Langmuir 2005, 21, 10021-10027
10021
What Is So Special about Aerosol-OT? Part IV. Phenyl-Tipped Surfactants† Sandrine Nave, Alison Paul, and Julian Eastoe* School of Chemistry, University of Bristol, Bristol BS8 1TS, U.K.
Alan R. Pitt Kodak R&D, Kodak Ltd., Headstone Drive, Harrow, HA1 4TY, U.K.
Richard K. Heenan ISIS-CLRC, Rutherford Appleton Laboratory, Chilton OXON OX11 0QX, U.K. Received March 23, 2005. In Final Form: April 29, 2005 Properties are reported for new phenyl-tipped anionic surfactants, which are aromatic chain relatives of the normal aliphatic aerosol-OT (AOT, sodium bis(2-ethyl-1-hexyl)sulfosuccinate). Variations in chain length and branching with these aromatic surfactants have important effects on aqueous and water-in-oil (w/o) microemulsion phase properties. In dilute aqueous systems, chain structure affects the cmc and surface tension behavior: compared to linear chain analogues, the branched-chain surfactants display lower surface tensions but also reduced packing as measured by molecular area at the cmc acmc. Owing to the phenyl-tipped structure, water-in-oil microemulsions were stabilized with aromatic toluene as an oil but not with aliphatic heptane; the latter is commonly used with normal AOT. Contrast variation small-angle neutron scattering (SANS) was used to characterize the microemulsion aggregates and adsorbed films. These SANS data show that water-in-toluene microemulsions stabilized by aromatic-AOTs contain mildly polydisperse spherical nanodroplets of similar structure to those found in systems containing normal AOT. Molecular areas at the air-water and toluene-water interfaces are found to be of similar magnitude and follow a trend that correlates with variations in surfactant chain structure. The new results with aromatic surfactants build on extensive studies of aliphatic AOT analogues (Nave, S.; Eastoe, J.; Penfold, J. Langmuir 2000, 16, 8733. Nave, S.; Eastoe, J.; Heenan, R. K.; Steytler, D.; Grillo, I. Langmuir 2002, 16, 8741. Nave, S.; Eastoe, J.; Heenan, R. K.; Steytler, D.; Grillo, I. 2002, 18, 1505), suggesting that the versatility of normal AOT originates from an optimized head and chain spacer group rather than from any specific effects of the 2-ethyhexyl chain structure.
Introduction Changes in the molecular structure of surfactants have important consequences for physicochemical solution properties and hence on potential applications. In addition, recent work has demonstrated how certain surfactant mixtures1 and custom-designed amphiphilic polymer additives2 can be used to boost microemulsion phase stability. The twin-tailed anionic surfactant aerosol-OT (AOT, sodium bis(2-ethyl-1-hexyl)sulfosuccinate) is widely used in the formulation of water-in-oil (w-o) microemulsions, which are beginning to find applications in the generation and stabilization of nanoparticles.3 There is a vast literature devoted to studying AOT w/o microemulsions using many different structural and physical techniques and over a wide range of experimental conditions. (Important example papers are given;4 see background references in refs 5-7.) Naturally, the question arises as to why this AOT molecule is so effective in microemulsion applications and whether more efficient surfactants could be designed. To address this, extensive †
Part of the Bob Rowell Festschrift special issue. * Corresponding author. E-mail:
[email protected]. Tel: +117 9289180. Fax: + 117 9250612. (1) Bumajdad, A.; Eastoe, J.; Nave, S.; Steytler, D. C.; Heenan, R. K.; Grillo, I. Langmuir 2003, 19, 2560. (2) Jakobs, B.; Sottmann, T.; Strey, R.; Allgaier, J.; Willner, L.; Richter, D. Langmuir 1999, 15, 6707. (3) Pileni, M. P. J. Phys. Chem. 1993, 97, 1. Pileni, M. P. Langmuir 1997, 13, 3266. Pileni, M. P. Nat. Mater. 2003, 2, 145. Lisiecki, I.; Pileni, M. P.; Langmuir 2003, 19, 9486.
