You are on page 1of 8

PHYSICAL REVIEW E

VOLUME 61, NUMBER 1

JANUARY 2000

Adiabatic compressibility of AOT †sodium bis„2-ethylhexyl…sulfosuccinate‡ reverse micelles: Analysis of a simple model based on micellar size and volumetric measurements
1

´ rou,1 W. Urbach,2 D. Valdez,1 and M. Waks1 A. Amararene,1 M. Gindre,1 J.-Y. Le Hue
´ trique, CNRS UMR 7623, 15 rue de l’Ecole de Me ´ decine, F-75006 Paris, France Laboratoire d’Imagerie Parame 2 Laboratoire de Physique Statistique de l’ENS, CNRS URA 1306, 24 rue Lhomond, F-75005 Paris, France ͑Received 13 April 1999͒ The self-assembly of amphiphilic molecules into supramolecular aggregates involves a number of complex phenomena and forces. Recent developments of highly sensitive, densimetric and acoustic methods on small volume samples have provided novel sensitive probes to explore the physical properties of these complex fluids. We have investigated, by high precision densimetry and ultrasound velocimetry, reverse micelles of ͓sodium bis͑2-ethylhexyl͒sulfosuccinate͔ in oil ͑isooctane and decane͒, at increasing water concentration and at variable micellar volume fractions. The size of these spherical micelles has been determined by small angle x-ray scattering. Using these results, in the framework of the effective medium theory, we have developed a simple model of micellar compressibility, allowing the calculation of physical parameters ͑aggregation number, volume, and compressibility͒ of the surfactant monomolecular film as well as that of the micellar waters. In particular, we show that the central aqueous core designated as ‘‘free’’ water, located at a distance from the oil-water interacting interface, is twice as compressible as ‘‘bulk’’ water. One notable feature of this work is the influence of the nature of the oil on the above parameters. PACS number͑s͒: 62.10.ϩs

I. INTRODUCTION

Water-in-oil microemulsions or reverse micelles have attracted the attention of physicists as early as 1943 ͓1͔. Since then, many characteristics of reverse micelles have been extensively studied and documented ͓2,3͔. Reverse micelles also display a number of versatile properties: they have been considered as effective membrane-mimetic systems ͓4͔, models of water-restrictive environment of biological relevance ͓5͔, microreactors for micellar enzymology or microchemistry ͓6,7͔, and recently a medium for membrane protein crystallization ͓8͔. Reverse micelles can be described as water microdroplets of variable size, dispersed in nonmiscible apolar solvents and stabilized by a monolayer of surfactant. Their nonpolar tails protrude into and are solvated by the organic solvent, while their polar head groups are in direct contact with the water core that solvates them. Thus, reverse ͑or inverted͒ micelles are inside out with respect to more common micelles. The architecture of these organized molecular assemblies delineates two distinct compartments: the smallest one, on the inside, contains all the water present, while the largest one is made by the bulk organic phase. The two compartments are separated by a boundary, the monomolecular layer of surfactant, an interacting interface between oil and water. One of the most common systems used by investigators is the ternary mixture ͓sodium bis͑2-ethylhexyl͒sulfosuccinate͔ ͑AOT͒ –water–apolar solvent. The size of these droplets, spherical in shape, which are thermodynamically stable structures, depends only on water concentration, defined as the water-to-surfactant molar ratio W 0 . When water is added, the micelles swell, their radius increases, and the number of surfactant molecules per droplet ͑aggregation number͒ grows with the size of the droplet. As
1063-651X/2000/61͑1͒/682͑8͒/$15.00 PRE 61

a consequence, the water content of the system can be precisely varied and experimentally controlled. Provided W 0 is held fixed, the volume fraction of the droplets can be varied without changing their shape and size. There is a wealth of information concerning reverse micelles, their structure, phase behavior ͓9–11͔, and the anomalous properties of micellar water, comprizing at least two populations in rapid equilibrium: ‘‘bound’’ interfacial and ‘‘free,’’ developing as hydration of the system is increased ͓12,13͔. In contrast, information concerning adiabatic compressibility of reverse micelles in organic solvents remains scarce ͓14͔. Recent developments in acoustic techniques ͑difference ultrasound velocimetry͒ have rendered possible high precision measurements with small solute volumes ͓15͔, in particular biopolymers ͓16,17͔ or complex fluids ͓18͔. It is now well established that compressibility is exquisitely sensitive to solute-solvent interactions; at low solute concentration, the measured volume and compressibility properties reflect therefore the contribution of both solvent and solute intrinsic properties. A major issue in interpreting the results thus originates from the difficulty to discriminate between the intrinsic compressibility of a solute and that due to interactions occurring at the interface in contact with the solvent. An additional complexity of reverse micelles is generated by the possible interplay with different organic solvents ͓10͔. One of the possible approaches of this interesting problem is to establish a microscopic model of the various compartments described above, which might account for experimental results. To achieve this goal, we report here the determination of density and ultrasound velocity of AOT reverse micellar solutions, at water concentration corresponding to W 0 values ranging between 0 and 30, and at variable micellar volume fraction. All experiments were carried out in isooctane and
682 ©2000 The American Physical Society

