Presence or Absence of Counterion Specificity in the Interaction of

Dec 16, 2005 - DNA Release from Cross-Linked DNA Gels and DNA Gel Particles. M Morán , Diana Costa , Maria da Graça , Björn Lindman. 2014,233-250 ...
0 downloads 0 Views 116KB Size
864

J. Phys. Chem. B 2006, 110, 864-870

Presence or Absence of Counterion Specificity in the Interaction of Alkylammonium Surfactants with Alkylacrylamide Gels Iseult Lynch*,†,‡ and Lennart Piculell‡ Irish Centre for Colloid Science and Biomaterials, Department of Chemistry, UniVersity College Dublin, Ireland, and Physical Chemistry 1, Center for Chemistry and Chemical Engineering, Lund UniVersity, PO Box 124, 22100 Lund, Sweden ReceiVed: May 30, 2005; In Final Form: September 22, 2005

Patterns in the interaction of cationic surfactants with nonionic polymer gels, which were inferred from a recent study from our laboratory, are confirmed by measurements of a series of alkylammonium surfactants with different counterions with a series of alkyl acrylamide gels of increasing hydrophobicity. Two swelling patterns were observed: Either the swelling continued above the surfactant critical micelle concentration (cmc) and the maximum swelling differed for different counterions and increased in the order of Br- < Cl< Ac-, in agreement with the degree of dissociation of these counterions and with the Hofmeister series or the swelling stopped below (or close to) the surfactant cmc, implying saturation binding, and the maximum swelling was similar for each of the three counterions. Binding studies confirmed that the amount of surfactant bound depends on the counterion in cases where saturation binding is not reached but not in cases where saturation binding is reached. The swelling/binding patterns could be rationalized in terms of a shift from partial binding to saturation binding with increasing gel hydrophobicity and an electrostatic intermicellar repulsion effectively independent of specific counterion binding, owing to “condensation” at the highly charged micellar surface.

Introduction Gel swelling experiments, where a piece of gel is immersed into an excess of surfactant solution, have been shown to be a simple but sensitive method to study polymer-surfactant interactions.1-10 The swelling isotherm is a plot of the degree of gel swelling vs the surfactant concentration, and it clearly shows the onset of surfactant binding and the cooperative nature of the surfactant binding process. Changes in the binding of the surfactant to the gel are accompanied by macroscopic changes in the gel volume, and the gel response is obtained as a function of the free surfactant concentration. For an uncharged gel, the occurrence of binding is evidenced by the gel swelling at a surfactant concentration called the critical association concentration or cac.1,11 The cac is typically close to, but below, the cmc (critical micelle concentration). The existence of a critical aggregation concentration reflects the cooperative nature of the binding process, which is actually a micelle formation at the polymer.12 The generic features of the swelling isotherm of a nonionic gel in solutions of an associating ionic surfactant were presented previously.1 For neutral gels, the gel volume is essentially constant until the critical association concentration (cac). At the cac, an onset of gel swelling is observed as the bound surfactant effectively transforms the gel into an ionic gel. The swelling results from an increase in the osmotic pressure due to the dissociation of the counterions to the surfactant. At an external surfactant concentration close to the cmc, the swelling either levels off or displays a maximum. As will be discussed in detail below, a maximum may reflect a saturation binding of surfactant * To whom correspondence should be addressed. E-mail: Iseult.Lynch@ fkeml.lu.se. † University College Dublin. ‡ Lund University.

to the polymer. However, regardless of whether the polymer is saturated with surfactant, the binding must level off above the cmc, since the chemical potential of the surfactant levels off when free micelles start to form. The swelling behavior of polymer gels is the macroscopic manifestation of the behavior of the corresponding linear polymers,13 where complexation has been shown to begin at the cac,14 and saturation, or a leveling off at the cmc, takes place at increasing surfactant concentration.15 Much work has been done to study and understand the interaction of neutral gels of various compositions with ionic surfactants. Gels such as hydroxyethylcellulose (HEC) and modifications such as hydrophobically modified HEC,1 poly(N-isopropylacrylamide) (NIPA),16,17 hydrophobically modified acrylamide gels (HM-AM),9,18 and so on have been mixed with cationic or anionic surfactants. In many of these cases, it was the presence of the side groups, or hydrophobic modifications that induced the surfactant binding, as the unmodified gel backbone did not bind surfactant at all. For example, AM gels do not bind SDS, whereas poly(ethylene oxide) grafted AM gels do.18 Hydrophobically modified AM was also found to bind cationic surfactants whereas the unmodified AM does not.9 Here, we were specifically interested to study the interactions of ionic surfactants with a series of gels of increasing backbone hydrophobicity where we could go systematically from a nonsurfactant binding backbone to a surfactant binding backbone. To this end, we prepared a series of alkylacrylamide gels from monomers of increasing hydrophobicitysusing the hydrophilic N,N-dimethylacrylamide (DAM), the slightly more hydrophobic N-isopropylacrylamide (NIPA), and the hydrophobic N-tert-butylacrylamide (BAM). Thus, the choice of gels for study was based on their simplicity and the possibility for systematic variation of the hydrophobicity.

