Liquid Interface Functionalized by an Ion Extractant: The Case of

Dec 29, 2010 - Phone: +334 6633 9288. Fax: +334 6679 7611. E-mail: [email protected]. Cite this:Langmuir 2011, 27, 5, 1653-1661 ...
0 downloads 0 Views 929KB Size
Download PDF
pubs.acs.org/Langmuir © 2010 American Chemical Society

Liquid Interface Functionalized by an Ion Extractant: The Case of Winsor III Microemulsions Caroline Bauer, Pierre Bauduin,* Olivier Diat, and Thomas Zemb Institut de Chimie S eparative de Marcoule, UMR 5257 (CEA/CNRS/UM2/ENSCM), BP 17171, F-30206 Bagnols sur C eze, France Received October 5, 2010. Revised Manuscript Received December 6, 2010 The present work shows for the first time that tributylphosphate (TBP), the major ion extractant used in the reprocessing of spent nuclear fuel, acts efficiently as a cosurfactant in the formation of three-phase microemulsions. The system is composed of water, dodecane, TBP, and an extremely hydrophilic sugar surfactant, n-octyl-β-glucoside. The investigation of the three-phase region (Winsor III), the so-called “fish-cut” diagrams, revealed that TBP exhibits cosurfactant behavior comparable to that of classical cosurfactants n-pentanol and n-hexanol. Upon increasing the cosurfactant/surfactant molar ratio, TBP appears to be more efficient than single-chain alcohols in raising the spontaneous curvature of the adsorbed surfactant film toward oil. This is a direct consequence of the different lateral packing of TBP and n-pentanol or n-hexanol in the mixed surfactant film, with TBP having three alkyl chains and so a higher hydrophobic volume than those n-alcohols. This property is underlined by the interfacial film composition, which is determined by the chemical analysis of the excess phases. It gives a surfactant to cosurfactant molar ratio of 1:1 for TBP and 1:3 for n-hexanol. Moreover, the local microstructure of the microemulsion becomes dependent on the addition of salt when n-alcohol is replaced by TBP. A specific salt effect is also observed and rationalized in terms of the complexing property of TBP and Hofmeister’s effects. Treatment of the small-angle neutron scattering (SANS) data gives access to (i) the length scales characterizing the microemulsions (i.e., the persistence length, ξ, and aqueous or organic domain sizes, D*) and (ii) the specific surface, Σ. It results that a subtle change is highlighted in the TBP microemulsion structure, in terms of connectivity, according to the type of salt added.

Introduction An effective strategy for cosolubilizing water and oil is the addition of amphiphilic molecules. The use of a hydrophilic surfactant in combination with a hydrophobic cosurfactant, typically n-alcohols with 3 < n < 8, permits to create microemulsions (μE’s) that are clear, isotropic, and thermodynamically stable dispersions of water and oil separated by a mixed surfactant/cosurfactant film.1,2 Therefore, this makes them interesting systems for longlife solubilization, and their particular small structural size provides a remarkably high specific surface. In recent years, microemulsions were designed to be more complex in terms of their composition (e.g., carrying polymers,3 biomolecules,4 complexing surfactant,5 nanoparticles,6 and more recently ionic liquids7). Small-angle scattering techniques remain the most suitable technique for probing the microstructure of microemulsions over a large scale from the interface up to the size of the water and oil domains.8 Under certain conditions, bicontinuous μE’s are formed in equilibrium with excess water and oil phases, so-called Winsor III systems. These *Corresponding author. Phone: þ334 6633 9288. Fax: þ334 6679 7611. E-mail: [email protected]. (1) K. Shinoda, K.; Kunieda, H. J. Colloid Interface Sci. 1973, 42, 381–387. (2) Bourrel, M.; Schechter, R. S. Microemulsions and Related System: Formulation, Solvency and Physical Properties; Surfactant Science Series 30; Marcel Dekker: New York, 1988. (3) Marchal, F.; Guenoun, P.; Daillant, J; Holley, D. W.; Mays, J. M. Soft Matter 2009, 5, 4006–4014. (4) Bauduin, P.; Touraud, D.; Kunz, W.; Savelli, M. P.; Pulvin, S.; Ninham, B. W. J. Colloid Interface Sci. 2005, 292, 244–254. (5) Nave, S.; Testard, F.; Coulombeau, H.; Baczko, K.; Larpent, C.; Zemb, T. Phys. Chem. Chem. Phys. 2009, 11, 2700–2707. (6) Tabor, R. F.; Eastoe, J.; Dowding, P. J.; Grillo, I.; Rogers, S. E. J. Colloid Interface Sci. 2010, 344, 447–450. (7) Liu, L.; Bauduin, P.; Zemb, T.; Eastoe, J.; Hao, J. Langmuir 2009, 25, 2055–2059. (8) Chevalier, Y.; Zemb, T. Rep. Prog. Phys. 1990, 53, 279–371.

Langmuir 2011, 27(5), 1653–1661

can be obtained around the phase inversion when the system changes from an oil-in-water μE with an excess oil phase (Winsor I) to a water-in-oil μE with an excess aqueous phase (Winsor II). Depending on the type of surfactant, these transitions are driven either by increasing temperature9,10 or by adding a hydrophobic cosurfactant that is incorporated into the interfacial monolayer.11,12 Both procedures increase the spontaneous curvature of the interface toward the oil phase. In ion-separation technology, the main process used is based on liquid-liquid extraction where ions are selectively transferred from an aqueous to an organic phase containing a solvent and an extractant molecule. In the nuclear industry, for instance, tributyl phosphate (TBP) is the benchmark extractant used in the socalled PUREX process (plutonium and uranium refining by extraction) where uranium and plutonium are selectively extracted from an acidic aqueous solution.13 TBP and other extractants are hydrophobic, nearly insoluble in water, and have an amphiphilic structure composed of a complexing polar part and one or more alkyl chains as the apolar part. Hence, extractants have all the properties required to play the role of a cosurfactant.12 Incorporating extractants into μE’s might permit us to functionalize the interfacial film artificially by a complexing group. Ultimately, such μE’s can be considered to be model systems for indirectly studying ion adsorption at the water/oil interface. In a similar approach, we recently proposed the use of lamellar phases as model systems for studying the repartitioning of extractants between the surface and the bulk of the lamellar structure.14 (9) Kahlweit, M.; Strey, R.; Firman, P. J. Phys. Chem. 1986, 90, 671–677. (10) Bauduin, P.; Touraud, D.; Kunz, W. Langmuir 2005, 21, 8138–8145. (11) Stilbs, P.; Rapacki, K.; Lindman, B. J. Colloid Interface Sci. 1983, 95, 583–585. (12) Zana, R. Adv. Colloid Interface Sci. 1995, 57, 1–64. (13) Rao, P. R. V.; Kolarik, Z. Solvent Extr. Ion Exch. 1996, 14, 955–993.

