventthe particle from reaching the target site and must be pre-vented. Steps toward this goal have already been taken withthe production of so-called “stealth” nanoparticles. Theseare nanoparticles which incorporate a biomimetic polymer,usually polyethylene glycol, into their structure to avoid elic-itation of an immune response [11–13]. Presently, such anidealized nanoparticle with these three important proper-ties has yet to be realized, but attempts have been made toattach some of the subsystems described.

The construction of an idealized particle must start withthe formation of a solid polymeric sphere with a diameter

ISBN: 1-58883-064-0/$35.00Copyright © 2004 by American Scientific PublishersAll rights of reproduction in any form reserved.

Encyclopedia of Nanoscience and NanotechnologyEdited by H. S. Nalwa

Volume 8: Pages (469–487)

470 Pharmaceutical Nanotechnology

Oraladministration

Intravenousadministration

intestinalwall

connectivetissue

vascularendothelium

targetcell

drugpayload

particle permeation

(a)

(b)

Figure 1. (a) A simplified overview depicting the barriers to drug local-ization at a target cell after oral administration into the gut or intra-venous administration to the systemic circulation. For example, thepath from a capillary to a target cell may involve crossing the vascularendothelium and various connective tissue structures. (b) An idealizeddraft of a multifaceted nanoparticulate system containing a drug pay-load that must also permeate in some way to the target cell. Pendentmolecules, such as a targeting ligand, add to the functionality of thedevice.

of approximately 200 nm. This is followed closely by load-ing the drug into the naked particle. This is dependent onthe physicochemical properties of the drug molecule, whichmay often present problems in terms of efficient loading,especially if the drug has appreciable water solubility. Thisscenario may not be uncommon given that the majority ofcandidate molecules for inclusion in nanoparticulate systemswill possess some degree of aqueous solubility. This chap-ter discusses the construction of nanoparticles and describestheir characterization. In particular, the process of incorpo-rating a water-soluble payload is discussed and the problemsassociated with this are used to illustrate the complexitiesencountered during the proposed construction of a device,as typified in Figure 1b.

2. DEFINITION OF NANOPARTICLESNanoparticles have been studied extensively as carriers fordrugs employed in a wide variety of routes of administra-tion, including parenteral [14], ocular [15], and peroral [16]pathways. The term nanoparticle is a collective name for anycolloidal carrier of submicrometer dimension and includesnanospheres, nanocapsules, and liposomes. They can all bedefined as solid carriers, approximately spherical and rang-ing in size from 10 to 1000 nm. They are generally polymericin nature (synthetic or natural) and can be biodegradable

or nonbiodegradable in character. Biodegradable polymersused include poly(alkylcyanoacrylates) [17, 18], poly(lactide-co-glycolide) [19, 20], chitosan [21, 22], and gelatin [23],while nonbiodegradable polymers include poly(styrene) [24],poly(acrylamide) [25], and poly(methylmethacrylate) [26].Nanospheres represent the simplest carrier and are solid,monolithic systems in which the drug is dissolved orentrapped throughout the particle matrix. Alternatively, itmay be adsorbed onto the surface. No continuous mem-brane surrounds the particle, as illustrated in Figure 2a.

Nanocapsules are reservoir type systems comprising anoily liquid core surrounded by a polymeric shell [27]. Thedrug is usually dissolved in this liquid core but may be moreclosely associated with the shell polymer and the exposedsurface, as illustrated in Figure 2b.

Liposomes are closely related to nanocapsules in struc-tural layout but consist of an aqueous core surrounded bya bilayer membrane composed of lipid molecules, such asphospholipids [28], as illustrated in Figure 2c. The drug canbe located in the aqueous core or in the bilayer membrane.

drug dispersed in particle drug adsorbed on to surface

drug molecule

drug dispersed in oily coredrug adsorbed on to surfaceor dispersed in shell

drug molecule

hydrophyllic drug inaqueous core

hydrophobic drug inlipid bilayer

(a)

(c)

drug molecule

(b)

Figure 2. (a) The morphology of nanospheres, where drug can be eitherdispersed throughout the polymeric matrix of the particle or adsorbedonto the surface. (b) The morphology of nanocapsules, where drug isdispersed mostly in a liquid core. Drug can also be associated withthe polymeric shell, either by dispersal through the polymer or adsorp-tion to it. (c) The morphology of a small single lamellar liposomeof submicrometer dimensions showing the possible location of a drugpayload.

Pharmaceutical Nanotechnology 471

From a historical prospective, liposomes have been themost extensively studied nanoparticulate carrier [29–32].However, their full development leading to extensive clin-ical use has been restricted by pharmaceutical problemsentailing drug leakage, stability, and problems with scale-up procedures. These all arise due to their liquid andsemisolid nature leading to a lack of rigidity and sub-sequent fragility. However, the important early work byCouvreur and co-workers [33] introduced nanoparticles pre-pared from biodegradable poly(alkylcyanoacrylates). Thishas prompted much research into nanoparticles made frompolymeric materials and, furthermore, into finding polymersthat are pharmaceutically acceptable, biodegradable, andnontoxic [34]. A wide variety of preparation methods fornanoparticles have now evolved, which encapsulate manydifferent types of drug compounds [35, 36].

3. PREPARATIONBoth nanospheres and nanocapsules are prepared fromeither a polymerization reaction of dispersed monomers orfrom a solvent dispersion procedure using preformed poly-mers. In many instances, the latter procedure using pre-formed polymer is desirable, as potential reactions betweendrug and monomer are avoided and the potential toxicity ofresidual monomers, surfactant, and initiator is reduced [37].The final properties of nanoparticles, such as their size, mor-phology, drug loading, release characteristics, and biodisti-bution, are all influenced by the method of preparation [38].

3.1. Nanoparticles Preparedfrom Polymerization Reactions

Most examples of polymerization used to create nano-particles occur by a free radical mechanism involving dis-tinct initiation, propagation, and termination processes [39].Polymerization occurs within a continuous liquid medium,which also comprises the monomer, initiator, and a surfac-tant. Four different polymerization techniques are describedto polymerize vinyl type monomers, namely:

• emulsion polymerization, resulting in 0.05–0.2 �m par-ticles

• dispersion polymerization, resulting in 0.3–10 �m par-ticles

• precipitation polymerization, resulting in nonspherical0.1–10 �m particles

• suspension polymerization, resulting in 20–2000 �mparticles

3.1.1. Emulsion Polymerization in aContinuous Aqueous Phase

Continuous aqueous phase emulsion polymerization is oneof the most widely used procedures to make nanoparticlesfor drug delivery purposes, especially those prepared fromthe alkylcyanoacrylate monomers. An oil-in-water emulsionsystem is employed where the monomer is emulsified in acontinuous aqueous phase containing soluble initiator andsurfactant [39, 40]. Under these conditions, the monomer ispartly solubilized in micelles (5–10 nm), emulsified as large

monomer droplets (1–10 �m), and partly dissolved in thecontinuous aqueous phase. The polymerization takes placein the outer continuous phase and not in the micelle inte-rior [41]. Initiation of polymerization occurs in the aqueousphase when a monomer molecule collides with an initia-tor molecule, which may be an ion or a free radical [42].Alternatively, the monomer can be struck by a high-energyphoton, such as gamma radiation and ultraviolet or intensevisible light. Propagation occurs when nascent monomericradicals interact with other monomer molecules forming pri-mary particles [39]. The polymerization is maintained by thecontinuous supply of monomer from the monomer reser-voirs, such as droplets and micelles, and is illustrated inFigure 3.

The surfactant is an important component of this pro-cess and acts to stabilize the growing polymeric particlesby surface adsorption. Phase separation and the formationof solid particles occur before or after termination of thepolymerization process [42]. Polymerization can occur insome systems without the presence of surfactants [40]. Vari-ous particulate systems have been prepared by this method,including poly(styrene) [43], poly(vinylpyridine) [44, 45],poly(acrolein) [46, 47], and poly(glutaraldehyde) [48–50].

Considerable work has been done using poly(alkylcyano-acrylate) (PACA) nanospheres due to their biodegradability,bioelimination, ease of preparation, low toxicity, and stabil-ity [18, 51, 52]. Although initially used as hemostatic aidsin surgery [53, 54], their preparation was first recorded byCouvreur et al. [33]. Monomer was emulsified in a stirred,acidic, aqueous medium containing a nonionic surfactant,as illustrated in Figure 4. The suspension obtained dis-plays a blue coloration due to the scattering of light bythe particles known as the Tyndall effect. Glucose, Tween,Dextran 70, and Pluronic F68 have been used to stabilizeand protect the particles from agglomeration [52]. Variousmonomers are available, such as methyl, ethyl, butyl, andisohexyl monomers, to prepare PACA nanoparticles, whichdiffer in the length of their alkyl side chains.

The polymerization is an anionic mechanism initiatedby hydroxide ions or any bases present [42]. The reactionscheme can be seen in Figure 5. The polymerization rate isregulated by hydroxyl ion concentration and hence is car-ried out at pH values below 3.5. Above this pH, the reactionrate is too rapid to allow discrete particle formation [55, 56].

monomerdroplet

nanoparticle

monomers in micellesgrowing polymerradical

dissolvedmonomers

Figure 3. Presentation of the emulsion–polymerization mechanism inan aqueous phase.

472 Pharmaceutical Nanotechnology

monomer addeddropwise

acidic polymerizationmedium containinga surfactant

mechanical stirrer

nanospheres formed anddisplaying the Tyndall effect

Figure 4. Preparation conditions used in the preparation of poly(alkyl-cyanoacrylate) nanoparticles.

The main advantage of this polymerization is that it can becarried out at room temperature without an energy input.PACA nanoparticles have been prepared reproducibly at asemi-industrial level and can be made aseptically [57].

Spherical particles of around 200 nm in diameter are pre-pared by this technique as demonstrated by scanning elec-tron microscopy [58, 59]. Particles as small as 10 nm wereobtained by altering the concentration of surfactant andby increasing amounts of sulphur dioxide [60, 61]. Freezefracture studies confirmed that PACA nanospheres have adense solid spherical core with no distinguishable surround-ing membrane [27]. The contents of the polymerizationmedium, such as pH, concentration, type of surfactant, andmonomer type have been shown to influence the particle size[55, 62–64]. A large variety of drugs, such as cytostatics andantibiotics, have been encapsulated in PACA nanospheressuccessfully, in part due to their porosity and high specificsurface area [35, 65–68]. Drugs are incorporated into thesenanospheres during polymerization or adsorbed onto thesurface of preformed particles [52]. The loading of lipophilicdrugs is limited by their solubility in the aqueous acidic con-tinuous phase. Acid labile drugs can be incorporated usingpoly(dialkylmethylidenemalonate) esters as they polymerizeat neutral pH [69, 70]. This is due to the presence of the lesselectrophilic alkoxycarbonyl group compared to the cyano

CH2C

CN

CO2R

HO HO CH2 C

CO2R

CN

HO CH2

n

C

CO2R

CN

CH2 C

CO2R

CN

HHO CH2

n

C

CO2R

CN

H

δ+ δ−- -

-

Furthermonomer

+

n+1 n

Figure 5. Anionic polymerization mechanism for alkylcyanoacrylatemonomers (R = n-butyl).

group of the alkylcyanoacrylates. Their use, however, hasbeen limited due to slow biodegradability [71].

3.1.2. Emulsion Polymerization in aContinuous Organic Phase

This procedure involves emulsifying a water-soluble mono-mer in a continuous organic phase containing a solubleinitiator and surfactant. This constitutes a water-in-oilemulsion, as the different phases are reversed from themore common continuous aqueous phase based methods.The first process reported for nanoparticle formation byBirrenbach and Speiser is an example of this procedure[72]. Poly(acrylic) nanoparticles were prepared by dissolv-ing acrylamide and N ,N ′-methylene bizacrylamide in anouter hexane phase containing anionic surfactants. The poly-merization was initiated by gamma, ultraviolet, or visibleradiation. Biologically active, antigenic materials wereencapsulated and showed intact biological activity and highantibody production after encapsulation [72]. Spherical par-ticles of 80 nm were produced based on electron microscopicstudies. Kreuter [59] reported that transmission electronmicroscopy (TEM) revealed these particles to have a soliddense interior, forming matrix type nanopellets. Ekmanand Sjoholm [73] investigated the polymerization of thesemonomers by radical initiation in a toluene/chloroform mix-ture, but higher particle sizes were obtained.

The main disadvantage in using poly(acrylamide) sys-tems is that they are not biodegradable and the monomersare toxic. Extensive purification is also required to removethe organic solvents, anionic surfactants, and residualmonomers. Edman et al. [74] produced biodegradablepoly(acryldextran) particles by incorporating dextran intothe poly(acrylamide) chain. These particulate systems weremetabolized and eliminated faster, both in vivo and in vitro,than poly(acrylamide) particles.

3.1.3. Dispersion PolymerizationIn dispersion polymerization, the monomer and initiator aredissolved in the continuous phase, which acts as a nonsol-vent for the developing polymer. The continuous phase canbe organic, aqueous, or a mixture of miscible phases. Twomethods of initiation have been employed, including gammaradiation [75] and chemical initiation by potassium perox-odisulphate [76]. As the polymer is formed, it precipitates asnanoparticles. These particles are not polymeric precipitatesas in precipitation polymerization. Rather, they are swollenby a mixture of the monomer and the continuous phase [39].

Kreuter and Speiser [77] developed a dispersion poly-merization producing adjuvant nanospheres of poly(methyl-methacrylate) (PMMA). The monomer is dissolved inphosphate buffered saline and initiated by gamma radi-ation in the presence and absence of influenza virions.These systems showed enhanced adjuvant effect over alu-minum hydroxide and prolonged antibody response. PMMAparticles could be distinguished by TEM studies and theparticle size was reported elsewhere to be 130 nm by pho-ton correlation spectroscopy [75]. The particle size couldbe reduced, producing monodisperse particles by inclu-sion of protective colloids, such as proteins or casein [40].Poly(methylmethacrylate) nanoparticles are also prepared

Pharmaceutical Nanotechnology 473

by chemical means using potassium peroxodisulphate as ini-tiator combined with heating at 65 or 85 �C [76]. Increas-ing monomer concentration and temperature resulted in anincrease in particle size.

Copolymerization of PMMA has been carried out for thepurpose of producing more hydrophilic particles [45, 78, 79].Methylmethacrylate has been copolymerized with metha-crylic acid, ethylene glycol dimethacrylate, hydroxypropyldimethacrylate, or 2-hydroxyethylmethacrylate, by emulsionpolymerization. Particle sizes from 30 to 340 nm in diame-ter can be produced by chemical initiation and 0.3–3.0 �mparticles produced using gamma radiation [45, 80]. Anti-bodies, amino acids, and fluorescent molecules can becovalently bound to the active surface groups on these parti-cles by cyanogen bromide, carbodiimide, and glutaraldehydemethods [78, 80]. The polymerization of PMMA and itscopolymers avoids the use of organic solvents and anionicsurfactants. However, PMMA is not biodegradable and,thus, persists in the body [18, 81]. The hydrolysis of thesepolymers requires an acid environment not generally foundin body fluids.

3.2. Nanospheres Preparedfrom Preformed Polymers

Synthetic and natural polymers have been investigated whichare biodegradable and biocompatible. The nanospheres areformed by precipitation of synthetic polymers or by denatu-ration/solidification of polymers of natural origin. Four tech-niques have been reported for preparing nanoparticles fromsynthetic preformed polymers. These include:

• emulsion–evaporation,• salting out technique,• emulsion–diffusion,• coacervation/phase separation technique.

3.2.1. Emulsion–EvaporationThis method was first reported by Vanderhoff [82] forthe preparation of artificial latexes. The polymer and drugare dissolved or dispersed in a volatile water-immiscibleorganic solvent, such as dichloromethane, chloroform, orethyl acetate. This is emulsified in an aqueous continuousphase containing a surfactant, such as poly(vinylalcohol), toform nanodroplets. The organic solvent diffuses out of thenanodroplets into the aqueous phase and evaporates at theair/water interface, as illustrated in Figure 6. The solvent isremoved under reduced pressure. The nanodroplets solidifyand can be separated, washed, and dried to form a free-flowing powder.

Poly(d,l-lactic acid) (PLA) nanospheres containingtestosterone, with a particle size of 450 nm, were first pro-duced by Vanderhoff et al. [83]. Microfluidization producedspherical particles of less than 200 nm with a narrow sizedistribution [84]. Krause et al. [85] produced PLA nano-particles of 500 nm using sonication. The polymer and drug,triamcinolone acetonide, were dissolved in chloroform andemulsified, with sonication, for 45 minutes at 15 �C in agelatin solution. The solvent was evaporated by heatingto 40 �C for 45 minutes under continuous stirring. TEM

polymer and drug dissolvedin water miscible solvent

aqueous continuousmedium containinga surfactant

mechanical stirrer

mechanical stirrer

evaporation underreduced pressure

1 semisolid nanodroplets formed

2 diffusion of solvent from nanodroplet particle

3 solid nanoparticles formed

Figure 6. Preparation scheme for the emulsion–evaporation method.

showed these particles to have a highly porous interior struc-ture. Scholes et al. [86] and Julienne et al. [87] optimizedthe particle size of poly(lactide-co-glycolide) (PLGA) nano-spheres by altering preparation conditions in the emulsion–evaporation method. These conditions included the phasevolume ratio of the emulsion formed, concentration andmolecular weight of the polymer and surfactant, as wellas homogenization time and pressure. Tice and Gilley [88]reported that size of the particles formed depended on theinitial size of the emulsion droplets.

It is desirable to use good film forming polymers for theemulsion–evaporation process [88]. For example, Eudragit®

RL, Eudragit® RS, ethyl cellulose, PMMA, and celluloseacetate butyrate polymers formed nanoparticles by thismethod [89]. Biodegradable poly(�-hydroxybutyrate) hasbeen used to prepare nanospheres by this process [90, 91].Particle sizes from 170 to 220 nm were obtained using highpressure emulsification with Brij® surfactants [90].

The successful entrapment of drugs by this preparationprocedure involves minimizing the loss of drug to the aque-ous continuous phase. Various strategies employed includethe use of drugs with low water solubility in the continu-ous aqueous phase, a high concentration of polymer in theorganic phase, and a fast precipitation rate of the polymer inthe continuous phase [92, 93]. By pH adjustment of the con-tinuous aqueous phase, the loss of ionizable drugs, such asquinidine, can be reduced [94]. The emulsion–evaporationmethod is most suited to the encapsulation of lipophilicdrugs, such as hydrocortisone, progesterone, diazepam, andindomethacin [84, 94].

Adding water miscible solvents to the organic phaseenhanced the encapsulation of water-soluble drugs due torapid polymer precipitation in the aqueous phase [93, 95].Bodmeier and McGinity [93] investigated using co-solvents

474 Pharmaceutical Nanotechnology

such as acetone, ethyl acetate, methanol, and dimethyl-sulphoxide (DMSO) with PLA to increase quinidine load-ings. Niwa et al. [95] produced nanospheres of PLA andPLGA enhancing peptide encapsulation by incorporatingacetone with DCM. Organic phases have been employedto improve the encapsulation of water-soluble drugs by theemulsion–evaporation method. Tsai et al. [96] first reportedusing light mineral oil as the continuous phase with Span 65as the surfactant, but microspheres were produced. This wasalso reported by other authors [38, 97, 98].

3.2.2. Salting Out TechniqueThe salting out technique was first proposed by Bind-schaedler et al. [99] This preparation method involvesadding a water-soluble polymer, such as poly(vinylalcohol)(PVA), to a concentrated solution of an electrolyte or non-electrolyte forming a viscous gel. This is emulsified in awater-miscible solvent, such as acetone containing the poly-mer and drug. The saturated aqueous phase prevented ace-tone from mixing with the water by a salting out process.On further addition of the gel to the organic phase, an oil-in-water emulsion is formed. Sufficient water is then addedto allow complete diffusion of the acetone into the aqueouscontinuous phase, thereby forming nanospheres [100, 101].Typically, magnesium chloride and magnesium stearate areused as salting out agents. This method avoids the use ofsurfactants and chlorinated solvents, which offers distinctadvantages over the emulsion–evaporation method. How-ever, the hydrocolloidal PVA is employed as a viscosityincreasing agent and stabilizing agent. The PVA, electrolyte,and acetone were removed from the nanospheres by crossflow filtration, as described by Allémann et al. [101].

The preparation conditions of the salting out techniquecould be altered to produce nanospheres of monomodaldistributions with particle sizes in the range 170 to900 nm using Eudragit® S as polymer [102]. The con-ditions investigated include homogenization, stirring rate,internal/external phase ratio, viscosity of the external phase,concentration of polymer in the organic phase, type of salt-ing out agent, and concentration and type of stabilizingagent in the aqueous phase. The size of the droplets in theemulsion determined the size of the final particles [102].Leroux et al. [103] prepared PLA particles of 70 nm usingbenzyl alcohol and high PVA concentrations.

Different polymers have been used to produce nano-particles by the salting out process, such as Eudragit® S,Eudragit® E, poly(dl-lactic acid), polycaprolactone, ethylcellulose, cellulose acetate and cellulose acetate butyrate[99, 100, 102, 103]. Other salting out agents, such as sucrose,and solvents, such as THF, ethyl acetate, and isopropyl alco-hol, also produced nanoparticles [102]. The salting out tech-nique was found to entrap lipophilic drugs successfully, suchas savoxepine and chlorambucil, due to the continuous aque-ous phase used [35, 103, 104]. Leroux et al. [103] com-mented on the pharmaceutical acceptability of the saltingout process as the use of acetone and large amounts of saltmay pose problems concerning salt recycling and compati-bility with some drug substances. These authors used benzylalcohol as the organic solvent and low amounts of salt andgelatin as the stabilizing hydrocolloid to improve the phar-maceutical acceptability.

3.2.3. Emulsion–DiffusionThis procedure is also referred to as the precipitationmethod and was first reported by Fessi et al. [105]. In thistype of preparation, as shown in Figure 7, the polymer isdissolved in a water-miscible solvent and then poured understirring into a nonsolvent, which is usually water. This leadsto the polymer precipitating as nanospheres. Prior emulsifi-cation and inclusion of surfactants are not necessary.

The proposed mechanism for formation of nanospheresby this technique is by a diffusion and stranding mecha-nism, as illustrated in Figure 10 (Section 4.2) [106]. Polymerglobules are formed as the polymer solution is added tothe aqueous phase. As the solvent diffuses into the waterfrom the polymer globule, the polymer precipitates out ofsolution and is stranded in the water in the form of fineemulsion droplets or “protonanoparticles.” Complete diffu-sion of solvent from these protonanoparticles will provokepolymer solidification in the form of nanospheres of about200 nm in diameter. Turbidity measurements are consistentwith this mechanism [106]. In summary, the diffusion of thesolvent causes regions of supersaturation from which nano-particles are formed, due to phase transformation in theseregions. The size of the resulting particles is dependent onthis process.

This precipitation procedure has been successfully appliedto a wide variety of polymers such as poly(lactide), poly-(lactide-co-glycolide), poly-(�-caprolactone), ethylcellulose,poly(alkylcyanoacrylate), and poly(styrene) [107]. Morerecent biodegradable polyesters, called poly(�-malic acid-co-benzyl malate) copolymers, have formed nanospheresof 200 nm in diameter by the precipitation method[108, 109]. Solvent/nonsolvent mixtures that have been usedinclude ethanol/water, acetone/water, and propylene car-bonate/water [110]. Murakawi et al. [19] employed solventmixtures, such as ethanol/acetone and methanol/acetone,providing high yields of PLGA nanometer sized particleswith less aggregation. This preparation method is best suitedto the encapsulation of lipophilic drugs as an aqueous con-tinuous phase is used [111]. The lipophilic drugs are usuallydissolved in with the polymer-rich phase. High encapsula-tion was reported for peptides, such as TRH and elcatonin,with this method [112].

polymer and solvent addeddropwise under constantstirring

precipitating aqueouscontinuous medium

mechanical stirrer

instant formationof nanospheres

polymerglobule

solidifiednanosphere

diffusion of solventfrom proto-nanoparticles

supersaturationregion

Figure 7. Preparation procedure for precipitating technique. Shownalso is the schematic illustration of the diffusion stranding mechanismfor nanosphere formation that occurs during the precipitation method.

Pharmaceutical Nanotechnology 475

A major advantage of this procedure is that a surfac-tant is not required. This avoids its removal by purifica-tion procedures. However, poly(vinyl alcohol) has been usedas a stabilizer in some instances and was found to adsorbstrongly, resisting washing procedures [19, 106]. Without sur-factant, nanoparticle formation and stability are dependenton the initial polymer concentration, the volume ratio ofthe solvent to nonsolvent, and the dielectric constant of thefinal mixture [111]. Ternary phase diagrams have been con-structed to find the optimal region of nanoparticle formation[110, 111]. Another advantage of this procedure from anindustrial point of view is that high-pressure homogenizersand ultrasonication are not required as spontaneous emulsi-fication occurs [106].

3.2.4. CoacervationCoacervation is a phase separation technique. It was firstintroduced by Bungenberg de Jong and Kruyt [113] whodescribed a process in which aqueous colloidal solutionsare separated into two liquid phases, a colloid-rich phaseor coacervate and a colloid-poor phase. Many syntheticpolymers are water-insoluble and consequently, nonaque-ous coacervation systems have been employed. Coacervationusually consists of binary and ternary systems [114, 115].Binary systems consist of one polymer dissolved in a solventand ternary systems of either a single polymer in a binarysolvent mixture or two polymers in a single solvent. Forbinary systems, phase separation can occur by changing thetemperature or pH or by the addition of electrolytes, suchas sodium sulphate, to gelatin [116]. For ternary systems,phase separation can be induced by adding a nonsolvent fora single polymer system or by adding a polymer incompati-ble with the second polymer for a two polymer system [117].Solidification of the particles produced by coacervation canbe effected by temperature, chemical cross-linking, adjust-ment of pH, or rinsing with a nonsolvent for the polymer.

The coacervation process is controlled by the interactionsbetween polymer/solvent/nonsolvent. The formation of sta-ble coacervates is influenced by the polymer type, molec-ular weight, and hydrophobicity and by the amount andviscosity of the added coacervating agent [118, 119]. By care-ful adjustment of these parameters, optimum coacervationconditions can be achieved by the use of phase diagrams.The classical coacervation method for nonaqueous systemshas been modified to produce nanoparticles by using anemulsion-phase separation method in an oil system [120].An aqueous solution of the drug, a peptide in this instance,was emulsified in a DCM–acetone mixture containing thedissolved PLGA using homogenization [112, 120]. Triesteroil (caprylate and caprate triglyceride) was added gradu-ally, inducing phase separation of the PLGA at the interfacebetween the aqueous and oily phase. The aqueous dropletsare covered with the coacervate droplets of PLGA and theevaporation of the organic solvent reduces the solubility ofthe PLGA and deposits it around the aqueous droplets con-taining the drug. This mechanism is illustrated in Figure 8.

Nanoparticles encapsulating 5-flurouracil were preparedby coacervation using cellulose derivatives [121]. An ethano-lic solution of ethyl cellulose was desolvated by stirring indistilled water producing particles with an average diam-eter of 472 nm, as determined using scanning electron

W/O Coacervate-deposited droplets

Coacervation

Additionof Triester

Evaporation oforganic solvent

Nanospheres

AqueousPhase

OrganicPhase

Figure 8. Preparation mechanisms of PLGA nanospheres by theemulsion-phase separation method in an oil system.

microscopy (SEM). Nanospheres of methyl cellulose wereprepared by desolvating an aqueous solution of the polymerwith sodium sulphate solution [121]. An average particle sizeof 540 nm was obtained.

The main drawback to coacervation is the toxicologicalimplication of residual solvents, coacervating agents, andhardening agents left in the particles. More biocompatiblecoacervating and hardening agents are being investigatedalong with approaches to minimize solvent residues [114].The triester oil used in the coacervation of PLGA by Niwaet al. [120] is nontoxic and biocompatible and an alternativeto silicone oil usually used in the coacervation of PLGA.Biocompatible hardening agents suggested include fatty acidesters, such as isopropyl myristate for PLGA [114]. For lowmolecular weight polymers, drying near the glass transitiontemperature of the polymer was found to be effective inlowering DCM content [114].

3.3. Nanoparticle ProductionUsing Supercritical Fluids

Supercritical fluids are defined as those substances whosetemperature and pressure have been raised above the pointwhere the densities of the liquid and gaseous states are thesame. When substances exist in a supercritical state, theycan be used as solvents or antisolvents and, hence, used inthe production of particulates in a similar manner to othersolvent systems. The use supercritical fluids has increasedsignificantly over the last few years, with many drugs eitherbeing encapsulated within, or being formed into, nano-particles using this technique [122–125]. However, there area number of advantages to using supercritical fluids overconventional solvent systems. Production is usually a one-step process, with little polydispersity in particle size and,perhaps, most importantly, much less organic solvent is used,with carbon dioxide being the most frequently used super-critical fluid. This is obviously of significance environmen-tally, especially if production is on the commercial scale.

A number of techniques are based on supercritical fluidtechnology. Three are of particular pharmaceutical interest,namely the supercritical antisolvent (SAS) system, the rapidexpansion of supercritical solution (RESS) method, and thegas antisolvent (GAS) technique [126].

The SAS method involves the introduction of the super-critical fluid into a solution of solute in an organic solvent.

476 Pharmaceutical Nanotechnology

At high pressures, the fluid acts as an antisolvent, becom-ing soluble enough in the organic solvent to precipitate outthe solute. At the final operating pressure, the supercriticalfluid flows through the precipitation vessel stripping awayany residual organic solvent. Finally, the vessel is depressur-ized and the solid product is collected. The GAS technique isa modified version of this method, with the organic solutionbeing rapidly introduced into the supercritical fluid througha fine nozzle [127, 128]. The solution is quickly extractedinto the fluid, resulting in instantaneous precipitation of thesolute as fine particles.

In the RESS method, the solute of interest is solubi-lized in a supercritical fluid, which is then rapidly expandedthrough a nozzle. As the fluid expands, it loses its solventcapabilities and the solute precipitates out. While this tech-nique has the advantage of not using any organic solvent, itis restricted by the generally poor solubility of most polymersin supercritical fluids. Indeed, polymers generally have to bebelow 10,000 MW in order to be eligible for this method ofparticle production [126].

4. CHARACTERIZATIONThe physicochemical characterization of a colloidal carrieris necessary because important characteristics, such as par-ticle size, hydrophobicity, and surface charge, determine thebiodistribution after administration [129–132]. Preparationconditions, such as the pH of the polymerization medium,monomer concentration, and surfactant concentration, caninfluence the physicochemical characteristics of the particles[60, 62, 64]. It is, therefore, essential to perform a compre-hensive physicochemical characterization of nanoparticles,which has been reviewed by Magenheim and Benita [133].

4.1. Particle Size

Clearly, the definitive characteristic of any nanoparticulatedrug delivery system will be its submicrometer diameter. Siz-ing such particles in the suboptical region can be difficultas the measuring technique itself may alter size and proper-ties by either hydrating or aggregating the particles. This willhave a profound influence on the size of the particle [59].Haskell [134] has discussed the various optical techniquesavailable to measure the size of nanoparticles.

Photon correlation spectroscopy (PCS) has been usedextensively for the sizing of submicrometer particles and isnow the accepted technique in most sizing determinations.PCS is based on the Brownian motion that colloidal particlesundergo, where they are in constant, random motion due tothe bombardment of solvent (or gas) molecules surround-ing them. The time dependence of the fluctuations in inten-sity of scattered light from particles undergoing Brownianmotion is a function of the size of the particles. Smaller par-ticles move more rapidly than larger ones and the amountof movement is defined by the diffusion coefficient or trans-lational diffusion coefficient, which can be related to size bythe Stokes–Einstein equation, as described by

d�H = kT

3� D(1)

where D is the diffusion coefficient, k is the Boltzmann’sconstant, T is the absolute temperature, is the viscosity,and d�H is the hydrodynamic diameter. The hydrodynamicdiameter, or Stokes diameter, measures how a particlemoves within a fluid. To obtain mass or volume diameters,the full Mie theory should be used, which requires knowl-edge of the refractive indices of the particulate material andsizing medium. The upper limit of size for PCS is deter-mined by the density of the particle and, most importantly,its onset of sedimentation, rather than limitations on thetechnique itself.

PCS is a rapid, nondestructive technique which enablesthe measurement of many types of spherically shaped enti-ties, including discrete living cells, without causing alterationor damage [133]. However, the use of PCS is not possiblewhen the size distribution of the sample is broad, the samplematerial is nonspherical, and multimodal distributions arepresent. PCS is also susceptible to errors from the presenceof larger particles, such as dust, microbial contamination,crystallization of ingredients, or secondary particle agglom-erates. Larger particles scatter more light than smaller parti-cles, as shown by the Rayleigh ratio, and thus will swamp thescattered light from smaller particles. Samples for PCS mea-surements must be clean and filtered to ensure removal ofcontaminating dust particles. Another disadvantage of thismethod is that the particle size is influenced by the sur-rounding medium, such as adsorbed surfactants or hydrationlayers.

PCS has been applied in nanoparticulate characterizationto investigate the effect of altering preparation parameterson the particle size of various nanoparticulate carriers, suchas poly(alkylcyanoacrylates) [55, 60, 61, 135]. The effect ofthe incorporation of drugs on nanoparticle size has also beeninvestigated [136, 137]. PCS has been used to examine thedegradation of PACA nanoparticles [138]. This techniquehas been useful in determining the influence on size of usingdifferent surfactants in the preparation process and also thethickness of the surfactant coating layer on the surface ofthe nanoparticle [139, 140].

Electron microscopy techniques, such as SEM and TEM,have been used to measure the particle size of nano-particles. TEM has the additional advantage in that interiorparticle morphology can be determined by freeze fractur-ing [133]. TEM has been used to size poly(acrylic) acidand cellulose nanoparticles and it was shown that thesizes obtained were comparable to PCS [59, 121]. TEMhas been used to optimize the preparation conditions ofpoly(methylmethacrylate) nanoparticles to obtain small par-ticle sizes [141]. SEM requires the particles to be coated withgold to make them conductive. This coating varies in thick-ness from 30 to 60 nm and has to be subtracted in order todetermine the size of the uncoated particles. On comparisonto other sizing techniques, the particle sizes of poly(acrylic)acid nanoparticles by SEM were in good agreement [59].Although electron microscopy allows the measurement ofindividual particles, this may be unrepresentative of thewhole sample. The low vacuum used, gold coating, and anysurfactants present may alter the particle size [59].

Pharmaceutical Nanotechnology 477

4.2. Surface Charge

Particles in a suspending medium carry a thin layer of ionsand solvent around them, which will cause the particle todrift when placed in an electric field. The surface separat-ing the stationary medium from the moving particle andits bound ions and solvent is called the surface of hydro-dynamic shear [142]. The potential at this surface is calledthe zeta potential, as illustrated in Figure 9. This surfacepotential can originate from the dissociation of surface poly-meric groups or preferential adsorption of ions or otherionic molecules from the aqueous suspending medium. Thezeta potential of the particle is dependent on electrolyteconcentration and pH of the suspending particle medium[142, 143].

The zeta potential can be measured by electrophoresis,which determines the velocity of particles in an electric fieldof known strength [144]. This particle velocity, v, can thenbe related to the electrical field strength, E, as the elec-trophoretic mobility, �. This is shown by

� = v

E(2)

The electrophoretic mobility, �, can be converted to azeta potential by using the Smoluchowski equation,

Zp = �4�

�(3)

where Zp is the zeta potential (mV), � is the electrophoreticmobility [(cm/s)/(Volt/cm)], is the dynamic viscosity of thedispersion medium (poise), and � is the dielectric constant.

Electrophoretic mobility measurements can be performedby laser Doppler anemometry (LDA). LDA is fast and capa-ble of high resolution of particle velocities [144]. It mea-sures particle velocity, which is measured in the stationary

d

Shea

r Pl

ane

Diffuse Layer

Inner Helmholtz

Outer Helmholtz

Stern

Y

z

Pote

ntia

lPa

rtic

le

ZP

Y i

Y

Figure 9. Formation of Stern plane and diffuse layer on particle surface(�O = surface or Nernst potential, �i = potential of inner Helmholtzplane, �� = Stern potential, � = thickness of Stern plane, ZP = zetapotential at surface of shear, d = distance from particle surface).

layer, where the particles are moving solely with a velocitydue to their charge [142]. This stationary layer is defined bythe crossing of two laser beams, as illustrated in Figure 10.Interference fringes are produced by the crossing of the twobeams. Particles inside the scattering volume interact withthese fringes to produce scattered light, which is detectedby the photomultiplier tube. The frequency of the scatteredlight by the particles differs from the frequency of the inci-dent laser beam. This shift in frequency is caused by theDoppler effect and is a function of the particle velocity [142].

Zeta potential measurements have been used to deter-mine the stability of particle suspensions. Particles with largenegative or positive zeta potentials (+30 and −30 mV)will form stable colloidal suspensions as they tend to repeleach other and reduce aggregation [143, 145]. Zeta potentialmeasurements have been used to estimate particle opsoniza-tion, a process whereby specific proteins in blood plasmaadsorb to the particle surface. It is relevant to this dis-cussion as such proteins make the particle more suscepti-ble to phagocytosis and subsequent unwanted removal bythe body. High negative zeta potential values in the pres-ence of serum are indicative of particles that have becomehighly opsonized [10] and may indicate a poor formula-tion. Zeta potential measurements have also been appliedto investigate drug/polymer association. Alonso et al. [67]found the negative charge of poly(alkylcyanoacrylate) nano-particles was neutralized by the adsorption of the polar drug,amikacin sulphate, indicating an electrostatic association.

An alternative technique to LDA for zeta potential mea-surement using Bragg cells is amplitude weighted phasestructuration (AWPS). It is a laser light scattering techniqueallowing the simultaneous determination of particle size andzeta potential [142]. The major difference between AWPSand LDA is in the theoretical treatment and signal process-ing. A major advantage of this technique over LDA is theability to measure at low field strengths, which can be impor-tant for sensitive biological samples [144].

4.3. Surface Hydrophobicity

Surface hydrophobicity can be evaluated using several tech-niques. One method involves determining the contact angleof a polymer film that is representative of the surface ofthe particle. This is performed by removing all the waterfrom nanoparticulate suspension, dissolving in a suitable sol-vent, and casting as a film on a microscope slide [59]. Thecontact angles can then be determined by a goniometer.The contact angles increase with increasing hydrophobicity.

lasersystem

beamsplitter

lens

pinhole

photomultipliersamplecell

Figure 10. Optical setup for LDA, consisting of a laser beam being splitand focused by a lens through a suspension of particles in a samplecell. Scattered light is focused through a pinhole and detected by aphotomultiplier tube.

478 Pharmaceutical Nanotechnology

This method has been used to determine the hydrophobicityof poly(methylmethacrylate) nanoparticles and acrylic acidcopolymer nanoparticles [146–148].

An alternative procedure is the adsorption of thehydrophobic dye Rose Bengal [144]. This particular dyeshows different degrees of affinity for the particle surfacedepending on the surface hydrophobicity. After incubatingthe particles in Rose Bengal, the dye undergoes partition-ing between the surface of the particles and the dispersionmedium and is determined by spectroscopic measurementsat 542.7 nm [142]. This enables the partition quotient ofRose Bengal for the particles to be calculated and plot-ted against the surface area of the particles. The slopeof this plot is a measure of hydrophobicity, with a steepgradient indicating high hydrophobicity. This approach wasused by Lukowski et al. [147] to investigate the decreasein hydrophobicity on increasing the acrylic acid content ofacrylic acid copolymer nanoparticles.

Hydrophobic interaction chromatography (HIC) is a col-umn chromatography technique which can determine par-ticle hydrophobicity by interaction with a hydrophobic gelmatrix [142, 149, 150]. Hydrophilic particles pass through thecolumn without interaction, whereas particles with increasedhydrophobicity show a retarded elution and are retainedby the column. Hydrophobicity measurements are used todetermine the hydrophobicity of nanoparticulate carriersand correlate this to their in vivo biodistribution [10, 149].

The components of the HIC apparatus are essentially sim-ilar to those found on basic high performance liquid chro-matographic systems and consist of a column, pump, andultraviolet spectrophotometer. Particles are detected by areduction in optical density at 400 nm. A miniaturized ver-sion of HIC with a 1 ml bed volume has been developedto establish a rapid and less costly screening method [151].Alkyl agarose gels are used as the column packing and theirhydrophobicity increases with alkyl-chain length. Phosphatebuffered saline is usually used as the elution medium. Thechromatograms obtained are analyzed with regard to theelution volume and the area under the elution peak, whichenables quantitation.

HIC enables the particles to be determined in their orig-inal suspending medium preserving their true surface prop-erties. Contact angle and Rose Bengal partitioning requireremoval of the particles from their dispersion medium,which could result in a modification of the surface proper-ties. These two methods only determine an average valuefor hydrophobicity and are unable to detect subpopulationsdiffering in surface hydrophobicity. The detection of sub-populations is possible with HIC, which also allows the poly-dispersity in surface properties to be determined.

4.4. Molecular Weight Characterization

The determination of the molecular weight of nanoparticlesis performed by gel permeation chromatography (GPC).The experimental setup consists of a high performance liq-uid chromatography system with a size exclusion column anda refractive index detector. The nanoparticles are usuallyfreeze-dried and dissolved in tetrahydrofuran for analysis onthe system. Poly(styrene) or poly(methylmethacrylate) stan-dards are used to calibrate the column, to enable the deter-mination of number average molecular weight (Mn), as in

Eq. (4), and the weight average molecular weight (Mw), asin Eq. (5),

�Mn =∑

Qi∑�Qi/Mi

(4)

�Mw =∑��Qi ·Mi∑

Qi

(5)

where Qi represents the amount of polymer having a molec-ular weight Mi. The polymer molecular weight distributioncan be estimated by calculating �Mw/ �Mn.

Molecular weight determinations have been used to eluci-date the polymerization mechanism of polymers [152, 153].Poly(butylcyanoacrylate) (PBCA) polymers were reported tobe made up of numerous oligomeric subunits rather thanone or a few polymer chains due to rapid termination atthe low pH values required for preparation [153]. The pHand monomer and surfactant concentrations were foundto influence the resulting molecular weight of the PBCAnanoparticles produced [153, 154]. Interaction between thedrug and polymer could be detected by molecular weightmeasurements [60, 137, 155]. The biodegradation of nano-particles can be monitored by molecular weight measure-ments and the mechanism of degradation can also beelucidated [156–158].

It is not possible to determine the molecular weight ofhighly cross-linked polymers and natural polymers by GPCas they do not dissolve in tetrahydrafuran. Accurate resultsfor molecular weight can only be obtained if the poly-mer standards used have similar properties to the polymeremployed to prepare the nanoparticles. In most cases, thissimilarity is absent.

4.5. Microscopic Characterization

The two main techniques used to characterize nanoparticlesby microscopy are scanning electron microscopy and trans-mission electron microscopy. SEM has been used to deter-mine particle size, morphology, surface roughness, andporosity of nanoparticles made from various materials[27, 33, 159, 160]. SEM can also provide information on thebehavior of adsorbed drugs [161]. Aprahamian et al. [162]characterized lipiodol nanocapsules by the X-ray emissionof iodine using SEM fitted with an energy-dispersive X-rayspectrometer. The preparation of samples for SEM involvesdrying a portion of the particle suspension at room temper-ature and then coating it with a thin metallic film, such asgold, which is usually 30–60 nm in thickness [59]. A 10 nmlayer was obtained by spraying the nanoparticle suspensiononto an aluminum foil with the aid of an atomizer [85].Samples for SEM analysis must be capable of withstandinga vacuum environment. This technique is limited for sizingdue to the interference from surfactants in the nanoparticles,but their presence is necessary to prevent the particles fromagglomerating [59].

TEM has been used to determine the shape and par-ticle size of nanoparticles [27, 33]. Samples are preparedby placing a drop of preparation on copper grids, fol-lowed by negative staining with an aqueous solution ofsodium phosphotungstate, phosphotungstic acid, or uranylacetate [27, 163, 164]. Freeze fracturing with TEM has been

Pharmaceutical Nanotechnology 479

used to investigate the internal structure of nanoparticlesand applied to samples which are not suitable for stain-ing or which melt or sinter when irradiated by the elec-tron beam. Nanospheres were shown to have a dense, solidinternal matrix and nanocapsules a hollow internal struc-ture [27]. The interior morphology and surface roughnesshave been characterized for various nanoparticulate carriers[33, 59, 165]. TEM has been used to establish the wall thick-ness of nanocapsules, which was found to be 3 nm for PBCAnanocapsules [135]. TEM is not useful for routine mea-surements as the sample preparation is time-consuming andlaborious.

Other more advanced microscopic techniques have beendeveloped, including near-field scanning optical microscopy[166] and scanning probe microscopy techniques, such asatomic force microscopy and scanning tunnelling microscopy[166, 167].

5. DRUG LOADINGDrugs may be incorporated into nanoparticles by addition tothe polymerization medium or by adsorption to preformedparticles [168]. Depending on the drug, polymer, and prepa-ration method used, the drug can exist as:

• a solid solution of the drug in the particle,• a solid dispersion of the drug in the polymer,• surface adsorption of the drug,• chemical binding of the drug to the polymer [169].

Determination of drug content in nanoparticles is mainlyconcerned with the effective separation of the free drugfrom the bound drug. Various techniques have beenemployed including:

• ultracentrifugation,• ultrafiltration,• gel filtration,• dialysis.

The separation must be as rapid as possible to preventdesorption of drug from the particle [170].

Ultracentrifugation is the most extensively used separa-tion technique, where the particle suspension is spun athigh speeds, usually at 100,000 g for one hour [171–173].The supernatant is carefully removed and analyzed for theamount of free drug present. This is subtracted from thetotal drug added to the system, to determine the amount ofdrug bound to the particles. However, the pellet itself canbe analyzed directly to measure the amount of drug boundto the particles [174, 175]. Usually, both the supernatant andpellet are analyzed together [176–178]. The main disadvan-tage of this technique is that drug desorption and caking ofthe pellet can occur due to the long centrifugation times andlarge centrifugal forces required. Undissolved free drug inthe form of nanocrystals may spin down with the sedimentedpellet, leading to inaccurate determination of loading [133].

Reszka et al. [168] used ultrafiltration to separate freemitoxantrone from PBCA nanoparticles. The particle sus-pension was passed through a cellulose nitrate membranewith a pore size of 20 nm with magnetic stirring and pres-surized nitrogen. The filtrate obtained was analyzed for freedrug content. Recovery of the particles is not possible with

this method and the particles themselves can clog the filter.Undissolved drug may be removed with the particulate car-rier and it must be ensured that drug adsorption to the filtermembrane does not occur.

Gel filtration employs a column containing a packingmaterial, such as Sephadex® media, to separate free drugfrom the particles. The particles elute first in the void vol-ume followed by the drug, which is retained on the column,thus enabling its separation. The eluted fractions are col-lected from the column and analyzed by ultraviolet/visiblespectrophotometry. Beck et al. [170] used gel filtration withSephadex® G50 packing to separate five model drugs fromPBCA nanoparticles. This media was also used to separatefree mitoxantrone and ampicillin from liposomes [168, 173].This technique is rapid and does not alter the particle size,but desorption of the bound drug can occur. It can be scaledup and equipment costs are low [179].

Dialysis has been used by Labhasetwar and Dorle [180]to determine free drug concentration in particle suspen-sions. The suspension is placed in a dialysis bag and thenpositioned in an aqueous continuous phase until there isno change in drug concentration. The main disadvantagesof dialysis are that the equilibration times are long, caus-ing drug desorption. Large volumes of continuous phase arerequired and particle aggregation can occur [170, 181].

A potentially new separation method called field flowfractionation has been introduced only on a preliminarybasis for the determination of adsorbed substances to par-ticles [182]. A novel method was developed by Illum et al.[68] to determine both free and bound drug on PBCAnanoparticles without prior separation. For example, abathochromic shift was observed in the ultraviolet/visible(UV/vis) spectrum of Rose Bengal after binding to PBCAnanoparticles. The amount of free drug was determined at540 nm and bound drug at 548 nm. This type of analysis isvery specific to the drug under investigation.

The end result of the separation techniques described isto calculate the particle drug loading. The drug loading usu-ally expresses the bound drug as a percentage of the totaldrug content used in the preparation or as a percentage ofthe weight of the dried nanoparticles. The total drug contentis usually taken as the amount of drug added to the prepara-tion medium initially [66, 136], but errors may result if drugdegradation occurs [183]. Alternatively, total drug contentcan be expressed by dissolving the whole nanoparticle sus-pension in a suitable solvent and analyzing both the free andbound drug contents [171, 184]. Total drug content used canalso be calculated from the bound and free drug contentsanalyzed separately in the supernatant and pellet [137, 185].The drug loading efficiency is a measure of the total amountof drug recovered on analyzing the particles compared tothe total amount of drug added initially at the preparationstage.

Various analytical techniques have been employed todetermine the drug content in nanoparticles after the sep-aration procedures. High performance liquid chromatogra-phy and UV/vis spectroscopy are two of the most extensivelyused techniques [133]. Other techniques used include scin-tillation counting [186], spectrofluorodensitometry [176],microbiological assays [136], spectrofluorimetry [187], andpolarization fluoroimmunanalysis [67].

480 Pharmaceutical Nanotechnology

6. INCORPORATION OFWATER-SOLUBLE DRUGS

When introducing a water-soluble drug into a nano-particulate system, a number of factors must be considered.Different preparation methods will give rise to particles witha wide range of surface properties, sizes, and drug incorpo-rations. Some methods may appear to offer the solution toa particular loading problem, but the efficiency of particleformation may be preclusive to their use. In our experience,there are a number of factors that can be optimized in termsof loading and particle manufacture, but unfortunately thelevel of optimization may not reach that which is deemedacceptable in a therapeutic context. Many of the problemsencountered by those attempting to encapsulate a water-soluble drug have been experienced in our own work study-ing the incorporation of 5-fluorouracil. As such, it representsa good model drug for testing the incorporation efficiencyof nanoparticle production methods.

Much information exists in the literature about the var-ious payloads introduced into nanoparticles. The choice ofpharmaceutically active compound that is incorporated isdictated by many considerations but is driven primarily bythe potency of the drug, its route of administration, andthe need to give the compound some protection againstenzymatic degradation. It must be remembered that nano-particles have a low capacity for carrying drugs and largedoses cannot be administered by this means. As can be seenfrom Table 1, the range of drugs that have been incorpo-rated is wide, with many of high potency and some requir-ing protection from proteolytic attack. Other applicationsexploit the unique properties of nanoparticles in enhancingdrug delivery to the eye, with many examples to be found inthe literature. This list is by no means absolute, with furtherwork produced regularly using new drug payloads for novelapplications.

7. INCORPORATIONOF 5-FLUOROURACILINTO NANOPARTICLES

5-Fluorouracil (5FU) is 5-fluoropyridrimidine-2,4(1H , 3H)-dione. Its structure is illustrated in Figure 11. The hydrogenin the naturally occurring pyrimidine, uracil, is substituted byfluorine in the 5 position. The presence of the heteroatomsin the structure imparts hydrophilicity to the compound asthey are capable of hydrogen bonding.

The incorporation of 5FU into particulate carriers madefrom both natural and synthetic polymers has been reportedextensively. The natural carriers investigated have includedalbumin [188], chitosan [189], casein [190], liposomes [191],and fibrinogen [192]. The drug-loaded particles are usu-ally prepared by an emulsion-cross-linking procedure, whichcan be performed chemically or by heat. The loading of5FU in albumin and chitosan particles by this procedurehas been reported to be low [193, 194]. Nanospheres of chi-tosan loaded with 5FU derivatives were coated with anionicpoly(saccharides) for targeting studies involving cell-specificrecognition [189]. Methyl and ethyl cellulose nanospherescontaining 5FU have been prepared by a desolvation tech-nique with drug loading efficiencies of 45% for ethyl

Table 1. The diverse range of pharmaceutically active compounds thathave been incorporated into nanoparticles or nanocapsules.

Active compound Ref.

Peptides and proteinsinsulin [207–219]muramyl dipeptide [220–222]salmon calcitonin [223–227]chymotrypsin [228–230]

Steroids [231]triamcinolone [232]hydrocortisone [233, 234]dexamethasone [235, 236]prednisolone [237, 238]progesterone [233, 239, 240]

Nonsteroidal anti-inflammatory drugsindomethacin [241–248]diclofenac [249–254]

Polysaccharides [255–268]

Antibiotic drugs [269–278]

Antisense [24, 279–284]

Beta receptor blocking drugs [285, 286]timolol [287]carteolol [288]

Anticonvulsant drugs [289–293]primidone [294]

Cytotoxic drugsfluorouracil [295–298]cisplatin [299–301]doxorubicin [302–310]paclitaxel [311–316]methotrexate [317, 318]vinblastine [319, 320]

Antiarrhythmic drugs [321]

Miscellaneous

chlorhexidine [322, 323]diazapam [324–326]cyclosporin [327–334]pilocarpine [335–337]primaquine [338–340]

cellulose and 11.6% for methyl cellulose [121]. Syntheticcarriers have also been used to incorporate 5FU, includ-ing PMMA [195], PLGA [196], PLA [197], PBCA [198],and poly(glutaraldehyde) (PGL) [50]. The most widely usedcarriers for 5FU are the PLA and PLGA polymers. Anemulsion–evaporation preparation technique is frequentlyused to prepare the drug-loaded particles [196, 199]. Sucha method loaded 5FU up to 30% w/w in PLGA parti-cles with a diameter of 20–540 �m, which were implicatedfor the treatment of cerebral tumors by implantation [200].

C4H3FN2O2

Molecular Weight = 130.1

CAS- 51-21-8

N

FN

O

OH

H

Figure 11. Chemical structure of 5FU.

Pharmaceutical Nanotechnology 481

Nanospheres of PLGA were prepared by emulsion–evaporation by employing homogenization and a solventsystem consisting of acetone in which both the drug andpolymer were dissolved [95]. The encapsulation of 5FU inthese nanospheres was low compared to the more hydropho-bic drug, indomethacin. It was reported that the hydrophilic5FU leaked to the aqueous phase. PGL nanoparticles wereprepared by aldol condensation loading 5FU with an effi-ciency of 14.32% [50].