we have been able to predict the volume and compressibility of the AOT monomolecular film itself. Taking into account the densities ( ␳ oil). were either pure or organic solvents. RESULTS AND DISCUSSION A. expressed in kg mϪ3. The x-ray apparent source had dimensions 0. identical cells enclosed in a single metal bloc thermostated at 25 Ϯ 0. as well as that of encased waters. in two chemically distinct apolar solvents: isooctane is branched and decane is linear. However. B. then. ͑6͒. The apparatus is automated using a PC equipped with an IEEE interface. V ͑2͒ where V is the volume of the solution and V mic the volume of micelles in the solution. A 10 MHz sine signal is gated by short pulses ͑400 ns͒. Ultrasound velocity measurements Improvements of the above expression have been obtained by including the effect of fluid viscosity and particle size. MATERIAL AND METHODS AOT was purchased from Sigma ͑SigmaUltra͒ 99% pure. Detailed characteristics of apparatus and experimental performances have been given elsewhere ͓18͔. Micellar density All the samples were prepared by weighing the solutes ͑surfactant and water͒ on a Model 1712 Sartorius balance. mass conservation law can be expressed as ͓22–24͔: ␳ ϭ ␳ oil␾ oilϩ ␳ mic␾ mic . we have modeled the micellar compressibility with the help of simple geometrical considerations.03 mg. Both cells are first filled with the reference solvent.00Ϯ 0. Small angle x-ray scattering „SAXS… leading to an alternative form of Eq. The dimensions of the beam on the counter are 3 mm vertically and 0. in precision volumetric flasks ͑class A Ϯ 0. With the total volume fraction being ͑4͒ ␾ micϩ ␾ oilϭ 1. Isooctane. 683 decane for comparizon of solvent effect. By doing so. with a precision of Ϯ 0. ␾ micϭ V mic . Sample preparation zontally. ͑1͒ The densities ␳ ( c ) of solutions. From these results. .1 mmϫ 0. The solvents used to make up the volume at 20 °C. a 50 mm useful length. the corrective term is irrelevant ͓25͔. relation ͑1͒ becomes ͑5͒ ␾ micϭ 1 Ϫ and ␳ Ϫ c mic ␳ oil ͑6͒ Samples were filled at 22 °C in Lindeman capillaries of 1 mm diameter and sealed. have been used to determine the waterpool radius R w . A nickel filter attenuates the K ␤ waves. The maximum error introduced by this procedure does not exceed the reported error limits. III. one can easily calculate ␾ mic .54 Å wavelength. A vertical mirror acts as a total reflector for the ␭ K ␣ ϭ 1. .01 °C.04 ml). Figure 1 displays the plot at a . as a function of the micellar volume fraction ␾ mic .01 °C using the vibrating tube Anton Paar DMA 58 digital density meter. The x-ray generator was a copper rotating anode machine operating at 40 kV and 25 mA. The distance between the sample and the counter is 802 mm ͓19͔. The size of droplets was determined by small angle x-ray scattering ͑SAXS͒. then the second cell is filled with the solution. II. was from Merck and decane Ͼ99% pure. using Eq.( ␾ mic) of oil and micelles respectively.3 mm hori- ␳ ϭ ␳ oilϪ ␾ mic͑ ␳ oilϪ ␳ mic͒ . The final precision in ultrasound velocity determination is better than 10Ϫ 5 . D. eliminates shorter wavelengths of the beam and directs the x rays on the positive proportional counter. ͑3͒ They were carried out on a custom-built apparatus. and a 200 ␮m spatial resolution. The counter has a window of 3 mm height.1 mm. analyzed using the general equation for the scattered intensity derived by Guinier and Fournet ͓20͔ as described by ͓21͔.( ␳ mic) and volume fractions ( ␾ oil). The issued signal drives simultaneously the emitting transducers ͑ceramic͒ of both cells and the output signals are captured by a digital scope. Each determination was carried out at least five times and averaged. The precision obtained is of the order of Ϯ 5 ϫ 10Ϫ 3 kg mϪ3. from Sigma. Ultrasound velocity is measured in two small ͑3 ml͒. due to the small size of micelles studied in this work ͑Ͻ100 Å͒ and the low ␾ values used in experiments. The micellar concentration c mic can be written as a function of mass of water and that of AOT: c micϭ m AOTϩ m H2O V . at increasing solute concentrations c. The size polydispersity was estimated between 15 and 20 % according to W 0 values.PRE 61 ADIABATIC COMPRESSIBILITY OF AOT . and used after desiccation in vacuum over phosphorus pentoxide ͑Sicapent from Merck͒. The obtained radius and the calculated volume of the sphere include the surfactant polar head group volume of identical electronic density. ␳ ϭ ␳ oil␾ oilϩ c mic . were determined at 25. by applying the effective medium theory. C. Pro Analysi grade. A. Water used in this study was of MilliQ purity. Volumetric measurements The density of AOT reverse micelles has been measured at the molar ratio of water-to-surfactant W 0 ϭ 11. We can now plot the measured density ␳ of the solution. ͑1͒. in bound or sequestered states. The SAXS data. ͑7͒ If the oil density ␳ oil is assumed independent of the micellar volume fraction ␾ mic .

