J . Chem. Soc., Faraday Trans. 1, 1986, 82, 77-88 Volumetric Investigation of the Hydrotropic Action of Aqueous Sodium Propionate Solutions on Pentanol at 298 K Olav Hans& and Jar1 B. Rosenholm* Department of Physical Chernistry,oAbo Akademi, Porthansgatan 3-5, SF-20500 Abo, Finland The miscibility range for water - sodium propionate - pentanol mixtures has been determined at 298 K. The phase equilibria are compared with ‘normal’ and ‘ detergentless ’ microemulsion systems. Sodium propionate imposes a very slight salting out action (if any) on pentanol in dilute aqueous solutions. However, pentanol gains full miscibility with concentrated salt solutions. Volumetry was chosen to detect the interactions between the components since it has been found to monitor both changes in the environment of the hydrocarbon chains as well as large shifts in the polar association equilibria.According to the apparent molar volumes the enhanced solubility of pentanol is apparently obtained through dissociation of the alcohol com- plexes. Interestingly the concentrated hydrotrope solvent seems to be apolar in character. The state of the hydrotrope is almost entirely determined by the relative amount of (hydration) water. The interaction between pentanol and sodium propionate also seems to be of importance in dilute pentanol solutions. The absence of extended association structures in the system explains the difference between hydrotropy and normal solubilization, which is also evidenced by the volumetric behaviour of pentanol. On the other hand, the restriction of the mutual miscibility of the hydrotrope and the alcohol to the region where all the water is considered to be bound as hydration water of the ions marks the distinction between these systems and ‘ detergentless microemulsions ’.The modern and now generally accepted view is that microemulsions may be defined as isotropic and thermodynamically stable solutions of small polymolecular aggregates (micelles) dispersed in a pure or mixed solvent. The number of components in the system is normally three or four including a surfactant, a cosurfactant and one or two solvent components, i.e. water and oi1.l In some cases the cosurfactant is considered to be the solvent or a part of it. Owing to their dualistic nature, the surfactants associate in aqueous media in such a way that the hydrocarbon chains collect in the core of the oil/water (o/w) micelles.29 In this local apolar environment a small amount of oil can be dissolved (solubilized) into the aqueous solution.Similarly small amounts of water can be brought into oil solvents. Here the surfactants are reversing their molecular arrangement to form small polar droplets of solubilized water (w/o) mi~elles.~ Much of the effort to elucidate the changes in stability of the association structures has concentrated on the role played by the cosurfactant. This approach seems natural, since it is known from the phase diagrams published by Ekwal15 that a shortening of the chain length of e.g. an alcohol cosurfactant destabilizes the high-order phase structures (liquid-crystalline phases) to give stable solutions over wide composition ranges.The starting point for such studies has been the use of a real surfactant which is characterized by a cooperative association in aqueous solutions.6 The influence of ascending a homologous series of alcohols on the mutual miscibility of the components is illustrated in fig. 1. Less attention has been paid to the fact that the main changes introduced by the variation of the chain length of the cosurfactant are also produced if the chain of the 7778 Action of Aqueous Sodium Propionate on Pentanol W Fig. 1. The miscibility region (in black) of (i) a water-hexanol mixture with an alkyltrimethyl- ammonium bromide of varying chain length at 301 K [ref. (7)] and (ii) a water-sodium octanoate mixture with an alkanol of varying chain length at 293 K [ref.( 5 ) ] , respectively. The figure is drawn as an aid for the eye and thus the concentrations (in wt :<) and the liquid-crystalline phases are omitted from the figures. surfactant is varied, keeping the cosurfactant the same.’