Published on Web 12/29/2010

DOI: 10.1021/la104005x

1653

Article

Bauer et al. Table 1. Composition of the Winsor III Systems Prepared for Neutron-Scattering Experiments and for Chemical Analysisa system

sample name

A1 C8G1/hexanol A2 C8G1/hexanol/LiCl A3 C8G1/hexanol/LiNO3 A4 C8G1/hexanol/Nd(NO3)3 B1 C8G1/hexanol B2 C8G1/hexanol/LiCl B3 C8G1/hexanol/LiNO3 B4 C8G1/hexanol/Nd(NO3)3 C1 C8G1/hexanol C2 C8G1/hexanol/LiCl C3 C8G1/hexanol/LiNO3 C4 C8G1/hexanol/Nd(NO3)3 D1 C8G1/TBP D2 C8G1/TBP/LiCl D3 C8G1/TBP/LiNO3 D4 C8G1/TBP/Nd(NO3)3 a γ denotes the cosurfactant/surfactant molar ratio.

nC8G1 (mol  104)

ncosurfactant (mol  104)

csalt (mol/L)

γ

7.9 7.9 7.8 7.5 8.2 8.1 8.1 7.8 8.1 8.1 8.0 7.6 6.4 6.1 6.4 6.0

9.0 9.0 9.0 9.0 15.1 15.1 15.1 15.1 21.0 21.0 21.0 21.0 12.0 12.0 12.0 12.0

0.0 0.3 0.3 0.3 0.0 0.3 0.3 0.3 0.0 0.3 0.3 0.3 0.0 0.3 0.3 0.3

1.14 1.14 1.15 1.20 1.84 1.86 1.86 1.94 2.59 2.59 2.62 2.76 1.89 1.97 1.88 2.00

In this article, TBP was selected with regard to its many implications in the nuclear field and because very few fundamental studies investigate “exotic” cosurfactants2,15-17 because most studies are based on n-alcohol cosurfactants.6,18 The main goal of this article is to show that TBP acts as an efficient cosurfactant for forming microemulsions with controlled curvature. We focus here on WIII-type microemulsions because the bicontinuous structure provides a large accessible oil-water interface that could promote both ion adsorption/depletion and ion transfer from the aqueous to the oil phase. The model system that we studied is composed of water, n-octyl-β-glucoside (C8G1) as the surfactant, dodecane, and TBP or n-alcohols (n =5, 6, and 8) as the cosurfactant. The evaluation of the cosurfactant behavior of TBP is done by comparison to the more standard n-alcohols systems. n-Octylβ-glucoside was chosen because of its high hydrophilicity that enables the hydrophobicity of TBP or chosen n-alcohols to be counterbalanced. First, the phase behavior was studied by establishing the so-called “fish cuts” that shows the realms of existence of Winsor-type systems. Second, the film composition that consists only of a surfactant and cosurfactant was determined by chemical analysis of the excess phases because surfactant and cosurfactant are distributed between excess and microemulsion phases. In addition, the microstructure was investigated by smallangle neutron scattering (SANS) that gives access to the characteristic length scales and specific area. Consequently, a structural description of the microemulsions on the molecular and submolecular scales is obtained in the bulk microemulsions and at the water/oil interface. These results are discussed in terms of the solubilization efficiency, salt effects, interface rigidity, packing of the surfactant/cosurfactant at the interface, and domain connectivity.

Experiments Materials. Surfactant n-octyl-β-glucopyranoside (C8G1, C14H28O6, purity >97%) was purchased from Anatrace. Hydrocarbon dodecane (C12H26, purity 99þ%) was provided by Fluka, and alcohols n-pentanol (C5H12O, purity 99þ%), n-hexanol (C6H15O purity þ98%), and n-octanol (C8H18O, purity þ99%) were (14) Banc, A.; Bauduin, P.; Diat, O. Chem. Phys. Lett. 2010, 494, 301–305. (15) Zech, O.; Thomaier, S.; Bauduin, P.; Rueck, T.; Touraud, D.; Kunz, W. J. Phys. Chem. B 2009, 113, 465–473. (16) Bauduin, P.; Wattebled, L.; Schr€odle, S.; Touraud, D.; Kunz, W. J. Mol. Liq. 2004, 115, 23–28. (17) Schulman, J. H.; Stoeckenius, W.; Prince, L. M. J. Phys. Chem. 1959, 63, 1677–1680. (18) Barnes, I. S.; Derian, P. J.; Hyde, S. T.; Ninham, B. W.; Zemb, T. N. J. Phys. Paris 1990, 51, 2605–2628.

1654 DOI: 10.1021/la104005x

purchased from Sigma-Aldrich. Tributylphosphate ((C4H9O)3PO, TBP, purity þ98%) was purchased from Prolabo. Salts were dissolved in distilled water (Milli-Q) or in D2O obtained from Sigma-Aldrich with a degree of deuteration of >99.9%. The following salts were used: lithium chloride (LiCl), lithium nitrate (LiNO3), and neodymium(III) nitrate hexahydrate (Nd(NO3)3 3 6H2O). The purity of all salts that were provided by SigmaAldrich was >99%. All chemicals were used as received.