close to the densities of the pure solvents. It is obvious that the density of the surfactant film decreases. Plot of the measured solution density ␳. ͑8͒ and ͑11͒ leads to the density of AOT monomolecular film itself.4 41. as determined in literature ͓28–30͔.2 22. ͑10͒ and ͑6͒. leading to Eq. one can write the mass conservation law for all micelles present.4 33. one can write c micϭ ␳ mic␾ micϭ c AOTϩ c H2O with c AOT ϭ m AOT . Experimental measurements of ␳ mic can be carried out at various ␾ mic values.684 A. Note that ␾ AOT and ␾ mic can be obtained from Eqs. 1. 3.9 12.6 11. Micellar water pool Now.27͔. 2. Aqueous core radius R w . ␾ AOT␾ mic ͑12͒ FIG. for example as a function of micellar size.e. W 0 value of 11: it is obvious that the increase in the solution density is linear with the increase of the micellar volume fraction. at W 0 ϭ 11. This result implies that ␳ mic is independent of ␾ mic as long as ␾ mic remains below 0. Note that in isooctane R w is continuously smaller than in decane. The error on ␳ mic is estimated at 5 ϫ 10Ϫ 3 . which have been carried out using two different solvents.5 32. Table I summarizes the results of SAXS measurements of R w as a function W 0 . 12. ͑10͒ Various experimental methods have been used to determine the variation of the micellar water pool radius R w with the water content. ͑3͒. AMARARENE et al.7 50. 1.. the two curves have a tendency to converge. The precision of density measurements is 10Ϫ 5 and that of volume fractions is 10Ϫ 3 . directly related to water concentration ( W 0 ). vs the micellar volume fraction ␾ mic . the curves extrapolate to the pure solvent values. 2.5 38.2 FIG. obtained by SAXS measurements as a function W 0 in isooctane and decane. We can therefore express ␳ AOT as a function of R w . ␳ micϭ ␳ AOT␾ AOTϩ ␳ H2O␾ H2O . The two curves display distinctive slopes. ͑7͒ one can obtain ␳ mic for different values of W 0 .1 55.6 Decane 8. In all of the figures the solvents are represented by the following symbols: ͑᭹͒ for isooctane and ͑᭡͒ for decane. ␳ AOT is close to the value .7 14. if we consider the densities and the volume fractions of water and of the surfactant AOT.26. From Eq. R w ϩ l c represents a value close to the micellar hydrodynamic radius R H . PRE 61 TABLE I. ͑8͒ ͑9͒ ␾ AOTϩ ␾ H2Oϭ 1.9 21. Such a plot is represented in Fig. down to a W 0 value between 10. ͑12͒. directly proportional to the water content W 0 .2 30.8 19.2 42.9 50. They are in good agreement with literature ͓7. R w ͑Å͒ W0 3 5 7 9 11 15 17 19 22 25 27 30 Isooctane 8.5 46. The error corresponds to Ϯ1 Å. Assuming identical spherical micelles leads to ␾ AOTϭ Rw V AOT ϭ1Ϫ V mic R wϩ 1 c ͩ ͪ 3 .3 15.0 27. i. ␳ AOT ϭ c AOT . For ␾ micϭ 0. the two curves extrapolate to the values of ‘‘dry’’ AOT aggregates in each solvent for W 0 ϭ 0. ␾ AOT and ␾ H2O being respectively the volume fractions of AOT and water within micelles. where l c represents the average length of the surfactant tail solvated by oil taken as 11 Å. At high water concentration as ␳ mic decreases. Plot of the density reverse micelles ␳ mic vs water content W 0 . and then reaches a different plateau value distinct for each solvent. V ͑11͒ Rearranging Eqs. Note that for W 0 ϭ 0.6 29.1 ͑as it is always the case throughout this work͒.9 36. as illustrated in Fig.8 45. and extrapolate at infinite dilution. The sum of the radius of the water pool and the length of the surfactant tail.8 17. If we introduce now the concentration of AOT ( c AOT) and that of water ( c H2O) in Eq.

