糖基表面活性剂的扩展行为

Journal of Colloid and Interface Science 337(2009)

211–217

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science

w w w. e l s e v i e r. c o m /l o c a t e /j c i

The spreading behaviour and spreading mechanism of new glucosamide-based trisiloxane on polystyrene surfaces

Yue Zhang a , Fu Han b, *

a b

Engineering Research Center of Fine Chemicals, Ministry of Education, College of Chemistry and Chemical Engineering, Shanxi University, Shanxi 030006, PR China College of Chemical and Environmental Engineering, Beijing Technology and Business University, Beijing 100037, PR China

a r t i c l e i n f o a b s t r a c t

In order to evaluate the spreading behaviour of new glucosamide-based trisiloxane surfactants on hydro-phobic surfaces, the time-and concentration-dependent spreading performance was investigated. Dynamic light scattering (DLS)was applied to investigate the role of surfactant aggregates on the spread-ing process. The results suggest that glucosamide-based trisiloxane surfactants show superspreading behaviour on hydrophobic surfaces; however, the spreading mechanism differs from that of ethylene oxide trisiloxane. The existence of large aggregates at low concentrations and relatively large spreading areas during the initial stages of wetting is helpful for generating the Marangoni effect, but the gemini-structure does not show good spreading characteristics. Synergistic effects between these ﬁvesurfactants were also investigated, and considerable synergism was observed. Furthermore, both the spreading behaviour of mixed and single component systems suggests that the hydrophilic–lipophilicbalance of surfactants play an important role in superspreading and an optimal HLB value is important for achieving good spreading behaviour.

Article history:

Received 12October 2008Accepted 22April 2009

Available online 6May 2009Keywords:

Superspreading Glucosamide Trisiloxane

Marangoni ﬂowHLB value

1. Introduction

The spreading of water on hydrophobic surfaces is important in many technological applications, such as coatings, cosmetics, and agriculture. Surfactants are often added to improve the spreading behaviour of aqueous solutions. Conventional ionic and non-ionic surfactants are able to enhance the spreading performance of aqueous drops on moderately hydrophobic surfaces effectively, but these surfactants often perform poorly on highly hydrophobic surfaces.

Trisiloxane is a relatively new class of surfactant, which has proven to be effective in enhancing spreading behaviour of aque-ous solutions on highly hydrophobic surfaces. The development and applications of trisiloxane surfactants has attracted great interest. Trisiloxane alkoxylate aqueous solutions show super-spreading behaviour on very hydrophobic surfaces [1], which includes very rapid wetting of fairly hydrophobic substrates, linear wetted area vs. time curves, a maximum in spreading rate on surfaces of intermediate hydrophobicity, and often a maximum in spreading rate as a function of surfactant concentration [2]. Numerous studies have been undertaken in order to determine the mechanisms of superspreading and the effect of molecular structure on spreading behaviour. The impact of ethylene oxide

*Corresponding author. Address:College of Chemical and Environmental Engi-neering, Beijing Technology and Business University, Beijing 100037, PR China.

(EO)chain length on the surface spreading behaviour of trisiloxane surfactants were investigated and provided a connection between chemical structure and superspreading phenomena [2–11]. Some assumptions underlying the mechanism of superspreading were also proposed. These assumptions mostly relate to the conﬁgura-tion of the trisiloxane molecules at an interface, the formation of special aggregates that induce turbidity of surfactant solutions [2,12,13], the formation of a precursor ﬁlmon solid surfaces [2,14,15], or positing that mechanism is a Marangoni-ﬂow-drivenprocess [16,17].

Despite these efforts, the mechanisms of superspreading and the effect of surfactant structure on spreading behaviour are not well understood. Additionally, as all of the suggested superspreading mechanisms are based on studies conducted with ethylene oxide (EO)trisiloxane surfactants, some questions are worth considering:Are trisiloxane surfactants with other kinds of hydrophilic groups able to show superspreading behaviour on hydrophobic surfaces? How do the other hydrophilic groups affect trisiloxane surfactants’spreading performance? Is the superspreading mechanism for the trisiloxane surfactants with EO hydrophilic groups the same as those having other hydrophilic groups?