Sample Preparations and Phase Diagram Determination. To study phase behavior, samples were prepared by mixing 2 mL of an aqueous solution containing C8G1 and salt with 2 mL of dodecane in sealed tubes. Aliquots of cosurfactant, 25 to 50 μL of either n-alcohols or TBP, were added stepwise to reach a given molar ratio γ of cosurfactant/surfactant (Table 1): γ ¼

ncosurf nsurf

ð1Þ

The samples were shaken using a vortex mixer for 2 min, centrifuged to enhance the phase separation, and thermostatted in a water bath at 25.0 ( 0.1 C for 24 h. Visual observation enables the determination of the phase diagram by determining the phase transition from Winsor I to Winsor III to Winsor II systems. The phase diagrams studied show the regions of existence of the Winsor-type systems. The so-called fish-cut phase diagram (Figure 2 and 3) were realized for different cosurfactants: pentanol, hexanol, octanol, and TBP. For neutron-scattering experiments and chemical analysis, water was replaced by D2O. The molar ratio γ of cosurfactant/ surfactant was fixed at around 1.1, 1.9, and 2.6 in order to be located inside the three-phase region for hexanol and TBP. The compositions of the samples used for the SANS experiments are listed in Table 1. Once phase equilibrium was reached, the three phases were separated with a microsyringe and stored in glass vials. SANS Measurements. Neutron-scattering experiments were carried out at the Helmholtz Zentrum Berlin (HZB) in Berlin, Germany on the V4 instrument.19 Two wavelengths were selected, 4.5 and 6 A˚. To access a q range of 0.03-4 nm-1, sample-to-detector distances of 1, 2, 4, 8, and 16 m were applied. Data were registered on a 64  64 two-dimensional detector, radially averaged, and converted into absolute units according to a standard procedure. Data treatment was done with BerSANS software available at HZB.19 Some of the SANS experiments were also carried out at the Laboratoire Leon Brillouin (LLB, Saclay, France) using the PAXE instrument. Karl Fischer Measurements. The determination of the water content of the μE’s was carried out by a Metrohm KF 764 coulometer with a limit of detection of between 10 μg and 10 mg (19) Keiderling, U.; Wiedenmann, A. Physica B 1995, 213, 895–897.

Langmuir 2011, 27(5), 1653–1661

Bauer et al.

Article

Figure 2. Schematic phase tetrahedron of a quaternary water/ surfactant/cosurfactant/oil system. A cut at the constant water/ oil volume fraction of R = 0.5 shows phase boundaries describing the shape of a fish. Figure 1. Density of cosurfactant/dodecane mixtures as a function of the cosurfactant mass fraction for hexanol (circles) and TBP (squares). These calibration curves are used for the determination of the hexanol and TBP mole fraction in the excess oil phase; the results are collected in Table 2. of water per measurement. The coulometric detection measures the iodine generated during the reaction with water, which is related to the initial water content. All titrations were carried out three times to minimize experimental error. The μE probe was controlled by a microliter syringe and rapidly injected via a septum into the Karl Fischer solution. Total Organic Carbon Analysis (TOC). Total organic carbon measurements were carried out with a Shimadzu TOCV CPH on the excess aqueous phases. This measurement is based on an interference-free method of detecting CO2 down to 1 μg/L (ppb). The diluted sample is injected onto a platinum catalyst at 680 C in an oxygen-rich atmosphere. The carbon dioxide generated is carried by carbon-free gas and measured with a nondispersive infrared (NDIR) detector. The integration of the peak area related to the region of adsorption of infrared light specific to CO2 (around 2350 cm-1) gives, after background subtraction, the concentration of the organic carbon concentration. All measurements were repeated three times. The calibration of the apparatus between 0 and 100 mg/L was done with the help of potassium hydrogen phthalate. Density Measurements of the Excess Oil Phases. Because the three macrophases are in thermodynamic equilibrium, the distribution of cosurfactant between the microemulsion and the excess oil phase can be investigated. The determination of the cosurfactant concentration in the excess dodecane phase is made by density measurement. Therefore, after phase separation the density of the excess oil phase was measured at 25 C using an Anton Paar DSA 5000 precise vibrating tube densitometer. Only cosurfactant is dissolved in the excess oil phase. The residual cosurfactant in the excess oil phase is determined using a calibration curve (Figure 1). By doing so, it is assumed that the monomer concentrations of (i) alcohol in the oil and (ii) surfactant in the aqueous phase inside the micro domains of water and oil of the microemulsion are equal to the concentrations in the excess phases, once thermodynamic equilibrium is reached.

Results and Discussion Fish Cut. The system studied contains four components: water or salt solution, oil, hydrophilic surfactant, and hydrophobic cosurfactant. A tetrahedron representation20,21 of the phase diagram of such a system is schematized in Figure 2. By adding cosurfactant (20) Reimer, J.; Soderman, O.; Sottmann, T.; Kluge, K.; Strey, R. Langmuir 2003, 19, 10692–10702. (21) Kahlweit, M.; Strey, R.; Busse, G. J. Phys. Chem. 1991, 95, 5344–5352.

Langmuir 2011, 27(5), 1653–1661

stepwise to a surfactant-water-oil mixture, two phases are observed at first. A surfactant-rich phase, identified as an oil-inwater microemulsion, is then in equilibrium with an excess oil phase (Winsor I, 2). With an increase in the cosurfactant content, the well-known three-phase body occurs. A microemulsion coexists with excess water and oil (Winsor III, 3). At higher cosurfactant content, the system is driven into a two-phase region where a water-in-oil microemulsion and an excess water phase coexist (Winsor II, 2). At sufficiently high surfactant concentrations, a one-phase microemulsion can be obtained (Winsor IV, 1). By fixing the water-to-oil volume ratio to 1:1, the schematic phase boundaries in the resulting 2D section of the tetrahedron describe the shape of a fish. The head of the fish represents the 2-3-2 phase transitions whereas the tail situated at high surfactant concentrations corresponds to the 2-1-2 phase transitions. Principally, the form of the fish head and the position of the intersection of the two critical lines (Figure 2) are of interest because they are directly related to the solubilization power of the amphiphile. For classical ternary systems composed of oil, water, and polyethoxylated surfactants, temperature is used to tune the curvature of the surfactant film in order to determine the fish cuts.21 Here, because neither the surfactant nor the cosurfactant presents a pronounced temperature dependence,22-24 the cosurfactant-to-surfactant ratio is varied. Hence, the cosurfactant is used to increase the curvature of the surfactant film from direct to reverse systems. The aim of this article is to study the cosurfactant behavior of a highly hydrophobic extractant molecule, here tri-n-butylphosphate (TBP), in comparison to that of classical n-alcohol cosurfactants. Therefore, fish cuts were realized with hydrophilic surfactant C8G1, water, and dodecane by varying the cosurfactant: n-pentanol, n-hexanol, n-octanol, and TBP (Figure 3). For the sake of clarity, the fish diagrams were only partially plotted; the full diagrams are presented in the Supporting Information. For the n-alcohol series, by increasing the chain length the fish is shifted both to (i) lower surfactant concentrations and (ii) lower cosurfactant-to-surfactant ratios. This can be interpreted respectively as (i) an increase in the solubilization efficiency25 (i.e., less surfactant is needed to cosolubilize water and dodecane) and (ii) an increase in the hydrophobicity of the cosurfactant (i.e., less (22) Ruiz, C. C., Ed. Sugar-Based Surfactants; Surfactant Science Series 143; Marcel Dekker: New York, 2009. (23) Kluge, K.; Stubenrauch, C.; Sottmann, T.; Strey, R. Tenside, Surfactants, Deterg. 2001, 38, 40–48. (24) Higgins, E.; Baldwin, W. H. Anal. Chem. 1960, 32, 236–240. (25) Kunieda, H.; Solans, C. How to Prepare Microemulsions: TemperatureInsensitive Microemulsions. In Industrial Applications of Microemulsions; Surfactant Science Series 66; Marcel Dekker: New York, 1997; p 21.