n AOT W0 3 5 7 9 11 15 17 19 22 25 27 30 Isooctane 43 64 88 114 143 212 253 298 376 468 538 658 Decane 51 77 104 132 163 228 264 302 363 429 476 552 ␴ ͑Å2͒ Isooctane 21 28 33 37 40 44 46 47 48 49 50 50 Decane 24 29 33 37 40 44 46 47 49 50 50 51 FIG. most strongly bonded water molecules ͓12͔. i.. n H2O 3 W 0 n AOT ͑15͒ From the previous results we can deduce the variation of the surfactant polar head group area with W 0 . M AOT N ͑13͒ where M AOT is the molecular mass of AOT ͑444. 3.PRE 61 ADIABATIC COMPRESSIBILITY OF AOT . Since the monolayer surfactant tails are solvated by two oils of different chemical nature. the aggregation number ( n AOT). Note that according to Collins ͓33͔. Such effect has been related to different degrees of penetration of the surfactant chain region by the solvents. n AOT . and N av the Avogadro number. is represented in Table II for isooctane and decane. The total volume of the water mol4 3 ␲Rw . Variation of the aggregation number n AOT and of the area per surfactant polar headgroup ␴ as a function W 0 . knowing the number of AOT molecules per micelle ( n AOT) the number of water molecules per micelle. affecting the bending elasticity constant ͓31͔. 3 water molecules hydrating the immediate surfactant sulfonate groups (SO3Ϫ). 4. calculated for each value of W 0 . 2. but as well on the characteristics of encased water itself. the agreement with values extracted from the literature is good ͓26. 685 TABLE II. are also considered in the strongest interaction with the anion. For the highest W 0 measured. those first 2. The area per surfactant polar headgroup ␴ is then given by We have plotted in Fig. . will be W 0 n AOT . The difference between isooctane and decane levels off after W 0 ϭ 12. Note that the minimum volume is obtained for W 0 ϭ 3. as a function of W 0 . Average volume of a water molecule At this point. The variation of the aggregation number. These curves clearly indicate the influence of the solvent chemical structure. but constantly larger in decane. If the volume 4 3 ␲RH . Variation of the micellar AOT density ␳ AOT as a function of W 0 . 4 the volume variation of a water molecule sequestered within an AOT reverse micelle in isooctane and decane. . Area per surfactant polar head group v wϭ 3 Rw Vw 4 ϭ ␲ . possibly due to a better fit between the branched solvent chains and surfactant tails ͓11͔. FIG. although the minimum is somewhat lower and the curve does not reach the ‘‘bulk’’ water volume. ␴͑ W0͒ϭ4␲ 2 Rw n AOT . it reaches in isooctane a value of about 28 Å3. 3. One can thus estimate the ecules in a micelle is v w ϭ 3 average volume v w occupied by a micellar water molecule. the number of surfactant molecules per micelle is defined as n AOT ϭ m AOT N av .e. these values reflect the isooctane or decane/AOT interactions. not only on micellar physical properties. . Plot of the volume of a micellar water molecule as a function of W 0 . and the total number of a spherical micelle is v micϭ 3 of micelles in solution is N ϭ V mic / v mic .27. ͑14͒ The values of ␴ as a function of increasing water content W 0 . are given in Table II.4 g moleϪ1͒. a value corresponding to the three nonfreezing ͓13͔. We observe in isooctane that the minimum molecular volume ͑21 Å3͒ is obtained for W 0 ϭ 3. In decane the overall curve profile is similar to that in isooctane. as a function of W 0 . close to that used by most investigators for ‘‘bulk water’’ ͑30 Å3͒. of the ‘‘dry’’ surfactant aggregate in each solvent ͓18͔. The values calculated for both solvents are of the same order of magnitude. n H2O .32͔.