? As shown in fig. 1 there is a systematic change in miscibility when the chain length of one of the surfactants is varied. Despite the relationship in phase behaviour, the ability of ionic short-chain surfactants to enhance the solubility of water-insoluble components may be considered as different from normal solubilization. In contrast to surfactants, the association tendency of hydrotropes is weak.9 They are thus incapable of normal solubilization. To mark this distinction the effect has been denoted hydrotropy by Neuberg.l0 The special behaviour of short-chain surfactant solutions has attracted much discussion in early investigations on association phenomenal1* l2 and serves as a bridge between microemulsion systems and normal so1utions.13 A new situation is also introduced if the cosurfactant, e.g.the alcohol, is mutually soluble with water. In favourable cases very polar alcohols can, mixed with water, but without any ‘real’ surfactant, dissolve large amounts of oil. This kind of ‘detergentless microemulsion’ is characterized, as are the hydrotrope systems, by the absence of any well defined aggregate structure^.^^^ l5 The present paper constitutes one contribution in a series of investigations to characterize thermodynamically the properties of simple and complex surfactant systems.16-20 In the present case both of the surface-active components have only a limited miscibility with water.The interactions found should then be typical for ionic hydrotrope systems. The ‘surfactant’ is unable to associate to micellesg and the alcohol is only sparingly mutually soluble with water. Of the parameters available, volumetric properties seem to be particularly suitable since electrostatic and polar interactions contribute only secondarily to the total changes in the apparent molar volumes of the surfactant component^,^^-^^ making this parameter0. Hans& and J. B. Rosenholm 79 especially useful in rough model calculations of surfactant systems.24 The present mixed solvent system is, of course, not suitable for the type of closed structure modelling indicated above.Experiment a1 Chemicals Both the sodium 1 -propionate (SIGMA Chem. Co., > 99 % ) and the pentan-1 -01 (Merck AG, zur analyse, > 99 % ) were used without further purification. The water was distilled and passed through an ion-exchange resin. Its conductivity was 0.5 $3 cm-l. Instruments The densities were measured with an Anton-Paar vibrating-tube densitometer equipped with an external measuring cell (DMA 601). The calibrations were carried out against air and water. The apparatus has a sensitivity of ca. 1 x lop6 g ~ m - ~ . In all the experiments the temperature was maintained at 298.15 K. Phase Equilibria The extension of the solution phase of the water - sodium propionate - pentanol mixture was determined by visual observations at 298 K.Before inspection all the solutions were allowed to equilibriate in a thermostatted water bath for 7 days. The estimated accuracy is ca. 1 % . Mathematical Operations The densities (p) obtained were converted into molar volumes of solution by employing the following expression : where M(i) denotes the molar mass and x(i) the mole fraction of component i: n(i) refers to the amount (number of moles) of substance i ( n = 1-3). volume according to the equation [" = ' (pure) or O0 (infinite dilution)] The molar volume of solution was then recalculated in terms of the molar excess V: = vrn -C x(i) V" (i) = C x(i) VE(i). ( 3 ) The reference values used were V'(H20) = 18.068 cm3 mol-1 V'(C,OH) = 108.65 cm3 molt1.These were derived by recalculating the densities of the pure liquids or by extrapolating the apparent molar volumes of the salt to infinite The values found are close to those reported, e.g. in ref. (1 6), (26) and (27) (alcohol) and (25) (salt). The relationship between the molar excess volume and the apparent molar volume is given by the general formula Vm(NaC302) = 53.6 cm3 mol-180 Action of Aqueous Sodium Propionate on Pentanol Fig. 2. The thermodynamically stable liquid phase of the system water-sodium propionate- pentanol at 298 K. The concentrations are given in wt % . The solid lines drawn within the liquid phase refer to the solution series measured: binary system H,O-C,OH (1 and A), n(C,OH)/n(NaC,O,) = 2.0 (2) and = 1.0 (3), binary system H,O-NaC,O, (4), n(H,O)/n(NaC,O,) = 6.5 (B), and a constant percentage by weight for