The process of loading 5FU into a nanoparticulate sys-tem proves to be far from simple and is representative ofother water-soluble drugs. The methods used to encapsu-late drugs tend toward low drug loading encapsulation effi-ciency by the nature of the solvent systems used. Our workfocused on the use of the emulsion–polymerization schemewith a continuous aqueous phase. In order to assess the mostsuitable parameters for particle formation, response surfacemethodology was used. Three variables were analyzed—surfactant concentration, pH of the polymerization medium,and monomer concentration. By varying these parametersand then analyzing the resultant particles produced, it waspossible to predict the set of conditions which would leadto the formation of an optimized nanoparticulate system.Using this methodology enabled the production of particlesthat could be tailored to meet specific requirements, suchas a certain diameter with a low polydispersity, and so on.When the optimum conditions were known, work involvingthe encapsulation of 5FU could proceed.

Preliminary results using the optimized system from theresponse surface model proved disappointing. Particle load-ings were below 1% and this value was even lower whena washing step was introduced. In fact, around 90% of thedrug which was “encapsulated” was removed upon the wash-ing of the particles, indicating a large surface bound fraction,which was freely soluble in the wash. A number of strate-gies were used in order to raise the loading efficiency ofthe system. Various surfactants were tried to assess whetherthere was an advantage in the use of one over another withrespect to drug loading. Indeed, there is evidence that thetype of surfactant used can influence drug loadings of nano-particles [62]. Sodium lauryl sulphate was found to be effec-tive in increasing the loading of the polar drug, amikacinsulphate, in PBCA nanoparticles [67], with the percentagedrug loading efficiency increasing from 4.76% to 60.47%.It was reported that this surfactant reduced the aqueoussolubility of the drug, thereby improving its incorporationinto nanoparticles. These authors also reported that polox-amer 188 (Pluronic F68) appeared to hinder the drug–polymer interaction for the sorption of amikacin sulphateonto freeze-dried PBCA nanoparticles. Similar results werealso obtained by Harmia et al. [201]. The most appropriatesurfactant depends on the drug under investigation. Surfac-tants can further affect the molecular weight and size of thedrug-loaded particles obtained as well as the drug release[62, 202]. However, in our work, despite definite advantagesbeing seen in the use of particular surfactants, the loadingcould not be increased significantly solely by their carefulselection.

Alterations to the concentration of 5FU, surfactant con-centration, pH, and monomer concentration were also madeto effect an increase in the loading. However, in each

case the changing of the preparatory conditions made lit-tle impact upon the loading of the particles. It was foundthat changes could lead to various particle size differencesand zeta potential changes, but any variation in loading wasin the manner of 1–2%. It was clear that the absence ofany dynamic for the water-soluble drug to move into themonomer-rich micelles during particle formation was funda-mental to the lack of success in the incorporation of 5FU.Hence, other methods of particle formation were employed,including an oil-in-oil emulsion/solvent diffusion method. Itwas reasoned that if all sources of water could be elimi-nated from the manufacturing process, there was a muchgreater likelihood of achieving significant incorporation intothe nanoparticles.

8. DRUG RELEASEDrug release can be defined as the fraction of drug releasedfrom the nanoparticulate system as a function of time afterthe system has been administered [203]. The release of drugfrom nanoparticles depends on the location and physicalstate of the drug loaded into the colloidal carrier [133]. Thedrug can be released by:

• desorption of surface-bound drug,• diffusion through the nanoparticulate matrix,• diffusion through the polymer wall of nanocapsules,• nanoparticle matrix erosion,• a combined erosion diffusion process [169].

Measurement of true release profiles requires good sinkor infinite dilution conditions so that the drug has the great-est opportunity to be released [204]. Therefore, the releasemust occur into a large volume of sink or continuous phase.However, as the sink volume is increased the concentrationof drug being measured decreases. As a compromise, it isrecommended that the drug concentration in the continu-ous phase be kept below 10% of saturation [203]. Goodsink conditions are required to avoid errors from the read-sorption of the drug to the carrier and saturating the sinkwith drug, especially if it is poorly water-soluble. Two tech-niques have been employed to determine drug release fromnanoparticulate carriers, which include membrane diffusiontechniques and the sample and separate techniques. Drugrelease from submicrometer carriers has been reviewed byWashington [203] and Magenheim and Benita [133].

8.1. Membrane Diffusion Techniques

A volume of the particle suspension is placed in a dial-ysis bag, which is sealed and placed into the continuoussink phase, which is usually phosphate buffered saline. Theentire system is kept at 37 �C with continuous magnetic stir-ring of the sink [203]. The drug in the particles diffusesinto the aqueous phase in the dialysis bag with a rate con-stant, ki. This drug concentration in the dialysis bag diffusesthrough the dialysis membrane to the sink phase with rateconstant km. The drug concentration in the sink is periodi-cally assayed to determine the amount of drug released. Themain disadvantage of this method is that the release of drugfrom particles in the dialysis bag is driven by the partitioningof the drug between the particle and its continuous phase in

482 Pharmaceutical Nanotechnology

the dialysis bag [204]. The release is determined by Kp, thepartition coefficent, and is independent of ki. This reducesthe amount of drug available for release through the dialy-sis membrane and km becomes the rate-limiting step in therelease. In contrast, the diffusion of free drug across themembrane is rapid, as the release is driven by the drug’sconcentration gradient as all the drug is available for release[204]. The rate constant, km, is fast in this instance. Anotherdisadvantage is that the membrane diffusion experiment isnot performed under sink conditions. This is because theparticulate sample itself is never diluted in the sink, since itis separated from it by the dialysis membrane. Therefore, themembrane diffusion technique is not a measure of the truerelease rate and has been found limited for the predictionof release from nanoparticles in vivo [205]. The techniqueis less prone to error where dissolution or disintegration ofthe carrier occurs [203]. However, the technique has beenextensively used to determine release from nanoparticulatecarriers in vitro for many different carriers, including lipo-somes and emulsions [159, 160, 206].

9. CONCLUSIONSNanoparticles represent a unique drug delivery system dueto their small size, which allows them to pass through thenarrowest components of the circulatory system. They alsorepresent a generic system that has the potential for adap-tation so that they can be tailor-made to specific applica-tions. Because of this, more sophisticated drug delivery ispossible, based on active targeting methods, allowing local-ization to sites that can be accessed from the circulation.An attached monoclonal antibody can be used to recognizean exposed antigen and so effect attachment. Consequentdrug release will be concentrated around the target cells andwill increase the probability that drug will diffuse into thecell of choice and not into those that do not require drugadministration.

Incorporating water-soluble drugs has proved to be prob-lematic. However, much work is underway to develop meth-ods to achieve loadings in the particle that can be usedin clinical problems. Development of particles with target-ing capabilities has been less successful and more inves-tigations are awaited, although one potential opportunityremains to be exploited and involves the integration oftechnology, more commonly seen in solid state-based elec-tronic technology, with the design of particles describedherein. Whether a particle can be controlled by onboardprocessing capabilities remains a debatable possibility, butan exciting one with no shortage of potential clinicalapplications.

GLOSSARYCoacervation The separation of two liquid phases in col-loidal systems.Liposome A spherical capsule consisting of a liquid coresurrounded by a lipid bilayer.Nanocapsule A spherical particle of submicrometer diam-eter consisting of an outer polymeric wall which encapsu-lates an inner core.

Nanoparticle Any solid particle ranging in size from 10 to1000 nm.Nanosphere A spherical particle of submicrometer diame-ter consisting of a solid monolithic polymer matrix with nodiscernable core.Photon correlation spectroscopy A technique for measur-ing the size of submicrometer particles by analyzing theirsize-dependent scattering of laser light.Supercritical fluid A substance whose temperature andpressure has been raised beyond the point where its gaseousand liquid states have equal density.Tyndall effect The scattering of light as it passes throughcolloidal suspensions.Ultracentrifugation The centrifugation of samples at veryhigh speeds.Zeta potential The potential existing between the suspend-ing medium and the effective electrical surface of a particle.

REFERENCES1. R. Sykes, Int. J. Antimicrobial Agents 14, 1 (2000).2. P. K. Gupta, J. Pharm. Sci. 79, 949 (1990).3. P. R. Lockman, R. J. Mumper, M. A. Khan, and D. D. Allen, Drug

Devel. Ind. Pharm. 28, 1 (2002).4. C. Damge, M. Aprahamian, G. Balboni, A. Hoeltzel, V. Andrieu,

and J. P. Devissaguet, Int. J. Pharm. 36, 121 (1987).5. L. Illum, P. D. E. Jones, J. Kreuter, R. W. Baldwin, and S. S. Davis,

Int. J. Pharm. 17, 65 (1983).6. S. S. Davis, L. Illum, S. M. Moghimi, M. C. Davies, C. J. H. Porter,

I. S. Muir, A. Brindley, N. M. Christy, M. E. Norman, P. Williams,and S. E. Dunn, J. Controlled Release 24, 157 (1993).

7. J. Kreuter, European J. Drug Metabolism Pharmacokinetics 3, 253(1994).

8. S. M. Moghimi, Adv. Drug Delivery Rev. 17, 1 (1995).9. M. Silhol, M. Tyagi, M. Giacca, B. Lebleu, and E. Vives, European

J. Biochem. 269, 494 (2002).10. R. H. Muller, K. H. Wallis, S. D. Troster, and J. Kreuter, J. Con-

trolled Release 20, 237 (1992).11. M. T. Peracchia, E. Fattal, D. Desmaele, M. Besnard, J. P. Noel,

J. M. Gomis, M. Appel, J. d’Angelo, and P. Couvreur, J. ControlledRelease 60, 121 (1999).

12. D. Bazile, C. Prud’homme, M. T. Bassoullet, M. Marlard, G. Spen-lehauer, and M. Veillard, J. Pharm. Sci. 84, 493 (1995).

13. Y. P. Li, Y. Y. Pei, Z. H. Zhou, X. Y. Zhang, Z. H. Gu, J. Ding,J. J. Zhou, X. J. Gao, and J. H. Zhu, Biol. Pharm. Bull. 24, 662(2001).

14. T. Verrecchia, G. Spenlehauer, D. V. Bazile, A. Murry-Brelier,Y. Archimbaud, and M. Veillard, J. Controlled Release 36, 49(1995).

15. C. L. Bourlais, L. Acar, H. Zia, P. A. Sado, T. Needham, andR. Leverge, Progr. Retinal Eye Res. 17, 33 (1998).

16. J. Kreuter, Adv. Drug Delivery Rev. 7, 71 (1991).17. P. Couvreur and C. Vauthier, J. Controlled Release 17, 187 (1991).18. J. Kreuter, Methods Enzymol. 112, 129 (1985).19. H. Murakami, M. Kobayashi, H. Takeuchi, and Y. Kawashima, Int.

J. Pharm. 187, 143 (1999).20. T. Govender, S. Stolnik, M. C. Garnett, L. Illum, and S. S. Davis,

J. Controlled Release 57, 171 (1999).21. K. A. Janes, M. P. Fresneau, A. Marazuela, A. Fabra, and M. J.

Alonso, J. Controlled Release 73, 255 (2001).22. P. Calvo, C. Remunan-Lopez, J. L. Vila-Jato, and M. J. Alonso,

Pharm. Res. 14, 1431 (1997).

Pharmaceutical Nanotechnology 483

23. E. Leo, M. Angela Vandelli, R. Cameroni, and F. Forni, Int. J.Pharm. 155, 75 (1997).

24. H. Fritz, M. Maier, and E. Bayer, J. Colloid Interface Sci. 195, 272(1997).

25. G.-H. Hsiue, S.-h. Hsu, C.-C. Yang, S.-H. Lee, and I.-K. Yang,Biomaterials 23, 457 (2002).

26. F. J. Papatheofanis and R. Barmada, J. Biomed. Mater. Res. 25, 761(1991).

27. J. M. Rollot, P. Couvreur, L. Roblot-Treupel, and F. Puisieux,J. Pharm. Sci. 75, 361 (1986).

28. R. R. C. New, “Liposomes: A Practical Approach.” IRL Press,Oxford, 1990.

29. G. Gregoriadas, “Liposomes as Drug Carriers: Recent Trends andProgress.” Wiley, New York, 1988.

30. G. Gregoriadis, A. T. Florence, and H. M. Patel, “Liposomes inDrug Delivery.” Harwood, London, 1993.

31. Y. Barenholz and D. J. Crommelin, in “Encyclopedia of Phar-maceutical Technology” (J. C. Boylan, Ed.), pp. 1–39. Dekker,New York, 1994.

32. T. M. Allen, Drugs 54, 8 (1997).33. P. Couvreur, B. Kante, M. Roland, P. Guiot, P. Bauduin, and

P. Speiser, J. Pharmacy Pharmacol. 31, 331 (1979).34. S. S. Davis, L. Illum, and S. Stolnik, Current Opinion Colloid Inter-

face Sci. 1, 660 (1996).35. E. Allemann, R. Gurny, and E. Doelker, European J. Pharm. Bio-

pharm. 39, 173 (1993).36. P. Couvreur, C. Dubernet, and F. Puisieux, European J. Pharm.

Biopharm. 41, 2 (1995).37. M. Gallardo, L. Roblot-Treupel, J. Manhuteau, I. Genin, P. Cou-

vreur, M. Plat, and F. Puisieux, Nanocapsules et nanospherèsd’alkyl-cyanoacrylate, interactions principe actif/polymère, in“1989 Proc. 5th Int. Conf. Pharm. Tech.” pp. D36–D45.

38. R. Jalil and J. R. Nixon, J. Microencapsulation 7, 25 (1989).39. R. Arshady, J. Microencapsulation 5, 101 (1988).40. J. Kreuter, Pharm. Acta Helvetiae 58, 196 (1983).41. R. M. Fitch, M. M. B. Prenosil, and K. J. Sprick, J. Polymer Sci. A

27, 95 (1969).42. J. Kreuter, in “Microcapsules and Nanoparticles in Medicine

and Pharmacy” (M. Donbrow, Ed.), pp. 125–245. CRC Press,Boca Raton, 1992.

43. V. U. E. Berg, J. Kreuter, P. P. Speiser, and M. Soliva, Pharm. Ind.48, 75 (1986).

44. A. Schwarz and A. Rembaum, in “Methods in Enzymology”(R. Green, Ed.), pp. 175–194. Academic Press, Orlando, 1985.

45. A. Rembaum, S. P. S. Yen, and R. S. Molday, J. Macromol. Sci. A13, 603 (1979).

46. M. Chang, G. Richards, and A. Rembaum, in “Methods in Enzy-mology” (R. Green, Ed.), pp. 150–164. Academic Press, Orlando,1985.

47. S. Margel and E. Wiesel, J. Polymer Sci. A 22, 145 (1984).48. S. Margel, S. Zisblatt, and A. Rembaum, J. Immunol. Methods 28,

341 (1979).49. A. D. McLeod, F. C. Lam, P. K. Gupta, and C. T. Hung, J. Pharm.

Sci. 77, 704 (1988).50. G. Mukherji, R. S. R. Murthy, and B. D. Miglani, Int. J. Pharm.

50, 15 (1989).51. S. J. Douglas, S. S. Davis, and L. Illum, CRC Critical Rev. Thera-

peutic Drug Carrier Syst. 3, 233 (1987).52. P. Couvreur and C. Vauthier, J. Controlled Release 17, 187 (1991).53. D. K. Ousterhout, W. T. Tumbusch, P. M. Margetis, and

F. Leonard, British J. Plastic Surgery 24, 23 (1971).54. J. A. Collins, K. C. Pani, M. M. Seidenstein, G. Brandes, and

F. Leonard, Surgery 65, 256 (1969).55. S. J. Douglas, L. Illum, S. S. Davis, and J. Kreuter, J. Colloid

Interface Sci. 101, 149 (1984).56. P. Couvreur, CRC Critical Rev. Therapeutic Drug Carrier Syst. 5, 1

(1988).

57. C. Verdun, P. Couvreur, and H. Vranckx, J. Controlled Release 3,205 (1986).

58. P. Maincent, J. P. Devissaguet, R. Leverge, P. A. Sado, andP. Couvreur, Appl. Biochem. Biotechnol. 10, 263 (1984).

59. J. Kreuter, Int. J. Pharm. 14, 43 (1983).60. B. Seijo, E. Fattal, L. Roblottreupel, and P. Couvreur, Int. J.

Pharm. 62, 1 (1990).61. V. Lenaerts, P. Raymond, J. Juhasz, M. A. Simard, and C. Jolicoeur,

J. Pharm. Sci. 78, 1051 (1989).62. G. Puglisi, G. Giammona, M. Fresta, B. Carlisi, N. Micali, and

A. Villari, J. Microencapsulation 10, 353 (1993).63. T. Harmia-Pulkkinen, A. Tuomi, and E. Kristoffersson, J. Microen-

capsulation 6, 87 (1989).64. P. A. McCarron, A. D. Woolfson, and S. M. Keating, Int. J. Pharm.

193, 37 (1999).65. J. Kreuter, in “Encyclopedia of Pharmaceutical Technology” (J. C.

Boylan, Ed.), pp. 1–29. Dekker, New York, 1988.66. N. Bapat and M. Boroujerdi, Drug Devel. Ind. Pharm. 18, 65

(1992).67. M. J. Alonso, C. Losa, P. Calvo, and J. L. Vilajato, Int. J. Pharm.

68, 69 (1991).68. L. Illum, M. A. Khan, E. Mak, and S. S. Davis, Int. J. Pharm. 30,

17 (1986).69. F. Lescure, C. Seguin, P. Breton, P. Bourrinet, D. Roy, and P. Cou-

vreur, Pharm. Res. 11, 1270 (1994).70. T. K. M. Mbela, J. H. Poupaert, P. Dumont, and A. Haemers, Int.

J. Pharm. 92, 71 (1993).71. J. L. De Keyser, J. H. Poupaert, and P. Dumont, J. Pharm. Sci. 80,

67 (1991).72. G. Birrenbach and P. P. Speiser, J. Pharm. Sci. 65, 1763 (1976).73. B. Ekman and I. Sjoholm, J. Pharm. Sci. 67, 693 (1978).74. P. Edman, B. Ekman, and I. Sjoholm, J. Pharm. Sci. 69, 838 (1980).75. J. Kreuter and P. P. Speiser, J. Pharm. Sci. 65, 1624 (1976).76. U. Berg, “Immunstimulation durch hochdisperse polymersuspen-

sionen.” ETH, Zurich, 1979.77. J. Kreuter and P. P. Speiser, Infection Immunity 13, 204 (1976).78. R. S. Molday, W. J. Dreyer, A. Rembaum, and S. P. S. Yen, J. Cell

Biol. 64, 75 (1975).79. A. Rolland, D. Gibassier, P. Sado, and R. Leverge, J. Pharm. Bel-

gique 41, 94 (1986).80. A. Rembaum, S. P. S. Yen, E. Cheong, S. Wallace, R. S. Molday,

and W. J. Deyer, Macromolecules 9, 328 (1976).81. J. Kreuter, U. Tauber, and V. Illi, J. Pharm. Sci. 68, 1443 (1979).82. J. W. Vanderhoff, M. S. El-Aasser, and J. Ugelstad, U.S. Patent

4, 177, 177, 1979.83. R. Gurny, N. A. Peppas, D. D. Harrington, and G. S. Banker, Drug

Devel. Ind. Pharm. 7, 1 (1981).84. R. Bodmeier and H. G. Chen, J. Controlled Release 12, 223 (1990).85. H. J. Krause, A. Schwarz, and P. Rohdewald, Int. J. Pharm. 27, 145

(1985).86. P. D. Scholes, A. G. A. Coombes, L. Illum, S. S. Davis, M. Vert,

and M. C. Davies, J. Controlled Release 25, 145 (1993).87. M. C. Julienne, M. J. Alonso, J. L. Gomez-Amoza, and J. P.

Benoit, Drug Devel. Ind. Pharm. 18, 1063 (1992).88. T. R. Tice and R. M. Gilley, J. Controlled Release 2, 343 (1985).89. R. Bodmeier and H. Chen, Preparation and evaluation of drug

containing polymeric nanosuspensions, in “1989 Proc. 5th Int.Conf. Pharm. Tech.,” pp. B265–268.

90. F. Koosha and R. H. Muller, Arch. Pharm. 320, 913 (1987).91. F. Koosha, R. H. Muller, S. S. Davis, and M. C. Davies, J. Con-

trolled Release 9, 149 (1989).92. R. Bodmeier and H. Chen, J. Controlled Release 10, 167 (1989).93. R. Bodmeier and J. W. McGinity, Int. J. Pharm. 43, 179 (1988).94. R. Bodmeier and J. W. McGinity, Pharm. Res. 4, 465 (1987).95. T. Niwa, H. Takeuchi, T. Hino, N. Kunou, and Y. Kawashima,

J. Controlled Release 25, 89 (1993).

484 Pharmaceutical Nanotechnology

96. D. C. Tsai, S. A. Howard, T. F. Hogan, C. J. Malanga, S. J.Kandzari, and J. K. Ma, J. Microencapsulation 3, 181 (1986).

97. R. Jalil and J. R. Nixon, J. Microencapsulation 7, 25 (1990).98. R. Wada, S. H. Hyon, and Y. Ikada, J. Pharm. Sci. 79, 919 (1990).99. C. Bindschaedler, R. Gurny, and E. Doelker, Swiss Patent 1497/88,

1988.100. H. Ibrahim, C. Bindschaedler, E. Doelker, P. Buri, and R. Gurny,

Int. J. Pharm. 87, 239 (1992).101. E. Allemann, E. Doelker, and R. Gurny, European J. Pharm. Bio-

pharm. 39, 13 (1993).102. E. Allemann, R. Gurny, and E. Dorlker, Int. J. Pharm. 87, 247

(1992).103. J. C. Leroux, E. Allemann, E. Doelker, and R. Gurny, European

J. Pharm. Biopharm. 41, 14 (1995).104. S. Massen, E. Fattal, R. H. Müller, and P. Couvreur, S.T.P. Pharma

Sci. 3, 11 (1993).105. H. Fessi, F. Puisieux, and J. P. Devissaguet, European Patent

274961, 1987.106. D. Quintanar-Guerrero, E. Allemann, E. Doelker, and H. Fessi,

Colloid Polymer Sci. 275, 640 (1997).107. P. Couvreur, C. Dubernet, and F. Puisieux, European J. Pharm.

Biopharm. 41, 2 (1995).108. S. Stolnik, M. C. Davies, L. Illum, S. S. Davis, M. Boustta, and

M. Vert, J. Controlled Release 30, 57 (1994).109. S. Stolnik, M. C. Garnett, M. C. Davies, L. Illum, M. Bousta,

M. Vert, and S. S. Davis, Colloids Surfaces A 97, 235 (1995).110. S. Stainmesse, A. M. Orecchioni, and E. Nakache, Modelling of an

original process to obtain biocompatible polymeric nanospheres,in “1992 6th International Conference on Pharmaceutical Tech-nology,” pp. 89–97.

111. S. Stainmesse, A. M. Orecchioni, E. Nakache, F. Puisieux, andH. Fessi, Colloid Polymer Sci. 273, 505 (1995).

112. Y. Kawashima, H. Yamamoto, H. Takeuchi, T. Hino, and T. Niwa,European J. Pharm. Biopharm. 45, 41 (1998).

113. H. G. Bungenberg de Jong and H. R. Kruyt, Chemistry coac-ervation (partial miscibility in colloid systems) (preliminary com-munication), in “1929 Proc. Konink. Nederl. Akad. Wetensch,”pp. 849–856.

114. C. Thomasin, N. T. Ho, H. P. Merkle, and B. Gander, J. Pharm.Sci. 87, 259 (1998).

115. M. Donbrow, “Microcapsules and Nanoparticles in Medicine andPharmacy.” CRC Press, Boca Raton, 1992.

116. P. B. Deasy, “Microencapsulation and Related Processes.” Dekker,New York, 1984.

117. L. M. Sanders, J. S. Kent, G. I. McRae, B. H. Vickery, T. R. Tice,and D. H. Lewis, J. Pharm. Sci. 73, 1294 (1984).

118. J. M. Ruiz, B. Tissier, and J. P. Benoit, Int. J. Pharm. 49, 69 (1989).119. C. Thomasin, H. P. Merkle, and B. Gander, J. Pharm. Sci. 87, 269

(1998).120. T. Niwa, H. Takeuchi, T. Hino, M. Nohara, and Y. Kawashima, Int.

J. Pharm. 121, 45 (1995).121. G. Mukherji, R. S. R. Murthy, and B. D. Miglani, Int. J. Pharm.

65, 1 (1990).122. E. Reverchon, I. De Marco, and G. Della Porta, Int. J. Pharm. 243,

83 (2002).123. P. Chattopadhyay and R. B. Gupta, Int. J. Pharm. 228, 19 (2001).124. P. Chattopadhyay and R. B. Gupta, AICHE J. 48, 235 (2002).125. P. Chattopadhyay and R. B. Gupta, Ind. Eng. Chem. Res. 40, 3530

(2001).126. K. S. Soppimath, T. M. Aminabhavi, A. R. Kulkarni, and W. E.

Rudzinski, J. Controlled Release 70, 1 (2001).127. U. Foerter-Barth, U. Teipel, and H. Krause, Formation of parti-

cles by applying the gas anti-solvent (GAS) process, in “Notting-ham, 1999 Supercritical fluids: Chemistry and Materials,” InstitutNational Polytechnique de Lorraine, pp. 175–180.

128. N. Elvassore, A. Bertucco, and P. Caliceti, J. Pharm. Sci. 90, 1628(2001).

129. L. Illum, S. S. Davis, C. G. Wilson, N. W. Thomas, M. Frier, andJ. G. Hardy, Int. J. Pharm. 12, 135 (1982).

130. D. J. Wilkins and P. A. Meyers, British J. Exp. Pathol. 47, 568(1966).

131. C. J. Van Oss, C. F. Gillman, and A. W. Neumann, “Phago-cytic Engulfment and Cell Adhesiveness as Surface Phenomena.”Dekker, New York, 1975.

132. L. Grislain, P. Couvreur, V. Lenaerts, M. Roland, D. Deprezde-campeneere, and P. Speiser, Int. J. Pharm. 15, 335 (1983).

133. B. Magenheim and S. Benita, STP Pharma Sci. 1, 221 (1991).134. R. J. Haskell, J. Pharm. Sci. 87, 125 (1998).135. N. A. Fallouh, L. Roblottreupel, H. Fessi, J. P. Devissaguet, and

F. Puisieux, Int. J. Pharm. 28, 125 (1986).136. S. Henry-Michelland, M. J. Alonso, A. Andremont, P. Maincen,

J. Sauzieres, and P. Couvreur, Int. J. Pharm. 35, 121 (1987).137. R. Gaspar, V. Preat, and M. Roland, Int. J. Pharm. 68, 111 (1991).138. R. H. Muller, C. Lherm, J. Herbot, and P. Couvreur, Biomaterials

11, 590 (1990).139. S. J. Douglas, L. Illum, and S. S. Davis, J. Colloid Interface Sci.

103, 155 (1985).140. R. H. Muller and K. H. Wallis, Int. J. Pharm. 89, 25 (1993).141. A. Malaiya and S. P. Vyas, J. Microencapsulation 5, 243 (1988).142. R. H. Muller, “Colloidal Carriers for Controlled Drug Delivery

and Targeting: Modification, Characterisation and In Vivo Distri-bution.” CRC Press, Stuttgart, 1991.

143. A. Zimmer, J. Kreuter, and J. R. Robinson, Proc. Int. Symp. Con-trol. Rel. Bioact. Mater. 18, 642 (1991).

144. R. H. Muller, S. S. Davis, L. Illum, and E. Mak, in “Targeting ofDrugs with Synthetic Systems.” (G. Poste, Ed.), Vol. 113. PlenumPress, New York, 1986.

145. S. E. Dunn, A. G. A. Coombes, M. C. Garnett, S. S. Davis, M. C.Davies, and L. Illum, J. Controlled Release 44, 65 (1997).

146. S. D. Troster and J. Kreuter, J. Microencapsulation 9, 19 (1992).147. G. Lukowski, R. H. Muller, B. W. Muller, and M. Dittgen, Int. J.

Pharm. 84, 23 (1992).148. G. Lukowski, R. H. Muller, B. W. Muller, and M. Dittgen, Colloid

Polymer Sci. 271, 100 (1993).149. H. Carstensen, B. W. Muller, and R. H. Muller, Int. J. Pharm. 67,

29 (1991).150. T. Blunk, E. Mak, and R. H. Muller, Pharm. Ind. 55, 612 (1993).151. K. H. Wallis and R. H. Muller, Pharm. Ind. 55, 1124 (1993).152. V. Bentele, U. E. Berg, and J. Kreuter, Int. J. Pharm. 13, 109

(1983).153. S. J. Douglas and S. S. Davis, British Polymer J. 17, 339 (1985).154. L. Vansnick, P. Couvreur, D. Christiaens-Leyh, and M. Roland,

Pharm. Res. 36 (1984).155. C. Damge, C. Michel, M. Aprahamian, P. Couvreur, and J. P.

Devissaguet, J. Controlled Release 13, 233 (1990).156. W. R. Vezin and A. T. Florence, J. Biomed. Mater. Res. 14, 93

(1980).157. F. Leonard, R. K. Kulkarni, G. Brandes, J. Nelson, and J. Cameron,

J. Appl. Polymer Sci. 10, 259 (1966).158. V. Lenaerts, P. Couvreur, D. Christiaens-Leyh, E. Joiris,

M. Roland, B. Rollman, and P. Speiser, Biomaterials 5, 65 (1984).159. M. S. El-Samaligy, P. Rohdewald, and H. A. Mahmoud, J. Pharm.

Pharmacol. 38, 216 (1986).160. H.-J. Krause, A. Schwarz, and P. Rohdewald, Drug Devel. Ind.

Pharm. 12, 527 (1986).161. M. El-Samaligy and P. Rohdewald, Pharm. Acta Helvetiae 57, 201

(1982).162. M. Aprahamian, C. Michel, W. Humbert, J. P. Devissaguet, and

C. Damge, Biol. Cell 61, 69 (1987).163. H. Fessi, F. Puisieux, J. P. Devissaguet, N. Ammouray, and

S. Benita, Int. J. Pharm. 55, R1 (1989).164. H. J. Krause and P. Rohdewald, Pharm. Res. 239 (1985).165. B. Kante, P. Couvreur, V. Lenaerts, P. Guiot, M. Roland, P. Baud-

huin, and P. Speiser, Int. J. Pharm. 7, 45 (1980).

Pharmaceutical Nanotechnology 485

166. G. Binnig, C. F. Quate, and C. Gerber, Phys. Rev. Lett. 56, 930(1986).

167. D. R. Denley, Ultramicroscopy 33, 83 (1990).168. R. Reszka, P. Beck, I. Fichtner, M. Hentschel, J. Richter, and

J. Kreuter, J. Pharm. Exp. Therapeutics 280, 232 (1997).169. J. Kreuter, in “Encyclopedia of Pharmaceutical Technology” (J. C.

Boylan, Ed.), pp. 165–190. Dekker, New York, 1994.170. P. Beck, D. Csherer, and J. Kreuter, J. Microencapsulation 7, 491

(1990).171. J. L. Grangier, M. Puygrenier, J. C. Gautier, and P. Couvreur,

J. Controlled Release 15, 3 (1991).172. F. Nemati, C. Dubernet, A. C. Deverdiere, M. F. Poupon, L. Tre-

upelacar, F. Puisieux, and P. Couvreur, Int. J. Pharm. 102, 55(1994).

173. E. Fattal, J. Rojas, L. Roblot-Treupel, A. Andremont, and P. Cou-vreur, J. Microencapsulation 8, 29 (1991).

174. G. Cavallaro, M. Fresta, G. Giammona, G. Puglisi, and A. Villari,Int. J. Pharm. 111, 31 (1994).

175. A. Sanchez, J. L. Vila-Jato, and M. J. Alonso, Int. J. Pharm. 99,263 (1993).

176. P. Maincent, R. Le Verge, P. Sado, P. Couvreur, and J. P. Devis-sageut, J. Pharm. Sci. 75, 955 (1986).

177. M. Fresta, G. Puglisi, G. Giammona, G. Cavallaro, N. Micali, andP. M. Furneri, J. Pharm. Sci. 84, 895 (1995).

178. M. Radwan, I. Zaghloul, and Z. Aly, European J. Pharm. Sci. 8,95 (1999).

179. P. Beck, D. Scherer, and J. Kreuter, J. Microencapsulation 7, 491(1990).

180. V. D. Labhasetwar and A. K. Dorle, J. Controlled Release 12, 113(1990).

181. J. J. Marty, R. C. Oppenheim, and P. Speiser, Pharm. Acta Helve-tiae 53, 17 (1978).

182. S. Levin, Israel J. Chem. 30, 257 (1990).183. J. Kreuter, M. Nefzger, E. Liehl, R. Czok, and R. Voges, J. Pharm.

Sci. 72, 1146 (1983).184. N. Brasseur, D. Brault, and P. Couvreur, Int. J. Pharm. 70, 129

(1991).185. M. A. Radwan, Drug Devel. Ind. Pharm. 21, 2371 (1995).186. C. Verdun, F. Brasseur, V. Vranckx, P. Couvreur, and M. Roland,

Cancer Chemother. Pharmacol. 26, 13 (1990).187. P. Calvo, J. L. Vila-Jato, and M. J. Alonso, J. Pharm. Sci. 85, 530

(1996).188. E. J. Truter, Artificial Cells Blood Substitutes Immobilization

Biotechnol. 23, 579 (1995).189. T. Ouchi, H. Kobayashi, and T. Banba, British Polymer J. 23, 221

(1990).190. M. S. Latha, A. Jayakrishnan, K. Rathinam, and M. Mohanty,

J. Pharm. Pharmacol. 46, 858 (1994).191. B. Elorza, M. A. Elorza, G. Frutos, and J. R. Chantres, Biochim.

Biophys. Acta 1153, 135 (1993).192. S. Miyazaki, N. Hashiguchi, M. Sugiyama, M. Takada, and

Y. Morimoto, Chem. Pharm. Bull. 34, 1370 (1986).193. K. Sugibayashi, Y. Morimoto, T. Nadai, Y. Kato, A. Hasegawa,

and T. Arita, Chem. Pharm. Bull. 27, 204 (1979).194. J. Akbuga and N. Bergisadi, J. Microencapsulation 13, 161 (1996).195. H. S. Yalabik-Kas, J. Kreuter, A. A. Hincal, and P. P. Speiser,

J. Microencapsulation 3, 71 (1986).196. M. Boisdron-Celle, J. M. Ruiz, and J. P. Benoit, Preparation

and characterisation of 5-fluorouracil-loaded microspheres asbiodegradable anti-cancer drug carriers, in “1992 6th InternationalConference on Pharmaceutical Technology,” pp. 52–61.

197. K. Ciftci, A. A. Hincal, H. S. Kas, M. T. Ercan, and S. Ruacan,European J. Pharm. Sci. 1, 249 (1994).

198. J. Kreuter and H. R. Hartmann, Oncology 40, 363 (1983).199. K. Ciftci, A. A. Hincal, and H. S. Kas, Microspheres of 5-fluorou-

racil using poly(dl-lactic acid), effect of polymer molecular weight

on release characteristics, in “1992 6th International Conferenceon Pharmaceutical Technology,” pp. 62–71.

200. M. Boisdron-Celle, P. Menei, and J. P. Benoit, J. Pharm. Pharma-col. 47, 108 (1995).

201. T. Harmia, P. Speiser, and J. Kreuter, Int. J. Pharm. 33, 45 (1986).202. M. A. Vandelli, M. Fresta, G. Puglisi, and F. Forni, J. Microencap-

sulation 11, 531 (1994).203. C. Washington, Int. J. Pharm. 58, 1 (1990).204. C. Washington, Int. J. Pharm. 56, 71 (1989).205. M. Y. Levy and S. Benita, Int. J. Pharm. 66, 29 (1990).206. N. Ammoury, H. Fessi, J. P. Devissaguet, F. Puisieux, and S. Benita,

STP Pharma Sci. 5, 647 (1989).207. H. Yuan, F. q. Hu, and X. y. Ying, Chinese Pharm. J. 37, 349

(2002).208. Z. Ma, H. H. Yeoh, and L. Y. Lim, J. Pharm. Sci. 91, 1396 (2002).209. Y. Pan, H. Xu, H. Y. Zhao, G. Wei, and J. M. Zheng, Acta Pharm.

Sinica 37, 377 (2002).210. N. Elvassore, A. Bertucco, and P. Caliceti, J. Pharm. Sci. 90, 1628

(2001).211. N. Behan and C. Birkinshaw, Macromole. Rapid Comm. 22, 41

(2001).212. P. Venugopalan, A. Sapre, N. Venkatesan, and S. P. Vyas, Phar-

mazie 56, 217 (2001).213. L. M. Ma, Q. Zhang, and Y. Z. Li, Chinese Pharm. J. 36, 38 (2001).214. Q. Zhang, Z. Shen, and T. Nagai, Int. J. Pharm. 218, 75 (2001).215. Z. N. Yin and B. Lu, Chinese J. Pharm. 31, 349 (2000).216. S. Watnasirichaikul, N. M. Davies, T. Rades, and I. G. Tucker,

Pharm. Res. 17, 684 (2000).217. Z. N. Yin, B. Lu, and G. J. Wang, West China J. Pharm. Sci. 15,

250 (2000).218. M. Aboubakar, P. Couvreur, H. Pinto-Alphandary, B. Gouritin,

B. Lacour, R. Farinotti, F. Puisieux, and C. Vauthier, Drug Dev.Res. 49, 109 (2000).

219. C. X. Zheng, M. X. Duan, Z. C. Yue, H. Ma, and J. H. Guo,Polymer Preprints 40, 351 (1999).

220. I. Seyler, M. Appel, J. P. Devissaguet, P. Legrand, and G. Barratt,Int. J. Immunopharmacol. 18, 385 (1996).

221. G. M. Barratt, C. Morin, M. Appel, and I. Seyler, Intracellu-lar delivery of a muramyl dipeptide derivative by poly(dl-lactide)nanocapsules, in “In vitro and ex vivo test systems to rationalizedrug design and delivery,” Paris, 1993, pp. 265–268.

222. C. Morin, G. Barratt, H. Fessi, and J. P. Devissaguet, Int. J.Immunopharmacol. 16, 451 (1994).

223. S. Sakuma, N. Suzuki, R. Sudo, K. I. Hiwatari, A. Kishida, andM. Akashi, Int. J. Pharm. 239, 185 (2002).

224. S. Sakuma, Y. Ishida, R. Sudo, N. Suzuki, H. Kikuchi, K. I.Hiwatari, A. Kishida, M. Akashi, and M. Hayashi, Int. J. Pharm.159, 181 (1997).

225. S. Sakuma, N. Suzuki, H. Kikuchi, K. I. Hiwatari, K. Arikawa,A. Kishida, and M. Akashi, Int. J. Pharm. 149, 93 (1997).

226. S. Sakuma, N. Suzuki, H. Kikuchi, K. I. Hiwatari, K. Arikawa,A. Kishida, and M. Akashi, Int. J. Pharm. 158, 69 (1997).

227. H. Vranckx, M. Demoustier, and M. Deleers, Eur. J. Pharm. Bio-pharm. 42, 345 (1996).

228. N. O. Fischer, C. M. McIntosh, J. M. Simard, and V. M. Rotello,Proc. Nat. Acad. Sci. USA 99, 5018 (2002).

229. Y. E. Shapiro and E. G. Pykhteeva, Appl. Biochem. Biotechnol. 74,67 (1998).

230. Y. E. Shapiro, E. G. Pykhteeva, and G. V. Fedorova, Russian J.Bioorganic Chem. 23, 156 (1997).

231. A. M. da Silveira, D. Duchene, and G. Ponchel, STP Pharma Sci.10, 309 (2000).

232. J. Valero, M. A. Egea, M. Espina, F. Gamisans, and M. L. Garcia,Drug Dev. Ind. Pharm. 22, 167 (1996).

233. R. Cavalli, E. Peira, O. Caputo, and M. R. Gasco, Int. J. Pharm.182, 59 (1999).

486 Pharmaceutical Nanotechnology

234. A. K. Zimmer, P. Maincent, P. Thouvenot, and J. Kreuter, Int. J.Pharm. 110, 211 (1994).

235. M. G. Cascone, P. M. Pot, Z. Zhu, and L. Lazzeri, J. Mater. Sci.Mater. Med. 13, 265 (2002).

236. C. X. Song, V. Labhasetwar, H. Murphy, X. Qu, W. R. Humphrey,R. J. Shebuski, and R. J. Levy, J. Control. Release 43, 197 (1997).

237. J. H. Chen, M. Shagufta, and S. S. Davis, Acta Pharm. Sinica 37,473 (2002).

238. A. Zur Muehlen and W. Mehnert, Pharmazie 53, 552 (1998).239. E. Lemos-Senna, D. Wouessidjewe, and D. Duchene, Colloids Sur-

faces B 10, 303 (1998).240. J. Jansen, C. Treiner, and C. Vaution, J. Colloid Interface Sci. 179,

578 (1996).241. A. Raffin Pohlmann, V. Weiss, O. Mertins, N. Pesce da Silveira,

and S. Staniscuaski Guterres, Eur. J. Pharm. Sci. 16, 305 (2002).242. C.W.Maboundou, G. Paintaud, Y. Ottignon, J. Magnette, Y. Bech-

tel, H. Fessi, F. Gouchet, P. Carayon, and P. Bechtel, Therapie 51,88 (1996).

243. F. Fawaz, F. Bonini, M. Guyot, A. M. Lagueny, H. Fessi, and J. P.Devissaguet, Int. J. Pharm. 133, 107 (1996).

244. P. Calvo, M. J. Alonso, J. L. Vila-Jato, and J. R. Robinson,J. Pharm. Pharmacol. 48, 1147 (1996).

245. M. Skiba, C. Morvan, D. Duchene, F. Puisieux, and D. Wouessid-jewe, Int. J. Pharm. 126, 275 (1995).

246. F. Fawaz, F. Bonini, M. Guyot, and A. M. Lagueny, Pharm. Res.10, 750 (1993).

247. N. Ammoury, M. Dubrasquet, H. Fessi, and J. P. Devissaguet, in“Biomedical Polymers” (D. Cohn and J. Kost, Eds.), Vol. 13, Num-ber 1/4, p. 121. Elsevier Applied Science, 1993.

248. P. Calvo, J. L. Vila-Jato, and M. J. Alonso, Corneal penetra-tion of 14C-indomethacin-loaded nanocapsules, in “Methods toOvercome Biological Barriers in Drug Delivery,” Kuopio, Finland,1993, Kuopio University, Department of Pharmaceutical Technol-ogy, p. 93.

249. C. R. Muller, S. R. Schaffazick, A. R. Pohlmann, L. de LuccaFreitas, N. Pesce da Silveira, T. Dalla Costa, and S. S. Guterres,Pharmazie 56, 864 (2001).

250. S. S. Guterres, H. Fessi, G. Barratt, F. Puisieux, and J. P. Devis-saguet, J. Biomater. Sci. Polymer Ed. 11, 1347 (2000).

251. S. S. Guterres, C. R. Muller, C. B. Michalowski, A. R. Pohlmann,and T. D. Costa, STP Pharma Sci. 11, 229 (2001).

252. A. R. Kulkarni, K. S. Soppimath, and T. M. Aminabhavi,J. Microencapsul. 17, 449 (2000).

253. O. Valls, M. L. Garcia, X. Pages, J. Valero, M. A. Egea, M. A.Salgueiro, and R. Valls, in “Trends in Colloid and Interface Sci-ence X,” (C. Solans, M. R. Infante, and M. J. Garcia-Celma, Eds.),Vol. 100, pp. 107–111. (Steinkopff-Verlag, 1996).

254. S. S. Guterres, H. Fessi, G. Barratt, and J. P. Devissaguet, Int. J.Pharm. 113, 57 (1995).

255. X. W. Li, J. L. Liu, Y. Q. Wei, A. S. C. Chan, and K. L. D. Lee,J. Pharm. Pharmacol. 51, 176 (1999).

256. K. Akiyoshi, S. Kurumada, E. C. Kang, J. Sunamoto, T. Principi,and F. M. Winnik, J. Network Polymer Japan 20, 171 (1999).

257. C. Passirani, G. Barratt, J. P. Devissaguet, and D. Labarre, LifeSci. 62, 775 (1998).

258. J. Sunamoto, S. Deguchi, Y. Kato, I. Taniguchii, Y. Sasaki,K. Kuroda, and K. Akiyoshi, in “Proceedings of the AmericanChemical Society Division of Polymeric Materials: Science andEngineering,” Vol. 79, pp. 230–231. Amer. Chem. Soc., Washington,DC, 1998.

259. X. G. Gu, M. Schmitt, A. Hiasa, Y. Nagata, H. Ikeda, Y. Sasaki,K. Akiyoshi, J. Sunamoto, H. Nakamura, and K. Kuribayashi, Can-cer Res. 58, 3385 (1998).

260. C. Passirani, G. Barratt, J. P. Devissaguet, and D. Labarre, Pharm.Res. 15, 1046 (1998).

261. N. C. Santos, M. J. E. Prieto, A. Morna-Gomes, D. Betbeder, andM. A. R. B. Castanho, Biopolymers 41, 511 (1997).

262. K. Akiyoshi, T. Nishikawa, S. I. Kobayashi, and J. Sunamoto, in“Biomedical Functions and Biotechnology of Natural and Artifi-cial Polymers” (M. Yalpani, Ed.), Vol. 3, pp. 115–126. Atl Press,1996.

263. I. Taniguchi, K. Akiyoshi, and J. Sunamoto, in “SPSJ 45th Sym-posium on Macromolecules,” Vol. 45, Number 2, p. E1019. TheSociety of Polymer Science, 1996.

264. K. Akiyoshi, S. Deguchi, H. Tajima, and T. Nishikawa, Proc. JapanAcad. Ser. B Phys. Biol. Sci. 71, 15 (1995).

265. K. Akiyoshi, T. Nishikawa, S. Shichibe, and J. Sunamoto, Chem.Lett. 707 (1995).

266. Y. Kakizawa, K. Akiyoshi, K. Nakamura, and J. Sunamoto, in“Enzymatic Activity of Lipase Complexed with Nanoparticle ofHydrophobized Polysaccharide,” Kyoto, Japan, 1998, p. III-806.Society of Polymer Science, Japan (Spsj).

267. S. Xu and M. Yonese, in “Charge Densities and Nanoparticle For-mation of Complexes Composed of Acid Polysaccharides and Pro-tein,” Nagoya, Japan, 1998, p. E1076.

268. X. W. Li, J. L. Liu, Y. Q. Wei, and A. S. C. Chan, Prepara-tion of polysaccharide nanoparticles for controlled gene release,in “British Pharmaceutical Conference,” p. 176. PharmaceuticalSociety of Great Britain, Cardiff, 1999.

269. N. S. Santos-Magalhaes, A. Pontes, V. M. Pereira, and M. N. Cae-tano, Int. J. Pharm. 208, 71 (2000).

270. P. Chattopadhyay and R. B. Gupta, Ind. Eng. Chem. Res. 40, 3530(2001).

271. H. Lin and J. Xue-hua, Chinese J. Antibiotics 25, 272 (2000).272. X. Jiang and G. Liao, Chinese J. Antibiotics 24, 35 (1999).273. H. Pinto-Alphandary, A. Andremont, and P. Couvreur, Int. J.

Antimicrob. Agents 13, 155 (2000).274. E. Reverchon and G. Della Porta, Powder Technol. 106, 23 (1999).275. M. E. Page-Clisson, H. Pinto-Alphandary, M. Ourevitch,

A. Andremont, and P. Couvreur, J. Control. Release 56, 23 (1998).276. A. E. Gulyaev, B. A. Ermekbaeva, G. Y. Kivman, T. G. Rad-

chenko, A. Y. Sherstov, and V. G. Shirinskii, Pharm. Chem. J. 32,115 (1998).

277. M. S. Espuelas, P. Legrand, J. M. Irache, C. Gamazo, A. M. Orec-chioni, J. P. Devissaguet, and P. Ygartua, Int. J. Pharm. 158, 19(1997).

278. G. Cavallaro, M. Fresta, G. Giammona, and G. Puglisi, Int. J.Pharm. 111, 31 (1994).

279. G. Lambert, J. R. Bertrand, E. Fattal, F. Subra, H. Pinto-Alphandary, C. Malvy, C. Auclair, and P. Couvreur, Biochem. Bio-phys. Res. Commun. 279, 401 (2000).

280. H. P. Zobel, M. Junghans, V. Maienschein, D. Werner, M. Gilbert,H. Zimmermann, C. Noe, J. Kreuter, and A. Zimmer, Eur. J.Pharm. Biopharm. 49, 203 (2000).

281. E. Fattal, C. Vauthier, I. Aynie, Y. Nakada, G. Lambert, C. Malvy,and P. Couvreur, J. Control. Release 53, 137 (1998).

282. H. P. Zobel, J. Kreuter, D. Werner, C. R. Noe, G. Kuemel, andA. Zimmer, Antisense Nucleic Acid Drug Dev. 7, 483 (1997).

283. O. Balland, T. Saison-Behmoaras, T. Garestier, and C. Helene, in“Targeting of Drugs: 5 Strategies for Oligonucleotide and GeneDelivery in Therapy” (G. Gregoriadis and B. McCormack, Eds.),Vol. 290, pp. 131–142. Plenum Press, New York, 1996.

284. I. Aynie, C. Vauthier, E. Fattal, and M. Foulquier, Alginatenanoparticles as a noval carrier for antisense oligonucleotides, in“Future Strategies for Drug Delivery with Particulate Systems,”Stuttgart, 1997, pp. 11–16.

285. E. Benoit, O. Prot, P. Maincent, and J. Bessiere, Pharm. Res. 11,585 (1994).

286. O. Prot, E. Benoit, P. Maincent, and J. Bessiere, J. Colloid InterfaceSci. 184, 251 (1996).

287. T. K. De and A. S. Hoffman, Artificial Cells Blood SubstitutesImmobilization Biotechnol. 29, 31 (2001).

288. L. Marchal-Heussler, D. Sirbat, M. Hoffman, and P. Maincent,Pharm. Res. 10, 386 (1993).

Pharmaceutical Nanotechnology 487

289. J. M. Irache, M. Berrabah, P. Verite, and S. Menager,Phenobarbitone-loaded poly-curly epsilon-caprolactone nano-capsules; In vitro kinetics and in vivo behaviour by the oral route,in “Formulation of Poorly-Available Drugs for Oral Administra-tion,” Paris, 1996, pp. 334–337.

290. M. Berrabah, D. Andre, F. Prevot, and A. M. Orecchioni, J. Pharm.Biomed. Anal. 12, 373 (1994).

291. H. Goto, T. Isobe, and M. Senna, Controlled dissolution of pheny-toin by complexing with silica nanoparticles modified by acrylamide, in “International Conference on Silica Science and Tech-nology,” Mulhouse, France, 1998, pp. 573–576.

292. H. Goto, T. Isobe, and M. Senna, Mater. Res. Soc. Symp. Proc. 501,321 (1998).

293. M. Fresta, G. Cavallaro, G. Giammona, E. Wehrli, and G. Puglisi,Biomaterials 17, 751 (1996).

294. V. Ferranti, H. Marchais, C. Chabenat, A. M. Orecchioni, andO. Lafont, Int. J. Pharm. 193, 107 (1999).

295. P. A. McCarron, A. D. Woolfson, and S. M. Keating, J. Pharm.Pharmacol. 52, 1451 (2000).

296. S. Keating, P. A. McCarron, and A. D. Woolfson, J. Pharm. Phar-macol. 52, 4 (2000).

297. B. T. Yu, Z. R. Zhang, and R. J. Zeng, Acta Pharm. Sinica 35, 705(2000).

298. H. S. Ch’ng, in “Formulation and Release Studies of 5-Fluorou-racil from Chitosan-Nanoparticles,” Penang, Malaysia, 1999,Malaysian Society of Pharmacology and Physiology/SingaporeUniversity Press, p. S56.

299. K. N. J. Burger, R. W. H. M. Staffhorst, H. C. de Vijlder, M. J.Velinova, P. H. Bomans, P. M. Frederik, and B. de Kruijff, Nat.Med. 8, 81 (2002).

300. K. Avgoustakis, A. Beletsi, Z. Panagi, P. Klepetsanis, A. G. Kary-das, and D. S. Ithakissios, J. Control. Release 79, 123 (2002).

301. M. A. Ecea, F. Gamisans, J. Valero, and M. E. Garcia, Farmaco49, 211 (1994).

302. L. Barraud, P. Merle, E. Soma, L. Lefrancois, S. Guerret,M. Chevallier, C. Dubernet, P. Couvreur, C. Trepo, and L. Vitvit-ski, J. Hepatol. 36, 82 (2002).

303. S. E. Gelperina, A. S. Khalansky, I. N. Skidan, Z. S. Smirnova,A. I. Bobruskin, S. E. Severin, B. Turowski, F. E. Zanella, andJ. Kreuter, Toxicol. Lett. 126, 131 (2002).

304. W. Ai-dong, C. Jiang-hao, and W. Dao-cheng, Chinese J. Antibiotics26, 464 (2001).

305. S. Mitra, U. Gaur, P. C. Ghosh, and A. N. Maitra, J. Control.Release 74, 317 (2001).

306. K. A. Janes, M. P. Fresneau, A. Marazuela, A. Fabra, andM. Alonso, J. Control. Release 73, 255 (2001).

307. H. S. Yoo, K. H. Lee, J. E. Oh, and T. G. Park, J. Control. Release68, 419 (2000).

308. S. C. Yang, H. X. Ge, Y. Hu, X. Q. Jiang, and C. Z. Yang, J. Appl.Polym. Sci. 78, 517 (2000).

309. A. Fundaro, R. Cavalli, A. Bargoni, D. Vighetto, G. P. Zara, andM. R. Gasco, Pharmacol. Res. 42, 337 (2000).

310. H. S. Yoo and T. G. Park, Polymer Preprints 41, 992 (2000).311. F. D. Kolodgie, M. John, C. Khurana, A. Farb, P. S. Wilson,

E. Acampado, N. Desai, P. Soon-Shiong, and R. Virmani, Circula-tion 106, 1195 (2002).

312. C. Fonseca, S. Simoes, and R. Gaspar, J. Control. Release 83, 273(2002).

313. D. B. Chen, T. Z. Yang, W. L. Lu, and Q. Zhang, Acta Pharm.Sinica 37, 58 (2002).

314. N. K. Ibrahim, N. Desai, S. Legha, P. Soon-Shiong, R. L. Theri-ault, E. Rivera, B. Esmaeli, S. E. Ring, A. Bedikian, and G. N.Hortobagyi, Clin. Cancer Res. 8, 1038 (2002).