DOI: 10.1021/la104005x

1655

Article

Figure 3. Fish cuts for cosurfactants pentanol (squares), hexanol (triangles), octanol (rhombi), TBP (arrows), and TBP with the addition of Nd(NO3)3 (0.3 M) (stars) as a function of the total surfactant mass fraction, wSURF. (A) Classical mass representation. (B) Mole representation with γ, the cosurfactant/surfactant molar ratio on the y axis.

cosurfactant is required to invert the system from direct to reverse microemulsions). Such an evolution as a function of the alcohol chain length is usually observed in microemulsions.25 Note also that the head of the fish becomes thinner as the length of the n-alcohol increases. This is usually interpreted by an increase in the interfacial film rigidity,26-28 which becomes too stiff to accommodate water and oil (i.e., to form a bicountinuous structure usually observed in Winsor III microemulsions). When rigidity diverges, phase diagrams become temperature-insensitive and phase limits are then imposed by local packing constraints.29 In the case of TBP, the fish diagram is comparable to the one obtained with n-pentanol but is only very slightly shifted to higher surfactant concentrations. Hence, TBP is identified as an efficient cosurfactant for forming microemulsions and is just a little less efficient in terms of solubilization power than n-pentanol. It is interesting that the fish diagrams in Figure 3a are represented as usual with the mass fraction of surfactant on the x axis. (26) De Gennes, P. G.; Taupin, C. J. Phys. Chem. 1982, 86, 2294–2304. (27) Andelman, D.; Cates, M. E.; Roux, D.; Safran, S. A. J. Chem. Phys. 1987, 87, 7229–7241. (28) Kegel, W. K.; Lekkerkerker, H. N. W. Colloids Surf., A 1993, 76, 241–248. (29) Zemb, T. C. R. Chim. 2009, 12, 218–224.

1656 DOI: 10.1021/la104005x

Bauer et al.

If a mole representation is used, for example, by plotting γ versus the mass fraction of surfactant, then the fish determined for TBP is shifted far below the “fishes” corresponding to the n-alcohol series (Figure 3b). Therefore, TBP is far more efficient on a per mole basis than n-alcohols (e.g., 3 times more efficient than n-pentanol) in changing the microemulsion curvature. Because TBP has three saturated hydrocarbon chains, this shows that it forms a mixed film with C8G1. The influence of the presence of salt on the fish cut was also studied for TBP by using 0.3 M Nd(NO3)3 instead of pure water (Figure 3a). Neodymium salt was chosen because it is supposed to have the greatest effect as a result of (i) its three charges and (ii) its strong interaction with TBP, which is known to form a complex with Nd3þ.30,31 A slight shift in the solubilization power to lower surfactant concentration is observed in the presence of Nd3þ. Moreover, the phase transition appears to occur slightly at higher cosurfactant-to-surfactant ratios. This is likely to be related to a salting-in effect, subsequent to ion adsorption at the TBP-functionalized interface. This leads to an increase in the hydrophilicity of organic molecules (i.e., it increases the film curvature toward water). Nevertheless, the addition of salt to the studied system affects the phase transitions only slightly. This is not surprising because salt addition is well known to have large effects on Winsor-type systems based on ionic surfactants and only slight effects on nonionic ones.32 Hence, this observation is confirmed here with TBP, a nonionic extractant, as a cosurfactant and with a nonionic surfactant. Using TBP in the presence Nd3þ does not affect the phase transitions very much. We will now go further by studying the microemulsion structure of alcohols and TBP. Microemulsion Structure and Effect of Salt. Determination of the Water Volume Fraction in the Microemulsion. The volume fraction of the aqueous phase in the microemulsion (φW) gives important information on the microstructure. The system structure tends to oil-in-water and water-in-oil systems when φW is above and below 0.5, respectively, whereas bicontinuous structure appears at φW of around 0.5. The water content was determined by the Karl Fischer method. By using the densities of the microemulsions and the corresponding aqueous phases containing salts at a concentration of 0.3 M, the water content (in g of H2O/g of microemulsion) was converted into volume fractions of the aqueous phase in the microemulsion (φW) with an accuracy of 1%. The values of φW decrease systematically with increasing γ (Figure 4). This reflects the evolution from direct water-rich swollen micelles (oil in water) to the oil-rich reverse micelles (water in oil) via the bicontinuous microemulsion. The water content of the microemulsion varies slightly with different salts. At γ = 2.0, the water content, φW, in the microemulsion is always higher for hexanol than for TBP as the cosurfactant (e.g., by comparing B1/D1 or B2/D2 samples; Table 1). This has to be related to the larger curvature toward the aqueous phase for the hexanol μE compared to that for the TBP μE. This result confirms that TBP is far more efficient than hexanol in changing the microemulsion curvature. SANS: Scattering Peak. To study the microstructure of the microemulsions and to access their length scales, SANS experiments were carried out. The goal here is to highlight (i) the effect of TBP and hexanol as cosurfactants and (ii) the effect of the nature of salt on the bicontinuous structure of microemulsions at (30) Keskinov, V. A.; Lishchuk, V. V.; Pyartman, A. K. Russ. J. Inorg. Chem. 2007, 52, 1144–1146. (31) Sawada, K.; Hirabayashi, D.; Enokida, Y. Prog. Nucl. Energy 2008, 50, 483–486. (32) Yamaguchi, Y.; Aoki, R.; Azemar, N.; Solans, C.; Kunieda, H. Langmuir 1999, 15, 7438–7445.

Langmuir 2011, 27(5), 1653–1661

Bauer et al.