water of 30 (C), respectively.where V*(i) denotes the volume of the pure component i. In the case of the salt the volume of sodium propionate infinitely diluted in water ( Va) was used instead. Note that the apparent molar properties of complex multicomponent systems may in some instances be distorted owing to compensation effects.20 However, the relatively scarce number of experimental points prevented us from partially deriving the molar excess volume of Results and Discussion Three-component Phase Diagrams As is commonly done for three-component systems, the compositions are expressed in a triangular diagram ; each component increases in concentration counterclockwise, thus ending up with the pure component at eachcorner.Fig. 2 illustrates the thermodynamically stable isotropic solution phase (intersected by solid lines) of the water - sodium propionate - pentanol system. The right-hand corner of the triangle is characterized by a very large solubility gap between the liquid phase and hydrated salt. The right-hand boundary is defined by the minimum amount of water needed to hydrate the sodium carboxylate group, being determined by the hydration requirement of the alkali-metal The minimum amount of water needed to hydrate one mole of sodium carboxylate groups is 5-7 mol. The left-hand side of the triangle gives the mutual solubilities of water and pentanol. The logarithm of the solubility of n-alcohols in water from butanol to octanol has been found to be a linear function of the number of carbon atoms in the chain.’ The solubility of these alcohols has successfully been correlated with the surface contact between their hydrocarbon residues and water.30 For pentanol the solubility was found to be 2.2-2.3 % by weight.It is important to distinguish the process when water is dissolved in alcohols from the reverse case. Thus the logarithmic dependence indicated above is not regenerated. Recently it was concluded that the origin of the water solubility may be found in the polarity of the alcohol solvent. For hydrophobic long-chain alcohols all the water is consumed in complex formation while an increasing part of the water is distributed freely in polar short-chain alcohol^.^^^ 31 If the molar ratio of alcohol to water is used to express0.Hansen and J . B. Rosenholm 81 the saturation point then the distribution of water in the short-chain alcohols decreases the ratio from the constant value of ca. 2.7 prevailing for the longer homologue~.~~ The molar ratio of pentanol to water is reported to be 1.7-1.8 (ca. 10% by The ternary left-hand boundary departs from the binary saturated solutions. In the ternary solutions all the polar head groups must be fully hydrated before the distribution of water into the solvent is activated.20 If the maximum amount of water needed to hydrate the ionic groups fully [n(H,O)/n(NaC,O,) = 11-14] is assumed to be fully consumed by the sodium propionate, then the vertical left-hand phase border may be taken as determined by hydration water and water complexated with pentanol molecules.The latter fraction of water is successively diminished as the concentration of sodium propionate increases. It was concluded above that one should distinguish the present hydrotrope from micelle-forming surfactants. As indicated in fig. 1, the left-hand phase boundary never extends to high water fractions if the hydrophobicity of the ionic surfactant is insufficient for cooperative association. This is also true even if the chain of the cosurfactant, e.g. the alcohol, is lengthened.12 The mutual solubility of the components reaches its maximum when both of the surface-active components are chemically compatible (about the same chain length, branching etc.). In ' real ' ionic surfactant systems the water-rich part remains thermodynamically stable until a full miscibility between the cosurfactant and water is obtained (fig.1). Clearly it is the capability of the surfactant to associate which secures phase stability at high water contents. The largest continuous solution phase is obtained when the extensive association structures are destabilized with a short-chain cosurfactant (e.g. with water-soluble alcohols; fig. 1). With chemically compatible cosurfactants the long-chain surfactants form stable