studies have been performed encompassing 12 different straight and branched aliphatic chain AOT analogues.5-7 This family of AOT-related compounds was synthesized and purified, and their properties were characterized at both air-water and aliphatic heptane-water interfaces by surface tensiometry, small-angle neutron scattering (SANS), and neutron reflection (NR). The main conclusions were the following: • For a given surfactant, limiting molecular areas were very similar at both air-water and oil-water interfaces; hydrocarbon structure and branching dictates interfacial packing, which correlates well with an empirical chain branching factor. • The stabilization of ternary w/o microemulsions requires a minimum conformational/structural disorder (4) (a) Cabos, C.; Delord, P. J. Appl. Crystallogr. 1979 12, 502. (b) Kotlarchyk, M.; Chen, S.-H.; Huang, J. S.; Kim, M. W. Phys. Rev. A 1984, 29, 2054. (c) Kotlarchyk, M.; Huang, J. S.; Chen, S.-H. J. Phys. Chem. 1985, 89, 4382. (d) Aveyard, R.; Binks, B. P.; Clark, S.; Mead, J. J. Chem. Soc., Faraday Trans. 1 1986, 82, 125. (e) Hou, M. J.; Kim, M.; Shah, D. O. J. Colloid Interface Sci. 1988, 123, 398. (f) Binks, B. P.; Meunier, J.; Abillon, O.; Langevin, D. Langmuir 1989, 5, 415. (g) Chen, S.-H.; Chang, S.-L.; Strey, R. J. Chem. Phys. 1990, 93, 1907. (h) Farago, B.; Huang, J. S.; Richter, D.; Safran, S. A,; Miller, S. T. Prog. Colloid Polym. Sci. 1990, 81, 60. (i) Eastoe, J.; Robinson, B. H.; Steytler, D. C.; Thorn-Leeson, D. Adv. Colloid Interface Sci. 1991, 36, 1. (5) Nave, S.; Eastoe, J.; Penfold, J. Langmuir 2000, 16, 8733. (6) Nave, S.; Eastoe, J.; Heenan, R. K.; Steytler, D.; Grillo, I. Langmuir 2000, 16, 8741. (7) Eastoe, J.; Nave, S.; Steytler, D. C.; Heenan, R. K.; Grillo, I. Langmuir 2002, 18, 1505.
10.1021/la050767a CCC: $30.25 © 2005 American Chemical Society Published on Web 09/03/2005
10022
Langmuir, Vol. 21, No. 22, 2005
Figure 1. Surfactants used in this study.
in the surfactant hydrophobic chains; this is demonstrated by the fact that unbranched linear chain surfactants (diCnSS) do not stabilize microemulsions alone, whereas the branched chain analogues do. • The magnitude of surfactant film curvature in the w/o microemulsions scales with surfactant hydrophobic size so that water solubilization capacity correlates with the degree of chain branching. • There is nothing special about the molecule AOT, which fits into a general pattern of behavior of other branched chain relatives; all result in high solubilization capacity for water in w/o systems. The only distinguishing feature of AOT is high room-temperature solubility in aqueous phases, and this in turn influences the location of w-o phase boundaries to make them conveniently centered on room temperature. • Normalizing for this latter effect shows that AOT and analogues display very similar water solubilization efficiencies in w/o systems. Because normal AOT behaves in a fashion very similar to that of other branched-chain analogues, the presence of a sulfosuccinate group supporting two branched hydrocarbon chains appears to be the key structural motif, rather then the 2-ethyhexyl chains themselves. The purpose of this article is to make a stringent test of this conclusion by replacing the terminal CH3- groups with an aromatic C6H5- group. The presence of delocalized electrons in chain-tip phenyl groups is known to radically affect the properties of aqueous systems compared to those of permethylated surfactants.8 Rosen8b has conducted an extensive review of how structural factors affect interfacial and aggregation properties in aqueous systems. The focus of this article is to compare the behavior of Ph- and Meterminated chain surfactants at oil-water interfaces and in w/o microemulsions. The compounds studied here are identified in Figure 1: normal AOT, sodium bis(4-phenyl-butyl)sulfosuccinate (di-PhC4SS), sodium bis(5-phenyl-pentyl)sulfosuccinate (di-PhC5SS), sodium bis(2-phenyl-propyl)sulfosuccinate (br-di-PhC3SS), and sodium bis(3-phenyl-2,2-dimeythylpropyl)sulfosuccinate (br-di-PhC5SS). With aqueous solutions, surface properties for anionic dichains and trichains (sodium sulfosuccinates and sulfotricarballylates) have (8) (a) Pitt, A. R.; Morley, S. D.; Burbidge, N. J.; Quickenden, E. L. Colloids Surf., A 1996, 114, 321-335. (b) Rosen, M. Surfactants and Interfacial Phenomena, 2nd Ed.; Wiley: New York, 1989.