Article

Figure 4. Volume fraction of water in the microemulsions (φW) vs γ, the cosurfactant/surfactant molar ratio for hexanol microemulsions (full symbols) and for TBP microemulsions (open symbols): without salt (stars), LiCl (squares), LiNO3 (circles), Nd(NO3)3 (triangles).

a given composition. To do so, Winsor III samples B and D were compared. Figure 5 shows the bulk contrast scattering curves for TBP and hexanol microemulsions at γ = 1.9 in the presence of 0.3 M LiCl, LiNO3, or Nd(NO3)3 plotted in a log-log representation. As it is usually observed with bicontinuous microemulsions, the scattering curves show a characteristic interaction peak. Interestingly, a structure peak is obtained with TBP as the cosurfactant, indicating that it forms microemulsions, as is the case with hexanol. The spectra are well adjusted with the Teubner-Strey model33 (eq 1), which can be used to describe small-angle scattering by microemulsions IðqÞ ¼

1 þ Iincoh a þ bq2 þ cq4

ð2Þ

where I(q) is the neutron scattering intensity on the absolute scale in cm-1, q is the wave vector defined as q = 4π/λ sin(θ/2) with θ being the scattering angle and λ being the wavelength of the incident beam, and Iincoh is the incoherent background. The fitting procedure of the scattering curves (Figure 5) then gives access to the characteristic length scales (i.e., the persistence length ξTS and the domain size D*TS of the bicountinuous structure, which are defined by eqs 3 and 4) " rffiffiffi # - 0:5 a b ð3Þ þ 0:25 ξ ¼ 0:5 c c "

rffiffiffiffiffiffiffiffi a b - 0:25 c c

# - 0:5

Figure 5. Neutron-scattering intensity on the absolute scale, I(q),

as a function of the wave vector, q, for samples B and D at γ = 1.9 containing either hexanol or TBP as the cosurfactant and different salts or no salt. For clarity, the curves were displaced by a factor of 2 for the hexanol microemulsion and 1/2 for TBP microemulsions. The solid lines correspond to the best fit obtained with the Teubner-Strey model.

Factor fa amounts to 1 for the disorder line, where the solution loses its quasiperiodic order.36 The liquid-crystalline lamellar phase corresponds to fa= -1.35 These values delimit the region where microemulsions may be found.37 Factor fa ranges from -0.9 to -0.7 for well-structured bicontinuous microemulsions.34,35 The values obtained here for microemulsions containing hexanol as a cosurfactant range from -0.6 to -0.7 (Table 2), which is close to the values obtained for well-structured microemulsions. Moreover, microemulsions for which the alcohol is replaced by the more bulky TBP slightly decrease the amphiphilic factor, here between -0.4 and -0.6. This can be explained by the decrease in the stiffness of the interfacial film for the case of TBP. The values of fa, D*, and ξ are listed in Table 2. The persistence length is around one-third of the domain size, ξTS/D* ≈ 1/3. This ratio appears to be slightly lower for TBP than for hexanol systems. This is mainly due to a lower persistence length of the interface for TBP, compared to hexanol, microemulsions whereas the domain size is less affected. The persistence length provides information about the stiffness of the surfactant/cosurfactant film: the shorter the persistence length, the softer the monolayer. The persistence length can be related to the bending constant (Kc) of the film by ξ ¼ t exp2πKc =kB T

ð6Þ

ð5Þ

where t is the film thickness and kBT is the thermal energy.26 For our systems, the persistence length for microemulsions without salt (samples B1 and D1) differs significantly, from about 1 nm, from hexanol to TBP at exactly the same sample composition. It can therefore be concluded that mixed surfactant films containing hexanol as a cosurfactant are less fluid than those containing TBP. This is in good agreement with the discussion of the fish-cut diagrams. Indeed, the conclusion is that the stiffness of the film is highly dominated by the chain length and the number of chains of the cosurfactant used to form the three-phase microemulsions.

(33) Teubner, M.; Strey, R. J. Chem. Phys. 1987, 87, 3195–3200. (34) Schubert, K. V.; Strey, R.; Kline, S. R.; Kaler, E. W. J. Chem. Phys. 1994, 101, 5343–5355.

(35) Engelskirchen, S.; Elsner, N.; Sottmann, T.; Strey, R. J. Colloid Interface Sci. 2007, 312, 114–121. (36) Chen, S. H.; Chang, S. L.; Strey, R. J. Chem. Phys. 1990, 93, 1907–1918. (37) Leitao, H.; Dagama, M. M. T.; Strey, R. J. Chem. Phys. 1998, 108, 4189–4198.

D ¼ 2π 0:5

ð4Þ

where a, b, and c are fitting parameters. The accuracy of ξTS and D*TS values is around 0.1 nm. Moreover, the so-called amphiphilic factor fa can also be determined according to eq 5.34,35 fa ¼

b ð4acÞ1=2

Langmuir 2011, 27(5), 1653–1661

DOI: 10.1021/la104005x

1657

Article

Bauer et al.

Table 2. Results of the Karl Fischer Measurements and the SANS Data Treatment of the System D2O/C8G1/Dodecane/Salt/Cosurfactanta sample name

ΔF (1010 cm-2)

Φw

Σ (106 cm-1)

Iincoh (cm-1)

Qtheo/Qexp

ξ (nm)

fa

D* (nm)

ξ/D*

B1 0.36 6.6 2.58 0.43 1.18 -0.69 3.7 10.8 0.35 B2 0.37 6.6 2.21 0.42 1.26 -0.69 3.8 10.9 0.35 B3 0.35 6.7 2.05 0.41 1.18 -0.65 3.9 11.3 0.35 B4 0.34 6.8 2.08 0.40 1.25 -0.66 4.3 12.5 0.35 D1 0.22 6.5 2.02 0.58 1.02 -0.50 2.8 10.4 0.27 D2 0.24 6.5 2.80 0.51 1.12 -0.44 2.7 9.7 0.28 D3 0.25 6.6 2.63 0.44 1.20 -0.48 2.9 9.8 0.29 D4 0.26 6.7 2.24 0.50 1.29 -0.58 3.3 10.2 0.32 a Water volume fraction of the microemulsion, φ W; difference in the scattering-length density between the aqueous phase (water/salt) and the organic phase (dodecane/cosurfactant), ΔF; specific surface, Σ; incoherent scattering, Iincoh; theoretical and experimental invariants, Qtheo and Qexp; amphiphilic factor, fa; correlation length, ξ; and domain size, D*.