aggregates. Then the aqueous phase does not extend to cosurfactant contents higher than ca. 10% by weight. The critical micellization concentration is visible as a break point in the phase boundary from the binary water-surfactant axis to solutions richer in additive.The strong-associate stability of ionic surfactants is also manifested by the lamellar liquid-crystalline phase protruding between the two liquid phases to very high water contents5* (the liquid-crystal phases are omitted from fig. 1). The tie lines in the two-liquid phase region (left-hand side of the continuous solution phase) gives the distribution of the components between the liquids in equilibrium. As shown by Lawrence, the hydrotrope is distributed for short-chain homologues in favour of the water-rich r e g i ~ n . ~ For matching surfactants the distribution is almost equal between the alcohol and the aqueous phase. (The tie lines run almost parallel to the left-hand side of the triangle.) In the present case the wide and narrow water-rich phase indicates a distribution of NaC,02 in favour of the aqueous solution.However, the increased ionic strength does not enforce any marked salting-out of pentanol from the aqueous phase. 33 The Binary Systems Water-Pentanol and Water-Sodium Propionate In the dilute aqueous solutions short-chain alcohols show complex behaviour appearing as a minimum in the apparent and partial molar volumes. The minimum is shifted towards lower concentrations when the chain length is extended from methanol to propan01.~~ Hvidt et al. explained the minimum on the basis of a 'primitive equilibrium mixture model '.35 The volumetric parameters found characterize the highly cooperative association of water around the non-polar alkyl groups (hydrophobic solvation), giving a positive contribution, and the strong increase in the extent of solvation of the hydroxy groups, giving rise to a negative contribution. The small change found for &(C50H) up to the phase-separation limit (fig. 3) indicates that an efficient compensation between the two abovementioned effects occur.Since the slope is negative, then according to the model of Hvidt et al. hydrophobic82 Action of Aqueous Sodium Propionate on Pentanol 0.00 0.02 0.04 0.06 0.08 - ' t 18.0 -- I 0.70 0.80 0.90 1.00 1001 t 1 , . I X(C5 OH) 0.70 0.80 0.90 1.00 Fig. 3. Apparent molar volumes of both components of the binary system water-pentanol at 298.15 K. The upper scale refers to dilute aqueous solutions of pentanol [squares are experimental results, dots are from ref. (37)] and the lower to water diluted in pentanol (diamonds).solvation predominates in the dilute aqueous region. Another view is that the hydrocarbon residue induces a water structurization which is further enforced when the hydration shells initially overlap.36 Jolicoeur et al. investigated the volumetric and heat-capacity behaviour of a large number of alcohols in water.37 They found that the concentration dependence of the apparent molar volume was linearly and negatively dependent on the limiting volume of the alcohols in water and concluded that this was related to the hydrophobicity of the alcohol. Our concentration dependence and limiting value of 102.82 cm3 mol-l and B, = - 5.07 cm3 mo1-2 kg, respectively, are in close agreement with their values of &? = 102.88 cm3 mol-l and B, = - 4.55 cm3 molt2 kg (fig.3). Dilute solutions of water in alcohols have a special significance when one is trying to interpret the intermolecular hydrogen-bond equilibria, owing to the comparatively large solubility of water in all liquid alcohol^.^! 3 2 9 33 To our knowledge there was no systematic study focusing on the thermodynamic state of water diluted in alcohols prior to the paper of Jolicoeur et all6 They showed that the limiting volume of water in pentanol at 288.15, 298.15 and 308.15 K was 16.60, 16.86 and 17.25 cm3 mol-l, respectively. The values0. Hanskn and J . B. Rosenholm 83 found are close to the limiting volumes of 16.53, 16.89 and 17.33 cm3 mol-l reported by Sakurai and N a k a g a ~ a . ~ ~ Our limiting value of water, 16.93 cm3 mol-l, is close to the values reported for 298.15 K.The low molar volume of water should be connected with strong intermolecular hydrogen bonding.16 Fig. 3 illustrates how the