J. CHEM. SOC. FARADAY TRANS., 1994, 90(18), 2703-2708 Phase Transitions in the Bilayers of Vesicles formed from Binary Mixtures of Symmetric Di+Alkylphosphates in Aqueous Solutions Michael J. Blandamer," Barbara Briggs and Paul M. Cullis Department of Chemistry, The University, Leicester, UK LEI 7RH Jan B. F. N. Engberts, Anno Wagenaar and Elly Smits Department of Organic & Molecular Inorganic Chemistry, University of Groningen, Nijenborgh 4, 9747 AG Groningen, The Netherlands Dick Hoekstra Department of Physiological Chemistry, University of Groningen , Bloemsingel 10,9712 KZ Groningen, The Netherlands Anna Kacperska Department of Physical Chemistry, University of Lodz, Pomorska 18,91416 Lodz, Poland Vesicles in aqueous solutions were prepared from binary equimolar mixtures of di-n-alkyl-phosphates (sodium and potassium), (R'0)2P02-M+ and (R20)2P02-M+.When the number of carbon atoms in R' and R2 differs by two and when R' or R2 = C12H25, C,,H2, , C1&& and C,,H,, the membranes undergo well defined gel to liquid crystal transitions at characteristic temperatures T, .The recorded T,s are intermediate between the melting temperatures for vesicles prepared from the respective single di-n-alkylphosphates. Furthermore, the extrema recorded by differential scanning microcalorimetry show that the vesicle membrane is made up of domains that differ in composition. For those vesicles produced from di-n-alkylphosphates where the number of carbon atoms in R' and R2 differs by more than two the plots recorded by the scanning microcalorimeter are complex.The scans show many extrema, suggesting that the bilayers are formed from many domains having different compositions. In all cases, the scan patterns are essentially repeated through several heat-cool- heat.. .cycles. The temperatures T, are increased relative to those of the component surfactants when K+ and Na+ salts are mixed, showing that the counter cations play an important role in determining the thermotropic properties of the vesicles reflecting the importance of electrical interactions in determining the packing within the bi layers. In aqueous solution vesicles' formed by sodium di-n-dodecylphosphate (DDP) undergo a gel to liquid crystal tran- sition near 34.8 "C. This transition produces a sharp well defined feature in the scan recorded by a differential scanning microcalorimeter2 (DS).Reproducible and, for the most part, reversible scans were obtained2 when DDP vesicles were pre- pared using a carefully defined protocol. The recorded extre- mum in the scan is characterised by a standard enthalpy of melting A,H" equal to 3.9 kcal (mol monomer)-' and a patch number, n, of 168. The latter describes the number of surfactant monomers which form domains in which coopera- tive melting takes place. In other words, the minimum fluc- tuation in the bilayers necessary to trigger the main phase transition involves 168 DDP monomers. With increase in alkyl chain length, T, increases, reaching 77.1 "C for the case where R' = R2 = C,,H3,. In the study reported here we measured the DSC scans for vesicles prepared using equi- molar mixtures of two symmetric di-n-alkylphosphates where the alkyl groups differ in chain length; e.g.mixtures of (R'O),PO,-Na+(aq) and (R20),P02-Na+, where R' and R2 are C,2H25, Cl4H29, C16H33 and C18H3,. Our interest in these mixed-vesicle systems was prompted by the recogni- tion that although synthetic amphiphiles4 such as DDP form bilayer structures similar to lipids,' naturally occurring lipid systems usually comprise mixtures of amphiphiles.6 We show that for a mixed-vesicle system formed from two di-n-alkyl- phosphates where R' and R2 differ by C2H4, the scan shows a single extremum at a temperature intermediate between the T,S characteristic of vesicles