315. B. Damascelli, F. Mattavelli, P. Tamplenizza, P. Bidoli, E. Leo,F. Dosio, A. M. Cerrolla, G. Di Tolla, L. F. Frigerio, and F. Garbag-nati, Cancer 92, 2592 (2001).

316. D. B. Chen, T. Yang, W. L. Lu, and Q. Zhang, Chem. Pharm. Bull.49, 1444 (2001).

317. M. G. Cascone, L. Lazzeri, C. Carmignani, and Z. Zhu, J. Mater.Sci. Mater. Med. 13, 523 (2002).

318. M. I. Llovet, M. A. Egea, J. Valero, and M. A. Alsina, Drug Dev.Ind. Pharm. 21, 1761 (1995).

319. M. Simeonova and M. Antcheva, Acta Physiol. Pharmacol. Bulg.20, 31 (1994).

320. M. Simeonova, M. Ilarionova, T. Ivanova, and C. Konstantinov,Acta Physiol. Pharmacol. Bulg. 17, 43 (1991).

321. A. Lamprecht, Y. Bouligand, and J. P. Benoit, J. Control. Release84, 59 (2002).

322. H. Lboutounne, J. F. Chaulet, C. Ploton, F. Falson, and F. Pirot,J. Control. Release 82, 319 (2002).

323. H. Lboutounne, F. Pirot, M. A. Belfihadj, G. Degobert, J. Y.Dusseau, and F. Falson, Antimicrobial activity and percutaneousabsorption of chlorhexidine base from poly-epsilon-caprolactonnanocapsules, “Perspectives in Percutaneous penetration,” p. 102.STS Publishing, La Grande Motte, France, 2000.

324. M. Sznitowska, M. Gajewska, S. Janicki, A. Radwanska, andG. Lukowski, Eur. J. Pharm. Biopharm. 52, 159 (2001).

325. P. Calvo, C. Remuan-Lopez, J. L. Vila-Jato, and M. J. Alonso,Colloid Polym. Sci. 275, 46 (1997).

326. R. Cavalli, O. Caputo, M. E. Carlotti, M. Trotta, C. Scarnecchia,and M. R. Gasco, Int. J. Pharm. 148, 47 (1997).

327. E. Ugazio, R. Cavalli, and M. R. Gasco, Int. J. Pharm. 241, 341(2002).

328. Q. Zhang, G. Yie, Y. Li, Q. Yang, and T. Nagai, Int. J. Pharm.200, 153 (2000).

329. Q. Zhang, G. Q. Ye, Y. Li, and Q. S. Yang, Acta Phara. Sinica 34,308 (1999).

330. J. Molpeceres, M. Chacon, M. Guzman, L. Berges, and M. delRosario Aberturas, Int. J. Pharm. 187, 101 (1999).

331. P. Calvo, A. Sanchez, J. Martinez, M. I. Lopez, M. Calonge, J. C.Pastor, and M. J. Alonso, Pharm. Res. 13, 311 (1996).

332. S. Bonduelle, M. Carrier, C. Pimienta, J. P. Benoit, andV. Lenaerts, Eur. J. Pharm. Biopharm. 42, 313 (1996).

333. E. Harris, Z. Ramtoola, and J. G. Kelly, Pharm. Technol. Conf. 1,301 (1995).

334. J. Miones, E. Yebra-Pimentel, O. Conde, and E. Iribarnegaray,Langmuir 10, 1888 (1994).

335. S. D. Desai and J. Blanchard, Drug Delivery 7, 201 (2000).336. A. Zimmer and J. Kreuter, in “Ocular Drug Delivery” (R. D.

Schoenwald, Ed.), Vol. 16, Number 1, p. 61. Elsevier Science Divi-sion, New York, 1995.

337. A. Zimmer, E. Mutschler, G. Lambrecht, and D. Mayer, Pharm.Res. 11, 1435 (1994).

338. B. Heurtault, P. Legrand, V. Mosqueira, J. P. Devissaguet, G. Bar-ratt, and C. Bories, Ann. Trop. Med. Parasitol. 95, 529 (2001).

339. J. M. Rodrigues, H. Fessi, C. Bories, F. Puisieux, and J. P. Devis-saguet, Int. J. Pharm. 126, 253 (1995).

340. J. M. Rodrigues, S. L. Croft, H. Fessi, and C. Bories, Trop. Med.Parasitol. 45, 223 (1994).

How Does The Colloid Mill Work?The product to be dispersed or emulsifi ed is fed by gravity or under pressure to a rapidly spinning rotor. This is set very closely and precisely to a matching stationary part called a stator. The distance between them is adjusted from 0.001 to 0.125”. Generally distances of not more than 0.020” are needed. This distance is commonly called the grinding gap. (See Fig. 2)

Premier Colloid Mills - The Ultimate In Dispersing and Emulsifying

Fig. 1 - Left: Ordinary oil-in-water emulsion made with propeller mixer, magnifi ed 1000 times. Note dispro-portionate size of particles. Right: Same emulsion process through a colloid mill. Note smaller and more uniform size of particles.

As the material comes in contact with the rotor it is fl ung out to the edge by centrifugal force. This force pushes the material down through the narrow gap between the rotor and stator. A particle material, while it is whirling around in this fi lm, is subjected to a great many eddy currents of tremendous force. This imparts high shear to the product which overcomes the surface forces tending to hold the particles together. The prod-uct makes its way through the shear zone and fi nally is fl ung out into an open area. At this point it passes out of the colloid mill through a suitable opening.

An almost limitless variety of materials can be processed. Viscosities can run from water thin to paste thick. As long as the product will fl ow (either by gravity or pressure) a Premier

Colloid Mill can handle it.

What Does A Colloid Mill Do?A colloid mill does its work by hydraulic shear, bringing to bear a tremendous amount of energy on a small portion of material in the form of a thick fi lm.

This action overcomes the strong polar forces which bind together small clumps of solids or which hold together drops of liquid. A colloid mill will not break down hard cyrstalline particles by fracturing them across the crystal planes as an im-pact type mill would do. It will, however, reduce these particles down to their ultimate crystal size by breaking up the agglom-erates into which they form.

In the case of emulsions, the same principle holds. As the par-ticles of the dispersed phase of the emulsion get smaller and it requires progressively more energy to overcome the surface tension holding them together. Enormous hydraulic shear is needed to do the job and a colloid mill is an ideal means of accomplishing it. (See Fig. 1)

Clearance

RotorStator

CoolantOutlet

CoolantInlet

Bearing

Areas showjacketing for heating and cooling with in-lets and outlets indicated.

CoolantOutlet

Hand Wheel

Spillway

SpillwayCover

RotorStator

Adjusting Ring

CoolantInlet

Inlet Funnel(Fig. 3)

All Premier Colloid Mills have securely held vertical shafts in a heavy duty housing called the Spillway. This permits products contained in the inlet funnel to imme-diately enter the high shear zone. (See Fig. 3)

The Rotor and StatorThe heart of a colloid mill is the rotor and stator. The best design is a rotor and a stator with smooth, ground and lapped faces. With this design, found in the Pre-mier Colloid Mill, the thin fi lm of material between them will be of uniform cross section and will be subjected to the maximum amount of shear as it whirls around. To ensure fast entry of product into the grinding gap all Colloid Mills are provided with impellers (two pins for low viscosity products and a special blade for thick material). The rapidly turning impeller creates centrifu-gal force to pull the material into the grinding zone.

(Fig. 2)

(Fig. 1)

The speed at which a colloid mill runs is extremely impor-tant. The linear speed at the rotor face, where the work is done, must be high enough to develop suffi cient hydrau-lic shear. To obtain superior peripheral speeds, the two small scale Premier Colloid Mills (3” and 4”) have high speed belt drives. The larger mills (6” to 10”) have large diameter rotors which provide high linear speed at the edges when directly coupled to 3600 rpm motors.

(Fig. 4)

Both rotor and shaft are made of “Invar”, an alloy with the lowest coeffi cient of expansion of any commercially avail-able material. If the product heats up during the grinding operation, the mill temperature change does not affect the grinding gap, which stays constant. Rotors are faced with a thick layer of Stellite, harder then most abrasive materials. A matching, Stellite-faced, Invar ring is set into the stainless steel stator, (See Fig. 4)

When the product contains fi bers, crystals or similar hard agglomerates - rough, abrasive surfaces are required. Premier uses the preferred material, silicon carbide. The stator has a specially made silicon carbide insert and the rotor a similar matching part. (See Fig. 4)

In addition to the standard silicon carbide components, fi ne grain and wide faced rotors and stators are available.

GAP ADJUSTMENT: The clearance between the rotor and stator can be increased, decreased, or locked in seconds - even while the mill is operating. No tools are needed. Loosen the hand wheels which allow the non-galling magnesium bronze calibrated adjusting ring to be easily rotated from 0.125” down to 0.001”.

LABYRINTH SEAL: The triple heavy-duty bearings are protected by the easy-to-maintain Labyrinth Seal. (See Fig.5). Product is prevented from entering the upper bearings by a tortuous fl ow path of two concentric rings in the rotor and seal cap. Material entering the outer ring is ejected by centrifugal force through the drain holes. Product cannot get into the bearings under normal use.

HEATING AND COOLING: All mills have the stator and spillway (main housing) jacketed for use with steam, hot or cold water. Fittings for these connectors are an integral part of the mill. Materials which have to be kept in a molten state can be easily handled. Heat build-up can be minimized by using a cooling fl uid in the jacket. The colloid mill can be operated at temperatures up to 350° F.

CLEANING: Product contact parts are stainless steel. A quick fl ush is suffi cient to clean the mill after use. For more diffi cult products or for more thorough cleaning, every bit of surface touched by the product can be ex-posed in less than one minute. No tools are needed. The adjusting ring is backed off. The stator is lifted out. The interior is now completely accessible.

Closed Continuous OperationAny Premier Colloid Mill may be easily converted to closed continuous operation. The inlet funnel is un-screwed and appropriate inlet piping screwed into the stator. The standard partial spillway cover is replaced with a full spillway cover having appropriate pipe fi tting. A pump may feed the mill or the mill may be mounted underneath a tank for gravity feeding. (Fig. 6)

(Fig. 5)

Rotor (Underside)

Drain Holes Seal Cap (Top)

Flow of Material

Shaft LabyrinthSeal Cap

StatorInsert

Rotor

Stainless Steel Stator

Stellite Insert Ring

Inlet

Silicon CarbideStator Insert

Invar Rotor

Stellite Facing Invar ShaftSilicon CarbideRotor

SMOOTH FACEDROTOR AND STATOR

SILICON CARBIDEROTOR AND STATOR

ZETASIZER NANOSZ BRIEF USER MANUAL

IMPORTANT: to use laser in an efficient way, it is recommended to turn the

instrument on and off once in a day.

Measurement limitations:

Size: 0,6nm-6um

Light source: 633nm laser light

Size measurement:

Cuvette type Solvent type for cuvette Max temp. for cuvette

PS

water

50°C Chemicals which PS is

resistant

Glass Other chemicals 90°C

1. Firstly, the instrument is turned on by pressing the button on the back of the

instrument and the LED lamp on the instrument become orenge. wait 20

minutes for the instrument is ready.

2. Run the software by clicking the shortcut of the Zetasizer on the desktop. The

LED lamp on the instrument has to be green.

3. To open file

If you open a new folder

new→new folder( enter name)→file name

If you open a new file (to your folder)

new→file name

If you open old file

Open→your folder→old file

NOT: don’t use another folder or file

4. Select measurement type

5. measure→manual

measurement type→size

sample name: same as file name

material:you can choose from list

if there is no your material you must add your material. click on the “….”

İcon,and add the refractive index and absorption value of your sample.

NOT: if you don’t know absorption value; Conc. solution: 0,1

Fogged solution: 0,01

Clear solution: 0,001

Dispersant:you can choose from list

If there is no your dispersant you must add your dispersant. Click on the

“…” icon. And enter the necessary informations.

If the solvent is a complex system, by following the path add→additive the

system can be added.

general options: don’t change

temperature: it can be adjusted to the value that is required.

Equilibration time: 5min + 1 min per degrees changed.

a. cell type: Disposible sizing cuvette

measurements:

angle: 173° default. (if the sample contains large particles and the back

scattering is required to be collected, it can be selected as ”dual”)

duration: autamatic default.

Number of measurements:suggested value is 3 of course arbitrary

Delay between measurements: no change

Advanced options:no change

data processing:

analysis mode:normal resolution (default)

click OK, and click start on the window opened.

2) To analyze the data:

If there is fluctuations in the count rate, there may be dust particles in the

sample.

If the correlation graph is not flat, there can be thermal change.

The longer the correlation graph stays flat, the larger the particle.

If the result is not what you expected, you can look at expert advice menu

To view size graphs:

tools→size→intensity PSD (M)

polydispersity index (PDI) α 1/intercept

widht: 1/3 below the peak point.

Peak point: the diameter of the particle which exists in the large quantity.

the measurements can be obtained in three different values; number, intensity

and volume

wheter the dispersities of the particules are evaluated according to the following

polydispersty index

0 to 0.05 - Only normally encountered with latex standards or particles made

to be monodisperse

0.05 to 0.08 - Nearly monodisperse sample. PCS cannot normally extract a

distribution within this range

0.08 to 0.7 - This is a mid-range polydispersity, it is the range over which the

distribution algorithm based on NNLS best operates over

Greater than 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

You should encounter correlation function during the analysis. Here is the optimum

graph and what the graph means:

Here is another graph which is not represents a good data

This graph means we have large particles medium polydispersity index and we have

some dust or aggregates.

11/15/2012

1

Dynamic Light Scattering

ZetaPALS w/ 90Plus particle size analyzer

Also equipped w/ BI-FOQELS & Otsuka DLS-700 (Rm CCR230)

Dynamic Light Scattering (DLS) Photon Correlation Spectroscopy (PCS) Quasi-Elastic Light Scattering (QELS)

1. Measure Brownian motion by … 2. Collect scattered light from suspended particles to … 3. Obtain diffusion rate to … 4. Calculate particle size

11/15/2012

2

Brownian motion • Velocity of the Brownian motion is defined by the

Translational Diffusion Coefficient (D) • Brownian motion is indirectly proportional to size

– Larger particles diffuse slower than smaller particles • Temperature and viscosity must be known • Temperature stability is necessary

– Convection currents induce particle movement that interferes with size determination

• Temperature is proportional to diffusion rate – Increasing temperature increases Brownian motion

Brownian motion

Random movement of particles due to bombardment of solvent molecules

11/15/2012

3

Stokes-Einstein Equation

dH = hydrodynamic diameter (m) k = Boltzmann constant (J/K=kg·m2/s2·K) T = temperature (K) η = solvent viscosity (kg/m·s) D = diffusion coefficient (m2/s)

D

kTdH

3

Hydrodynamic diameter

• The diameter measured by DLS correlates to the effective particle movement within a liquid

• Particle diameter + electrical double layer

• Affected by surface bound species which slows diffusion

Hydrodynamic diameter

Particle diameter

11/15/2012

4

Nonspherical particles

Equivalent sphere

Hydrodynamic diameter is calculated based on the equivalent sphere with the same diffusion coefficient

Rapid

Slow

=

Experimental DLS

• Measure the Brownian motion of particles and calculate size

• DLS measures the intensity fluctuations of scattered light arising from Brownian motion

• How do these fluctuations in scattered light intensity arise?

11/15/2012

5

What causes light scattering from (small) particles?

• Explained by JW Strutt (Lord Rayleigh) – Electromagnetic wave (light) induces oscillations of

electrons in a particle – This interaction causes a deviation in the light path

through an angle calculated using vector analysis – Scattering coefficient varies inversely with the fourth

power of the wavelength

6

2

12

2

12

4

2

2

022

12

2

cos1

d

nn

nn

RII

Interaction of light with matter Rayleigh approximation

• For small particles (d ≤ λ/10), scattering is isotropic

• Rayleigh approximation tells us that

I α d6

I α 1/λ4

where I = intensity of scattered light d = particle diameter λ = laser wavelength

11/15/2012

6

Mie scattering from large particles

• Used for particles where d ~ λ0

• Complete analytical solution of Maxwell’s

equations for scattering of electromagnetic radiation from spherical particles

• Assumes homogeneous, isotropic and optically linear material

Stratton, A. Electromagnetic theory, McGraw-Hill, New York (1941) www.lightscattering.de/MieCalc

Brownian motion and scattering

Constructive interference

Destructive interference

11/15/2012

7

Brownian motion and scattering Multiple particles

Instrument layout

Laser

Correlator

11/15/2012

8

Intensity fluctuations • Apply the

autocorrelation function to determine diffusion coefficient

• Large particles – smooth curve

• Small particles – noisy curve

Determining particle size

• Determine autocorrelation function • Fit measured function to G(τ) to calculate Γ • Calculate D, given n*, θ, and Γ • Calculate dH, given T* and η*

*User defined values.

11/15/2012

9

How a correlator works

• Random motion of small particles in a liquid gives rise to fluctuations in the time intensity of the scattered light

• Fluctuating signal is processed by forming the autocorrelation function • Calculates diffusion

How a correlator works

• Large particles – the signal will be changing slowly and the correlation will persist for a long time

• Small, rapidly moving particles – the correlation will disappear quickly

11/15/2012

10

The correlation function • For monodisperse particles the correlation function is

• Where – A= baseline of the correlation function – B=intercept of the correlation function – Γ =Dq2

• D=translational diffusion coefficient • q=(4πn/λ0)sin(θ/2)

– n=refractive index of solution – λ0=wavelength of laser – θ=scattering angle

BAG )2exp()(

The correlation function

• For polydisperse particles the correlation function becomes

where g1(τ) is the sum of all exponential decays contained in the correlation function

2

1 )(1)( BgAG

11/15/2012

11

Broad particle size distribution

• Correlation function becomes nontrivial – Measurement noise, baseline drifts, and dust

make the function difficult to solve accurately • Cumulants analysis

– Convert exponential to Taylor series – First two cumulants are used to describe data

• Γ = Dq2 • μ2 = (D2*-D*2)q4

• Polydispersity = μ2/ Γ2

Cumulants analysis

• The decay in the correlation function is exponential

• Simplest way to obtain size is to use cumulants analysis1

• A 3rd order fit to a semi-log plot of the correlation function

• If the distribution is polydisperse, the semi-log plot will be curved

• Fit error of less than 0.005 is good.

1ISO 13321:1996 Particle size analysis -- Photon correlation spectroscopy

11/15/2012

12

Cumulants analysis • Third order fit to correlation function

– b = z-average diffusion coefficient – 2c/b2 = polydispersity index

• This method only calculates a mean and width

– Intensity mean size – Only good for narrow, monomodal samples – Use NNLS for multimodal samples

2)( cbaCLn

Cumulants analysis

11/15/2012

13

Polydispersity index • 0 to 0.05 – only normally encountered w/ latex standards

or particles made to be monodisperse

• 0.05 to 0.08 – nearly monodisperse sample

• 0.08 to 0.7 – This is a mid-range polydispersity

• >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 intensity-

weighted 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

11/15/2012

14

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

11/15/2012

15

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/agglomerates (noisy baseline)

11/15/2012

16

Data interpretation Correlograms

• Very large particles • High polydispersity • Presence of large

particles/agglomerates (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

11/15/2012

17

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

Laser

Detector

11/15/2012

18

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

11/15/2012

19

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

11/15/2012

20

Sample concentration Small particles

• For particle sizes <10 nm, one must determine the minimum concentration to generate enough scattered light

• Particles should generate ~ 10 kcps (count rate) in excess of the scattering from the solvent

• Maximum concentration determined by the physical properties of the particles – Avoid particle/particle interactions – Should be at least 1000 particles in the scattering

volume

Sample preparation • When possible, perform DLS on as prepared sample

• Dilute aqueous or organic suspensions – Alcohol and aggressive solvents require a glass/quartz cell

– 0.0001 to 1%(v/v)

• Dilution media (1) should be the same (or as close as possible) as the synthesis media, (2) HPLC grade and (3) filtered before use – Chemical equilibrium will be established if diluent is taken from

the original sample

• Suspension should be sonicated prior to analysis

11/15/2012

21

Checking instrument operation

• DLS instruments are not calibrated

– Measurement based on first principles

– Verification of accuracy can be checked using standards

• Duke Scientific (based on TEM) • Polysciences

Count rate and z-average diameter Repeatability

• Perform at least 3 repeat measurements on the same sample

• Count rates should fall within a few percent of one another

• z-average diameter should also be with 1-2% of one another

11/15/2012

22

Count rate Repeatability problems

• Count rate DECREASES with successive measurements – Particle sedimentation – Particle creaming – Particle dissolution or breaking up

• Resolution

– Prepare a better, stabilized dispersion – Get rid of large particles – Coulter

Count rate Repeatability problems

• Count rate is RANDOM with successive measurements – Dispersion instability – Sample contains large particles – Bubbles

• Resolution

– Prepare a better, stabilized dispersion – Remove large particles – De-gas sample

11/15/2012

23

Z-average diameter Repeatability problems

• Size DECREASES with successive measurements – Temperature not stable – Sample unstable

• Resolution

– Allow plenty of time for temperature equilibration

– Prepare a better, stabilized dispersion

Repeatability of size distributions

• The sized distributions are derived from a NNLS analysis and should be checked for repeatability as well

• If distributions are not repeatable, repeat measurements with longer measurement duration

11/15/2012

24

References • http://www.bic.com/90Plus.html • http://www.brainshark.com/brainshark/vu/view.asp?text=M913802&pi

=62212 • http://www.malvern.co.uk/malvern/ondemand.nsf/frmondemandview • http://www.brainshark.com/brainshark/vu/view.asp?text=M913802&pi

=96389 • http://www.brainshark.com/brainshark/vu/view.asp?text=M913802&pi

=73504 • http://physics.ucsd.edu/neurophysics/courses/physics_173_273/dyna

mic_light_scattering_03.pdf • http://www.brookhaven.co.uk/dynamic-light-scattering.html • Dynamic Light Scattering: With Applications to Chemistry, Biology,

and Physics, Bruce J. Berne and Robert Pecora, DOVER PUBLICATIONS, INC. Mineola. New York

• Scattering of Light & Other Electromagnetic Radiation, Milton Kerker, Academic Press (1969)

11/15/2012

25

Evaluating the correlation function

• If the intensity distribution is a fairly smooth peak, there is little point in conversion to a volume distribution using Mie theory

• However, if the intensity plot shows a substantial tail or more than one peak, then a volume distribution will give a more realistic view of the importance of the tail or second peak

• Number distributions are of little use because small error in data acquisition can lead to huge error in the distribution by number and are not displayed

Correlogram from a sample of small particles

11/15/2012

26

Correlogram from a sample of large particles

Extra • Time-dependent fluctuations in the scattered intensity

due to Brownian motion • Constructive and destructive interference of light • Decay times of fluctuations related to the diffusion

constants --- particle size • Fluctuations determined in the time domain by a

correlator • Correlation – average of products of two quantities • Delay times chosen to be much smaller than the time

required for a particle to relax back to average scattering

Surfactants: Fundamentalsand Applications

in the Petroleum Industry

Laurier L. SchrammPetroleum Recovery Institute

PUBLISHED BY THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGEPUBLISHED BY THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGE

The Pitt Building, Trumpington Street, Cambridge, United Kingdom

CAMBRIDGE UNIVERSITY PRESSCAMBRIDGE UNIVERSITY PRESS

The Edinburgh Building, Cambridge CB2 2RU, UK http://www.cup.cam.ac.uk40 West 20th Street, New York, NY 10011-4211, USA http://www.cup.org10 Stamford Road, Oakleigh, Melbourne 3166, AustraliaRuiz de AlarcoÂn 13, 28014 Madrid, Spain

# Cambridge University Press 2000

This book is in copyright. Subject to statutory exceptionand to the provisions of relevant collective licensing agreements,no reproduction of any part may take place withoutthe written permission of Cambridge University Press.

First published 2000

Printed in the United Kingdom at the University Press, Cambridge

Typeset in New Caledonia 10.75/12pt, in 3B21 [PNPN]

A catalogue record for this book is available from the British Library

Library of Congress cataloguing in publication dataSurfactants: fundamentals and applications in the petroleum industry / Laurier L. Schramm, editor.

p. cm.Includes index.ISBN 0 521 64067 91. Surface active agents ± Industrial applications. 2. Petroleum industry and trade.I. Schramm, Laurier Lincoln.

TN871.S76784 2000665.5Ðdc21 99-15820 CIP

ISBN 0 521 64067 9 hardback

CONTENTS

Preface vii

SURFACTANTURFACTANT FUNDAMENTALSUNDAMENTALS

1. Surfactants and Their Solutions: Basic Principles 3Laurier L. Schramm and D. Gerrard Marangoni

2. Characterization of Demulsifiers 51R.J. Mikula and V.A. Munoz

3. Emulsions and Foams in the Petroleum Industry 79Laurier L. Schramm and Susan M. Kutay

SURFACTANTSURFACTANTS ININ POROUSOROUS MEDIAEDIA

4. Surfactant Adsorption in Porous Media 121Laura L. Wesson and Jeffrey H. Harwell

5. Surfactant Induced Wettability Alteration in Porous Media 159Eugene A. Spinler and Bernard A. Baldwin

6. Surfactant Flooding in Enhanced Oil Recovery 203Tor Austad and Jess Milter

7. Scale-Up Evaluations and Simulations of Mobility ControlFoams for Improved Oil Recovery 251Fred Wassmuth, Laurier L. Schramm, Karin Mannhardt, andLaurie Hodgins

OILWELLILWELL, NEAREAR-WELLELL, ANDAND SURFACEURFACE OPERATIONSPERATIONS

8. The Use of Surfactants in Lightweight Drilling Fluids 295Todd R. Thomas and Ted M. Wilkes

9. Surfactant Use in Acid Stimulation 329Hisham A. Nasr-El-Din

10. Surfactants in Athabasca Oil Sands Slurry Conditioning,Flotation Recovery, and Tailings Processes 365Laurier L. Schramm, Elaine N. Stasiuk, and Mike MacKinnon

ENVIRONMENTALNVIRONMENTAL, HEALTHEALTH, ANDAND SAFETYAFETY APPLICATIONSPPLICATIONS

11. Surfactant Enhanced Aquifer Remediation 433Varadarajan Dwarakanath and Gary A. Pope

v

12. Use of Surfactants for Environmental Applications 461Merv Fingas

13. Toxicity and Persistence of Surfactants Used in thePetroleum Industry 541Larry N. Britton

GLOSSARYLOSSARY ANDAND INDEXESNDEXES

14. Glossary of Surfactant Terminology 569Laurier L. Schramm

Author Index 613

Affiliation Index 614

Subject Index 615

vi Contents

1Surfactants and Their Solutions:Basic Principles

Laurier L. Schramm1,2 and D. Gerrard Marangoni3

1Petroleum Recovery Institute, 100, 3512 ± 33rd St. NW, Calgary, AB,Canada T2L 2A62 University of Calgary, Dept. of Chemistry, 2500 University Drive NW,Calgary, AB, Canada T2N 1N43St. Francis Xavier University, Dept. of Chemistry, PO Box 5000,Antigonish, NS, Canada B2G 2W5

This chapter provides an introduction to the occurrence, proper-ties and importance of surfactants as they relate to the petroleumindustry. With an emphasis on the definition of important terms,the importance of surfactants, their micellization and adsorptionbehaviours, and their interfacial properties are demonstrated. Itis shown how surfactants may be applied to alter interfacialproperties, promote oil displacement, and stabilize or destabilizedispersions such as foams, emulsions, and suspensions. Under-standing and controlling the properties of surfactant-containingsolutions and dispersions has considerable practical importancesince fluids that must be made to behave in a certain fashion toassist one stage of an oil production process, may requireconsiderable modification in order to assist in another stage.

Introduction

Surfactants are widely used and find a very large number of applicationsbecause of their remarkable ability to influence the properties of surfacesand interfaces, as will be discussed below. Some important applications ofsurfactants in the petroleum industry are shown in Table 1. Surfactantsmay be applied or encountered at all stages in the petroleum recovery andprocessing industry, from oilwell drilling, reservoir injection, oilwellproduction, and surface plant processes, to pipeline and seagoing trans-portation of petroleum emulsions. This chapter is intended to provide anintroduction to the basic principles involved in the occurrence and uses ofsurfactants in the petroleum industry. Subsequent chapters in this bookwill go into specific areas in greater detail.

3

All the petroleum industry's surfactant applications or problems havein common the same basic principles of colloid and interface science. Thewidespread importance of surfactants in general, and scientific interest intheir nature and properties, have precipitated a wealth of publishedliterature on the subject. Good starting points for further basic informa-tion are classic books like Rosen's Surfactants and Interfacial Phenomena[1] andMyers' Surfactant Science and Technology [2], and the many otherbooks on surfactants [3±19]. Most good colloid chemistry texts containintroductory chapters on surfactants. Good starting points are references[20±23], while for much more detailed treatment of advances in specificsurfactant-related areas the reader is referred to some of the chaptersavailable in specialist books [24±29]. With regard to the occurrence ofrelated colloidal systems in the petroleum industry, three recent books

Table 1. Some Examples of SurfactantApplications in the Petroleum Industry

Gas/Liquid SystemsProducing oilwell and well-head foamsOil flotation process frothDistillation and fractionation tower foamsFuel oil and jet fuel tank (truck) foamsFoam drilling fluidFoam fracturing fluidFoam acidizing fluidBlocking and diverting foamsGas-mobility control foams

Liquid/Liquid SystemsEmulsion drilling fluidsEnhanced oil recovery in situ emulsionsOil sand flotation process slurryOil sand flotation process frothsWell-head emulsionsHeavy oil pipeline emulsionsFuel oil emulsionsAsphalt emulsionOil spill emulsionsTanker bilge emulsions

Liquid/Solid SystemsReservoir wettability modifiersReservoir fines stabilizersTank/vessel sludge dispersantsDrilling mud dispersants

4 SURFACTANTSURFACTANTS: FUNDAMENTALSUNDAMENTALS ANDAND APPLICATIONSPPLICATIONS ININ THETHE PETROLEUMETROLEUM INDUSTRYNDUSTRY

describe the principles and occurrences of emulsions, foams, and suspen-sions in the petroleum industry [30±32].

Definition and Classification of Surfactants4

Some compounds, like short-chain fatty acids, are amphiphilic or amphi-pathic, i.e., they have one part that has an affinity for nonpolar media andone part that has an affinity for polar media. These molecules formoriented monolayers at interfaces and show surface activity (i.e., theylower the surface or interfacial tension of the medium in which they aredissolved). In some usage surfactants are defined as molecules capable ofassociating to form micelles. These compounds are termed surfactants,amphiphiles, surface-active agents, tensides, or, in the very old literature,paraffin-chain salts. The term surfactant is now probably the mostcommonly used and will be employed in this book. This word has asomewhat unusual origin, it was first created and registered as a trade-mark by the General Aniline and Film Corp. for their surface-activeproducts.5 The company later (ca. 1950) released the term to the publicdomain for others to use [33]. Soaps (fatty acid salts containing at leasteight carbon atoms) are surfactants. Detergents are surfactants, orsurfactant mixtures, whose solutions have cleaning properties. That is,detergents alter interfacial properties so as to promote removal of a phasefrom solid surfaces.

The unusual properties of aqueous surfactant solutions can beascribed to the presence of a hydrophilic head group and a hydrophobicchain (or tail) in the molecule. The polar or ionic head group usuallyinteracts strongly with an aqueous environment, in which case it issolvated via dipole±dipole or ion±dipole interactions. In fact, it is thenature of the polar head group which is used to divide surfactants intodifferent categories, as illustrated in Table 2. In-depth discussions ofsurfactant structure and chemistry can be found in references [1, 2, 8, 34,35].

The Hydrophobic Effect and Micelle Formation

In aqueous solution dilute concentrations of surfactant act much asnormal electrolytes, but at higher concentrations very different behaviourresults. This behaviour is explained in terms of the formation of organizedaggregates of large numbers of molecules called micelles, in which the

4 A glossary of frequently encountered terms in the science and engineering ofsurfactants is given in the final chapter of this book.5 For an example of one of GAF Corp's. early ads promoting their trademarkedsurfactants, see Business Week, March 11, 1950, pp. 42±43.

1. SCHRAMMCHRAMM & MARANGONIARANGONI Basic Principles 5

Table 2. Surfactant Classifications

Class Examples Structures

Anionic Na stearate CH3(CH2)16COO7Na+

Na dodecyl sulfate CH3(CH2)11SO47Na+

Na dodecyl benzene sulfonate CH3(CH2)11C6H4SO37Na+

Cationic Laurylamine hydrochloride CH3(CH2)11NH3+Cl7

Trimethyl dodecylammonium chloride C12H25N+(CH3)3Cl

7

Cetyl trimethylammonium bromide CH3(CH2)15N+(CH3)3Br

7

Nonionic Polyoxyethylene alcohol CnH2n+1(OCH2CH2)mOHAlkylphenol ethoxylate C9H19ÐC6H4Ð(OCH2CH2)nOHPolysorbate 80 HO(C2H4O)w (OC2H4)xOHw+ x+ y+ z= 20,R= (C17H33)COO

CH(OC2H4)yOH|CH2(OC2H4)zR

Propylene oxide-modifiedpolymethylsiloxane

(CH3)3SiO((CH3)2SiO)x(CH3SiO)ySi(CH3)3|

EO= ethyleneoxy CH2CH2CH2O(EO)m(PO)nHPO=propyleneoxy

Zwitterionic Dodecyl betaine C12H25N+(CH3)2CH2COO7

Lauramidopropyl betaine C11H23CONH(CH2)3N+(CH3)2CH2COO7

Cocoamido-2-hydroxy-propyl sulfobetaine CnH2n+1CONH(CH2)3N+(CH3)2CH2CH(OH)CH2SO3

7

lipophilic parts of the surfactants associate in the interior of the aggregateleaving hydrophilic parts to face the aqueous medium. An illustrationpresented by Hiemenz and Rajagopalan [22] is given in Figure 1. Theformation of micelles in aqueous solution is generally viewed as acompromise between the tendency for alkyl chains to avoid energeticallyunfavourable contacts with water, and the desire for the polar parts tomaintain contact with the aqueous environment.

A thermodynamic description of the process of micelle formation willinclude a description of both electrostatic and hydrophobic contributionsto the overall Gibbs energy of the system. Hydrocarbons (e.g., dodecane)and water are not miscible; the limited solubility of hydrophobic speciesin water can be attributed to the hydrophobic effect. The hydrophobicGibbs energy (or the transfer Gibbs energy) can be defined as thedifference between the standard chemical potential of the hydrocarbonsolute in water and a hydrocarbon solvent at infinite dilution [36±40]

DG8t = m8HC7m8aq (1)

where m8HC and m8aq are the chemical potentials of the hydrocarbondissolved in the hydrocarbon solvent and water, respectively, and DG8t is

Figure 1. Schematic representation of the structure of an aqueousmicelle showing several possibilities: (a) overlapping tails in the centre,(b) water penetrating to the centre, and (c) chains protruding andbending. (From Hiemenz and Rajagopalan [22]. Copyright 1997 MarcelDekker Inc., New York.)

1. SCHRAMMCHRAMM & MARANGONIARANGONI Basic Principles 7

the Gibbs energy for the process of transferring the hydrocarbon solutefrom the hydrocarbon solvent to water. In a homologous series ofhydrocarbons (e.g., the n-alcohols or the n-alkanes), the value of DG8tgenerally increases in a regular fashion

DG8t = (a7 bnc)RT (2)

where a and b are constants for a particular hydrocarbon series and nc isthe number of carbon atoms in the chain. The transfer Gibbs energy, DG8t,can be divided into entropic and enthalpic contributions

DG8t =DH8t7TDS8t (3)

where DH8t and DS8t are the enthalpy and entropy of transfer, respectively.A significant characteristic of the hydrophobic effect is that the entropyterm is dominant, i.e., the transfer of the hydrocarbon solute from thehydrocarbon solvent to water is accompanied by an increase in the Gibbstransfer energy (DG4 0) [41]. The decrease in entropy is thought to bethe result of the breakdown of the normal hydrogen-bonded structure ofwater accompanied by the formation of differently structured water, oftentermed icebergs, around the hydrocarbon chain. The presence of thehydrophobic species promotes an ordering of water molecules in thevicinity of the hydrocarbon chain. To minimize the large entropy effect,the ``icebergs'' tend to cluster [38], in order to reduce the number of watermolecules involved; the ``clustering'' is enthalpically favoured (i.e.,DH5 0), but entropically unfavourable. The overall process has thetendency to bring the hydrocarbon molecules together, which is knownas the hydrophobic interaction. Molecular interactions, arising from thetendency for the water molecules to regain their normal tetrahedralstructure, and the attractive dispersion forces between hydrocarbonchains, act cooperatively to remove the hydrocarbon chain from thewater ``icebergs'', leading to an association of hydrophobic chains.

Due to the presence of the hydrophobic effect, surfactant moleculesadsorb at interfaces, even at low surfactant concentrations. As there willbe a balance between adsorption and desorption (due to thermalmotions), the interfacial condition requires some time to establish. Thesurface activity of surfactants should therefore be considered a dynamicphenomenon. This can be determined by measuring surface or interfacialtensions versus time for a freshly formed surface, as will be discussedfurther below.

At a specific, higher, surfactant concentration, known as the criticalmicelle concentration (cmc), molecular aggregates termed micelles areformed. The cmc is a property of the surfactant and several other factors,since micellization is opposed by thermal and electrostatic forces. A lowcmc is favoured by increasing the molecular mass of the lipophilic part ofthe molecule, lowering the temperature (usually), and adding electrolyte.

8 SURFACTANTSURFACTANTS: FUNDAMENTALSUNDAMENTALS ANDAND APPLICATIONSPPLICATIONS ININ THETHE PETROLEUMETROLEUM INDUSTRYNDUSTRY

Surfactant molar masses range from a few hundreds up to severalthousands.

The most commonly held view of a surfactant micelle is not muchdifferent than that published by Hartley in 1936 [41, 42] (see Figure 1). Atsurfactant concentrations slightly above the cmc value, surfactants tend toassociate into spherical micelles, of about 50±100 monomers, with aradius similar to that of the length of an extended hydrocarbon chain.The micellar interior, being composed essentially of hydrocarbon chains,has properties closely related to the liquid hydrocarbon.

Critical Micelle Concentration

It is well known that the physico-chemical properties of surfactants varymarkedly above and below a specific surfactant concentration, the cmcvalue [2±9, 13, 14, 17, 35±47]. Below the cmc value, the physico-chemicalproperties of ionic surfactants like sodium dodecylsulfate, SDS, (e.g.,conductivities, electromotive force measurements) resemble those of astrong electrolyte. Above the cmc value, these properties change drama-tically, indicating a highly cooperative association process is taking place.In fact, a large number of experimental observations can be summed up ina single statement: almost all physico-chemical properties versus concen-tration plots for a given surfactant±solvent system will show an abruptchange in slope in a narrow concentration range (the cmc value). This isillustrated by Preston's [48] classic graph, shown in Figure 2.

In terms of micellar models, the cmc value has a precise definition inthe pseudo-phase separation model, in which the micelles are treated as aseparate phase. The cmc value is defined, in terms of the pseudo-phasemodel, as the concentration of maximum solubility of the monomer in thatparticular solvent. The pseudo-phase model has a number of short-comings; however, the concept of the cmc value, as it is described interms of this model, is very useful when discussing the association ofsurfactants into micelles. It is for this reason that the cmc value is,perhaps, the most frequently measured and discussed micellar parameter[39].

Cmc values are important in virtually all of the petroleum industrysurfactant applications. For example, a number of improved or enhancedoil recovery processes involve the use of surfactants including micellar,alkali/surfactant/polymer (A/S/P) and gas (hydrocarbon, N2, CO2 orsteam) flooding. In these processes, surfactant must usually be present ata concentration higher than the cmc because the greatest effect of thesurfactant, whether in interfacial tension lowering [30] or in promotingfoam stability [31], is achieved when a significant concentration ofmicelles is present. The cmc is also of interest because at concentrations

1. SCHRAMMCHRAMM & MARANGONIARANGONI Basic Principles 9

above this value the adsorption of surfactant onto reservoir rock surfacesincreases very little. That is, the cmc represents the solution concentra-tion of surfactant from which nearly maximum adsorption occurs.

Cmc Measurements. The general way of obtaining the cmcvalue of a surfactant micelle is to plot some physico-chemical property of

Figure 2. Illustration of the dramatic changes in physical propertiesthat occur beyond the critical micelle concentration. (From Preston [48].Copyright 1948 American Chemical Society, Washington.)

10 SURFACTANTSURFACTANTS: FUNDAMENTALSUNDAMENTALS ANDAND APPLICATIONSPPLICATIONS ININ THETHE PETROLEUMETROLEUM INDUSTRYNDUSTRY

interest versus the surfactant concentration and observe the break in theplot. Table 3 lists the most common cmc methods. Many of these methodshave been reviewed by Shinoda [11] and Mukerjee and Mysels [49]. Itshould be noted that different experimental techniques may give slightlydifferent values for the cmc of a surfactant. However, Mukerjee andMysels [49], in their vast compilation of cmc values, have noted that themajority of values for a single surfactant (e.g., sodium dodecyl sulfate, orSDS, in the absence of additives) are in good agreement and the outlyingvalues are easily accounted for.

For petroleum industry processes, one tends to have a special interestin the cmc's of practical surfactants that may be anionic, cationic, nonionicor amphoteric. The media are typically high salinity, high hardnesselectrolyte solutions, and in addition, the cmc values of interest span thefull range from ambient laboratory conditions to oil and gas reservoirconditions of temperature and pressure. Irrespective of aiming forprocess development and optimization under realistic (reservoir) condi-tions of temperature and pressure, it remains common to determine cmc'sexperimentally at ambient laboratory conditions and assume that thesame hold even at elevated temperatures and pressures. This can be anextremely dangerous assumption.

The nature and limits of applicability of specific methods for deter-mining critical micelle concentrations vary widely. Most methods havebeen developed for a relatively small set of pure surfactants involving verydilute electrolyte solutions and only ambient temperature and pressure.The determination of cmc at elevated temperature and pressure isexperimentally much more difficult than for ambient conditions andcomparatively little work has been done in this area. Most high tempera-ture cmc studies have been by conductivity measurements and havetherefore been limited to ionic surfactants. For example, cmc's at up to166 8C have been reported by Evans and Wightman [50]. Some work hasbeen reported using calorimetry, up to 200 8C by Noll [51], and using 19F

Table 3. Some Common Cmc Methods

UV/Vis, IR spectroscopyFluorescence spectroscopyNuclear magnetic resonance spectroscopyElectrode potential/conductivityVoltametryScattering techniquesCalorimetrySurface tensionFoaming

1. SCHRAMMCHRAMM & MARANGONIARANGONI Basic Principles 11

NMR, up to 180 8C by Shinoda et al. [52]. Some work has been reportedinvolving cmc determination by calorimetry (measuring heats of dilutionor specific heats). Archer et al. [53] used flow calorimetry to determinethe cmc's of several sulfonate surfactants at up to 178 8C. Noll [51]determined cmc's for dodecyltrimethylammonium bromide and commer-cial surfactants in the temperature range 25±200 8C using flow calorime-try. Surface tension is the classical method for determining cmc's butmany surface tension methods are not suitable for use with aqueoussolutions at elevated temperatures. Exceptions include the pendant,sessile, and captive drop methods which can be conducted with high-pressure cells [54, 55].

For any of the techniques applied it appears (Archer et al. [53]) thatthe uncertainties in the experimental cmc determinations increase withincreasing temperature because at the same time the surfactant aggrega-tion number decreases and the aggregation distribution increases. That is,the concentration range over which micellization occurs broadens withincreasing temperature. Almost all of the elevated temperature cmcstudies have involved carefully purified surfactants (not commercialsurfactants or their formulations) in pure water or very dilute electrolytesolutions. Conducting cmc determinations at elevated pressure, as well astemperature, is even more difficult and only a few studies have beenreported, mostly employing conductivity methods (La Mesa et al. [56];Sugihara and Mukerjee [57]; Brun et al. [58]; Kaneshina et al. [59];Hamann [60]) which, again, are unsuitable for nonionic or zwitterionicsurfactants and for use where the background electrolyte concentrationsare significant.

In the case where one needs to be able to determine cmc's for nonionicor zwitterionic surfactants, in electrolyte solutions that may be veryconcentrated, and at temperatures and pressures up to those that maybe encountered in improved oil recovery operations in petroleumreservoirs, most of the established methods are not practical. Onesuccessful approach to this problem has been to use elevated tempera-ture and pressure surface tension measurements involving the captivedrop technique [8] although this method is quite time-consuming.Another approach is to use dynamic foam stability measurements.Foaming effectiveness and the ease of foam formation are related tosurface tension lowering and to micelle formation, the latter of whichpromotes foam stability through surface elasticity and other mechanisms[61]. Accordingly, static or dynamic foam height methods generally showthat foam height increases with surfactant concentration and thenbecomes relatively constant at concentrations greater than the cmc(Rosen and Solash [62]; Goette [63]). Using a modified Ross-Miles staticfoam height apparatus, Kashiwagi [64] determined the cmc of SDSat 40 8C to be 7.08 mM which compared well with values attained

12 SURFACTANTSURFACTANTS: FUNDAMENTALSUNDAMENTALS ANDAND APPLICATIONSPPLICATIONS ININ THETHE PETROLEUMETROLEUM INDUSTRYNDUSTRY

by conductivity (7.2 mM) and surface tension (7.2 mM). Rosen andSolash [62] also found that foam production was related to cmc usingthe Ross-Miles method at 60 8C when they assessed SDS, potassiumtetradecyl sulfonate, potassium hexadecyl sulfonate, and sodium hexa-decyl sulfate.

Morrison et al. [65] describe a dynamic foam height method for theestimation of cmc's that is suitable for use at high temperatures andpressures. This method is much more rapid than the surface tensionmethod, and is applicable to a wide range of surfactant classes, includingboth ionic and amphoteric (zwitterionic) surfactants. The method issuitable for the estimation of cmc's, for determining the minimum cmcas a function of temperature, for identifying the temperature at which theminimum cmc occurs, and for determining how cmc's vary with signifi-cant temperature and pressure changes. The method has been used todetermine the temperature variation of cmc's for a number of commercialfoaming surfactants in aqueous solutions, for the derivation of thermo-dynamic parameters, and to establish useful correlations [55].

Cmc Values. Some typical cmc values for low electrolyte con-centrations at room temperature are:

Anionics 1073±1072 MAmphoterics 1073±1071 MCationics 1073±1071 MNonionics 1075±1074 M

Cmc values show little variation with regard to the nature of the chargedhead group. The main influence appears to come from the charge ofthe hydrophilic head group. For example, the cmc of dodecyltrimethyl-ammonium chloride (DTAC) is 20 mM, while for a 12 carbon nonionicsurfactant, hexaethylene glycol mono-n-dodecyl ether (C12E6), the cmc isabout 0.09 mM [39, 41, 49]; the cmc for SDS is about 8 mM, while thatfor disodium 1,2-dodecyldisulfate (1,2-SDDS) is 40 mM [66]. In additionto the relative insensitivity of the cmc value of the surfactant to the natureof the charged head group, cmc's show little dependence on the nature ofthe counter-ion. It is mainly the valence number of the counter-ion thataffects the cmc. As an example, the cmc value for Cu(DS)2 is about1.2 mM, while the cmc for SDS is about 8 mM [49, 67].

Cmc values often exhibit a weak dependence on both temperature[68±70] and pressure [59, 71], although, as shown in Figure 3, somesurfactant cmc's have been observed to increase markedly with tempera-ture above 100 8C [55, 65]. The effects of added substances on the cmcare complicated and interesting, and depend greatly on whether theadditive is solubilized in the micelle, or in the intermicellar solution. Theaddition of electrolytes to ionic surfactant solutions results in a well

1. SCHRAMMCHRAMM & MARANGONIARANGONI Basic Principles 13

established linear dependence of log (cmc) on the concentration of addedsalt [72±76]. For nonionic micelles, electrolyte addition has little effect oncmc values. When non-electrolytes are added to the micellar solution, theeffects are dependent on the nature of the additive. For polar additives(e.g., n-alcohols), the cmc decreases with increasing concentration ofalcohol, while the addition of urea to micellar solutions tends to increasethe cmc, and may even inhibit micelle formation [77, 78]. Nonpolaradditives tend to have little effect on the cmc [79].

Figure 3. Temperature variation of the critical micelle concentrations ofthree amphoteric surfactants in 2.1% total dissolved solids brine solu-tions. (From Stasiuk and Schramm [55]. Copyright 1996 Academic Press,New York.)

14 SURFACTANTSURFACTANTS: FUNDAMENTALSUNDAMENTALS ANDAND APPLICATIONSPPLICATIONS ININ THETHE PETROLEUMETROLEUM INDUSTRYNDUSTRY

The Krafft Point

The solubilities of micelle-forming surfactants show a strong increaseabove a certain temperature, termed the Krafft point (Tk). This isexplained by the fact that the single surfactant molecules have limitedsolubility whereas the micelles are very soluble. Referring to the illustra-tion from Shinoda [11] in Figure 4, below the Krafft point the solubility ofthe surfactant is too low for micellization so solubility alone determinesthe surfactant monomer concentration. As temperature increases thesolubility increases until at Tk the cmc is reached. At this temperature arelatively large amount of surfactant can be dispersed in micelles andsolubility increases greatly. Above the Krafft point maximum reduction insurface or interfacial tension occurs at the cmc because the cmc thendetermines the surfactant monomer concentration. Krafft points for anumber of surfactants are listed in references [1, 80].

Nonionic surfactants do not exhibit Krafft points. Instead, the solubilityof nonionic surfactants decreases with increasing temperature, and thesesurfactants may begin to lose their surface active properties above atransition temperature referred to as the cloud point. This occurs becauseabove the cloud point a surfactant rich phase of swollen micellesseparates, and the transition is usually accompanied by a marked increasein dispersion turbidity.

Figure 4. Example of a ``phase behaviour'' diagram for a surfactant inaqueous solution, showing the cmc and Krafft points. (From Shinoda et al.[11]. Copyright 1963 Academic Press, New York.)

1. SCHRAMMCHRAMM & MARANGONIARANGONI Basic Principles 15

Analysis

Numerous methods have been developed for the quantitative determina-tion of each class of surfactant. The analysis of commercial surfactants isgreatly complicated by the fact that these products are mixtures. They areoften comprised of a range of molar mass structures of a given structuralclass, may contain surface-active impurities, are sometimes intentionallyformulated to contain several different surfactants, and are often supplieddissolved in mixed organic solvents or complex aqueous salt solutions.Each of these components has the potential to interfere with a givenanalytical method. Therefore surfactant assays may well have to bepreceded by surfactant separation techniques. Both the separation andassay techniques can be highly specific to a given surfactant/solutionsystem. This makes any substantial treatment beyond the scope of thepresent chapter. Good starting points can be found in the several books onsurfactant analysis [81±86]. The characterization and analysis of surfactantdemulsifiers is discussed in Chapter 2 of this book. Table 4 shows sometypical kinds of analysis methods that are applied to the differentsurfactant classes.

Table 4. Typical Methods of Surfactant Analysis

Surfactant Class Method

Anionicalkyl sulfates and sulfonates Two-phase or surfactant-electrode monitored

titrationpetroleum and lignin sulfonates Column or gel permeation chromatographyphosphate esters Potentiometric titrationsulfosuccinate esters Gravimetric or titration methodscarboxylates Potentiometric titration or two-phase titration

Nonionicalcohols NMR or IR spectroscopyethoxylated acids Gas chromatographyalkanolamides Gas chromatographyethoxylated amines HPLCamine oxides Potentiometric titration

Cationicquaternary ammonium salts Two-phase or surfactant-electrode monitored

titration, or GC or HPLC

Amphotericcarboxybetaines Low pH two-phase titration, gravimetric analysis,

or potentiometric titrationsulfobetaines HPLC

16 SURFACTANTSURFACTANTS: FUNDAMENTALSUNDAMENTALS ANDAND APPLICATIONSPPLICATIONS ININ THETHE PETROLEUMETROLEUM INDUSTRYNDUSTRY

There are a number of reviews available for surfactants in specificindustries [87], and for specific surfactant classes. References [81±90]discuss methods for the determination of anionic surfactants, which areprobably the most commonly encountered in the petroleum industry.Most of these latter methods are applicable only to the determination ofsulfate- and sulfonate-functional surfactants. Probably the most commonanalysis method for anionic surfactants is Epton's two-phase titrationmethod [91, 92] or one of its variations [93, 94]. Related, single-phasetitrations can be performed and monitored by either surface tension [95]or surfactant-sensitive electrode [84, 85, 96±98] measurements. Grons-veld and Faber [99] discuss adaptation of the titration method to oleicphase samples.

Surfactants and Surface Tension

In two-phase dispersions, a thin intermediate region or boundary, knownas the interface, lies between the two phases. The physical properties ofthe interface can be very important in all kinds of petroleum recovery andprocessing operations. Whether in a well, a reservoir or a surfaceprocessing operation, one tends to encounter large interfacial areasexposed to many kinds of chemical reactions. In addition, many petro-leum industry processes involve colloidal dispersions, such as foams,emulsions, and suspensions, all of which contain large interfacial areas;the properties of these interfaces may also play a large role in determiningthe properties of the dispersions themselves. In fact, even a modestsurface energy per unit area can become a considerable total surfaceenergy. Suppose we wish to make a foam by dispersion of gas bubbles intowater. For a constant gas volume fraction the total surface area producedincreases as the bubble size produced decreases. Since there is a freeenergy associated with surface area, this increases as well with decreasingbubble size. The energy has to be added to the system to achieve thedispersion of small bubbles. If this amount of energy cannot be provided,say through mechanical energy input, then another alternative is to usesurfactant chemistry to lower the interfacial free energy, or interfacialtension. The addition of a small quantity of a surfactant to the water,possibly a few tenths of a percent, would significantly lower the surfacetension and significantly lower the amount of mechanical energy neededfor foam formation. For examples of this simple calculation for foams andemulsions, see references [61] and [100] respectively.