which where a and b are obtained by the linear fit of ␤ mic( ␴ ). ␾ AOTϷ 1 and Eq. Eq. for W 0 ϭ 11. allow the determination of adiabatic compressibility ␤ of the solution using Laplace’s equation. From Eq. Figure 7 also shows that the slope b is significantly higher in isooctane than in decane. ␤. Hence. R w ). ␤ ϭ ␤ mic␾ micϩ ␤ oil␾ oil . ͑18͒ In Fig. assuming that ␤ AOT varies according to Eq. ␤ H2O are the respective compressibilities of AOT and micellar water. underlining once more the difference between AOT-isooctane and decane interactions.15). Thus. Plot of solution compressibility ␤ vs the micellar volume fraction ␾ mic . at a given W 0 ͑or micellar water radius. This result has been confirmed for different values of W 0 ͑not shown͒. ͑16͒ In this work we make use of the effective medium theory ͓34͔. B. and increases then proportionally to W 0 . we can also conclude that ␤ mic is independent of the micellar volume fraction. Each of the constituent phases is described by parameters corresponding to the pure phase. the oil phase and the inclusion phase. one can assume as a first approximation. we obtain thus average values for ␤ H2O(0) ϭ (60Ϯ 4) ϫ 10Ϫ 11 PaϪ1 in isooctane and . u. displays again distinctive features for isooctane and decane. where water is tightly bound to the AOT sulfonate head groups. AMARARENE et al.686 A. the curves extrapolate to compressibility of the pure solvents. evidencing once more the importance of the three first water molecules. Micellar compressibility FIG. For the former. The precision of compressibility values is estimated to 1%. it is obvious that the linear relation is no more valid for ␤ mic( ␴ ) and the contribution of ␤ H2O has to be taken into account. The results are represented in Fig. 5. applied at the level of the micelle itself. and then a very slow increase. We take one phase to be the continuum fluid. The error is Ϯ 1 ϫ 10Ϫ 11 PaϪ1. Plot of micellar compressibility ␤ mic . since the relevant acoustic wavelength is always orders of magnitude larger than the size of micelles. ͑19͒ shows that micellar compressibility is very close to that of the monomolecular film of AOT. ͑19͒ where ␾ AOT is defined by Eq. to account for the behavior of reverse micelles. ␤ mic displays lower values under W 0 ϭ 3. ͑17͒ becomes ͑17͒ ␤ mic͑ ␴ ͒ ϭ a ϩ b ␴ ͑ W 0 ͒ Ϸ ␤ AOT͑ ␴ ͒ . Experimental determination of the density of the solution and measurements of sound velocity. ͑19͒ then becomes ␤ mic͑ W 0 ͒ ϭ ␤ AOT͑ ␴ ͒ ␾ AOTϩ ␤ H2O͑ W 0 ͒ ␾ H2O . 6. ͑14͒. so that: ␤ micϷ ␤ AOT . 6. We can thus plot ␤ mic versus the surfactant polar head group area ␴. The variation of ␤ mic as a function of W 0 can be interpreted by making use of the effective medium theory. ͑10͒ and ␤ AOT . as a function of ␾ mic . between 10 and 30. ␤ micϭ ␤ AOT␾ AOTϩ ␤ H2O␾ H2O . we have plotted the measured solution compressibility. 8: for the lowest W 0 studied. Modeling water compressibility ␤ϭ 1 ␳u 2 . It is intriguing that the two curves intersect at W 0 ϭ 3. Eq. If we assume that ␤ oil is independent from ␾ mic . PRE 61 FIG. The isooctane and decane curves cross around W 0 ϭ 3. Thus. we can relate the solution compressibility ␤ to both micellar ( ␤ mic) and solvent ( ␤ oil) compressibilities ͓22–24͔. ͑21͒ The above equation allows the calculation of ␤ H2O as a function of the water content W 0 . it is obvious that the relationship is linear. For the latter solvent. We describe each of the constituent components. as given in Eq. For W 0 Ͻ 10. ͑20͒ ␤ ϭ ␤ oilϩ ␾ mic͑ ␤ micϪ ␤ oil͒ . the value of ␤ remaining always higher in isooctane than in decane. ␾ mic . with ␾ oilϩ ␾ micϭ 1. C. as a function of the relative volumes of the constituent phases. while the second phase consists of micellar inclusions randomly embedded in the continuous phase. corresponds to the upper limit of ‘‘bound’’ water. ͑20͒. one can observe a sharp increase of the curve until W 0 ϭ 10. 5. very tightly bound to the surfactant polar head groups. For ␾ micϭ 0. Such a plot shown in Fig. For W 0 Ͼ 10 ( ␾ H2OϾ 0. At a W 0 ϭ 11. in terms of several parameters. ␤ mic remains constant. as a function of W 0 . The compressibility decreases linearly with ␾ mic . ͑18͒ one can calculate ␤ mic as a function of W 0 . Figures 7͑a͒ and 7͑b͒ illustrates such a plot: there is a linear relation between compressibility and ␴ ( W 0 ).