In order to answer these questions, a series of glucosamide-based trisiloxane surfactants, which were considered to be more dermato-logically compatible, biodegradable, and less toxic [18–20], were synthesised. As shown in Fig. 1, surfactant I has a glucosamide-based hydrophilic group instead of an EO chain. A short hydrophilic side-chain was added to adjust the HLB (hydrophilic–lipophilic

212Y. Zhang, F. Han /Journal of Colloid and Interface Science 337(2009)211–217

Fig. 1. The structure of trisiloxane surfactants I, II, III, IV and V.

balance) property of the molecule (II,III and IV). In order to deter-mine the effect of gemini-structure on spreading behaviour, we syn-thesised a surfactant molecule consisting of, two surfactant I molecules connected with a (CH2CHOHCH 2OCH 2CH 2OCH 2CHO-HCH 2) group (V).In this article, we investigated the time-dependent and the concentration-dependent spreading performance of the ﬁveglucosamide-based trisiloxane surfactants on a polystyrene surface. The synergism between these surfactants is also discussed. Further-more, dynamic light scattering (DLS)was applied to investigate the role of surfactant aggregation on the spreading process. Based on these experiments, we are able to partially answer the above stated questions.

2. Experimental 2.1. Materials

The structure of the siloxane surfactants used in our study is shown in Fig. 1. The surfactants, I, II, III, IV and V, were synthes-ised in our laboratory and were structurally characterised by 1H and 13C NMR spectra and elemental analysis. In all cases, the spectra and elemental analysis acquired were consistent with the assigned structures of the compounds. The details of these spectral characterisations and elemental analysis can be found in [18–20]

Y. Zhang, F. Han /Journal of Colloid and Interface Science 337(2009)211–217213

Sterile polystyrene (PS)petri dishes, purchased from Greiner Bio-One GmbH, were used as hydrophobic surface in the spreading and dynamic contact angle experiments. The surface was pre-washed with double distilled water and dried at room temperature (environmenthumidity 30–40%).Surfactants solutions were pre-pared by manually dispersing the synthesised compounds in dou-ble distilled water and were used within two hours of preparation to avoid hydrolysis.

2.2. Spreading and contact angle measurements

The optical arrangement of the experimental set-up is shown in Fig. 2. The PS surface was placed on a movable stage in front of a microscope, which was connected to a CCD camera. The vertical and horizontal CCD cameras monitored the droplet shape and dy-namic contact angle individually (30frames per second). The vol-umes of surfactants solution used in the spreading measurements and dynamic contact angle measurements were 5l L and 2l L, respectively. The droplet shape and contact angles were estimated with on-screen protractor software (PowereachCo., Shanghai, Chi-na). The temperature of these experiments was kept constant at 25±1°C, and environmental humidity was also kept constant at 50±5%.Every experiment was repeated at least three times until the reproducibility was satisfactory to ensure minimal relative er-ror. The contact angle and spreading area are the average of the areas obtained in each set of measurements. 2.3. Dynamic light scattering

Dynamic light-scattering measurements were performed at 25°C using a Malvern Zetasizer Nano ZS. Double distilled water was ﬁltratedby 0.45-l m and 0.2-l m ﬁlters(GelmanSciences) to remove any dust particles, and then solutions were prepared. The apparatus measures the scattering information close to 180°(back-scatter detection). The correlation curves were analysed using the CONTIN [21]software provided by Malvern. Every measurement was repeated at least ﬁvetimes. 2.4. Surface tension

The surface tension (c ) of aqueous surfactant solutions were measured by pendant drop shape analysis (DSA100,KRÜSS).An environmental chamber was equipped to minimise evaporation, and the temperature was kept constant at 25±0.5°C. We consid-ered the surface tension to be at equilibrium when the surface ten-sion reached a constant value and remained at a constant value

