pubs.acs.org/Langmuir © 2010 American Chemical Society
)
Surface and Solution Properties of Anionic/Nonionic Surfactant Mixtures of Alkylbenzene Sulfonate and Triethyleneglycol Decyl Ether I. Tucker,*,† J. Penfold,‡,§ R. K. Thomas,§ C. C. Dong,§ S. Golding,† C. Gibson,† and I. Grillo †
)
Unilever Research and Development Laboratory, Port Sunlight, Quarry Road East, Bebington, Wirral, U.K., ‡ ISIS Facility, STFC, Rutherford Appleton Laboratory, Chilton, Didcot, OXON, U.K., §Physical and Theoretical Chemistry Laboratory, Oxford University, South Parks Road, Oxford, U.K., and Institute Laue Langevin, 6 Rue Jules Horowitz, F-38042, Grenoble, Cedex 09, France Received February 27, 2010. Revised Manuscript Received April 14, 2010
The surface adsorption behavior and the solution microstructure of mixtures of the C6 isomer of anionic surfactant sodium para-dodecyl benzene sulfonate, ABS, with nonionic surfactant monodecyl triethyleneglycol ether, C10E3, have been investigated using a combination of neutron reflectivity, NR, and small-angle neutron scattering, SANS. In solution, the mixing of C10E3 and ABS results in the formation of small globular micelles over most of the composition range (100:0 to 20:80 ABS/C10E3). Planar aggregates (lamellar or unilamellar vesicles, ULV) are observed for solution compositions rich in the nonionic surfactant (>80 mol % nonionic). Prior to the transition to planar aggregates, the micelle aggregation number increases with increasing nonionic composition. The lamellar-phase region is preceded by a narrow range of composition over which mixtures of micelles and small unilamellar vesicles coexist. The variation in surface absorption behavior with solution composition shows a strong surface partitioning of the more surface-active component, C10E3. This pronounced departure from ideal mixing is not readily explained by existing surfactant mixing theories. In the presence of Ca2þ ions, a more complex evolution of solution phase behavior with solution composition is observed. The lamellar-phase region occurs over a broader range of solution compositions at the expense of the smallvesicle phase. The phase boundaries are shifted to lower nonionic compositions, and the extent to which the solutionphase diagrams are modified increases with increasing calcium ion concentration. The SANS data for the large planar aggregates are consistent with large polydisperse flexible unilamellar vesicles. In the presence of Ca2þ ions, the surface adsorption patterns become more consistent with ideal mixing in the nonionic-rich region of the surface-phase diagram. However, in the ABS-rich regions the surface behavior is more complex because of the spontaneous formation of more complex surface microstructures (bilayers to multilayers). Both in water and in the presence of Ca2þ ions the variations in the surface adsorption behavior and in the solution mesophase structure do not appear to be closely correlated.
Introduction Surfactants are commonplace ingredients of many medical, home, and personal care products. In solution, they self-assemble into a range of well-defined structures that define their fundamental rheological, dispersion, and colloidal properties.1,2 Mixtures of surfactants are usually preferred because of synergistic enhancements to many aspects of performance and behavior and greater flexibility in processing and formulation and because many commercial surfactants are inherently mixtures because of impurities arising from their synthesis. In recent years, the understanding of surfactant mixing both in solution and at interfaces has advanced greatly because of a combination of developments in modern experimental techniques3,4 and in advanced theoretical treatments.5,6 Nevertheless, there remain many aspects that are either poorly understood or relatively unexplored, especially where strong interactions occur and where significant departures from ideal mixing can arise. Hence, the study of surfactant *Corresponding author. E-mail:
[email protected] (1) Rosen, M. In Phenomena in Mixed Surfactant Systems; Scamehorn, J. F., Ed.; ACS Symposium Series 311; American Chemical Society: Washington, DC, 1988. (2) Scamehorn, J. F. In Mixed Surfactant Systems; Ogino, K., Abe, M., Eds.; Marcel Dekker: New York, 1992. (3) Penfold, J.; Tucker, I.; Thomas, R. K.; Staples, E.; Schuermann, R. J. Phys. Chem. B 2005, 109, 10760. (4) Lu, J. R.; Thomas, R. K.; Penfold, J. Adv. Colloid Interface Sci. 2000, 84, 143. (5) Holland, P. M.; Rubingh, D. N. In Cationic Surfactants; Rubingh, D. N., Holland, P. M., Eds.; Surfactant Science Series; Marcel Dekker: New York, 1990; Vol 37. (6) Puvvada, S.; Blankschtein, D. J. Phys. Chem. 1992, 96, 5567.
10614 DOI: 10.1021/la100846b
mixing, particularly under these extreme solution conditions, is still a rich and interesting area. Linear alkyl benzene sulfonates, ABS, are an important class of anionic surfactants. They are one of the most commonly used commercial surfactants and usually exist as a mixture of alkyl chain homologues with a range of headgroup positional isomers. In application, they are normally used in combination with a range of different ethoxylated nonionic surfactants. In spite of their importance and widespread use, there has been relatively little published on such surfactants. The exception to this is the recent work of Ma et al.7 and Penfold et al.8 In a recent paper, Ma et al.7 studied the surface and solution properties of the different positional isomers of ABS in detail. Penfold et al.8 studied the surface and solution behavior of mixtures of ABS with nonionic cosurfactants monododecyl octaethylene glycol, C12E8, and monododecyl triiscosaethylene glycol, C12E23. Although the role of the electrolyte in modifying surfactant adsorption and self-assembly and its effect on surfactant mixing has been extensively studied,3 the emphasis has been primarily on simple monovalent counterions. An exception to this is the study of the effect of aromatic counterions, such as benzoate, salicylate, and tosylate,9 where the impact on self-assembly can be particularly dramatic, inducing highly elongated structures at relatively (7) Ma, J.-G.; Boyd, B. J.; Drummond, C. J. Langmuir 2006, 22, 8646. (8) Penfold, J.; Thomas, R. K.; Dong, C. C.; Tucker, I.; Metcalfe, K; Golding, S.; Grillo, I. Langmuir 2007, 23, 10140. (9) Penfold, J.; Tucker, I.; Staples, E.; Thomas, R. K. Langmuir 2004, 20, 8054.
Published on Web 04/27/2010
Langmuir 2010, 26(13), 10614–10626
Tucker et al.
Article
low surfactant concentrations. The impact of multivalent counterions is also significant, as recently demonstrated by Alargova et al.,10 where the strong counterion binding induces substantial changes in micellar morphology. In the context of detergencybased applications, hard water (and hence the presence of Ca2þ and Mg2þ counterions) is known to have a profound effect on different aspects of performance. Penfold and co-workers have previously observed the effect of the preferred curvature of different nonionic cosurfactants on the phase behavior in ABS/C12E8 and ABS/C12E23 mixtures, in the presence and absence of calcium ions.8 In the presence of calcium ions, ABS forms a lamellar-phase dispersion in solution, which corresponds to the observations of pronounced multilayer formation at the air/water interface. The addition of low levels of the larger ethylene oxide-length surfactants produced a rich range of surface behavior from bilayers to multilayers, whereas the bulk solution remained a mixture of lamellar and micellar phases. Higher levels of ethoxylated surfactants were shown to suppress the formation of this surface complex and preserve the micellar form of the aggregates in solution. The motivation in this article is to extend this study to include a nonionic of similar alkyl chain length to ABS whose preferred curvature would tend to promote the formation of lamellar phases. The aim is to understand how that lower preferred curvature would modify the surface and solution behavior. Previous studies demonstrated the important role of multivalent counterions8 and the relative insensitivity to equivalent amounts of monovalent electrolyte. Hence the impact of multivalent counterions is also an important focus of this study. Small-angle neutron scattering, SANS, was used to investigate the variation in solution microstructure at relatively low surfactant concentrations and thus determine the nature of the selfassembly in solution. Neutron reflectometry, NR, was used, in combination with D/H isotopic substitution, to determine the amounts adsorbed and the composition and structure of the adsorbed layers at the air-solution interface and hence to determine the surface adsorption behavior directly.