10.1021/jp0528562 CCC: $33.50 © 2006 American Chemical Society Published on Web 12/16/2005

Counterion Specificity of Alkylammonium Surfactant The interaction of a series of synthetic alkyl acrylamide gels of increasing hydrophobicity with hexadecyl trimethylammonium halide (C16TA+) surfactants, where the counterions were bromide (Br-), chloride (Cl-), or acetate (Ac-), has been studied previously.3 The gels used were DAM, NIPA, and a copolymer of NIPA and BAM (20:80 BAM/NIPA), where the hydrophobicity of the gels increased in the order of DAM < NIPA < 20:80 BAM/NIPA. The cmc values of the surfactants with the different counterions increase in the order of Br- < Cl- < Ac-. The results suggested that there is a minimum threshold hydrophobicity of the gels, below which a surfactant will not bind and that the hydrophobicity threshold depends on the surfactant and is also sensitive to the counterion to the surfactant. The appearance of a hydrophobicity threshold is a reflection of the nature of the binding process between neutral gels and surfactantssthe binding is between the polymeric gel backbone and the surfactant tails, via a hydrophobic interaction, with counterion specificity being apparent only in cases where saturation binding did not occur.3 Thus, C16TA+ did not bind to DAM gels regardless of the counterion. NIPA gels were found to interact strongly with all C16TA+ surfactants, with the degree of swelling increasing in the order of Br- < Cl- < Ac-. The introduction of 20 mol % BAM groups into NIPA gels resulted in more hydrophobic gels, all of which interacted strongly with C16TA+, and without any dependence of the degree of swelling on the counterion. Additionally, we found that once there is enough incentive to bind (i.e., once the hydrophobicity threshold has been reached) binding can occur in one of two patternss swelling continues above the cmc and the gel does not necessarily become saturated with surfactant or saturation of the gel with surfactant occurs below (or close to) the cmc, at which point no further swelling occurs.3 Moreover, the presence or absence of counterion specificity in the degree of swelling seemed to correlate with the swelling pattern. However, some questions remained from this previous work, in particular, relating to the origin of the observed presence or absence of counterion specificity in the swelling behavior. Additionally, one may question whether this reduction of the specific ion effects upon saturation of the gel with surfactant is indeed a general phenomenon that appears with a range of surfactants. In this work, we aim to provide answers to these questions. To this end, swelling isotherms of NIPA and BAM/NIPA (20:80) gels with C14TA+ (tetradecyl trimethylammonium halide) and C12TA+ (dodecyl trimethylammonium halide) surfactants with each of the three counterions (Br-, Cl- and Ac-) have been investigated here. Additionally, binding studies have been conducted using the C16TA+ surfactants to determine quantitatively the amount of surfactant bound by the gels of increasing hydrophobicity. The observed trends can be rationalized in terms of accepted physical models of polymer-surfactant binding and of counterion binding to surfactant micelles. Experimental Section Materials. N-isopropylacrylamide (NIPA) monomer (purity >99%) from Phase Separations Ltd. (Clwyd, U.K.) was recrystallized twice from hexane. The following chemicals were used as supplied: N-tert-butylacrylamide (BAM) and N,Ndimethylacrylamide (DAM), from Fluka (Dorset, U.K.); N,N1methylene-bisacrylamide (BisAM) cross-linker (purity 99%+) from Aldrich (Dorset, UK); ammonium peroxydisulfate (APS) initiator (purity 98%+) and N,N,N1,N1-tetramethylethylenediamine (TEMED) promotor, from Sigma (MO); N,N-azobisisobutyronitrile (AIBM) from Phase Separation Ltd. (U.K.);