(±0.02mN/m)for at least 2min. The critical aggregation concen-trations (CAC)of these surfactants were determined by equilibrium surface tension vs. surfactant concentration. 3. Results and discussion

3.1. Spreading behaviour on polystyrene surfaces

The spreading factor (5s) is the ratio of the spreading area of the surfactant solution at 5s vs. the spreading area of the same vol-ume of distilled water. The effect that varying the concentration (bothabove and slightly below the CAC) of surfactants has on the spreading factor can be seen in Fig. 3. Surfactants I, II, III and IV showed some superspreading behaviour on the hydrophobic surface, and V showed the worst wetting behaviour. The greatest achievable spreading area for an aqueous droplet containing I (0.5wt%)and II (0.1wt%)can be as much as 15times greater than that of pure water. Moreover, the spreading factor goes through a maximum value at approximately 0.1wt%for surfactants I, II, III and IV.

The existence of a maximum in the spreading factor as a func-tion of the surfactant concentration supports the Marangoni stress hypothesis [16]. At low surfactant concentration, the surface ten-sion gradient generated on the droplet surface is too low to gener-ate the Marangoni ﬂow,which leads to a low spreading factor. As the surfactant concentration increases, a larger surface tension gra-dient is generated and results in larger spreading factors. A further increase in surfactant concentration reduced the surface tension gradient resulting in a higher rate of surfactant diffusion from the bulk to the surface. As a result, the spreading factors of droplets proceed through a maximum as the surfactant concentration in-creases. We also observed that the drop containing I, II, III and IV spreads on the PS surface in the form of a ‘‘ﬂatpan”,and the dy-namic contact angle is large ($30°) at the beginning of spreading (Fig. 4), indicating that the surface tension at the edge of the drop is high and the spreading is not driven by imbalanced capillary forces on the three-phase contact line.

The gemini-structure surfactant V showed poor wetting behav-iour on the PS surface. Although the spreading factor of V increased with concentration, the wetting area is only four times greater than that of pure water. Considering the larger dynamic contact angle during the whole wetting process (Fig. 4), we suspect that surfac-tant V might be driven by imbalanced capillary forces.

In order to better understand the spreading mechanism, the spreading of the droplets during 0–2s are characterised by ﬁttinga power law of the form [22–24]

Fig. 2. Experimental set-up to study droplet spreading and dynamic contact angle.

214Y. Zhang, F. Han /Journal of Colloid and Interface Science 337(2009)211–217

R ðt Þ¼Ct n

In the power-law equation, R is the radius of the spreading area, C is the coefﬁcientof spreading, t is the spreading time and n is the spreading exponent. As the drop shape is axisymmetric and the three-phase contact region along the drop periphery is regular dur-ing 0–2s, the radius of spreading area can be obtained. The value of n gives an indication of the forces dominant during the spreading process. For example, in the case of small drops driven by capillary forces, n =0.1(Tanner’slaw); whereas for larger gravity driven drops, n =0.125[25].

The spreading exponent n for the power-law ﬁtsof the radius vs. time are summarised in Fig. 5. The exponent n of surfactants I, II, III and IV gradually reaches a maximum value and decreases again with increasing surfactant concentration. For the higher sur-factant concentrations, a relation R (t ) /t 1/4is obtained, and the exponent n of II reaches 0.25at lower concentrations for compar-ison with surfactants II, III and IV.

According to the literature [16,23,24], the exponent n =0.25again suggests that the spreading of surfactants I, II, III and IV at higher concentrations is driven by Marangoni ﬂow,and the surface tension gradient may be radial. The maximum of exponent n is the result of surface tension gradient diminishing with a further in-crease in the surfactant concentration. The exponent n of I, II, III

and IV with lower concentrations (slightlyabove and below the CAC) is close to 0.1, which indicates capillary force-driven spreading [22–25]. Moreover, for the crossover regime between capillary force-driven spreading and surface tension gradient-driven spread-ing (0.1

The exponent n of surfactant V increases from approximately 0.05to 0.1with increasing surfactant concentration, which indi-cated that the wetting of V droplet on the PS surface might be dri-ven by an imbalanced capillary force. That the exponent n for the V spreading is smaller than that of Tanner’slaw (n =0.1) then sug-gests that dynamic surface tension needs to be taken into account.