Nave et al.
been reported before8 in a broad-based study of structureperformance relationships. It was found that the chemical nature of the final groups at the hydrophobic chain tips has a profound influence on the magnitude of the limiting post-cmc surface tension γcmc at the air-water (a-w) interface.8 Varying from terminal fluoroalkyl (CF3[CF2]nand H[CF2CF2]n-) through alkyl (with methyl branches) to aryl groups gave a progressive decrease in the effectiveness of the surfactant (i.e., an increase in γcmc values). Fluorinated chains and highly methylated tertbutyl hydrocarbon tips promote the lowest γcmc between 18 and 30 mN m-1; typically, linear hydrocarbon chains (only one terminal CH3- group) give rise to 35-40 mN m-1; at the upper end phenyl-tipped surfactant analogues generate γcmc values of up to 55 mN m-1, depending on the exact structure. Also reported were properties of nonionic sugar-based surfactants containing aryl-ended tails and n-alkyl tails (classified as two-tail bisgluconamides) of chemical structure [CH2NHCO(CHOH)4CH2OH]2-C-R2 with R ) phenyl, (CH2)3, and n-C6H13. Therefore, welldocumented correlations between the identity of hydrophobic groups and the ability to reduce aqueous surface tension have been established.8 Here, these effects are examined in more detail through an evaluation of molecular areas at both a-w and w-o microemulsion interfaces for the compounds shown in Figure 1. At the air-water surface, molecular areas acmc have been determined from pre-cmc surface tension data as detailed elsewhere;5 packing area at w-toluene interfaces ah has been assessed through a detailed analysis of small-angle neutron scattering (SANS) data, as previously reported.6,7 The results contribute a better understanding of the links between surfactant chemical structure and surface and interfacial properties, which has implications for the rational design of surfactants for speciality applications, such as non-hydrocarbon emulsions9 and dispersions in green solvents such as supercritical carbon dioxide.10 Experimental Section a. Materials and Equipment. Surfactants were synthesized and purified as outlined in refs 5-7 using 4-phenyl-butan-1-ol, 5-phenyl-pentan-1-ol, 2-phenyl-propan-1-ol, or 3-phenyl-2,2diemythylpropan-1-ol (all from Aldrich) to generate di-PhC4SS, di-PhC5SS, br-di-PhC3SS, and br-di-PhC5SS, respectively. The surfactants were characterized by 1H and 13C NMR (JEOL Lambda 300) and elemental analyses that confirmed the expected surfactant chain structures and compositions.5-7,11 All surfactant materials were stored over refreshed P2O5 in a desiccator cabinet until used. Deuterated solvents toluene-d8 (Aldrich, >99% D atom) and D2O (Fluorochem, 99.9% D atom) were used as received. Normal toluene-h8 was Sigma HPLC grade, and H2O was of ultrahigh purity (RO100HP Purite water system or Millipore Milli-Q Plus system). Microemulsion phase equilibria were determined from a visual inspection of samples made up in clean, stoppered 5 mL volumetric flasks that were thermostated by a heater-cooler-circulator water bath accurate to (0.1 °C. The composition parameter, w, is given by [water]/[surfactant], and for the phase diagrams, the surfactant concentration was kept constant at 0.10 mol dm-3. For tensiometry, glassware was cleaned in 40% nitric acid solution, rinsed well, soaked in Micro critical cleaning solution, rinsed repeatedly with water, and finally checked for cleanliness over a steam bath. Samples were made up by mass, and concentrations were corrected for density (Paar digital density meter DMA 35). Surface tensions were measured using a drop volume instrument (Lauda TVT1) operating in dynamic mode. (9) Patel, N.; Marlow, M.; Lawrence, M. J. J. Colloid Interface Sci. 2003, 258, 345. (10) Gold, S.; Eastoe, J. Phys. Chem. Chem. Phys. 2005, 7, 1352. (11) Nave, S. Ph.D. Thesis, University of Bristol, Bristol, U.K., 2002.
What Is So Special about Aerosol-OT?