initial limiting state is changed upon addition of water. As is seen, the slope of the apparent molar volume of water is positive, while a slight decrease is found for pentanol. One may interpret this observation as a partial liberation of water from the previously denser state to be dispersed freely in the solvent. In accordance with the negative slope of &(C205) part of the water remains tightly bound to the intermolecular complexes (J. B. Rosenholm, unpublished results). The point of saturation of water in pentanol is not reached in fig. 3. Its logarithmic temperature dependence is zero33 which, assuming ideal behaviour, implies that the phase separation is characterized by zero enthalpy or that both the enthalpy and entropy of the phase transition are zero.Similarly as for the alcohols, the thermodynamic behaviour of a number of straight- chain, branched, cyclic and aromatic surfactants having an equal number of carbon atoms have been rationalized using the surface area of the hydrocarbon residue.39 Fig. 4 illustrates the initial concentration dependence of aqueous sodium propionate solutions. The initial slope is only slightly positive ( B , = 0.9 cm3 mo1-2 kg) when corrected for ionic interactions, as is seen from the extended Debye-Huckel limiting slope (insert).25 This behaviour is typical for most surfactants in the range where no soiute-solute interaction is expected to occur.4o Since propionate has been found to be the last representative of the homologous series of sodium alkyl carboxylates not showing extended as~ociation,~~ 417 42 the interaction should be predominantly ionic in all solutions.The hydrocarbon residue is, however, still large enough to initiate typical hydrophobic effects on water s t r ~ c t u r e . ~ ~ - ~ ~ If the model of Hvidt is applied to this binary system also, the increase in the apparent molar volume of sodium propionate may be interpreted as a polar solvation effect. This is also evident from the positive change in the apparent molar volume of water as the hydrotrope concentration is increased (fig. 5). The Ternary System Water-Sodium Propionate-Pentanol The changes recorded accompanying the dilution of solutions of different molar ratios of pentanol to sodium propionate with water are shown in fig.5. The binary systems discussed above represent both extremes of an infinite and a zero molar ratio. The lines were chosen to record the changes occurring when the pentanol continuous phase is converted into an hydrotrope solution. Starting from the binary system of sodium propionate in water (fig. 5, line 4) it may be concluded that the overall change found (ca. 8 cm" mol-l) is close to values found for surfactants upon micellization.48 Although it seems unjustified to assume any extended association for propionate, the large positive change in apparent molar volume indicates significant changes in the environment of the hydrotrope. The full miscibility with pentanol indicates that most of the water may be assumed to be bound to the ions.The predominance of ionic hydration is supported by the nearly identical dependence of the apparent molar volumes on x(H,O) found for water and the salt (independent if the molar ratio of pentanol to sodium propionate is increased; lines 2-4 in fig. 5). The apparent molar volume of water seems to increase to very high values, which may be taken as evidence for a bulky ice-like hydration structure. When investigating the apparent molar volume of pentanol one may conclude that larger changes are produced when changing the molar ratio of pentanol to hydrotrope than upon dilution with water. In comparison with the binary system given in fig. 3 we note that the limiting volume of pentanol in water is 102.82 cm3 mol-l, i.e.some 6 cm3 mol-i lower than in the pure state. For the solutions discussed (lines 2 and 3, fig. 5) we observe a twofold but positive change.84 Action of Aqueous Sodium Propionate on Pentanol 64'62 .O 1 a a a a 0.00 0.04 0.08 0.12 0.16 X(NaC, 0,) Fig. 4. Apparent molar volume of aqueous sodium propionate solutions at 298.15 K in the most dilute region [asterisks and circles from ref. (25), triangles from ref. (16)]. The extended Debye-Hiickel limiting slope is given above: 4: = 53.4 cm3 mol-l, B, = 0.9 cm3 mol-* kg [data