3.2. Dynamic surface tension

Fig. 6illustrates the dynamic surface tension of surfactants I, II, III, IV and V at concentrations close to CAC (1.3ÂCAC). Although a similar equilibrium tension ($21mN/m)is reached in all the sur-factants at concentrations above the CAC, the time required to at-tain the equilibrium surface tension is obviously different. Surfactants I and II reached equilibrium surface tension in a shorter time and showed better spreading behaviour than surfactants III and IV.

Kumar et al. [26]explained the reason for observing different surface tensions at the perimeter and the apex of drop during a Marangoni-driven spreading:As shown in Fig. 7, although a large interface area is created during a rapid spreading process, the ﬂuidlayer under the apex is not continually depleted of surfactant by adsorption onto the solid surface and its thickness is greater than at the perimeter. The inventory of surfactant available for adsorp-tion may be larger than at the advancing perimeter, and, thus, the ﬂuxof surfactant to the surface may be greater at the apex than at the perimeter. This difference in transport rates induces a surface tension gradient between the perimeter and the apex of drop. According to this theory, we suspect that the faster diffusion of sur-factants I and II can induce a lower surface tension at the centre of droplet than that of III, IV or V and can result in a larger surface tension gradient between the edge and centre of the drop. As a re-sult, surfactants I and II show better spreading behaviour on the PS surface.

As shown in Fig. 6b, although surfactant V with gemini-struc-ture attained equilibrium surface tension in a similar time with

Y. Zhang, F. Han /Journal of Colloid and Interface Science 337(2009)211–217215

Fig. 7. Schematic of the generation of surface tension gradient.

that of surfactant II, and shows a lower equilibrium surface ten-sion, this surfactant had the worst wetting characteristics over the entire concentration range studied. It should be noted that the dynamic surface tension of V at this concentration (1.3ÂCAC) is larger than the critical surface tension of the PS surface (33mN/m [27]) for a long time (about20s). As a result, surfactant V can just wet the PS surface during this period. The limited interfacial area and higher dynamic surface tension of surfactant V at the ini-tial stage of wetting cannot generate a larger surface tension gradi-ent between the perimeter and the apex of drop. Surfactant V did not show a similar spreading behaviour to that of surfactant I or II even after the equilibrium surface tension was reached. This ﬁndingrequires further research. 3.3. Hydrodynamic radius (DLS)

As some research has reported that spreading behaviour can be related to surfactant aggregation [1,2,14], dynamic light-scattering (DLS)measurements of the aggregates in solution are performed. Fig. 8shows the relationship between the average values (ﬁvemeasurements) of apparent hydrodynamic radius and the concen-trations (0.01–0.1wt%).As Fig. 8illustrates, the variation trend of particle size is quite different:The apparent hydrodynamic radius of surfactant I increases form 169.5nm (0.005wt%)to 284nm (0.08wt%).A large size of 555.1nm is abruptly seen at 0.01wt%,and, at this time, the solution appears to be to slightly turbid. The apparent hydrodynamic radius of surfactant II decreases from 357.8nm (0.01wt%)to 172.7nm (0.1wt%)with increasing con-centration. The variation in particle size for surfactants III and IV is not as obvious as that of I and II, and the particle size of III de-creases from 184.4nm (0.01wt%)to 119.6nm (0.1wt%).The apparent hydrodynamic radius of V also increases with concentra-tion, reaches a stable value of approximately 300nm at 0.03wt%,and is almost two times the size of III at the same concentration. Comparing the size of common surfactant micelle, the radius of 150–200nm indicated that large aggregates existed in these glu-cosamide-based trisiloxane aqueous solution. However, large aggregates are not always accompanied by superspreading behav-iour in our experiments. For example, the hydrodynamic radius of

V aggregates in all concentrations is about 200–300nm but shows poor spreading behaviour.