prepared from equimolar solu- tions containing single di-n-alkylphosphates. Where R' and R2 differ by more than C2H4, the scans are complicated showing several extrema, although the patterns are repro- duced through several heat-cool-heat...cycles. Experimental Materials The di-n-alkylphosphates were prepared as described pre- viously. Differential Scanning Microcalorimetry A MicroCal (USA) differential scanning microcalorimeter was used as previously described.2 The reference cell was filled with water. Scans were normally recorded between 15 and 90°C. After the first scan had been recorded, the solution was allowed to cool slowly to 15°C and the scan again recorded from 15 to 90 "C. This protocol was repeated several times, the aim being to determine if the same scan pattern was recorded over several heat-cool cycles.The scan data were recorded on 3.5 in7 discs and later analysed using the Origin (MicroCal) software. Preparation of Vesicles Previously we ~howed~?~ the importance of developing a pro- tocol for vesicle preparation which leads to DSC scans which are reversible and repeatable. The excellent sensitivity of the scanning microcalorimeter in these studies2 confirmed that this same consideration applied to solutions prepared using binary mixtures of surfactants. For the purpose of compari-son, we report here the properties of vesicle solutions con- taining fixed total concentration of surfactant, 8.42 x mol dm-3.The appropriate masses of the two surfactants t 1 in x 2.54 cm. were weighed to produce a suspension (volume, 2.2 cm3) con- taining (R10)2P02-M,+ (aq; ca. 4.21 x mol dm-3) and (R20)2P02-M1+ (as; 4.21 x mol dm-3). The aqueous suspension was heated to a temperature T* (see below) and held at that temperature for 1 h. The resulting solution was cooled and carefully placed in the sample cell of the calorimeter, ensuring that no air bubbles were trapped in the cell. The solution was cooled in the sample cell to 15 "C, which was the starting temperature for the scans. In detail, the temperature T* was as follows; for (i) R' = C12,H25, R2 = C14H29, M1+ = M2+ = Na', (ii) R' = R2 = C12H25, M,+ = Na+, M2+ = K+, and (iii) R' = R2 = C14H29, M1+ = Na+, M2+ = K+, T* x 60°C; for (iv) R' = C12H25, R2 = C16H33, M1+ = M2+ = Na+, (v) R' = C14H29, R2 = C16H33, M1+ = M,' = Na', (vi) R' = R2 = C16H33, M1+ =Na+, M2+ =K+, T*=70"C; for (vii) R' = C12H25, R2 = C1BH37, M1+ = M2+ = Na', (viii) R' = C14H29, R2 = C18H37, M1+ = M2+ = Na+, (ix) R' = C16H33, R2 = C18H37, M1+ = M2+ = Na+ and (x) R' = C16H33, R2 = trans-CH3(CH2),CH=CH(CH,), , M1+ = M2+ = Na+, T* x 75°C.Results The scans recorded for aqueous solutions containing vesicles formed from mixtures of di-n-alkylphosphates, (R'0)2P02-Ml + and (R20)2P02 -M, were strikingly dif- + ferent, on the one hand for vesicles where R' and R2 differ by two CH, groups and, on the other hand, where R' and R2 differ by more than two CH, groups.Where the difference was two CH, groups, the scans showed extrema intermediate between those characteristic of solutions containing the single di-n-alkylphosphate at the same overall concentration; Fig. 1A. This pattern is illus-trated by vesicles containing sodium di-n-alkylphosphates, where R' = C12H25 and R2 = C14H29. Solutions containing a single di-n-alkylphosphate have extrema at 34.8 & 0.2 "C (R = C12H25) and 52.2 f0.1 "C (R = C14H29) where the quoted means of estimates for T, (Table 1) are calculated using at least four successive scans on the same solution; Fig. 1A. The equimolar mixture produced a broad extremum at 40.8 & 0.1 "C; Fig. 1A. The same broad pattern was produced by the mixture in five repeat scans taken over a period of 17 h, showing that the processes involved are reversible equi- librium properties of the mixed system having short