J. Phys. Chem. B, Vol. 110, No. 2, 2006 865 dodecyl trimethylammonium bromide (C12TABr) from Eastman Kodak Company (NY); dodecyl trimethylammonium chloride (C12TACl) (purity 98%) from Acros Organics (Belgium); tetradecyl trimethylammonium bromide (C14TABr) (purity 98%) from Aldrich (Germany); tetradecyl trimethylammonium chloride (purity >98%) from TCI Europe (Belgium); hexadecyl trimethylammonium bromide (C16TABr) (purity 98.5%) from Merck (Germany); and hexadecyl trimethylammonium chloride (C16TACl) from TCI-EP (Japan). Dodecyl trimethylammonium acetate (C12TAAc) was prepared from C12TABr, tetradecyl trimethylammonium acetate (C14TAAc) was prepared from C14TABr, and hexadecyl trimethylammonium acetate (C16TAAc) was prepared from C16TABr by an ion-exchange process according to the method of Svensson et al.19 Dowex 1 ion-exchange resin (Sigma, U.S.A.) was activated by stirring in 1 M NaOH solution followed by rinsing with copious amounts of deionized water immediately prior to use. A portion of 1 M acetic acid was prepared from a standard solution kit. Cylindrical moulds were Duron Ring caps from Hirschumann Laborgora¨te (Germany), with a 1.4 mm internal diameter. All water used was of Milli-Q (Millipore) quality. The cmc values for C12TAAc and C14TAAc were determined by conductance measurements using a Metrohm 712 conductometer, with a cell constant of 0.83 cm-1. The cmc of C12TAAc was determined to be 27 mM and that of C14TAAc was 5.4 mM, both at 23 °C. Synthetic hydrogels of increasing hydrophobicity were prepared by a standard method.20,21 An aqueous solution containing 700 mM monomer (C0) and 8.6 mM cross-linker (BisAM) was degassed under vacuum, and 15 µL of TEMED was added. The initiator concentration was 40 mg in 1 mL of water, of which 100 µL was added. The gelation reaction proceeded over 24 h at 4 °C. Due to the hydrophobic nature of N-tert-butylacrylamide, it was not possible to dissolve more than 140 mM of it in aqueous solution, and thus, a copolymer of it and NIPA was prepared. The amount of leachable product (unreacted monomer and short chain polymer not incorporated into the gel network) was determined to be negligible (6 months) and even then the results were inconclusive. Kokufuta et al. suggested that the distribution of bound surfactant in a gel may be localized at the gel surface, making the overall surfactant concentration in the gel lower than would be expected and

Counterion Specificity of Alkylammonium Surfactant TABLE 1: Amount (mol/g) of Surfactant Bound to NIPA or 15:85 BAM/NIPA Determined by Ion-Specific Titration PNIPAM 85:15 PNIPAM/BAM