. CONCLUSION ␤ H2O͑ W 0 ͒ ϭ ␤ F ␾ F ϩ ␤ B ␾ B . not obtained in this work due to phase instability. Straight lines correspond to bound water. Our model has its own limitations. In particular the variation of the hydrodynamic radius R H ͑or the radius of the water core R w ) with the aggregation number n AOT . In the present work. where water is often designated as ‘‘free. Such effects have been related to different solvent penetration of the surfactant chain region. ͑22͒. be it bound or sequestered. of protein-water interface. one obtains. 60ϫ 10Ϫ 11 and 54ϫ 10Ϫ 11 PaϪ1 in isooctane and decane respectively. A second external water-shell of thickness d. in particular in the first hydration shell of proteins ͓37͔. this might indicate that the model is no longer valid. and water have led to an abundant literature. Within a reverse micelle. we can conjecture the existence of a ␤ F maximum with a subsequent decrease for much higher values of W 0 .’’ but yet remains sequestered in a restricted volume. discussed in the preceding section. using Eq. The relation is linear to about W 0 ϭ 10. These numbers cannot be obtained experimentally. in agreement with observations by x-ray and neutron scattering experiments. At high W 0 . we did not take into account the apparent polydispersity of solutions since it is at the low- . we can calculate ␾ F . affecting the bending elasticity constant. we present an approach that makes possible the interpretation of micellar adiabatic compressibility by the use of a simple model. In any case. The first significant finding of this work is that in AOT reverse micelles. such as aggregation number. water is present as a mixture of ‘‘bound’’ and ‘‘free’’ form. its compressibility is ␤ B . at intermediate W 0 values.e. compressibility reaches 80ϫ 10Ϫ 11 and Ϫ 11 Ϫ1 Pa for the same solvents. oil. as well as on that of the encased micellar water ͑volume and compressibility͒. ͑b͒ in decane.3͔ that all water present is tightly bound to the surfactant monolayer. we have favored the simplest approach. the compressibility of the central water core at the highest water content measured ( W 0 ϭ 30): these numbers are ␤ F ϭ 85ϫ 10Ϫ 11 and 67 ϫ 10Ϫ 11 PaϪ1 for the two solvents. water can be represented by a first central aqueous sphere of adiabatic compressibility ␤ F . polar headgroup area. we have used high precision density measurements to obtain a direct estimation of the micellar volume fraction. IV. Since the relation of ␴ with W 0 also depends on the nature of the solvent. Using ultrasound velocity measurements. is in tight interaction with the surfactant anionic headgroups and is designated as ‘‘bound’’. In this work. to tend toward the compressibility of ‘‘bulk’’ water. Variation of micellar compressibility ␤ mic with the surfactant polar headgroup area ␴. ͑22͒ FIG.3. In addition. Assuming the bound water thickness d ϭ 4 Å ͓26͔. but since we do not observe this behavior in Fig. The dashed line represents the compressibility of ‘‘bulk’’ water. consequently ␤ B Ϸ ␤ H2O .PRE 61 ADIABATIC COMPRESSIBILITY OF AOT . 8. . but can be extracted from our model. we rule out the existence of a compressibility maximum. (54Ϯ 4) ϫ 10Ϫ 11 PaϪ1 in decane. Plot of the micellar water compressibility ␤ H2O vs W 0 in isooctane and decane.. 26 range͒. ͑see Table I͒. and on the first approximation ␾ B Ϸ 1. if at a high water content. i. have been documented in AOT reverse micelles ͓2. note that the compressibilities of pure solvents are 121ϫ 10Ϫ 11 and 90ϫ 10Ϫ 11 PaϪ1 for isooctane and decane. 687 FIG. the nature of apolar solvent exerts a distinct action on the properties of the surfactant. It could be intuitively expected for these values. For example. Our results point also to the higher density of bound water. there are different formulations describing ultrasonic propagation in complex fluids. Taking for ␤ B the values found above. respectively. By applying again the effective medium theory to the water pool of the reverse micelle.36͔. As a matter of comparizon. respectively. the corresponding W 0 values are indicated on the top of each curve: ͑a͒ in isooctane. one can write Simple geometrical considerations lead to ␾ F ϭ ͓ 1 Ϫ ( d / R w ) ͔ 3 and ␾ B ϭ 1 Ϫ ␾ F . and the surface of the polar headgroup with the molar water-to-surfactant ratio W 0 . 7. For the highest W 0 measured ͑in the 23. For W 0 Ͻ 5 there is a general agreement ͓2. In addition. 67ϫ 10 We can also model the variation of the aqueous compressibility ␤ H2O as a function of W 0 . the observed variation can be attributed to water molecules in rapid exchange between the two aqueous compartments ͓35͔. Efforts to understand the physical properties of molecular assemblies constituted of surfactant. In addition. 8. the above mentioned compressibilities are substantially higher than that of ‘‘bulk’’ water (45ϫ 10Ϫ 11 PaϪ1).