Experimental Details Neutron Reflectivity. Specular neutron reflectivity measurements were made on the SURF reflectometer at the ISIS pulsed neutron source at the Rutherford Appleton Laboratory, U.K. The measurements were made using a single detector at a fixed angle, θ, of 1.5 and neutron wavelengths, λ, in the range of 0.5-6.8 A˚ to access a Q range (Q is the wave vector transfer normal to the surface, defined as Q = (4π/λ)sin θ, where λ is the neutron wavelength and θ is the grazing angle of incidence) of 0.048 to 0.5 A˚-1 using what are now well established experimental procedures.11 The basis of a neutron reflectivity experiment is that the variation in specular reflection, R(Q), with Q is related to the composition or density profile in a direction normal to the interface. In the kinematic or Born approximation, the reflectivity is related to the square of the Fourier transform of the scattering length density profile, F(z), normal to the interface,4 that is, 16π2 RðQÞ ¼ 2 Q
Z 2 - iQz dz FðzÞe
ð1Þ
P where F(z) = i ni(z)bi, ni(z) is the number density of the ith nucleus, and bi is its scattering length. The strength of this technique for studying mixtures of surfactants adsorbed at (10) Alargova, R. G.; Petkov, J. T.; Petsev, D. N. J. Colloid Interface Sci. 2003, 261, 1. (11) Penfold, J.; et al. J. Chem. Soc., Faraday Trans. 1997, 93, 3899.
Langmuir 2010, 26(13), 10614–10626
interfaces is the ability to substitute deuterium for hydrogen on part or all of the amphiphile and hence manipulate the scattering length density or neutron refractive index profile (the neutron refractive index is defined as n = (1 - λ2F(z))/2π) at the interface without any significant change in the physical properties of the amphiphile. The use of isotopic substitution in combination with NR has been powerfully demonstrated for the study of surfactant adsorption (for the determination of adsorbed amounts and surface structure) in a wide range of surfactants, surfactant mixtures, and polymer-surfactant mixtures.4 Furthermore, at the air/water interface D/H isotopic substitution can be used to index match the solvent to the air phase (null reflecting water, nrw, 8 vol % D2O/ H2O mixture with a refractive index of 1.0 for neutrons) such that the only contribution to the scattering at the interface arises from the deuterated surfactant that is adsorbed at the interface. This reflected signal can be analyzed in terms of the adsorbed amount at the interface and the thickness of the adsorbed layer. The most direct procedure for determining the surface concentration of surfactant is to assume that it is in the form of a single layer of uniform composition. The measured reflectivity can then be fitted by comparing it with a profile calculated using the optical matrix method for this simple structural model.12 The parameters obtained for such a model fit are the scattering length density, F, and the thickness, τ, of the layer. The area per molecule is then given by P A¼
bi
i
Fτ
ð2Þ
P where bi is the scattering length of the adsorbed surfactant molecule. In the case of a binary mixture, as studied here, eq 2 can be extended as P
P b1 b2 F¼ þ A1 τ A2 τ
ð3Þ
where bi and Ai are the scattering lengths and area/molecule of each component in the binary mixture. Making three different reflectivity measurements, with both surfactants deuteriumlabeled and with either of the two surfactants deuterium-labeled, provides a self-consistent estimate of the surface composition; a minimum requirement is two measurements with each component of the mixture selectively deuterium-labeled. For the neutron reflectivity data reported here, there are regions of surfactant concentration where the adsorbed layer is well described as a thin monolayer of uniform composition (density) and regions where the surface structure is more complex. In these cases, the simplest model consistent with the data is used to describe the surface structure. Where a pronounced interference fringe is observed, two or three layers are mostly sufficient to describe the data. In cases where a pronounced Bragg peak is observed, indicative that more extensive multilayer formation is occurring at the air-water interface, a more sophisticated approach is required to evaluate the surface structure. In the kinematic approximation and consistent with the approach of Tidswell et al.13 and Sinha et al.,14 the specular reflectivity for such a multilayer structure at the interface can be written as !2 N 2 16π2 X 2 σi RðQÞ ¼ 2 ðFi - Fi þ 1 Þ expð- iQdi Þ exp - Q Q i¼0 2 ð4Þ (12) Penfold, J. In Neutron, X-ray, and Light Scattering; Lindner, P., Zemb, T., Eds.; Elsevier: New York, 1991. (13) Tidswell, I. M.; Ocko, B. M.; Pershan, P. S.; Wasserman, S. R.; Whitesides, G. M.; Axe, J. D. Phys. Rev. B 1990, 41, 1111. (14) Sinha, S. K.; Sanyal, M. K.; Satija, S. K.; Majkrzak, C. F.; Neumann, D. A.; Homma, H.; Szpala, S.; Gibaud, H.; Morkov, H. Physica B 1994, 198, 72.
DOI: 10.1021/la100846b
10615
Article
Tucker et al.