The origin of surface tension may be visualized by considering themolecules in a liquid. The attractive van der Waals forces betweenmolecules are felt equally by all molecules except those in the interfacialregion. This imbalance pulls the latter molecules towards the interior ofthe liquid. The contracting force at the surface is known as the surface

1. SCHRAMMCHRAMM & MARANGONIARANGONI Basic Principles 17

tension. Since the surface has a tendency to contract spontaneously inorder to minimize the surface area, bubbles of gas tend to adopt aspherical shape: this reduces the total surface free energy. For emulsionsof two immiscible liquids a similar situation applies to the droplets of oneof the liquids, except that it may not be so immediately obvious whichliquid will form the droplets. There will still be an imbalance ofintermolecular force resulting in an interfacial tension, and the interfacewill adopt a configuration that minimizes the interfacial free energy.Physically, surface tension may be thought of as the sum of the contract-ing forces acting parallel to the surface or interface. This point of viewdefines surface or interfacial tension (g), as the contracting force per unitlength around a surface. Another way to think about surface tension is thatarea expansion of a surface requires energy. Since the work required toexpand a surface against contracting forces is equal to the increase insurface free energy accompanying this expansion, surface tension mayalso be expressed as energy per unit area.

There are many methods available for the measurement of surface andinterfacial tensions. Details of these experimental techniques and theirlimitations are available in several good reviews [101±104]. Table 5 showssome of the methods that are used in petroleum recovery processresearch. A particular requirement of reservoir oil recovery processresearch is that measurements be made under actual reservoir conditionsof temperature and pressure. The pendant and sessile drop methods arethe most commonly used where high temperature/pressure conditions arerequired. Examples are discussed by McCaffery [105] and DePhilippis etal. [106]. These standard techniques can be difficult to apply to themeasurement of extremely low interfacial tensions (51 to 10 mN/m).For ultra-low tensions two approaches are being used. For moderatetemperatures and low pressures the most common method is thatof the spinning drop, especially for microemulsion research [107]. Forelevated temperatures and pressures a captive drop method has beendeveloped by Schramm et al. [108], which can measure tensions as low as0.001 mN/m at up to 200 8C and 10,000 psi. In all surface and interfacialtension work it should be appreciated that when solutions, rather thanpure liquids, are involved appreciable changes can occur with time at thesurfaces and interfaces, so that techniques capable of dynamic measure-ments tend to be the most useful.

When surfactant molecules adsorb at an interface they provide anexpanding force acting against the normal interfacial tension. Thus,surfactants tend to lower interfacial tension. This is illustrated by thegeneral Gibbs adsorption equation for a binary, isothermal systemcontaining excess electrolyte:

Gs =7(1/RT)(dg/d lnCs) (4)

18 SURFACTANTSURFACTANTS: FUNDAMENTALSUNDAMENTALS ANDAND APPLICATIONSPPLICATIONS ININ THETHE PETROLEUMETROLEUM INDUSTRYNDUSTRY

Table 5. Surface and Interfacial Tension Methods used in Petroleum Research

Static Dynamic Surface Interfacial High T, PMethod Values Values Tension Tension Contact Angle Capability

Capillary rise [ & [ ` `, need y= 0 `

Wilhelmy plate [ & [ ` [, need to know g `

du Nouy ring [ ` [ ` `, pure liquids only `

Drop weight [ ` [ [ `, need y= 0 [

Drop volume [ ` [ [ `, need y= 0 [

Pendant drop [ [ [ [ ` [

Sessile drop [ [ [ [ [ [

Oscillating jet [ [ [ ` ` `

Spinning drop [ & [ [ ` `

Captive drop [ [ [ [ `, forces y= 0 [

Maximum bubble pressure [ & [ ` ` `

Surface laser light scattering [ [ [ & ` [

Tilting plate [ & ` ` [ `

where Gs is the surface excess of surfactant (mol/cm2), Cs is the solutionconcentration of the surfactant (M), and g may be either surface orinterfacial tension (mN/m). This equation can be applied to dilutesurfactant solutions where the surface curvature is not great and wherethe adsorbed film can be considered to be a monolayer. The packingdensity of surfactant in a monolayer at the interface can be calculated asfollows. According to equation 4, the surface excess in a tightly packedmonolayer is related to the slope of the linear portion of a plot of surfacetension versus the logarithm of solution concentration. From this, the areaper adsorbed molecule (aS) can be calculated from

aS = 1/(NAGs) (5)

where NA is Avogadro's number. Numerous examples are given by Rosen[1].

When surfactants concentrate in an adsorbed monolayer at a surfacethe interfacial film may take on any of a number of quite differentproperties which will be discussed in the next several sections. Suitablyaltered interfacial properties can provide a stabilizing influence indispersions such as emulsions, foams, and suspensions.

Surface Elasticity

As surfactant adsorbs at an interface the interfacial tension decreases (atleast up to the cmc), a phenomenon termed the Gibbs effect. If asurfactant stabilized film undergoes a sudden expansion, the immediatelyexpanded portion of the film must have a lower degree of surfactantadsorption than unexpanded portions because the surface area hasincreased. This causes an increased local surface tension which producesimmediate contraction of the surface. The surface is coupled, by viscousforces, to the underlying liquid layers. Thus, the contraction of the surfaceinduces liquid flow, in the near-surface region, from the low tensionregion to the high tension region. The transport of bulk liquid due tosurface tension gradients is termed the Marangoni effect [27]. In foams,the Gibbs±Marangoni effect provides a resisting force to the thinning ofliquid films.

The Gibbs±Marangoni effect only persists until the surfactant adsorp-tion equilibrium is re-established in the surface, a process that may takeplace within seconds or over a period of hours. For bulk liquids and inthick films this can take place quite quickly, however, in thin films theremay not be enough surfactant in the extended surface region to re-establish the equilibrium quickly, requiring diffusion from other parts ofthe film. The restoring processes are then the movement of surfactantalong the interface from a region of low surface tension to one of high

20 SURFACTANTSURFACTANTS: FUNDAMENTALSUNDAMENTALS ANDAND APPLICATIONSPPLICATIONS ININ THETHE PETROLEUMETROLEUM INDUSTRYNDUSTRY

surface tension, and the movement of surfactant from the thin film intothe now depleted surface region. Thus the Gibbs±Marangoni effectprovides a force to counteract film rupture in foams.

Many surfactant solutions show dynamic surface tension behaviour.That is, some time is required to establish the equilibrium surface tension.After the surface area of a solution is suddenly increased or decreased(locally), the adsorbed surfactant layer at the interface requires some timeto restore its equilibrium surface concentration by diffusion of surfactantfrom, or to, the bulk liquid (see Figure 5, [109]). At the same time, sincesurface tension gradients are now in effect, Gibbs±Marangoni forces actin opposition to the initial disturbance. The dissipation of surface tensiongradients, to achieve equilibrium, embodies the interface with a finiteelasticity. This explains why some substances that lower surface tensiondo not stabilize foams [21]; they do not have the required rate of approachto equilibrium after a surface expansion or contraction. In other words,they do not have the requisite surface elasticity.

At equilibrium, the surface elasticity, or surface dilational elasticity,EG, is defined [21, 110] by

EG � dgd lnA

�6�where g is the surface tension and A is the geometric area of the surface.This is related to the compressibility of the surface film, K, by K=1/EG.EG is a thermodynamic property, termed the Gibbs surface elasticity. Thisis the elasticity that is determined by isothermal equilibrium measure-ments, such as the spreading pressure±area method [21]. EG occurs invery thin films where the number of molecules is so low that thesurfactant cannot restore the equilibrium surface concentration afterdeformation. An illustration is given in [61].

The elasticity determined from nonequilibrium dynamic measure-ments depends upon the stresses applied to a particular system, isgenerally larger in magnitude than EG, and is termed the Marangonisurface elasticity, EM [21, 111]. For foams it is this dynamic property thatis of most interest. Surface elasticity measures the resistance againstcreation of surface tension gradients and of the rate at which suchgradients disappear once the system is again left to itself [112]. TheMarangoni elasticity can be determined experimentally from dynamicsurface tension measurements that involve known surface area changes,such as the maximum bubble pressure method [113, 115]. Although suchmeasurements include some contribution from surface dilational viscosity[112, 114] the results are frequently simply referred to in terms of surfaceelasticities.

Numerous studies have examined the relation between EG or EM andfoam stability [111, 112, 115]. From low bulk surfactant concentrations,

1. SCHRAMMCHRAMM & MARANGONIARANGONI Basic Principles 21

Figure 5. Illustration of the Gibbs±Marangoni effect in a thin liquidfilm. Reaction of a liquid film to a surface disturbance. (a) Low surfactantconcentration yields only low differential tension in film. The thin film ispoorly stabilized. (b) Intermediate surfactant concentration yields astrong Gibbs±Marangoni effect which restores the film to its originalthickness. The thin film is stabilized. (c) High surfactant concentration(4cmc) yields a differential tension which relaxes too quickly due todiffusion of surfactant. The thinner film is easily ruptured. (From Pugh[109]. Copyright 1996 Elsevier, Amsterdam.)

22 SURFACTANTSURFACTANTS: FUNDAMENTALSUNDAMENTALS ANDAND APPLICATIONSPPLICATIONS ININ THETHE PETROLEUMETROLEUM INDUSTRYNDUSTRY

2622 E. Ruckenstein and R. Nagarajan

Critical Micelle Concentration. A Transition Point for Micellar Size Distribution

E. Ruckenstein. and R. Nagarajan

Faculty of Engineering and Applied Sciences, State University of New York at Buffalo, Buffalo, New York 14214 (Received July 21, 1975)

Publication costs assisted by the State University of New York at Buffalo

A critical concentration is defined which separates two kinds of behavior of the size distribution of micellar aggregates of surfactant molecules. Below this concentration, the size distribution is a monotonic decreas- ing function of size; above this concentration, the size distribution is a function exhibiting two extrema and the contribution of larger aggregates becomes important. Quantitative results are obtained from the condi- tion of thermodynamic equilibrium, decomposing the standard chemical potential per amphiphile of vari- ous aggregates into size-dependent and size-independent terms, and using for these terms the empirical ex- pressions given by Tanford.

Introduction Micellization has been treated either as a stepwise asso-

ciation phenomenon or as a phase transition process.lV2 In the first approach the micellar aggregates and the single molecules of surfactant are assumed to be in association- dissociation equilibrium and the law of mass action is ap- plied. The critical micelle concentration (crnc) is defined as the concentration above which any added surfactant mole- cules appear with high probability as micellar aggregates. In the latter approach, micellization is regarded as a phase separation starting a t cmc which now represents the satu- ration concentration of the phase containing single mole- cules of surfactant. Although both approaches can explain some features of the micelle formation, the available exper- imental evidence seems to be in agreement with the first of these. In the region close to the cmc, the physico-chemical properties do indeed change rapidly, but continuously, and the concentration of the single surfactant molecules in- creases slowly. The addition of surfactant above the cmc not only leads to an increase in the number of aggregates but dso gives rise to larger average sizes of the aggregates. It is therefore reasonable to assume that for all concentra- tions of surfactant there are micellar aggregates of various sizes. At low concentrations, however, the number and size of these aggregates is small. As the concentration increases, both the number and the average size of the aggregates in- crease. The particular kind of behavior occurring a t cmc is caused by an essential change in the shape of the size dis- tribution of the aggregates; above the cmc the participation of the larger sizes is more important.

The model suggested here is similar to that provided by the first approach. The nature of the transition taking place at cmc is, however, better identified. In the present paper it is shown that a t low concentrations of surfactant the size distribution of the aggregates is a monotonic de- creasing function of the size. As the concentration in- creases, the size distribution of the aggregates changes from a monotonic decreasing function to one which has two extrema, a minimum and a maximum.

From this description a well-defined critical concentra- tion emerges as that total amphiphile concentration corre- sponding to a transition from the monotonic decreasing size distribution function to a size distribution function ex- hibiting two extrema. This critical concentration corre- sponds to a surfactant solution with no appreciable amount

of aggregates, whereas the cmc, as defined in the earlier pa- pers based upon the first approach, corresponds to systems with about 5-10% of amphiphiles in micellar form. Hence, the critical concentration defined here as the transition point separating the two different behaviors of the aggre- gate size distribution function constitutes a close lower bound on the observed value of the cmc.

In the present paper the equilibrium condition defined by the minimum of the total free energy determines the size distribution of aggregates for a given concentration of the single surfactant molecules. The standard chemical po- tential per amphiphile of aggregates is decomposed into size-independent and size-dependent parts for which the empirical expressions given by Tanford3 are used. The value of the critical concentration is computed as the con- centration for which the size distribution curve has an in- flection point. More refined computations, based on statis- tical mechanics, following the methods of Poland and Sher- aga4 and of Hoeve and Benson? will be published in anoth- er paper.

Thermodynamic Formulation The standard chemical potential of a single molecule of

the solvent is denoted by ps and that of a single molecule of amphiphile in the aqueous medium by PA. The number of molecules of the solvent and the monomeric amphiphile are denoted by N s and NA, respectively, and the number of ag- gregates containing g molecules each by Ng. The standard chemical potential of the micelle per amphiphile is sepa- rated into two components: one is independent of the size of the aggregate and the second is size dependent. The size- independent part of the standard chemical potential for a single amphiphile, p ~ , accounts for the liquidlike behavior of the micellar interior made up of hydrocarbon tails of am- phiphiles. The size of the aggregate and the compactness of the packing of the amphiphiles affect the repulsive interac- tions between the head groups, as well as the magnitude of the hydrocarbon surface area exposed to the aqueous medi- um which does not take part in the hydrophobic bonding. The size-dependent part of the standard chemical potential accounts for these contributions and is denoted by pg for an aggregate of size g.

Assuming ideal behavior of the solution, the free energy of the system is given by

The Journal of Physical Chemistry, Vol. 79, No. 24, 1975

Transition Point for Micellar Size Distribution 2823

unity from n, here accounts for the fact that the CH2 group closest to the hydrophilic head group lies in its hydration sphere and, hence, does not contTibute to the hydrophobic bonding energy. The term I depends upon the character of the head group. For nonionic head groups, it is virtually constant a t a value of -700 cal/mol while for ionic head groups, it may be taken as having a value of approximately -420 cal/mol. The term S depends upon the specificity of the head group, particularly for nonionic micelles, but may be assumed to be of the order of -2000 cal/mol.

The size-dependent term Pg/kT includes (a) the decrease in attractive hydrophobic bonding between hydrocarbon tails of amphiphiles, due to their partial exposure to the aqueous medium, and (b) the repulsive interaction between the hydrophilic head groups of the amphiphiles. This re- pulsive interaction is caused by steric repulsion between head groups in nonionic micelles and by electrical repulsion between the ionic head groups in ionic micelles.

The positive standard free-energy contribution due to contact between hydrocarbon tails of amphiphiles and the aqueous medium is given by T a n f ~ r d ~ ~ as

where m

F = N s + N A + Ng ( 2 ) g=2

k is the Boltzmann constant and T the absolute tempera- ture. The total number of amphiphiles present as monomer and as aggregates is

m

N = NA + gNg = constant ( 3 ) g=2

The equilibrium condition leads to the equation

which can be rewritten as

Denoting by 6 the quantity

the size distribution function is expressed in the form

(7)

The decomposition of the standard chemical potentials per amphiphile of the aggregates permits the size distribu- tion function to be written as a product of exp [ - ( k g / k T ) ] , which is a decreasing function of g , and another factor con- taining the size independent quantity (. If 6 is small, the size distribution is a monotonic decreasing function of g . If, however, .$ is sufficiently large compared to unity, the size distribution can have a maximum. Therefore, a critical value of 6 exists, separating the two kinds of behavior. To compute this critical value as well as the corresponding critical concentration of amphiphile, explicit expressions for the terms of the standard chemical potentials are need- ed.

Explicit Expressions for the Free Energy Terms The size-independent free-energy change PB - PA corre-

sponds to the change in the standard free energy in trans- ferring an amphiphile from the dilute aqueous phase to the hydrocarbon phase of the micellar core. The transfer of an amphiphile into a micelle differs from the transfer into a bulk hydrocarbon phase because of the orientation of the amphiphile within the micelle with the head groups in con- tact with the aqueous medium. This free-energy change has significantly different values for ionic and nonionic mi- celles and also depends on the specificity of the hydrophilic head groups.

For illustrative purposes, the expression given by Tan- ford3b is used for the size-independent part WB - PA of the standard free-energy change.

In general, for an alkyl chain with n, carbon atoms

W B - PA S + I(n, - 1 ) -- - k T R T

k= I 25(A - 21)g k T R T

where A is the surface area in (&ngstroms)2 per amphiphile in an aggregate of size g . For a spherical aggregate of radius

(10 ) 1 A = - 4?rro2 g

The radius ro of the aggregate, calculated from the volume uo of the amphiphile, is

ro

(11 )

The volume of the hydrocarbon tail of an amphiphile is de- termined from3b

( 1 2 )

The assumption of a spherical shape for the micelle is not a restrictive one and is used here for the sake of simplicity only. Any other shape of the aggregate or even changes in the shape of the aggregates with the growth in size can be incorporated in the formulation.

The second contribution to the size-dependent standard free-energy term arises from the repulsive head group in- teractions. The magnitude of this interaction depends on the separation between the head groups; the available area per head group is used as a measure of this separation. Tanford has shown3b that an expression of the form

uo = 27.4 + 26.9nC A3

k T A R T g (13 )

can be used to represent this interaction, where a is a con- stant independent of g . The constant a has different values for ionic and nonionic micelles and depends, among other parameters, on the nature of the hydrophilic head group and the ionic strength. The size-dependent part of the free energy can now be written in the form

Expression for Critical Concentration The value of the parameter .$ at which the size distribu- where R has a value of 2 cal/mol K. The subtraction of

The Journal of Physical Chemistry, Vol. 79, No. 24, 1975

2624 E. Ruckenstein and R. Nagarajan

tion function NgIF has a point of inflection is denoted as Fcrit and the corresponding value of the aggregation number is denoted as gcrit. The values of fcrit and gcrit are obtained from the solution of the system of equations ’ (In $) = In - - (”) = 0 at g = gcrit (15a)

3 = exp [“(PB;; p ~ ) F ] exp (- $1 = tmaxg exp (- $1

(23) where PB’ is the chemical potential per molecule of the bulk amphiphilar phase. The value of corresponding to this phase transition can be calculated from

d2 d2 lntmax=-= ( - kT ) - (y) (24)

The term PB’ - PA is the standard free-energy change for the transfer of an amphiphile molecule from the aqueous medium to the bulk amphiphilar phase. Since no informa- tion is available for this term a t the present time, it is set

d dg dg kT

and PB’ - PB PB‘ - P A

kT - ( l n $ ) = - - ( h ) = O dg2 kT atgagcrit (15b) dg2

Substituting expression 14 for the size-dependent free-en- ergy term bg and eq 10 and 11 for the surface area A Per amphiphile in eq 15a and 15b yields for the inflection point

and

0 (17)

The aggregation number at the inflection point is obtained from eq 17 and is given by

(18)

The value of &it a t the inflection point, obtained from eq 18 and eq 16, is

800 112 In Ecrit = ((9) - 525)lRT (19)

The total amphiphile concentration at this value o f t = fcrit gives the critical concentration Ccrit.

Replacing the summation by an integral, eq 20 becomes

(20)

Ccrit = tcrit exp ( y) +

There is some analogy between the critical concentration as defined here and the critical temperature predicted by the van der Waals equation of state, since each of them separates two kinds of behavior of the size distribution function and pressure-volume relationship, respectively.

Phase Separation There is a saturation value of the concentration of non-

aggregated amphiphile at which phase separation occurs. The nonaggregated amphiphile and aggregates in solution are in this case in thermodynamic equilibrium with a bulk amphiphilar phase. The corresponding thermodynamic equilibrium condition introduces an upper limit, [max, for 6. The equilibrium condition leads to

and

equal to the standard free-energy change for transferring a hydrocarbon chain from an aqueous solution of monomer to a pure hydrocarbon phase.3a Consequently

(25) PB’ - P A (400 to 2000) - 800nc -- -

k T R T

Results and Discussion The size distribution of micellar aggregates N,IF is plot-

ted against the aggregation number, g, for an amphiphile with an octyl hydrocarbon tail and for a = 2 x lo4 cal A2/ moPb (Figure 1). Equation 19 leads to [crjt = 3.88. For t < &it, the size distribution is a monotonic decreasing func- tion of g. A t f = &, the size distribution function has an inflection point. At ,$ > [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 in- crease.

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 vol- ume of the hydrophobic core. In reality the surface area should account for the surface roughness due to the pres- ence of the head groups. The actual surface area can be cal- culated 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 lead- ing to the equations for gcrit and [crit more complex. For this case, the value of &it can be computed only by numer- ical methods. Figure 2 represents the size distribution func- tion for 6 = 3 A and a = 8 X lo4 cal A2/mol for an amphi- phile with an octyl hydrocarbon tail.

The mole fractions of amphiphiles present both in non- aggregated form and as aggregates for various values of are presented in Table I for the same parameters as in Fig- ure 2. Table I shows that below Ecrit = 27.82, the concentra- tion of micellar aggregates remains extremely small. At = &it, the concentration of micellar aggregates begins to in- crease even though their contribution to the total amphi- phile 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

The Journal of Physical Chemistry, Vol. 79, No. 24, 1975

2625 Transition Point for Micellar Size Distribution

E . 3 7 3

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 4 /mol, n, = 8.

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 ex- perimental cmc of amphiphiles with hexoxyethy- lene glycol monoether head groups and various hydrocar- bon chain lengths. I t 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. I t 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 re- pulsive interaction term a and remains a constant for am-

TABLE I: Variation of the Mole Fraction of the Aggregates as a Function of the Mole Fraction of the Nonaggregated Amphiphilea

5 N J F ~

22.14 2.09 x 10-4 1.53 x 10-1'

3.61 x 10-9 39.18 3.67 x l V 4 1.28 x 10-4

2.34 x 10-4 1.97 X lo-'' 24.98 27.82 2.61 x 10-4 2.54 X 1 r 1 ' 33.50

44.86 4.21 X l V 4 8.59' Mole fraction of

nonaggregated amphiphile. Mole fraction of the aggregates. See the explanation in the text.

3.14 x 10-4

=n, = 8, a = 8 X lo4 cal A2/mol, 6 = 3 A.

TABLE I IVar ia t ion of the Critical Concentration as a Function of Hydrocarbon Chain Lengtha

Exptl cmcbpc % 5 cri t Calcd Ccritb

2.64 x 10-3 2.76 x 10-3 8 27.8 2.61 x 10-4 2.59 x 10-4

10 29.0 2.58 X l V 5 2.44 x 10-5

14 30 2.49 X l V 7 2.16 x 10-7

6 27

12 29.5 2.52 x lom6 2.30 x 1 V 6

16 31 2.45 x 1W8 2.04 x Q 6 = 3 A, a = 8 X 104 cal AZ/mol. In mole fraction units.

C 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 tmax which 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 f m a x can be obtained from the condition

N~ + e ~ , + N~ = F 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 non- spherical 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 mono- mer concentration is always present.

The explicit expressions used here for the various free- energy terms are of an empirical nature. Of course, it is possible to obtain expressions for the free energy on theo- retical 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

2626 Yohji Achiba and Katsumi Kimura

Conclusions A well-defined transition point in the aggregation pro-

cess emerges as that separating two different types of be- havior of the aggregate size distribution function. A critical concentration corresponding to this transition point is de- fined which is a close lower bound on the cmc values usual- ly 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,

(5) C. A. J. Hoeve and G. C. Benson, J. Phys. Chem., 61, 1149-1 158 (1957). 273-283 (1966).

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 Received June 23, 1975)

Publication costs assisted by 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 com- plexes, 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 ob- tained by analyzing the rise curve of radical ions produced as a result of electron transfer. The first direct, spectro- scopic 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 fol- lowed 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 triplet- state ionic photodissociation of CT complexes.2-6

In the present work, we considered it interesting to de- termine the rate constants of the CT-triplet ionic photodis- sociation using a nanosecond laser flash technique. It may also be interesting to study the effects of solvent and elec- tron donor on the rate constant. The reason that PMDA has been used as an electron acceptor is that its CT com- plexes 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 sep- arated from the T-T absorption bands. Naphthalene and its derivatives (1-methyl-, 2-methyl-, 1-chloro-, 2,3-di-

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 dis- tilled. Acetonitrile was refluxed over phosphorus pentoxide and distilled. Naphthalene and 2-hydroxynaphthalene were recrystallized from ligroin, and 2-methyl- and 2,3- dimethylnaphthalene 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 phosphores- cence spectra on a Hitachi MPF-PA fluorescence spectro- photometer. Laser pulse excitation experiments were car- ried 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-

The Journal of Physical Chemistry, Vol. 79, No. 24, 1975

07/ 11/ 12 KRÜSS: CM C M easur em ent

1/ 2www. 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

Determination of the critical micelle formation concentration

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 consistof 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 willthe surfactant molecules start to form agglomerates inside the liquid; these agglomerates are known as micelles. Thesurfactant 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 aregathered together whereas the hydrophobic tails project inwards. The washing effect of surfactants is based on the fact thathydrophobic 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 aseries of different surfactant concentrations. Surfactants exhibit a specific surface tension curve as a function of theconcentration. Initially the surfactant molecules increasingly enrich themselves at the water surface. During this phase thesurface tension decreases linearly with the logarithm of the surfactant concentration. When the CMC is reached, i.e. whenthe surface is saturated with surfactant molecules, a further increase in surfactant concentration no longer has anyappreciable influence on the surface tension.

This means that in order to determine the CMC the two linear sections formed by the measuring points obtained from theseries of different concentrations must be determined. The CMC is obtained from the intersection of the straight lines for thelinear 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 isgenerated 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.

07/ 11/ 12 KRÜSS: CM C M easur em ent

2/ 2www. 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

Range of conventional CMC method Range of extended CMC method

Reverse CMC measurement

For reverse CMC measurements not the solvent but the parent solution is first put into the sample vessel and then dilutedwith 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 theclouding 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 thesurfactant. With standard CMC measurements the region of interest is also the region where only small amounts of thesample solution are dosed and where therefore the largest error by means of dosing inaccuracy would occur. With reverseCMC measurement low concentrations are reached with large amounts of the solvent and so this error is reduced to aminimum.

Another advantage is that the dosimat only gets in contact with the pure solvent and can run in continuous operation. Thismakes 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 rangeis increased many times over.

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

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

Corresponding Author: Dr. Carlota de Oliveira Rangel Yagui, Av.Prof. Lineu Prestes, 580 – Bloco 16, CEP: 05508-950 – São Paulo, SP.corangel@usp.br

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 pow-erful alternative for dissolving hydrophobic drugs inaqueous environments. In this work, we provide aninsight into this subject. METHODS: A concisereview of surfactants and micelles applications in phar-macy was carried out. RESULTS: Initially, a descrip-tion of surfactants and aqueous micellar systems ispresented. Following, an extensive review on micellardrug solubilization, including both the principlesinvolved on this phenomenon and the work alreadydone regarding solubilization of drugs by micelles ispresented. The application of micelles in drug delivery,in order to minimize drug degradation and loss, to pre-vent harmful side effects, and to increase drug bioavail-ability, is also presented. Special emphasis is given tothe more recent use of polymeric micelles. Finally, webriefly discuss the importance of surfactants andmicelles as biological systems models as well as itsapplication in micellar catalysis. CONCLUSIONS:As can be seen from the review presented, the use ofmicelles in pharmacy is an important tool that findsnumerous applications.

INTRODUCTION

Surfactants are known to play a vital role in many pro-cesses of interest in both fundamental and applied sci-ence. One important property of surfactants is theformation of colloidal-sized clusters in solutions,known as micelles, which have particular significancein pharmacy because of their ability to increase the sol-ubility of sparingly soluble substances in water (1).Micelles are known to have an anisotropic water distri-bution within their structure. In other words, thewater concentration decreases from the surfacetowards the core of the micelle, with a completelyhydrophobic (water-excluded) core. Consequently, thespatial position of a solubilized drug in a micelle will

depend on its polarity: nonpolar molecules will be sol-ubilized in the micellar core, and substances with inter-mediate polarity will be distributed along thesurfactant molecules in certain intermediate positions.

On the other hand, numerous drug delivery and drugtargeting systems have been studied in an attempt tominimize drug degradation and loss, to prevent harm-ful side effects, and to increase drug bioavailability (2-6). Within this context, the utilization of micelles asdrug carriers presents some advantages when comparedto other alternatives such as soluble polymers and lipo-somes. Micellar systems can solubilize poorly solubledrugs and thus increase their bioavailability, they canstay in the body (blood) long enough to provide grad-ual accumulation in the required area, and their sizespermit them to accumulate in areas with leaky vascula-ture (7).

In general, surfactants play an important role in con-temporary pharmaceutical biotechnology, since theyare largely utilized in various drug dosage forms tocontrol wetting, stability, bioavailability, among otherproperties (8). It is important to notice that lyophobiccolloids, such as polymers, require certain energy to beapplied for their formation, are quite unstable from thethermodynamic point of view, and frequently formlarge aggregates. Association colloids such as micelles,on the other hand, can form spontaneously under cer-tain conditions (self-assembling systems), and are ther-modynamically more stable towards both dissociationand aggregation (9).

Therefore, the study of surfactants and their role inpharmacy is of paramount importance, especially withrespect to their ability of solubilizing hydrophobicdrugs. In this work, we provide a review of micellarsolubilization of drugs in surfactant systems, blendingit with basic information on surfactants structure andproperties, as well as the applications for drug delivery.

147

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

Surfactants and Micelles

Surfactants are amphiphilic molecules composed of ahydrophilic or polar moiety known as head and ahydrophobic or nonpolar moiety known as tail. Thesurfactant head can be charged (anionic or cationic),dipolar (zwitterionic), or non-charged (nonionic).Sodium dodecyl sulfate (SDS), dodecyltrimethylammo-nium bromide (DTAB), n-dodecyl tetra (ethyleneoxide) (C12E4) and dioctanoyl phosphatidylcholine (C8-lecithin) are typical examples of anionic, cationic, non-ionic and zwitterionic surfactants, respectively (Figure1). The surfactant tail is usually a long chain hydrocar-bon residue and less often a halogenated or oxygenatedhydrocarbon or siloxane chain (16, 17).

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 asystem, adsorbs onto surfaces or interfaces significantlychanging the surface or interfacial free energy. Surfac-tants usually act to reduce the interfacial free energy,although there are occasions when they are used toincrease it (17). When surfactant molecules are dis-solved in water at concentrations above the criticalmicelle concentration (cmc), they form aggregatesknown as micelles. In a micelle, the hydrophobic tailsflock to the interior in order to minimize their contactwith water, and the hydrophilic heads remain on theouter surface in order to maximize their contact withwater (see Figure 2) (18,19). The micellization processin water results from a delicate balance of intermolecu-lar forces, including hydrophobic, steric, electrostatic,hydrogen bonding, and van der Waals interactions.The main attractive force results from the hydropho-bic effect associated with the nonpolar surfactant tails,and the main opposing repulsive force results fromsteric interactions and electrostatic interactionsbetween the surfactant polar heads. Whether micelliza-

tion occurs and, if so, at what concentration of mono-meric surfactant, depends on the balance of the forcespromoting micellization and those opposing it (19, 20).

Figure 2: Schematic illustration of the reversiblemonomer-micelle thermodynamic equilibrium. The blackcircles represent the surfactant heads (hydrophilicmoieties) and the black curved lines represent thesurfactant tails (hydrophobic moieties).

The determination of a surfactant cmc can be made byuse of several physical properties, such as surface ten-sion (γ), conductivity (κ) – in case of ionic surfactants,osmotic pressure (π), detergency, etc. When theseproperties are plotted as a function of surfactant con-centration (or its logarithm, in case of surface tension),a sharp break can be observed in the curves obtainedevidencing the formation of micelles at that point (16)(Figure 3).

Figure 3: Changes in the physical properties detergency,conductivity (κ), osmotic pressure (π) and surfacetension (γ) of an aqueous solution of surfactant as afunction of surfactant concentration. The break in thecurve of each property corresponds to the CriticalMicelle Concentration (cmc).

148

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

Another important parameter that characterizesmicelles is the aggregation number, Nag, that corre-sponds to the average number of surfactant monomersin each micelle of a micellar solution. Usually, in amicellar solution the aggregation number is approxi-mately constant for a broad total concentration range(up to about 100 times the cmc), with the number ofmicelles varying (21). However, at certain conditionsmicelles can grow with the aggregation number vary-ing with the surfactant concentration. (22).

Micelles are labile entities formed by the noncovalentaggregation of individual surfactant monomers. There-fore, they can be spherical, cylindrical, or planar (discsor bilayers). Micelle shape and size can be controlledby changing the surfactant chemical structure as well asby varying solution conditions such as temperature,overall surfactant concentration, surfactant composi-tion (in the case of mixed surfactant systems), ionicstrength and pH. In particular, depending on the sur-factant type and on the solution conditions, sphericalmicelles can grow one-dimensionally into cylindricalmicelles or two-dimensionally into bilayers or discoi-dal micelles. Micelle growth is controlled primarily bythe surfactant heads, since both one-dimensional andtwo-dimensional growth require bringing the surfac-tant heads closer to each other in order to reduce theavailable area per surfactant molecule at the micellesurface, and hence the curvature of the micelle surface(18, 22).

For all these micellar structures in aqueous media, thesurfactant molecules are oriented with their polarheads towards the water phase and their tail away fromit. In ionic micelles, the interfacial region between themicelle and the aqueous phase contains the ionic headgroups, the Stern Layer of the electrical double layerrelated to these groups, approximately half of thecounter ions associated with the micelle, and water.The remaining counter ions are contained in theGouy-Chapman portion of the double layer thatextends further into the aqueous phase. The length ofthe double layer is a function of the ionic strength ofthe solution and it can be highly compressed in thepresence of electrolytes (23). For the nonionic surfac-tants having a polyethylene oxide (PEO) head group,the structure is essentially the same, except that thecounter ions are not present in the outer region, butrather coils of hydrated polyethylene oxide chains.

The interior of the micelle containing the hydrophobicgroups presents a radius of approximately the length ofthe fully extended hydrophobic chain (17). Anotherimportant characteristic of micelles is that the aqueousphase penetrates into the micelle beyond the hydro-philic head groups, and the first few methylene groupsadjacent to the head are considered in the hydrationsphere. Therefore, we can divide the interior region ofthe micelle in an outer core penetrated by water and inan inner core completely water-excluded (22).

Based on the geometry of various micellar shapes andthe space occupied by the hydrophilic and hydropho-bic groups of the surfactants, it is possible to estimatethe structure of a micelle (20). Accordingly, the param-eter VH/lcao can determine the shape of the micelle,with VH corresponding to the volume of the hydro-phobic group in the micellar core, lc is the length of thehydrophobic group in the core and ao the cross-sec-tional area occupied by the hydrophilic group at themicelle-solution interface. Based on Tanford (19), VH

= 27.4 + 26.9n Å, where n is the number of carbonatoms 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 (Table1).

Table 1: Correlation between the parameter VH/lcao and

the structure of the micelle.

Micellar Solubilization

An important property of micelles that has particularsignificance in pharmacy is their ability to increase thesolubility of sparingly soluble substances in water. Inthis context, solubilization can be defined as the spon-taneous dissolving of a substance by reversible interac-tion with the micelles of a surfactant in water to form athermodynamically stable isotropic solution with

149

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

reduced thermodynamic activity of the solubilizedmaterial (17). If we plot the solubility of a poorly solu-ble compound as a function of the concentration ofsurfactant, as shown in Figure 4, usually what happensis that the solubility is very low until the surfactantconcentration reaches the cmc. At surfactant concen-trations above the cmc the solubility increases linearlywith the concentration of surfactant, indicating thatsolubilization is related to micellization.

Figure 4: Schematic plot of the concentration of a poorlysoluble compound as a function of the surfactantconcentration in aqueous solution.

From the thermodynamic point of view, the solubili-zation can be considered as a normal partitioning ofthe drug between two phases, micelle and aqueous, andthe standard free energy of solubilization (ΔGS

º) can berepresented by the following expression (1):

(1)

where R is the universal constant of the gases, T is theabsolute temperature, and P is the partition coefficientbetween the micelle and the aqueous phase.

Usually, the solubilization of a molecule by a surfac-tant can be evaluated based on two descriptors that arethe molar solubilization capacity, χ, and the micelle-water partition coefficient, P (24). The χ value isdefined as the number of moles of the solute (drug)that can be solubilized by one mol of micellar surfac-tant, and characterizes the ability of the surfactant tosolubilize the drug. It can be calculated based on thegeneral equation for micellar solubilization:

(2)

where Stot is the total drug solubility, SW is the waterdrug solubility, Csurf is the molar concentration of sur-factant in solution, and cmc is the critical micelle con-centration (25). Since above the cmc the surfactantmonomer concentration is approximately equal to thecmc, the term (Csurf – cmc) is approximately equal tothe surfactant concentration in the micellar form and,therefore, χ is equal to the ratio of drug concentrationin the micelles to the surfactant concentration in themicellar form.

On the other hand, the micelle-water partition coeffi-cient is the ratio of drug concentration in the micelleto the drug concentration in water for a particular sur-factant concentration, as follows:

(3)

Combining Equations (2) and (3), we can relate thetwo solubility descriptors. Accordingly, for a givensurfactant concentration:

(4)

As can be seen, P is related to the water solubility ofthe compound, in contrary to χ (25). In order to elimi-nate the dependence of P on the surfactant concentra-tion, a molar micelle-water partition coefficient (PM),corresponding to the partition coefficient when Csurf =1 M, can be defined as follows:

(5)

The lower is the cmc value of a given surfactant, themore stable are the micelles. This is especially impor-tant from the pharmacological point of view, sinceupon dilution with a large volume of the blood, con-sidering intravenous administration, only micelles ofsurfactants with low cmc value still exist, whilemicelles from surfactants with high cmc value may dis-sociate into monomers and their content may precipi-tate in the blood (26).

There are a number of possible loci of solubilizationfor a drug in a micelle, as represented in Figure 5.

150

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

Figure 5: Possible loci of solubilization of drugs insurfactant micelles, depending on the drughydrophobicity. The black bold lines (⎯) represent thedrug at different sites in the micelle. The black circlesrepresent the surfactant heads, the black bold curvedlines represent surfactant heads consisting of PEO, andthe light black curved lines represent the surfactant tails.

Accordingly, hydrophilic drugs can be adsorbed on thesurface of the micelle (1), drugs with intermediate solu-bility should be located in intermediate positionswithin the micelle such as between the hydrophilichead groups of PEO micelles (2) and in the palisadelayer between the hydrophilic groups and the first fewcarbon atoms of the hydrophobic group, that is theouter core (3), and completely insoluble hydrophobicdrugs may be located in the inner core of the micelle(4) (7,17). The existence of different sites of solubiliza-tion in the micelle results from the fact that the physi-cal properties, such as microviscosity, polarity andhydration degree, are not uniform along the micelle(27).

Mukerjee and Cardinal (28) studied the microenviron-ments of benzene, some of its derivatives, Triton X-100, and naphthalene when solubilized in micelles atlow solubilizate to surfactant ratios and proposed theexistence of at least two states (loci) of solubilizationwith different polarity. According to the authors, thetotal uptake by micelles could be divided approxi-mately into an “adsorbed” fraction (location at themicelle-water interface) and a “dissolved” fraction(location in the hydrocarbon core). When adsorptiontakes place the solubility increases beyond the solubil-ity power of the hydrocarbon core. In fact, numerousstudies indicate that the solubility of slightly polar sub-stances and aromatic compounds tend to be consider-ably higher than the solubility of aliphatic compoundspresenting similar molar volumes, despite the fact thatthe later are expected to be more compatible with thealiphatic hydrocarbon core of most micelles (28).

The capacity of surfactants in solubilizing drugsdepends on numerous factors, such as chemical struc-ture of the surfactant, chemical structure of the drug,temperature, pH, ionic strength, etc (7). Nonionic sur-factants usually are better solubilizing agents thanionic surfactants for hydrophobic drugs, because oftheir lower cmc values. For polar drugs it is more com-plicate to establish a general relationship between thedegree of solubilization and the chemical structure ofthe surfactant, since solubilization can be in both theinner and the outer regions of the micelle. Krishna andFlanagan (29) observed that, for the antimalarial drugβ-Arteether (an endoperoxide containing a sesquiter-pene lactone), nonionic surfactants showed muchlower solubilization power than ionic surfactants.They suggested that the solubilization of this drug maynot only involve incorporation into the micellar inte-rior, but may be substantially due to adsorption at themicelle-water interface.

Regarding the influence of structure of the drug, crys-talline solids generally show less solubility in micellesthan do liquids of similar structure (17). For polardrugs, the depth of penetration into the micelle varieswith the structure of the drug. Usually, the less polarthe drug (or the weaker its interaction with either thepolar head of the surfactant in the micelle or the watermolecules at the micelle-water interface) and the longeris the chain length, the smaller its degree of solubiliza-tion, reflecting its deeper penetration into the palisadelayer (17,23).

The extent of solubilization into a particular micelledepends upon the locus of solubilization and thereforethe shape of the micelle. As described previously, theshape of the micelle is determined by the value of theparameter VH/lcao and as this parameter increases themicelle becomes more asymmetrical and the volume ofthe inner core increases relative to that of the outerportion. Therefore, one can expect that the solubiliza-tion of drugs in the core will increase with increase inasymmetry, whereas the solubilization of drugs in theouter region will decrease (17). In fact, it was observedthat for alkyl sodium sulfates and alkyl trimethylam-monium bromides, the solubilization of β-Arteetherincreases with the increase in the alkyl chain length,due to the larger micellar size (29).

151

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

However, Barry and El Eini (30) studying the solubili-zation of non-polar steroidal drugs in aqueous solu-tions of long-chain polyoxyethylene nonionicsurfactants have observed that the molar solubilizingefficiency of surfactants increased as the length of thePEO chain increased while micellar sizes are known todecrease with the increase in PEO chain length. Theauthors suggested that, although the inclusion of non-polar steroids into the micelles decreases as the PEOhydrophilic chain increases, the number of micelles inequimolar amounts of surfactants increases and conse-quently the total amount of steroid per mole of surfac-tant is greater, hence the observed increase insolubilizing efficiency with increased hydrophilicchain length when molar concentrations are consid-ered.

Ong and Manoukian (31) have studied the solubiliza-tion of timobesone acetate, a corticosteroid used ininflammatory therapy, in nonionic surfactants solu-tions and observed that the solubilization capacityincreased with increasing length of the hydrophobictail of the surfactants. Therefore, timobesone wasassumed to be solubilized in the hydrophobic core ofthe micelles. This observation was also confirmed bythe fact that the length of the PEO chain of the surfac-tants studied did not affect the solubilization capacity,for a given tail length, and thus solubilization shouldnot have occurred in the palisade layer or among thePEO heads.

In general, the amount of drug solubilized in a micellarsystem increases with the increase in temperature.Alkhamis et al. (32) studied the solubilization of thedrug gliclazide, a second-generation sulfonylurea usedin the treatment of non-insulin dependent diabetesmellitus. The drug solubility was determined as a func-tion of the concentration of different surfactants at 25and 37oC and, for all the ionic surfactants studied, thesolubilization was higher at 37oC than at 25oC. Thiswas attributed to the increase in thermal agitation,which results in an increase in the space available forsolubilization in the micelle, in addition to the increaseof gliclazide solubility in water at higher temperatures.For the polyoxyethylene nonionic surfactants, theeffect of the temperature on the extent of drug solubili-zation may depend on whether the drug is locatedinside the hydrophobic core or in the palisade layer. Inthis same work, the solubility of gliclazide was found

to decrease with temperature for the nonionic surfac-tants studied. Barry and El Eini (30) also observed a sig-nificant decrease in the micelle/water molar partitioncoefficient, PM, obtained for nonpolar steroidal drugsin PEO surfactants solutions when the temperaturewas increased from 10 to 50oC. The drugs are believedto be located preferentially in the palisade layer, andthe increase in temperature causes dehydration of thePEO groups, bringing them closer and consequentlyreducing the space available for the drugs in this regionof the micelle. Nevertheless, the solubility of drugslocated preferentially in the inner core of PEOmicelles is expected to increase as the temperature israised, due to micellar growth (23).

The ionic strength can influence significantly the solu-bilization of a drug in micellar solutions, especially incase of ionic surfactants. The addition of smallamounts of salts decreases the repulsion between thesimilarly charged ionic surfactant head groups, therebydecreasing the cmc and increasing the aggregation num-ber and volume of the micelles. The increase in aggre-gation number favors the solubilization ofhydrophobic drugs in the inner core of the micelle. Onthe other hand, the decrease in mutual repulsion of theionic head groups causes closer packing of the ionicsurfactant molecules in the palisade layer decreasingthe volume available for solubilization of polar drugs.The addition of salts to solutions of PEO nonionic sur-factants may also increase the extent of solubilizationof hydrophobic drugs because of the increase in aggre-gation number (17).

The pH of micellar solutions can also show significantinfluence on the extent of solubilization of drugs, sinceit may change the equilibrium between ionized andmolecular forms of some drugs. LI et al. (33) studiedthe solubility of the ionized and un-ionized forms offlavopiridol in polysorbate solutions at different pHvalues. This drug is a weakly basic (pKa = 5.68) deriva-tive of rohitukine that has been developed for breastcancer treatment. The authors observed that the high-est total drug solubility was achieved at pH 4.3 wheremost of the drug was ionized. More recently, Li andZhao (34) studied the solubilization of flurbiprofen, anon-steroidal antinflamatory drug used in rheumatoidarthritis, in polysorbate solutions at different pH val-ues. This drug is a weak acid, with a pKa of 4.17. It wasobserved that the drug solubility increases with the

152

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

increase in pH for pH values over the pKa, due to theincrease in the ionized form of the drug. The authorshave also proposed an equilibrium-based model tocharacterize drug-surfactant interactions in pH-con-trolled systems, reflecting both interactions and inter-dependence among all drug-containing species:unionized drug in water, ionized drug in water, union-ized drug in micelles, and ionized drug in micelles. Themodel proposed yielded reasonably good estimationwhen compared to experimental data.

Regarding ionic surfactants, a particular kind of behav-ior can be observed for the solubility of drugs at differ-ent pH values. Enhanced solubility of a drug may beobserved at pH values at which the drug is foundmostly ionized, when surfactant and drug are oppo-sitely charged. This behavior is a consequence of theelectrostatic interactions between the surfactant mole-cules and the charged drug that causes a decrease in therepulsive forces between the head groups of the surfac-tant molecules, contributing to the micellization pro-cess and thus decreasing the cmc value. In fact, an earlystudy has demonstrated that the drug chlorpromazinecan form mixed monomolecular films with phospho-lipids such as L-α-dipalmitoyl phosphatidylethanola-mine and L-α-dipalmitoyl phosphatidyl-choline (35).

More recently, Caetano et al. (36) observed a comicelli-zation phenomenon for the negatively charged surfac-tant SDS and trifluoperazine, an amphiphilic cationicdrug used as antipsychotic and tranquilizer. Theauthors demonstrated, based on SAXS (Small Angle X-ray Scattering) studies, that the presence of the proto-nated drug mediates the effect that the counter ion hason the SDS micelle, in such a way that the drug is ableto promote micellar surface charge screening. More-over, the electrostatic interaction between the posi-tively charged drug and the negatively charged SDSmust cause a decrease in the repulsive forces betweenthe head groups of the surfactant.

One interesting approach is to combine micellar solu-bilization with other properties that may be improvedin a drug solution. In this context, recently Palma et al.(37) combined the solubilization properties of a surfac-tant with the ascorbic acid antioxidant property thatprotects drugs from degradation by light, heat, dis-solved oxygen and other radical producing species, bymeans of synthesizing an ascorbyl-decanoate surfac-

tant. It was observed that micellar solutions of the sur-factant obtained significantly improved the solubilityof hydrophobic drugs with respect to pure water, byincluding these molecules in the hydrophobic micellarcore, as well as protected them from degradation. Itwas also observed that the drug solubilization wasmore effective for the most hydrophobic drugs (Dan-thron and Griseofulvin) than for more hydrophilicones (Phenacetin).

A nonionic surfactant that deserves special attention isCremophor EL (CrEL), which has been used for solu-bilization of a wide variety of hydrophobic drugs suchas anaesthetics, photosensitizers, sedatives, immuno-suppressive agents and anticancer drugs. This heteroge-neous surfactant is a result of the reaction of castor oilwith ethylene oxide, with polyoxyethylene glycerolricinoleate 35 as the major component identified (38).Formulations containing CrEL have been shown topresent important biological side effects, includingsevere anaphylactic hypersensitivity reactions, hyper-lipidaemia, abnormal lipoprotein patterns, aggregationof erythrocytes and peripheral neuropathy (39-44).

One of the most recognized applications of CrEL isthe pharmaceutical formulation of paclitaxel, a hydro-phobic drug active against several murine tumors thathas its development suspended for several years due tosolubilization problems. Various studies have shownthat CrEL influences the pharmacokinetics of manydrugs including paclitaxel that presents a nonlinear dis-position when formulated with this surfactant (38).Recently, Sparreboom et al. (45) proposed that theeffect of CrEL on paclitaxel pharmacokinetics is associ-ated with micellar solubilization, i.e. encapsulation ofthe drug within CrEL micelles, with the micelles act-ing as the principal carrier of paclitaxel in the systemiccirculation.

In paclitaxel I.V. infusions, an exceptionally largeamount of CrEL is necessary, resulting in importantbiological events that can lead to serious acute hyper-sensitivity reactions and neurological toxicity. There-fore, large variety CrEL-free formulation vehicles forpaclitaxel are currently in (pre)clinical development,including liposomes, nanocapsules and microspheres(46).

153

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

Polymeric Micelles and Drug Delivery

Long-circulating pharmaceuticals and drug carriers rep-resent a growing area of medical and pharmaceuticalresearch. There are several reasons for the search forlong-circulating pharmaceuticals and drug carriers,such as:

(i) Long-circulating particles may be used to main-tain a required level of a pharmaceutical agent in theblood for extended time intervals for better drug avail-ability. Moreover, long-circulating diagnostic agentsare of primary importance for blood pool imaging (47).

(ii) Long-circulating particles of nanoscopic size canslowly accumulate in pathological sites with affectedand leaky vasculature (such as tumors, inflammations,and infarcted areas) and improve or enhance drugdelivery in those areas. This phenomenon is usuallycalled enhanced permeability and retention effect,EPR, known also as “passive” targeting or accumula-tion via an impaired filtration mechanism (48,49).

(iii) Prolonged circulation can help to achieve a bettertargeting effect for specific ligand-modified drugs anddrug carriers, since it increases the total quantity of tar-geted drug/carrier passing through the target, and thenumber of interactions between the drug and the tar-get (50).

As stated before, micellar systems present some advan-tages when compared to other drug carriers. For exam-ple, micelles can be obtained in an easy andreproducible manner in large scale and specific ligandscan be attached to their outer surface in order to opti-mize the controlled releasing and specificity of phar-macological effect (7). Polymeric carriers might lead toprecipitation in water, since the drug-polymer interac-tion can result in conversion of functional water-solu-ble groups of the drug into more hydrophobicgroups. Micelles, on the other hand, offer a core/shellstructure and, therefore, stay water-soluble (51).

According to Kabanov et al. (52), the ideal self-assem-bling drug delivery system should spontaneously formfrom drug molecules, carrier components and targetingmoieties; their size should be of around 10 nm in orderto enable them to penetrate various tissues and evencells; they should be stable in vivo for a sufficiently

long period of time without provoke any biologicalreactions; should release the drug upon contact withtarget tissues/cells; and the components of the carrier(surfactant molecules) should be easily removed fromthe 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 gapbetween such drug carriers as individual macromole-cules (antibodies, albumin, and dextran) with sizebelow 5 nm, and particles such as liposomes and micro-capsules with size of 50 nm and up. The most usualsize of a pharmaceutical micelle is between 10 and 80nm and the optimal cmc value should be in a low milli-molar region.

In drug delivery, special attention has been given to theso-called polymeric micelles (5,7,53-57). Polymericmicelles are formed from copolymers consisting ofboth hydrophilic and hydrophobic monomer units,such as PEO and PPO (polypropylene oxide), respec-tively. These amphiphilic block co-polymers with thelength of the hydrophilic block exceeding the length ofthe hydrophobic block can form spherical micelles inaqueous solution. The micellar core consists of thehydrophobic blocks and the shell region consists of thehydrophilic blocks (53). The PEO coating has beenshown to prevent opsonization and subsequent recog-nition by the macrophages of the reticuloendothelialsystem (RES), allowing the micelles to circulate longerand deliver drugs more effectively to the desired sites.(58). Another advantage of polymeric micelles refers tothe ease of sterilization via filtration and safety foradministration (59, 60). Figure 6 presents a schematicrepresentation of the mechanism of polymeric micellesformation.

As aforementioned, micelles are subject to extremedilution upon intravenous injection into humans.However, the slow dissociation of kinetically stablepolymeric micelles allows them to retain their integ-rity and perhaps drug content in blood circulationabove or even below the cmc for some time, creatingan opportunity to reach the target site before decayinginto monomers (51,61). In addition, some polymericmicelles seems to present better solubilization capacitywhen compared to surfactant micelles due to thehigher number of micelles and/or larger cores of theformers (62).

154

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

Figure 6: Formation of polymeric micelles from differenttypes of amphiphilic block co-polymers (Extracted fromTorchillin, 2001).

Besides the solubilization of a drug by physical encap-sulation, polymeric micelles can be loaded with hydro-phobic molecules that are conjugated or complexedwith the polymeric backbone (63). In case of drug con-jugates, there should be a cleavage (hydrolysis) of thecovalent bond between drug and polymer. Therefore,the release may be dependent on the rate of micellardissociation, since water diffusion into the hydropho-bic micellar core must be restricted, resulting in a sus-tained drug release (64).

Most studies and applications that have been con-ducted are based on block copolymers of PEO andPPO blocks, commercially known as Pluronics® (65).Studies for the solubilization of drugs such as haloperi-dol, indomethacin, doxorubicin (DOX), amphotericinB and digoxin have been reported (52, 66-70) and aparenteral formulation of DOX in these polymericmicelles has entered the phase I of clinical trials in Can-ada.