According to Schulman et al.17 and De Gennes et al.,26 cosurfactants increase the disorder of the interfacial film, which consequently makes the film more fluid. This seems to be even more pronounced for a bulky cosurfactant such as TBP compared to single-chain alcohols. For TBP and hexanol microemulsions at the same composition but with salt in the aqueous phase, differences in the length scales of the microemulsions are observed. Microemulsions formed with hexanol and containing salt (LiCl, LiNO3, and Nd(NO3)3 at 0.3 M, samples B2, B3, and B4, respectively) show an increase in both the persistence length and the domain size, resulting in a constant ξTS/D* value. In the case of TBP microemulsions (samples D2, D3, and D4), different behavior is observed. As the persistence length increases by varying the type of salt, the domain size of the aqueous phase or the organic phase remains unchanged. As was already discussed above, this might be a consequence of the decreased stiffness of the interfacial film. A softer interface does not necessarily force the system to enlarge the domain size whereas a more rigid interface tends to form larger domain sizes. We can compare our findings to SANS measurements performed by Wellert et al. on three-phase microemulsions composed of an alkyloligoglucoside surfactant, pentanol, and biodiesel oil.38 In this study, the persistence lengths are about 4.5 nm and the domain sizes are also slightly higher than in our case, about 13 nm. As for our microemulsions with hexanol, the ξTS/D* values are about 1/3. This means (i) that the microemulsion structure observed in ref 38 is comparable to the hexanol microemulsions studied here and (ii) that TBP induces a change in the microemulsion structure that is sensitive to the addition of salts. This is due to the presence of the three alkyl chains and the chelating polar part of TBP. SANS: Porod Region and the Invariant. At large q values, the scattering cross section decreases as q-4, known as Porod decay, until the incoherent background scattering, Iincoh, is reached. The internal interface or specific surface (S/V or Σ) can be determined from the large q part of the spectrum after background (Iincoh) subtraction. In our case, the Porod regime is observed even without subtraction of the background (Figure 5). lim IðqÞq4 ¼ 2πΔF2 Σ

qf¥

ð7Þ

Equation 7 describes the Porod limit where ΔF is the scatteringlength-density differences between the polar deuterated and the apolar hydrogenated part.39,40 ΔF values (Table 2) were estimated from the compositions of the aqueous and oil phases (Supporting (38) Wellert, S.; Karg, M.; Imhof, H.; Steppin, A.; Altmann, H. J.; Dolle, M.; Richardt, A.; Tiersch, B.; Koetz, J.; Lapp, A.; Hellweg, T. J. Colloid Interface Sci. 2008, 325, 250–258. (39) Porod, G. Kolloid Z. 1951, 124, 83–114. (40) Kratky, O.; Glatter, O. Small Angle x-Ray Scattering; Academic Press: London, 1982.

1658 DOI: 10.1021/la104005x

Information). The accuracy of the ΔF and Σ values are estimated to be around (0.1  1010 cm-2 and (0.05  10-6 cm-1, respectively. The Porod regime is obtained only if a sharp interface separates two media of different scattering-length densities.39 This hypothesis is verified by evaluating the so-called experimental invariant (Qexp, eq 8), which is a general property of scattering spectra and is calculated from the scattering intensity.41 Z ¥ Qexp ¼ IðqÞq2 dq ð8Þ 0

Qexp is then calculated after background subtraction over the whole accessible q range. The invariant can also be calculated theoretically (Qtheo, eq 9) from the volume fraction of the polar phase φW and the scattering-length-density contrast ΔF. When both invariants are known with good confidence, Qtheo/Qexp represents the accuracy of the calibration. Qexp and Qtheo are indeed comparable (Table 2). Qtheo ¼ 2π2 φð1 - φÞΔF2

ð9Þ

Dilution Plot. Another approach to describe microemulsion structures is to treat the problem purely geometrically by assuming that the cost of deviations from the optimal volume fractions, interfacial area, and curvature is sufficient to justify treating these as constraints.18,29 The practical method of studying microemulsion structures is to draw a dilution plot of dimensionless quantity ΣD* versus φW and to compare the experimental values to theoretical predictions that describe different topologies (Figure 6). A critical description of existing models for microemulsions is given in ref 29. The most pertinent models for the description of microemulsions are exposed here. The cubic random cell (CRC) model assumes a set of cubes filled randomly with aqueous and organic phases and gives this simple analytical expression: ΣD* = 6φW(1 - φW). This model holds for bicountinuous structures and usually gives a good approximation for flexible interfaces. If curvature constraints come into play, then the gradual transformation from droplets to locally connected cylinders and finally to disordered lamella (sponge or vesicles) is more relevant. This consideration is taken into account in the disordered connected models (DOC) for cylinders and lamella that are valid whenever fluctuations are not dominant. The DOC models are based on a Vorono_i cell tessellation of space, yielding a complete set of microstructures ranging from isolated spheres via connected cylinders to disordered lamellae. The input parameters are φW and Σ, obtained respectively from Karl Fischer measurement and the Porod treatment of the SANS data, and the surfactant (C8G1) chain length in an extended conformation (11.6 A˚). For the DOC lamella model, the (41) Spalla, O. In Neutrons, X-rays and Light: Scattering Methods Applied to Soft Condensed Matter; Lindner, P., Zemb, Th., Eds.; Elsevier: Amsterdam, 2002.

Langmuir 2011, 27(5), 1653–1661

Bauer et al.

Figure 6. Dilution plot, ΣD* vs φW, where the experimental values are given for the hexanol microemulsions (open symbols) and for the TBP microemulsions (solid symbols): without salt (triangles), LiCl (circles), LiNO3 (diamonds), and Nd(NO3)3 (squares). Theoretical predictions that describe different microemulsion topologies are shown by lines: CRC model (full lines), repulsive spheres (dashed lines), DOC cylinder model (dotted lines) with a connectivity of Z = 2, 3, and 5, and the DOC lamellae model (dasheddotted lines).