Nevertheless, we still ﬁndsome effect of the aggregation size on spreading behaviour. We suspect that the formation of larger aggregates at low concentration favours spreading generated by the Marangoni effect:As new surface are created continuously, lar-ger aggregates can disintegrate and provide an efﬁcientdelivery of surfactant molecules to the air–liquidinterface behind the advanc-ing drop, resulting in high interfacial concentrations of surfactant in this region [10]. As a result, more obvious surface tension gradi-ents will be generated between this region and the region with bulk surfactant concentration. A good example is seen in the sur-factant II system:Comparing the 150–200nm particle size of I, III, IV and V at 0.01wt%(Fig. 8), surfactant II with the apparent hydrodynamic radius of 357.8nm shows the best spreading behav-iour and more obvious Marangoni effect.

3.4. Synergistic effect and HLB (hydrophilic–lipophilicbalance) value 3.4.1. Synergistic effect

In our synergistic effect experiments, the total concentration is always 0.1wt%.The spreading rates of aqueous mixtures of glu-cosamide-based trisiloxane surfactants with different weight frac-tions X B in the mixtures are plotted in Fig. 7, where X B =B /(A +B ). As it showed in Fig. 9, although the curve of spreading rate vs. weight fraction between these surfactants is not very regular, con-siderable synergistic effects were observed in the surfactant I/II,I/III, and I/IVsystems (Fig. 9a).

216Y. Zhang, F. Han /Journal of Colloid and Interface Science 337(2009)211–217

tension data (Fig. 6) of the glucosamide-based trisiloxane indicate that the surface tension gradients of the mixtures investigated may be due to the difference in diffusion rates of the two individual component surfactants. Moreover, disintegrating aggregates on surface also generated a surface tension gradient to some extent. As a result, the Marangoni effect is induced when these surface tension gradient are large enough.

For the mixtures containing surfactant V, the worse synergistic effect (Fig. 9) can be explained in two ways:(1)The adsorption rate of V is too slow to generate a low surface tension, and at the same time, the formation of mixed micelle containing V and other sur-factants might inhibit the adsorption of other surfactants on sur-face. (2)The difference in adsorption rates between V and other surfactants are too big to generating large surface tension gradi-ents. Both of these effects result in diminished synergism and poor spreading/wettingbehaviour.

The aggregates of these systems are investigated with dynamic light scattering at different concentration and different weight ra-tio, but no speciﬁcaggregates or distinct variation rules were iden-tiﬁedas being responsible for the synergistic effect.

3.4.2. HLB effect

In our experiments, we ﬁndthe turbidity of surfactants I, II, III and IV at the same concentration is decreasing with increasing with the number of EO segments and the surfactant with a moder-ate turbidity between I and III shows the best spreading behaviour. Considering that the turbidity of a solution is tightly connected with the HLB value of the surfactant, we try to ﬁndif there is a rela-tionship between the HLB value and spreading behaviour.

The HLB values of the ﬁveglucosamide-based trisiloxanes are calculated and listed in Table 1. It is worth noting that all those mix-tures showing positive synergistic effects are composed of pairs of surfactants with one having a HLB value above 10and the another with HLB value under 10(Fig. 9). Moreover, the larger the difference in their HLB values, the more synergism was observed (Fig. 10). Weak synergistic effects were observed in those both with HLB val-ues above 10. All these phenomena indicated that good spreading

