Effects of Temperature, Salt, and Deuterium Oxide on the Self

Department of Physical Chemistry 1, Centre of Chemistry and Chemical Engineering, Lund University, P.O. ... S-100 44 Stockholm, Sweden, and Product De...
0 downloads 0 Views 126KB Size
Download PDF
Langmuir 2004, 20, 1401-1408

1401

Effects of Temperature, Salt, and Deuterium Oxide on the Self-Aggregation of Alkylglycosides in Dilute Solution. 1. n-Nonyl-β-D-glucoside Caroline A. Ericsson,† Olle So¨derman,† Vasil M. Garamus,‡ Magnus Bergstro¨m,§ and Stefan Ulvenlund*,†,| Department of Physical Chemistry 1, Centre of Chemistry and Chemical Engineering, Lund University, P.O. Box 124, S-221 00 Lund, Sweden, GKSS Research Centre, Max Planck Street, D-215 02 Geesthacht, Germany, Department of Chemistry, Surface Chemistry, Drottning Kristinas va¨ g 51, Royal Institute of Technology, S-100 44 Stockholm, Sweden, and Product Development, AstraZeneca R&D Lund, S-221 87 Lund, Sweden Received August 29, 2003. In Final Form: November 13, 2003 The influence of salt, temperature, and deuterium oxide on the self-aggregation of n-nonyl-β-D-glucoside (β-C9G1) in dilute solution has been investigated by static and dynamic light scattering, neutron scattering, and tensiometry. Scattering data show that the micelles can be described as relatively stiff, elongated structures with a circular cross section. With a decrease of temperature, the micelles grow in one dimension, which makes it surprising that the critical micelle concentration (cmc) shows a concomitant increase. On the other hand, substitution of D2O for H2O causes a large increase in micelle size at low temperatures, without any appreciable effect on cmc. With increasing temperature, the deuterium effect on the micelle size diminishes. The effects of salt on the micelle size and cmc were found to follow the Hofmeister series. Thus, at constant salt concentration, the micelle size decreased according to the sequence SO42- > Cl- > Br- > NO3- > I- > SCN-, whereas the effect on cmc displays the opposite trend. Here, I- and SCN- are salting-in anions. Similarly, the effects of cations decrease with increasing polarizability in the sequence Li+ > Na+ > K+ > Cs+. At high ionic strength, the systems separate into two micellar phases. The results imply that the size of β-C9G1 micelles is extremely sensitive to changes in the headgroup size. More specifically, temperature and salt effects on effective headgroup size, including intermolecular interactions and water of hydration, are suggested to be more decisive for the micelle morphology than the corresponding effects on unimer solubility.

Introduction The increased environmental awareness among consumers and industry has triggered extensive research directed toward the identification of surfactants that are nontoxic, biodegradable, and synthesized from renewable resources. Alkylglycosides, which contain one or more sugar moieties as the hydrophilic group, promise to meet these demands. Due to the favorable environmental profile and the “mild” character of alkylglycosides, they have found applications in cosmetic products, as food emulsifiers, and as solubilizing agents for membrane proteins. 1-4 In-depth understanding of alkylglycoside self-aggregation is of obvious, direct relevance for these applications, as well as for the attempts to identify new ones. In particular, the influence of salt and temperature on self-aggregation and general phase behavior has a direct bearing on dispersing and solubilizing capacity of a given surfactant in various applications. For nonionic surfactants based on poly(ethylene oxide) (PEO), the effects of temperature, salt, * To whom correspondence should be sent. E-mail: [email protected]. Fax: +46-46-337128. † Lund University. ‡ GKSS Research Centre. § Royal Institute of Technology. | AstraZeneca R&D Lund. (1) Kiwada, H.; Nimmura, H.; Fujisaki, Y.; Yamada, S.; Kato, Y. Chem. Pharm. Bull. 1985, 33, 753. (2) Hughes, F. A.; Lew, B. W. J. Am. Oil Chem. Soc. 1970, 47, 162. (3) Matsumura, S.; Imai, K.; Yoshikawa, S.; Kawada, K.; Uchibori, T. J. Am. Oil Chem. Soc. 1990, 67, 996. (4) Baron, C.; Thompson, T. E. Biochim. Biophys. Acta 1975, 382, 276.