asymmetry parameter Ψ was fixed at 0.3, corresponding to the case of an asymmetric sponge structure. For the DOC cylinder model, the connectivity Z was varied from 2 to 3 to 5, corresponding respectively to cylinders, connected cylinders, and a network of cylinders. For the case of repulsive spheres, the analytical relation is ΣD* = 4.326φW2/3.42 All of these models are plotted in Figure 6 with the experimental data for the hexanol and TBP microemulsions with or without salts. For the hexanol microemulsions, it can be noted (i) that the ΣD* values are close to the prediction of the DOC cylinder model with a connectivity of 2 and (ii) that the effect of adding salt, as well as the type of salt added, does not significantly influence the ΣD* values. On the contrary, TBP microemulsions show large differences in their ΣD* values according to the nature of the salt added. The experimental ΣD* values range from the DOC cylinder model with a connectivity from 2 to 5 to the DOC lamella model, respectively, for microemulsions containing Nd(NO3)3, no salt, or LiNO3 and LiCl. Consequently, the addition of LiNO3 has nearly no effect on the microstructure whereas Nd(NO3)3 and LiCl decrease the connectivity of the cylinders and lead to the formation of connected lamella. Interestingly, we have shown recently that large cylindrical objects, detected by the q-1 dependence of the SANS pattern, could form near the Winsor II region at higher γ values.43 The DOC cylinder model locally describes cylinder structures that do not show a q-1 dependence in the SANS or SAXS pattern. The specific salt effect observed here can be discussed in terms of the Hofmeister series of ions.44,45 In these series, anions or cations can be classified according to their relative effectiveness over a wide range of phenomena: colloid stability, ion binding (42) Zemb, T. In Neutrons, X-rays and Light: Scattering Methods Applied to Soft Condensed Matter; Lindner, P., Zemb, Th., Eds.; Elsevier: Amsterdam, 2002. (43) Bauer, C.; Bauduin, P.; Diat, O.; Zemb, T. Tenside, Surfactants, Deterg. 2010, 47, 307–311. (44) Kunz, W. Curr. Opin. Colloid Interface Sci. 2010, 15, 34–39. (45) Kunz, W. Specific Ion Effects; World Scientific Publishing Company: Singapore, 2010. (46) Petrache, H. I.; Zemb, T.; Belloni, L.; Parsegian, V. A. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 7982–7987.

Langmuir 2011, 27(5), 1653–1661

Article

to interfaces,46 protein activity,47 and crystallization. Ions with a strong propensity to adsorb at an interface or to be depleted from an interface are usually referred to as chaotropes (structure breaking, such as NO3-) and kosmotropes (structure making, such as Liþ and Nd3þ). Chloride ions are usually considered to be neutral in terms of these two denominations. Consequently, LiCl has a kosmotropic effect and can lead to depletion at the water/oil interface that induces the transition from connected cylinders to local lamellar structure by decreasing the curvature. LiNO3 is composed of a chaotropic anion and a kosmotropic cation and then has a rather neutral Hofmeister effect that does not influence the microemulsion structure (i.e., ΣD* remains constant). Concerning the effect of Nd(NO3)3, it is known that TBP strongly interacts with Nd3þ,30,31 hence the resulting salt effect is based more on a complexation effect rather than a second-order Hofmeister effect. The influence of Nd(NO3)3 on the microstructure is then coherent with the adsorption of Nd3þ leading to a decrease in the cylinder connectivity. Determination of the Film Composition. The description of the interfacial film on the molecular level was investigated by performing quantitative analysis in order to determine the composition of the mixed surfactant/cosurfactant film. In three-phase Winsor-type microemulsions, the cosurfactant is distributed among the excess oil (mcosurf,oilexcess), the oil microphase inside the microemulsion (mcosurf,oilμE), and the interfacial film (mcosurf,filmμE). Therefore, a mass action law can be expressed (eq 10) μE μE mcosurf , total ¼ mexcess cosurf , oil þ mcosurf , oil þ mcosurf , film

ð10Þ

with mcosurf,total being the total mass of the cosurfactant in the sample, mcosurf,oilexcess being the mass of the cosurfactant in the excess oil, mcosurf,oilμE being the mass of the cosurfactant in the oil of the microemulsion, and mcosurf,oilμE being the mass of the cosurfactant in the microemulsion surfactant film. Through density measurements of the excess oil phase and via a calibration curve (Experiments section), the residual cosurfactant concentration was obtained. It is assumed with good confidence that the composition of the bulk oil microphase inside the microemulsion is the same as the composition of the excess oil phase. The amount of cosurfactant implied in the formation of the surfactant film, in the microemulsion, is then calculated from the difference between the total cosurfactant mass fraction in dodecane (wcosurf, total) and the cosurfactant mass fraction in the excess oil phase (wcosurf, excess) (eq 11) wcosurf , total - wcosurf , excess ¼ wcosurf , film ¼

ncosurf , film Mcosurf moil, total

ð11Þ

with ncosurf,film being the number of moles of cosurfactant in the surfactant film, Mcosurf being the molar mass of the cosurfactant, and moil,total being the total mass of oil in the sample. In the same way, a mass action law for the surfactant can be expressed (eq 12). The surfactant is distributed among excess water (msurf,waterexcess), the water microdomains inside the microemulsion (msurf,waterμE), and the surfactant film (msurf,filmμE). μE μE msurf , total ¼ mexcess surf , water þ msurf , water þ msurf , film

ð12Þ

The main feature of three-phase microemulsions is that the surfactant has a very strong affinity for the interface. When equal quantities of oil and water are cosolubilized, those systems are (47) Bauduin, P.; Renoncourt, A.; Touraud, D.; Kunz, W.; Ninham, B. W. Curr. Opin. Colloid Interface Sci. 2004, 9, 43–47.

DOI: 10.1021/la104005x

1659

Article

Bauer et al.

Table 3. Comparison of the Initial Cosurfactant/Surfactant Molar Ratio, γ initial, and the Molar Ratio in the Interfacial Film, γ interface, for the Case of Hexanol and TBP Taking into Account the Cosurfacant or Surfactant Concentrations in the Bulk Excess Phasesa sample name

TOC (g/L)

γ initial

ncosurf excess oil (mmol)

ncosurf interface (mmol)

ncosurf total (mmol)

γinterface

1.570 1.550 1.585 1.574 1.220 1.219 1.228 1.200

0.928 0.982 1.020 0.980 0.325 0.373 0.383 0.314

B1 3.56 1.8 0.863 0.707 B2 3.28 1.9 0.807 0.743 B3 3.71 1.9 0.819 0.767 B4 3.22 1.9 0.862 0.712 D1 n.d. 1.9 1.029 0.190 D2 2.79 2.0 1.100 0.215 D3 2.59 1.9 0.997 0.231 D4 2.49 2.0 1.020 0.180 a The total organic carbon (TOC) of the excess aqueous phases is expressed in grams of carbon per liter.