Langmuir, Vol. 21, No. 22, 2005 10023
Prior to determining the adsorption isotherm, measurements were carried out at a constant surfactant concentration using EDTA (Sigma, 99.5% ethylenediaminetetraacetic acid tetrasodium salt hydrate) to sequester any polyvalent cations. The appropriate molar ratio of surfactant/EDTA was determined to optimize the removal of contaminant cations but not exert any electrolyte effect. This procedure is described elsewhere.5 Surface tension measurements were then carried out with the appropriate molar ratio of surfactant/EDTA at 25.0 ( 0.1 °C. (Surfactant/ EDTA ratios were 120:1, 300:1, 1800:1, 95:1, and 160:1 for AOT, di-PhC4SS, di-PhC5SS, br-di-PhC3SS, and br-di-PhC5SS, respectively.) Foam fractionation was not carried out to purify the surfactants because it has been shown that this treatment has only a minor effect on the adsorption parameters for similar sulfosuccinates.5 Mean ionic activities were estimated using the Debye-Hu¨ckel approximation. Polarizing light microscopy studies were performed using a Nikon Optiphot-2 microscope equipped with polarizing filters and a Nikon Optizoom ×0.8-×0.2 lens. The samples were held between a cover slip and a microscope slide and thermostated to 25 °C on a Linkam hot-cold stage. Optical textures were captured on a PC. A small amount of surfactant was placed on a microscope slide, under a cover slip. The slide was mounted on the heating stage, and the temperature was raised until the sample became fluid and entirely isotropic. After slowly cooling the sample (1.0 °C min-1) to 25 °C, a drop of water was added at the edge of the cover slip. As the water penetrated the surfactant, a concentration gradient was produced from water on one side to pure surfactant on the other, enabling the whole range of mesophases to be observed. The SANS experiments were carried out on the time-of-flight LOQ instrument at ISIS, U.K., where incident wavelengths are 2.2 e λ e 10 Å.12 The momentum transfer Q is (4π/λ)sin(θ/2) with the scattering angle θ yielding 0.009-0.249 Å-1. Absolute intensities for I(Q) (cm-1) were determined to better than (8% using a partially deuterated polymer standard.13 Accepted procedures for data treatment were employed.12 Samples were preequilibrated at the appropriate temperature for about 5 h prior to the SANS measurements, which were made at 25 °C. b. SANS Data Analysis. The microemulsion droplets were treated as spherical core-shell particles with a Schultz distribution in the core radius (Supporting Information). Full accounts of the scattering laws are given elsewhere,1,4b,c,6,7,14,15 and only a summary is necessary here. For polydisperse spherical droplets at volume fraction φ, radius Ri, volume Vi, and coherent scattering length density Fp dispersed in a medium of Fm, the normalized SANS intensity I(Q) (cm-1) may be written as
∑V P(Q, R ) X(R )]S(Q, R
I(Q) ) φ(Fp - Fm)2[
i
i
i
hs,
φhs) (1)
i
where P(Q, Ri) is the single-particle form factor. The Schultz distribution X(Ri) defines the effective polydispersity using an average radius Rav and a root-mean-squared deviation σ, with z being a width parameter. This effective polydispersity function takes into account what are believed to be two dominant contributions to the observed distribution (e.g., ref 4b, f, and h): a natural-sized polydispersity and thermally excited shape fluctuations. S(Q, Rhs, φ hs) is the structure factor, and a hardsphere model modified for polydispersity was used:6,12 the constraints were φhs ) φd and Rhs ) Rav drop together with the known F values for solvents. (Subscripts drop and hs denote the droplet and hard-sphere volume fraction, respectively.) For any given set of parameters, the core volume fraction φc defines the absolute scattering intensity, and (10% of the known value was allowed in the modeling. Using the approach of Ottewill et al., eq 1 can be modified to allow for sharp-step shells built onto a spherical core.15 (12) Information about SANS and data processing can be found at http://isise.rl.ac.uk/LargeScale/LOQ/loq.htm, http://www.ill.fr and http:// www.ncnr.nist.gov/resources/n-lengths/. (13) Wignall, G. D.; Bates, F. S. J. Appl. Crystallogr. 1987, 20, 28. (14) Eastoe, J.; Hetherington, K. J.; Sharpe, D.; Dong, J.; Heenan, R. K.; Steytler, D. C. Langmuir 1996, 12, 3876. (15) Markovic, I.; Ottewill, R. H.; Cebula, D. J.; Field, I.; Marsh, J. Colloid Polym. Sci. 1984, 262, 648.
Figure 2. Surface tension isotherms for di-PhC4SS (4), diPhC5SS (]), br-di-PhC3SS (O), and br-di-PhC5SS (2) at 25 °C. Quadratic lines fitted to the pre-cmc data were used to generate molecular areas shown in Figure 3. The least-squares FISH program was used to analyze the SANS data.16 For each surfactant, at a common w value two contrasts were fitted simultaneously: core contrast D2O/H-surf/H-alkane (D/H/H) and shell contrast D2O/H-surf/D-alkane. Because the scattering-length densities F and concentrations of the components are all known, three adjustable parameters were required in the analysis: the most probable core radius and polydispersity av width, Rav c and σ/Rc , respectively, as well as the apparent film thickness, ts. The water concentration defines φc, the core volume fraction; adding in the surfactant gives an overall value φd for the droplets. For each component, the scattering-length density F was calculated from eq 2
F)
bi
∑V i
(2)
m
where bi are the nuclear scattering lengths as given in standard tables (e.g., ref 12) and Vm is the molecular volume, which can be obtained from mass density (for surfactants 1 g cm-3 was assumed). In addition, the individual core SANS data sets for a variation of w values were analyzed on their own using this model.