and other factors on micelle size and morphology are well-understood and may, in the general case, be directly linked to effects on PEO solubility.5 Similarly, the correlation between physical factors and the details of micellization is well-known for ionic surfactants.5,6 For alkylglycosides, on the other hand, the information and understanding of the micelle formation are much less extensive. Furthermore, many of the features of the selfaggregation of ionic and PEO-based surfactants are difficult or impossible to apply to alkylglycosides in a direct and meaningful manner. More specifically, alkylglycosides carry no charge, but still have relatively rigid headgroups. Thus, they share features with both PEObased and ionic surfactants. Some information about the morphology of alkylglycoside micelles and its dependence on temperature is possible to extract from the literature. n-Octyl-β-Dmaltoside (β-C8G2) and n-dodecyl-β-D-maltoside (β-C12G2) have both been shown to form discrete, nearly spherical micelles whose size shows no, or little, temperature dependence.7,8 However, the micellization of n-octyl-β-Dglucoside (β-C8G1) has been reported to be considerably more complex. In this case, the micelle size shows a flat maximum at 40 °C corresponding to a micellar hydro(5) Jo¨nsson, B.; Lindman, B.; Holmberg, K.; Kronberg, B. Surfactants and Polymers in Aqueous Solution; John Wiley & Sons Ltd.: New York, 1998. (6) Laughlin, R. G. The Aqueous Phase Behaviour of Surfactants, 1 ed.; Academic Press: London, 1994. (7) He, L.-Z.; Garamus, V. M.; Funari, S. S.; Malfois, M.; Willumeit, R.; Niemeyer, B. J. Phys. Chem. B 2002, 106, 7596. (8) Aoudia, M.; Zana, R. J. Colloid Interface Sci. 1998, 206, 158.

10.1021/la035613e CCC: $27.50 © 2004 American Chemical Society Published on Web 01/15/2004

1402

Langmuir, Vol. 20, No. 4, 2004

dynamic radius of ca. 28 Å.9,10 These data are inconsistent with the expected size for spherical micelles, and NMR self-diffusion data imply that nonspherical aggregates are actually formed at concentrations directly above the cmc.11 The NMR data are consistent with an axial ratio of 3:1 when the micelles are modeled as prolate objects. On the other hand, a combination of small-angle neutron scattering (SANS) and small-angle X-ray scattering (SAXS) data suggests that the micelle shape can rather be described by a cylindrical core-shell form factor, with dimensions dependent upon surfactant concentration.12 In the β-C8G1/D2O system, the isotropic, micellar domain extends to ca. 60 wt % of surfactant, where it is followed by a hexagonal phase below 22 °C.13 Increasing the alkyl chain length from C8 to C10 changes the micellization and phase behavior of the corresponding alkylglucoside in a dramatic fashion. Thus, D2O solutions of n-decyl-β-D-glucoside, β-C10G1, display separation into two micellar phases at concentrations between 0.1 and 17 wt % of surfactant.14 The more dilute phase contains discrete aggregates with an estimated aggregation number between 200 and 400, whereas the concentrated phase contains large aggregates with aggregation numbers of >600. These large aggregates have been suggested to be either discrete, wormlike aggregates, or branched micelles that form a bicontinuous structure.14 At higher concentrations, an isotropic one-phase fluid region is formed, which extends to 67 wt % of surfactant, where it is in equilibrium with a lamellar liquid crystalline phase. The phase separation of the dilute micellar phase has been rationalized as a consequence of dilution of a bicontinuous micellar network. According to this line of reasoning, such a network cannot be diluted beyond a certain point, due to the free energy penalty associated with the increase of the curvature of the surfactant film upon dilution.14 Thus, the phase separation can be regarded as an effect of a strong tendency of β-C10G1 to form branched, infinite micelles. Recent, detailed studies of the β-C9G1/β-C10G1/ water system shed further light on this type of phase separation.15 When the two surfactants are mixed in different ratios, the corresponding phase diagrams show a progression from the phase diagram of β-C9G1 to that of β-C10G1, with the two-phase region increasing and the critical temperature decreasing as the ratio of β-C10G1 in the mixture increases. However, there are strong indications that threadlike micelles exist in both micellar phases at equilibrium. A possible explanation for the phase separation in this case is that one phase contains branched micelles, whereas the other comprises no, or at least substantially less, micellar branching. In the present work, we have focused on micelles of β-C9G1, since this surfactant represents a borderline case between the discrete and infinite micellar systems formed in dilute solution by β-C8G1 and β-C10G1, respectively. Consequently, it is reasonable to assume this surfactant to be a key system for the understanding of the micellization of the alkylglycosides as a class. Nilsson et al. have published the binary β-C9G1/water phase diagram and (9) Focher, B.; Savelli, G.; Torri, G.; Vecchio, G.; McKenzie, D. C.; Nicoli, D. F.; Bunton, C. A. Chem. Phys. Lett. 1989, 158, 491. (10) Kameyama, K.; Takagi, T. J. Colloid Interface Sci. 1990, 137, 1. (11) Nilsson, F.; So¨derman, O.; Johansson, I. J. Colloid Interface Sci. 1998, 203, 131. (12) Zhang, R.; Marone, P. A.; Thiyagarajan, P.; Tiede, D. M. Langmuir 1999, 15, 7510. (13) Nilsson, F.; So¨derman, O.; Johansson, I. Langmuir 1996, 12, 902. (14) Nilsson, F.; So¨derman, O.; Hansson, P. Langmuir 1998, 14, 4050. (15) Whiddon, C. R.; So¨derman, O.; Hansson, P. Langmuir 2002, 18, 4610.