called “balanced” or at “optimal formulation”. It appears when the surfactant has exactly the same affinity for the oil as for the aqueous phase.48 Therefore, in Winsor III systems the surfactant is mainly present at the interface and is supposed to have a negligible concentration in the water phase, meaning that msurf,waterexcess and msurf,waterμE could be neglected over msurf,filmμE. This was checked by measuring the total organic carbon (TOC) in the excess water phase; see the results in the Supporting Information. TOC is a sensitive quantitative method for detecting the presence of carbon coming from residual surfactant and cosurfactant in water. It was found that the carbon residual concentration ranges from 2.5 to 3.5 g/L, which is less than the critical micelle concentration (cmc) of C8G1 in water (cmc < 10-2 M). This means that only a few monomers stay in the bulk water. This is coherent with values found in the literature for other Winsor III systems.49 Hence, nearly all of the surfactant is located at the water/oil interface. Furthermore, it is assumed that the compositions of the excess water phase and the water microdomains in the μE are the same. From mass balance, the composition of the interfacial film expressed as the molar ratio of cosurfactant over surfactant at the interface, γ interface, is calculated and compared to the initial cosurfactant/surfactant molar ratio values, γ initial (Table 3). For microemulsions prepared with hexanol or TBP as the cosurfactant, γ interface values of 0.92-1 and 0.32-0.37 are obtained, respectively. Microemulsions containing TBP have 3 times fewer moles of cosurfactant at the water/oil interface than microemulsions containing hexanol. The comparison between the two systems can be made because they exhibit similar phase behavior and the initial conditions such as the surfactant, nature of the oil, and salinity are identical. Tentatively, this difference can be explained (i) in terms of the size of the cosurfactant because TBP has a cross-sectional area that is 3 times larger than hexanol and (ii) in terms of the affinity of the cosurfactant for the oil. TBP and hexanol are both very soluble in the oil. Consequently, there are two competing mechanisms in the phase equilibrium: the incorporation of TBP at the surfactant interface and the solubilization in the oil. This effect can be visualized in the fish cut. In the lowsurfactant-concentration regime, the fish-cut diagrams are distorted toward lower cosurfactant/surfactant molar ratios. This means that a higher fraction of alcohol is needed to obtain a balanced interfacial film. A horizontal fish could therefore be obtained by presaturation of the oil phase with the cosurfactant.50 Stubenrauch et al. studied microemulsion systems containing water, C8G1, cyclohexane, and geraniol, with the latter being a doubly unsaturated monoterpenoid alcohol.49 They studied the properties of geraniol as a cosurfactant and cosolvent in 1:1 oil-to(48) Salager, J. L.; Anton, R. E.; Sabatini, D. A.; Harwell, H. J.; Acosta, E. J.; Tolosa, L. I. J. Surfactants Deterg. 2005, 8, 3–21. (49) Stubenrauch, C.; Paeplow, B.; Findenegg, G. H. Langmuir 1997, 13, 3652– 3658. (50) Stubenrauch, C. Curr. Opin. Colloid Interface Sci. 2001, 6, 160–170.

1660 DOI: 10.1021/la104005x

Figure 7. Scheme of the possible molecular organization of C8G1/ TBP (a) and C8G1/n-hexanol (b) in the interfacial layer of balanced microemulsions.

water volume fraction microemulsions. They found a mole ratio of geraniol in the interfacial film of around 0.28 corresponding to a geraniol-to-C8G1 ratio of about 2:5. Penders et al. reported for the system water/n-octane/C8E5/1-octanol a ratio of alcohol to surfactant of 2:3.51 We find for hexanol typically a mole ratio of 1 corresponding to a ratio of hexanol molecules to C8G1 of 1:1 whereas for TBP the ratio is 1:3. Note, however, that comparisons to other data in the literature should be made only for comparable systems because the nature of the oil or surfactant affects the spontaneous curvature. An illustration of the molecular organization of C8G1 with TBP in balanced interfacial layers is proposed in Figure 7. This is compared to a film of similar curvature containing n-hexanol, with an orientation of the hydrophilic phosphate groups toward the water domains and that of the three n-butyl chains toward the oil domains.

Conclusions It was proven here that TBP, a benchmark ion extractant used in liquid/liquid extraction process, can efficiently play the role of a cosurfactant in the formation of balanced microemulsions. The phase diagram of TBP microemulsions in the three-phase region is comparable to the ones formulated with the more classical cosurfactants: n-pentanol and n-hexanol. However, TBP appears to be 3 times more efficient per mole than hexanol in the formation of three-phase microemulsions. This effect is reflected in the interfacial film composition with a surfactant to cosurfactant mole ratio of 1:1 for TBP and 1:3 for n-hexanol. This is directly related to the difference in the packing of TBP and hexanol, with TBP having three alkyl chains compared to one for n-hexanol. It is interesting that the film composition is “adjusted” in order to obtain a nearly zero curvature of the interfacial film, a condition necessary for the formation of three-phase microemulsions. Moreover, it was observed that replacing hexanol by TBP also induces a salt dependence on the microemulsion structure. (51) Penders, M. H.; Strey, R. J. Phys. Chem. 1995, 99, 10313–10318.

Langmuir 2011, 27(5), 1653–1661

Bauer et al.

Article Table 4. List of the Abbreviations Used

abbreviation C8G1 TBP μE ξ D* Σ or S/V φW PUREX SANS γ TOC I(q) q

abbreviation octyl-β-glucoside tributylphosphate microemulsion persistence length aqueous or organic domain size specific surface water volume fraction in the microemulsion plutonium and uranium refining by extraction small-angle neutron scattering cosurfactant to surfactant molar ratio total organic carbon neutron-scattering intensity on the absolute scale wave vector

A subtle change in the microemulsion structure according to the type of salt added is highlighted in the ΣD* versus φW dilution plot. This effect was rationalized in terms of the Hofmeister series and the complexation property of TBP.

fa Kc t kT Iincoh ΔF Qexp and Qtheo CRC DOC Ψ Z w cmc

amphiphilic factor bending constant surfactant film thickness thermal energy incoherent background scattering scattering-length-density difference experimental and theoretical invariants cubic random cell disordered connected model asymmetry parameter connectivity mass fraction critical micellar concentration

contact at Laboratoire Leon Brioullin, for valuable support with respect to the small-angle scattering experiments and Veronique Dubois (ICSM) for the TOC measurements of the samples and for her assistance with the Karl Fischer measurements.

Nomenclature The abbreviations used in this work are listed in Table 4. Acknowledgment. We thank Dr. Sylvain Prevost, local contact at Helmholtz Zentrum Berlin, and Dr. Jacques Jestin, local

Langmuir 2011, 27(5), 1653–1661

Supporting Information Available: Fish-cut diagrams for the different cosurfactants. Scattering-length-difference calculation. This material is available free of charge via the Internet at http://pubs.acs.org.

DOI: 10.1021/la104005x

1661