where Fi is the scattering length density of the ith layer, i = 0 represents the subphase, di is the distance of the interface between the ith and (i þ 1)th layers from the subphase, di = Σi li, li is the thickness of the ith layer, σi is the roughness between the ith and (i þ 1)th layers, F(N þ 1) is the upper bulk phase (air), and N is the number of layers (N/2 is the number of bilayers). Small-Angle Neutron Scattering. SANS measurements were made on both the LOQ diffractometer15 at the ISIS pulsed neutron source at the Rutherford Appleton Laboratory and on D2216 at the Institute Laue Langevin, Grenoble, France. The measurements on LOQ were made using the white beam timeof-flight method in the wave vector transfer, Q, range of 0.008 to 0.25 A˚-1. The measurements on D22 were made using a wavelength of 8 A˚ (Δλ/λ ≈ 10%) and two different detector/collimation distance combinations (3.5/5.6 m, 16.5/17.6 m) to cover the Q range of ∼0.003 to 0.25 A˚-1. The samples were contained in Starna 1-mm-path-length quartz spectrophotometer cells and maintained at a temperature of 25 C. The data were corrected for background scattering, detector response, and the spectral distribution of the incident neutron beam and converted to an absolute scattering cross-section (I(Q) in cm-1) using standard procedures.17,18 In SANS, the scattering cross section or scattering intensity for colloidal aggregates in solution can be written as19 Z 2 3 IðQÞ ¼ N ðFp ðrÞ - Fs Þ exp iQr d r v
ð5Þ
where Fp and Fs are the aggregate and solvent scattering length densities and N is the number of aggregates per unit volume. In the micellar phase, the micelle structure is determined by analyzing the scattering data using a standard, well-established model for globular micelles.19 For a solution of globular polydisperse interacting particles (micelles), the scattering intensity can be written in the decoupling approximation19 as h i IðQÞ ¼ n SðQÞjÆFðQÞæQ j2 þ ÆjFðQÞj2 æQ - jÆFðQÞæQ j2
ð6Þ
where the averages denoted by ÆQæ are averages over particles size and orientation, n is the micelle number density, S(Q) is the structure factor, and F(Q) is the form factor. The micelle structure (form factor) is modeled using a standard core þ shell model,20 where the form factor is FðQÞ ¼ V1 ðF1 - F2 Þ F0 ðQR1 Þ þ V2 ðF2 - Fs Þ F0 ðQR2 Þ
ð7Þ
4πRi3/3, 3
and R1 and R2 are the core and shell radii, Vi = F0(QRi) = 3j1(QRi)/(QR) = 3[sin(QR) - QR cos(QR)]/(QR) , F1, F2, and Fs are the scattering length densities of the micelle core and shell and of the solvent, and j1(QRi) is a first-order spherical Bessel function. The micelle core þ shell model19 comprises an inner core made up of the alkyl chains only and is constrained to space fill a volume limited by a radius, R1, and the fully extended chain length of the surfactant, lc. For larger aggregation numbers, ν, volumes greater than that defined by R1 (as is found in this study) are accommodated by a prolate elliptical distortion with dimensions of R1, R1, and eR1 (where e is the elliptical ratio). The outer shell, with dimensions of R2, R2, and eR2, contains the headgroups and the (15) Heenan, R. K.; King, S. M.; Penfold, J. J. Appl. Crystallogr. 1997, 30, 1140. (16) Neutron beam facilities at the high-flux reactor available for users, ILL, Grenoble France, 1994. (17) Heenan, R. K.; King, S. M.; Osborn, R.; Stanley, H. B. RAL Internal Report; RAL-89-128, 1989 (18) Ghosh, R. E.; Egelhaaf, S. U.; Rennie, A. R. ILL Internal Report; ILL 98GH14T, 1998 (19) Hayter, J. B.; Penfold, J. Colloid Polym. Sci. 1983, 261, 1032. (20) Thomas, R. K.; Golding, S. Private communication, 2009.
10616 DOI: 10.1021/la100846b
corresponding water of hydration. Representative hydration values for the EO headgroup, the cation, and the bound counterions are included as fixed values, and the modeling is not particularly sensitive to variations in hydration. From the known molecular volumes and neutron scattering lengths, the scattering length density (F) for the core, shell, and solvent can be estimated.19 For the mixtures, the two surfactant components in the binary mixture are accommodated by assuming ideal mixing, which has been shown to be consistent with previous observations for concentrations well in excess of the mixed cmc.21 The interparticle interactions are included using the rescaled mean spherical approximation, RMSA, calculated for a repulsive screened Coulombic potential22,23 defined by the surface charge, z, the micelle number density, n, the micelle diameter, and the DebyeHuckel inverse screening length, κ-1.22 The model parameters refined are then ν, z, and e, and an acceptable model fit requires the shape of the scattering to be reproduced and the absolute value of the scattering intensity to be predicted to within (10%. For mixtures containing components with a more complex geometry than the single alkyl chain, the simple constraints described above are not sufficient, and this was previously observed for the dialkyl chain cationic/nonionic surfactant mixtures.24,27 To accommodate the disruption to the simple packing arguments, an additional model parameter, ext, is included. This allows a modification to the constraint that the inner dimension of the micelle, R1, is limited by the extended alkyl chain length such that it can be greater or smaller than that value. Although much of the scattering data is in the micellar phase, in the presence of CaCl2 there are mixed phase (micellar/lamellar) and lamellar regions in the phase behavior. This data has not been analyzed quantitatively, but the form of the data has been used qualitatively to identify the form of the microstructure and to distinguish between pure one-component and mixed-phase regions to try to correlate the solution and surface behaviors. There are regions where small unilamellar vesicles exist, not micelles or a lamellar phase. A quantitative analysis of the small unilamellar vesicles, denoted Lsv, was made using a core þ shell model, where the scattering is also defined by eqs 6 and 7 (similar to that used for micelles but without the model constraints specific to the micelle structure). In this case, the core þ shell model has a solvent core (D2O) and a surfactant mixture shell. The key model parameters, in this case, are the inner and outer radii, R1 and R2, the polydispersity, σ, the surface charge, z, and the scattering length density of the shell, F2. Materials and Measurements Made. The 6-isomer of ABS, sodium para-dodecyl benzene sulfonate, was custom synthesized at Oxford/Unilever R & D20 in two isotopic forms, that is, with and without the dodecyl alkyl chain and deuterium-labeled benzene ring, and is termed d-ABS and h-ABS, respectively. The purity of ABS was verified from cmc measurements derived from surface tension measurements and using neutron reflectivity. Surface tension data (measured on a Kruss K10T maximum pull digital tensiometer with a du Nouy ring) gave a cmc for ABS of ∼1.5 mM, which is broadly consistent with other reported literature values,7 and the surface tension data showed no pronounced minimum that would be a clear signature of impurities. The protonated C10E3 was obtained from Nikkol and used without further purification, and the deuterated C10E3 was custom synthesized and purified by Dr. R. K. Thomas using established procedures. All of the solutions for the SANS measurements were made in D2O, which was obtained from Fluorochem. The solutions for the neutron reflectivity measurements were made in nrw, and high-purity water (Elga Ultrpure) was used with D2O. Analytical-grade (>99.9% purity) CaCl2 3 2H2O was used for (21) Staples, E.; Penfold, J.; Thompson, L.; Tucker, I.; Hines, J.; Thomas, R. K.; Lu, J. R. Langmuir 1995, 11, 2479. (22) Hayter, J. B.; Penfold, J. Mol. Phys. 1981, 42, 109. (23) Hayter, J. B.; Hansen, J. P. Mol. Phys. 1982, 42, 651. (24) Tucker, I.; Penfold, J.; Thomas, R. K.; Bradbury, R.; Grillo, I. Langmuir 2009, 25, 4943.
Langmuir 2010, 26(13), 10614–10626
Tucker et al.
Article
the solutions in electrolyte. All glassware and sample cells were cleaned using alkali detergent (Decon 90), followed by copious washing in high-purity water. SANS measurements were made at concentrations of 25, 12.5, and 2.5 mM in the composition range of 100:0 to 0:100 mol % C10E3 in the absence and presence of 0.5, 1, and 2.0 mM CaCl2 using the hydrogenated versions of both surfactants in D2O to provide a contrast. Neutron reflectivity measurements were made on 2 mM solutions of surfactants in nrw using either chaindeuterated ABS and hydrogenated C10E3 or hydrogenated ABS and chain-deuterated C10E3, termed dh and hd, respectively.