More recently, biodegradable block copolymers withpolyester core-forming structures have been developed.For example, micelles of PEO-poly(D,L-lactic acid-co-caprolactone) (PEO-PDLLA) have been used to encap-sulate paclitaxel and shown similar in vitro toxicity,fivefold increase in maximum tolerable dose andincreased efficacy after intraperitoneal injection inmurine P388 leukemia model when compared to thestandard formulation with Cremophor EL (71).

Polymeric micelles made of poly(ethylene oxide)-b-poly(L-amino acid) (PEO-b-PLAA) has been suggestedas synthetic analogs of natural carriers presenting aunique ability for chemical modification, since the freefunctional groups of PLAA blocks constitute sites toattach drugs. In addition, these PLAA blocks are ofincreasing interest once they may generate biocompati-ble monomers after hydrolysis and/or enzymatic deg-radation (61).

Yokoyama et col. studied PEO-b-poly(L-aspartic acid)-DOX conjugates and, according to the results, thesuperiority of the block copolymer-drug conjugateover the free drug was a result of the lower toxicity ofthe former (72-75). Cisplatin (CIS) has also been com-plexed with PEO-b-poly(Asp), demonstrating increasein cytotoxic concentration against B16 melanoma cellsand lower nefrotoxicity (76). In addition, a PEO-b-poly(α-glutamic acid)-CIS complex was investigatedand presented greater stability, prolonged circulationin blood stream and improved accumulation in tumorsite when compared to the previous complex (77).

Despite the several block copolymer-drug conjugatesstudies, physical encapsulation of drugs within poly-meric micelles offers a great alternative, since conjuga-tion of the drug may lead to changes in the biologicalproperties of the drug and consequently difficult thecharacterization and regulatory approval of the drug.However, physical encapsulation may present lowcapacity and/or rapid release of the encapsulated drug(51).

Other alternative that emerges in the field of poly-meric micelles refers to polyion complex micelles.Oppositely charged macromolecules, such as peptidesand DNA, can complex with the charges of the sidechains of some PLAA blocks resulting in the requiredamphiphilic character for micellization of the complexand leading to stabilization against digestive enzymessuch as nucleases (51). These systems seem to be prom-ising and have been receiving significant attention (78-81).

Recently, polymeric micelles incorporating CIS wereprepared through polymer-metal complex formationbetween CIS and poly(ethylene glycol)-poly(glutamicacid) block copolymers, and showed remarkably pro-longed blood circulation and effective accumulation in

155

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

solid tumors (82). Other polyion complex micellescomposed of a porphyrin dendrimer and PEG-b-poly(aspartic acid) were evaluated as new photosensi-tizers for photodynamic therapy in the Lewis lung car-cinoma cell line and resulted in reduced dark toxicityof the cationic dendrimer porphyrin, probably due tothe biocompatible PEG shell of the micelles (83).

In another work, α-lactosyl-PEG-poly(2(dimethy-lamino)ethyl methacrylate) block copolymer (lactose-PEG-PAMA) was synthesized to construct a polyioncomplex micelle-type gene vector potentially useful forselective transfection of hepatic cells, by spontaneouslycomplexion with plasmid DNA encoding luciferase(pGL3-Luc). The lactose-PEG-PAMA-pDNA micellerevealed enhanced transfection compared to the con-trol polyion complex micelle without the ligand (lac-tose) at a lower pDNA dose (84).

One of the drawbacks of polyion complex micelles isthe sensitivity to environment changes such as dilutionand ionic strength. To overcome these, polymericmicelles prepared from PEG-poly(α,β,aspartic acid)and the cationic protein trypsin were cross-linked withglutaraldehyde through the Schift base formation, con-ferring stability to high salt concentrations and increas-ing the stability of the protein (85).

Micelles, Biological Systems and Micellar Catalysis

The study of cell membranes and of the roles it playsin living cells contributes significantly to the under-standing of cellular function. Membranes have beenshown to consist of lipids in association with proteinsand glycoproteins (16). The present accepted model ofa biomembrane is that the phospholipids are organizedin a bilayer structure, resulting in a fluid lipid matrix ofvarying composition and fluidity. Embedded in thismatrix are the integral proteins that are able toundergo lateral and rotational diffusion. A wide vari-ety of lipids is found in biological membranes, with thephospholipids being among the most common (86).

Many biological processes occur at membrane surfacesor within their hydrophobic moiety. Owing to theionic head groups of the lipids, the surface of biologicalmembranes frequently presents a net charge, givingrise to different binding properties of charged anduncharged forms of molecules such as drugs (87-88). In

this sense, the relationship between the binding prop-erties of a drug and its active form, as well as its mem-brane location, deserves attention. Despite the effortaiming at an understanding of drugs mechanism ofaction at the molecular level, demonstrate by the num-ber of studies on the interaction of drugs with biologi-cal membranes, more studies involving model systemsare necessary (87,89).

Surfactants have a far-ranging use in membrane studies.Because surfactants are amphiphilic molecules, like lip-ids, some of the same rules governing lipid behavioralso apply to the surfactants. Among the membranemodels utilized, micellar systems can be considered aninteresting alternative to study the interactions of dif-ferent compounds with membranes because of the rela-tive simplicity of these systems, and therefore havebeen used with this purpose (13,20).

Reactions behavior observed at surfactant interfacesare expected to be more representative of many biolog-ical reactions than are reactions studied in dilute aque-ous solutions (90). In this sense, micellar catalysis ofreactions is important because of the parallel withenzymes behavior. Catalysis by both normal micellesand reversed micelles is possible. In normal micelles inaqueous medium, enhanced reaction of the solubilizedsubstrate generally occurs at the micelle-water inter-face; in reversed micelles in non-polar medium thisreaction occurs deep in the inner core (17).

Micellar catalysis in aqueous solution is generallyexplained in terms of distribution of reactants betweenwater and micelles, with reactions occurring in bothmedia. Therefore, it is possible to treat the rate-surfac-tant profiles in terms of the concentrations of reactantsin the aqueous and micellar pseudo-phases and the rateconstants in each pseudo-phase (91).

There are different kinetic models to explain micellarcatalytic effects in aqueous medium (92-95). In thepseudo-phase kinetic model (92), the kinetics of a nthorder reaction is analyzed by considering the partition-ing of the reactants between the two pseudo-phases.

The reactants (A and B) may be distributed as shownin Figure 7.

156

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

Figure 7: Schematic representation of a micellarcatalyzed reaction according to the pseudo-phase kineticmodel. AW, BW and AM, BM correspond to theconcentrations of the reactants in the aqueous (W) andmicellar (M) phases; KA and KB are the biding constantsof the reactants to the micelles, and kW and kM are therate constants in the aqueous and micellar paths.

Therefore, a quantitative rate expression for a bimolec-ular 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 reactantsA and B, respectively; C is the total surfactant molarconcentration minus the cmc; V is the partial molarvolume of the surfactant in the micelle and, therefore,CV and (1 – CV) stand for the volume fractions of themicellar and aqueous phases, respectively. The bindingconstants (KA and KB) are related to the partition coeffi-cient (P), as follows:

(7)

For dilute surfactant solutions, where the volume frac-tion 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)

In this model, no distinction is made between the vari-ous regions of the micelles, although reactions gener-ally occur in the Stern layer, at the micelle/waterinterface, rather than in the hydrocarbon-like core ofthe micelle. Nevertheless, the pseudo-phase modelexplains many features of micellar rate effects and itcan be applied, at least qualitatively, to a variety ofreactions in colloidal assemblies (96).

Since the binding constant K depends on the extent ofhydrophobic bonding between surfactant and sub-strate, it can be expected that K will increase withincrease in the chain length of both the surfactant andthe substrate. However, if the hydrophobic group ofthe substrate is too long, it may be solubilized sodeeply in the micelle that access to its reactive site by areagent in aqueous solution phase is hindered and,therefore, solubilization will inhibit the reaction (17).

The charge of surfactant head group also influences thecatalytic power of micelles. Thus, catalysis of somenucleophilic aromatic substitution reactions is morepronounced by dicationic micellar surfactants than bycationic micellar hexadecyltrimethylammonium bro-mide (96). Yu et al. (97) observed that cationic micellesinhibit, anionic micelles accelerate and nonionicmicelles show no appreciable effect on metal ionhydrolysis of p-nitrophenyl picolinate. The higher theelectron charge of metal ions, the greater these effectsare, indicating that electrostatic interactions are themajor contribution for this reaction in micellar solu-tion. In another work, it was observed an increase inthe oxidation rate of L(+)arabinose by chromic acidwith the addition of SDS and Triton X-100 concentra-tions. On the other hand, the addition of ammonium,lithium and sodium bromides in SDS micelles resultedin rate decrease (98).

Plots of rate constant versus surfactant concentrationoften show a maximum at some surfactant concentra-tion above the CMC. One of the reasons for this is thatthe number of micelles increases with increase in thesurfactant concentration. When the number ofmicelles exceeds that required to solubilize all of thesubstrate there is a dilution of the substrate concentra-tion per micelle with further increase in surfactant con-centration, leading to a reduction in the rate constant.Moreover, the charge surface of an ionic micelle inaqueous solution may cause not only the concentra-

157

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

tion of an oppositely charged reactant at the micelle-solution interface, but the adsorption of that reactanton it or even the solubilization into micelles, resultingin a decrease in the reactant activity in the solutionphase. Therefore, an increase in surfactant concentra-tion over that required to complete solubilization ofthe substrate may result in a decrease in the rate con-stant (17).

There is strong evidence to believe that most micelle-catalyzed reactions occur on the surface of the ionicmicelles, at or near the charged double layer that sur-rounds the hydrocarbon core. Typically, reactionsbetween very hydrophobic substrates and hydrophilicanions seem to have lower second-order rate constantsin the micellar pseudo phase than in water, because theanions are located in the Stern layer at the micelle/water interface whereas the substrate may be, on aver-age, more deeply in the micelle (96).

Micelles also allow co-solubilization of compounds ofvery different hydrophobic and hydrophilic characterand, as a result, chemical reactions can be developedwhich otherwise would proceed only with difficulty.An example is the formation of o-phthaldehydeadducts from water-insoluble amines of high molecularmass (99). The presence of micelles can also result inthe formation of different reaction products. A diazo-nium salt in an aqueous micellar solution of sodiumdodecyl sulfate, for example, yielded the correspond-ing phenol from reaction with OH- in the bulk phase,but the corresponding hydrocarbon from material sol-ubilized in the micelles (100).

One interesting approach refers to functional surfac-tants, which are surfactants containing a reactive resi-due, usually at the head group, that can be micellizedor co-micellized with a chemically inert, non-func-tional surfactant. In this sense, micelles functionalizedwith groups that model the amino acid side chainsresponsible for enzyme activity are generally impres-sive catalysts (96).

Final Considerations

Considering the importance of micellar systems in thepharmaceutical field and the many applications that itpresents, our group have been carrying research on thismatter, with special attention to the solubilization of

drugs in aqueous micellar solutions. We study the solu-bilization of model drugs, such as the non-steroidalanti-inflammatory ibuprofen, as well as of potentialdrugs such as p-substituted benzhydrazides compoundsin solutions of different surfactants.

Recently, we investigated the solubilization of ibupro-fen (IBU) in micellar solutions of three surfactants pos-sessing the same hydrocarbon tail but differenthydrophilic head groups, namely sodium dodecyl sul-phate (SDS), dodecyltrimethylammonium bromide(DTAB), and n-dodecyl octa(ethylene oxide) (C12E8)(101). The results obtained showed that, irrespective ofthe surfactant type, the solubility of IBU increases lin-early with increasing surfactant concentration, becauseof the association between the drug and the micelles.Nonionic surfactants were shown to provide a combi-nation of good molar solubilization capacity and highmicellar concentration, due to their low cmc, resultingin increased solubility of IBU. On should keep in mindthat the low toxicity of nonionic surfactants makesthem particularly interesting for solubilization anddrug delivery purposes. In addition to these studies,Small Angle X-ray Scattering (SAXS) studies on theinteraction of ibuprofen with micelles of SDS, DTABand C12E8 are in progress, aiming at a deeper under-standing on the nature of these interactions as well ason the properties of the aggregates obtained.

The p-substituted benzhydrazides are hydrophobiccompounds synthesized by Taveres et col. that repre-sent potential drugs with anti-staphylococci and anti-trypanosome activities. Quantitative Structure-Activ-ity Relationships (QSAR) studies and experimentalresults have shown a dependency between biologicalactivity and hydrophobicity of these compounds, withthe more hydrophobic molecules presenting higheractivity (102,103). However, the more hydrophobicthe compound, the more difficult the solubilization inthe culture media used for activity determination.Therefore, the possibility of micellar solubilization ofthese molecules should contribute to more precisedeterminations of biological activity. We are currentlycarrying on experiments on the solubilization of p-sub-stituted benzhydrazides in aqueous micellar solutionsof n-dodecyl octaethylene oxide (C12E8) and n-hexade-cyl octaethylene oxide (C16E8), two nonionic surfac-tants possessing the same hydrophilic head groups butdifferent hydrophobic tails.

158

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

ACKNOWLEDGEMENTS

Carlota Rangel-Yagui is grateful for the Post-Doctoralfellowship and financial support from FAPESP(Fundação de Amparo à Pesquisa no Estado de SãoPaulo). We acknowledge CAPES (Coordenação deAperfeiçoamento de Pessoal de Nível Superior –Brazil)and CNPq (Conselho Nacional de DesenvolvimentoCientífico e Tecnológico – Brazil) for financial sup-port.

REFERENCES

[1] Mall, S., Buckton, G., Rawlins, D.A. Dissolutionbehaviour of sulphonamides into sodium dodecyl sul-phate micelles: A thermodynamic approach. J PharmSci, 85(1):75-78, 1996.

[2] Allen, T.M., Hansen, C.B., Menenez, D.E.L. Phar-macokinetics of long-circulating liposomes. Adv.Drug Deliv Rev, 16:267–284, 1995.

[3] Canto, G.S., Dalmora, S.L., Oliveira, A.G. Piroxi-cam encapsulated in liposomes: characterization andin vivo evaluation of topical anti-inflammatoryeffect. Drug Dev Ind Pharm, 25:1235-1239, 1999.

[4] Gref, R., Minamitake, Y., Peracchia, M.T., Tru-betskoy, V.S., Torchilin, V.P., Langer, R. Biodegrad-able long-circulating polymeric nanospheres. Science,263:1600–1603, 1994.

[5] Jones, M.C., Leroux, J.C. Polymeric micelles — anew generation of colloidal drug carriers. Eur J PharBio Pharm, 48:101–111, 1999.

[6] Torchilin, V.P., Trubetskoy, V.S., Whiteman, K.R.,Caliceti, P., Ferruti, P., Veronese, F.M. New syn-thetic amphiphilic polymers for steric protection ofliposomes in vivo. J Pharm Sci, 84:1049-1053, 1995.

[7] Torchilin, V.P. Structure and design of polymericsurfactant-base drug delivery systems. J Control Rel,73:137-172, 2001.

[8] Martin, A. Physical Pharmacy, 4ed., Williams andWilkins, Baltimore, USA, pp 396–398, 1993.

[9] Hunter, R.J. Introduction to Modern Colloid Science.Oxford University Press, Oxford, 1993.

[10] Ford, J.M., Hait, W.N. Pharmacology of drugs thatalter multidrug resistance in cancer. Pharmacol Rev,42(3):155-199, 1990.

[11] Chakraborty, H., Banerjee, R., Sarkar, M. Incorpora-tion of NSAIDS in micelles: implication of structuralswitchover in drug-membrane interaction. BiophysChem, 104(1):315-325, 2003.

[12] Ferreira, H., Lucio, M., De Castro, B., Gameiro, P.,Lima, J.L.F.C., Reis, S. Partition and location ofnimesulide in EPC liposomes: a spectrophotometric

and fluorescence study. Anal Bioanal Chem,377(2):293-298, 2003.

[13] Fresta, M., Guccione, S., Beccari, A.R., Furneri,P.M., Puglisi, G. Combining molecular modelingwith experimental methodologies: Mechanism ofmembrane permeation and accumulation of ofloxa-cin. Bioorg Med Chem, 10(12):3871-3889, 2002.

[14] Israelachvili, J.N., Mitchell, D.J., Ninham, B.W. The-ory of self-assembly of hydrocarbon amphiphiles intomicelles and bilayers. J Chem Soc Far Trans II,72:1525-1568, 1976.

[15] Suwalsky, M., Hernandez, P., Villena, F., Aguilar, F.,Sotomayor, C.P. Interaction of the anticancer drugtamoxifen with the human erythrocyte membraneand molecular models. Zeitschrift fur NaturforschungC - J Biosciences, 53(3-4):182-190, 1998.

[16] Jones, M.N., Chapman, D. Micelles, monolayers andbiomembranes. Wiley-Liss, New York, 1995.

[17] Rosen, M.J. Surfactants and interfacial phenomena,2ed., John Wiley & Sons, New York, 1989.

[18] Chevalier, Y., Zemb, T. The structure of micelles andmicroemulsions. Rep Prog Phys, 53:279-371, 1990.

[19] Tanford, C. The hydrophobic effect: Formation ofmicelles and biological membranes. Wiley, New York,1980.

[20] Israelachvili, J.N. Intermolecular and surface forces,2ed., Academic Press, London, 1991.

[21] Hiemenz, P.C., Rajagopalan, R. Principles of colloidand surface chemistry, 3ed. Marcel Decker, NewYork, 1997.

[22] Puvvada, S., Blankschtein, D. Molecular-thermody-namic approach to predict micellization, phasebehavior, and phase separation of micellar solutions.I. Application to nonionic surfactants. J Chem Phys,92(6):3710-3724. 1990.

[23] Florence, A.T., Attwood, D. Physicochemical Princi-ples of Pharmacy, 3rd ed. The MacMillan Press, Lon-don, 2003.

[24] Attwood, D., Florence, A.T. Surfactant systems: Theirchemistry, pharmacy and biology. Chapman and Hall,New York, 1983.

[25] Alvarez-Núñez, F.A., Yalkowsky, S.H. Relationshipbetween polysorbate 80 solubilization descriptorsand octanol-water partition coefficient of drugs. Int JPharm, 200:217-222, 2000.

[26] Yokoyama, M. Block copolymers as drug carriers.CRC Crit Rev Ther Drug Carrier Syst, 9:213-248,1992.

[27] Dutt, G.B. Rotational diffusion of hydrophobicprobes in Brij-35 micelles: Effect of temperature on

159

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

micellar internal environment. J Phys Chem B,107:10546-10551, 2003.

[28] Mukerjee, P., Cardinal, J.R. Benzene derivatives andnaphthalene solubilized in micelles. Polarity ofmicroenvironment, location and distribution inmicelles, and correlation with surface activity inhydrocarbon-water systems. J Phys Chem,82(14):1620-1627, 1978.

[29] Krishna, A.K., Flanagan, D.R. Micellar solubiliza-tion of a new antimalarial drug, β-arteether. J PharmSci, 78(7):574-576,1989.

[30] Barry, B.W., El Eini, D.I.D. Solubilization of hydro-cortisone, dexamethasone, testosterone and progest-erone by long-chain polyoxyethylene surfactants. JPharm Pharmac, 28:210-218, 1976.

[31] Ong, J.T.H., Manoukian, E. Micellar solubilizationof timbesone acetate in aqueous and aqueous propy-lene glycol solutions of nonionic surfactants. PharmRes, 3(11):704-708, 1988.

[32] Alkhamis, K.A., Allaboun, H., Al-Momani, W.Y.Study of the solubilization of gliclazide by aqueousmicellar solutions. J Pharm Sci, 92(4):839-846, 2003.

[33] Li, P., Tabib, E., Yalkowsky, S.H. Solubilization ofionized and un-ionized flavopiridol by ethanol andpolysorbate 20. J Pharm Sci, 88(5):507-509, 1999.

[34] Li, P., Zhao, L. Solubilization of flurbiprofen in pH-surfactant solutions. J Pharm Sci, 92(5):951-956, 2003.

[35] Zografi, G., Auslander, D.E. Surface activity of chlo-rpromazine and chlorpromazine sulfoxide in thepresence of insoluble monomolecular films. J PharmSci, 54(9):1313-1318, 1965.

[36] Caetano, W., Gelamo, E.L., Tabak, M., Itri R. Chlor-promazine and sodium dodecyl sulphate mixedmicelles investigated by Small Angle X-Ray Scatter-ing. J Colloid Interf Sci, 248:149-157, 2002.

[37] Palma, S., Manzo, R.H., Allemandi, D., Fratoni, L.,Lo Nostro, P. Drug solubilization in ascorbyl-decanoate micellar solutions. Colloid Surf A, 212:163-173, 2003.

[38] Gelderblom, H., Verweij, J., Nooter, K., Spar-reboom, A. Cremophor EL: the drawbacks andadvantages of vehicle selection for drug formulation.Eur J Cancer, 37:1590-1598, 2001.

[39] Adams, J.D., Flora, K.P., Goldspiel, B.R., Wilson,J.W., Arbuck, S.G.I. Taxol: a history of pharmaceuti-cal development and current pharmaceutical con-cerns. J Natl Cancer Inst Monogr, 15:141-147, 1993.

[40] Watkins, J., Ward, A.M., Appleyard, T.N. Adversereactions in intravenous anaesthetic induction agents.Br Med J, 2:1084-1085, 1997.

[41] Huttel, M.S., Olesen, A.S., Stoffersen, E. Comple-ment-mediated reactions to diazepam with cremo-phor as solvent (steroid MR). Br J Anaesth 52(1):77-79, 1980.

[42] Kongshaug, M., Cheng, L.S., Moan, J., Rimington,C. Interaction of Cremophor-EL with humanplasma. Int J Biochem, 23(4):473-478, 1991.

[43] Woodburn, K., Kessel, D. The alteration of plasma-lipoproteins by Cremophor EL. J Photoch PhotobiolB, 22(3):197-201, 1994.

[44] Windebank, A.J., Blexrud, M.D., Degroen, P.C.Potential neurotoxicity of the solvent vehicle forcyclosporine. J Pharmacol Exp Ther, 268(2):1051-1056,1994.

[45] Sparreboom, A., van Zuylen, L., Brouwer, E., Loos,W.J., de Bruijn, P., Gelderblom, B., Pillay, M.,Nooter, K., Stoter, G., Verweij, J. Cremophor EL-mediated alteration of paclitaxel distribution inhuman blood: Clinical pharmacokinetic implica-tions. Cancer Res, 59(7):1454-1457, 1999.

[46] van Zuylen, L., Verweij, J., Sparreboom, A. Role offormulation vehicles in taxane pharmacology. InvNew Drug, 19(2):125-141, 2001.

[47] Torchilin, V.P., Trubetskoy, V.S. Which polymerscan make nanoparticulate drug carriers long-circulat-ing? Adv Drug Deliv Rev, 16:141–155, 1995.

[48] Palmer, T.N., Caride, V.J., Caldecourt, M.A., Twick-ler, J., Abdullah, V. The mechanism of liposomeaccumulation in infarction. Biochim BiophysActa,797:363–368, 1984.

[49] Maeda, H., Wu, J., Sawa, T., Matsumura, Y., Hori,K. Tumor vascular permeability and the EPR effectin macromolecular therapeutics: a review. J ContrRel, 65:271–284, 2000.

[50] Torchilin, V.P. How do polymers prolong circula-tion time of liposomes? J Liposome Res, 6:99–116,1996.

[51] Lavasanifar, A., Samuel, J., Kwon, G.S. Poly(ethyl-ene oxide)-block-poly(L-amino acid) micelles for drugdelivery. Adv Drug Deliv Rev, 54:169-190, 2002.

[52] 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(oxy-ethylene)-poly(oxypropylene) block copolymers asmicrocontainers for drug targeting from blood tobrain. J Contr Rel, 22:141–158, 1992.

[53] Kwon, G.S., Kataoka, K. Block copolymer micellesas long-circulating carriers for delivery of therapeuticand diagnostic agents. Adv Drug Deliv Rev, 16:295-309, 1995.

160

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

[54] Inoue, T., Chen, G.H., Nakamae, K., Hoffman, A.S.An AB block copolymer of oligo(methyl methacry-late) and poly(acrylic acid) for micellar delivery ofhydrophobic drugs. J Control Rel 51(2-3):221-229,1998.

[55] Discher, B.M., Won, Y.Y., Ege, D.S., Lee, J.C.M.,Bates, F.S., Discher, D.E., Hammer, D.A. Polymer-somes: Tough vesicles made from diblock copoly-mers. Science, 284(5417):1143-1146, 1999.

[56] Rapoport, N., Marin, A., Luo, Y., Prestwich, G.D.,Muniruzzaman, M. Intracellular uptake and traffick-ing of pluronic micelles in drug-sensitive and MDRcells: Effect on the intracellular drug localization. JPharm Sci, 1:157-170, 2002.

[57] Soo, P.L., Luo, L., Maysinger, D., Eisenberg, A.Incorporation and release of hydrophobic probes inbiocompatible polycaprolactone-block-poly(ethyleneoxide) micelles: implications for drug delivery. Lang-muir, 18:9996-10004, 2002.

[58] Lee, J.H., Lee, H.B., Andrade, J.D. Blood compatibil-ity of polyethylene oxide surfaces. Prog Polym Sci,20(6):1043-1079, 1995.

[59] Kwon, G.S., Kataoka, K. Block copolymer micellesas long circulating drug vehicles. Adv Drug Deliv Rev,16:295–309, 1995.

[60] Kwon, G.S., Okano, T. Polymeric micelles as newdrug carriers. Adv Drug Deliv Rev, 21:107–116, 1996.

[61] Adams, M.L., Lavasanifar, A., Kwon, G.S.Amphiphilic block copolymers for drug delivery. JPharm Sci, 92(7):1343-1355, 2003.

[62] Gadelle, F., Koros, W.J., Schechter, R.S. Solubiliza-tion of aromatic solutes in block copolymers. Macro-molecules, 28:4883–4892, 1995.

[63] Kataoka, K., Harada, A., Nagasaki, Y. Block copoly-mer micelles for drug delivery: design,characteriza-tion and biological significance. Adv Drug Deliv Rev,47:113–131,2001.

[64] Li, Y., Kwon, G.S. Methotrexate esters of poly(ethyl-ene oxide)-block-poly(2-hydroxyethyl- -asparta-mide). Part I: effects of the level of methotrexateconjugation on the stability of micelles and on drugrelease. Pharm Res, 17:607–611, 2000.

[65] Mortensen, K. PEO-related block copolymer surfac-tants. Coll Surf A, 183-185:277-292, 2001.

[66] 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 neurolepticactivity of haloperidol increases after its solubiliza-tion in surfactant micelles. FEBS Lett, 258:343–345,1989.

[67] Lin, S., Kawashima, Y. Kinetic studies on the stabil-ity of indomethacin in alkaline aqueous solutionscontaining poly(oxyethylene)poly(oxypropylene)surface active block copolymers. Pharm Acta Helv,60:345–350, 1985.

[68] Baratkova, E.V., Dorodnich, T.Y., Klinskii, E.Y.,Kliushnenkova, E.N., Shemchukova, O.B., Gon-charova, O.N., Arjakov, V.Y., Alakhov, V.,Kabanov, A.V. Anthracycline antibiotics nonco-valently incorporated into the block copolymermicelles: in vivo evaluation of anticancer activity. Br JCancer, 74:1545–1552, 1996.

[69] Forster, D., Washington, C., Davis, S.S. Toxicity ofsolubilized and colloidal amphotericin B formula-tions to human erythrocytes. J Pharm Pharmacol,40:325–328, 1988.

[70] Baratkova, E.V., Miller, D.W., Li, S., Alakhov, V.,Kabanov, A.V., Elmquist, W.F. Pluronic P85enhances the delivery of digoxin to the brain: in vitroand in vivo studies. J Pharmacol Exp Ther, 269:551–557, 2001.

[71] 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 bio-distribution of polymeric micellar paclitaxel. CancerChemother Pharmacol, 40:81–86, 1997.

[72] Yokoyama, M., Kwon, G.S., Okano, T., Sakurai, Y.,Sato, T., Kataoka, K. Preparation of micelle-formingpolymer–drug conjugates. Bioconjug Chem, 3:295–301, 1992.

[73] Kwon, G.S., Yokoyama, M., Okano, T., Sakurai, Y.,Kataoka, K. Biodistribution of micelle-forming poly-mer–drug conjugates. Pharm Res, 10:970–974, 1993.

[74] Kataoka, K., Kwon, G.S., Yokoyama, M., Okano, T.,Sakurai, Y. Polymeric micelles as novel drug carriersand virus mimicking vehicles. In: J. Kahovec (Ed),Macromolecules, 1992(VSP):267–276, 1993.

[75] Yokoyama, M., Okano, T., Sakurai, Y., Ekimoto,H., Shibazaki, C., Kataoka, K. Toxicity and antitu-mor activity against solid tumors of micelle-formingpolymeric anticancer drug and its extremely long cir-culation in blood. Cancer Res, 51:3229–3236, 1991.

[76] Mizumura, Y., Matsumura, Y., Hamaguchi, T., Nish-iyama, N., Kataoka, K., Kawaguchi, T., Saito, T.,Kakizoe, T. Cisplatin-incorporated polymericmicelles reducing nephrotoxicity, while maintainingantitumor activity. Jpn J Cancer Res, 92:328–336,2001.

[77] Nishiyama, N., Kato, Y., Sugiyama, Y., Kataoka, K.Development of cisplatin-loaded polymeric micelleswith a prolonged circulation in the bloodstream andan enhanced accumulation in the solid tumor. Pro-

161

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

ceedings of the 20th International Symposium on Con-trolled Release of Bioactive Compounds, 28:5155–5156,2001.

[78] Kabanov, A.V., Kabanov, V.A. Interpolyelectrolyteand block ionomer complexes for gene delivery:physico-chemical aspects. Adv Drug Deliv Rev,30:49–60, 1998.

[79] 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. Blockand graft copolymers and nanogel copolymer net-works for DNA delivery into cell. J Drug Target, 8:91–105, 2000.

[80] Itaka, K., Yamauchi, K., Harada, A., Nakamura, K.,Kawaguchi, H., Kataoka, K. Polyion complexmicelles from plasmid DNA and poly(ethylene gly-col)-poly(L-lysine) block copolymer as serum tolera-ble polyplex system: physicochemical properties ofmicelles relevant to gene transfection efficiency. Bio-materials, 24(24):4495-4506, 2003.

[81] Berret, J.F., Vigolo, B., Eng, R., Herve, P., Grillo, I.,Yang, L. Electrostatic self-assembly of oppositelycharged copolymers and surfactants: A light, neu-tron, and X-ray scattering study, Macromolecules,37(13):4922-4930, 2004.

[82] Nishiyama, N., Okazaki, S., Cabral, H., Myiamoto,M., Kato, Y., Sugiyama, Y., Nishio, K., Matsumura,Y., Kataoka, K. Novel cisplatin-incorporated poly-meric micelles can eradicate solid tumours in mice.Cancer Res, 63(24):8977-8983, 2003.

[83] Zhang, G.D., Harada, A., Nishiyama, N., Jiang,D.L., Koyama, H., Aida, T., Kataoka, K. Polyioncomplex micelles entrapping cationic dendrimer por-phyrin: effective photosensitizer for photodynamictherapy of cancer. J Control Rel, 93(2):141-150, 2003.

[84] Wakebayashi, D., Nishiyama, N., Yamasaki, K.,Itaka, K., Kanayama, N., Harada, A., Nagasaki, Y.,Kataoka, K. Lactose-conjugated polyion complexmicelles incorporating plasmid DNA as a targetablegene vector system: their preparation and gene trans-fecting efficiency against cultured HepG2 cells. J Con-trol Rel, 95(3):653-664, 2004.

[85] Jaturanpinyo, M., Harada, A., Yuan, X.F., Kataoka,K. Preparation of bionanoreactor based on core-shellstructured polyion complex micelles entrappingtrypsin in the core cross-linked with glutaraldehyde.Bioconjugate Chem, 15(2):344-348, 2004.

[86] Yeagle, P.L. The membranes of cells, 2nd ed. NewYork: Academic Press, 1993.

[87] Caetano, W., Tabak, M. Interaction of chlorprom-azine and trifluoperazine with ionic micelles: elec-

tronic absorption spectroscopy studies. Spectrochim.Acta A. 55(12):2513-2528, 1999.

[88] Louro, S.R.W., Nascimento, O.R., Tabak, M.Charge-dependent and ph-dependent binding-sites fordibucaine in ionic micelles - a fluorescence study. Bio-chim Biophys Acta, 1190:319-328, 1994.

[89] Yushmanov, V.E., Perussi, J.R., Imasato, A.C., Rugi-ero, M. Tabak, M. Ionization and binding equilibriaof papaverine in ionic micelles studied by H-1-NMRand optical-absorption spectroscopy Biophys Chem,52:157-163, 1994.

[90] Price, S.E., Jappar, D., Lorenzo, P., Saavedra, J.E.,Hrabie, J.A., Davies, K.M. Micellar catalysis of nitricoxide dissociation from diazeniumdiolates. Lang-muir, 19:2096-2102, 2003.

[91] Arikan, B., Tunçay, M. Micellar effects and reactantincorporation in reduction of toluidine blue by ascor-bic acid. Dyes and Pigments. 64:1-8, 2005.

[92] Martinek, K., Yatsimirskii, A.K. Levashov, A.V.,Berezin, I.V. In: Micellization, Solubilization andMicroemulsions, Ed: Mittal, K.L., Plenum, NewYork, vol.2, p.489, 1977.

[93] Menger, F.M., Portnoy, C.E. On chemistry of reac-tions proceeding inside molecular aggregates. J AmChem Soc, 89(18): 4698-&, 1967.

[94] Piszkiewicz, D. Cooperativity in bimolecularmicelle-catalyzed reactions - inhibition of catalysis byhigh-concentrations of detergent. J Am Chem Soc,99(23):7695-7697, 1977.

[95] Dash, A.C., Prusti, J., Pradhan, J., Das, P.K. Micellareffects upon the reactions of complex-ions in solution.4. Kinetics of aquation and base hydrolysis of somecis-(chloro)(amine)bis (ethylenediamine)cobalt(III)complexes in the presence of neutral and anionic sur-factants in an aqueous-medium. J Chem Soc Far Trans,86(3):507-510, 1990.

[96] Carreto, M.L., Rubio, S., Pérez-Bendito, D. Organicmicroheterogeneous systems in kinetic analysis. Selfassembled systems – A review. Analyst, 121:33R-44R,1996.

[97] Yu, X.Q., Jiang, B.Y., Cheng, S.Q., Huang, Z., Zeng,X.C. Comparative reactivities of metal cation-cata-lyzed hydrolysis of p-nitrophenyl picolinate in micel-lar solutions. J Disp Sci Technol, 24(6):761-765, 2003.

[98] Kabir-ud-Din, Morshed, A., Khan, Z. Oxidative deg-radation of L(+)arabinose by chromium (VI) inabsence and presence of sodium dodecyl sulphate andTX-100 micelles. Oxid Commun, 26(1):59-71, 2003.

[99] Memon, M.H., Worsfold, P.J. The use of microemul-sions in flow-injection analysis - spectrofluorimetric

162

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

determination of primary amines. Anal Chim Acta,183:179-185, 1986.

[100] Abe, M. Suzuki, N., Ogino, K. Replacement reactionof diazonium salt in the presence of sodium dodecyl-sulfate micelles. J Coll Interface Sci, 93:285-288, 1983.

[101] 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 sol-ubilization. RBCF – Braz J Pharm Sci, 2004. (In Press)

[102] Furlanetto, M., Masunari, A., Tavares, L. C. Valida-tion of carbonyl group absorption frequency , Vc=o,as a structural descriptor for QSAR/QSPR studiesIn: 15th European Symposium on Quantitative Struc-ture-Activity Relationships & Molecular Modeling,2004, Istanbul. Program & Abstracts , p.142, 2004.

[103] Masunari, A., Rezende, P., Tavares, L. C. QSARstudies of nifuroxazide analogs with antimicrobialactivity against multidrug-resistant Staphylococcusaureus In: 15th European Symposium on QuantitativeStructure-Activity Relationships & Molecular Modeling,2004, Istanbul. Program & Abstract, p.123, 2004.

163

07/11/12 Lyotropic l iquid crystal - Wikipedia, the free encyclopedia

1/3en.wikipedia.org/wiki/Lyotropic_liquid_crystal

A highly viscous cubic phase gel

made of polysorbate 80, water, and

liquid paraffin.

Lyotropic liquid crystalFrom Wikipedia, the free encyclopedia

A liquid crystalline material is called lyotropic if phases having long-rangedorientational order are induced by the addition of a solvent. Historically theterm was used to describe materials composed of amphiphilic molecules. Suchmolecules comprise a water-loving 'hydrophilic' head-group (which may beionic or non-ionic) attached to a water-hating 'hydrophobic' group. Typicalhydrophobic 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

Amphiphilic molecules form aggregates through a self-assembly process that isdriven by the 'hydrophobic effect' when they are mixed with a solvent, which isusually water. The aggregates formed by amphiphilic molecules arecharacterised 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 criticalconcentration (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 withmonomeric amphiphiles in solution, and with no long ranged orientational or positional (translational) order. Thesedispersions 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 ofamphiphile in water is increased beyond the point where the micellar aggregates are forced to be disposed regularly inspace. For amphiphiles that consist of a single hydrocarbon chain the concentration at which the first liquid crystallinephases 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 amphiphileconcentrations the micelles fuse to form cylindrical aggregates of indefinite length, and these cylinders are arranged on along-ranged hexagonal lattice. This lyotropic liquid crystalline phase is known as the 'hexagonal phase', or morespecifically 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 lipidsin cell membranes. For most amphiphiles that consist of a single hydrocarbon chain, one or more phases having complexarchitectures are formed at concentrations that are intermediate between those required to form a hexagonal phase andthose that lead to the formation of a lamellar phase. Often this intermediate phase is a bicontinuous cubic phase.

07/11/12 Lyotropic l iquid crystal - Wikipedia, the free encyclopedia

2/3en.wikipedia.org/wiki/Lyotropic_liquid_crystal

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 theformation 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 thathave at least two hyrocarbon chains attached to a headgroup. The most abundant phospholipids that are found in cellmembranes 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 0-12-

437760-2.

Fennell Evans D. and Wennerström H. (1999). The Colloidal Domain. New York: Wiley VCH. ISBN 0-471-24247-0.

Retrieved from "http://en.wikipedia.org/w/index.php?title=Lyotropic_liquid_crystal&oldid=513648868"

Categories: Chemical properties Phases of matter Liquid crystals

07/11/12 Lyotropic l iquid crystal - Wikipedia, the free encyclopedia

3/3en.wikipedia.org/wiki/Lyotropic_liquid_crystal

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. SeeTerms of Use for details.

Wikipedia® is a registered trademark of the Wikimedia Foundation, Inc., a non-profit organization.

MTLNNLNO qÜÉ= ÜÉêÉòçî =i ~Ä=J=i ` m=oÉëçì êÅÉë

NLOÅÜÉêÉòçî KëÅêáééëKÉÇì LêÉëçì êÅÉëKÜíã

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 emulsif ier

Biosensors

Separation of biomolecules

Chemical synthesis

Protein crystallization

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.

Figure 2. Cartoon pictures of three bicontinuous LCPs with 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 f irst 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 native-

like 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 membrane

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 multi lamellar l ipid

portal, which feeds the growing crystal

with the protein molecules.

Introduction

home research LCP resources publications members links contact

INTRO

LCP Equations

LCP Lipids

LCP Protocols

LCP Tools

Crystal Gallery

Structures from LCP

LCP Publications

LCP Links

MTLNNLNO qÜÉ= ÜÉêÉòçî =i ~Ä=J=i ` m=oÉëçì êÅÉë

OLOÅÜÉêÉòçî KëÅêáééëKÉÇì LêÉëçì êÅÉëKÜíã

<< back to top next >>

CONTACT US: The Scripps Research Institute | Cherezov Lab | 10550 North Torrey Pines Road | GAC1200 | La Jolla | CA 92037 | vcherezo@scripps.edu

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 formnanosized aggregates when the individual polymer chains (“unimers”) are directly dissolved inaqueous solution (dissolution method)above a threshold concentration (critical micelle concentration orCMC) and solution temperature (critical micelle temperature or CMT) (Fig. 1). Amphiphilic polymerswith very low water solubility can alternatively be dissolved in a volatile organic solvent, then dialyzedagainst 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 individualunimers are designed to be biodegradable and/or have a low enough molecular mass (< ~40 kDa) to beeliminated by renal clearance, in order to avoid polymer buildup within the body that can potentiallylead to toxicity. The most developed amphiphilic block copolymers assemble into spherical coreshellmicelles approximately 10 to 80 nm in diameter, consisting of a hydrophobic core for drug loading anda 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 commonhydrophilic block used to form the hydrophilic shell is the FDAapproved excipient poly(ethyleneglycol) (PEG) or poly(ethylene oxide) (PEO). PEG or PEO consists of the same repeating monomersubunit CH2CH2O, and may have different terminal end groups, depending on the synthesisprocedure, 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 extremelysoluble and has a large excluded volume. This makes it especially suitable for physically interferingwith intramicelle interactions and subsequent micelle aggregation. PEG also blocks protein and cellsurface 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 stericprotection by the hydrophilicshell is a function of both the density and length of the hydrophilic PEGblocks.

Fig. 2. Polymer micelle structures.

Unlike the hydrophilic block, which is typically PEG or PEO, different types of hydrophobic blockshave been sufficiently developed as hydrophobic drug loading cores. Examples of diblock copolymersinclude (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 (inPluronics/poloxamers). The choice of hydrophobic block is largely dictated by drug compatibility withthe hydrophobic core (when drug is physically loaded, as described later) and the kinetic stability of themicelle. The selfassembly of amphiphilic copolymers is a thermodynamic and, consequently, areversible process that is entropically driven by the release of ordered water from hydrophobic blocks; itis either stabilized or destabilized by solvent interactions with the hydrophilic shell. As such, thestructural potential of amphiphilic copolymer unimers to form micelles is determined by the mass ratioof hydrophilic to hydrophobic blocks, which also affects the subsequent morphology if aggregates areformed. If the mass of the hydrophilic block is too great, the copolymers exist in aqueous solution asunimers, whereas, if the mass of the hydrophobic block is too great, unimer aggregates withnonmicellar morphology are formed. If the mass of the hydrophilic block is similar or slightly greaterthan 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 andsubsequent extreme dilution in the vascular compartment. This is because the polymer micelles must bestable enough to avoid burst release of the drug cargo, as in the case of a physically loaded drug, uponsystemic administration and remain as nanoparticles long enough to accumulate in sufficientconcentrations at the target site. The relative thermodynamic stability of polymer micelles (which isinversely related to the CMC) is primarily controled by the length of the hydrophobic block.Anincrease in the length of the hydrophobic block alone significantly decreases the CMC of the unimerconstruct (i.e. increases the thermodynamic stability of the polymer micelle), whereas an increase in thehydrophilic block alone slightly increases the CMC (i.e. decrease the thermodynamic stability of apolymer micelle).Although the CMC indicates the unimer concentration below which polymer micelleswill begin to disassemble, the kinetic stability determines the rate at which polymer micelle disassemblyoccurs. Many diblock copolymer micelles possess good kinetic stability and only slowly dissociate intounimers after extreme dilution. Thus, although polymer micelles are diluted well below typical unimerCMCs (10~6107M) after intravenous injection, their relative kinetic stability might still be suitable fordrug delivery. The kinetic stability depends on several factors, including the size of a hydrophobicblock, 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 polymermolecules with covalently linked amphiphile chains. For example, copolymers with starlike or dendriticarchitecture, depending on their structure and composition, can either aggregate into multimolecularmicelles,or exist as unimolecular micelles. Dendrimers are widely used as building blocks to prepareunimolecular micelles, because they are highlybranched, have welldefined globular shape andcontroled surface functionality. For example, unimolecular micelles were prepared by couplingdendritic hypercores of different generations with PEO chains. The dendritic cores can entrap variousdrug molecules. However, due to the structural limitations involved in the synthesis of dendrimers ofhigher generation, and relatively compact structure of the dendrimers, the loading capacity of suchmicelles is limited. Thus, to increase the loading capacity, the dendrimer core can be modified withhydrophobic block, followed by the attachment of the PEO chains. For example, Wang et al. recentlysynthesized an amphiphilic 16arm star polymer with a polyamidoamine dendrimer core and armscomposed of inner lipophilic poly(ecaprolactone) block and outer PEO block. These unimolecularmicelles were shown to encapsulate a hydrophobic drug, etoposide, with high loading capacity.Multiarm starlike block copolymers represent another type of unimolecular micelles. Star polymers aregenerally synthesized by either the armfirst or corefirst methods. In the armfirst method,monofunctional living linear macromolecules are synthesized and then crosslinked either throughpropagation, using a bifunctional comonomer, or by adding a multifunctional terminating agent toconnect precise number of arms to one center. Conversely, in the corefirst method, polymer chains aregrown 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 fattyacids as a lipophilic inner block and PEO as a hydrophilic outer block. These polymers were directlydispersible in aqueous solutions and formed unimolecular micelles. The size and solubilizing capacity ofthe 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 thepolymer chains. These resulting crosslinked micelles are, in essence, single molecules of nanoscale sizethat are stabile upon dilution, shear forces and environmental variations (e.g. changes in pH, ionicstrength, solvents etc.). There are several reports on the stabilization of the polymer micelles bycrosslinking either within the core domain or throughout the shell layer. In these cases, the crosslinkedmicelles maintained small size and coreshell morphology, while their dissociation was permanentlysuppressed. Stable nanospheres from the PEObpolylactide micelles were prepared by usingpolymerizable group at the core segment. In addition to stabilization, the core polymerized micellesreadily solubilized rather large molecules such as paclitaxel, and retained high loading capacity evenupon dilution. Formation of interpenetrating network of a temperaturesensitive polymer(polyNisopropylacrilomide) inside the core was also employed for the stabilization of the Pluronicmicelles. The resulting micelle structures were stable against dilution, exhibited temperatureresponsiveswelling behavior, and showed higher drug loading capacity than regular Pluronic micelles. Recently, anovel type of polymer micelles with crosslinked ionic cores was prepared by using block ionomercomplexes as templates. The nanofabrication of these micelles involved condensation ofPEObpoly(sodium methacrylate) diblock copolymers by divalent metal cations into spherical micellesof coreshell morphology. The core of the micelle was further chemically crosslinked and cationsremoved by dialysis. Resulting micelles represent hydrophilic nanospheres of coreshell morphology.The core comprises a network of the crosslinked polyanions and can encapsulate oppositely chargedtherapeutic and diagnostic agents, while a hydrophilic PEO shell provides for increased solubility.Furthermore, these micelles displayed the pH and ionic strengthresponsive hydrogellike behavior, dueto the effect of the crosslinked ionic core. Such behavior is instrumental for the design of drug carrierswith controled loading and release characteristics.

Comments»

No comments yet — be the first.

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

www.aspbs.com/enn

Encyclopedia ofNanoscience andNanotechnology

Preparation of Vesicles (Liposomes)

Peter WaldeETH Zürich, Zürich, Switzerland

CONTENTS

1. What are Vesicles (Liposomes)?2. Vesicles and the Liquid Crystalline States

of Surfactants3. Methods for Preparing Normal Vesicles4. Preparation of Reversed Vesicles5. Characterizations and Applications

of Vesicles6. Concluding Remarks

GlossaryReferences

1. WHAT ARE VESICLES (LIPOSOMES)?

1.1. Introduction

Vesicles—more precisely “normal vesicles”—are a particulartype of polymolecular aggregate (polymolecular assembly)of certain amphipathic molecules, formed in aqueous solu-tion. A vesicle is composed of one or more closed shells(which are usually 4–5 nm thick) that entrap a small volumeof the aqueous solution in which the vesicle is formed. Vesi-cles are often spherical (under osmotically balanced condi-tions) and can have diameters between about 20 nm andmore than 0.1 mm. If an analogous type of aggregate isformed in a water-immiscible, apolar solvent, the aggregateis called a reversed vesicle (see Section 4). In the following,the term “vesicle” or “lipid vesicle” always stands for a nor-mal type of vesicle and not for a reversed vesicle.

For the sake of simplicity, a so-called unilamellar vesicleis first considered. It is a closed lamella with an aqueousinterior. The lamella is composed of amphipathic molecules,compounds that comprise at least two opposing parts, ahydrophilic part (which is soluble in water) and a hydropho-bic part (which is not soluble in water but is soluble in anorganic solvent that is not miscible with water, in this con-text also called “oil”). Amphipathic molecules have a “sym-pathy” as well as an “antipathy” for water. Because of themixed affinities within the same chemical structure, amphi-pathic molecules are also called amphiphiles (meaning “both

loving,” water as well as oil). They are surfactants, whichstands for “surface active agents” and means that they accu-mulate at the surface of liquids or solids. The accumulationof surfactant molecules on the surface of water (at the water-air interface) leads to a reduction in the surface tension ofwater, as a result of an alteration of the hydrogen bondsbetween the interfacial water molecules.

The aqueous solution in which vesicles form is presentoutside of the vesicles as well as inside. Figure 1 is aschematic representation of a unilamellar spherical vesicleformed by the amphiphile POPC (1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine) in water at room temperature.The vesicle drawn in Figure 1 is assumed to have an outerdiameter of 100 nm. The bilayer has a thickness of onlyabout 4 nm, which corresponds in a first approximation toabout two times the length of an extended POPC molecule.

Lipid vesicles in the size range of 0.1 �m �=100 nm� canbe visualized by electron microscopy, for example, by thefreeze-fracture technique [1, 2], or by cryofixation [2, 3] (seeFig. 2A and B). For vesicles in the micrometer range, lightmicroscopy can be applied [4] (see Fig. 2C).

Based on simple geometric considerations, one can calcu-late the approximate number of lipid molecules present ina particular defined vesicle. In the case of the unilamellarvesicle shown in Figure 1 (outer diameter 100 nm), about8�1 × 104 POPC molecules form the shell of one vesicle,and all of these molecules are held together by noncova-lent bonds. The single lamella of the giant vesicle shown inFigure 2C (diameter ∼ 60 �m) contains about 2�6 × 1010

POPC molecules. The shell is a molecular bilayer with anarrangement of the POPC molecules in such a way thatthe hydrophobic (“water-hating”) acyl chains are in the inte-rior of the bilayer and the hydrophilic (“water-loving”) polarhead groups are on the two outer sites of the bilayer, indirect contact with either the trapped water inside the vesicleor with the bulk water in which the vesicle is dispersed. Sincethe bilayer shell in a sphere is necessarily curved, the num-ber of amphiphiles constituting the inner layer is expectedto be smaller than the number of amphiphiles present in theouter layer. For the 100-nm vesicle of Figure 1, the calcu-lated number of POPC molecules is 3�74× 104 in the innerlayer and 4�36 × 104 in the outer layer, assuming a meanhead group area of one POPC molecule of 0.72 nm2 [5]

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

Encyclopedia of Nanoscience and NanotechnologyEdited by H. S. Nalwa

Volume 9: Pages (43–79)

44 Preparation of Vesicles (Liposomes)

1

2

3

4

5

aqueousexterior

aqueousinterior

100 nm

~ 4 nm

O

O

O P OO

O

HN

+

CH3CH3

CH3

O

O

1

2

3910 1

cis

hydrophobic hydrophilic

POPC

1

2

3

4

5

aqueousexterior

aqueousinterior

100 nm

~ 4 nm

O

O

O P OO

O

HN

+

CH3CH3

CH3

O

O

1

2

3910 1

cis

hydrophobic hydrophilic

POPC

Figure 1. Schematic representation of the cross section through a par-ticular unilamellar, spherical vesicle that has an assumed outer diam-eter of 100 nm and is formed in water from the surfactant POPC(1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine). POPC constitutingthe single, closed lamellar shell of the vesicle is represented as a filledcircle to which two tails are connected. The tails stand for the twohydrophobic (water-insoluble) chains of POPC, and the filled circlesymbolizes 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 isa naturally occurring glycerophospholipid, is also shown. The glycerolmoiety with its stereospecifically numbered carbon atoms is localizedinside 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 diffu-sion about the axis perpendicular to the surface of the bilayer; (4) lat-eral diffusion within the bilayer plane; (5) vertical vibrations, out of thebilayer plane; and (not shown) collective undulations of the membrane.See text for nomenclature details and [392] for a detailed descriptionof the vesicle membrane dynamics.s

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

Figure 2. Electron and light microscopic visualization of vesicles pre-pared from POPC in water at ∼ 25 �C. (A and B) Transmission electronmicrographs of a suspension containing LUVs (FAT-VET100), preparedby the extrusion technique (see Section 3.8 and Fig. 3). The sampleswere analyzed by the freeze-fracture technique (A) and by cryo-fixation(B). The length of the bars corresponds to 100 nm. (C) Light micro-graph (differential interference contrast mode) of a single GUV pre-pared by the electroformation method (see Section 3.3). Length ofthe 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 micrographwas taken by R. Wick at the Department of Materials Science at theETH Zürich.

the chains are flexible enough to accommodate these asym-metric constraints [7].

The mean head group area of a surfactant, abbreviateda0, corresponds to the area that is occupied on average bythe polar head group packed within the vesicle’s bilayer asa consequence of the actual molecular size and the interac-tion between the neighboring surfactant molecules. A simplegeometric model that, in addition to a0, takes into accountthe overall geometric shape of the surfactant as a criticalpacking parameter (or “shape factor”), p �=v/�a0lc)), hasbeen developed [7–10] and successfully applied in a usefulsimple theory toward an understanding of molecular self-assembling systems at large [11]; v is the volume of thehydrophobic portion of the molecule and lc is the critical

Preparation of Vesicles (Liposomes) 45

length of the hydrophobic tails, effectively the maximumextent to which the chains can be stretched out. Accordingto this simple theory, lipid bilayers and vesicles can be pre-pared if p for a particular amphiphile has a value between0.5 and 1.0 (conditions under which the surfactants will packinto 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 presentin biological membranes, glycerophospholipids, lipids thatcontain a glycerol-3-phosphate unit. Actually, the geometricstructure of vesicles as spherulites—which is the term origi-nally used [12]—containing one (or more) concentric bilayershell(s) was elaborated for the first time with vesicle prepa-rations made from (mixtures of) naturally occurring glyc-erophospholipids [12, 13]. This is why it has been proposedto call these aggregates lipid vesicles [14] or liposomes (actu-ally meaning “fat bodies”) [15, 16]. Since the type of aggre-gate shown in Figures 1 and 2 has not so much to do witha “fat body,” it is in principle more appropriate to use theterm lipid vesicle or just vesicle instead of liposome [17]. Inany case, all of the terms liposome, lipid vesicle, and vesicleare used here for the same type of polymolecular aggregate.Sometimes, however, one also finds the term synthetic vesi-cles [18, 19], referring to vesicles formed by synthetic, oftencharged, nonnatural surfactants. Others use the term vesi-cle exclusively for a closed unilamellar (not multilamellar)aggregate of amphiphiles [20, 21], such as the one shownschematically in Figure 1. Other names appearing in the lit-erature are niosomes (vesicles prepared from nonionic sur-factants) [22, 23], polymer vesicles or polymersomes (vesiclesprepared from polymeric surfactants) [24–26], and so on.Table 1 summarizes different terms to describe a particulartype of vesicle. Whenever one uses one of these terms, oneshould specify how it is used and how it is actually defined,to avoid any possible confusion or misunderstanding.