Results and Discussion a. Aqueous Systems: Surface Tensions, Surfactant Packing, and Lyotropic Mesophases. Figure 2 shows γ-ln activity plots for the aromatic compounds shown in Figure 1, respectively; breaks at the cmc were clearly defined. Adsorption isotherms were derived from the precmc data using the Gibbs equation with a prefactor of 2, as detailed elsewhere,17 and the surface excesses were converted to molecular areas acmc at the respective cmc’s. Parameters derived from these tension data, cmc’s, limiting surface tensions γcmc and effective headgroup areas acmc, are given in Table 1. Figure 3 shows the variation acmc as a function of the effective chain length n-C (each phenyl group was accounted for by 3.5 × CH2groups). Also shown in Figure 3 are results for normal AOT and the series of linear chain di-CnSS sulfosuccinates taken form ref 5 so that 10 different but structurally related compounds are compared. Clearly, replacement of the terminal methyl end group with a phenyl changes these surfactant properties significantly. From a more general viewpoint, acmc values correlate with molecular geometric features (i.e., the presence of the phenyl group (16) Heenan, R. K. The “FISH” Data Fitting Program Manual; Rutherford Appleton Laboratory Report RAL-89-129, 1989, (17) Eastoe, J.; Nave, S.; Rankin, A.; Paul, A.; Downer, A.; Tribe, K.; Penfold, J. Langmuir 2000, 16, 4511.
10024
Langmuir, Vol. 21, No. 22, 2005
Nave et al.
Table 1. Parameters Derived from Surface Tension Measurements of the Linear and Branched Phenyl-Tipped Sodium Sulfosuccinatesa surfactant AOTa di-PhC4SS di-PhC5SS br-di-PhC3SS br-di-PhC5SS
cmc/(mmol dm-3) γcmc ((1/mN m-1) acmc ((2/Å2) 2.56 ( 0.3 4.21 ( 0.4 0.778 ( 0.06 13.8 ( 0.4 1.72 ( 0.15
30.8 44.1 37.0 38.4 34.3
75 69 71 100 98
a Data taken from ref 5. The uncertainties in the cmc have been estimated from repeat measurements.
Figure 4. Polarizing light microscopy textures showing the progression of lyotropic phases for di-PhC3SS at 25 °C. Dominant phases are identified as H2 reversed hexagonal, V2 bicontinuous cubic, and LR lamellar.
Figure 3. Limiting molecular areas at the respective cmc for various sulfosuccinate surfactants. Values for AOT and linear chain di-CnSS analogues were taken from ref 5.
increases the molecular volume of the tail in comparison to that of a simple linear alkyl chain, thus giving rise to a less dense adsorbed monolayer (increased acmc)). The largest effects are seen with the branched Ph-tipped compounds, for which molecular area increases on the order of 50-60% are noted as compared with those for analogous di-CnSS compounds. A striking difference observed upon replacement of the methyl end group with a phenyl group is the change in limiting surface tension, γcmc. The introduction of the relatively polarizable phenyl group that contains a large area of delocalized electrons increases γcmc to above 44 mN m-1 (Table 1). Phenyl end groups thus decrease the effectiveness of both the linear and branched surfactants. In all cases, the introduction of chain branching reduces γcmc but increases acmc, highlighting the fact that surface packing alone is not the only factor controlling limiting surface tension and that chain chemistry has an important role to play. An overall comparison of all of the doubletailed sulfosuccinates studied in this work therefore gives the following trend in terms of the effectiveness and chemistry of the terminal group: CH3CH2 f -CH2-CH2 f phenyl-. This observation agrees with previously reported results.8 Phase penetration polarizing light microscopy (PLM) studies revealed the mesophase progressions as the pure aromatic surfactants were diluted with water. Figure 4 shows a phase-cut optical texture for di-PhC3SS-water mesophases. The same phase sequence is found for normal AOT and was seen with all of these Ph-tipped compounds: at the higher surfactant concentrations, reverse hexagonal (H2) displaying fanlike optical textures; bicon-
Figure 5. Water-in-toluene microemulsion phase stability diagrams with phenyl-tipped sulfosuccinate surfactants. The phase boundary for AOT (- - -) in toluene is given for comparison purposes. In all cases, lines represent a Winsor II (solubilization) boundary. [surf] ) 0.10 mol dm-3.