time constants; Fig.2. With increasing alkyl chain length but holding the differ- ence between R' and R2 at two methylene groups, the tem- perature T, of the pure solutions moves to higher temperatures; Fig. 1B and Table 1. Furthermore, the tem- perature difference between T,s decrease; e.g. for R = C16H33,T, = 66.3 f0.1 "c.The scans for equimolar mix- tures, R, = C14H29 and R2 = C16H33, show an extremum between the two T,s at 57.6 f0.1 "C; Fig. 1B. An important finding was that with increase in chain length the scan recorded for the mixture broadens and then separates into two components. The latter pattern is particularly clear-cut for the mixture where R' = C16H33 and R2 = C18H37;Fig.1C. For solutions containing the single surfactant R = C18H37, T, = 77.1 f0.1 "C. For these mixtures, the pattern shown by the scans was broadly similar when recorded several times over a period of 6 h. In the case of the mixture Fig. 1 Dependences on temperature of the differential heat capa- cities (reference = water) for aqueous solutions containing vesicles prepared from the sodium salts of di-n-alkylphosphates, (RO) PO,-Na+ [total concentration = 8.4 x (mol monomer) dm-33 and from equimolar [ca. 4.2 x lo-' (mol monomer) dm-3] of two sodium di-n-alkylphosphates, (R'O),PO, -Na+ and J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 0.06LA I I I I 20 40 60 TPC 1 I I 1 40 60 TrC 0.06 0.04 0.02 I I I I 60 80 T/OC (R20),P0,-Na+.A, Scan (a) for R = C,,H,, where T, = 343°C; scan (b) for R' = C,,H,, and R2 = C14H29 where T, = 40.8"C; scan (c) for R = C,,H,, where T, = 52.2"C. B, Scan (a) for R = C,,H,, where T, = 52.2"C; scan (b) for R = C16H33 where T, = 66.3"C; scan (c) for R' = C14H,, and R2 = C16H33 where T, = 57.6"C. C, Scan (a)for R = C,,H,, where T, = 66.3 "C; scan (b) for R = C,,H,, where T, = 77.1 "C; scan (c) for R' = C,,H,, and R2 = C18H37where T, = 67.6 and 757°C. J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 I* 1 lB 40 60 60 80 TpC Fig. 2 Dependences on temperature of the differential heat capa- cities for equimolar mixtures [total concentration 8.4 x low3(mol monomer) dm -3] of two sodium di-n-alkylphosphates, (R'O),PO,-Na+ and (R20),P02-Na+.A, R' = C12H25 and R2 = C14H29 where the recorded scans (a)-(d) were recorded consecutively and scan (e) was recorded after the solution had stood at 15°C for 8 h; B, R' = C16Hj3and R2 = C18H3,where the scans were recorded as in A except that scan (e) was recorded after the solution had stood at 15 "Cfor 6 h. where R' = C16H3, and R2 = C18H37, the scans were char- acterised by two well resolved extrema at 67.2 and 77.4"C. After each repeat scan the extremum originally at 77.4"C moved to lower temperatures. After 6 h the extremum at the higher temperature had moved to 75.7"C, the extremum at the lower temperatures being slightly shifted to 67.6 "C; Fig. 2.The patterns emerging from the scans are much more com- plicated when the number of CH, groups in the alkyl chains differs by more than two. A typical example is shown in Fig. 3A, which records the DSC scans for vesicles prepared using, in one case (Cl,H2,0),P02-Na+ and, in the other case, (C16H3J0)2P02-Na+.These scans are compared with that recorded for vesicles in an aqueous solution containing equi- molar amounts of the two di-n-alkylphosphates. The contrast is striking in that the new scan shows a broad almost feature- less plot. Interestingly, the same types of pattern were recorded in successive scans for solutions containing + +(C ,H2 ,O),PO, -Na and (C ,H 70)2PO, -Na [Fig. 3(b)], showing that within a time period of ca.8 h the changes contributing to the differential heat capacities are reversible. Nevertheless, over a longer time period, changes had taken place producing new extrema near 40°C. For mixtures con- taining (C14H,50)2P02-Na+ and (C18H370)2P02-Na+ a broad extremum with partially resolved maxima at 55.7 and 67.4"C was recorded; Table 1. Repeat scans of the equimolar mixtures (cf. Fig. 3B) containing (Cl,H2,0)2P0,-Na+ and (C16H330)2P02-Na+ showed a similar pattern with one slight difference. In the first scan, an extremum was observed near 38.5 "C,but this feature disappeared over seven suc- cessive scans, reappearing on an eighth scan after the solution had stood for 3 h at 15°C. Otherwise, all remaining features in the complicated pattern were repeated over eight recorded scans.In the examples quoted above, the counter cation was fixed at Na+, whereas the length of the alkyl chains was varied. In the next series of experiments the alkyl chain was fixed but the mixtures were prepared using equimolar amounts of sodium and potassium salts. The T,s of the mixtures were generally higher than for those solutions containing either sodium or potassium salts. For the (Cl,H,,0),P02 -system, T,(K+) = 52.0 & 0.1 "C, T,(Na+) = 52.2 & 0.1 "C but T, for the Na+ + K+ mixture = 53.2 & 0.1 "C; Fig. 4A. A similar pattern was recorded for the solution containing (C16H,,0),P02-where the recorded T,s are 66.3 & 0.1 (Na+), 65.2 & 0.1 (K') and 66.5 & 0.1 (K+ + Na+)"C; Fig.4B. The processes responsible for the extrema are reversible, as confirmed by the scans in Fig. 4C. Table 1 Derived parameters characterising the gel to liquid crystal transitions for di-n-alkylphosphate vesicles formed from (R'O),PO,-M + and (R20)PO -M + AmH"/kcal patch number (mol monomer) -n 34.8 f0.2 3.9 168 52.2 f0.1 5.6 140 66.3 f0.1 7.9 91 77.1 f0.1 9.1 45 Q 40.8 f0.1 3.7 158 f1 57.6 f0.1 5.1 137 k 7 67.6 0.1; 75.7 f0.1 8.4b 149 f15 C d e e 33.3 f0.1 4.4 153 52.0 f0.1 6.3 90 65.2 f0.1 8.3 84 35.0 f0.1 3.3 435 13 53.2 f0.1 5.6 308 f9 66.5 f0.1 7.8 277 & 7 Four important maxima in the DSC scan at 25.2 f 0.1, 42.3, 47.4 0.1, 50.9 f0.1.Two transitions. ' Many weak extrema observed of which those at 32.8, 54.0 and 57.8 "C are the most striking in the DSC scan. Three important extrema in the DSC scans at 33.6 f0.1, 63.1 0.1 and 66.9 f0.1 "C. Enthalpy of transition refers to a mean molar mass for the mixtures of dialkylphosphates. J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 0.06 c I -0.04 m0 \ $ 0.02 OL I I 20 40 TPC 60 80 I I I I-0.04' 20 40 60 80 TfT Fig. 3 Dependences on temperature of the differential heat capa- cities (reference = water) for aqueous solutions containing vesicles prepared from the sodium salts of di-n-alkylphosphates, (R0)2P02-Na+ [total concentration = 8.4 x (mol monomer) dm-3] and from equimolar [ca.4.2 x (mol monomer) dm-3] mixture of two sodium di-n-alkylphosphates, (R'O),PO,-Na+ and (R20),P02-Na+. A, R = C12H,, where T, = 343°C; (b) mixture where R' = C1,H25 and R2' = C16H33; (c) R = CI6H33 where T, = 663°C. B, Differential scans of the mixture, R' = C12H2, and R2 = CI8H3, recorded successively [scans (a)-(d)]; scans recorded after standing at 15"C for 6 h [scan (e)] and for 12 h [scan (f)]. Fig. 4 Dependences on temperature of the differential heat capa- cities (reference = water) for aqueous solutions containing vesicles prepared from sodium and potassium di-n-alkylphosphates, [total concentration = 8.4 x (mol monomer) dm-3] and from equi- molar [ca. 4.2 x (mol monomer) dm-'] mixtures of the sodium and potassium salts. A, (a) (C,4H,,0)2P0,-K+ [aq; 8.4 x lop3 (mol monomer) dm-3]; (b) (C,,H,,O),PO,-Na+ [aq; 8.4 x (mol monomer) dm-3] and (c) equimolar mixture where the concen- tration of both salts is ca.4.2 x lo-' (mol monomer) dm-j. B, (a) (C16H330)2PO;K+ [aq; 8.4 x (mol monomer) dm-3]; (b) (Cl,H330)2P02-Na+ [as; 8.4 x (mol monomer) dm-3] and (c) equimolar mixture where the concentration of both salts is ca. 4.2 x (mol monomer) dm-3. C, Repeat