Ericsson et al.

found it to be very similar to that of the shorter analogue β-C8G1.14 Consequently, it displays a large micellar region, which, upon increasing surfactant content, is followed by a hexagonal phase, a bicontinuous cubic phase, and a lamellar phase. The hexagonal phase melts at 11 °C, so at temperatures above 11 °C, the micellar phase is directly followed by a bicontinuous cubic phase with the space group Ia3d. NMR self-diffusion data across the phase boundary between the cubic and micellar phase display no discontinuity, which infers the existence of some form of connected micellar network, at least close to the phase boundary.14 Whereas it thus seems clear that β-C9G1 forms threadlike micelles at high concentration, there is disagreement as to the micelle morphology in dilute solution. Isothermal calorimetry and differential scanning calorimetry (DSC) measurements suggest aggregation numbers consistent with spherical micelles at concentrations close to the critical micelle concentration (cmc),16 whereas NMR diffusion studies suggest that nonspherical aggregates are formed at concentrations immediately above the cmc.14 The nonspherical aggregates have been modeled both as prolates with an axial ratio of 11:1 by means of diffusion NMR14 and as a cylindrical core-shell structure using a combination of SAXS and SANS data.12 The SAXS and SANS study suggests much longer aggregates as compared to the NMR diffusion study. The first part of the present work is devoted to a detailed study of the influence of temperature and surfactant concentration on the self-aggregation and micelle morphology of β-C9G1, whereas the second part concerns the influence of deuterium oxide. Deuterium isotope effects are generally small in surfactant systems. For PEO-based surfactants, the critical temperature is typically lowered 2-4 K when H2O is replaced by D2O.17 However, replacing normal water with deuterium oxide has recently been shown to have unexpected and truly dramatic effects on micelles of some alkylglucosides. For instance, the twophase loop in the ternary system β-C9G1/β-C10G1/water (see above) grows dramatically when D2O is substituted for H2O.18 Similarly, the micelles of β-C8G1 are substantially larger in D2O than those in H2O, which has been explained in terms of one-dimensional growth.12 However, such a marked deuterium isotope effect is not found for all types of alkylglycosides. In β-C8G2, there is no such effect and the micelles have the same size both in D2O and in H2O.7 Since we expect the deuterium oxide effect to be relevant for the general understanding of alkylglycoside micellization, micelles of β-C9G1 were studied in both H2O and D2O in the present work. The final part of the work concerns the effects of various salts on micelles of β-C9G1 in terms of both the surface activity and the aggregation behavior of the surfactant. Alkylglycosides, as a class, have often been claimed to have a high tolerance toward salt, particularly when compared to PEO-based surfactants.19 The extensive investigation of salt effects on the surface tension of alkylmaltoside solutions reported by Zhang et al. substantiate this claim.19 The study revealed that the packing of the molecules at the air-water interface was not affected by the nature of the salt. However, the effects of cations on the surface activity and cmc of the surfactant were found to be different from those of anions. Furthermore, the effects of the salts were found to follow the well-known (16) Majhi, P. R.; Blume, A. Langmuir 2001, 17, 3844. (17) Martı´n, A.; Lesemann, M.; Belkoura, L.; Woermann, D. J. Phys. Chem. 1996, 100, 13760. (18) Whiddon, C. R.; So¨derman, O. Langmuir 2001, 17, 1803. (19) Zhang, L.; Somasundaran, P.; Maltesh, C. Langmuir 1996, 12, 2371.