1.3. Nomenclature and ChemicalStructures of Vesicle-FormingGlycerophospholipids

Although the basic principles of vesicle formation are forall types of vesicle-forming amphiphiles in the end thesame—independently of whether they are charged, neutral,or polymeric—the descriptions in Sections 2 and 3 focuson just one particular group of phospholipids, the so-calledphosphatidylcholines (PCs). POPC (see Fig. 1) is a partic-ular PC, namely 1-palmitoyl-2-oleoyl-phosphatidylcholine,more precisely, 1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-choline (also called 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphorylcholine). Since the nomenclature of glycerophos-pholipids (such as POPC) is often not so well known tothose not working with this particular type of biologicalmolecule, a short overview of the main nomenclature ruleis given next. The nomenclature of lipids outlined andused here has been proposed by the International Unionof Pure and Applied Chemistry–International Union ofBiochemistry (IUPAC-IUB) [27, 28].

POPC is a chiral phospholipid with one chiral centerat that carbon atom that is localized in the middle ofthe glycerol moiety; POPC belongs to the glycerophospho-lipids, the quantitatively most important structural groupwithin the class of phospholipids. Glycerophospholipids havea glycerol backbone to which a phosphate group is boundthrough a phosphoric acid ester linkage to one of the glyc-erol hydroxyl groups. To designate the configuration of thisglycerol derivative, the carbon atoms of the glycerol moi-ety are numbered stereospecifically (indicated in the chemi-cal name as prefix -sn-). If the glycerol backbone is writtenin a Fischer projection (see, for example, [29]) in such away that the three carbon atoms are arranged vertically andthe hydroxyl group connected to the central carbon atom ispointing to the left, then the carbon atom on top is C-1,the carbon atom in the middle is C-2 (this is the actualchiral center common to all glycerophospholipids), and thecarbon atom at the bottom is C-3. With this convention,the chemical structure of 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine is defined. It is the naturally occurring formof the two possible enantiomers. The mirror image of POPCis 2-oleoyl-3-palmitoyl-sn-glycero-1-phosphocholine. In theCahn-Ingold-Prelog nomenclature system (also called theR/S convention) [29], the configuration of C-2 in POPCis R. (Another shorthand description of POPC sometimesused is PamOleGroPCho or PamOlePtdCho [27, 28]; Pamstands for palmitoyl, Ole for oleoyl, Gro for glycerol, P forphosphate, Cho for choline, and Ptd for phosphatidyl).

All naturally occurring diacylglycerophospholipids thathave a zwitterionic phosphocholine head group are alsocalled 3-sn-phosphatidylcholines, or just phosphatidyl-cholines. In the biochemical and biophysical literature, phos-phatidylcholine is also called lecithin (or l-�-lecithin becauseof its stereochemical relationship to the naturally occurringl-�-glycerol-phosphate). Therefore, POPC is a lecithin. Thename lecithin implies that egg yolk (lekithos in Greek) con-tains large amounts of phosphatidylcholines. According tothe IUPAC-IUB, the use of the term lecithin is permitted butnot recommended [27]. It is indeed better to avoid using thisterm since it may have another meaning in the food tech-nology literature. The International Lecithin and Phospho-lipid Society (ILPS) of the American Oil Chemists Society(AOCS) defines lecithin as a mixture of lipids obtained fromanimal, vegetable, or microbial sources, which includes PCsbut also contains a variety of other substances, such as sphin-gosylphospholipids, triglycerides, fatty acids, and glycolipids[30].

Egg yolk is one of the cheapest commercial sources forthe isolation of phosphatidylcholines; the other one is soy-beans. Phosphatidylcholines from egg yolk contain a num-ber of chemically different phosphatidylcholines. All of thesePCs have, however, the same glycerol backbone and thesame polar head group (phosphocholine). They only differin the acyl chains esterified with the glycerol hydroxyl groupsat C-1 and C-2 (see Table 2) [31].

Many studies on the preparation and characterization oflipid vesicles have been carried out with egg yolk PC. Onaverage, in position C-1 in egg yolk PC is often palmitic acid,and in position C-2, oleic acid [31]. POPC is therefore arepresentative PC molecule for egg yolk PCs. In contrast toegg 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 Meaning and use of the term in the literature

Algosome Vesicle prepared on the basis of 1-O-alkylglycerol [485].Archaesome Vesicles prepared from archaebacterial, bolaamphiphilic lipids [359, 486].Bilosome Vesicle prepared from a particular mixture of non-ionic surfactants (1-monopalmitoyl-glycerol), cholesterol, dihexade-

cylphosphate (5:4:1 molar ratio), and bile salt (particularly deoxycholate) [487].Catanionic vesicle Vesicle prepared from a mixture of a cationic and an anionic surfactant [59, 488].Cerasome Vesicle with a silicate framework on its surface [366, 489].Ethosome Vesicle that contains in the final preparation a considerable amount of ethanol (prepared by a particular method described

in Section 3.20) [229–231].Fluorosome SUV containing a fluorescent dye embedded in its bilayer to monitor the entry of molecules into the bilayer [490–492].Hemosome Hemoglobin-containing vesicle [493].Immunoliposome 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].Lipid vesicle Vesicle prepared from amphiphilic lipids [31, 69].Liposome Vesicle prepared from amphiphilic lipids [31, 69].Magnetoliposome Vesicle containing magnetic nanoparticles (e.g., magnetite Fe3O4) [495–498].Marinosome Vesicle based on a natural marine lipid extract composed of phospholipids (PCs and phosphatidylethanolamines) contain-

ing a high amount (∼65%) polyunsaturated aycl chains [499].Niosome 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].Novasome 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].Phospholipid vesicle Vesicle prepared from (amphiphilic) phospholipids [71].PLARosome Phospholipid-alkylresocinol liposome: phospholipid vesicle containing resorcinolic lipids or their derivatives [504].Polymer vesicle Vesicle prepared from polymeric amphiphiles, such as di- or triblock copolymers [24, 26].Polymerized vesicle Vesicle prepared from polymerizable amphiphiles that were (partially) polymerized after vesicle formation [33, 352, 353,

505].Polymersome Vesicle prepared from polymeric amphiphiles, such as di- or triblock copolymers [24, 26].Proliposomes 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].Proniosomes A dry, granular product containing mainly (but not exclusively) non-ionic surfactants which, upon the addition of water,

disperses to form MLVs [507].Reversed vesicle Inverted vesicle formed in a water-immiscible, apolar solvent in the presence of a small amount of water (Section 4) [508].Spherulite Onion-like vesicle prepared with spherulite technology, which involves the use of shear forces (Section 3.11).Sphingosome Vesicle prepared on the basis of sphingolipids present in human skin [69, 509].Stealth liposome 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 immunesystem (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].

Synthetic vesicle Vesicle prepared from synthetic surfactants (surfactant mixtures) that are not present in biological membranes. Thesurfactants usually have a single hydrophobic tail [17, 488].

Toposome Vesicle that has a surface that is site-selectively (toposelectively) modified in a stable manner at specific and deliberatelocations (e.g., through chemical modifications or chemical functionalizations) [513].

Transfersome Ultradeformable ethanol-containing mixed lipid/detergent vesicle claimed to transfer water-soluble molecules acrosshuman skin (Section 3.28) [275, 276].

Ufasome Vesicle prepared from unsaturated fatty acid/soap mixtures [330].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 ofan eukaryotic cell [97].

Virosome Vesicle containing viral proteins and viral membranes, reconstituted from viral envelopes, the shells that surround thevirus [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 suitedthan the egg yolk PC mixture. For applications, however, thecheaper egg yolk PCs may be advantageous.

Although lipid vesicles prepared from egg yolk PCsare similar in many respects to vesicles prepared from

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

Preparation of Vesicles (Liposomes) 47

Table 2. Main approximate fatty acid content in egg yolk and soybean PCs (see [31]).

Relative abundance in phospholipids

Egg yolk Soybeans

at at at atTotal sn-1 sn-2 Total sn-1 sn-2

Fatty acid (trivial name) Abbreviation (%) (%) (%) (%) (%) (%)

Hexadecanoic acid C 16:0 ∼35 ∼69 ∼2 ∼17 ∼34(palmitic acid)

Octadecanoic acid C 18:0 ∼14 ∼26 ∼1 ∼8(stearic acid)

Cis-9-octadecenoic C 18:1c9 ∼27 ∼5 ∼49 ∼23 ∼30 ∼16acid (oleic acid)

Cis, cis-9,12-octadecadienoic acid 18:2c9c12 ∼48 ∼24 ∼71(linoleic acid)

Cis, cis-6,9- 18:2c6c9 ∼6 ∼11octadecadienoic acid

All cis-9,12,15- 18:3c9c12c15 ∼9 ∼4 ∼13octadecatrienoic acid

All cis-5,8,11,14,17- 20:5c5c8c11c14c17 ∼4 ∼7eicosapentaenoic acid

All cis-4,7,10,13,16,19- 22:6c4c7c10c13c16c19 ∼13 ∼25docosahexaenoic acid

Note: The abbreviation 16:0, for example, indicates that the linear fatty acid has 16 carbon atoms without any doublebonds; 18:1c9 indicates that the linear fatty acid is composed of 18 carbon atoms with one cis double bond in position9,10 (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 ofthe molecules constituting the vesicle shell(s) as reportedin Table 1, but also according to their size, lamellarityand morphology, and method of preparation (see Table 3).Small unilamellar vesicles (SUVs) have one lamella anddiameters of less than about 50 nm. So-called large unil-amellar vesicles (LUVs) have one lamella and diametersbetween about 50 nm and about 500 nm (see Fig. 1 andFig. 2A and B). Giant vesicles (GVs) can be observed bylight microscopy and have diameters of more than about0.5–1 �m (Fig. 2C). Oligolamellar vesicles (OLVs) have afew and multilamellar vesicles (MLVs) have many concen-trically arranged lamellae. Multivesicular vesicles (MVVs)contain nonconcentrically arranged vesicles within a largervesicle.

As described in detail in Section 3, the preparation ofvesicle suspensions generally involves the use of a particulartechnique, a particular preparation method. Depending onthe technique applied, the vesicle suspensions are charac-terized by a certain degree of homogeneity, a certain meansize and mean lamellarity, and a certain trapped volume.The trapped volume is the aqueous volume that is encapsu-lated by the lipid vesicles, expressed as microliters of aque-ous solution per micromole of lipid (=liters/mol). A trappedvolume of 1 �l/�mol means that in a vesicle suspension con-taining 1 �mol lipid, for example, only 1 �l of the aqueoussolution is trapped by the vesicles. The spherical unilamel-lar POPC vesicle shown in Figure 1 has a trapped volumeof about 3 �l/�mol, as calculated based on simple geomet-ric considerations. This means that in a vesicle suspension

prepared from 10 mM POPC (=7�6 g/liter), only 30 �l outof 1 ml is trapped by the vesicles (3 volume %). The totallipid-water interfacial area in this vesicle suspension is 4�3×103 m2!

Since, in many cases, the vesicle suspensions preparedby one particular technique are not further characterizedwith respect to mean size, size distribution, and lamellarity,the vesicles are just named according to the method used.Examples include REVs (reversed-phase evaporation vesi-cles, vesicles prepared by the so-called reversed-phase evap-oration method), VETs (vesicles prepared by the so-calledextrusion technique), etc.; see some entries in Tables 1and 3.

2. VESICLES AND THE LIQUIDCRYSTALLINE STATESOF SURFACTANTS

2.1. Introduction

Since in most cases, lipid vesicles can be considered as dis-persions of a liquid crystalline state of a vesicle-formingsurfactant, it is useful to give a short introduction to someof 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 thermody-namic equilibrium states of most glycerophospholipids underthe conditions in which vesicle formation is observed. For arecent 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 Characteristics

DRV Dehydrated-rehydrated vesicle (ordried-reconstituted vesicle)

Vesicle prepared by the dehydration-rehydration method (Section 3.7) [31]

EIV Ethanol-injected vesicle Vesicle prepared by the ethanol injection method (Section 3.18)FATMLV Vesicle prepared by repeatedly

freezing and thawing a MLV sus-pension

Equilibration and homogenization procedure (Section 3.6) [133]

v FPV Vesicle prepared with a Frenchpress

Unilamellar vesicle or OLV prepared with a French press for vesicle size homogeniza-tion (Section 3.8) [72]

GUV Giant unilamellar vesicle Unilamellar vesicle with a diameter larger than about 500 nmGV Giant vesicle Vesicle with a diameter larger than about 500 nmIFV Interdigitation-fusion vesicle Vesicle prepared by the interdigitation-fusion method (Section 3.21) [232, 517]LUV Large unilamellar vesicle Unilamellar vesicle with a diameter between about 50 nm and 500 nmLUVET Large unilamellar Vesicle pre-

pared by the extrusion techniqueVesicles prepared by the extrusion technique are usually large and mainly unilamellar

(Section 3.8)MLV Multilamellar vesicle The vesicle contains several concentrically arranged lamellae osmotically stressed after

formation because of an exclusion of solute molecules during their formation [76]MVL Multivesicular liposome A large vesicle that contains internal, nonconcentrically arranged vesicular compart-

ments; also called MVV [211, 212]MVV Multivesicular vesicle A large vesicle that contains internal, nonconcentrically arranged vesicular compart-

ments; also called MVL [210]OLV Oligolamellar vesicle The vesicle contains a few concentrically arranged lamellaeREV Reversed-phase evaporation

vesicleVesicle prepared by the reversed-phase evaporation technique (Section 3.14)

RSE vesicle Rapid solvent exchange vesicle Vesicle prepared by the rapid solvent exchange method [238]SPLV Stable plurilamellar vesicle Similar to MLV but not osmotically stressed after its formation [76]SUV Small (or sonicated) unilamellar

vesicleUnilamellar vesicle with a diameter of less than about 50 nm, as typically obtained by

sonicating MLVs (Section 3.5)ULV Unilamellar vesicle Vesicle with only one lamellaUV Unilamellar vesicle Vesicle with only one lamellaVET Vesicle prepared by the extrusion

techniqueVesicles 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 asubstance is a state between a pure crystal (characterized bya high order of rigid molecules) and a pure liquid (charac-terized by rapid molecular motions of disordered molecules)[32]. There are dozens of different liquid crystalline states,all characterized by a different degree of molecular mobilityand order. Liquid crystals can be produced either by heat-ing a particular crystalline solid—called a “thermotropic liq-uid crystal”—or by dissolution of particular substances in asolvent—called a “lyotropic liquid crystal.” Many surfactantsin water form lyotropic liquid crystalline phases, such as thelamellar phase (L· �), the so-called gel phase (L· � or L· ′�), thenormal or reversed hexagonal phase (HI or HII), or one ormore of the known cubic phases (Ia3d, Pn3m, Im3m). Thetype of phase formed depends on the chemical structureof the surfactant used and on the experimental conditions(such as concentration and temperature, or the presence ofother compounds).

In the L· �-phase (also called the liquid-analogue [33] orliquid-disordered state [34, 35]), the surfactant moleculesare arranged in bilayers, frequently extending over large

distances (1 �m or more) [32]. The hydrophobic chains arerather disordered, with a lot of gauche conformations inthe saturated hydrocarbon parts of the hydrophobic chains,making the bilayers fluid, characterized by fast lateral androtational diffusions of the surfactant molecules, similar toa liquid. Comparable molecular motions are also presentin the liquid-disordered state of vesicles. In the case ofSUVs prepared from POPC, for example, the lateral diffu-sion coefficient seems to be on the order of 3–4×10−8 cm2/s,as determined between 5 �C and 35 �C [36]. The rotationalcorrelation time is on the order of 10−9 to 2 × 10−8 s [37].

The L· � (or L· ′�-)-phase of surfactant molecules (also calledsolid-analogue [33] or solid-ordered state [34, 35]) closelyresembles the L· �-phase in the sense that the surfactantmolecules are also arranged in bilayers. The viscosity isvery high, however. This is a consequence of the rigid-ity of the individual surfactant molecules which are mostlypresent with all-trans conformations in the saturated hydro-carbon parts of the hydrophobic chains. The motion of themolecules is rather restricted, similar to the molecules ina crystal. To specify the relative arrangement of the lipidmolecules, the gel phase may be abbreviated as L· ′� or L· �,

Preparation of Vesicles (Liposomes) 49

depending on whether the alkyl chains are tilted (P· ′�) or nottitled (L· �) with respect to the normal of the lipid bilayer.If it is not known whether the chains are tilted, L· � is oftenused 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, eitherDPPC or POPC, the P· ′�-phase is formed at thermodynamicequilibrium at temperatures at least 5–10� below a lipid spe-cific temperature called the main phase transition temper-ature (or lamellar chain melting temperature) �Tm�. Tm isalso called the lamellar gel-to-liquid crystalline phase transi-tion temperature and can be determined, for example, as theendothermic peak maximum in heating scans of differentialscanning calorimetry (DSC) measurements [41–43]. AboveTm the lipids are in the L· �-phase.

Between the L· �- (or P· ′�-) phase and the L· �-phase, anintermediate gel phase, called the ripple phase (abbreviatedas P· ′�) is often observed at high water content in the caseof PCs [44]. This particular lipid phase takes its name fromthe fact that in freeze-fracture electron micrographs, a “rip-ple” structure can be seen if the lipid dispersion is rapidlyfrozen from the particular temperature interval in which theripple phase is formed [44–47]. The transition from the “gelphase” to the “ripple phase” is called pretransition.

With respect to certain practical aspects in the methodsfor lipid vesicle preparations described in Section 3, the Tmvalue of the lipid (or the lipid mixtures) used is important toknow. In the case of dilute POPC-water systems, for exam-ple, Tm is around −3 �C [48, 49]. If the water content isdecreased below ∼10 wt%, Tm increases above 0 �C, until itreaches a value of 68 �C in the anhydrous system [49].

A list of different Tm values for a number of dilute aque-ous phosphatidylcholine systems (MLVs) is given in Table 4.For a more detailed list of Tm values, including other gly-cerophospholipids, see [38, 40, 50]. Please note that in thecase of phospholipids with charged head groups, the Tm val-ues depend on the degree of protonation and may dependconsiderably on the chemical nature of the counter-ionspresent [51]. Furthermore, measurements carried out withSUVs give values about 4–5� lower than the Tm valuesobtained from MLVs [43, 52, 53].

3. METHODS FOR PREPARINGNORMAL VESICLES

3.1. Introduction

The thermodynamic equilibrium state of glycerophospho-lipids (and many other bilayer-forming amphiphiles) inwater (or in a particular aqueous solution) is—probablyunder most experimental conditions—a stacked bilayerarrangement of the surfactant molecules, either as L· �-phase(above Tm) or as L· �-, P· ′�- (or P�′)-phase (below Tm) in equi-librium with excess aqueous phase (see Section 2).

Table 4. Main P�′ -L· � phase transition temperature (Tm) values ofdilute aqueous dispersions of certain common bilayer-forming phos-phatidylcholines, data taken from [39] and [518] (for soybean PCs).

Phosphatidylcholine Tm (�C)

DMPC (1,2-dimyristoyl-sn-glycero-3- 23.6 ± 1.5phosphocholine), 14:0/14:0

DPPC (1,2-dipalmitoyl-sn-glycero-3- 41.3 ± 1.8phosphocholine), 16:0/16:0

DSPC (1,2-distearoyl-sn-glycero-3- 54.5 ± 1.5phosphocholine), 16:0/16:0

POPC (1-palmitoyl-2-oleoyl-sn-glycero-3- −2.5 ± 2.4phosphocholine), 16:0/18:1c9

SOPC (1-stearoyl-2-oleoyl-sn-glycero-3- 6.9 ± 2.9phosphocholine), 18:0/18:1c9

DOPC (1,2-dioleoyl-sn-glycero-3- −18.3 ± 3.6phosphocholine), 18:1c9/18:1c9

Egg yolk PCs (see Table 2) −5.8 ± 6.5Soybean PCs (see Table 2) −15 ± 5Hydrogenated soybean PCs 51–52

Note: 16:0/18:1c9, for example, indicates that the linear acyl chain at sn-1has 16 carbon atoms without any double bonds; the linear acyl chain at sn-2 has18 carbon atoms with one cis double bond in position 9,10 (see also Table 2).

Upon dispersing in an aqueous solution, vesicles gener-ally form from an amphiphile (or a mixture of amphiphiles)that forms a lamellar L· �-phase at thermodynamic equilib-rium. Depending on how the dispersion is actually prepared(in other words, which method or technique is applied), thevesicles formed by the dispersed amphiphiles are either veryheterogeneous or rather homogeneous and are mainly small(below about 50 nm), mainly large (between about 50 nmand 500 nm), or mainly very large (above about 500 nm). Itall depends on the lipid (or lipid mixture) used, on the aque-ous solution, and particularly on the preparation method.

In the following, the principles of some of the best knownand widely used methods for the preparation of lipid vesiclesuspensions—mainly on a laboratory scale of a few millilitersup to about 100 ml—will be described. For each method,a different vesicle preparation with different typical generalcharacteristics is obtained.

It is important to point out once more that in most ofthe cases the resulting vesicle suspension is not at ther-modynamic 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 vesicledispersion prepared is therefore physically (with respect tovesicle size and lamellarity) not indefinitely stable; it mayslowly transform into the thermodynamically most stablestate (stacked bilayers), as a result of a so-called aging pro-cess [54, 55]. This aging may occur either through the fusionof vesicles or because of an exchange of amphiphiles thatare not aggregated (free, monomeric surfactant) [55]. Thislatter process—called Ostwald ripening (in analogy to thecorresponding process occurring in emulsion systems)—isexpected to be particularly relevant in the case of syntheticshort-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 processis less likely in these cases.

50 Preparation of Vesicles (Liposomes)

Although often only metastable, vesicle suspensions maybe stable for a prolonged period of time, for example, fordays, weeks, or months, provided that the vesicle-formingamphiphiles are chemically stable during this period of time[57, 58].

In the presence of other amphiphilic molecules (cosurfac-tants), the situation may change, particularly if the cosur-factant tends to form micellar aggregates, characterized bya packing parameter p ∼ 1/3 (relatively large head groupcross-sectional area, a0). In this case mixed micelles mayexist at thermodynamic equilibrium (in equilibrium withcosurfactant monomers), if the micelle-forming surfactant ispresent to a large extent. Such mixed micellar systems areused as a starting solution in the case of the so-called deter-gent depletion method described in Section 3.27. Further-more, there are also known cases where there appear to bethermodynamically stable vesicles (particularly composed ofmixtures of surfactants) [59, 60] (see Section 6).

The presence of cholesterol (or other sterols)—moleculesthat are water insoluble and do not form vesicles alone—may also influence the properties of lipid vesicles, dependingon the amount of cholesterol present [61, 62]. In the case ofDPPC, for example, up to 33 mol% cholesterol, the Tm valueof hydrated bilayers changes only slightly [63]. With increas-ing cholesterol concentration, however, the phase transitiontemperature is completely eliminated at 50 mol% (1:1 molarratio of DPPC to cholesterol). The fluidity of the bilayermembrane is thereby changed, resulting in an increase in thefluidity below Tm of the PC and a decrease in fluidity aboveTm, a state of the membranes that is intermediate betweensolid-ordered and liquid-disordered (see Section 2.3). Thisstate is called liquid-ordered [34].

For all of the methods outlined in the following, moredetailed descriptions can be found in the original litera-ture cited. Furthermore, most of the generally known meth-ods have already been summarized before—more or lessextensively—in review articles or books about lipid vesicles(liposomes) [14, 31, 64–75].

Most of the methods can be roughly divided into twogroups:

(i) Methods that are based on the simple swelling of ini-tially dried, preorganized lipids and the mechanicaldispersion and mechanical manipulation of the dis-persed bilayers (Sections 3.2–3.12).

(ii) Methods that involve the use of (a) a cosolvent inwhich the lipids are soluble (Sections 3.14–3.26 and3.30), (b) an additional non-bilayer-forming “helperamphiphile,” a coamphiphile (Sections 3.27 and 3.30),or (c) certain ions that influence the initial aggre-gational state of the lipids (Sections 3.13 and 3.29).All three type of molecules control the assembly pro-cess of the bilayer-forming amphiphile in a particu-lar way during the vesicle preparation process, andall three types of molecules may at the end be dif-ficult to remove completely from the final vesiclesuspensions.

3.2. MLVs, GUVs, or Myelin FiguresFormed by Thin Lipid Film Hydration

One of the easiest ways to prepare a vesicle suspensionis 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 inwhich the amphiphile is soluble (usually chloroform). Thissolution is then placed inside a round-bottomed flask, andthe solvent is completely removed by rotatory evaporationunder reduced pressure, followed by high-vacuum dryingovernight. The remaining amphiphiles form a dry thin filmthat is oriented in such a way as to separate hydrophilic andhydrophobic regions from each other [31]. If an aqueoussolution is added to this film at a temperature above themain phase transition temperature Tm (see Section 2.3),the head groups of the dry lipids become hydrated andhydrated bilayers form. The hydration and swelling processis usually speeded up by gentle or vigorous shaking (using avortex mixer), thereby dispersing the bilayers in the aqueoussolution, resulting in the formation of mainly MLVs, whichare very heterogenous with respect to size and lamellarity.Lipid film thickness and extent of shaking have an influenceon the properties of the resulting vesicle suspension. Theinterlamellar repeat distance in equilibrated, completelyhydrated 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 isabout 2.5 nm thick (taking into account a bilayer thicknessof 4 nm). On average, a MLV may be composed of up to10 bilayers [81].

The formation of closed bilayers (vesicles), in contrast toopen bilayers, can be easily understood on the simple basisthat interactions between the hydrophobic chains of theamphiphiles and the water molecules—as would be the casein open bilayers—are energetically unfavorable and there-fore rather unlikely.

The preparation of MLVs by the dispersal of a dried lipidfilm (also called hand-shaken MLVs [31]) is often a first stepin the preparation of more defined vesicle suspensions (see,for example, Sections 3.5 and 3.8). With respect to the equi-libration of water-soluble molecules between the bulk aque-ous medium and the inner aqueous compartments of MLVs,ionic species may be unevenly distributed [76]. A more evendistribution can be achieved by applying freeze/thaw cycles(see Section 3.6).

The experimental conditions under which a dried lipidfilm is hydrated affect the resulting lipid aggregatesobtained. In the case of phospholipid mixtures contain-ing 90 mol% PC and 10 mol% of a negatively chargedphospholipid (egg yolk phosphatidylglycerol, bovine brainphosphatidylserine, or bovine heart cardiolipin), the driedlipid film prepared inside a test tube can first be prehydratedat 45 �C with water-saturated nitrogen gas for 15–25 min.Afterward, an aqueous solution containing water-solublemolecules to be entrapped (e.g., 100 mM KCl and 1 mMCaCl2) is gently added, and the tube is sealed under argonand incubated at 37 �C for 10–15 h. During this incubation,the lipid film is completely stripped from the glass surfaceand forms vesicular aggregates as a kind of white, float-ing precipitate in the aqueous solution. The analysis of this

Preparation of Vesicles (Liposomes) 51

precipitate indicated the presence of many mainly unilamel-lar giant vesicles (not MLVs) with diameters on the orderof tens of micrometers up to more than 300 �m [82]. Inaddition, much smaller vesicles, large multilamellar vesiclesas well as myelin figures (see below) and undispersed lipidmaterial, could also be observed [82]. In the case of eggPC alone (no negatively charged phospholipids present), nogiant unilamellar vesicles (GUVs) formed under the experi-mental procedure used [83]. It therefore seems that electro-static repulsions between the charged head groups facilitatethe formation of unilamellar membranes by opposing theintrinsic adhesive force between the membranes [83]. If diva-lent cations (1–30 mM Ca2+ or Mg2+) are present, giantvesicle formation is also observed with zwitterionic phospho-lipids alone (POPC), with the use of a procedure almostidentical to the one just described [84]. Divalent cationsseem to promote giant unilamellar vesicle formation in thecase of POPC due to a binding of the ions to the free phos-phate oxygen of the lipid head group, which is known toalter the mean head group conformation [85] and the fluid-ity of the lipid bilayer [86], and which makes a zwitterionicPC molecule positively charged overall [87].

In an independent study and with a different experimentalprocedure, the formation of giant vesicles from PCs (DOPCor soybean PCs) in the presence of Mg2+ (<10 mM) hasalso been observed [87].

The general role of ions (including buffer ions)—as wellas dissolved gas molecules—in vesicle formation and inphysical chemistry at large is an open question [88] (seeSection 6).

The preparation of MLVs by thin lipid film hydration gen-erally involves the use of round-bottomed flasks and gentleshaking. If flat-bottomed flasks are used instead and if addi-tionally the lipid swelling process occurs undisturbed (noshaking) above Tm of the lipids [89] over a period of severalhours, the vesicles formed are no longer mainly multilamel-lar but rather mainly unilamellar, with diameters between0.5 and 10 �m [90] or even at 300 �m [91]. These are GUVs.

The preparation of GUVs by simple hydration of cer-tain amphiphiles deposited from an organic solution (e.g.,a chloroform-methanol mixture) on a flat surface can beconsiderably improved to yield a higher fraction of unilamel-lar vesicles by using a roughened flat disk of Teflon (poly-tetrafluoroethylene) [92]. The vesicles formed (e.g., fromDMPC after a slow swelling at a temperature of 30–35 �C)can be harvested by gentle pipette aspiration for furtherindividual investigation and micromanipulation [92, 93]. Thesizes of the vesicles thus formed are in the range of 20–40 �m [93� 94].

The slow swelling of dried phosphatidylcholines with theaddition of water had already been observed and investi-gated by light microscopy in the middle and at the end ofthe nineteenth century by Rudolf Virchow [95] and OttoLehmann [96]. The elongated, tubular structures that areobserved to grow from a deposit of PC molecules (e.g., eggyolk PCs) with the addition of water under undisturbed con-ditions (no shaking) are called myelin figures, as namedby Virchow [95] while making these observations with a(PC-containing) lipid extract from myelin, the isolating lipidsheath surrounding the elongated portion of nerve cells [97].Myelin figures are structures dozens of micrometers long

and a few micrometers thick. In the case of egg yolk PCsand 25 �C, water addition leads to the formation of myelinfigures that have a diameter of about 20 �m and a length ofseveral hundred micrometers [98]. Myelin figures are cylin-drical, rodlike structures composed of many lamellae formedby the amphiphiles, stacked coaxially around the rod axis,with water between all of the bilayers [98, 99].

The conventional solid surface on which the lipid film isinitially dried—glass, as described above—has been replacedby a support made from microcrystals (such as zeolite X witha crystallite size of 400 nm) [100]. Using very dilute mixturesof egg yolk PC and the positively charged amphiphile hex-adecyltrimethylammonium bromide, the swelling in water(or 5 mM NaCl) of the very thin dried film deposited on zeo-lite X led neither to the formation of MLVs nor to GUVs ormyelin figures, but to uniform SUVs with diameters around22 nm [100]. In general, the size distribution can be con-trolled to some extent by the topography of the surface uponwhich the phospholipid film is deposited [101].

3.3. GUVs Preparedby the Electroformation Method

If the swelling of PCs (or other phospholipids) in water (orin an aqueous solution with low ionic strength) is carriedout in a controlled way by applying an alternating electricfield, instead of myelin figures, GUVs form, with diameterstypically between 5 and 200 �m, depending on the chemicalstructure of the lipids used, on the lipid composition, on theswelling medium, and on the external electric field parame-ters [102–104]. The vesicles are often formed under a lightmicroscope in a specially designed cell on platinum wiresthat are positioned at a certain fixed distance [103, 104]. Thisso-called electroformation method has been proved to berather powerful for the investigation (including microinjec-tion [105, 106]) of individual GUVs of defined size by lightmicroscopy [103–110].

The mechanism of GUV electroformation in an alternat-ing electric field is not fully understood. It is likely, however,that the electroosmotic motion of the water molecules isresponsible for a controlled swelling and separation of thebilayers deposited on the platinum wires, leading to the for-mation of unilamellar vesicles on or close to the metal wires[103, 109]. The individual giant vesicles formed seem to beconnected to the platinum wires through thin lipid bilayertubes (so-called tethers [111, 112]) and possibly myelin-likeprotrusions [113].

In comparison with other methods for the preparation ofgiant unilamellar PC vesicles, the electroformation methodoffers several advantages with respect to reproducibility andlater vesicle manipulations [114].

3.4. MLVs Prepared by Hydrationof Spray-Dried Lipids

Instead of the preparation of a thin lipid film first, followedby hydration (Sections 3.2 and 3.3), the lipids can first bespray-dried and then hydrated [115]. With respect to thereproducibility and mass production of MLVs, the spray-drying method has several advantageous over the thin-filmdispersion method, although it seems that lipids with low Tm

52 Preparation of Vesicles (Liposomes)

(e.g., egg yolk PCs) cannot be used, because of adhesion tothe containers in which the vesicles are prepared [115].

In the spray-drying method, the vesicle-formingamphiphiles are first dissolved in chloroform or methylenechloride (in which, advantageously, mannitol is dispersed)and then spray-dried with a commercial spray-dryer. Vesicleformation is observed after hydration of the spray-driedlipids and vortexing above Tm of the amphiphile used.The presence of the mannitol prevents adhesion of the(saturated) lipids to the reaction containers and leads to abetter hydration behavior. (The role of sugars in surfactantassembly at large has hardly been touched on.)

The vesicles prepared are rather heterogeneous withrespect to size and lamellarity. One of the homogenizationprocedures described below may follow the spray-drying andhydration step.

3.5. SUVs (and Possibly LUVs) Preparedby Sonication (and Storage)

The treatment of a MLV suspension with ultrasound at atemperature at least about 5 �C above Tm leads to a homog-enization of the vesicles by reducing the size of most of thevesicles to probably the smallest possible diameter (about20 nm in the case of egg yolk PC mixtures [10, 116] orPOPC), due to simple molecular packing considerations.The vesicles thus prepared are unilamellar and called SUVs,an abbreviation that stands for sonicated unilamellar vesiclesor small unilamellar vesicles. If the sonication is performedbelow Tm, structural defects within the bilayers are observedthat result in an increased bilayer permeability [117, 118].

So-called probe sonication (the insertion into the vesiclesuspension of a metal rod that releases ultrasound wavesfrom the tip of the rod) [119–121] is more efficient thanbath sonication [31, 122]. In both cases, however, the sizes ofthe vesicles (20–50 nm usually) in the sonicated suspensiondepend on the sonication time [122], on the lipid composi-tion, and on whether other compounds (such as cholesterol)are present [14, 119, 123]. Furthermore, the vesicle suspen-sion may not be free from larger MLVs, and a separationstep, for example, size exclusion chromatography (using, forexample, sepharose 4B) or high-speed centrifugation, is usu-ally necessary [119, 121, 124]. A centrifugation step may alsobe used in the case of probe sonication to remove the pos-sibly released metal particles from the tip of the metal rod[125].

The ultrasound used in the treatment of MLVs consistsof pressure waves with frequencies around 20 kHz, and theultrasound propagation gives rise to periodic changes inlocal pressure and temperature [126]. Therefore, the heatgenerated during sonication has to be controlled and com-pensated for by cooling, otherwise the vesicle-forming lipidsmay undergo a partial chemical degradation [125, 127]. Fur-thermore, it has been reported that nonvesicular aggregatesmay be generated as a result of the high-energy input intothe system, and the ultrasound treatment may lead to a dras-tic decrease in the surface tension of sonicated MLV suspen-sions due to the accelerated movement of lipid moleculesform the vesicles to the water-air interface [128]. This pro-cess is usually much slower in the case of vesicles that have

been prepared under milder conditions, such as by extrusion(Section 3.8) [128].

Based on this latter observation, more uniform andreproducible SUV preparations may be obtained by anneal-ing at elevated temperatures (50 �C in the case ofDMPC:1,2-dimyristoyl-sn-glycero-3-phosphoserine vesicles),followed by a removal of possibly present large aggregatesby centrifugation at 100�000 × g (30 min at 37 �C in theparticular case) [129].

If carefully prepared SUVs from the saturated DPPC ata concentration in the range of 20–120 mM (in a 20 mMpiperazine-N ′�N ′′-bis(2-ethanesulfonic acid) buffer, pH 7.4,containing 10 mM NaCl) are kept at a temperature consid-erably lower than the Tm of DPPC (which is around 41 �C;see Table 4), then the vesicles fuse into uniform LUVs witha diameter of about 70 nm (after 7 days at 4 �C) or 95 nm(after 3–4 weeks at 4 �C) [130]. At higher temperatures(21 �C) but still below Tm, the fusion process is slower [131],whereas at 50 �C (above Tm), no appreciable fusion occursover a period of at least 5 days [130]. A similar fusion ofSUVs below Tm is observed for DSPC, resulting in vesiclesof about 60- and 100-nm diameter [132], with an apparentlyhigher fusion rate [130].

The implications of these experimental observations aretwofold: (a) LUVs of about 70 or 100 nm can be preparedfrom saturated PCs by simple storage of SUVs below Tm.(b) SUVs should not be stored below Tm, if one likes tokeep the vesicle sizes small; otherwise the vesicles may fuseto form larger vesicles.

In general, SUVs are often prepared in a first step tofuse them in a second step to LUVs by using anothermethodology, for example, in the case of the cochleate-cylinder method (see Section 3.13) or in the case of theinterdigitation-fusion method (see Section 3.21).

3.6. MLVs and MVVs (and PossiblyLUVs) Prepared by RepetitiveFreezing and Thawing

A MLV suspension—as prepared by the thin-film disper-sion method—is put, for example, inside a thick-walled testtube and repetitively (3–6 or, better, 10 times) completelyfrozen in liquid nitrogen (at −195 �C) and then thawed ina water bath kept at a temperature above Tm. In this way,the vesicle suspension undergoes a sort of equilibration pro-cedure caused by the water crystals and transient rigidifi-cation of the lipid molecules. It is a kind of homogeniza-tion process, and it has been reported that the populationof MVVs may increase [133] and that the amount of verysmall vesicles tends to decrease [134, 135], depending on thelipid used and on the salt content [136]. It may also be thatfreeze-thaw cycles lead to a fragmentation of large MLVsinto large or small unilamellar vesicles [136, 137]; it all verymuch depends on the experimental conditions, such as typeand concentration of lipid (or lipid mixture) and salt content[136, 138].

A further effect of freezing and thawing of vesicle sus-pensions is the removal of dissolved gas [88], which has notyet been explored. It is known that removal of dissolved gasstrongly affects hydrophobic interactions and colloidal sta-bility, as well as the structure of water [88, 139].

Preparation of Vesicles (Liposomes) 53

Freeze-thaw cycles have been applied to a series ofphospholipid vesicular suspensions, particularly focusing onphospholipids with unsaturated acyl chains (DOPC) andDOPC/DOPA mixtures (DOPA is the abbreviation forthe negatively charged 1,2-dioleoly-sn-glycero-3-phosphate)[136]. It has been shown that freeze-thaw cycles can beapplied successfully to the preparation of mainly LUVs start-ing from MLVs, depending on the experimental conditions[136]. With the use of a 0.1 M Hepes (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid) buffer of pH 8.0 contain-ing 0.1 M KCl and 5 mM EDTA, 10 freeze-thaw cycles ofaqueous suspensions of DOPC (or POPC) MLVs, for exam-ple, led to suspensions that contained about 98% unilamel-lar vesicles with diameters below 200 nm. The remaining2% were multilamellar. The vesicles obtained by repeti-tively freezing and thawing MLVs are often abbreviated asFATMLVs [140]. In the case of DOPC/DOPA (80:20 mol%),almost no MLVs could be detected after the freeze-thawprocedure; the vesicles again had diameters below 200 nmafter 10 freeze-thaw cycles [136]. After 50 cycles, the vesi-cle diameters decreased to below 100 nm [136]. In thecase of saturated PCs (DMPC or DPPC), MLV suspensionsremained essentially multilamellar, even after 10 freeze-thaw cycles under the conditions used [132].

Fusion of SUVs into larger vesicles upon repetitive freez-ing and thawing [31, 136, 140, 141] seems to occur efficientlyonly in pure water (without added salt) [136]. Furthermore,freeze-thaw cycles may not only affect the mean vesicle sizeand lamellarity, but also the entrapment efficiency for water-soluble molecules [142–145]. In some cases, the trappingefficiency may be dramatically increased [142]. Furthermore,the composition of the aqueous solution may generally influ-ence the freeze-thaw behavior of the vesicles; all of thefreeze-thaw effects depend on the experimental conditions[136, 138].

Depending on the phospholipid(s) used, freezing andthawing of SUVs may lead to the formation of giant MLVswith diameters ranging from 10 to 60 �m in the presence ofsalt (30–500 mM KCl, in the case of phospholipid mixturescontaining egg yolk phosphatidylethanolamine and bovinebrain phosphatidylserine, 7:3, mol/mol, at pH 7.0, for exam-ple) [146].

Although the process is not completely understood,freeze-thaw cycles often precede a further downsizing of thevesicles, for example, by extrusion (Section 3.8).

3.7. MLVs Preparedby Dehydration and Rehydration

The dehydration of a vesicle suspension followed by a con-trolled rehydration at a temperature above Tm leads to thefusion of small vesicles present in the suspension, result-ing finally in the formation mainly of MLVs [147–149]. Thisprocedure is based on the instability of dried SUVs anda lipid lamellae formation (SUV fusion) during the waterremoval. The resulting vesicle suspensions are referred toas dehydration-rehydration vesicles or dried-reconstitutedvesicles (DRVs) [31]. DRVs are often prepared to obtainhigh entrapment efficiencies (high encapsulation yields) forwater-soluble enzymes and other water-soluble molecules.The entrapment efficiency is defined as the percentage

amount of the encapsulated molecule in relation to the totalamount of molecules present during the vesicle formationand entrapment process.

In a typical, originally described preparation of DRVs[148], SUVs of the appropriate lipid composition are firstprepared in distilled water by probe sonication. After cen-trifugation to remove large vesicles and titanium particlesreleased from the sonicator probe (see Section 3.5), theSUV suspension is mixed with an equal volume of an aque-ous solution of the compounds to be trapped. After rapidfreezing of the mixture, it is lyophilized with the use of acommercial freeze-dryer at reduced pressure. After freeze-drying, the preparation is rehydrated with a volume of dis-tilled water equivalent to one-tenth of the total volume ofthe initial SUV suspension. The use of a small volume nec-essarily has a positive effect on the entrapment yield. Therehydration is usually aided by light vortexing and equilibra-tion for 30 min. Nonentrapped molecules are removed bycentrifugation or dialysis after an appropriate dilution of thevesicle preparation [148].

Since DRVs are generally not uniform with respect tosize and lamellarity, a second downsizing process often fol-lows a DRV preparation, such as extrusion (Section 3.8) ormicrofluidization (Section 3.9).

The essential process in the preparation of DRVs is thedisintegration of the initially prepared SUVs during the dehy-dration step, which finally results in high entrapment yieldsof the DRVs. If the experimental conditions during dehydra-tion (during freeze-drying) are such that the vesicles remainlargely intact, the rehydration will not lead to exception-ally high entrapment yields of DRVs. One particular suchcase is the use of lyoprotectants, chemicals that protect thelipid vesicles against drying stress (also called cryoprotec-tants [150]), such as trehalose [151], glucose [152], sucrose[153], or maltose [154], used typically at concentrations of10% (wt/wt) [153]. For a particular lyoprotectant, the lyopro-tection effect is very much dependent on the type of vesi-cle-forming lipid used, on the bilayer composition, on thesize of the vesicles, on the temperature at which the vesi-cles are kept before rehydration, and on the freezing rate[153, 155, 156]. Lyoprotectants are thought to prevent theincrease in the Tm of the lipid during dehydration (see Sec-tion 2.3), to bind water molecules and to interact with thepolar head groups of the lipids. In the latter case, watermolecules around the head groups would be replaced by thelyoprotectants, thereby protecting the vesicles from aggrega-tion and fusion during the freeze-drying process [157, 158].In the particular case of POPC, the presence of sucrose orsorbitol results in a value of Tm in an almost dry state of lessthan 6 �C, in comparison with about 60 �C in the absence oflyoprotectant [159] (see Section 2.3).

Lyoprotectants are only effective in protecting the lipidvesicles if the proper concentrations of the added lyopro-tectants are used. If, for example, the potentially effectivelyoprotectant sucrose is used at a concentration below theone known to preserve the stability of vesicles during dehy-dration [157], DRVs can be prepared from SUVs, whichare characterized by high entrapment yields and a size nogreater than that of the initially prepared SUVs, but alsonot as large as the size obtained in the absence of sucrose:90–200 nm in the presence of low sucrose concentrations

54 Preparation of Vesicles (Liposomes)

(<150 mM, with an optimal entrapment efficiency at a molarratio of sucrose to lipid of 1), in comparison with about1–6 �m in the absence of sucrose [160].

3.8. LUVs Preparedby the Extrusion Technique

This method is one of the most popular for the reproduciblepreparation of rather homogeneous vesicle suspensions con-taining LUVs, often with a mean diameter of about 70 nm or100 nm [161–163]. The principle of the method is the follow-ing (see Fig. 3): a MLV suspension is passed under moderatepressure repetitively (usually 10 times) at a temperatureat least 5 �C above the Tm [164, 165] through the poresof track-etched polycarbonate membranes. The pores arealmost cylindrical, and vesicles (unilamellar or multilamel-lar) that are larger than the mean pore diameter are reducedin size and lamellarity during the passage through the pores,resulting in a final mean vesicle size that corresponds ina first approximation to the mean size of the pores. Vesi-cles smaller than the pore diameters pass through the poreswithout significant size change. Usually a MLV suspension isfirst passed 10 times through polycarbonate membranes witha relatively large mean pore diameter of 400 nm, followedby a passage 10 times through membranes with 200-nmpores and, finally, 100-nm pores. If desired, membranes with50-nm pores or even 30-nm pores can be used for final extru-sions. The corresponding vesicle suspensions obtained areabbreviated as VET400, VET200, VET100, VET50, or VET30.VET stands for “vesicles prepared by the extrusion tech-nique.” The subscript indicates the pore size of the mem-branes used for final filtrations. The mean vesicle diameterof VET200 is usually less than 200 nm because of the pres-ence of vesicles considerably smaller than 200 nm in theoriginal vesicle suspension [166, 167], unless the very smallvesicles are first eliminated by freeze-thaw cycles [162, 163](see Section 3.6). If freezing-thawing is used as a homoge-nization method before extrusion, the resulting vesicles areoften abbreviated as FAT-VET (e.g., FAT-VET100: repeti-tively frozen and thawed vesicles that have been repetitivelyextruded through polycarbonate membranes with mean porediameters of 100 nm for final extrusions).

In the early stages of the development of the extrusiontechnique, freeze-thaw cycles were not performed [166, 168,169], and, depending on the application, it may even be bet-ter to avoid it, for example, in the case of the preparationof vesicles containing entrapped enzymes that may be inac-tivated during freeze/thaw cycles [75].

VET100- or FAT-VET100 suspensions are generally rathermonodisperse, containing vesicles with a mean diameterclose to 100 nm. The mean vesicle diameter of VET50 orVET30 is generally larger than 50 nm or 30 nm, respectively.This observation can be explained on the basis of the pos-sible mechanism by which the vesicles transform during thepassage through the pores [170–172]: most likely, sphericalvesicles change their shape to cylindrical structures withinthe pores. Because of a velocity profile inside the pores—with a low velocity close to the pore wall and a high velocitytoward the center of the pores—the cylinders pearl off ellip-soidal vesicles, which then relax in size and transform intospheres upon leaving the pores. Figure 3 shows a schematic

~ 6000 nm

~ 100 nm

A

B

Figure 3. Illustration of the extrusion technique, which is the mostpopular technique for the reproducible preparation of sub-micrometer-sized unilamellar lipid vesicles from a heterogeneous multilamellarvesicle suspension (see Section 3.8). (A) Schematic representationshowing (more or less) cylindrical pores with a mean diameter of 100nm (as an example), spanning a polycarbonate membrane of about 6�m thickness; MLVs; a MVV; and LUVs before and LUVs after repet-itive extrusions (passages through the pores). (B) Transmission electronmicrograph of a commercial polycarbonate membrane (Nucleopore)with pores that have a nominal mean diameter of 100 nm. Length ofthe bar: 100 nm. The electron micrograph was taken by E. Wehrli, Ser-vice Laboratory for Electron Microscopy, at the Department of Biologyat the ETH Zürich.

drawing of the extrusion method (Fig. 3A) together with anelectron micrograph of a top view of a section of a 100-nmpolycarbonate membrane (Fig. 3B).

There are several commercially available devices thathave been developed for the preparation of extruded vesiclesuspensions [173] with volumes between 0.2 ml and 1.0 ml(e.g., LiposoFast [174]), between 1.0 ml and 10 ml (called theExtruder [163]), between 100 ml and 800 ml (the Extruder),or between 200 ml and 1 l (e.g., the Maximator [175]).A comparison between lipid vesicles prepared with the dif-ferent devices (and the same set of lipids) has been per-formed [173], and it has been found—with the exception ofFAT-VET50—that the vesicles prepared with the LiposoFastwere smaller than the vesicles prepared with the Maximator,

Preparation of Vesicles (Liposomes) 55

most likely because of higher flow rates and pressure drops.In the case of FAT-VET50, the vesicles had the same meandiameter (around 50–60 nm) [173].

Since the vesicles prepared by the extrusion techniquesare large (LUVs as defined in Table 3) and mainly unil-amellar, extruded vesicle suspensions are often abbreviatedas LUVET (e.g., LUVET100, meaning large unilamellarvesicles prepared by the extrusion technique, using forfinal extrusions polycarbonate membranes with 100-nm porediameters) [161]. FAT-VET200 and particularly FAT-VET400suspensions usually contain a considerable amount of oligo-lamellar vesicles [162].

The extrusion technique can be applied as final downsizing and homogenization in principle for any type of vesi-cle suspension, MLVs, DRVs (Section 3.7), or REVs (Sec-tion 3.14), resulting in MLV-VET, DRV-VET, or REV-VETsuspensions, respectively.

With a so-called French press cell—a device that doesnot contain any polycarbonate membranes at all but is oftenapplied for the disintegration of biological cells—a vesiclesuspension can be “extruded” through a small orifice athigh pressure, resulting in the formation of unilamellar oroligolamellar vesicles with diameters between about 30 and80 nm, depending on the pressure [31, 66, 71, 176, 177].These vesicles are referred to as vesicles prepared by theFrench press (FPVs) [72]. The mechanism of the change invesicle size in a French press is very different from the actualextrusion technique using polycarbonate membranes [66]. Inthe French press, shear forces seem to play a particular role[66] (see also Section 3.11).

3.9. LUVs and OLVs Preparedby the Treatment of a MLV Suspensionwith a Microfluidizer

Vesicle suspensions containing a mixture of mainly largeMLVs can be downsized at a temperature above Tm witha high-pressure homogenizer (a microfluidizer) [178–181].The resulting vesicles usually have a relatively narrow sizedistribution centered around a value between about 50 and300 nm [179–181]. The mean size of the vesicles obtained(LUVs and OLVs) depends on the experimental conditions,such as homogenization pressure [182], number of cycles[181–183], ionic strength [181], lipid concentration [183],and lipid composition [182]. Relatively high lipid concentra-tions can be used, and a large-scale production is possible[181, 184]. To achieve high entrapment yields, a microflu-idization step can be performed on vesicles first prepared bythe dehydration-rehydration method, which itself results inhigh entrapment yields [180] (see Section 3.7).

3.10. Preparation of Oligolamellarand Multilamellar Novasomes

Novasomes, also called novasome vesicles [185], can be man-ufactured from a variety of amphiphiles (including PCs andmany nonphospholipid surfactants) by a unique and cheapprocess on a laboratory scale (a few milliliters) or for indus-trial bulk applications.

The bilayer-forming amphiphile (including any bilayer-modulating molecules to be added, such as cholesterol) is

first heated above the melting temperature (above about70 �C in the case of POPC; see Section 2.3) to give a liq-uid. This liquid is then injected at high velocity (10–50 m/s)through small channels with 1–3-mm diameters (or througha needle) with turbulent mixing into excess aqueous phaseand immediately cooled to room temperature. The rapidinjection leads to the formation of tiny droplets of theamphiphiles that are quickly converted into small aggre-gates. The subsequent cooling under conditions of contin-ued turbulence then leads to the formation of vesicles (theNovasomes) within milliseconds [185]. Novasomes are OLVsor MLVs with diameters typically between 200 nm and1 �m, depending on the experimental conditions. The sizeof the vesicles is controlled by the chemical structure of theamphiphile used and by the hydrodynamic shearing duringthe fusion of the small aggregates during the cooling pro-cess. Furthermore, the diameter of the exit channel plays arole [185].

3.11. Preparation of MultilamellarSpherulites

Spherulites (also called onions or multilamellar spherulites)are relatively uniform, densely packed MLVs that are pre-pared by shearing of a lyotropic lamellar phase under con-trolled conditions [186, 187]. The MLVs formed show verylittle polydispersity and can have mean diameters some-where within 100 nm and 20 �m [188] or within 200 nm and50 �m [187], depending on the formulation (i.e., the chem-ical structure of the amphiphile and possible additives) andthe shear rate.