Results and Discussion Solution Microstructure. The solution microstructure of the ABS/C10E3 mixtures has been determined using data from SANS measurements in the presence and absence of CaCl2. Measurements were made at surfactant concentrations of 2.5, 12.5, and 25 mM for compositions from 100:0 to 0:100 in D2O and at 20 mol % intervals in the presence and absence of 2 mM CaCl2. In the Absence of Electrolyte. Figure 1 shows the scattering from 25 mM solutions of ABS/C10E3 mixtures, where the solution composition varied from 100:0 to 0:100 ABS/C10E3 at 20 mol % intervals. Similar data were obtained for 12.5 and 2.5 mM surfactant concentrations and are not presented here. In the composition range of 100:0 to 20:80 ABS/C10E3, the data are consistent with relatively small interacting globular micelles whereas the data for 100% C10E3 have a Q-2 dependence and are consistent with planar structures. In the micellar region, the data are well described using the micelle model described earlier, and typical model fits for the 25 mM data are shown in Figure 2. The key model parameters are summarized in Table 1. At all three of the concentrations measured, the addition of the nonionic surfactant induces micelle growth and the micelles become progressively more elliptical as the solution composition becomes richer in C10E3. The scattering for 100% C10E3 has no pronounced features other than the Q-2 dependence associated with planar structures, and hence no detailed quantitative analysis was possible. For the data at the lower solution concentrations (12.5 and 2.5 mM) in the micellar region, model fits similar to those made at 25 mM were obtained, and the key model parameters are listed in Table S1 in the Supporting Information. The variation in aggregation number obtained from this quantitative analysis of the experimental data is shown in Figure 3. The increase in the micelle aggregation number with the increasing addition of C10E3 is similar for all three surfactant concentrations measured. Over most of the composition range, the aggregation number does not appear to vary significantly with solution concentration (within an experimental error of (5). The exceptions to this are the measurements at the highest nonionic content (80 mol % C10E3) where the aggregation number increases with increasing surfactant concentration. There are two main contributing factors to the increase in the aggregation number as the solution becomes richer in C10E3. C10E3 has a lower preferred curvature, and its combination with ABS will hence induce micellar growth toward more elongated and eventually planar structures. In addition, the incorporation of the nonionic surfactant reduces the impact of the interheadgroup electrostatic interactions of the ABS, and this further promotes growth. This has been previously observed and reported in other ionic/nonionic surfactant mixtures.3 To explore in more detail the marked transition between micellar and planar scattering between solution compositions of 20:80 and 0:100 ABS/C10E3, a further series of measurements Langmuir 2010, 26(13), 10614–10626
Figure 1. Scattering intensity I(Q) (cm-1) vs wave vector transfer,
Q, (A˚-1) for 25 mM ABS/C10E3 in D2O over the composition range of 100:0 (black), 80:20 (red), 60:40 (yellow), 40:60 (green), 20:80 (cyan), and 0:100 (blue).
Figure 2. Scattering intensity, I(Q) (cm-1) for 25 mM ABS/C10E3 in D2O for (red) 100:0 and (blue) 40:60. The solid lines are model fits to the experimental data using the model for interacting globular micelles and the model parameters summarized in Table 1. Table 1. Model Parameters for 25 mM ABS/C10E3 Mixed Micelles solution R2 ext ee scale composition R1 (mole% ABS) ν ((5) z ((2) ((1 A˚) ((1 A˚) ((0.1) ((0.1) factor 100 90 80 60 40 20 0
39 45 52 72 102 247 Lv
17 19 17 20 24 39
14 14 14 15 16 18
16 17 17 18 20 22
1.1 1.1 1.1 1.1 1.2 1.3
1.6 1.6 1.8 1.9 2.0 3.4
1.1 1.1 1.0 1.0 1.0 0.9
were made at concentrations of 25, 18, and 12.5 mM over the composition range of 30:70 to 0:100 ABS/C10E3 and at 5% composition intervals between 0:100 and 20:80. Figure 4 shows the evolution of the scattering for 25 mM surfactant over this composition range. The data for solution concentrations of 18 and 12.5 mM were similar and are not shown here. This data shows a more complex evolution of the form of the scattering with composition over this narrower range of DOI: 10.1021/la100846b
10617
Article
Figure 3. Variation in aggregation number with solution composition for ABS/C10E3 mixtures at solution concentrations of 2.5, 12.5, and 25 mM.
Tucker et al.
Figure 5. Scattering data for (red) 25 mM 5:95 and (blue) 12.5 mM 5:95 ABS/C10E3. The solid lines are fits to the experimental data using the small unilamellar vesicle model, as described in the text, using the parameters summarized in Table 2. Table 2. Model Parameters for the Small Vesicles for ABS/C10EO3 Mixtures
Figure 4. Evolution of bulk solution scattering for 25 mM ABS/ C10E3 mixtures: (black) 0:100, (red) 5:95, (yellow) 10:90, (green) 15:85, (blue) 20:80, and (cyan) 30:70. Data are shifted vertically for clarity.
composition. At a solution composition of 15:85, the scattering is still micellar. However, at a solution composition of 10:90 there is a significant change in the form of the scattering. The scattering data for ABS/C10E3 solution compositions of 30:70, 20:80, and 15:85 in Figure 4 are consistent with small globular interacting micelles, as previously described. At a solution composition of 0:100, the scattering data has a Q-2 dependence consistent with planar aggregates, as discussed earlier in the context of Figure 1. The absence of modulations and Bragg structures in the Q-2 dependence suggests that there are no significant correlations between any adjacent lamellae in the planar structure. Hence, it is most likely that the scattering arises from relatively large, polydisperse, flexible unilamellar vesicles. The scattering data for ABS/C10E3 solution compositions of 5:95 and 10:90 have a different form. Although superficially similar to micellar scattering, the scattering arises from larger globular interacting aggregates. This is indicated by the pronounced minimum in the scattering at intermediate Q values (∼0.05 to 0.07 A˚-1) that is absent in the micellar scattering. This implies a larger size than is physically reasonable for globular micelles, and the well-defined minimum implies a relatively low 10618 DOI: 10.1021/la100846b
surfactant concentration (mM)
solution composition (mole fraction nonionic)
25.0 25.0 12.5 12.5
0.95 0.90 0.95 0.90
R2 polydispersity, R1 ((1 A˚) ((1 A˚) σ ((0.05) 43 32 38 25
70 43 71 35
0.15 0.25 0.16 0.30
polydispersity. Consistent with recent observations in related systems,24-26 the data is attributed to the formation of small, relatively monodisperse unilamellar vesicles (i.e., nanovesicles). Figure 5 shows the scattering data for 25 mM 5:95 ABS/C10E3 and 12.5 mM 5:95 ABS/C10E3 and the associated model fits for small unilamellar vesicles. The model is similar to the core þ shell model used for the micellar scattering (eqs 6 and 7), where the inner core comprises solvent and the shell is the mixed ABS/C10E3 bilayer. The key model parameters (summarized in Table 2) are the inner and outer radii, R1 and R2, the polydispersity, σ, (modeled here as a Schultz distribution), the surface charge, z, (which in this case, because the solution compositions where the small vesicles form are predominantly nonionic, is close to zero), the scattering length density of the shell, F, and the concentration of vesicles in solution. The concentration of vesicles and F are constrained by the overall surfactant concentration and composition and by assuming space filling in the vesicle bilayer. In this narrow region of composition where the small monodisperse vesicles are formed, the overall size (R2) varies from ∼40 to 70 A˚ and the polydispersity is in the range of 0.15 to 0.3. This is comparable to the dimensions of the small vesicles formed by DHDAB/C12E12/alcohol mixtures24 and by the DDAB/C12E4 mixtures.25 From the qualitative interpretation and the quantitative analysis (where it has been possible) of the scattering data, a solutionphase diagram (at the relatively low surfactant concentrations measured) for the ABS/C10E3 mixture has been determined, and this is shown in Figure 6. (25) Grillo, I.; Penfold, J.; Tucker, I.; Cousin, F. Langmuir 2009, 25, 3932. (26) Oberdisse, J.; et al. Langmuir 1996, 12, 12.