The spreading factors of aqueous mixtures of glucosamide-based trisiloxane (totalconcentration always 0.1wt%)with differ-ent weight fractions (X B ) are plotted in Fig. 8, in order to give us a better intuition about synergistic effects in these systems. The most impressive spontaneous spreading effects are observed in the I/IVsystem, in which the droplet shape rapidly changes from a round to a large irregular shape at the ﬁnalstage of spreading (2–5s), and the rim of the advancing droplet is very unstable with many small ‘‘ﬁngers”formed. The spreading area achieved by the I/IV system can be as much as 29times greater than that of pure water (Fig. 10), which is much larger than that of individual II sys-tem. Some inconspicuous synergistic effect in II/III,II/IVand III/IVsystems are also found. However, no synergistic effects are found in the mixtures containing surfactant V (Fig. 9b).

We speculate that the synergistic effects in these mixture systems are caused by surface tension gradients. Dynamic surface

Table 1

Some parameters of surfactants.

CAC/CMC(10À4mol/L)

7.8221.69.7548

II 5.6520.810.9082

III 6.0621.011.0965

IV 7.3221.311.7424

V 0.62120.210.3595

Silwet L-77Ò1.09(EO8)[30]20.5(EO8)[14]10.1765(EO7)10.9371(EO8)

A 12E n [28,29]23(n =4) 26(n =5) 29.2(n =4) 31.0(n =5) 9.8338(n =4) 10.9717(n =5)

c CAC/CMC(mN/m)

HLB value a

Calculated with the Molecular Modelling Pro V6[31].

Y. Zhang, F. Han /Journal of Colloid and Interface Science 337(2009)211–217217

behaviour is tightly connected with the average HLB value of the mixture, and the optimal value of HLB value should be about 10. This opinion was also conﬁrmedin the one-component systems of the ﬁveglucosamide-based trisiloxane compounds. Combined with the spreading behaviour seen in Figs. 3and 5, it is easy to see that surfactant II, which has a medium HLB values between I and III, shows the best spreading behaviour on the PS surface. The spreading factors of III and IV decreased with increasing HLB values. This relationship between spreading behaviour and HLB values is similar to the behaviour exhibited by trisiloxane ethoxylate and alcohol surfactant systems [2,7]. The HLB values of trisiloxane eth-oxylate surfactant Silwet L-77(EO7,EO8) and ethoxylated alcohol surfactant A 12E n (n =4, 5) are also listed in Table 1, as both are re-garded as showing the highest spreading rate within their homolo-gous compounds [2,7]. Although the molecular structures of these surfactants are different, it is interesting to note that all of the sur-factants showing the best spreading behaviour have a HLB value about 10. The results suggest that an optimal hydrophilic–hydro-phobic balance is important for achieving good spreading behav-iour, and the optimal value of HLB value should be about 10. 4. Conclusion

Glucosamide-based trisiloxane surfactants can show super-spreading behaviour on hydrophobic surfaces (PSsurface). The hydrophilic–hydrophobicbalance of these surfactants affected their spreading behaviour, and an optimal HLB value of about 10is important for achieving good spreading behaviour. At the same time, molecular structure also greatly affects the spreading behav-iour of glucosamide-based trisiloxane surfactant. At least, the gem-ini-structure is not favoured with good wettability characteristics. The spreading of these glucosamide-based trisiloxane was dri-ven by the imbalanced capillary force and Marangoni effect at dif-ferent time scales and different concentrations. Moreover, we suspected that the formation of larger aggregates at low concentra-tion favours generating the Marangoni effect. Acknowledgments

The ﬁnancialsupport of the Natural Science Foundation of Shanxi Province (No.2008021016) and the ScientiﬁcResearch

and Development Program of Universities by Shanxi Provincial Education Ministry (No.20080003) is gratefully acknowledged. References

[1]R.M. Hill, Curr. Opin. Colloid. Interface Sci. 3(1998)247.

[2]T. Stoebe, Z. Lin, R.M. Hill, M.D. Ward, H.T. Davis, Langmuir 13(1997)7270. [3]L.R.R. Wagner, R. Wersig, G. Schmaucks, B. Weiland, J. Weissmuller, J. Reiners,

Appl. Organomet. Chem. 10(1996)421.