Robinson. Rev. 76. Compressibility thus correlates with other physical properties of confined water. H. P. 155. P. Engin. PCV 98N21/0125͒. Biochem. Nicot. Europhys. J. ͓4͔ C. Chem. J. Yarwood. Binks. Le Hue bach. 12 708 ͑1994͒. Acad. sequestered water contributes substantially to micellar compressibility. Chatenay. ͓9͔ H. Breslauer. J. We believe that similar events may occur in the first water solvation shell of proteins. Phys. Oldfield. S. M. Ober for his help in SAXS measurements. Howe. A. 453 ͑1986͒. Ober. Phys. and P. R. ͓1͔ T. H. Belton. A Textbook of Sounds ͑Bell and Sons. ͑iii͒ The model allows one to predict the variation of the average volume of a single micellar water molecule solvating the surfactant polar head group. or multiple scattering theory ͓39͔. Acoust. Zulauf and H. Small Angle Scattering of X-Rays ͑John Wiley and Sons. ͓5͔ M. Chieux. ͓13͔ H. C. Schurtenberger. Luisi. C. Phys. ͓25͔ W. and M. ͓7͔ M. C. Chem. Rev. the multiple scattering can be neglected for volume fractions lower than 0. Klyachko. Gallay. our model allows one to discriminate the contribution of waters tightly interacting with the polar headgroups. Such clarification should ultimately lead to a better understanding of the physical properties of biological macromolecules entrapped in the micellar aqueous core ͓43͔. H. 53 ͑1991͒. A 44. Funct. Chem. Dore.-Y. Christ and P. ͓31͔ B. M.. Biochem. Malliaris. S. 8249 ͑1991͒. 1946 ͑1988͒. ͓30͔ S. by hydrating AOT polar head groups. 1 83. Phys.-D. 6961 ͑1993͒. J. Ping Sheng. Schulman. Huang. ͓14͔ L. Colloid Interface Sci. 1941͒. ͓23͔ R. C. Berezin. Howe. A. If these few water molecules do not contribute substantially to the overall micellar compressibility. 1012 ͑1996͒. Am. An alternative treatment would be to use the variational principle ͓40͔. Annu. 25. P. 92. Sci. B 101. Lang. ´ rou. New York. 985 ͑1987͒. 16. Indeed. ͓17͔ T. Merdas. Ament. structured by anomalous hydrogen bonding and ion solvation. Urbach. 480 ͑1979͒. from that due to ‘‘free’’ water. J. Y. Levashov. Hills. Genet. B. J. M. 7. J. P. Biochemistry 24. B. W. Maitra. with the drag of one phase with another. R. 5122 ͑1984͒. Haering. Proc. Pande. and M. J. Gindre. Eicke. Waks. V. Urbach. Jada. 20. L. 102 ͑1943͒. T. ACKNOWLEDGMENTS This research was supported by Centre National de la Recherche Scientifique ͑Grant No. J. Urick. 20. Chem. PRE 61 est level at W 0 values. 321 ͑1991͒. M. and M. J. 151 ͑1982͒. Chem. Nicot. 93. N. Biotechnol. Vacher. ͓11͔ P. and I. Phys. Biophys. J.688 A. Ur͓18͔ A. J. 83. W. A. 1048 ͑1999͒. Kellay. and J. 1955͒. L. Waks. Nicot. Espinat. ͑ii͒ At higher water concentration ( W 0 Ͼ 10). further investigation will be required to fully understand the intriguing properties of this complex system. Nicot. Luisi and L. J. 13 ͑1984͒. Chem. Chem. Giasson. Vincent. F. Phys. R. Wood. Natl. 98. 153. where most of our experiments have been carried out. 255 ͑1994͒. J. In summary. ͓24͔ A. Chem. Sarvazyan. J. Pessah. which accumulates within the center of the micelle as W 0 increases. ͓2͔ P. J. Proteins: Struct. ͓6͔ K. A. . V. and P. Phys. D.-F. Martinek. and to discriminate the properties of micellar water. Technol. A. Phys. 638 ͑1985͒. Ye. Revs. L. 97. corresponding to ‘‘bound’’ water. 231 ͑1986͒. 7024 ͑1985͒. and A. 2198 ͑1987͒. Soc. Biophys. 152. ͓16͔ A. 1. Faraday Trans.15 and for particle radii lower than 8 nm at frequencies of several MHz or less. J. Soc. Robinson. Toprakcioglu. Amararene. We show that the latter is about twice as compressible as ‘‘bulk’’ water. 68. they act as surfactant plasticizers and lubricants ͓41͔. W. J. 355 ͑1992͒. AMARARENE et al.. J. Fletcher. Waks. Palermo. Phys. Chalikian and K. Phys. 7869 ͑1989͒. Rev.-P. Lett. 409 ͑1986͒. M. ͓8͔ H. M. Phys. P. Christopher. J. and M. which is expected for water in a confined geometry ͓42͔. ͓22͔ A. ͓26͔ A. J. occur at temperatures and W 0 values higher than that reported in this work. 465 ͑1992͒. Another approach would consist in using hydrodynamic equation balancing momentum and continuity of the phases. Morizur. M. Chem. 12. for complex fluids with acoustic impedance of inclusions close to that of the solvent. and M. M.-F. Fournet. ͓3͔ C. Waks. J. Mackie. Waks. ͓21͔ A. Gindre. Vacher. it has been shown that. J. The microemulsion clustering and the deviations of spherical shape observed by ͓38͔. London. Colloid Interface Sci. ͓15͔ A. J. Appl. Chem. ͓12͔ D. Wille and S. 15 180 ͑1996͒. Ultrasonics 20.-H. the adiabatic compressibility of the micellar solution is proportional to the AOT monolayer polar head group area. Khmelnitski. J. and B. Magid. Sarvazyan. The au´ on for the precision of his thors are grateful to Morvan Le work. P. as a function of the water content of the system. 440 ͑1979͒. J. Nevertheless. Faraday Trans. J. Hoar and J. V.-C. Eur. Biophys. Hauser. We expect that our results will help to elucidate new properties of these organized molecular assemblies. 10 751 ͑1997͒. Pileni. Chem. ͓10͔ B. J. Dispersion Sci. Robins. the effective medium approach is the most convenient ͓39͔. Genet.. M. Guinier and G. Chem. 91. Ober. ͓28͔ M. USA 93. Eicke.. We find the minimum molecular volume for W 0 Ϸ 3 in both isooctane and decane. 88. Colloid Interface Sci. They wish to thank R. it has enabled us to shed light on several interesting and unsolved problems: ͑i͒ For the lowest amounts of micellar water encapsulated ( W 0 Ͻ 10). 4 ͑1986͒. S. 100. Crit. ͓19͔ S. I. and M. ͓27͔ J. Nevertheless. 1 80. 983 ͑1947͒. Weitz. J. For upper values of W 0 . 18. and B. G. corresponding to three nonfreezing water molecules. in direct contact with the above water. In any case our model is operational since we have used it to extract a number of micellar parameters in good agreement with literature. ͓20͔ A. Nature ͑London͒ 152. Prusiner. A. ͓29͔ D. Meunier. Soc.