tinuous cubic (V2), which is nonbirefringent; a lamellar (LR) parallel streak pattern, sometimes with Maltese-crossshaped spherulites; and a region of myelins where the lamellar phase gives way to an L1 optically isotropic micellar solution. b. Stability of Water-in-Toluene Microemulsions. Water-in-oil composition-temperature phase stability diagrams of diphenylsodium sulfosuccinates were studied using n-heptane and toluene. Surprisingly, none of the aromatic compounds formed a single microemulsion phase with heptane, although this is a very good solvent for normal AOT. However, water-in-toluene microemulsions were stabilized both with the Ph-tipped compounds and also with normal AOT. These observations highlight the importance of favorable tail-solvent interactions for promoting low interfacial tensions necessary for microemulsion formation. Figure 5 shows that the linear and branched phenyl sulfosuccinates exhibit limited water-in-toluene phase stability in the temperature window studied (5-50 °C). In toluene, normal AOT itself shows poor microemulsifying efficiency (Figure 5) as compared with that of the standard n-heptane system.6,7 This is likely due to less-favorable oil/chain interactions with toluene. With the aromatic
What Is So Special about Aerosol-OT?
sulfosuccinates over the range 5-50 °C, only a solubilization (Winsor II) boundary was observed (i.e., separating a single L2-microemulsion phase from a two-phase L2microemulsion + excess water phase as temperature was increased). Hence, below the phase boundary lines in Figure 5 transparent, homogeneous single phases are present, whereas biphasic Winsor II systems are located above the boundaries. This contrasts with the wide range of branched aliphatic AOT analogues, which in water/ heptane microemulsions display distinctly different lowtemperature (Winsor II) and high-temperature (cloudpoint) phase boundaries.6,7 Upper phase boundaries could be observed for the branched compounds only at very high temperatures (approximately >70 °C using a hot-air gun), indicating possible “AOT-like” phase behavior. However, the evaporation of toluene limited the accuracy and reproducibility of these high-temperature experiments. c. Structural Studies by Small-Angle Neutron Scattering. In view of the more limited water-in-toluene phase stability for br-di-PhC3SS (Figure 5), only diPhC4SS, di-PhC5SS, and br-di-PhC5SS were selected for detailed SANS studies. Measurements were carried out at 25 °C at various w values within the single microemulsion phase (shown in Figure 5). i. Droplet and Interfacial Structure. For normal AOT, the water droplets are well described as core-shell Schultz polydisperse spheres with a hard-sphere structure factor to account for spatial correlations:4b,6,7 for all of the Ph surfactants, SANS data could be well fitted with this common model. To reveal detailed structural information, separate but complementary core (water-surfactanttoluene D/H/H) and shell (D/H/D) contrast experiments were performed at a common w value (w ) [water]/[surf] ) 16) and the data sets were fitted simultaneously. Including the shell contrast in the analysis allows an accurate determination of the polydispersity through the depth and breadth of interference fringes at high Q. Example scattering curves for the core and shell contrasts are shown in Figure 6, along with the lines from the simultaneous analysis. The fitted parameters for the water core radius, Rc, polydispersity, σ/Rav c , and film thickness, ts, at w ) 16 are given in Table 2. As observed with related sulfosuccinate surfactant series6,7 at a constant w value, the droplet size was found to decrease slightly with increasing surfactant chain length n, whereas the polydispersity remained approximately unchanged. The polydispersities are very similar, as would be expected on the basis of the film rigidity (bending energy) model,18 which should (approximately) scale inpredicts that σ/Rav c versely with film thickness ts. Because the effective chain length spans only a limited range, from C6 (AOT) to C8.5 (br-di-PhC5SS), only minimal changes in polydispersity are to be expected. In any case, the order of σ/Rav c essentially follows the changes in chain length. It is interesting to compare isomers di-PhC5SS and brdi-PhC5SS: the branched compound solubilizes slightly less water although its effective tail is shorter. This indicates that a certain length is required to give a wellbalanced molecule, with respect to the hydrophobic and hydrophilic moieties. Adding the methyl branch increases the efficiency of the surfactant in solubilizing water and considerably reduces the Krafft temperature. One limitation associated with these systems is the small w range available (see Figure 5, generally wmax < (18) Szleifer, I.; Kramer, D.; Ben-Shaul, A.; Gelbart, W. M.; Safran, S. A. J. Chem. Phys. 1990, 92, 6800. Gradzielski, M. Langmuir 1998, 14, 6037. Gradzielski, M.; Langevin, D.; Sottmann, T.; Strey, R. J. Chem. Phys. 1997, 106, 8232.