Langmuir 2010, 26(13), 10614–10626
Tucker et al.
Figure 6. Bulk solution-phase diagram for ABS/C10E3 mixtures at 25 C. Each dot represents a point at which a SANS measurement was made.
Figure 7. Scattering data for a 25 mM ABS/C10E3 mixture in the presence of 2 mM CaCl2: (black) 100:0, (red) 90:10, (yellow) 80:20, (green) 60:40, (blue) 40:60, (cyan) 10:90, and (pink) 0:100.
In the Presence of CaCl2. The effect of the addition of CaCl2 on ABS/C10E3 mixtures was investigated using SANS for CaCl2 concentrations of 0.5, 1.0, and 2.0 mM and for surfactant concentration of up to 25 mM. Figure 7 shows the SANS data for 25 mM ABS/C10E3 mixtures in 2 mM CaCl2. The form of the scattering in 2 mM CaCl2 shown in Figure 7 is very different from that shown in Figure 1 in the absence of electrolyte and demonstrates that the addition of divalent counterions has a significant impact upon the form and evolution of the aggregate structure. At 25 mM ABS/C10E3/2 mM CaCl2 (Figure 7), there is now a Q-2 component to the scattering at all solution compositions (from 100:0 to 0:100), consistent with planar structures at all compositions. In the absence of CaCl2, planar structures were present only for 100% C10E3. At 100% ABS, the scattering also has a significant micellar contribution in addition to the Q-2 component, so the microstructure is in the form of lamellar/ micellar coexistence. From previous arguments, we attribute the Q-2 component of the scattering to unilamellar vesicles. With increasing nonionic contribution (in the composition range from 90:10 to 40:60), there is an increasing contribution to the scattering from the planar aggregates. However, there Langmuir 2010, 26(13), 10614–10626
Article
remains evidence of a micellar contribution at high Q values, and this mixed phase behavior is assigned as Lv/L1. At these solution compositions, the scattering at the lower Q values also has some effect on the Q-2 dependence. Consistent with previous recent studies on vesicle structure,24,25,27-30 this is interpreted as arising from relatively rigid bilamellar vesicles, blv, or multilamellar vesicles, mlv. At compositions even richer in C10E3, the scattering is still Q-2 but there is no longer any evidence of a coexisting micellar contribution. Furthermore, here the Q-2 scattering dependence has little or no structure (in contrast to that observed at compositions less rich in C10E3) and (following previous arguments) is consistent with large polydisperse unilamellar vesicles that have a high degree of flexibility. At the lower surfactant concentration of 12.5 mM, the trend toward the formation of relatively flexible planar structures is even more pronounced (data not shown). Apart from 100% ABS and 90:10 ABS/C10E3, the data is entirely Q-2 with no pronounced interference features and is hence also attributed to the scattering from large, flexible polydisperse unilamellar vesicles. For 100% C10E3 at 12.5 mM and for all ABS/C10E3 compositions at 2.5 mM, the scattering still has a Q-2 form but is of reduced intensity, and this is indicative of the onset of precipitation. The lack of pronounced features in the Q-2 dependence of the scattering in the region of predominantly planar structures and the extensive vesicle/micelle coexistence precludes any detailed quantitative analysis in those regions. Hence, we restrict the analysis to purely a determination of the variation in the phase behavior in those regions. The one exception to this is the scattering for 25 mM 80:20 ABS/C10E3 in 2 mM CaCl2. This is in the Lv/L1 coexistence region but is predominantly Lv. For this scattering, there are pronounced interference oscillations on the Q-2 dependence of the scattering at the lower Q values. These are associated with a relatively rigid structure arising from bilamellar or multilamellar vesicles.27-29 Here we have made an approximate analysis by combining the core þ shell micellar model (described earlier) with the approach developed by Nallet et al.30 to describe the lamellar/vesicle scattering contribution. In the Nallet formulation, which is described in detail elsewhere,30 the lamellar scattering is accounted for by dσ 2π 1 PðQÞ SðQÞ ðQÞ ¼ dΩ d Q2
ð8Þ
which takes into account the lamellar form factor, P(Q), and the structure factor, S(Q), and includes membrane fluctuations and resolution and assumes “powderlike” averaging; d is the bilayer spacing. The membrane rigidity is defined by the Caille parameter, η,31 η¼
Q0 2 kT pffiffiffiffiffiffiffiffi 8π KB
ð9Þ
B and K are the bilayer compressibility and bending modulus of the bilayer assembly, Q0 = 2π/d, and d = δ/j (where j is the volume fraction and δ is the bilayer width). Figure 8 shows the approximate quantitative analysis of the scattering data for 25 mM 80:20 ABS/C10E3 in 2 mM CaCl2. (27) Tucker, I.; Penfold, J.; Thomas, R. K.; Grillo, I.; Barker, J. G.; Mildner, D. F. R. Langmuir 2008, 24, 6509. (28) Tucker, I.; Penfold, J.; Thomas, R. K.; Grillo, I.; Barker, J. G.; Mildner, D. F. R. Langmuir 2008, 24, 10089. (29) Tucker, I.; Penfold, J.; Thomas, R. K.; Grillo, I.; Barker, J. G.; Mildner, D. F. R. Langmuir 2008, 24, 7674. (30) Nallet, F.; Laversanne, R.; Roux, D. J. Phys. II 1993, 3, 487. (31) Caille, A. C. R. Hebd. Sci. Acad. Sci., Ser. B 1972, 274, 891.
DOI: 10.1021/la100846b
10619
Article
Figure 8. Scattered intensity, I(Q) (cm-1), vs Q for 25 mM 80:20 ABS/C10E3 in 2 mM CaCl2. The solid line is a model fit for coexisting bilamellar vesicles and micelles (as described in the text).