Svergun. Helv.-F. M. Willard. 1576 ͑1988͒. Minton. Biophys. L. Acad. Top. Biophys. J. Kuprin. Lions ͑Springer-Verlag. 72. D. Le Hue Waks. 2267 ͑1998͒. 689 ͓32͔ H. D. Temple. R. Phys. P. ´ rou. I. Sayers. Chem. A104 ͑1999͒. Barron. A. 13 144 ͑1997͒. Sager. Sci. and N. D 21. Haracker and J. Biochemistry 36. Weaver. 1311 ͑1995͒. M. Chem. D. C. Levinger. and W. Smeets. J. Biophys. Sheng. Kinderleher. Homogenization and Effective Modulii of Materials and Media. 105 ͑1985͒. ͓35͔ R. M. G. H. ͓41͔ L. W. ͓42͔ A. F. S. J. 102. ͓43͔ A. J. L. New York. C. Curr. ͓37͔ D. 65 ͑1997͒. Phys. D. .-F. Eicke and J. ͓34͔ P. . Kohn. and G. Chim Acta 59.PRE 61 ADIABATIC COMPRESSIBILITY OF AOT . 2883 ͑1976͒. Zaccai. J. 99. J. H. Eriksen. Wilson. Chem. E. Wave Motion 7. ͓39͔ A. J. edited by J. and G. and J. 68. Nicot. . Proc. Rehak. Phys. 13 291 ͑1995͒. ͓38͔ G. 87. Z. Urbach. Natl. and D. L. Collins. 85 ͑1980͒. USA 95. ͓33͔ K. M. ͓40͔ R. Eicke. Gindre. Riter. E. ͓36͔ H. Koper. S. J. Amararene. Richard. J. 76. L.-Y. J. 2705 ͑1998͒. Bedeaux. 1983͒. M. Koch. Hecht.