from ref. (25)]. In order to analyse further the state of the sparingly soluble pentanol we measured the change in volume as x(C,OH) was varied, keeping the molar ratio of water to sodium propionate constant (fig.6). Initially the apparent molar volume of pentanol remains constant upon dilution with the aqueous hydrotrope solution. When a high concentration of aqueous sodium propionate solution is approached there is, however, a dramatic increase ir. the volume of pentanol. Only by considering that the change upon micellization of surfactants is at the most of the order of 12-1 3 cm3 mol-1 can the increase be appreciated. There is, on the other hand, a very small variation in the apparent molar volumes of water and propionate over the same concentration range. This indicates that they are only marginally affected by mixing with pentanol. The small pocket observable in the dependence of the volumes of water and salt on the pentanol content is probably due to inaccuracy in the measurements.0.Hans& and J . B. Rosenholm 85 1 0 0 o! i i / I I , , , I I 0 0 0 0 0 0 t' 0 (D 0 0 9 7 0 286 Action of Aqueous Sodium Propionate on Pentanol N 0, 0 0, I0. Hans& and J . B. Rosenholm 87 The interaction between pentanol and sodium propionate was checked by keeping the weight fraction of water constant (line C, fig. 6). Although the mole fraction of water varies we note that the lines corresponding to the apparent molar volumes of both water and sodium propionate are almost vertical, illustrating the prime importance of water for hydration of the salt. Interestingly, the volumetric state of pentanol is almost equally dependent on the pentanol mole fraction, as it was when the molar ratio of water to hydrotrope was kept constant (line B).The key to the effect thus probably lies in the dissociation of the pentanol complexes when diluted in the hydrotrope solutions. The change in volume resembles the situation found when diluting alcohols in h y d r o c a r b o n ~ , ~ ~ - ~ ~ - 49 but is much larger than the change recorded when pentanol is diluted in water (fig. 3). Furthermore, the solubilization of alcohols in true surfactant solutions produces very small changes in the apparent molar volume from the molar volume of the pure The overall behaviour may thus be interpreted by the pentanol being dissolved in an apolar (aggregate) solvent. According to the results of Treszczanowicz and Benson, the change in volume increases when the size of the solvent molecules (aggregates) increases,22! 23 but the volume may also increase owing to steric mismatching of the components.21$ 49 The large magnitude of the change in &(C,OH) and the slight variation in &(NaC,O,) suggest, however, that the specific interaction between these two components is also of considerable importance for their volumetric state.Conclusions The solution phase of the water - sodium propionate - pentanol system extends from aqueous to pentanolic solutions. The propionate is unable to associate into micelles, and pentanol is only sparingly mutually soluble with water. Since pentanol is, on the other hand, fully miscible with concentrated aqueous proprionate solutions the behaviour may be considered typical of ionic hydrotrope systems.Volumetry was chosen as a probe of interactions between the components since it sensitively monitors both large shifts in the association equilibria of complexing agents and changes in the environment of the hydrocarbon chains. The volumetric state of water and salt was predominantly determined by their mutual interaction. Pentanol, on the other hand, experiences no change in its volumetric state over most of the phase area, provided that the hydration of the propionate is constant. Below x(C,OH) x 0.2 a large, positive change is observed in the apparent molar volume of pentanol. This effect may be related to dissociation of the pentanol complexes when diluted in the apolar hydrotrope solution. The relative changes in the apparent molar volumes of pentanol and sodium propionate suggest, however, that specific interactions between these two components contribute significantly to their volumetric states.The results presented suggest that, although normal solubilization is related to the hydrotropic action, there are well founded reasons to distinguish between these two effects. The restriction of the ternary solution phase to the range where all the water may be considered bound to the ionic hydrotrope, independent of the nature of the semi-polar solubilizate, marks the difference from detergentless microemulsion systems. J. B. R. thanks the Finnish Research Council for Natural Sciences for financial support. References 1 I. Danielsson and B. Lindman, Colloids Surt, 1981, 3, 391. 