The form of the scattering is consistent with that of a relatively fluid bilayer. To confirm that the Nallet model was applied to determine the magnitude of the Caille parameter, using only the assumption that a bilayer alone was responsible for the undulations in the SANS data and therefore the number of layers, N was set equal to 2 in the expression for S(Q).30 The key model parameters used to describe the lamellar contribution were d = 380 A˚, δ = 32 A˚ (calculated from the high-Q-region scattering of 100:0 ABS/C10E3) (N = 2) (bilamellar), and η ≈ 0.7. A micellar contribution to the scattering (with an aggregation number of υ = 40 and a surface charge of z = 10) was included, as discussed earlier, to provide an adequate description of the data at high Q. The Caille parameter is larger than the value normally encountered in the more rigid membrane structures.27-30 It is more consistent with the values obtained for the more flexible nonionic-based membrane structures. For example, Safinya et al.32 quote values for the fluctuation-stabilized C12E5 membranes in the range of 0.3-1.5. This is also consistent with the observation of the complete absence of structure in the data for compositions more rich in C10E3, which corresponds to even more flexible structures. SANS measurements were also made at surfactant concentrations of 12.5 and 25 mM at lower CaCl2 concentrations of 0.5 and 1.0 mM. The data measured at a surfactant concentration of 25 mM and in 0.5 mM CaCl2 are shown in Figure 9. In 0.5 mM CaCl2, the solution microstructure is initially micellar for 100% ABS. As the nonionic composition increases, there is a pronounced change in the form of the scattering. At a composition of 80:20, the scattering is strongly Q-2 but with a micellar component still evident at high Q values and hence corresponds to lamellar/micellar coexistence. At compositions of 60:40 and 40:60, the data are broadly similar to that at 80:20. However, the scattering for a solution composition of 20:80 is significantly different and is characteristic of the scattering from small unilamellar vesicles, as discussed earlier. Finally, at a composition of 0:100 the scattering has a purely Q-2 dependence, consistent with the scattering from planar objects. In the composition range of 80:20 to 40:60, the interference fringes visible in the data at low Q are consistent with relatively rigid bilamellar/ multilamellar vesicles, but there also remains a significant micellar contribution. (32) Safinya, C. R.; Sirota, E. B.; Roux, D.; Smith, G. S. Phys. Rev. Lett. 1989, 62, 1134.
10620 DOI: 10.1021/la100846b
Tucker et al.
Figure 9. SANS scattering data for 25 mM ABS/C10E3 mixtures in the presence of 0.5 mM CaCl2 in D2O: (red) 100:0, (blue) 80:20, (green) 60:40, (yellow) 40:60, (cyan) 20:80, and (black) 0:100. The data are shifted vertically for clarity.
In 1.0 mM CaCl2, the scattering for solutions in the composition range of 40:60 to 0:100 has a more dominant Q-2 dependence than at the lower CaCl2 concentration. This is consistent with the onset of planar lamellar structures occurring at lower nonionic compositions. At compositions of 100:0 and 80:20, the scattering still retains a micellar scattering component. At a composition of 100:0, it is still completely micellar, and at a composition of 80:20, the lamellar component is more abundant than at the lower CaCl2 concentrations. For a surfactant concentration of 25 mM and in 0.5, 1.0, and 2.0 mM CaCl2, there is a region of ABS/C10E3 compositions where the Q-2 -dependent scattering arising from planar structures in solution has additional structure. This is, as discussed earlier, consistent with strong correlations between adjacent lamellae in bilamellar or multilamellar vesicles. Between solution compositions of 80/20 and 40/60, there is vesicle/micelle coexistence with resolvable structure in the Q-2 vesicle component of the scattering. Because of the micellar contribution to the overall scattering, a reliable quantitative analysis (using eq 9) is not generally possible. In Figure 8, we have illustrated an approximate analysis incorporating both the vesicle and micellar contributions for 25 mM 80:20 ABS/C10E3 in 2 mM CaCl2, and this gives some indication of the parameters associated with the vesicles in this region. However, making an approximate estimation of the lamellar spacing from the data, using d ≈ 2π/Q0, illustrates an interesting trend in the lamellar spacing with surfactant composition and CaCl2 concentration, as illustrated in Figure 10. With increasing nonionic content, the d spacing of the bilamellar or multilamellar vesicle structure increases. The variations with surfactant composition in 0.5 and 1.0 mM CaCl2 are similar in magnitude, as is their dependence on solution composition. For 2.0 mM CaCl2, a much stronger dependence on solution composition is observed. Superficially, this dependence could imply that the increasing addition of C10E3 more effectively dilutes the lamellar structure and that the effect is most pronounced in 2 mM CaCl2 than for the lower CaCl2 concentrations. However, the increase in the d spacing is entirely consistent with the increasing flexibility of the membrane as C10E3 is added. If the membrane structure is increasingly dominated by repulsive interactions arising from Helfrich fluctuations, as demonstrated by Safinya et al.32 and others in related systems, then as the transition from electrostatically stabilized to fluctuation-stabilized membranes occurs, d will increase. Langmuir 2010, 26(13), 10614–10626
Tucker et al.
The qualitative and quantitative interpretations of all of the scattering data at these relatively low surfactant concentrations have been assembled into the form of phase diagrams, as shown in Figure 11.
Figure 10. Variation in lamellar spacing with solution composition (mole fraction of nonionic surfactant) for 25 mM ABS/C10E3 in 0.5, 12.0, and 2 mM CaCl2.
Article
The phase behavior in 2 mM CaCl2 and at low surfactant concentrations appears to contradict the previously observed phase behavior.8 However, it should be noted that in the current work samples tended to precipitate far more readily than was previously observed with the higher-molecular-weight ethoxylate nonionic surfactant. This was assigned as a LR phase in the previous study for 100% ABS, primarily because of its optical texture. In fact, from the more detailed measurements made here it is evident that under these conditions precipitation effects are likely Surface Adsorption. Neutron reflectivity measurements were made for ABS/C10E3 mixtures in nrw and in 0.5 and 1 mM CaCl2 for the isotopic combinations d-ABS/h-C10E3 and h-ABS/d-C10E3 and at 2 mM total surfactant concentration. Measurements were made for solution composition of 95:5, 90:10, 80:20, 70:30 60:40, 50:50, 40:60, and 20:80 and for the pure ABS and C10E3 components. In these measurements, the hydrogeneous component is effectively matched to the solvent (nrw) and only the deuterated component contributes to the reflectivity. The reflectivity curves obtained in the absence of CaCl2 were in the form of monolayers, and typical data are shown in Figure 12 for the two different contrasts. The data are well described as a thin monolayer of uniform composition ∼20 ( 1 A˚ thick. The data were analyzed as described earlier to provide a thickness and scattering length
Figure 11. Phase diagrams for ABS/C10E3 in 0.5, 1, and 2.0 mM CaCl2. Each spot marks a point in the composition, concentration space where a scattering measurement was made. Langmuir 2010, 26(13), 10614–10626
DOI: 10.1021/la100846b
10621
Article
Tucker et al.