[4]L.R.R. Wagner, J. Weibmuller, J. Reiners, K.D. Klein, D. Schaefer, S. Stadtmuller,

Appl. Organomet. Chem. 11(1997)617.

[5]Y.W.R. Wagner, H.V. Berlepsch, H. Zastrow, B. Weiland, L. Perepelittchenko,

Appl. Organomet. Chem. 13(1999)845.

[6]Y.W.R. Wagner, H.V. Berlepsch, F. Rexin, T. Rexin, L. Perepelittchenko, Appl.

Organomet. Chem. 13(1999)621.

[7]Y.W.R. Wagner, G. Czichocki, H.V. Berlepsch, B. Weiland, F. Rexin, L.

Perepelittchenko, Appl. Organomet. Chem. 13(1999)611.

[8]Y.W.R. Wagner, G. Czichocki, H.V. Berlepsch, F. Rexin, L. Perepelittchenko, Appl.

Organomet. Chem. 13(1999)201.

[9]Y.W.R. Wagner, H.V. Berlepsch, L. Perepelittchenko, Appl. Organomet. Chem.

14(2000)177.

[10]T. Stoebe, Z. Lin, R.M. Hill, M.D. Ward, H.T. Davis, Langmuir 13(1997)

7282.

[11]T. Stoebe, Z. Lin, R.M. Hill, M.D. Ward, H.T. Davis, Langmuir 12(1996)337. [12]K.P. Ananthapadmanabhan, E.D. Goddard, P. Chandar, Colloids Surf. A 44

(1990)281.

[13]R.M. Hill, M. He, H.T. Davis, L.E. Scriven, Langmuir 10(1994)1724.

[14]S. Zhu, W.G. Miller, L.E. Scriven, H.T. Davis, Colloids Surf. A 90(1994)63.

[15]N.V. Churaev, N.E. Esipova, R.M. Hill, V.D. Sobolev, V.M. Starov, Z.M. Zorin,

Langmuir 17(2001)1338.

[16]A.D. Nikolov, D.T. Wasan, A. Chengara, K. Koczo, G.A. Policello, I. Kolossvary,

Adv. Colloid Interface Sci. 96(2002)325.

[17]A. Chengara, A. Nikolov, D. Wasan, Colloids Surf. A 206(2002)31. [18]F. Han, G. Zhang, J. Surf. Det. 7(2004)175.

[19]F. Han, G. Zhang, Colloids Surf. A 237(2004)79. [20]F. Han, G. Zhang, Tenside Surf. Det. 40(2003)332.

[21]S.W. Provencher, Comput. Phys. Commun. 27(1982)213.

[22]K.S. Lee, N. Ivanova, V.M. Starov, N. Hilal, V. Dutschk, Adv. Colloid Interface Sci.

144(2008)54.

[23]S. Rafai, D. Bonn, Physica A 358(2005)58.

[24]S. Rafai, D. Sarker, V. Bergeron, J. Meunier, D. Bonn, Langmuir 18(2002)10486. [25]L.H. Turner, J. Phys. D 12(1979)1473.

[26]N. Kumar, A. Couzis, C. Maldarelli, J. Colloid Interface Sci. 267(2003)272. [27]B.W. Cherry, Cambridge Solid State Science Series:Polymer Surfaces,

Cambridge University Press, Cambridge, UK, 1981.

[28]A.A. Ali, B.A. Mulley, J. Pharm. Pharmacol. 30(1978)205.

[29]M.J. Rosen, A.W. Cohen, M. Dahanayake, X.Y. Hua, J. Phys. Chem. 86(1982)

541.

[30]T. Svitova, R.M. Hill, Y. Smirnova, A. Stuermer, G. Yakubov, Langmuir 14(1998)

5023.

[31]Y. Zhang, G.Y. Zhang, F. Han, Colloids Surf. A 276(2006)100.