Self-Aggregation of Nonylglucoside

Hofmeister series.20,21 Zhang et al. rationalize the comparably high salt tolerance of alkylmaltosides in terms of the strong hydration of the sugar headgroup.19 According to this idea, the strong hydration would make it difficult for electrolytes to interact with the headgroup. The explanation builds on earlier studies, which suggest that the strong hydration makes it difficult for neighboring molecules to approach the headgroup.22 Another reason for the salt tolerance has been claimed to be the rigidity of the sugar moieties, which is proposed to make it difficult for the electrolyte to penetrate the headgroup region.13 Both explanations essentially suggest the same thing, namely, that the electrolytes cannot penetrate the headgroup region thereby making the surfactant more tolerant toward electrolytes.19 Clearly, this would serve to rationalize the difference in salt effect between PEO- and sugarbased surfactants. However, the low electrolyte sensitivity of pure alkylglycosides does not seem to apply to technicalgrade alkylpolyglucosides (APG).23 These more complex systems show a pronounced tendency to phase separate at higher temperatures, and all inorganic salts have been found to induce a distinct reduction in the phase separation temperature (“cloud point”). It has been proposed that the effects are due to a small but significant negative charge of APG micelles, while micelles of ethoxylated surfactants are more or less uncharged.23

Langmuir, Vol. 20, No. 4, 2004 1403 primary beam. The two different instruments gave consistent results within limits of experimental error. Prior to measurement, all the samples were filtered through Acrodisc filters with a pore size of 0.1 or 0.2 µm. A number of samples were also re-examined over a time of several days in order to rule out kinetic effects or “cluster aggregation” of the type that has been previously observed for sugar surfactants with an amid linkage.24 Analysis of DLS Data. DLS data recorded on the Brookhaven instrument were analyzed by means of the software supplied with the instrument (Brookhaven DLS Software, ver. 2.13). The analysis is based on the method of cumulants first proposed by Koppel.25 In this method, each size is assumed to contribute its own exponential to the normalized time correlation function of the field, g(1)(τ), as expressed in eq 1.

∫

|g(1)(τ)| )

∞

0

G(Γ) exp( -Γτ) dΓ

(1)

Here, G(Γ) is the experimental time autocorrelation function of the scattered intensity. The exponential in eq 1 is then expressed as a polynomial in the delay time τ by means of a Taylor series expansion. The coefficients in the series expansion are the cumulants of the distribution. If G(Γ) is reduced, the first term in the Taylor expansion is (of course) unity, whereas the factors of the second term are given in eqs 2 and 3.

Γ)D h q2

(2)

µ2 ) (D 2 - D h 2)q4

(3)

Experimental Details Materials. The surfactant, n-nonyl-β-D-glucoside (β-C9G1) was purchased from Anatrace Inc. (Maumee, OH) and was of ANAGRADE quality. It was used without further purification. The supplier states a purity of >99%, and high-performance liquid chromatograms provided by the supplier confirm this claim. Solutions were prepared by dissolving carefully weighed amounts of surfactant powder in double distilled water, D2O, or salt solutions. All concentrations are given in grams of solute per liter of solution. The D2O used for dynamic light scattering experiments was purchased from Aldrich (Milwaukee, WI) with an isotopic purity stated as 100%. Surface tension experiments utilized D2O with an isotopic purity of 99% purchased from Cambridge Isotope Laboratories. The salts, NaCl, NaSCN, Na2SO4, NaI, NaNO3, KCl, KNO3, and CsCl were obtained from Merck (Darmstadt, Germany) and were of analytical grade except NaSCN, which has a purity of g98.5%. LiCl was purchased from Kebolab (Sweden) and was of purum quality. Methods. Tensiometry. A KSV Sigma 70 instrument equipped with a DuNouy ring made of platinum was used to determine the air/liquid surface tension as a function of surfactant concentration. The temperature of the system was controlled by a circulating water bath. At each measurement, the surface tension was recorded after a maximum of 1 h of equilibration. A small ( Cl- > NO3- > I- > SCN-, where I- and SCN- are saltingin anions. However, when comparing the data it should

Self-Aggregation of Nonylglucoside

Langmuir, Vol. 20, No. 4, 2004 1407

Figure 8. Surface tension vs surfactant concentration at 20 °C for β-C9G1 in water (filled circles), 1.5 M NaCl (open squares), and 1.5 M NaSCN (open circles). Table 2. cmc Values, Molecular Area, A0, and Surface Tension at Concentrations above cmc, γcmc, at 20 °C for β-C9G1, As Determined by Tensiometrya salt

cmc (g L-1)

A0 (Å2)