Figure 12. Neutron reflectivity for 2 mM ABS/C10E3 for solution compositions of (a) 90:10 and (b) 20:80 and isotopic combinations (red) dh and (blue) hd. Table 3. Model Parameters for Monolayer Surface Adsorption for 2 mM ABS/C10E3 in Null Reflecting Water solution composition, mole fraction of C10E3 0.0 0.05 0.10 0.20 0.3 0.40 0.5 0.60 0.80 1.0
Γ ( 10-10 mol cm-2)
Γtotal ((0.3)
surface composition, mole fraction of C10E3 ((0.02)
54
3.1
3.1
0.0
2.2 1.2 2.2 1.2 2.2 1.4 1.8 1.8 1.6 1.9 1.6 2.0 1.3 2.3 0.8 2.5
69 ( 3 80 ( 8 73 94 82 73 91 63 99 59 117 54 139 50 230 41
2.41 2.0 2.3 1.8 2.0 2.3 1.8 2.6 1.7 2.8 1.4 3.1 1.2 3.3 0.7 4.1
4.4
0.45
4.1
0.43
4.3
0.53
4.5
0.59
4.5
0.63
4.5
0.68
4.5
0.74
4.8
0.85
3.4
34
4.9
4.9
contrast
d ((1 A˚)
F ( 10-6 A˚-2) ( 0.1
dh hd dh hd dh hd dh hd hd dh dh hd dh hd dh hd dh hd dh hd
21
3.1
23 24 21 20 20 24 21 21 22 21 19 21 19 20 19 23 21
1.0
density from which (using eq 2 or 3) an adsorbed amount and surface composition were obtained. The key model parameters, adsorbed amounts, and surface compositions are summarized in Table 3. The total adsorption and amount of each surfactant adsorbed obtained from the data in Table 3 are presented in Figure 13a, and the variation in surface composition with solution composition is shown in Figure 13b. The variation in the total adsorption and in the surface composition with solution composition shows some interesting trends. From ABS-rich to C10E3-rich, the total adsorption increases (Figure 13a), and this in part reflects the differences in the total adsorption of the pure ABS and C10E3 components. At 2 mM, ABS has an area/molecule of ∼55 A˚2 whereas C10E3 has an area/molecule of ∼34 A˚2. Hence for ideal mixing and in the absence of any nonideal mixing and synergistic enhancements of the adsorption, that general trend would be expected. The variation in surface composition with solution composition (Figure 13b) shows that the behavior is quite extreme. The data are consistent with a pronounced deviation from the solution composition such that over most of the composition range and especially for solutions rich in ABS the surface is dominated by 10622 DOI: 10.1021/la100846b
A (A˚2)
the more-surface-active component, C10E3. At concentrations just above the mixed cmc, it would be expected, from both nonideal and ideal mixing, that the surface would be rich in the more-surface-active component, in this case, C10E3. With increasing concentration, the surface composition will evolve toward the solution composition (represented by the diagonal line in Figures 13b and 14). The detailed nature of the trends with concentration and composition will depend upon the cmc of the individual components, the mixed cmc, and the degree of departure from ideal mixing (as described in RST by an interaction parameter, β). The data presented here are broadly consistent with the general qualitative observations of surfactant mixing theories.5 What is different here is the extreme nature of the dominance of the C10E3 component at the surface. The cmc of C10E3 is ∼3 10-5 M, and that of LAS is ∼1.5 mM in the absence of electrolyte. Hence, even for a modest negative interaction parameter, the mixed cmc over most of the composition range will be far lower than the concentration of the measurements (2 mM), and this will be especially true in the presence of electrolyte. In Figure 13b, the surface composition, calculated from RST using β = -2.0 and -4.0, is included. A comparison of the data with those calculated curves demonstrates Langmuir 2010, 26(13), 10614–10626
Tucker et al.
Article
Figure 13. (a) Variation in the total and the amount of each surfactant adsorbed at the air-water interface for 2 mM ABS/C10E3 mixtures with solution composition. (b) Surface composition versus solution composition (circles) together with predictions of RST for β values of -2kBT and -4kBT and -2 for Ca-representative cmc’s.
Figure 14. Variation in surface composition with solution composition for 2 mM ABS/C10E3 mixtures in the absence and presence of 0.5 and 1.0 mM CaCl2.
the extreme nature of the surface mixing. Hence in detail the qualitative variation in the surface composition is not consistent with ideal mixing or with existing theoretical treatments of nonideal mixing based on RST.5 The neutron reflectivity measurements were made for 2 mM ABS/C10E3 mixtures in 0.5 and 1.0 mM CaCl2. Because the SANS data for ABS/C10E3/2 mM CaCl2 indicated precipitation effects, reflectivity measurements were not made at this higher CaCl2 concentration. For solution compositions of 60:40 ABS/C10E3 and richer in C10E3, the reflectivity data were consistent with the adsorption of a monolayer at the surface, similar to that observed in the absence of CaCl2. The data were analyzed in the same way to obtain a thickness and scattering length density, from which the adsorbed amount and surface composition were determined (using eqs 2 and 3). The key model parameters are summarized in Table S2 in the Supporting Information. In Figure 14, the variations in surface composition with solution composition in the absence of CaCl2 and in 0.5 and 1.0 mM CaCl2 are compared. The addition of 0.5 mM CaCl2 does not significantly alter the pattern of adsorption behavior. However, the addition of 1 mM CaCl2 has a more substantial effect. In this case, the surface and Langmuir 2010, 26(13), 10614–10626
solution compositions are much more similar. Included in Figure 14b is also an RST calculation for the variation in the surface composition with solution composition, where the impact of the electrolyte on the LAS cmc has been taken into account. An interaction parameter of -2.0 has been assumed, and it is estimated that the cmc of the LAS (in the presence of 1 mM CaCl2) has been reduced to ∼0.3 mM. A consequence of this is that the measurements are now made well in excess of the mixed cmc over the entire composition range and the calculated curve is now much closer to the measured data. Hence in water and in 0.5 mM CaCl2, there is a more significant departure from ideal mixing at the interface than would be expected from the existing theories of nonideal mixing.5 Marked departures were also observed in the surface adsorption of the dialkyl chain cationic/ nonionic surfactant mixtures of DHDAB/C12E6 (C12E12) in the absence of electrolyte.33 In that case, different behavior was observed in that below a critical composition the surface was totally dominated by DHDAB. This coincided and was correlated with a change in the bulk phase behavior, from micellar to micellar/lamellar coexistence. It was hence assumed that this change in phase behavior resulted in a change in the monomer concentrations and composition that is in equilibrium with the surface. This is not what is observed here because upon comparing the phase behavior in Figure 11 and the surface behavior in Figures 13 and 14 there is no evident or obvious correlation between the two patterns of behavior. The surface behavior in 1 mM CaCl2 is different, and the surface composition is very similar to the solution composition, corresponding to more ideal mixing. Increasing the electrolyte concentration will reduce the ABS cmc and make the cmc values for ABS and C10E3 much closer to each other and the mixed ABS/ C10E3 cmc much lower. This would promote a trend toward more ideal behavior, but significant effects are usually observed only at much higher electrolyte concentrations for monovalent counterions. However, the addition of CaCl2 promotes more planar structures in solution over most of the composition range as the spontaneous curvature of ABS is reduced. This would drive the surface and solution compositions to be more equivalent. In the presence of 0.5 and 1 mM CaCl2 and for solution compositions richer in ABS than 70:30 ABS/C10E3, the surface (33) Tucker, I.; Penfold, J.; Thomas, R. K.; Tildesley, D. J. Langmuir 2009, 25, 3924.
DOI: 10.1021/la100846b
10623
Article
Tucker et al. Table 4. Variation in Surface Microstructure with Solution Composition for 100:0 to 60:40 Composition Solutions and for 0.5 and 1.0 mM CaCl2a composition (mole % ABS/C10E3) 100:0 95:5 90:10 85:15 80:20 70:30 60:40 CaCl2 (mM) 0.0 1 1 0.5 M M 1.0 M M M a M, multilayer; B, bilayer; 1, monolayer.