Langmuir, Vol. 21, No. 22, 2005 10025
Figure 6. (Upper) D2O core contrast data and fits (-) to a Schultz polydisperse sphere model for Winsor II microemulsions of di-PhC4SS in toluene-h8. [surf] ) 0.10 mol dm-3, and T ) 25 °C. Data and fits have been multiplied for clarity of presentation: w ) 16 (O), 18 × 2 (×), 22 × 4 (4), and 26 × 8 (*). (Lower) Surfactant shell contrast data and fits (-) to the Schultz coreshell model for di-PhCnSS microemulsions in toluene. w ) 16, [surf] ) 0.10 mol dm-3, and T ) 25 °C. Data and fits have been multiplied as follows: di-PhC4SS (O), di-PhC5SS × 3.5 (×), and br-di-PhC5SS × 20 (4). Table 2. Parameters Derived from Simultaneous Analyses of Core-Shell Contrasts at w ) 16' surfactant
σ/Rav c ( 0.01
Rc ( 1/Å
ts ( 1/Å
AOT di-PhC4SS di-PhC5SS br-di-PhC5SS
0.22 0.20 0.18 0.19
28.5 29.5 28.8 24.7
9.1 8.2 9.0 6.0
25) so that scattering intensities are relatively weak and the fitting is less accurate than for normal systems such as water-AOT-n-heptane. However, as shown by the values in Table 2, film thicknesses change in line with the expected effective chain lengths. This shows that such subtle structural changes can be identified by SANS, and even for systems of this kind, true molecular resolution can be claimed with carefully designed contrast variation experiments. ii. w Variation and Molecular Areas. To evaluate molecular areas at the water-toluene interface, a series of w-variation experiments were performed in core contrast (example data for di-PhC4SS shown in Figure 6, upper panel). The fitted mean radii as a function of water solubilization are shown in Figure 7, demonstrating a linear Rc versus w swelling law. As shown in Supporting Information, the headgroup area ah can be estimated from this plot using
R(p)Rav c )
3υh 3υw w+ ah ah
(3)
10026
Langmuir, Vol. 21, No. 22, 2005
Nave et al. Table 3. Headgroup Areas of Sulfosuccinate Surfactantsa headgroup area/Å2 surfactant
ah(3υww/Rc)
ah (Porod)
acmc
AOT di-PhC4SS di-PhC5SS br-di-PhC5SS
74 63 66 85
69 59 58 72
75 69 71 98
a Headgroup areas at the air-water interface a cmc are from Table 1. Uncertainties: ah(3υww/Rc), (3 Å2; ah(Porod), (15 Å2; acmc, (2 2 Å.
Figure 7. Microemulsion droplet swelling of average radius Rc as a function of water content w. Linear fits used to generate molecular areas given in Table 3 using eq 3.
where υw is the volume of a water molecule, p is the polydispersity index (σ/Rav c ), and for a Schultz distribution it has been verified that R(p) ) 1 + 2p2 (see Supporting Information and ref 4b). Therefore, assuming that the polydispersity is independent of w the lines in Figure 6 will have a slope that depends on the interfacial molecular area ah and an intercept that depends on the average headgroup volume υh via the surfactant headgroup radius rh.
rh )
( ) 3υh 4π
1/3
(4)
For all surfactants, similar intercepts were found (Ro ) 7.0 Å for AOT, 7.5 Å for br-di-PhC5SS, and 7.7 Å for diPhC4SS and di-PhC5SS) resulting in υh and rh values spanning 162-210 Å3 and 3.4-3.7 Å, respectively. These dimensions are consistent with previous studies on related compounds6,7 and suggest that the hydrocarbon chain has little effect on the internal structure of the dry reversed micelles, which is apparently dominated by the sodium sulfosuccinate group itself. An alternative method used to estimate ah is to evaluate high Q core contrast data with the Porod equation
{I(Q)Q4} ) 2π∆F2Σ
(5)
where ∆F (FD2O - Foil) is the contrast step and Σ is the total area per unit volume.19 Assuming that all N surfactant molecules per unit volume are adsorbed at the oil-water interface (i.e., monomer solubility in water and toluene is negligible compared to the total concentration), then ah(Porod) ≈ Σ/N. (Limitations of this approach have been documented elsewhere,6 and Porod plots are shown in Supporting Information.) As given in Table 3, values of ah calculated using eqs 3 and 4 agree reasonably well given the assumptions involvedspolydisperse spheres for eq 3 and sharp-step interfaces for eq 4. As for the series of AOT analogues studied elsewhere,6,7 the branching effect is clearly seen with an increase in ah of about 25-27 Å2 (see di-PhC5SS vs br-di-PhC5SS). The variation in headgroup areas (19) Porod, G. Kolloid-Z. 1951, 124, 82.