C16TABr

C16TACl

1.59 × 10-4 mol/g 1.40 × 10-3 mol/g

4.95 × 10-4 mol/g 1.37 × 10-3 mol/g

resulting in less gel swelling. However, this was found to occur only in the case where the gel was already swollen in pure water upon mixing with the surfactant solution.16 In all our studies, both the swelling and binding isotherms, the gels were dried before the addition of the surfactant solution, so we could rule this out as the source of the anomalous binding behavior but could not explain it further. Considering the problems with the experiment, it was decided to use linear polymers of similar composition (100% NIPAM and 15:85 BAM/NIPA) and to dialyze the polymer solutions against the different surfactant solutions. Polymers in solution undergo Brownian motion, and thus, even though no stirring was used, it was assumed that thorough mixing would have occurred and, more importantly, that the surfactant concentration inside and outside the dialysis tube would have equilibrated. Table 1 shows the binding data for the C16TABr and C16TACl surfactants to NIPAM and 15:85 BAM/NIPA linear polymers. With the NIPAM polymers, more surfactant is bound in the chloride case than in the bromide case. With the 15:85 BAM/NIPAM polymers, the amount of surfactant bound is the same in the two cases (within experimental error). The binding trends thus follow the same patterns as the swelling trends, with the amount of surfactant bound being dependent on the counterion and increasing in the order of the Hofmeister series with the NIPA polymer and the amount of surfactant bound being similar for all counterions with the BAM/NIPA polymer. Thus, the swelling differences/similarities may be ascribed to differences/similarities in the degree of surfactant binding. Discussion The results in this and our previous study3 have lead to the following general picture of the binding of surfactants to nonionic polymer gels. As is generally accepted, surfactant molecules bind to the polymer chains not individually but as self-assembled micellar aggregates. Alternatively, one may describe the association as an adsorption of the polymer chains to the micellar surfaces. The binding requires a net attraction between the surfactant micelles and the polymer chains. If this attraction is strong enough, then it can induce micelle formation on the polymer chains at the cac, below the cmc. The binding is hydrophobic in nature, and a more hydrophobic polymer chain gives a more effective lowering of the cmc (i.e., a lower cac). If the hydrophobic interaction is sufficiently weak, then binding will not occur below the cmc, hence, the notion of a hydrophobicity threshold. Binding may, in the case of a weaker binding, occur above the cmc, at least in principle. However, the degree of binding should then be very low, since the chemical potential of the surfactant increases very slowly with increasing surfactant concentration above the cmc. The experimental evidence, furthermore, is that saturation binding can occur. This should happen when the density of bound micelles on the polymer chains becomes so large that the (electrostatic) repulsion between bound micelles becomes prohibitively high. However, saturation does not necessarily occur below the cmc. If it does not occur, then there should nevertheless be a leveling off of the binding above the cmc, since (see above) the chemical potential of the surfactant then increases very slowly with increasing surfactant concentration. This leveling off was clearly seen in the detailed swelling

J. Phys. Chem. B, Vol. 110, No. 2, 2006 869 isotherms in Figure 4 of ref 3. Additionally, further increasing the gel hydrophobicity above the saturation hydrophobicity threshold does not result in significant additional surfactant binding. How, then, do the surfactant chain length and counterion dependences, and the resulting swelling behavior, fit into the above picture? This work and a previous paper from our group22 have shown that shorter surfactant chain lengths result in weaker binding. The ratio of cac/cmc increases as the chain length becomes shorter, meaning that there is less incentive to form polymer complexes. The simplest explanation for this finding is that the surface area of the micelle becomes smaller with a decreasing aggregation number as the surfactant tail length decreases. Hence, the number of available contact points for the adsorbed polymer decreases. The stronger curvature of small micelles may also play a role for semirigid polymer chains, as computer simulations have shown that as the stiffness of the polymer increases the number of contact points between the polymer and the macroion (micelle) decreases.23 This brings us to the issue of counterion specificity. The general understanding is that ions may be classified in terms of their “polarizability” or “hardness”, both of which reflect their affinity for water and the degree of attraction they feel for, for example, the surface of a micelle.24 The specific ion effects manifested in the Hofmeister series of ions, such as the formation of rodlike micelle structures by C16TABr vs the formation of spherical micelles by C16TACl, have been related to these differences in counterion polarizability.25 Bromide ions reside closer to the micellar surface than acetate ions, which behave as if strongly dissociated.26 As the hydrophilicity of the counterions decreases, they provide better screening of the charged surfactant headgroups. This leads to higher aggregation numbers and lower values of cmc in surfactants with counterions of lower hydrophilicity (i.e., the cmc of CTAAc is higher than that of CTABr3, and the cmc of LiDS is higher than that of NaDS27). However, contrary to the increase of cmc with increasing ion hydrophilicity is the observation here and in our previous work that surfactants with acetate as the counterion have a lower hydrophobicity threshold (cac) for binding to the gels than do surfactants with bromide as the counterion. This was true for C16TA+ with HEC gels, where the chloride and acetate surfactants were bound by the gel but the bromide surfactant was not3 and for C12TA+ gels with the 20:80 BAM/NIPA gel where the chloride and acetate surfactants were bound but bromide surfactant was not. The question is why more hydrophobic counterions give less association of a slightly hydrophobic polymer to the surfactant micelles. This may seem counterintuitive. We suggest that this is a result of a competition between the polymer and the counterions to bind to the micellar surface. There is less to be gained by polymer adsorption if the chain has to partially replace bound counterions. This interpretation does not necessarily mean that counterions and polymer chains bind to the same part of the micellar surface, although this might actually be the case. (The binding may in both cases be a nonspecific adsorption of the binding molecules to the exposed hydrophobic core which takes up a considerable fraction of the surface area of a typical ionic micelle; see, for example, Figure 1.2 in ref 29 (Evans and Wennerstro¨m).) Even if the binding takes place to different sites, the accumulation of nonbinding polymer segments close to the micellar surface may provide a hindrance to counterion binding. This would explain the order of binding of surfactants with different counterions above cac but below saturation.