Figure 15. Neutron reflectivity for 2 mM d-ABS/h-C10E3 in 0.5 mM CaCl2: (red) 60:40, (blue) 80:20, and (green) 95:5. The solid lines are model fits as described in the text.
structure is more complex and is no longer a simple monolayer. This is illustrated in Figure 15 for 2 mM ABS/C10E3 in 0.5 mM CaCl2 at solution compositions of 60:40, 80:20, and 95:5. In 0.5 mM CaCl2 and at a solution composition of 60:40 ABS/ C10E3, the surface adsorbed layer is a monolayer, as describe earlier (Table S2 in Supporting Information), with a thickness of ∼25 A˚. At a solution composition of 80:20, the form of the reflectivity has changed and a pronounced interference fringe is present. This corresponds to a thicker bilayer structure at the surface and is described by three layers (with d, F values of 20 A˚, 4.2 10-6 A˚-2; 17, 1.3 10-6; and 10, 1.9 10-6). At a solution composition of 95:5, the reflectivity is different again and a pronounced Bragg peak is evident in the data at relatively high Q values. This corresponds to multilayer formation at the interface, similar to that observed in ABS/C12E8 mixtures.8 This profile was analyzed using the approach outlined earlier and described by eq 4. The data are consistent with N = 40 (40 bilayers), 2d = 32.5 A˚, and ΔF = 1.5 10-5 A˚-2. This is similar to that reported in previous studies.8 Furthermore, the bilayer thickness here is similar to that reported earlier in the paper from the analysis of the SANS data, where the Nallet analysis gave a bilayer thickness (δ) of 32 A˚. In 1.0 mM CaCl2, the formation of surface multilayer structures occurs over a wider composition range and extends to a solution compositions of 80:20, and this is summarized in Table 4. A broadly similar CaCl2 concentration dependence was observed in the ABS/C12E8 mixture.8 The development of the more complex multilayer surface structures is time-dependent, as previously reported,8 and although the pattern of behavior summarized in Table 4 is reproducible, it is not clear in all cases that the equilibrium structures have been obtained. With this in mind, we have not attempted or presented a more extensive quantitative analysis of the rest of the multilayer data. A more detailed study of the kinetics of multilayer formation will be reported separately in the future.34 General Discussion. In a previous related study,8 we have investigated the surface and solution mixing behavior of ABS with two different nonionic cosurfactants with different but high intrinsic preferred curvature, C12E8 and C12E23. At low surfactant concentrations, ABS forms globular micelles but its packing constraints put it close to the transition to larger structures. (34) Penfold, J.; Thomas, R. K.; Tucker, I.; Petkov, J.; Webster, J. R. P.; Morgan, C. To be submitted for publication, 2010.
10624 DOI: 10.1021/la100846b
1 B M
1 1
1 1 1
Hence at higher surfactant concentrations and in the presence of divalent counterion Ca2þ, planar structures dominate. The addition of Ca2þ was demonstrated to be an important parameter for tuning and adjusting the surface and solution behavior. The motivation for the current study was to contrast that situation with ABS/nonionic surfactant mixtures where the nonionic cosurfactant has a much lower preferred curvature, C10E3, which is consistent with planar structures. Of particular importance also is the role of Ca2þ counterion in tuning that behavior. The incorporation of C10E3 as a cosurfactant with ABS, compared to C12E8 or C12E23, has a pronounced impact upon the associated phase behavior. In the absence of CaCl2, micellar structures still dominate most of the composition range, except for C10E3-rich compositions that are planar, in the form of ulv. The notable feature, in the absence of CaCl2, is the narrow region in composition (between 80 and 90 mol % C10E3) where small monodisperse unilamellar vesicles (i.e., nanovesicles) are formed. Similar nanovesicle structures have been reported in other mixed surfactant systems26 and by us in DHDAB/C12E12/octanol (decanol),24 DDAB/C12E4,25 and DHDAB/C12E12/linalool35 mixtures. From those previous studies,24-26 it was evident that the criteria for the formation of these very small unilamellar vesicles are a relatively high bending energy and a finite spontaneous curvature. It is hence evident that the incorporation of a small amount of ABS into C10E3 indeed satisfies those criteria and provides a direct analogy to the formation of nanovesicles by the DDAB/C12E4 mixture.25 In 0.5 mM CaCl2, the nanovesicle region still exists, but for 1.0 and 2.0 mM CaCl2, the preferred structure becomes increasingly planar in the form of large vesicles, either ulv, blv, or mlv. Hence the increased binding of the Ca2þ ions has decreased the bending energy and/or the spontaneous curvature such that the criteria for nanovesicle formation are no longer sustained. In the predominantly planar (vesicle) region of the ABS/C10E3/ CaCl2 phase behavior, there are some interesting trends associated with the vesicle structure. For the compositions richest in C10E3, the Q-2-dependent scattering from the planar structures has no further structure from the bilamellar/multilamellar nature of the vesicles. This has previously been attributed to the formation of relatively large polydisperse unilamellar vesicles. At intermediate ABS/C10E3 compositions, the Q-2 scattering now has additional structure that is associated with strong correlations between adjacent lamellae in the bilamellar or multilamellar vesicles. Although a detailed quantitative analysis was not feasible because these regions exist mostly as a micelle/vesicle mixed phase, a representative analysis (for 25 mM 80:20 ABS/C10E3/2 mM CaCl2 reported earlier) showed that these are relatively rigid structures. C10E3 alone (as described earlier) demonstrates relatively flexible structures, and hence the addition of a small amount (35) Penfold, J.; Tucker, I.; Green, A.; Grainger, D.; Jones, C.; Ford, G.; Roberts, C.; Hubbard, J.; Petkov, J.; Thomas, R. K.; Grillo, I. Langmuir 2008, 24, 12209.
Langmuir 2010, 26(13), 10614–10626
Tucker et al.
of charge is sufficient to induce some increased rigidity such that interlamellar correlations are evident. This is not entirely unexpected, as in the DHDAB/C12E3 mixture28 the addition of C12E3 was seen to enhance the membrane rigidity because of an increased electrostatic contribution from DHDAB, whereas the addition of C12E6 or C12E12 had the more traditionally expected impact and induces an increase in the membrane flexibility. The other notable feature of the mixed micelle/vesicle behavior is illustrated in Figure 10, which showed the approximate variation in the bilayer spacing with composition in 0.5, 1.0, and 2.0 mM CaCl2. The bilayer spacing increases as the solution becomes richer in C10E3. The dependence of the bilayer spacing on composition is similar in 0.5 and in 1.0 mM CaCl2 but is more pronounced in 2 mM CaCl2. This is initially unexpected because the behavior occurs at constant concentration and is in a region of the phase diagram where the microstructure is predominantly but not exclusively planar. It is observed that as C10E3 is added the membrane becomes more flexible, and this increase in d would then be expected as the membrane becomes stabilized by fluctuations32 and not charge. As the repulsion from the Helfrich fluctuations increases, d will increase. This is analogous to the increased membrane fluidity observed in the SDS/pentanol mixture,32 where the addition of an alkyl alcohol promotes fluidity. Hence C10E3 could be considered to be a membrane lubricant. It also provides an important insight into the transition from relatively rigid blv/mlv structures for ABS-rich compositions to relatively flexible ulv for C10E3-rich mixtures. That is, the d spacing increases up to the point where the vesicle becomes sufficiently flexible that the transition from mlv (blv) to ulv occurs. The strong interaction and complexation of the divalent counterion with the ABS headgroup drives the surface multilayer formation at relatively low surfactant concentrations (∼2 mM) and electrolyte (