The starting lamellar phase is concentrated and composedof a stack of bilayers of the amphiphiles, separated by waterlayers. This lamellar phase can be composed of a largenumber of different amphiphiles or mixtures of amphiphiles[189–192], including soybean PCs [193]. During the shearing,which is performed in specially prepared Couette cells [186],the lamellar phase reorganizes into spherical or polyhedralmicrodomains [186], which can be dispersed in an excessamount of aqueous phase. The process is called spherulitetechnology [186, 187, 189]. Spherulites can be considereddroplets of dispersed lamellar phase [188, 193], and the tech-nology allows the entrapment of a variety of different water-soluble compounds at high yields, including enzymes [194],nucleic acids [195], and metal ions for nanoparticle synthesis[187, 190].

The formation of spherulites is a direct consequence ofglobal packing constraints [196].

3.12. MLVs Prepared by the Bubble Method

This method is based on the bubbling of an inert gas (nitro-gen) over several hours through a coarse dispersion of (ini-tially) nonhydrated lipid particles (containing phospholipidsor non-ionic surfactants) and results in vesicle preparationswith a clearly non-uniform size distribution [197]. The actualbubbling unit consists of a round-bottomed flask with threenecks, one used for a water-cooled reflux, one for a ther-mometer, and one for the gas supply. A continuous streamof gas bubbles is generated at the bottom of the flask.A coarse vesicle dispersion is first formed usually with a

56 Preparation of Vesicles (Liposomes)

homogenizer, and the gas bubbling is carried out at a tem-perature above the Tm of the lipids (actually at 80 �C inthe case of a mixture of hydrogenated soybean PCs anddicetylphosphate (10:1, mol:mol). The Tm of hydrogenatedsoybean PCs is about 51–52 �C (see Table 4). Dependingon the experimental conditions, the resulting vesicles have amean diameter between 200 and 500 nm [197].

3.13. LUVs and GUVs Preparedby the Cochleate Cylinder Method

A so-called cochleate cylinder is a precipitate of a cylindri-cal, cigar-like phospholipid aggregate that looks like a snailwith a spiral shell (its Greek name is cochleate [198]).

In the standard procedure, cochleate cylinders form uponstepwise addition of Ca2+ to SUVs prepared by bath son-ication from phospholipid mixtures containing negativelycharged amphiphiles, such as phosphatidylserine (or phos-phatidylglycerol) [31, 199].

Cochleate cylinders are rolled bilayers that do not containan interlamellar aqueous space. The divalent cation bringsthe negatively charged lipids into close contact and excludeswater. A majority of the lipid present must therefore be neg-atively charged [198], and an excess of Ca2+ with respect tothe negative charges present has to be added.

Instead of adding Ca2+ to the SUVs prepared, the vesiclesuspension can also be dialyzed against an aqueous solutionthat contains the required amount of Ca2+ ions [198]. Fur-thermore, instead of SUVs, unilamellar vesicles preparedby the detergent-depletion method (see Section 3.27), withd-glucopyranoside as a detergent, can be applied [198].

Once cochleate cylinders are prepared, LUVs are formedunder the appropriate conditions upon removal of the Ca2+

ions by the complexing agent EDTA (ethylenediaminete-traacetate) at a controlled pH of around 7.4. Experimentally,the cochleate cylinders are first separated from the bulkCa2+-containing solution by centrifugation to form a tightpellet, which is further used and made free from lipid-boundCa2+ by one of three different procedures, with the use ofEDTA (direct EDTA addition, rotatory dialysis, or agaroseplug diffusion [198]). EDTA binds Ca2+, which leads to anunrolling of the bilayers and then to a formation of (not veryuniform) unilamellar vesicles with diameters in the range ofbelow 1 �m to about 10 �m, depending on the experimentalconditions used, such as the Ca2+ complexation procedure(see above) [198]: while the direct EDTA addition results invesicle sizes below 1 �m, rotatory dialysis yields vesicles withan average size of 0.5–1 �m, and the agarose plug diffusionmethod gives vesicles below 1 �m or 5–10 �m, dependingon the procedure.

3.14. LUVs, OLVs, MLVs, and SPLVsPrepared by the Reversed-PhaseEvaporation Technique

For vesicles prepared by the so-called reversed-phase evap-oration technique, cosolvents are used [31, 200]. Thebilayer-forming amphiphile (e.g., POPC) is first dissolvedin a water-immiscible organic solvent of low boiling point(e.g., diethylether, isopropylether, or mixtures of theseethers with chloroform or methanol). This solution is then

mixed with an aqueous solution in which the final vesiclesare to be formed. After vortexing and brief sonication in abath-type sonifier, a reverse emulsion (a so-called water-in-oil emulsion, abbreviated w/o emulsion) is formed in whichaqueous droplets are stabilized for a certain period of timeby the amphiphiles at the interface between the dropletsand the bulk organic solution. The organic solvent is thenremoved under reduced pressure at a temperature aboveTm. During the solvent removal, the reversed emulsion col-lapses and is transferred into bilayered vesicles. The result-ing vesicles (abbreviated REVs) are often unilamellar oroligolamellar vesicles [200], and the size is usually not veryuniform, ranging from about 200 nm to 1 �m [200] or evenmore. Therefore, a homogenization step, such as extrusion(Section 3.8), often follows REV preparation [169, 201].

Vesicles prepared by the reversed-phase evaporation tech-nique are useful for obtaining high encapsulation yields forwater-soluble molecules (similar to the DRV preparationdescribed in Section 3.7).

In a modification of the originally developed REVmethod, the experimental conditions are altered in such away that the phospholipid/water ratio in the water-in-oilemulsion is changed; and the vesicles formed are no longermainly uni- or oligolamellar, but mainly MLVs with sizesabove 600 nm, thereby allowing higher encapsulation yieldsfor water-soluble compounds [202–204].

With the aim of applying the principles of the REVmethod to large scale productions, the original method hasbeen modified with the use of the nontoxic and cheap super-critical fluid CO2 [205]. CO2 has a critical temperature of31 �C and a critical pressure of 73.8 bar [205], above whichCO2 exists as liquid. For the preparation of DPPC vesicles,the temperature was set at 60 �C (well above Tm; see Table 4)and the pressure was kept at 200 bar [205]. The resultingsuspension contains mainly unilamellar vesicles with diame-ters between ∼0.1 and 1.2 �m and high trapping efficiencies[205].

Another procedure that is rather similar to the origi-nally developed preparation of REVs has been described;it resulted in the formation of stable MLVs (SPLVs, sta-ble plurilamellar vesicles [76], to distinguish them from theMLVs obtained by the thin-film method, as described in Sec-tion 3.2). In the case of SPLVs, a dry phospholipid film isfirst dissolved in diethyl ether. After the addition of a buffersolution, the two-phase mixture formed is emulsified witha sonication bath, during which a gentle stream of nitro-gen gas is passed over the mixture, until the ether is almostcompletely evaporated. The resulting mass (“cake”) is resus-pended in buffer solution, followed by a pelleting by cen-trifugation and washing with buffer [76].

3.15. LUVs Prepared FromW/O- and W/O/W-Emulsions

This multistep procedure for the formation of vesicles[206, 207] is very similar to the reversed-phase evapora-tion method (Section 3.14). It starts with the preparationof a w/o emulsion containing the water-soluble moleculesto be entrapped in the final vesicle preparation. The emul-sion is formed from soybean PCs, cholesterol, and benzene

Preparation of Vesicles (Liposomes) 57

(or methylene chloride). This w/o emulsion is then emulsi-fied with another aqueous phase to form a water-in-oil-in-water emulsion (w/o/w emulsion), paralleled by evaporationof the solvent molecules, which is speeded up by mechani-cal agitation and a stream of nitrogen gas [207]. The mainlyunilamellar vesicles thus obtained have a reported meandiameter of about 400 nm [207]. The size of the vesicles(apparently always between 50 and 500 nm) depends on theexperimental conditions, such as the intensity of the secondemulsification step (to form the w/o/w emulsion) [206] andthe chemical nature of the organic solvent used [207]. Withincreasing boiling point of the solvent, the mean size of thevesicles tends to decrease [207].

3.16. MLVs, GUVs, and MVVs Prepared bythe Solvent-Spherule (W/O/W Emulsion)Method or DepoFoam Technology

This particular method is called the solvent-spherule methodby its inventors [208], since solvent spherules (surfactant-stabilized o/w emulsion droplets) are the starting systemfrom which the oil (the solvent) is removed in a particularway, resulting in the formation of micrometer-sized MLVs[208]. The method is conceptually similar to the methodsdescribed in Sections 3.14 and 3.15.

The vesicle-forming amphiphiles (necessarily containinga small amount of negatively charged surfactants) are firstdissolved in a 1:1 (v/v) mixture of chloroform and diethylether. This lipid solution is placed under the surface of anaqueous solution (5% glucose) with a glass capillary pipette.After agitation for about 1 min, surfactant-stabilized sol-vent spherules (droplets) form in the aqueous phase (w/o/wemulsion). The volatile solvent mixture is then removed ina particular way by careful dropwise addition to a flask towhich a stream of nitrogen gas is added. The flask is keptat 37 �C and gently swirled. The average size of the MLVsformed is affected by the lipid concentration and the size ofthe lipid spherules formed by mechanical agitation.

In a modification of the method, which involves the addi-tional use of triolein and a certain complex way of mixing,evaporation and centrifugation steps, GUVs in the 5–10-�msize range can be prepared, depending on the experimentalconditions (strength and duration of the initial vortexing toform the o/w emulsion droplets) [31, 209].

In a further variation of the method, multivesicular vesi-cles (MVVs), with sizes between about 5 and 30 �m, canbe prepared by a similar stepwise procedure (which involvespelleting by centrifugation), using a particular lipid mix-ture and, as solvents chloroform, diethyl ether and triolein[31, 210]. The technology for the preparation of this type ofMVV is known as DepoFoam technology [211].

DepoFoam technology is a double emulsification processthat has been developed based on initial observations [210]of the formation of MVVs. The vesicles formed are mul-tivesicular vesicles (also called multivesicular liposomes),micrometer-sized vesicles that contain internal, nonconcen-trically arranged compartments. The internal packing iscomparable to the way gas bubbles are packed in a gas-liquidfoam [211, 212]. The contacts between the compartmentsexhibit a tetrahedral coordination.

The first step in the formation of DepoFoam MVVs isthe preparation of a w/o emulsion by dissolving a mixtureof vesicle-forming amphiphiles (e.g., phospholipids) con-taining at least one neutral lipid (e.g., triolein) in one ormore volatile, water-immiscible organic solvents (e.g., chlo-roform) and the addition of an aqueous solution containingwater-soluble molecules to be entrapped in the final vesiclesformed. In a second step, the w/o emulsion is mixed witha second aqueous solution, followed by mechanical mixingto yield solvent spherules suspended in the second aqueousphase (a w/o/w emulsion [212]). The organic solvent is thenremoved from the spherules by evaporation at reduced pres-sure or by passing a stream of nitrogen gas over or throughthe suspension [211]. The properties of the MVVs formed(such as captured volume) depend on the experimental con-ditions, such as molar fraction of the neutral lipid [210].

The presence of a neutral lipid like triolein is importantsince it allows a particular type of compartmentation [211].Triolein acts as a hydrophobic space filler at bilayer inter-section points and stabilizes these junctions. Furthermore,triolein is also present as oil droplets dispersed in the encap-sulated aqueous space [211].

3.17. GVs Prepared from an Organic/AqueousTwo-Phase System

Giant vesicles—claimed to be unilamellar [213]—can beprepared rapidly by first dissolving the vesicle-formingamphiphiles in a chloroform-methanol solution in a round-bottomed flask, and then adding carefully along the flaskwalls an aqueous solution (water or buffer) that may alsocontain water-soluble molecules to be entrapped. Afterremoval of the organic solvent in a rotatory evapora-tor under reduced pressure and at elevated temperature(40 �C), a suspension is obtained that contains many giantvesicles with diameters up to 50 �m, which can be removedfor further investigation [213, 214].

3.18. SUVs and LUVs Preparedby the Ethanol Injection Method

This method is a rather simple one that uses ethanol asa cosolvent and does not require homogenization devices[31, 215, 216]. The bilayer-forming amphiphile (e.g., POPCor another PC) is first dissolved at a certain concentra-tion in ethanol (or methanol [217]). A transparent solu-tion is obtained. If a small amount of this ethanolic (ormethanolic) lipid solution is now rapidly added at a temper-ature above the Tm to an aqueous solution, the formationof vesicles is observed. The reason for vesicle formation isthe miscibility of ethanol (or methanol) with water and themigration of alcohol molecules—originally surrounding thelipid molecules—away from the lipids into the bulk solution.Depending on the experimental conditions (e.g., lipid con-centration in the alcohol, speed of adding the alcoholic lipidsolution, final concentration in the aqueous suspension, andstirring rate), the vesicles formed are more or less homoge-neous with respect to size and lamellarity. The most impor-tant factor seems to be the concentration of the lipid in thealcohol injected into the buffer solution.

58 Preparation of Vesicles (Liposomes)

Vesicles prepared from DMPC (at 35 �C) or DPPC (at55 �C), for example, were mainly unilamellar vesicles withdiameters between about 30 nm and about 120 nm if the PCconcentration in the ethanolic solution was varied betweenabout 3 mM and about 40 mM (with an ethanol concen-tration in the buffer after the addition of 2.5–7.5%, v/v)[210]. Similar results were obtained with a mixture of soy-bean phospholipids [218].

In the case of POPC injected as a methanolic solution,POPC concentrations in methanol up to 25 mM result invesicles with diameters between 40 and 70 nm [217].

The ethanol (or methanol) present in the final vesicle sus-pension may be removed almost completely by dialysis ifrequired [216].

One of the limitations of the method is the limitedsolubility of the phospholipids in the alcohol (e.g., 40 mMsoybean PC in ethanol [218]). This necessarily results in rel-atively dilute vesicle suspensions (a few millimolar). Fur-thermore, whereas entrapment yields for ethanol-solublesubstances are high [218], the encapsulation efficiency forwater-soluble compounds is low [218], unless a sophisti-cated cross-flow injection technique is used [219], which alsoallows an upscaling to at least several 100 ml [220].

The ethanol injection method has also been combinedwith high-speed homogenization [221, 222], thereby allow-ing the preparation of uniform vesicles with a diameter of170–200 nm on an industrial scale [222].

In a further modification of the method, an ethanolic lipidsolution is not injected into an aqueous solution, but water ispoured into a concentrated lipid-ethanol solution, followedby the removal of the ethanol in an evaporator and the addi-tion of water [223]. The particular lipid mixture containeda defined amount of soybean PC, cholesterol, �-sitosterol�-d-glucoside, and oleic acid, and the resulting polydispersevesicles had mean diameters between about 150 nm and1.3 �m, depending on the experimental conditions [223].

3.19. ULVs and OLVs Preparedby the Proliposome Method

This method is related to the ethanol injection methoddescribed in Section 3.18 in the sense that ethanol is alsoused as a cosolvent. An initial mixture (called prolipo-some mixture [224]) containing vesicle-forming amphiphiles(egg PCs [225], soybean PCs, or hydrogenated soybean PCs[224]), ethanol (or glycerol or propyleneglycol), and water isconverted into vesicles by a dilution step [224–226]. It is amethod that seems to be particularly applicable to the bulkproduction of lipid vesicles.

The vesicles (liposomes) only form after water additionsince the proliposome mixture does not contain enoughwater to trigger vesicle formation. The proliposome mixtureis probably built up of extended hydrated lipid bilayers thatare separated by an ethanol-rich hydrophilic medium [224].

The encapsulation efficiencies for water-soluble or bilayersoluble compounds are rather high [224, 226]. The vesiclesformed by the proliposome method may be predominantlyunilamellar or oligolamellar vesicles with a broad size distri-bution between 20 nm and about 400 nm [218]. The meansize and lamellarity of the vesicles obtained seem, however,to depend on the actual experimental conditions. In the case

of vesicles formed from a “proliposome mixture” containingegg yolk PC:ethanol:water at a ratio of 100:80:20 (w/w/w),the mean size varied between 100 nm and 1.2 �m, and mostvesicles were oligo- or multilamellar [225].

In a modification of the actual proliposome method, alarge-scale production of lipid vesicles could be achieved bydiluting about 10–20 times—with an aqueous phase usinga dynamic mixing device above Tm of the lipids—a water-miscible solvent mixture composed of N -methylpyrrolidoneand tert-butyl alcohol (1:4, v/v) containing POPC:1,2-dioleoyl-sn-glycero-3-phosphoserine (7:3, mol/mol) and adrug to be entrapped [227]. The resulting mean size of themainly unilamellar vesicles formed after the dilution variedwith the composition of the aqueous phase between about50 and 150 nm [227].

Please note that the term “proliposome” has also beenused for particular dry granular phospholipid preparationswhich, upon dispersion in water, result in the formation ofMLVs [31, 228]. These preparations do not contain ethanolat all.

3.20. Preparation ofMultilamellar Ethosomes

With a third method in which ethanol plays an importantrole, so-called ethosomes can be prepared. Ethosomes arelipid vesicles that contain in the final preparation a consid-erable amount of ethanol. Ethosomes are prepared by firstdissolving a phospholipid (such as soybean PC mixtures) inethanol. Water is then slowly added in a fine stream withconstant mixing to a specially prepared container, followedby an equilibration of the system at 30 �C. The final vesiclepreparation contains 2% (w/w) soybean PCs and 30% (w/w)ethanol and seems to be particularly useful in pharmaceu-tical applications for drug transport across the skin (trans-dermal drug delivery) [229–231]. The vesicles formed aremainly MLVs and are apparently relatively monodisperse,with a mean reported diameter of about 150 nm [229]. Thesize of the vesicles seems to increase with decreasing ethanolconcentration [229]. At 20% (w/w) ethanol, the mean etho-some diameter is around 190 nm; at 45% (w/w) it is around100 nm [229]. The mean vesicle size is also dependent onthe lipid concentration. At 30% (w/w) ethanol, the vesicles’mean diameter varies from about 120 nm to about 250 nmon going from 0.5% (w/w) to 4% (w/w) soybean PCs [229].

In transdermal drug delivery applications, the ethanolpresent in an ethosome preparation may act as a skin per-meation enhancer because of the interaction with the lipidlayers of the skin’s horny layer (stratum corneum), therebyallowing the passage of drugs across the skin.

3.21. GVs Prepared by theInterdigitation-Fusion Method

A fourth method that uses ethanol is based on the factthat under certain conditions certain glycerophospholipidsare known to form bilayers that have interpenetrated (inter-digitated) hydrophobic chains. This means that the methylgroups localized at the end of the hydrophobic chains of amonolayer in a bilayer are in contact with the methylenegroups of the hydrophobic chains of the other monolayer

Preparation of Vesicles (Liposomes) 59

and vice versa. Such interdigitated structures (interpene-trated lamellar sheets) are formed upon the addition ofethanol to SUVs prepared from specific saturated PCs (e.g.,DPPC), at a temperature below Tm. When the temperatureis increased above Tm, the interdigitated lamellar sheets fuseand transform into vesicles that are mainly unilamellar andhave diameters above 1 �m (called IFVs, vesicles preparedby the interdigitated fusion method [232]).

The trapped volume of the IFVs depends on the chemicalstructure of the lipid, the concentration of ethanol used toinduce interdigitation fusion, and the size of the precursorSUVs [232].

3.22. ULVs and MLVs Preparedby the Coacervation Technique

The starting system in this method of vesicle preparationis a mixture of naturally occurring egg yolk phospholipids(including about 81% PCs), an alcohol in which the phos-pholipids are soluble (methanol, ethanol, n-propanol, or2-propanol), and water [233]. Under appropriate conditions,phase separation is observed in this three-component sys-tem, corresponding to a region in the phase diagram thatis related to a so-called coacervation. Coacervate is an oldterm used in colloid chemistry [234]. It refers to a systemin which an amphiphile-rich aqueous phase is in equilibriumwith an amphiphile-poor aqueous phase. It seems that coac-ervates actually correspond to the so-called sponge phase(the L3-phase), a disordered version of the bilayered bicon-tinuous cubic phase [235].

After the initial coacervation system is dialyzed againstwater, vesicles form that seem to be either relatively homo-geneous and unilamellar vesicles (in the case of methanolor propanol) or mainly MLVs (in the case of ethanol), rang-ing in size from about 100 nm to 1 �m, depending on theexperimental conditions used [233].

3.23. Vesicles Prepared by the SupercriticalLiposome Method

With the use of a specially designed, technically rathercomplex apparatus, vesicles with an average size of about200 nm can be prepared by mixing at low pressure anaqueous solution with supercritical CO2 (kept at high pres-sure (25 MPa) and 60 �C) containing the vesicle-formingamphiphile (POPC:cholesterol, 7:3, mol/mol). The meanvesicle sizes vary with the experimental conditions, such asgeometric dimensions of the important parts of the appara-tus [236].

3.24. Vesicles Prepared by the EtherInjection Method

Vesicles can be prepared by slowly injecting (at 0.2 ml/ml)a diethyl ether/phospholipid solution into an aqueous phasethat has been warmed to a temperature (55 �C) above theboiling point of diethyl ether [31, 237]. The diethyl ethervaporizes upon contact with the aqueous phase, and thedispersed lipids form preferentially (but not entirely) uni-lamellar vesicles. These vesicles can then be sized down byextrusion (see Section 3.8) or simple Millipore filtration. In

the latter case, the reported vesicle diameters are in therange of 100 to 300 nm [237].

3.25. OLVs Prepared by the RapidSolvent Exchange Method

The rapid solvent exchange (RSE) method has been specif-ically designed for the preparation of vesicles from phos-pholipid-cholesterol mixtures containing high amounts ofcholesterol [238]. The method is based on a rapid transferof the vesicle-forming lipids from an organic solvent to anaqueous buffer solution in which the vesicles are meant tobe formed. This rapid solvent exchange avoids the transientformation of solid lipid mixtures, which often demix (phaseseparate) and result in inhomogeneous vesicle preparations.

The lipids and the membrane soluble additives (i.e.,cholesterol) are first dissolved in a solvent that is not misci-ble with water (e.g., chloroform or methylene chloride). Thislipid solution is then added to an aqueous solution abovethe Tm of the lipids in a particular manner at reduced pres-sure with a specially designed apparatus in such a way thatthe solvent is rapidly (within 1 min) and almost completelyevaporated, as a result of pressure changes during the injec-tion process.

Vesicles prepared by the RSE strategy from POPC, forexample, are oligolamellar with an expected lamellar repeatdistance of 6.5 nm (see Section 3.2) and an entrapped vol-ume of about 4.5 liters/mol [238].

3.26. Vesicles Prepared from an InitialO/W Emulsion

In this simple method [239], an o/w emulsion is first formedby bath sonication from a vesicle-forming amphiphile (eggyolk PC and cholesterol), an aqueous solution (containingthe water-soluble molecules to be entrapped), and n-decane.This o/w emulsion is then transferred to a second aqueoussolution, which gives two separated phases, an upper organicphase and a lower aqueous phase. The two-phase system iscentrifuged at 3500 × g for 10 min, resulting in the move-ment of the amphiphiles from the organic phase into theaqueous phase and, as a consequence, the formation of vesi-cles in the lower aqueous phase, which is separated from theupper phase. The resulting vesicles, which may contain smallamounts of n-decane in the bilayer, have a diameter in therange of 50–200 nm and are characterized by relatively highencapsulation yields [239].

3.27. ULVs Preparedby the Detergent-Depletion Method

If a bilayer-forming lipid is mixed in an aqueous solu-tion with a micelle-forming surfactant (often called deter-gent, from the Latin word detergere, meaning to wipeoff or to clean) under such conditions that the deter-gent molecules “dominate,” mixed detergent/lipid micellesare formed [240–243]. These aggregates are composedof bilayer-forming amphiphiles as well as micelle-formingamphiphiles and are disc-like, sheet-like, or cylindricalstructures.

60 Preparation of Vesicles (Liposomes)

In the detergent-depletion method (also called thedetergent dialysis or detergent removal method), the startingsystem from which the vesicles are formed is mixed deter-gent/lipid micelles. The micelle-forming detergent molecules(with their large a0; see Section 1.1) are expected to bedistributed in the mixed-micelle aggregate in such a waythat they particularly occupy the highly curved edges ofthe aggregates [31]. The micelle-forming surfactant is alsopresent in a relatively high amount in the bulk phase asnonaggregated, monomeric detergent, at a concentration

hydration,mechanical treatment

at T > Tm

'oil' removalat T > Tm

detergentremovalat T > Tm

solventexchangeat T > Tm

(i)

(ii)w/o-emulsion w/o/w-emulsion

(iii) (iv)

'oil'

aqueous solution

Figure 4. Simplified schematic representation of the principal pathwaysfor the preparation of (normal) lipid vesicles in the case of “conven-tional amphiphiles” (surfactants that do not form a true vesicle phaseat thermodynamic equilibrium). The pathways involve, as starting stateof the amphiphiles, (i) preorganized dry lipids, which are hydratedand (possibly) mechanically manipulated above the Tm of the lipids;(ii) preorganized lipids in w/o emulsions (or w/o mircoemulsions orreversed micelles) or w/o/w emulsions prepared in a volatile solventthat is removed during the vesicle preparation procedure above Tm ofthe lipids; (iii) preorganized lipids in the presence of micelle-formingdetergents (mixed detergent/lipid micelles) existing in dynamic equilib-rium with free detergent monomers that are removed during the vesiclepreparation procedure above the Tm of the lipids; or (iv) lipids dissolvedin a solvent that is miscible with water and is exchanged with water dur-ing the vesicle preparation procedure above the Tm of the lipids. Once avesicle suspension is formed, the mean vesicle size and size distributioncan always be altered by mechanical treatments above Tm.

corresponding in a first approximation to a value a bitlower than the detergent’s cmc, the critical concentration formicelle formation determined separately under comparableconditions.

The amount of monomeric detergent in the mixedmicellar system is important, as it is this nonaggregatedamphiphile that is removed from the solution during thedetergent removal process, which finally leads to the for-mation of vesicles [31, 66, 69, 71, 244–247]. The principleof the detergent-depletion method is the following: mixeddetergent/lipid vesicles, present in rapid equilibrium withdetergent monomers, are put into a dialysis bag or anotherdialysis device [31] at a temperature above Tm of the lipidused [248]. The dialysis membrane is characterized by a per-meability for the monomers, whereas the much larger mixedmicelles cannot pass the membrane. Then, at a temperatureabove Tm [248], the dialysis device is put in contact with abuffer solution in which the mixed micelles were formed.Since the monomers can pass the dialysis membrane, theamount of monomers in the solution inside the dialysisdevice continuously and slowly decreases, and detergentmonomers move from the mixed micellar aggregate into thebulk solution. This process continues until the amount ofdetergents in the micellar aggregates is so low that mixedmicelles can no longer exist, and extended mixed bilayerfragments (sheets) and finally mixed lipid/detergent vesi-cles form. Extensive dialysis leads to the formation of vesi-cles that are almost (but not necessarily completely [249])free from detergent molecules. These vesicle suspensions areoften to a large extent unilamellar and have a narrow sizedistribution. The mean size depends on the experimentalconditions, such as type of detergent used, initial lipid anddetergent concentrations in the mixed micellar solution, andspeed of detergent removal [247, 250–252].

Table 5 lists detergent molecules that are often usedfor the preparation of vesicles by the detergent-depletionmethod, together with characteristic size ranges of themainly unilamellar vesicles formed.

In the case of a system containing egg yolk PC andthe bile salt sodium taurochenodeoxycholate (which aggre-gates itself stepwise into a particular type of unconven-tional, small micelles [253–255]), the mixed micelle-mixedvesicle transformation process—initiated by a rapid dilutionprocess—has been investigated by time-resolved static anddynamic light-scattering measurements [256]. The scatter-ing data analysis indicates that the key kinetic steps duringvesicle formation are the rapid appearance of disc-like inter-mediate micelles, followed by growth of these micelles andclosure of the large discs formed into vesicles [256].

In addition to detergent removal through dialysis, gelpermeation chromatography [256, 257] (which is based onthe partitioning of detergent monomers into the pores of aswollen 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 andoctyl-�-d-glucopyranoside, the originally developed deter-gent-dialysis method has been modified slightly because ofthe relatively high Tm value of these lipids (see Table 4)[248]. The important modification is a slow dilution stepbefore the actual dialysis procedure [248], resulting inmainly unilamellar vesicles with a mean diameter of 98 nm

Preparation of Vesicles (Liposomes) 61

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.

cmc at Method for removing Reported approximateDetergent 25 �C (mM) the detergent vesicle diameters (nm) Refs.

Sodium cholate ∼11 Gel permeation chromatography 30 [257]Dialysis 60 [244]

70 [252]60–80 [245]80–100 [251]50–150 [519]

Sodium glycocholate ∼10 Dilution and dialysis 30–100 [143]Sodium deoxycholate ∼4 Dialysis 150 [252]Sodium chenodeoxycholate ∼5 Dialysis 160 [252]n-Octyl-�-d-glucopyranoside ∼23 Dialysis 180 [245, 524]

230 [525]Bio Beads 300–500 [246]

250 [526]C12EO8 (n-dodecyl ∼0.08–0.09 Bio Beads 60–90 [247]

octaethylenglycolmonoether)

25–80 [527]120 [526]

CHAPS (3-[(3- ∼5–10 Dialysis 380 [519]cholamidopropyl)dimethylammonio]-1-propane sulfonate)

Note: The size of the vesicles may very much depend on the experimental conditions (see text). In particular, the resulting sizes mayalso depend on the presence of bilayer soluble substances (e.g., cholesterol [245, 246]) or in particular cases (cholate) on the presence ofdivalent cations (e.g., Ca2+) [519]. The approximative cmc values given in the table are taken for the bile salts from [520]; for C12E8 from[521, 522]; for CHAPS from [522, 523]; and for n-octyl-�-d-glucopyranoside from [522].

(DMPC) and 94 nm (DPPC) under the corresponding con-ditions used [248].

It is quite generally possible to first simply dilute themixed micellar system, followed by dialysis to completelyremove the detergent [260]. If the detergent is not com-pletely removed, the vesicle preparation by simple dilutionwill always contain detergent molecules, even if the dilutionis 100- or 200-fold [250].

In a particular case [261], SUVs with a mean diameter of23 nm were first prepared by sonication from egg yolk PCat a concentration of 20 mM. The detergent sodium deoxy-cholate was then added to give an aqueous mixture at a ratioof deoxycholate to PC of 1:2 (mol/mol). This mixture con-tained vesicles that were considerably larger than the SUVsused, because of the uptake of the detergent molecules.Deoxycholate was then removed to about 96–98% first bygel filtration and then almost completely by a second gelfiltration. The final preparation—containing less than onedeoxycholate molecule per PC molecule—were dispersedunilamellar vesicles with a mean diameter of 100 nm [261].The same mean vesicle sizes were also obtained if instead ofSUVs 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 ispossible (e.g., [262]), and commercial devices are availableunder the trade names Liposomat and Mini Lipoprep.

The detergent depletion method is the method ofchoice for the reconstitution of water-insoluble membrane-associated proteins, which in a first step are extracted fromthe biological membrane by a mild detergent that does notlead to an irreversible protein denaturation [258, 263]. The

membrane protein-containing mixed detergent/lipid micellesare then converted to membrane protein-containing vesiclesby one of the detergent removal techniques described above.

3.28. (Mixed) Vesicles Preparedby Mixing Bilayer-Formingand Micelle-Forming Amphiphiles

As mentioned in Section 3.27, mixed lipid/detergent vesi-cles form transiently during detergent removal fromdetergent/lipid micelles. Such mixed vesicles can alsobe prepared by adding to preformed lipid vesicles(prepared by any method) above Tm an appropriateamount of a particular detergent [264, 265] or by sim-ply diluting a mixed detergent/lipid micellar solution [266].Examples of detergents that have been used includesodium cholate [267], n-octyl-�-D-glucopyranoside [268],or 3-[(3-cholamidopropyl)dimethylammonio]-1-propane sul-fonate (CHAPS) [269]. The size of the resulting mixedlipid/detergent vesicles depends on the experimental condi-tions, 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 detergentconcentration.

Phosphatidylcholines like POPC, SOPC, DMPC, DPPC,or the mixtures extracted from egg yolk or soybeans are theamphiphiles whose mixed lipid/detergent vesicle (bilayer)formation has been studied most extensively. These PCs allhave two long hydrophobic acyl chains containing 14 to 18or 22 carbon atoms (see Tables 2 and 4).

62 Preparation of Vesicles (Liposomes)

PCs with two short acyl chains, containing fewer than10 carbon atoms, do not form bilayers (vesicles) indilute aqueous solution, but rather micelles (as long asthe concentration is kept above the cmc) [270, 271].While a very short-chain PC (1,2-dibutanoyl-sn-glycero-3-phosphocholine) forms more or less spherical micelles, anincrease in the chain length up to eight carbon atoms leadsto the formation of extended, elongated micellar struc-tures. All of these short-chain PCs are detergents in thesense of the term used here (Section 3.27) [272], and unil-amellar vesicles composed of long-chain PCs and short-chain PCs can easily be prepared, for example, by theaddition of a micellar solution of 1,2-diheptanoyl-sn-glycero-3-phosphocholine (final concentration 5 mM) to a DPPCdispersion (final concentration 20 mM) in an aqueous solu-tion [273]. The resulting vesicles are rather stable and havemean diameters below as well as above 1 �m [274], depend-ing on the experimental conditions, such as the ratio of theshort-chain PC to the long-chain PC and cholesterol content[274].

One particular type of mixed lipid/detergent vesicularsystem (also containing ethanol)—not prepared by deter-gent dialysis, but by simple detergent and lipid mixing—arethe so-called Transfersomes, vesicles that are claimed to beable to transfer water-soluble drugs through the skin to theblood circulation (transdermal drug delivery), particularlythrough the outermost physical barrier of the skin, the hornylayer, called the stratum corneum. Transfersomes seem tobe ultradeformable and may squeeze through the stratumcorneum pores, which have diameters only one-tenth of thesize of the Transfersomes. In this way, encapsulated water-soluble drugs may be transported across the skin [231].Transfersomes are prepared by mixing an ethanolic solutionof soybean PCs with (for example) sodium cholate to yielda suspension typically containing 8.7% (w/w) soybean PCsand 1.3% (w/w) cholate as well as approximately 8.5% (v/v)ethanol [275, 276]. For a reduction in vesicle size to about500 nm, the suspension was treated by sonication, freeze-thaw cycles, and processing with a homogenizer.

In the case of the particular non-ionic industrialamphiphile Brij 30 (n-dodecyl tetraethylene glycol ether,abbreviated as C12E4 or C12EO4, also called Laureth-4), vesicles have been prepared by simple dilution fromso-called hydrotrope solutions, which contained in additionto C12EO4 the micelle-forming surfactant sodium xylene sul-fonate [277]. Through a detailed elaboration and analysisof the three-component phase diagram containing C12EO4,sodium xylene sulfonate, and water, the conditions underwhich vesicles form have first been established. Appropri-ate starting conditions were then chosen, and a correspond-ing dilution resulted finally in the desired vesicular system,which contained not only the vesicle-forming amphiphile,but also the hydrotrope (as constituent of the aqueousphase) [277].

3.29. Vesicles Prepared from Lipidsin Chaotropic Ion Solutions

This method is related to the detergent-depletion methoddescribed in Section 3.27. Chaotropic substances (e.g.,urea, guanidinium hydrochloride, potassium thiocyanate,

trichloroacetate, or trichlorobromate) are known to disturb(break) the structure of water, and lipids are soluble inchaotropic solutions [278]. Solutions of trichloroacetate, forexample, dissolve PCs as micellar solution [279], througha binding of the ions to the head groups, which leads toan increase in a0 (see Section 6). If a SUV suspensionprepared from phospholipids in buffer solution by probesonication is mixed with an aqueous solution containingsodium trichloroacetate, a micellar solution is first obtained(at about 1–3 M sodium trichloroacetate in the case of0.1–0.5 mM phospholipids). This solution is then exten-sively dialyzed against a buffer solution to remove all ofthe chaotropic ions, resulting in the formation of uni- oroligolamellar vesicles with diameters between 10 and 20 �m,although many small vesicles are also present [278]. Iffreeze-thaw cycles are used, the amount of chaotropic ionsneeded to generate giant vesicles is lowered considerably[278].

3.30. GVs Prepared from a W/O Emulsionwith the Help of a Detergent

This method involves the use of nonphospholipidamphiphiles in a four-step procedure [280]. First, a w/oemulsion is formed by homogenization of an aqueoussolution, n-hexane and a mixture of soybean phospholipidsand the non-ionic surfactant Span 80 (sorbitan monooleate).Second, n-hexane is removed by rotatory evaporation fromthe emulsion under reduced pressure, resulting in theformation of a water-in-lipid mixture. Third, this mixture ismixed with an aqueous solution containing a water-solubledetergent (such as sodium dodecyl sulfate, hexadecytrimethy-lammonium bromide, or Tween 80 (POE (20) sorbitanmonooleate, which is a modified sorbitan monooleate thatcontains on average a total of 20 hydrophilic oxyethyleneunits). Fourth, the highly water-soluble detergent is removedby dialysis, leading to the formation of micrometer-sizedvesicles.

4. PREPARATION OFREVERSED VESICLES

There seems to be a kind of symmetry among most self-organized surfactant aggregates with regard to the distri-bution of the hydrophilic and the hydrophobic parts ofthe surfactants in the aggregate [281–283]. There are (nor-mal) micelles (L1-phase) and reversed micelles (L2-phase),a (normal) hexagonal phase (HI� and a reversed hexagonalphase (HII�, and (normal) emulsions (oil-in-water, o/w) andreversed emulsions (water-in-oil, w/o) [235]. In the case ofvesicular aggregates, the existence of reversed vesicles hasbeen demonstrated [283], for example, for a system con-taining the non-ionic surfactant tetraethyleneglycol dodecylether (n-dodecyl tetraethyleneglycol monoether, abbreviatedR12EO4, C12EO4, or C12E4�, the solvent n-dodecane, andwater [281, 282]. These reversed vesicles form first uponthe preparation of a mixture composed of 1 wt% water and99 wt% of a n-dodecane solution which contains 2.5 wt%C12EO4 [281]. After equilibration at 25 �C, a two-phase sys-tem forms that is composed of a lamellar liquid crystalline

Preparation of Vesicles (Liposomes) 63

phase that is in equilibrium with an excess n-dodecanephase. When these two phases are mixed by hand-shaking, aheterogeneous dispersion of mainly multilamellar reversedvesicles forms with diameters of the reversed vesicles fromless than 1 �m to 10–20 �m [281]. Strong vortexing leads toa reduction in size below 200 nm [282]. Without water, noreversed vesicles form.

A reversed vesicle formed in an oil (in a water-immiscible,apolar solvent) is composed of an oily core and one or sev-eral closed “reversed” surfactant bilayer shells. The reversedbilayers are organized in such a way that the hydratedhydrophilic head groups of the amphiphiles are facingtoward the center of the reversed bilayer, while the hydro-carbon chains are in contact with the oil [281].

The particular type of reversed vesicles formed fromR12EO4, water, and n-dodecane seem to be rather unstableand convert back into the thermodynamically stable two-phase system within hours or days [281]. Considerably morestable reversed vesicles can be obtained from a mixture ofthe non-ionic surfactants sucrose monoalkanoate (which is,to at least 95%, a monoester containing 10 wt% tetrade-canoyl, 40 wt% hexadecanoyl, and 50 wt% octadecanoylchains [283]), hexaethyleneglycol hexadecyl ether (abbrevi-ated R16EO6, C16EO6, or C16E6�, n-decane, and water [284].The two surfactants are mixed at a weight ratio of sucrosemonoalkanoate to C16EO6 of 85:15, and vesicle formationis observed at typically 3 wt% surfactant in n-decane with atypical weight ratio of surfactant to water of 6 [283]. Treat-ment of the vesicle suspension with a probe sonicator at30 �C leads to a reduction in the reversed vesicle size toabout 150–300 nm, depending on the experimental condi-tions, such as the water-to-surfactant ratio and the surfactantconcentration [284].

The formation by vortexing of micrometer-sized reversedvesicles (and reversed myelin figures) has also been observedin m-xylene at a surfactant-to-water weight ratio around 0.8for a number of polyethyleneglycol oleoyl ethers (abbrevi-ated R18�1EOx or C18�1EOx, with mean values of x vary-ing between 10 and 55) [285]. Furthermore, it seems thatsome phospholipids can also form reversed vesicles underparticular conditions, as reported in the case of a systemcontaining soybean phosphatidylethanolamine and triolein,saturated with water through vapor equilibration [286]. Theparticles formed—which are most likely reversed vesicles—have diameters in the range of 500 nm to 1.7 �m [286].

5. CHARACTERIZATIONS ANDAPPLICATIONS OF VESICLES

5.1. Characterizations of Vesicles

There are a number of typical properties that characterizea vesicle suspension, particularly the mean vesicle size, thevesicle size distribution, the mean number of lamellae pervesicle, and the chemical and physical stability [287]. All ofthese properties depend on (i) the chemical structure of theamphiphiles (or mixtures of amphiphiles) from which thevesicles are formed; (ii) the solution in which the vesicles areformed (e.g., salt and buffer content in the case of (normal)vesicles and aqueous solutions); and (iii) the method of vesi-cle preparation (see Section 3). A minimal characterization

of a vesicle suspension prepared is always needed. In manycases, a routine size measurement is appropriate, particu-larly by photon correlation spectroscopy (PCS), also calleddynamic light scattering or quasi-elastic light scattering. PCSprovides information about the homogeneity of a vesiclesuspension, and it is relatively straightforward to find outwhether the vesicles in a preparation are relatively monodis-perse or whether they are polydisperse. For monodisperseand spherical vesicles, the mean hydrodynamic diameterof the vesicles can be determined relatively easily. In thecase of polydisperse vesicle preparations and/or nonspher-ical vesicles, the interpretation of the scattering curves ismore delicate, and it is recommended to apply additionalmethods as well, such as electron microscopy. A most proba-ble vesicle size can be obtained by cryotransmission electronmicroscopy (cryo-TEM), although this method is limited tosizes below about 250 nm, and the sizes and morphologiesobtained have to be taken with caution [288–290]. The useof different independent methods is always recommended.

Table 6 lists a number of methods that are used for char-acterizing vesicle suspensions. In many cases, fluorescentlylabeled amphiphiles are used in the vesicle membranes ata concentration of 1–2 mol%. These amphiphiles containfluorescent groups, and the behavior of these molecules isthen taken as a measure for the behavior of the bulk vesicle-forming amphiphiles, for example, for the determination ofthe lateral diffusion coefficient of the amphiphiles. Since thefluorescent groups are sometimes rather bulky, the kineticconstants obtained with these labeled molecules may be dif-ferent from the actual membrane-forming lipids [291, 292].The same is true for nitroxide-labeled amphiphiles usedfor ESR (electron spin resonance) measurements [292] (seeTable 6).

For the detection of the internal aqueous volume in (nor-mal) vesicles—usually expressed as microliter of trappedaqueous solution per micromole of amphiphile (�l/�mol)—dye molecules are often used, which can be easily quantifiedspectroscopically. Meaningful results, however, can only beobtained if the dye molecules do not interact with the vesi-cle membrane (no incorporation inside the membrane, noadsorption to the membrane).

In the case of permeability measurements, the use of dyemolecules is convenient, and the release of vesicle-trappedmolecules into the external medium can easily be quanti-fied. In this case, however, one should always be aware thatthe permeability measured is the permeability of the partic-ular dye molecules used under the particular experimentalconditions. In this respect, NMR methods that are basedon the use of paramagnetic shift reagents are better. Theexternal addition of the shift reagents allows a distinctionbetween the particular solute molecules present inside thevesicles and the solute molecules present outside of the vesi-cles, presuming that the shift reagent does not permeate themembranes. This has first to be proved. The permeabilitymay very much depend on the experimental conditions, suchas the presence of certain buffer species [293].

The list in Table 6 is certainly not complete. However, itgives a few hints about the general principles and problemsof the vesicle characterization and lists some of the meth-ods and techniques used. Depending on the vesicle prepara-tion, 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 Methods used and comments Ref.

Mean vesicle sizeand size distribution

PCSAnalysis 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 non-spherical vesicles (e.g., caused by an osmotic imbalance between the inside and theoutside of the vesicles), the size analysis is rather complex and difficult

[2, 31, 528–531]

SLSAverage 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]

FFEMReplicas 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-TEMThe 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—torepresent 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. Onlyapplicable for vesicles with sizes below ∼250–500 nm. Not free of “artefacts” (“arti-facts”), i.e., formation of microstructures due to specimen preparation, electron optics,or radiolytic effects [288–290]

[2, 3, 536]

LMOnly for GVs

[4, 66, 89, 91, 536]

Size exclusion chromatographyConventional and high-performance liquid chromatography (HPLC) separation based on

the principle that the partitioning behavior of vesicles in the pores of a solid columnmatrix depends on the size of the vesicles

[66, 386, 537]

NMRSize determinations possible from 31P-NMR spectra of phospholipid vesicles that have

sizes below 1 �m with the help of a comparison with simulated spectra. For spectrasimulation, the dynamics of the vesicles and the phospholipids in the vesicle membranesare considered

[136]

AFMThe vesicles have to be deposited on a solid surface, which may lead to vesicle deformations

and vesicle aggregations

[538, 539]

SANSMeasurements in D2O that may influence the vesicle size via altered head group interac-

tions. Relative complex analysis involving fitting of experimental data

[540, 541]

Bilayer thickness X-rayBased on the interaction of X-rays with the electrons of the amphiphiles in the vesicles

[80, 300, 542]

Cryo-TEM [543]

SANSMeasurements in D2O. 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]

Vesicle shapeand morphology

LMOnly for GVs. Detection of MVVs

[547–552]

FFEMDetection of MVVs

Negative staining EMHeavy 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

Preparation of Vesicles (Liposomes) 65

Table 6. Continued

Property Methods used and comments Ref.

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

[553]

Lamellarity Cryo-TEMOnly for vesicles with sizes below ∼250–500 nm

[3, 293]

Negative Staining EMAddition of heavy metals (osmium tetroxide) needed. The number of lamellae detected

may not represent the “true” number of lamellae

[12, 31]

NMR31P-NMR in the case of phospholipid vesicles. Externally added membrane-

impermeable Mn2+ influences the 31P-NMR signal of the phospholipids in the out-ermost monolayers by broadening the resonance beyond detection. The resonancesof the inner phospholipids are unaffected. Paramagnetic shift reagents may also beused (e.g., Pr3+ and Eu3+)

[161, 162, 293, 554]

Fluorescence quenchingAn appropriate amphiphilic fluorescent probe molecule is used in the vesicle prepara-

tion, and the fluorescence of the amphiphiles in the outermost layers is quenchedby adding to the vesicles a membrane-impermeable reagent

[555]

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

[556]

Chemical modificationFunctional 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 chem-ically modified by externally added reagents that do not permeate the membranesand 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. Thechemically modified amphiphiles are then quantified

[31]

Lipid domains in the mem-branes (rafts)

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

[557]

NMR31P-, 13C-, or 2H-NMR

[558, 559]

Confocal fluorescence microscopyDistinction 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]

Phase transition tempera-ture (Tm)

DSCA vesicle sample and an inert reference are heated independently to maintain an iden-

tical temperature in each. In endothermic solid-ordered/liquid-ordered phase tran-sitions, heat is required in excess in the vesicle sample over the heat required tomaintain the same temperature in the reference

[43, 300, 540]

FluorescenceFluorescence polarization measurements of a membrane soluble fluorescent probe

(e.g., 1,6-diphenylhexatriene, DPH)

[31, 561, 562]

CDInduced circular dichroism below the Tm of an achiral probe (1,6-diphenylhexatriene,

DPH) embedded inside the vesicle membrane constituted by chiral amphiphiles. NoCD signal above Tm

[563]

Mobility of theamphiphiles in the vesi-cle membranes

NMR1H-, 2H-, 13C-, 31P-NMR

[37, 564–567]

ESRUse 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.

FluorescenceUse of fluorescent amphiphiles

[571]

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

[560]

Fluorescence recovery after photobleachingLateral diffusion of fluorescently labeled amphiphiles determined by first bleaching

the lipids with an intense laser pulse, and then analyzing the fluorescence recoverykinetics in the bleached area, due to the diffusion of unbleached molecules into thepreviously bleached area

[37]

Molecular conformation of theamphiphiles in the vesiclemembranes

NMR13C- or 2H-NMR using labeled phospholipids

[300, 559, 564, 565, 572]

FT-IR and RamanDetermination of the equilibrium conformational characteristics of the amphiphiles

[37, 300, 573]

CDFor vesicles composed of chiral amphiphiles

[563]

Membrane fluidity and order FluorescenceUse of fluorescent membrane probes. Measurements of static or time-resolved fluores-

cence anisotropy

[37]

NMRUse of partially and specifically deuterated amphiphiles

[300, 564]

ESRUse 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(flip-flop)

ESRUse of spin-labeled amphiphiles (nitroxide radical)

[292, 574]

FluorescenceUse of fluorescently labeled amphiphiles

[292]

Surface charge Zeta potential measurements, microelectrophoresisProblematic, 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 backgroundsalt conditions (usually NaCl)

Internal volume and entrap-ment (encapsulation) yield

Use of dye moleculesThe 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 afterseparation of the nonentrapped molecules from the vesicles or by addition to thevesicles of a membrane-impermeable reagent, which leads to a complete quenchingof 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]

[31, 66, 575–577]

NMR17O-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 17O resonancecorresponding to the external water. The internal water peak remains the same.Above Tm, only one peak is observed because of rapid water equilibration

[554]

Membrane permeability Use of dyes or radioactively labeled moleculesDetermination of the release of dye molecules entrapped inside the aqueous interior

of the vesicles as a function of time under particular storage conditions

[31, 578]

NMRUse of 1H- or 17O-NMR and an externally added membrane-impermeable param-

agnetic shift reagent (Mn2+, Pr3+) to distinguish between internal and externalpermeants

[579–582]

continued

Preparation of Vesicles (Liposomes) 67

Table 6. Continued

Property Methods used and comments Ref.

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

[583]

Chemical stability Thin-layer chromatographyAnalysis of possible degradation product due to hydrolysis and oxidation reactions

[72]

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

[72]

Physical stability Turbidity, PCS, FFEM, Cryo-TEM, and othersAnalysis 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, differentialscanning calorimetry; EM, electron microscopy; ESR, electron spin resonance (also called electron paramagnetic resonance, EPR); FFEM, freeze-fracture electronmicroscopy; FT-IR, Fourier transformed infrared; LM, light microscopy; NMR, nuclear magnetic resonance; PCS, photon correlation spectroscopy (also called dynamiclight 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., 31P-NMR obviously cannot be used for the determina-tion of the lamellarity of vesicles that are not composed ofphosphorous-containing amphiphiles).

5.2. Applications of Vesicles

Lipid vesicles are used successfully in many different fieldsas interesting and versatile submicrometer- or micrometer-sized compartment systems [69, 294–297]. This wide appli-cability of vesicles and the broad interest in vesicles can beunderstood at least on the basis of the following four rea-sons:

(i) Lipid vesicles can be considered membrane orbiomimetic systems [298, 299], since the molec-ular arrangement of conventional vesicle-formingamphiphiles in a vesicle is a (more or less curved)bilayer, like the lipid matrix in biological cell mem-branes [97, 300] or in the outer coat of certain viruses[97].

(ii) Vesicles prepared from amphilphiles present in bio-logical systems allow applications as biocompatibleand biodegradable systems.

(iii) Water-soluble as well as certain water-insolublemolecules can be entrapped inside the aqueous orhydrophobic domains of the vesicles, allowing theuse of vesicles as carrier systems and nanometer- ormicrometer-sized reaction compartments.

(iv) Vesicles can be prepared not only from the conven-tional PC type of bilayer-forming amphiphiles, butalso from a large number of different nonphospho-lipid surfactants (or mixtures of surfactants), allowingthe preparation of application-tailored and specifi-cally designed and functionalized systems.

Among the more than 30,000 known surface active com-pounds [301], a large number of surfactants and surfactantmixtures (including many nonphospholipid amphiphiles)have been reported to form vesicles. The basic principlesthat lead to the formation of vesicles are the same for all,that is, the requirements of (i) an effective packing param-eter p (=v/a0l�eff ≈ 1 (see Section 1.1), (ii) chain flexibility(T > Tm), and (iii) sufficiently low amphiphile concentration

(global packing constraints) [7]. Examples of vesicle-formingamphiphiles include

• cationic di-n-alkyldimethylammonium ions [302, 303](such as di-n-dodecyldimethylammonium bromide(DDAB) [304] and other di-n-dodecyldimethylammo-nium halides [305] and di-n-octadecyldimethylammo-nium chloride [306–308] or bromide [308, 309] or othercounter-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-do-decylphosphate [317], di-n-hexadecylphosphate [318–322], and various di-polyprenylphosphates [323, 324]);

• anionic linear and branched monoalkylphosphates(such as different polyprenylphosphates [325], n-dodecylphosphate [326], 6-propylnonylphosphate [327],4-butyloctylphosphate [327], and 2-pentylheptyl-phosphate [327]);

• anionic tridecyl-6-benzene sulfonate in the presence ofsalt (sodium chloride) [328];

• anionic fatty acid/soap mixtures (e.g., n-octanoic acid/n-octanoate [329], n-decanoic acid/n-decanoate [329–331], oleic acid/oleate [332–334]);

• anionic surfactant/alcohol mixtures (such as sodiumn-dodecylsulfate/n-dodecanol [329, 335] and sodiumn-decanoate/n-decanol [336], or sodium oleate/n-octanol [337], which can form a highly viscous phase ofdensely packed vesicles);

• mixtures of cationic and anionic surfactants, so-calledcatanionic mixtures (such as n-hexadecyltrimethyl-ammonium tosylate/sodium n-dodecylbenzenesulfonate[20, 59], n-hexadecyltrimethylammonium bromide/sodium n-octylsulfate [338–342], and sodium n-dode-cylsulfate and di-n-dodecyldimethylammonium bro-mide [343, 344]);

• ganglioside GM3 [345, 346];• phosphatidylnucleosides (such as 1-palmitoyl-2-oleoyl-

sn-glycero-3-phosphocytidine [347, 348]);• diblock copolymers [24] (such as the ethylene oxide

(EO)/butyleneoxide (BO)diblockcopolymersEO6BO11,EO7BO12, EO11BO11, EO14BO10, and EO19BO11 [349]);

68 Preparation of Vesicles (Liposomes)

• triblock copolymers (such as the ethylene oxide(EO)/propylene oxide(PO)/ethylene oxide (EO) tri-block copolymer EO5PO68EO5 (called Pluronic L121)[350] and the polymerizable poly(2-methyloxazoline)/poly(dimethylsiloxane)/ poly(2-ethyloxazoline) triblockcopolymers [351]);

• polymerizable amphiphiles [33, 352, 353];• perfluorated surfactants [354] (such as short-chain per-

fluorophosphocholines [355] and perfluoroalkyl PCs[356]);

• bolaamphiphiles, which are membrane-spanning amphi-philes with two hydrophilic head groups at the two endsof the molecule and actually form vesicles containingnot bilayered but monolayered shells [357–360];

• gemini surfactants that contain two hydrophobic tailsand two hydrophilic head groups linked together witha short linker [361, 362];

• industrial, not very well defined surfactant/cosurfactantmixtures (such as N -methyl-N -alkanoyl-glucamine/octanol or oleic acid mixtures [363]);

• calixarene-containing [364] or cryptand-containingamphiphiles [365];

• fullerene-containing amphiphiles with two hydrophobicchains and two charged head groups [366];

• double-chain amphiphiles with a polar alkoxysilyl headgroup, allowing the preparation of a kind of organic-inorganic hybrid vesicle, called cerasomes [367];

• triple-chain amphiphiles containing three hydrophobicchains and two charged polar head groups [365, 368];

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 moretypes of amphiphiles may contain vesicles of another typeof amphiphile), based on principles that include the specificmolecular recognition between different types of vesicles ini-tially prepared by one of the established methods describedin Section 3 [370–372].