870 J. Phys. Chem. B, Vol. 110, No. 2, 2006 Next, we have to understand why, when there is saturation binding, the difference in binding and in gel swelling disappears for the different ion forms. First, we must recall from basic theory of colloidal stability28 that both the repulsive force that acts between two neighboring micelles (opposing more binding) and the ionic contribution to the osmotic pressure of the micellar solution (which gives rise to gel swelling) are given by the concentration of counterions halfway between the micelles. For highly charged micelles, this concentration is, in fact, quite insensitive to whether there is specific binding of the ions to the micellar surface.29 This is because even for ions that are not specifically bound, a large fraction “condense” on a highly charged micellar surface, so that the total fraction of bound and condensed ions is rather constant. Recent Monte Carlo simulations demonstrate this effect for the specific case of C16TA+ micelles.29 From Figure 13 in ref 29, we infer that even if, say, 25% of the micellar charges would be effectively neutralized by bound counterions, this would only reduce the counterion contribution to the osmotic pressure by less than 4%, which is too small a difference to be observed in our swelling studies. This effect thus provides a physical explanation why both the degree of saturation binding and the degree of gel swelling at saturation binding should be insensitive to specific counterion binding to the highly charged polymer-bound micelles. Finally, some comments on how the findings and conclusions from this work may be related to previous findings of other authors. Sakai et al. observed that changing the counterion to the surfactant did not alter the transition temperature of NIPA gels, and they concluded that changing the surfactant counterion did not change the amount of surfactant bound for either trimethylammonium or dodecyl sulfate/phosphate surfactants.30 However, they show clear evidence of surfactant binding, as the transitions temperature increased from 34 °C in the absence of surfactant to 38.9 °C with C12TABr and 38.0 °C with C12TACl. We have shown that at 25 °C, far below the critical solution temperature of NIPA, there is no binding of C12TABr and very little binding of C12TACl (see Figure 3). Due to the nature of N-isopropylacrylamide gels, which have inverse solubility in water, they become more hydrophobic at higher temperatures, and thus, it is possible that the hydrophobicity threshold for the onset of binding may be reached by increasing the temperature. Thus, binding of the C12TA+ surfactants does evidently occur at the higher temperatures used by Sakai et al. where the NIPA gels are more hydrophobic, and in light of our findings, it is possible that even saturation-type binding occurs, since no counterion specificity was observed for the two counterions tested. A similar effect was observed by Carlsson et al., where the amount of C12TABr bound to EHEC increased with increasing temperature, which was explained in terms of the increased hydrophobicity of the EHEC polymer above its cloud point temperature, leading to a stronger interaction between the polymer and the surfactant.31 In conclusion, we find that the results and interpretations presented here complement the existing findings and provide a basis to explain the processes occurring in these systems, and the factors affecting the degree of surfactant binding, the onset of surfactant binding, and the counterion effect on surfactant binding. Conclusions The different swelling behavior observed with alkylacrylamide gels of increasing hydrophobicity and alkyl trimethylammonium surfactants with different tail lengths and different counterions (CnTA+) can be explained in terms of a hydrophobically driven binding, where the degree of hydrophobicity