γcmc (mN m-1)

1.5 M NaSCN 1.5 M NaCl

2.12 1.59 0.80

45.6 53.4 41.8

30.4 32.4 29.6

a The molecular area is calculated from the Gibbs adsorption isotherm.

be borne in mind that the ionic strength of the divalent salt Na2SO4, at a given concentration, is higher than those of salts comprising monovalent ions. Indeed, if the effects are stated in terms of ionic strength, the chloride is the most efficient salting-out agent. In the system comprising 1.0 M of Na2SO4, a phase separation into two isotropic liquid phases was observed. The upper phase has a very small volume, which hampered further analysis. DLS studies of the lower phase indicate an effective micellar hydrodynamic diameter of >900 Å. The effects of NaCl and NaSCN on β-C9G1 were also investigated by tensiometry. Both salts were found to decrease cmc, although the effect was more pronounced for the salting-out anion Cl- (Figure 8, Table 2). Also, both salts were found to affect the slope of the surface tension isotherm (dγ/d ln c), as well as the surface tension above cmc (Figure 8, Table 2). Again, both effects follow the Hofmeister series. A similar investigation was performed for cations by studying the effects of chloride, nitrate, and sulfate salt of the alkali metals Li, Na, K, and Cs. In terms of micelle size, the effects of the salts decrease with increasing polarizability (increasing ionic radius) of the cation (Figure 9). In this respect, the effect of cations directly parallels that of the anions. Discussion The tensiometric cmc determinations in the present work clearly suggest that the unimer solubility of β-C9G1 is affected by temperature and salt, but not by a substitution of D2O for H2O. The difference in polarity between H2O and D2O often affects the hydrophobicity of a surfactant and commonly results in a somewhat lower cmc in D2O than in H2O.33 This effect has previously been observed for β-C8G1, where the cmc in H2O and D2O is (33) Kresheck, G. C. J. Am. Chem. Soc. 1998, 120, 10964.

Figure 9. The effect of cations in homologous series of salts on effective micellar hydrodynamic diameter in 10 g/L solutions of β-C9G1 at 25 °C. The dashed line represents the diameter at zero salt concentration.

25.2 and 22.3 mM, respectively.13,34 However, such relatively small differences may be within the limit of error of the tensiometric method employed in the present work. The decreasing unimer solubility with increasing temperature observed in the present study in the range 1040 °C corroborates previous DSC studies, which show the cmc of β-C9G1 to pass through a shallow minimum at ca. 40 °C upon increasing temperature.16 A similar minimum is often evident for anionic surfactants,16 whereas PEObased surfactants rather display a monotonic decrease in solubility with increasing temperature.5 For anionic surfactants, the minimum in cmc has been explained in terms of the temperature effects on the solubility of the hydrocarbon part in water,5 and the same explanation may be applicable to β-C9G1. Comparison of the cmc data with the data on micelle size leads to the somewhat surprising conclusion that both the micelle size and the unimer solubility of β-C9G1 decrease with increasing temperature in the range 1040 °C. Obviously, decreased unimer solubility (i.e., an increased overall hydrophobicity of the surfactant) would rather be expected to induce a micellar growth, so that surfactant-water contacts are decreased. Such a temperature-induced growth is also favored by the larger molar volume of the alkyl chain at higher temperatures, which increases the critical packing parameter (cpp). The expected correlation between aggregate size and unimer solubility is indeed observed for nonionic surfactants comprising poly(ethylene oxide) (PEO) as the hydrophilic part. When rationalizing the more complex behavior of β-C9G1, it is therefore illustrative to compare PEO-based and sugar-based surfactants in a more general context. For PEO-based surfactants, the temperature effects on cpp and those on unimer solubility both favor increased aggregate size upon increasing temperature. Furthermore, the temperature effects on the hydration of PEO are impossible to separate from the effects on headgroup conformation and size, since the collapse of the flexible PEO chain at higher temperatures is a direct reflection of the fact that water becomes a less good solvent for PEO with increasing temperature. In contrast, β-C9G1 carries a rigid, strongly hydrated headgroup, and temperatureinduced conformational changes may, consequently, be assumed to be marginal at best. Instead, we may envisage a situation where the hydration and hydrogen bonding of (34) Topgaard, D. Unpublished Work.