It is rather obvious that one consequence of the fact thatvesicle formation is observed from so many different classesof amphiphiles (or mixtures of amphiphiles) is the verybroad range of applications in very different fields. Vesiclesare applied—or investigated for potential applications—atleast in the following areas:

• 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, asimmunoadjuvants, and as diagnostics;

• in immunoassays [382–385];• in chromatographic separations using immobilized vesi-

cles [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 reconstitutionand use of membrane-soluble proteins [4, 69, 258, 263,392];

• in research on membrane-soluble ion channels [393–399];

• in research related to the question of the origin andevolution of life, as models for the precursor structuresof the first cells [400];

• in research aimed at constructing artificial (or minimal)cells [401, 402], for example, for potential biotechno-logical applications [402];

• in food technology and nutrition as carrier systemsfor food additives and ingredients and for the controlof certain food processes (e.g., cheese ripening) [403–405];

• in agrochemistry [406];• as nanometer- or micrometer-sized bioreactors contain-

ing 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 functional-ities (e.g., as catalysts for the unfolding [421] or folding[387, 422] of proteins);

• in biosensor developments [423–425], particularly forthe controlled preparation of bilayers adsorbed to solidsurfaces [426–429];

• in the extraction of heavy metal ions with the help offunctionalized, 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 thepreparation of hollow polymer capsules [356, 439–441];

• as templates for modifying the distribution of reactionproducts, for example, in reactions that lead to prod-ucts that are only sparingly soluble in the absence, butsoluble in the presence, of vesicles [442, 443];

• as a medium for the preparation of size-defined poly-mer particles from monomers that are soluble in thevesicle shell [444];

• as supramolecular, self-assembly-based devices [33,298, 299, 358, 445], for example, for the conversion oflight energy into chemical energy (artificial photosyn-thesis) [446, 447], for signaling and switching [448, 449],for the construction of molecular wires [450], and for anumber of different redox processes [418];

and many more (e.g., [295, 299]).One illustration of vesicles loaded with water-soluble

molecules is shown in Figure 5. A cryo-transmission elec-tron micrograph of POPC vesicles containing the proteinferritin is shown. The vesicles were prepared in the pres-ence of ferritin, and the nonentrapped protein moleculeswere separated from the vesicles by size exclusion chro-matography after vesicle preparation [451, 452]. Since thisparticular protein has a dense iron core with a size of about8 nm, it is visible by electron microscopy, and the actualnumber of protein molecules per individual vesicle can bedirectly counted. This type of loaded vesicles can be used,for example, in basic studies on vesicle transformation pro-cesses [451, 452]. If the vesicles contain catalytically activeproteins (enzymes), they may be used in drug delivery or assmall bioreactors [75, 410, 411].

Preparation of Vesicles (Liposomes) 69

Figure 5. Cryo-transmission electron micrograph of two unilamel-lar vesicles that have been prepared from POPC by the reversed-evaporation technique (see Section 3.14) loaded with the iron storageprotein ferritin, followed by extrusion (REV-VET100). Nonentrappedferritin molecules were removed by size-exclusion chromatography.Each black spot inside the vesicles represents one iron core (the coreof an individual protein). The length of the bar corresponds to 100 nm.The electron micrograph was taken by M. Müller and N. Berclaz, Ser-vice Laboratory for Electron Microscopy, at the Department of Biologyat the ETH Zürich. See [451, 452] for experimental details and for anapplication of ferritin-containing vesicles in the investigation of vesicletransformation processes.

Conceptually, an interesting principle of vesicle applica-tions is vesicle transformation (at the site of application) toanother type of surfactant assembly as a result of tempera-ture changes that lead to changes in the surfactant packingparameter p (see Section 1.1). One particular example isthe 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 willincrease with increasing molecular complexity of the vesicu-lar systems. Vesicles—and other surfactant assemblies (e.g.,hexagonal and cubic phases)—will be applied more andmore in all fields related to nanoscience and nanotechnol-ogy. The title of a recent publication in the field of vesi-cle drug delivery is an example of one of the directionsin which vesicle research and application may go: “Biotiny-lated Stealth Magnetoliposomes” [457], a particular vesiclepreparation that combines steric stabilization (Stealth) withmolecular recognition (biotin) and magnetic nanoparticleproperties. In other words, it is likely that the complexityof the vesicles prepared and investigated will increase tomake them functional, possibly by combining the principlesof supramolecular chemistry and surfactant self-assembly, toprepare nanometer- or micrometer-sized synthetic systemsthat may carry out some of the functions biological systemsdo rather efficiently [458, 459].

6. CONCLUDING REMARKSThere are many different methodologies that have beendescribed in the literature for the preparation of lipidvesicles. Some (but not all), and certainly the most promi-nent and well known, are mentioned in this review, focusingon normal lipid vesicles (Section 3), although reversed vesi-cles are briefly mentioned as well (Section 4).

As outlined in Sections 1 and 2, the use of a numberof different terms in the field of vesicles (and surfactantassemblies at large) is often rather confusing, and confusioneven exists over the general use of the term “surfactant.”It is worth pointing out that a reduction in the surface ten-sion of water by surfactants can be achieved not only withclassical micelle-forming, single-chain amphiphiles (such assodium n-dodecyl sulfate, SDS) [460], but also with typicalvesicle-forming double-chain phospholipids (such as SOPCor DPPC). Certainly, the kinetics and extent of adsorption ofa particular surfactant at the water-air interface very muchdepend on the precise experimental conditions [461]. In thecase of DPPC, for example, the equilibrium surface tensionand rate of monolayer formation at the water-air interfacedepend on the temperature [462]. Below Tm, in the solid-ordered state of the molecules (see Section 2), the decreasein the surface tension of water by DPPC is considerablylower than that above Tm [462]. However, there is no doubtthat glycerophospholipids like SOPC, POPC, or DPPC aresurfactants in the sense of the definition of this term givenin Section 1.1.

Since in dilute aqueous solution many of the bilayer-forming surfactants known (at least the most studied phos-pholipids) at thermodynamic equilibrium do not form a truevesicle phase (with a defined vesicle size, size distribution,and lamellarity), but under the appropriate conditions ratherform a lamellar phase (L· �, L· �, L· ′� or P′

�) with extendedstacked bilayers, the preparation of vesicle dispersions nec-essarily involves the use of a particular method, a particulartechnology.

Depending on the experimental conditions, as mentionedin Section 3, depending on the chemical structure of theamphiphile or mixtures of amphiphiles used, the amphiphileconcentration, the substances that may be encapsulated inthe vesicle’s aqueous interior or in the vesicle membrane,etc., there may be one particular method that is more advan-tageous in comparison with others.

Some of the methods described in Section 3 involve theuse of organic solvents. This may be a problem for cer-tain applications. Some methods can be scaled up to a bulkpreparation [287], some methods involve mechanical treat-ments that may cause an inactivation of sensitive moleculesone may like to entrap [75]; some methods are cheaper thanothers; 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 arerelatively time consuming, etc.

There is no “best method.” The choice of a certainmethod that may be useful very much depends on the par-ticular problem one is trying to solve [67]. There are, at leastin the case of the conventional type of phospholipids (or inthe case of lipid mixtures containing phospholipids, partic-ularly PC), a few general findings that are often valid; butone should be aware of the existence of possible exceptions:

• The hydration of a dried thin lipid film above the Tm ofthe lipids usually leads to the formation of MLVs if thefilm is dispersed vigorously (vortexing or hand-shaking)(see Section 3.2).

• The hydration of a dried thin lipid film above the Tmof the lipids usually leads to the formation of GUVs if

70 Preparation of Vesicles (Liposomes)

the film is undisturbed while being hydrated slowly (seeSection 3.2).

• The electroformation method can be used for thepreparation of GUVs with diameters between about 10and 50 �m (or more) in the case of certain lipids (andlipid mixtures) and buffers with low ionic strength (Sec-tion 3.3).

• Relatively homogeneous preparations of SUVs withdiameters around 50 nm or below can often beobtained by probe sonication (Section 3.5), althoughdegradation of the surfactants and consequences of thepossibly present metal particles (which may be a sourceof nucleation for vesicle transformation processes) haveto be taken into account.

• Rather homogeneous preparations of LUVs with diam-eters around 50 or 100 nm can be obtained by theextrusion technique as FAT-VET50 or FAT-VET100 (Sec-tion 3.8, Fig. 3).

• The dehydration-rehydration method usually results inhigh encapsulation yields (Section 3.7).

• Freeze-thaw cycles may lead to vesicle size homoge-nization and solute equilibration between the externalbulk solution and the trapped aqueous solution (Sec-tion 3.6).

• The detergent-depletion method often results in ratheruniform vesicles with sizes below 100 nm, although thepossibility of an incomplete removal of the detergentshould be considered (Section 3.27, Table 5).

• The use of volatile cosolvents (oils) during vesiclepreparation is often based on the principle that eitherw/o or w/o/w emulsions are formed from which the sol-vent is removed, limiting the solvents to all those thathave a boiling point considerably lower than the boilingpoint of water (Sections 3.14–3.17, 3.24, and 3.25).

The physicochemical properties (such as chemical andphysical stability, membrane permeability) of the vesiclesformed very much depend on the surfactant (or surfactantmixtures) used. Depending on the surfactant, the charac-teristic properties may be very different from those of theconventional PC type of vesicles. Furthermore, the wholeequilibrium phase behavior may be different, and cases areknown where even true vesicle phases seem to form “spon-taneously” [7, 463–465] and seem to exist at thermodynamicequilibrium [7, 59, 60, 463]. In these cases, vesicles of a cer-tain size and lamellarity just form by mixing, independentof the method of preparation. The stability of the vesiclesis understood in the case of the so-called catanionic vesicles(which are composed of a mixture of positively and neg-atively charged amphiphiles) on the basis of a most likelyuneven distribution of the amphiphiles in the two bilayers.This allows the required differences in curvature and molec-ular packing in the two halves of a bilayer [60, 466], simi-lar to the case of lipid vesicles prepared from mixtures oflong-chain and short-chain PCs (Section 3.28) [273, 274].With respect to the spontaneity in the formation of this typeof vesicles, it has been argued, however, that shear forcespresent during the preparation (mixing of solutions) play animportant role in vesicle formation [467].

An interesting case of unilamellar vesicles as thermo-dynamic equilibrium state has been described in the case

of certain ionized phospholipids (e.g., 1,2-dimyristoyl-sn-glycero-3-phosphoglycerol, DMPG) [468–472]. Under theexperimental conditions used, unilamellar DMPG vesiclesapparently only form at the critical temperature (T ∗) of31.6 �C [471], which is different from Tm. Above T ∗ MLVsare formed; below T ∗ the phospholipids arrange into asponge phase [471]. Further investigations are needed tofully understand this critical phenomenon and to clarifywhether this unilamellar vesicle formation is a particularcase or whether it can be more generally observed.

With the exception of a few cases—such as the detergentdepletion method (Section 3.27) (e.g., [256])—the generalmechanism of vesicle formation is not yet completely under-stood in its details, although general principles have beenelaborated [7, 69, 473–477].

From a more practical point of view, and by looking at allof the methods described in Section 3, it is evident that thevesicles often form from a preorganized state of the lipids.This preorganization may be

(i) Lamellar sheets present in a dry film deposited on asolid surface (Sections 3.2 and 3.3) or in the ethanolicpro-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-depletionmethod (Section 3.27) or micelles in the case ofchaotropic ion solutions (Section 3.29).

In a few cases, the vesicle formation is triggered as aresult of a direct contact of a nonorganized state of thelipids with an aqueous environment, as in the case of theNovasome technology (Section 3.10), the ethanol injectionmethod (Section 3.18), or the ether injection method (Sec-tion 3.24). A simple summarizing schematic representationof some of the different pathways for the formation of (nor-mal) lipid vesicles is given in Figure 4. It is expected thatmore methods for the preparation of vesicles will be devel-oped, although the general pathways may remain the same.The range of pre-organized starting systems from which vesi-cles can be formed may be expanded, and more will probablysoon be understood for the vesicle formation at large.

(Normal) vesicles are just a particular state of aggre-gation of surfactants (or surfactant mixtures) in an aque-ous solution that are topologically closed with an internalaqueous space. Vesicles can be unilamellar or multilamellar,and the general principles of surfactant assembly have beenoutlined [7–11]. As mentioned above, vesicles sometimesappear 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 (includ-ing interaggregate interactions), and (iii) flexibility of thehydrophobic chain(s). All of the methods described essen-tially depend on satisfying these criteria.

Physicochemical conditions of inside and outside of vesi-cles often differ greatly. This is a consequence of the closedtopology. This fact, together with adverse packing condi-tions, can often result in a stable state of so-called supra-self-assembly (e.g., surfactant micelles existing inside surfactantvesicles) [478–480].

Preparation of Vesicles (Liposomes) 71

The basic principles that underlie self-assembly of vesiclesare quite general and are firmly based in thermodynamicsand statistical mechanics [7, 9, 10, 481]. The requirementsfor “designing” a vesicle are conceptually simple: a packingparameter close to unity (which means effectively a double-chain surfactant or mixed single-chain surfactants), flexiblehydrophobic chains (a temperature above Tm), and controlof inter- and intraaggregate interactions. Local and globalpackings are the key principles to consider.

Vesicles, however, are a very small part of a much largerclass of self-assembled surfactant aggregates that includecubic phases, which are usually bicontinuous structures ofzero average curvature. “Bicontinuous” means that both theaqueous and “oily” parts of the structure are continuouslyconnected over the whole system. Cubic phases or bicontinu-ous structures in general are ubiquitous in biology for direct-ing biochemical traffic [11]. Likewise, hexagonal phases andmicrotubules are close to lamellar (and vesicular) phases ina phase diagram [11].

Although these things are known and are even beginningto be understood, quantitative predictions remain a prob-lem. This can be traced to the fact that the underlying the-ory of molecular forces that underpins physical chemistryand colloid and surface science is flawed [482, 483]. Previ-ous theories cannot deal with specific ion effects (so-calledHofmeister series), dissolved gas, and other solutes thatchange the water structure. There is rapid development inthis area at the moment, which is likely to provide pre-dictability in vesicle design [482–484], which is certainly whatone is aiming for.

GLOSSARYAmphiphile A molecule that comprises at least two oppos-ing parts, a solvophilic (for example “hydrophilic”, meaningwater-loving) and a solvophobic (for example “hydrophobic”,meaning water-hating). Amphiphiles are surfactants.Detergent A surfactant that in dilute aqueous solutionforms micelles, spherical or non-spherical aggregates thatcontain in the interior of the aggregate the hydrophobic partof the surfactant and on the surface the hydrophilic part ofthe surfactant.Glycerophospholipid A particular phospholipid that con-tains a glycerol backbone to which a phosphate group isbound.Lamellar phase, L� A particular liquid crystalline equi-librium state of surfactant molecules, also called liquid-disordered state. In an aqueous environment, the surfactantmolecules are arranged in layers in which the hydrophobicpart of a surfactant is in the interior of the layer and thehydrophilic part on the two surfaces of the layer, exposed tothe aqueous environment.Liposome Vesicle prepared from amphiphilic lipids.Phase transition temperature, Tm Characteristic tempera-ture (also called lamellar chain melting temperature) ofsurfactants that form a lamellar phase. Above the phasetransition temperature, the surfactant molecules are in aliquid-disordered state, below in a solid-ordered state.

Phosphatidylcholine, PC A particular glycerophospholipidthat contains a choline group in the hydrophilic part, boundto the phosphate.Phospholipid A surfactant that is present in some of thebiological membranes and contains at least one phosphategroup.Reversed vesicle Inverted vesicle formed in a water-immiscible, apolar solvent in the presence of a small amountof water.Surfactant A molecule that is surface active, meaning thatit accumulates at the surface of liquids or solids. Surfactantsare amphiphiles.Vesicle General term to describe any type of hollow, sur-factant-based aggregate composed of one or more shells. Inthe biological literature, the term vesicle is used for a par-ticular small, membrane-bounded, spherical organelle in thecytoplasm of an eukaryotic cell.

ACKNOWLEDGMENTSThe author thanks Barry W. Ninham (Department ofApplied Mathematics, Australian National University, Can-berra, Australia) for extensive discussions on the princi-ples of surfactant self-assembly and for critically readingthe manuscript. Manuscript reading and literature adviceby Pasquale Stano (Department of Materials Science, ETH,Zürich, Switzerland) are also acknowledged. Furthermore,discussions with Saša Svetina (Institute of Biophysics, Uni-versity 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 ColloidScience, Wageningen University, Wageningen, The Nether-lands) are appreciated.

REFERENCES1. H. W. Meyer and W. Richter, Micron 32, 615 (2001).2. S. U. Egelhaaf, E. Wehrli, M. Müller, M. Adrian, and P. Schurten-

berger, 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, Amster-dam, 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 Preparation of Vesicles (Liposomes)

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 Lipo-

somes” (D. D. Lasic and Y. Barenholz, Eds.), Vol. I, p. 1. CRCPress, 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, “Introduc-tion to Organic Chemistry,” 4th ed. Macmillian, New York, 1992.

30. A. Wendel, in “Kirk-Othmer Concise Encyclopedia of ChemicalTechnology,” Vol. 15, 4th ed., p. 192. Wiley, New York, 1995.

31. R. R. C. New, Ed., “Liposomes: A Practical Approach.” OxfordUniversity Press, New York, 1990.

32. S. Hassan, W. Rowe, and G. J. T. Tiddy, in “Handbook of AppliedSurface 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, Bio-phys. 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. Lipids78, 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 Ther-

apeutic Applications” (C. G. Knight, Ed.), p. 105. Elsevier/North-Holland Biomedical Press, Amsterdam, 1981.

42. R. N. A. H. Lewis and R. N. McElhaney, in “The Structure ofBiological Membranes” (P. Yeagle, Ed.), p. 73. CRC Press, BocaRaton, 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 Struc-ture 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).

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. Acta817, 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 Principlesand Models.” Wiley, New York, 1987.

52. P. Machy and L. Leserman, “Liposomes in Cell Biology and Phar-macology.” John Libbey Eurotext Ltd./INSERM, London, 1987.

53. D. Lichtenberg, E. Freire, C. F. Schmidt, Y. Barenholz, P. L. Fel-gner, 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. Crom-melin, 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 DrugDelivery 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 DeliverySystems” (J. Kreuter, Ed.), p. 73. Dekker, New York, 1994.

73. Y. Barenholz and D. J. A. Crommelin, in “Encyclopedia of Phar-maceutical 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, Biochem-

istry 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) 73

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 ScienceSeries (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. Ver-

lagsgesellschaft, Leipzig, 1911.97. B. Alberts, A. Johnson, J. Lewis, M. Raff, K. Roberts, and P. Wal-

ter, “Molecular Biology of the Cell,” 4th ed. Garland Science, NewYork, 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 Supramolecu-

lar 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, Bio-phys. J. 78, 830 (2000).

107. M. I. Angelova, S. Soléau, P. Méléard, J.-F. Faucon, and P. Both-orel, 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 Vesi-

cles,” Perspectives in Supramolecular Chemistry (P. L. Luisi andP. 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. Lipids105, 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, Nature256, 584 (1975).

118. R. Lawaczek, M. Kianosho, and S. I. Chan, Biochim. Biophys. Acta443, 313 (1976).

119. C. Huang, Biochemistry 8, 344 (1969).

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. Com-

mun. 296, 1352 (2002).130. M. Wong, F. H. Anthony, T. W. Tillack, and T. E. Thompson, Bio-

chemistry 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. CRCPress, 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 Technol-

ogy” (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. Gre-

goriadis, Ed.), Vol. 1, 2nd ed., p. 229. CRC Press, Boca Raton, FL,1993.

74 Preparation of Vesicles (Liposomes)

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. CRCPress, 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. Acta986, 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. Papahad-jopoulos, Biochim. Biophys. Acta 557, 9 (1979).

169. F. Szoka, F. Olson, T. Heath, W. Vail, E. Mayhew, and D. Papa-hadjopoulos, 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, andR. J. Havel. J. Lipid Res. 21, 981 (1980).

178. E. Mayhew, R. Lazo, W. J. Vail, and A. M. Green, Biochim. Bio-phys. 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, BocaRaton, 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).

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, BocaRaton, 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. USA75, 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, Biochem-

istry 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, andNovel 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) 75

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. Wiley-VCH, New York, 1999.

236. L. Frederiksen, K. Anton, P. van Hoogevest, H. R. Keller, andH. 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, BocaRaton, 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. Bio-phys. Acta 1508, 1 (2000).

265. M. Almgren, Biochim. Biophys. Acta 1508, 146 (2000).

266. M. Ollivon, S. Lesieur, C. Grabielle-Madelmont, and M. Paternos-tre, 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. ElectronMicrosc. 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 Applicationsof Liposomes.” Elsevier, Amsterdam, 1998.

297. A. S. Janoff, Ed., “Liposomes: Rational Design.” Dekker, NewYork, 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 Func-

tion.” Springer-Verlag, New York, 1989.301. I. Ash, “Handbook of Industrial Surfactants.” Synapse Informa-

tion Resources, New York, 2000.302. Y. Okahata, R. Ando, and T. Kunitake, Ber. Bunsenges. Phys.

Chem. 85, 789 (1981).

76 Preparation of Vesicles (Liposomes)

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. Carmona-

Ribeiro, 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. Mac-

Donald, 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, Langmuir10, 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, Lang-

muir 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. Conde-Frieboes, 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, Bio-

chemistry 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).

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 Applica-tions 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 Assem-

blies: The Synkinetic Approach.” Royal Society of Chemistry,Cambridge, 1994.

359. G. B. Patel, B. J. Agnew, L. Deschatelets, L. P. Fleming, andG. 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. ColloidInterface Sci. 243, 491 (2001).

363. H. Hoffmann, C. Thunig, and D. Miller, Colloids Surf. A: Physic-ochem. 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. Leck-

band, and J. A. Zasadzinski, Science 264, 1753 (1994).371. J. A. Zasadzinski, E. Kisak, and C. Evans, Curr. Opin. Colloid Inter-

face Sci. 6, 85 (2001).

Preparation of Vesicles (Liposomes) 77

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., “Lipo-

some Dermatics.” Springer-Verlag, Berlin, 1992.389. R. M. Handjani-Vila, A. Ribier, and G. Vanlerberghe, in “Lipo-

some Technology” (G. Gregoriadis, Ed.), Vol. 2, 2nd ed., p. 201.CRC Press, Boca Raton, FL, 1993.

390. G. Vanlerberghe, in “Handbook of Nonmedical Applications ofLiposomes” (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. Lipowskyand 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. Fet-tinger, and J. T. Davis, J. Am. Chem. Soc. 124, 2267 (2002).

400. P. L. Luisi, P. Walde, and T. Oberholzer, Curr. Opin. Colloid Inter-face 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 andG. A. Reineccius, Eds.), p. 113. Am. Chem. Soc., Washington, DC,1995.

404. P. Dufour, E. Laloy, J.-C. Vuillemard, and R. E. Simard, in “Hand-book of Nonmedical Applications of Liposomes” (D. D. Lasic andY. 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 Appli-

cations of Liposomes” (D. D. Lasic and Y. Barenholz, Eds.),Vol. 4, p. 127. CRC Press, Boca Raton, FL, 1996.

407. G. Mossa, M. C. Annesini, and L. Di Marzio, in “Handbookof Nonmedical Applications of Liposomes” (D. D. Lasic andY. Barenholz, Eds.), Vol. 1, p. 273. CRC Press, Boca Raton, FL,1996.

408. F. Cioci and R. Lavecchia, in “Handbook of Nonmedical Applica-tions 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 Reactiv-ity,” 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 Lipo-

somes” (D. D. Lasic and Y. Barenholz, Eds.), Vol. 2, p. 349. CRCPress, 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 Preparation of Vesicles (Liposomes)

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, andR. 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 Dis-

cuss. 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).

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. Rosen-zweig, 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 Lipo-

somes” (D. D. Lasic and D. Papahadjopoulos, Eds.), p. 75. Else-vier, Amsterdam, 1998.

515. T. Daemen, L. Burgener, A. Huckriede, and J. Wilschut, J. Lipo-some 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. Acta1146, 247 (1993).

Preparation of Vesicles (Liposomes) 79

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. Bio-

phys. 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, Biomacro-

molecules 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. Kil-

lany, 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. Tim-mis, 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).

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, Bio-

chemistry 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. Acta1467, 219 (2000).

1. Introduction

Liposomes are spherical, selfclosed vesicles of colloidal dimensions, in which (phos-pho)lipid bilayer sequesters part of the solvent, in which they freely float, into theirinterior [1]. In the case of one bilayer encapsulating the aqueous core one speakseither of small or large unilamellar vesicles while in the case of many concentricbilayers one defines large multilamellar vesicles [2].

Due to their structure, chemical composition and colloidal size, all of which can bewell controlled by preparation methods, liposomes exhibit several properties whichmay be useful in various applications. The most important properties are colloidalsize, 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 behav-ior, its mechanical properties and permeability, charge density, presence of surfacebound or grafted polymers, or attachment of special ligands, respectively. Addi-tionally, due to their amphiphilic character, liposomes are a powerful solubilizingsystem for a wide range of compounds. In addition to these physico-chemical prop-erties, 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 sol-ubilizers for difficult-to-dissolve substances, dispersants, sustained release systems,delivery systems for the encapsulated substances, stabilizers, protective agents, mi-croencapsulation systems and microreactors being the most obvious ones. Liposomescan be made entirely from naturally occurring substances and are therefore nontoxic,biodegradable and non immunogenic. In addition to these applications which hadsignificant impact in several industries, the properties of liposomes offer a veryuseful model system in many fundamental studies from topology, membrane bio-physics, photophysics and photochemistry, colloid interactions, cell function, signaltransduction, 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 analyt-ical biochemistry, solubilizers for various ingredients as well as support matrix forvarious 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 whichis related to their bending elasticity. A number of new theoretical concepts weredeveloped to understand their conformational behaviour [4]. On the other hand they

493

494 D.D. Lasic

Table 1Applications of liposomes in the sciences.

Discipline Application

Mathematics Topology of two-dimensional surfaces in three-dimensional space governedonly by bilayer elasticity

Physics Aggregation behaviour, fractals, soft and high-strength materialsBiophysics Permeability, phase transitions in two-dimensions, photophysicsPhysical Chemistry Colloid behaviour in a system of well-defined physical characteristics, inter-

and intra-aggregate forces, DLVOChemistry Photochemistry, artificial photosynthesis, catalysis, microcompartmentalizationBiochemistry Reconstitution of membrane proteins into artificial membranesBiology Model biological membranes, cell function, fusion, recognitionPharmaceutics Studies of drug actionMedicine Drug-delivery and medical diagnostics, gene therapy

Table 2Liposomes in the pharmaceutical industry.

Liposome Utility Current Applications Disease States Treated

Solubilization Amphotericin B, minoxidil Fungal infectionsSite-Avoidance Amphotericin B – reduced nephrotoxicity, Fungal infections, cancer

doxorubicin – decreased cardiotoxicitySustained-Release Systemic antineoplastic drugs, hormones, Cancer, biotherapeutics

corticosteroids, drug depot in the lungsDrug protection Cytosine arabinoside, interleukins Cancer, etc.RES Targeting Immunomodulators, vaccines, antimalarials, Cancer, MAI, tropical parasites

macophage-located diseasesSpecific Targeting Cells bearing specific antigens Wide therapeutic applicabilityExtravasation Leaky vasculature of tumours, inflammations, Cancer, bacterial infections

infectionsAccumulation Prostaglandins Cardiovascular diseasesEnhanced Penetration Topical vehicles DermatologyDrug Depot Lungs, sub-cutaneous, intra-muscular, ocular 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. Theiraggregation leads to fractal clusters. In addition they can serve as a model to studyvesiculation, including vesicle shedding and endo- and exo-cytosis, of living cells(table 1).

Despite their widespread application, the mechanism of liposome formation is notyet well understood. The equilibrium calculations of the shapes of giant unilamellarvesicles [7, 8] and their observations (fig. 1A) [9, 10], however, offer a qualitativeguidance in the modeling of structural transformations in the various vesiculationprocesses. Figure 1B shows that similar shapes occur also in multilamellar aggregatesand it is reasonable to assume that the gradient of hydration across the stack ofconcentric lamellae causes also the gradient of surface areas of polar heads in theconsecutive 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 thera-peutic 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 thatthe therapeutic concentration is not much lower than the toxic one. In several casesthe toxicity can be reduced or the efficacy enhanced by the use of an appropriatedrug carrier which changes the temporal and spatial distribution of the drug, i.e. itspharmacokinetics and biodistribution.

The benefits and limitations of liposome drug carriers critically depend on the in-teraction of liposomes with cells and their fate in vivo after administration. In vitroand in vivo studies of the interactions with cells have shown that the predominantinteraction of liposomes with cells is either simple adsorption or subsequent endocy-tosis. Fusion with cell membranes is much more rare. The fourth possible interactionis exchange of bilayer constituents, such as lipids, cholesterol, and membrane boundmolecules with components of cell membranes. These interactions, schematicallyshown in fig. 3, determine also the fate of liposomes in vivo.

The body protects itself with a complex defense system. Upon entering into thebody, larger objects cause thrombus formation and eventually their surface is passi-vated by coating with biomacromolecules while smaller particles, including microbes,bacteria, and colloids are eaten up by the cells of the immune system. This responseof the immune system has triggered substantial efforts in the development of bio-compatible and nonrecognizable surfaces and has also, on the other hand, narrowedthe spectrum of applications of microparticulate drug carriers only to targeting of thevery same cells of the immune system.

Fig. 3. Schematic presentation of liposome interactions with cells. Endocytosis is shown in the upperleft 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 locatedmainly in the liver, spleen, and bone marrow.

3.1. Modes of liposome action

Liposomes as drug delivery systems can offer several advantages over conventionaldosage forms especially for parenteral (i.e. local or systemic injection or infusion),topical, and pulmonary route of administration. The preceding discussion showsthat liposomes exhibit different biodistribution and pharmacokinetics than free drugmolecules. In several cases this can be used to improve the therapeutic efficacy ofthe encapsulated drug molecules. The limitations can be reduced bioavailability ofthe drug, saturation of the cells of the immune system with lipids and potentiallyincreased toxicity of some drugs due to their increased interactions with particularcells. The benefits of drug loaden liposomes, which can be applied as (colloidal) so-lution, aerosol, or in (semi) solid forms, such as creams and gels, can be summarizedinto seven categories:

(i) Improved solubility of lipophilic and amphiphilic drugs. Examples includePorphyrins, Amphotericin B, Minoxidil, some peptides, and anthracyclines,respectively; furthermore, in some cases hydrophilic drugs, such as anticanceragent Doxorubicin or Acyclovir can be encapsulateded in the liposome interiorat concentrations several fold above their aqueous solubility. This is possibledue to precipitation of the drug or gel formation inside the liposome withappropriate substances encapsulated [17];

(ii) Passive targeting to the cells of the immune system, especially cells of themononuclear phagocytic system (in older literature reticuloendothelial sys-tem). Examples are antimonials, Amphotericin B, porphyrins and also vac-cines, immunomodulators or (immuno)supressors;

(iii) Sustained release system of systemically or locally administered liposomes.Examples are doxorubicin, cytosine arabinose, cortisones, biological proteinsor peptides such as vasopressin;

(iv) Site-avoidance mechanism: liposomes do not dispose in certain organs, suchas heart, kidneys, brain, and nervous system and this reduces cardio-, nephro-,and neuro-toxicity. Typical examples are reduced nephrotoxicity of Ampho-tericin B, and reduced cardiotoxicity of Doxorubicin liposomes;

(v) Site specific targeting: in certain cases liposomes with surface attached ligandscan bind to target cells (‘key and lock’ mechanism), or can be delivered intothe target tissue by local anatomical conditions such as leaky and badly formedblood vessels, their basal lamina, and capillaries. Examples include anticancer,antiinfection and antiinflammatory drugs;

(vi) Improved transfer of hydrophilic, charged molecules such as chelators, antibi-otics, plasmids, and genes into cells; and

(vii) Improved penetration into tissues, especially in the case of dermally appliedliposomal dosage forms. Examples include anaesthetics, corticosteroids, andinsulin.

Applications of liposomes 505

Fig. 5. Comparison of blood clearance of various liposome formulations in rats following intravenousadministration 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 (1900PEG-DSPE). Clearance of the free label,

Galium desferal is also shown. Courtesy M.C. Woodle and M. Newman.

3.3.1. Sterically stabilised liposomesThe fate of liposomes, i.e. their rapid clearance from the body, was realized ratherearly. First attempts to alter their biodistribution by either surface ligands or mem-brane composition were undertaken in the late 70’s. The results showed that liposomedisposition can be altered, but predominantly within the mononuclear phagocytic sys-tem including the intrahepatic uptake itself. Blood circulation times were prolongedbut the first substantial improvements were achieved by the incorporation of gan-glioside GM1 or phosphatidylinositol at 5–10 mol% into the bilayer [33, 34]. Thebest results were obtained by substituting these two lipids with synthetic polymercontaining lipids. The longest circulation times were achieved when polyethyleneglycol covalently bound to the phospholipid was used. It seems that intermediatemolecular weights, from 1500 to 5000 Da are the optimum [35]. Figure 5 showsblood clearance profiles of several different formulations.

It was suggested [36] that the presence of a steric barrier reduces adhesion and ad-sorption (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 schematicallyshown 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 thebilayer 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 inbiological environments. The presence of the polymer mushroom or brush [68] reduces the adsorptionand adhesion rate of immunoglobulins and other antibodies (A) and plasma (lipo) proteins (P) whicheither 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 profilesof polyethylene glycol grafted liposomes have shown enhanced bilayer repulsion [38,39] in agreement with the hypothesis that reduced surface adhesion and adsorptionstabilises liposomes. Recent theoretical work also explained the experimentally well-known fact that increased concentrations of longer chains start to reverse the effect atparticular polymer density. This is due to the so-called collapse of the brush whichoccurs at certain polymer density and results in polymer selfaggregation [40], a well-known fact from the experimental polymer science. Longer chains can also exhibitincreased attractive and bridging forces with macromolecules [41]. Of course, thein vivo and in vitro stability are not necessarily correlated and, for example, in vitrovery stable formulations, such as highly charged ones, or the ones with chargedbrush, are cleared in vivo very rapidly. Another factor which may differ betweenthe two tests is the role of chain flexibility on the interactions with particles andproteins. It is possible that the decreased mobility of chains in the denser brushregimes, when the chain motion correlation times may approach times required forprotein binding, can account for the weak physisorption of proteins.

3.3.2. Medical applications of stealth liposomesSterically stabilised vesicles can act either as long circulating microreservoirs ortumour (or site of inflammation and infection) targetting vehicles. The former ap-plications requires larger liposomes (∼ 0.2 μm) while the latter one is due to theability of small vesicles to leave the blood circulation. The prolonged presence ofsmall 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 rhodamine-phosphatidylethanolamine labelled long circulating liposomes approximately one hour after injection.The difference between the healthy and tumour tissue, which shows accumulation of fluorescent lipo-somes in extravascular sites can be clearly observed. Plates C and D show accumulation of Doxorubicinloaded Stealth liposomes in mammary adenocarcinoma tumour one hour and one day, respectively, aftertail vein bolus injection in female Fischer rat skin flap window chamber model. In this study fluorescenceof the encapsulated drug Doxorubicin was used as a marker. The intensity is much lower, however, dueto the self quenching effect of highly concentrated drug inside the liposomes. Due to this high concentra-tion, which is 3–5 fold above its aqueous solubility the drug is precipitated or gelled with counterions inthe vesicle interior. Because these liposomes practically do not leak their contents the blood vessels canbe barely seen either immediately after injection or one hour later (C). Few focal points of accumulationwhich are outside blood circulation (as can be easily verified by the bright field light microscopy ofthe same area) show that after one hour there is already some extravasation (C). The increase of fluo-rescence intensity at 24 hours, however, indicates not only high accumulation of the drug but also thefact 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 equilibriumlarge doses of liposomes can accumulate due to their adhesion or immobilization. (Inanalogy with biocompatible surfaces we can speculate that PEG chains effectivelyreduce the adsorption of proteins while for the prevention of cell adhesion muchlonger chains would be required [42]. At present, it is still not known if such longchains can be effectively incorporated into liposomes.) This allows larger doses ofliposome loaden drugs to be delivered to malignant tissues. For instance more then10% of the injected dose of stealth liposome encapsulated Doxorubicin was foundin 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 inthe fluorescence microscope of the biodistribution of fluorescent molecules or labelledliposomes in vivo, in this case dorsal tissue of rat clipped between special microscopicslides [39]. Extensive liposome localization in the tumours was observed. Healthytissue did not accumulate any signal which was due to Doxorubicin fluorescen-ce [39].

Efficacy studies in various mice tumour models, such as implanted solid C26carcinoma and inoculated mammary carcinoma, have shown dramatic improvements[44–48]. Solid C26 colon tumour is practically resistant to free drug, conventionalEpirubicin (a very similar drug to Doxorubucin) liposomes, and mixtures of freedrug and empty stealth liposomes. Stealth Epirubicin and Doxorubicin liposomesresulted, however, in complete remissions of tumours in the early treatment scheduleand 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 micewith recent implants from various tumours but also in reducing the incidence ofmetastases originating from these intra mammary implants. Similarly, several foldincreased drug accumulation was observed also in sites of infections which are alsocharacterized by the enhanced vascular permeability. For instance, in mice withinfected lungs 10 fold more antibiotic drug accumulated in the infected lung ascompared to the noninfected one [49].

Sterically stabilised liposomes may act also as a sustained drug release systemeither 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 wereimplanted 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 eachgroup 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): Epirubicinin stealth liposomes (S-Epi) at 6 mg/ml and, (D): at 9 mg/ml (which resulted in no observable tumourat all). (E): mixture of free drug and empty liposomes. (F): the average values of ten animals in eachexperiment 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 treatmentof lymphoma [48] while the subcutaneous/intramuscular sustained release systemwas demonstrated by the action of polypeptide vasopressin [50]. Its action wasprolonged up to a month as compared to few days for a free drug and a week forthe peptide encapsulated in conventional liposomes. It is important to note thatthese concepts are becoming more and more important with the introduction ofgenetically engineered polypeptides and proteins which are hampered by the rapidblood clearance, degradation and/or deactivation in the body.

The altered biodistribution of stealth liposomes, in addition to the accumulation atthe sites characterised with porous blood capillaries, such as in tumours, inflamma-tions, and infections, may benefit several other applications. In the intact vasculaturethe distribution of stealth liposomes is shifted from the liver, spleen, and bone mar-row towards skin. This opens the opportunity to deliver antivirals and dermatologicalagents to these sites. On the other hand, and while it was shown that the adminis-tration of empty stealth liposomes is well tolerated [5], it requires careful toxicologyand tolerability studies when liposomes loaden with potent drugs are used.

3.3.3. Applications of Stealth liposomes in manThe encouraging results of Doxorubicin encapsulated in Stealth liposomes in pre-clinical studies were observed also in clinical trials in humans. Blood circulationtimes around 45 hours were found [51] and at reduced toxicity very good response inAIDS patients with Kaposi sarcoma was observed [52, 53]. Long circulation timessignificantly, i.e. 200-fold, increased the area under curve of drug concentration vstime and accumulation in various tumours was proportional to the liposome circula-tion times [51]. The drug remained encapsulated in circulating liposomes up to oneweek after injection while at tumour sites drug metabolites were found indicatingthat it had been released from liposomes. The concentration of the drug in tumourswas 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 lesionswhile total flattening was observed in 25% of the cases [52]. The high efficacy wasdue to the approximately ten fold higher drug concentration in lesions as comparedto the administration of free drug (table 3) [53].

In conclusion, it seems that stealth liposomes loaded with anticancer drugs willachieve substantial improvements in the treatments of various tumours. In addition, it

Table 3Localization of Doxorubicin in Kaposi Sarcoma lesions after intravenousinjection of free drug and drug encapsulated in Stealth liposomes (from

ref. [53]).

dose Doxorubicin concentration Selectivity[mg/m2] μg/g tissue Index

free drug in stealth liposome

10 0.18 2.06 11.420 0.31 1.61 5.240 0.72 7.11 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 ingredientsin cosmetics. In addition, liposomes as a carrier itself offers advantages becauselipids are well hydrated and can reduce the dryness of the skin which is a primarycause for its ageing. Also, liposomes can act as a supply which acts to replenishlipids and, importantly, linolenic acid.

In general the rules for topical drug applications and delivery of other compoundsare less stringent than the ones for parenteral administration and several hundredcosmetic products are commercially available since Capture (C. Dior) and Niosomes(L’Oreal) were introduced in 1987. They range from simple liposome pastes whichare used as a replacement for creams, gels, and ointments for do-it-yourself cosmet-ical products to formulations containing various extracts, moisturizers, antibiotics,and to complex products containing recombinant proteins for wound or sunburn heal-ing. Most of the products are anti-ageing skin creams. Unrinsable sunscreens, longlasting perfumes, hair conditioners, aftershaves and similar products, are also gaininglarge 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 fielddo not agree on the mechanism of action. While some claim enhanced permeabilityinto the skin the others claim mostly that liposomes are a noninteractive, skin-non-irritating, water based matrix (without alcohols, detergents, oils and other non-naturalsolubilizers) for the active ingredients.

In addition to the natural lipids, either phospholipids or ‘skin lipids’, which containmostly sphingolipids, ceramides, oleic acid, and cholesterol sulphate, liposomes madefrom synthetic lipids are also being used. They include mostly liposomes made fromnonionic surfactant lipids, which can be chemically more stable. Some of these

Table 4Some liposomal cosmetic formulations currently on the market. According to the manufacturers, lipo-somes may deliver moisture and a novel supply of lipid molecules to skin tissue in a superior fashionto 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 Cristian Dior Liposomes in gel with ingredientsEfect du Soleil L’Oreal Tanning agents in liposomesNiosomes Lancome (L’Oreal) Glyceropolyether with moisturizersNactosomes Lancome (L’Oreal VitaminsFormule Liposome Gel Payot (Ferdinand Muehlens) Thymoxin, hyaluronic acidFuture Perfect Skin Gel Estee Lauder TMF, vitamins E, A palmitate, cerebroside

ceramide, phospholipidSymphatic 2000 Biopharm GmbH Thymus extract, vitamin A palmitateNatipide II Nattermann PL Liposomal gel for do-it-yourself cosmeticsFlawless finish Elizabeth Arden Liquid make-upInovita Pharm/Apotheke Thymus extract, hyaluronic acid, vitamin EEye Perfector Avon Soothing cream to reduce eye irritationAquasome LA Nikko Chemical Co. 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.

“It was the odd pattern of a well-drawn drop ofblood that initiated my curiosity and eased my careeraway from morbid anatomy to that of the physicalchemistry of cell surfaces (1).”

A few individuals change the course of a scientific lifeif one is fortunate enough to work with them. Alec

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

1 Correspondence: E-mail: deamer@soe.ucsc.edudoi: 10.1096/fj.10-0503

Alec Bangham (1921-2010)

1308 0892-6638/10/0024-1308 © FASEB

2010, Alec died peacefully at his home in Great Shel-ford, England, after a brief illness.

Alec Bangham is best known for his seminal researchon 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 laboratoryand decided that a more descriptive term was needed.He proposed “liposomes,” defined as microscopic vesi-cles composed of one or more lipid bilayers (2). Thename stuck, and over the next 30 years liposomes grewinto a minor industry, with the word incorporated intoa number of book titles and a scientific journal, twosuccessful companies, a drug delivery agent, a treat-ment for infantile respiratory distress syndrome, andeven cosmetic formulations by Dior and Lancome. ByMarch 9, 2010 PubMed had listed 35,604 articles onliposomes.

The liposome story began with a paper in 1964,published in the Journal of Molecular Biology, in whichBangham and Horne showed electron microscopicimages of multilamellar phospholipid vesicles (3). By1965, Bangham and his co-workers had done the cru-cial experiments and in twin publications reported thatthe lipid bilayers of the vesicles could maintain concen-tration gradients of ions such as potassium and sodium.Moreover, when natural or synthetic detergents per-turbed the bilayer, the gradient was disrupted (4, 5).This evidence, along with the planar bilayer modelsbeing developed at the same time, established that lipidbilayers are the primary permeability barrier of all cellmembranes. It was the membrane equivalent of findingthe double helix structure of DNA, another Cambridgediscovery in the life sciences.

In 1975, I decided to spend a sabbatical in Alec’slaboratory at Babraham, at the time little more than avillage a few miles south of Cambridge. I knew thatseveral years earlier, Alec had given a talk at BristolUniversity with the title “Membranes Came First!” inwhich he proposed that something like liposomes musthave been available to house the first forms of cellularlife. Inspired by this idea, I returned to my homecampus and soon found that phospholipids could besynthesized in simulated prebiotic conditions (6) andthat membrane-forming amphiphilic molecules werepresent in the Murchison meteorite (7). Furthermore,simple cell-like structures could be produced by encap-sulating functional nucleic acid polymerases in lipidvesicles (8). With Jack Szostak, I am editing a book to bepublished in 2010 by Cold Spring Harbor Press inwhich expert authors give their perspectives on theorigins of life, several of whom echo Alec’s prescientassertion that membranes came first.

Throughout his time at Babraham, Alec’s laboratoryconsisted of a set of connected barrack structures leftover 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 neverfit very well into the British scientific establishment. Hisresearch was curiosity-driven, and he boldly followed

his ideas wherever they led. At first, liposomes seemedto be just a laboratory curiosity, but the excitementgenerated by Alec’s discovery finally penetrated theestablishment and he was named a Fellow of the RoyalSociety in 1977.

Alec probably never intended to be a research scien-tist. Instead, he trained as a clinician, and as a youngmedic he was posted to Israel where the British forceswere being attacked in the chaotic struggle from whichthe nation of Israel emerged. He returned for a shortvisit in 1998 while I was working at the WeizmannInstitute, and we drove around the countryside nearRehovot that he remembered clearly from his experi-ences as a young physician half a century earlier. Oneof the sites had become a geriatric hospital, and Alecpointed to a tree under which he had buried a pile ofmedical equipment when the British were forced toabandon Palestine.

In 1952, Alec accepted a research position at theInstitute of Animal Physiology at Babraham, where hespent the rest of his scientific career. By this time hehad met Rosalind, who had her own practice as aphysician. They were soon wed, and settled in nearbyGreat Shelford to raise their family of four children,Andrew, Janet, Oliver, and Daniel. Andrew now serveson the faculty of the University of East Anglia, special-izing in computational biology; Oliver runs a privateconsulting firm for industrial management in London;Janet is a well-known water color artist in Cambridge;and Daniel has earned an international reputation forhis superb hand-crafted woodwind instruments.

Alec’s first application of liposomes as a modelmembrane system was to test a hypothesis related to theaction of general anesthetics, which proposed thatanesthetics partitioned into the lipid bilayer moiety andin some way inhibited nervous function. In a collabo-rative study with Sheena Johnson and Keith Miller (9),Alec demonstrated that liposomes exposed to generalanesthetics became significantly more permeable toionic solutes. It had previously been shown that anes-thesia could be reversed by hyperbaric pressure, but thereason was not at all clear. Alec and his co-authorsdemonstrated that pressure could also reverse thepermeabilizing effect of anesthetics on liposome mem-branes. This suggested that anesthetic action could beunderstood in thermodynamic terms, in which pres-sure, temperature, and ionic permeability all must betaken into account. Keith Miller, now in the Depart-ment of Anesthesia, Massachusetts General Hospital,later commented: “To a field whose most powerfulmodel nearly seven decades ago had been a jar of oliveoil, the liposome’s arrival was a liberating force (10).”Alec recently wrote an account of this research for TheFASEB Journal (11).

Alec also applied liposomes to other clinical ques-tions, one of which involved infantile respiratory dis-tress syndrome. It is known that a monomolecular layerof a phospholipid called dipalmitoylphosphatidylcho-line (DPPC) coats the inner surface of lung alveoli. Atbirth, the decreased surface tension produced by the

1309MILESTONE: A MEMOIR OF ALEC BANGHAM, 1921–2010

monolayer is essential for expansion of the lung withthe first breath of the infant. However, particularly inpremature newborns, there is sometimes an insufficientamount of DPPC so that their lungs are unable toexpand. Alec had the idea that a mixture of twophospholipids he called “artificial lung expanding com-pound” (note the acronym) could be delivered to theairways to supply the monolayer (12). To quote theconcluding sentence of a later paper describing aclinical trial of the preparation: “Artificial surfactant(ALEC) given to very premature babies at birth signif-icantly reduces their mortality and the respiratory sup-port needed and should prove a valuable addition totreatment (13).”

Alec’s creative instincts were not limited to his re-search, but were also on display when he gave talksabout liposomes. I once invited him to give a seminar atUC Davis, where I was a member of the biology faculty.Alec asked if he could borrow my Teflon Langmuirtrough, some water saturated with diethyl ether, asample of phospholipid, a graduate student, and a fireextinguisher. The lecture hall was packed, and Alecbegan by introducing the barrier properties of bilayermembranes, pointing out that even a lipid monolayercould inhibit molecular diffusion across an interface.He poured the ether-water into the trough, stationedthe graduate student and fire extinguisher nearby, andheld a match to the trough which erupted in flamingether! He then dipped a glass rod into the phospho-lipid sample and touched it to the water surface. In afew seconds, as the monolayer spread over the trough,the flames were magically extinguished. Alec clearedthe monolayer and lit the ether again. “Any surface-active substance works,” he said, then stuck his finger inhis ear to get a bit of earwax with which he againextinguished the flaming ether. Memorable!

Last year Alec wrote to several of his friends in greatexcitement. He had been pondering another utterlynovel idea: that mixtures of volatile materials he calledbouquets could affect the immune response by alteringthe surface charge on membranes. Its editor recom-mended that he submit the manuscript to The FASEBJournal, and Alec’s last sole author paper was publishedin 2009 (1). I know that this publication gave himimmense pleasure, a remarkable achievement for an88-year-old scientist and something we can all hope toemulate.

I was able to visit Alec late in 2009. Rosalind had dieda few months earlier, and this loss clearly affected himdeeply. However, he still managed to prepare a pot ofhis favorite meringue biscuits from egg whites (see thephotograph) and ordered in a curry for lunch. Wespent a couple of hours discussing surface charges oncell membranes and how bouquets of volatile com-pounds might be able to hide the cells from theimmune response. Once again, Alec has given mesomething to think about. I will miss him.

REFERENCES

1. Bangham, A. (2009) The physical chemistry of self/non-self:jigsaws, transplants and fetuses. FASEB J. 23, 3644–3646

2. Sessa, G., and Weissmann, G. (1968) Phospholipid spherules(liposomes) as a model for biological membranes. J. Lipid Res. 9,310–318

3. Bangham, A. D., and Horne, R.W. (1964) Negative staining ofphospholipids and their structural modification by surfaceactive agents as observed in the electron microscope. J. Mol. Biol8, 660–668

4. Bangham, A. D., Standish, M. M., and Watkins, J. C. (1965)Diffusion of univalent ions across the lamellae of swollenphospholipids. J. Mol. Biol. 13, 238–252

5. Bangham, A. D., Standish, M. M., and Weissmann, G. (1965)The action of steroids and streptolysin S on the permeability ofphospholipid structures to cations. J. Mol. Biol. 13, 253–259

6. Hargreaves, W.W., Mulvihill, S. J., and Deamer, D.W. (1977)Synthesis of phospholipids and membranes in prebiotic condi-tions. Nature 266, 78–80

7. Deamer, D.W. (1985) Boundary structures are formed by or-ganic compounds of the Murchison carbonaceous chondrite.Nature 317, 792–794

8. 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

9. Johnson, S. M., Miller, K. W., and Bangham, A. D. (1973) Theopposing effects of pressure and general anaesthetis on thecation permeability of liposomes of varying lipid composition.Biochim. Biophys. Acta. 307, 42–57

10. Miller, K. W. (1983) Anaesthetized Liposomes. In: LiposomeLetters, pp. 251–9. (A. D. Bangham, ed.) Academic Press, Lon-don

11. Bangham, A. D. (2005) Liposomes and the physico-chemicalbasis of unconsciousness. FASEB J. 19, 1766–1768

12. Bangham, A. D., Miller, N. G. A., Davies, R. J., Greenough, A.,and Morley, C. J. (1984) Introductory remarks about artificiallung expanding compounds (ALEC). Colloids and Surfaces Phys-ical chemistry of colloids and interfaces: Biotechnologies anddrug research. 10, 337– 341

13. Ten Centre Study Group (1987) Ten Centre trial of artificialsurfactant (artificial lung expanding compound) in verypremature 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 anddo 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 journals@faseb.org.

1310 Vol. 24 May 2010 DEAMERThe FASEB Journal � www.fasebj.org