Lynch and Piculell determines the amount of surfactant bound and the binding pattern. The first threshold hydrophobicity, which is surfactant specific, is required for binding to occur. The amount of surfactant bound above this threshold depends on the counterion to the surfactant and increases in the order of the Hofmeister series of ions (Br- < Cl- < Ac-). The second threshold hydrophobicity indicates that the gel is sufficiently hydrophobic to induce saturation binding of the surfactant below the surfactant cmc. The amount of surfactant bound is then independent of the counterion to the surfactant. Due to the high charge density at the surface of the gel-surfactant micelles, and the resulting high proportion of “condensed” counterions, the differences between the polarizabilities of the three counterions studied here are no longer observed as differences in the swelling or binding. Thus, we can conclude that, for a given surfactant, specific ion effects are only important in a narrow hydrophobicity regimesin the range of gel hydrophobicities strong enough to give rise to binding (a cac) but still weak enough so that saturation binding does not occur before the cmc. Acknowledgment. I.L. is funded by a grant from the Health Research Board of Ireland and L.P. by a grant from the Swedish Research Council (VR). References and Notes (1) Sjo¨stro¨m, J.; Piculell, L. Langmuir 2001, 17, 3836. (2) Sjo¨stro¨m, J.; Piculell, L. Colloids Surf., A 2001, 183-185, 429. (3) Lynch, I.; Sjo¨stro¨m, J.; Piculell, L. J. Phys. Chem. B 2005, 109, 4252. (4) Lynch, I.; Sjo¨stro¨m, J.; Piculell, L. J. Phys. Chem. B 2005, 109, 4258. (5) Khokhlov, A. R.; Kramarenko, E. Yu.; Makhaeva, E. E.; Starodubstev, S. G. Macromolecules 1992, 25, 4779. (6) Mironov, A. V.; Starodoubtsev, S. G.; Khokhlov, A. R.; Dembo, A. T.; Yakunin, A. N. Macromolecules 1998, 31, 7698. (7) Shirahama, K.; Sato, S.; Niino, M.; Takisawa, N. Colloids Surf., A 1996, 112, 233. (8) Kokufuta, E.; Suzuki, H.; Yoshida, R.; Kaneko, F.; Yamada, K.; Hirata, M. Colloids Surf., A 1999, 147, 179. (9) White, B. B. H.; Kwak, J. C. T. Colloid Polym. Sci. 1999, 277, 785. (10) Philippova, O. E.; Chtcheglova, L. A.; Karybiants, N. S.; Khokhlov, A. R. Polym. Gels Networks 1998, 6, 409. (11) Wada, N.; Kajima, Y.; Yagi, Y.; Inomata, H.; Saito, S. Langmuir 1993, 9, 46. (12) Chu, D.; Thomas, J. K. J. Am. Chem. Soc. 1986, 108, 6270. (13) Brackman, J. C.; Engberts, J. B. F. N. Chem. Soc. ReV. 1993, 22, 85. (14) Jones, N. M. J. Colloid Interface Sci. 1967, 23, 36. (15) Shirahama, K.; Ide, N. J. Colloid Interface Sci. 1976, 54, 450. (16) Kokufta, E.; Suzuki, H.; Sakamoto, D. Langmuir 1997, 13, 2627. (17) Sayil, C.; Okay, O. Polym. Bull. 2000, 45, 175. (18) Piculell, L.; Hourdet, D.; Iliopoulos, I. Langmuir 1993, 9, 3324. (19) Svensson, A.; Piculell, L.; Karlsson, L.; Cabane, B.; Jo¨nsson, B. J. Phys. Chem., B 2002, 106, 1013. (20) Matsuo, E. S.; Tanaka, T. J. Chem. Phys. 1988, 89, 1695. (21) Lynch, I. Ph.D. Thesis, University College Dublin, Dublin, Ireland, 2000. (22) Rose´n, O.; Sjo¨stro¨m, J.; Piculell, L. Langmuir 1998, 14, 5795. (23) Akinchina, A.; Linse, P. J. Phys. Chem. B 2003, 107, 8011. (24) Subramanin, D. Langmuir 2000, 16, 4447. (25) Brady, J. E.; Evans, D. F.; Warr, G.; Griesser, F.; Ninham, B. W. J. Phys. Chem. 1986, 90, 1853. (26) Evans, D. F.; Ninham, B. W. J. Phys. Chem. 1983, 87, 5025. (27) Joshi, J. V.; Aswal, V. K.; Bahadur, P.; Goyal, P. S. Curr. Sci. 2002, 83, 47. (28) Evans, D. F.; Wennerstro¨m, H. The Colloidal Domain: Where chemistry, Biology and Technology meet, 2nd ed.; Wiley-VCH: New York, 1999. (29) Svensson, A.; Piculell, L.; Karlsson, L.; Cabane, B.; Jo¨nsson, B. J. Phys. Chem. B 2003, 107, 8119. (30) Sakai, M.; Satoh, N.; Tsujii, K.; Zhang, Y.-Q.; Tanaka, T. Langmuir 1995, 11, 2493. (31) Carlsson, A.; Karlstro¨m, K.; Lindman, B. J. Phys. Chem. 1989, 93, 3673.