1408

Langmuir, Vol. 20, No. 4, 2004

Ericsson et al.

the headgroup determine the effective headgroup size. Thus, if increased temperature is assumed to increase the effective headgroup size by influencing hydration and hydrogen bonding, then this effect would tend to favor decreased micelle size at higher temperature, thus competing with the temperature effects on unimer solubility. In addition, we need to consider the general tendency of smaller aggregates to be entropically favored at higher temperatures. The idea that the micellar size of β-C9G1 is extremely sensitive to small changes in headgroup size finds support in the observation that an exchange of D2O for H2O has a profound influence on micelle size. Considering that the cmc is independent of solvent, within limits of experimental error, the higher polarity of D2O is unlikely to be the cause of the observed effects. Rather, as has previously been pointed out, an exchange of O-D bonds for the longer O-H bonds leads to a decrease in headgroup size and thus favors larger aggregates.18 The identical slope (dγ/d ln c) in the surface tension isotherms would seem to suggest that an exchange of O-H bonds for O-D bonds does not affect the headgroup size. However, the difference in size is expected to be a mere 2%,35 which is clearly too minute to allow for tensiometric determination. Nevertheless, previous results from the β-C9G1/β-C10G1 system show that it is quite possible for this minute change in headgroup size to affect micelle morphology in a dramatic way.35 The results from the study of salt effects on the micellization provide further evidence that headgroup size is more decisive than surfactant solubility for the micelle size in the β-C9G1 system. Both the salting-out anion Cland the salting-in anion SCN- decrease the cmc, yet the former increases the micelle size, whereas the latter makes it decrease. Again, we find no correlation between unimer solubility and micelle size. On the other hand, the salt effects on the headgroup size correlate with the effects on the micelle size, as does the surface tension. In a more general context, the results in this work corroborate the idea that salting-in/salting-out effects in surfactant systems are due primarily to adsorption/desorption of ions at the water-solute interface (which would influence headgroup size), rather than to effects on solvent “quality”.36

dimensionally with decreasing temperature. In this context, it is thus somewhat surprising that the unimer solubility (as determined by tensiometric measurements) increases with decreasing temperature. This anomalous relation between unimer solubility and micelle size is possible to understand if one assumes that temperature effects on headgroup size are more important for micelle morphology than the effects on unimer solubility. According to this line of reasoning, temperature effects on headgroup hydration and hydrogen bonding increase the headgroup size with increasing temperature, which tends to favor smaller aggregates. This effect counteracts and overruns the effects of decreasing unimer solubility, which would tend to favor larger aggregates at higher temperatures. Data from systems in which deuterium oxide was substituted for water support the idea that even moderate changes in headgroup size give rise to substantial effects on micelle size. The O-D bond is somewhat shorter that the O-H bond and the headgroup is therefore expected to be smaller in D2O than in H2O. This effect serves to explain the observation that the micelles are considerably larger in the former solvent. Tensiometric data show that it is less attractive to explain this pronounced deuterium effect in terms of unimer solubility, since the cmc is the same in the two solvents. The size of the micelles also shows a pronounced dependence of salt. The effects of salt follow the Hofmeister series SO42- > Cl- > Br- > NO3- > I- > SCN-. Here, Iand SCN- are salting-in anions which give rise to a moderate decrease in the micelle size, as compared with neat water. Similarly, the effects of cations on the micelle size decrease with increasing polarizability in the sequence Li+ > Na+ > K+ > Cs+. The salt effects on micelle size, cmc, and headgroup area corroborate the idea that headgroup size is more decisive than surfactant solubility for the micelle size of β-C9G1. The salting-out anion Cldecreases the cmc, increases the micelle size, and decreases the headgroup area at the air-water interface. The salting-in anion SCN-, on the other hand, also decreases the cmc but increases the headgroup area and decreases the micelle size. In other words, we again find that the micelle size correlates with headgroup area, rather than with monomer solubility.

Conclusions

Acknowledgment. This research has been funded by the Centre of Competence for Surfactants Based on Natural Products (SNAP) and AstraZeneca R&D Lund. Jo¨rgen Jansson is acknowledged for his assistance during DLS measurements.

Light scattering and neutron scattering data from dilute solutions are consistent with the idea that β-C9G1 forms elongated, fairly inflexible micelles with a circular cross section. The data also show that the micelles grow one(35) Whiddon, C. Green Colloid Chemistry. Lund University, 2003, pp 31-32. (36) Kabalnov, A.; Olsson, U.; Wennerstro¨m, H. J. Phys. Chem. 1995, 99, 6220.

Supporting Information Available: Comparison of SANS data for 10 and 50 g/L solutions at 20 °C. This material is available free of charge via the Internet at http://pubs.acs.org. LA035613E