obtained for the phenyl systems seems reasonable and corresponds well to changes in surfactant molecular structure. d. Comparison of Films at Air-Water and Toluene-Water Interfaces. Also given in Table 3 are values for acmc, the molecular area demand in air-water films at the aqueous cmc (from Table 1), hence a clear comparison with the water-toluene interface can be made. Keeping in mind the different chain structures involved (Figure 1), insight into the film packing requirements can be gained. First, headgroup areas (Table 3) correlate well with the extent of branching in the chain: the two branched surfactants AOT and br-di-PhC5SS factors pack less efficiently at both interfaces as compared to the “linear” di-PhC4 and di-PhC5-SS. This reduced packing efficiency can explain the lower water uptake into the water-in-toluene microemulsions for the branched compounds as compared to that for the linear di-PhC4 and di-PhC5-SS surfactants (see phase behavior in Figure 5). Second, packing in the curved water-toluene film is very close to that found at the planar air-water surface (see values for acmc), indicating that the microemulsion interface can be described as a liquidlike condensed surfactant film. This has been seen previously with numerous AOT analogues, suggesting that in terms of surfactant the nature of the two interfaces is essentially the same; there is nothing special about the oil-water interface as compared to the reference air-water surface, apart from the quite weak penetration of oil molecules into the former.14,20 Summary Taken together with previous studies,6,7 extremes of surfactant chain structure have been explored with 16 different surfactants related to aerosol-OT. This work reveals the remarkable versatility of sulfosuccinate surfactants. So as to make robust comparisons between interfacial properties at air-water and oil-water interfaces, it has been important to work with surface chemically pure surfactants and employ surface-sensitive techniques (tensiometry and contrast variation SANS). Then it has been possible to make general observations about surfactant structure-performance relationships for this important class of amphiphiles. (1) In dilute aqueous systems, surface tensions and surface excess respond not only to dramatic changes in the chemical structure of the hydrophobic groups (CH3vs C6H5-tipped compounds) but also to subtle variations in chain group branching for aliphatic compounds (Figures 2 and 3). (2) For concentrated aqueous systems, the switch from aliphatic to aromatic chains has no notable effect on the lyotropic mesophase progression (Figure 4). (20) Eastoe, J.; Dong, J.; Hetherington, K. J.; Steytler, D. C.; Heenan, R., K. J. Chem. Soc., Faraday Trans. 1996, 92, 65.
What Is So Special about Aerosol-OT?
(3) Now, in w/o microemulsions a strong effect of chaintip structure in terms of chemical compatibility with the oil phase is observed (Figure 5). (4) For toluene-continuous w/o systems, the general microemulsion structure is essentially composed of polydisperse spherical nanodroplets, irrespective of the nature of the surfactant chain termini (Figure 6). The phases obey regular swelling laws (Figure 7), and surfactant packing at the water-toluene interface responds to the molecular structure of the chain in the same order as found at the air-water interface. There is a clear correlation between the surfactant packing in the film and the phase stability as measured by the extent of water uptake, hence with these surfactants, system properties are strongly influenced by molecular structure. In summary, the chemical nature of surfactant chain tips has a profound effect on performance at air-water and oil-water interfaces. Important tip-solvent interactions operating in microemulsions (and presumably emulsions) dominate the ability to promote interfacial tensions low enough for stabilizing microemulsions. Surfactants based on the sulfosuccinate motif are very versatile; they can be custom designed for a broad range
Langmuir, Vol. 21, No. 22, 2005 10027
of applications such as the efficient microemulsification of aromatic oils shown here, reaction templates for inorganic nanoparticles,3 and the formation of reversed micelles and microemulsions of water-in-supercritical fluid, as demonstrated elsewhere.10 Therefore, sulfosuccinate amphiphiles represent a central paradigm in colloid and interface science that can be employed in future advances with new and emerging chemical technologies requiring the self-assembly and modification of fluid interfaces. Acknowledgment. S.N. thanks the University of Bristol for a Ph.D. scholarship. We also acknowledge CLRC for the allocation of beam time at ISIS and grants toward consumables and travel. We thank Kodak for initial supplies of surfactants for evaluation. Supporting Information Available: Explanation of the origin of eq 3 and an example of Porod scattering plots. This material is available free of charge via the Internet at http://pubs.acs.org. LA050767A