2 B. Lindman and H. Wennerstrom, Top.Curr. Chem., 1980, 87, 1 3 H. Wennerstrom and B. Lindman, Phys. Rep., 1979, 52, 1. 4 H-F. Eicke, Top. Curr. Chem., 1980, 87, 84.88 Action of Aqueous Sodium Propionate on PentanoL 5 P. Ekwall, Advances in Liquid Crystals, ed. G. H. Brown (Academic Press, New York, 1975), vol. 1, 6 P. Mukherjee, J. Pharm. Sci., 1974, 63, 972. 7 A. S. C. Lawrence, Mol. Cryst. Liq. Cryst., 1969, 7, 1. 8 J. B. Rosenholm, Kemia-Kemi, 1981, 10, 661. 9 J. B. Rosenholm, P. Stenius and I. Danielsson, J. Colloid Interface Sci., 1976. 57, 551. p. 1. 10 C. Neuberg, Biochem. Z . , 1916,76, 107. 11 P. A. Winsor, Trans. Faraday Soc., 1948, 44, 376. 12 E. C. Lumb, Trans. Faraday Soc., 1951,47, 1049. 13 C-U. Herrmann, G. Klar and M. Kahlweit, J. Colloid Interface Sci., 1981, 82, 6.14 N. Ito, T. Fujiyama and Y. Udagava, Bull. Chem. Soc. Jpn, 1983, 56, 379. 15 N. F. Borys, S. L. Holt and R. E. Barden, J . Colloid Interfizce Sci., 1979, 71, 526. 16 Y. De Grandpre, J. B. Rosenholm, L. L. Lemelin and C. Jolicoeur, Solution Behavior of Surfactants, 17 J. B. Rosenholm and C. Jolicoeur, to be published. 18 P. Saris, J. B. Rosenholm, E. Sjoblom and U. Henriksson, J . Phvs. Chem., in press. 19 J. B. Rosenholm, R. B. Grigg and L. G. Hepler, J. Chem. Thermodyn., in press. 20 J. B. Rosenholm, Surface Chemistry, ed. K. S. Birdi (Technical University of Lyngby, Denmark, 1981), 21 A. J. Treszczanowicz and G. C. Benson, J. Chem. Thermodyn., 1978, 10, 967. 22 A. J. Treszczanowicz and G. C . Benson, J. Chem. Thermodyn., 1980, 12, 173. 23 A. J. Treszczanowicz, 0.Kiyohara and G. C. Benson, J . Chem. Thermodyn., 1981, 13, 253. 24 J. Biais, P. Bothorel, B. Clin and P. Lalanne, J . Colloid Interface Sci.. 1981, 80, 136. 25 J. B. Rosenholm and L. G. Hepler, Thermochim. Acta, 1985, 81, 381. 26 R. C. Wilhoit and B. J. Zwolinski, J. Phys. Chem. Ref. Data, 1973, 2, suppl. no. 1. 27 E. Vikingstad, J . Colloid Interface Sci., 1979, 72, 75. 28 K. Blomqvist, Zs. Bedo and J. B. Rosenholm, Colloid Polym. Sci., 1980, 258. 398. 29 P. Ekwall and L. Mandell, Acta Chem. Scand., 1968, 22, 699. 30 G. L. Amidon, S. H. Yalkowsky and S. Leung, J . Pharm. Sci., 1974, 63, 1858. 31 P. Stilbs, K. Rapacki and B. Lindman, J . Colloid Interface Sci., 1983, 95, 583. 32 S. Backlund, H. Heriland and 1. Wikholm, J. Solution Chem., 1984, 13, 749. 33 R. Stephenson, J. Stuart and M. Tabak, J. Chem. Eng. Dutu, 1984. 29, 287. 34 G. C. Benson and 0. Kiyohara, J. Solution Chem., 1980,9, 791. 35 C. Dethlefsen, P. G. Sorensen and A. Hvidt, J. Solution Chem., 1984, 13, 191. 36 F. Franks and D. S. Reid, in Water: A Comprehensive Treatise, ed. F. Franks (Plenum Press, New York, 37 C. Jolicoeur and G. Lacroix, Can. J . Chem., 1976. 54, 624. 38 M. Sakurai and T. Nakagava, J . Chem. Thermodyn., 1984, 16, 171. 39 C. Jolicoeur, J. Paquette, Y. Lavinge and R. Zana, Solution Behavior of Surfuctants, ed. K. L. Mittal and E. J. Fendler (Plenum Press, New York, 1982), vol. 1, p. 389. 40 P. A. Leduc and J. E. Desnoyers, Can. J. Chem., 1973, 51, 2993. 41 S. Lindenbaum, J. Chem. Thermodyn., 1971, 3, 625. 42 F. Eriksson, Thesis (Abo Akademi, SF-20500 Abo 50, Finland, 1982). 43 H. Snell and J. Greyson, J . Phys. Chem., 1970, 74, 2148. 44 J. Davies, S. Osmondroyd and M. C. R. Symons, J. Chem. Soc., Faraday Trans. 2, 1972, 4, 686. 45 0. D. Bonner, R. K. Arisman and C. F. Jumper, Z . Phys. Chem. (Leipzig). 1977, 258, 49. 46 J. Paquette and C. Jolicoeur, J. Solution Chem., 1977, 6. 1403. 47 C. Jolicoeur, J. Paquette and M. Lucas, J. Phys. Chem., 1978, 82, 1051. 48 J. B. Rosenholm, Colloid Polym. Sci., 1981, 259, 11 16. 49 K. N. Marsh and C. Burlitt, J . Chem. Thermodyn., 1975, 7, 955. ed. K. L. Mittal and E. J. Fendler (Plenum Press, New York, 1982), vol. 1, p. 431. p. 279. 1975), vol. 2, p. 339. Puper 5/370; Received 4th Murch, 1985