scans for the mixture described in B where successive scans were recorded in curves (a)-(d) and where scan (e) was recorded after standing at 15 "C for 5 h in the sample cell. 50 60 TPC B 0.06 0.04 0.02 0 C 0.04 0.02 0 -0.02 u70 6o TPC J. CHEM. SOC. FARADAY TRANS., 1994, VOL.90 A slightly more subtle experiment involved recording the scans for aqueous mixtures of the sodium salts of (Cl,H370)2P02- and [trans-CH,(CH,),CH= CH(CH2),0],P02 -;Fig. 5. As previously rep~rted,~ the pattern produced by the unsatu- rated di-n-alkylphosphate [aq; 8.4 x (mol monomer) dm-3] showed several extrema, those at (i) 25.2 f0.1, (ii) 42.3 & 0.1, (iii) 47.4f0.1 and (iv) 50.9& 0.1 "Cbeing particu- larly intense. Vesicles formed from (C,8H370)2P02-Na + [as; 8.4 x lo-, (mol monomer) dm-3] produce a single sharp extremum near 77.1 & 0.1 "C; Table 1. Interestingly, five features are apparent in the scan for the mixture with an almost one-to-one matching with the features recorded for the separate solutions. In common with the systems described above, five successive scans showed broadly similar patterns indicating that many of the processes responsible for the extrema in the differential scans are reversible.Analysis The dependences of differential heat capacity 6Cp on tem- perature were analysed in three stage^.^ First, a previously recorded water-water baseline was subtracted from the recorded scan. Secondly, a chemical baseline was subtracted from the derived scan. In contrast to the analysis of scans for micellar systems7 this subtraction was straightforward. The traces showed no overall changes in heat capacities over a range of temperatures below and above the extrema centred on T,. In the final stage of the analysis the dependence of calculated molar isobaric heat capacity C,,on temperature was fitted to eqn. (1) using three variables; (i) the patch number (cf.concentration of patches), (ii) the temperature T, and (iii) calorimetric enthalpy of reaction, the last being brought into agreement with the calculated van't Hoff en- thalpy of reaction, Av,Ho using the patch number, where K is the equilibrium constant at temperature T. In some cases the shape of the recorded envelope did not conform to the simple bell-shape required by eqn. (1).In the cases discussed below the envelope could be fitted assuming that one or more independent processes of the form X(aq)e Y(aq) contributed to the recorded scan. In these cases the 0.04 .-I Y-m0 \ (;ip 0.02 TrC Fig. 5 Dependences on temperature of the differential heat capa- cities (reference = water) for aqueous solutions containing vesicles prepared from sodium salts of saturated and unsaturated C,,-di-n- alkylphosphates; (a)[trans-CH,(CH,), = (CH,),O],PO,-Na+ [aq; 8.4 x lo-' (mol monomer) drn-,]; (b) (C,,H,,O),PO,-Na+ [aq; 8.4 x lo-, (mol monomer) dm-'1; (c) an equimolar mixture of these di-n-alkylphosphates [aq; 4.2 x lo-' (mol monomer) dm-'1.shape of the recorded envelope could be accounted for in terms of the sum of the separate independent transitions. Discussion In principle, the recorded extrema associated with the gel to liquid crystal transition yield three pieces of information, patch number n, the melting temperature T, and the enthalpy changes, van't Hoff and calorimetric.The latter is given in the first instance by the area under the scan envelope. When these parameters are calculated from the scans recorded for aqueous solution containing (i) (RlO),PO2-M+, (ii) -(R20),P02-M+ and (iii) (R10)2P02M i-+ (R20),PO2-M+, interesting patterns emerge particularly when R' and R2 differ by two CH, groups; Table 1. The recorded T, for mixtures where R' = C12H2, and R2 = C14H29, 40.8"C, is intermediate between the T,s for solutions containing single surfactants, 34.8 (C12H24 and 52.2 (C14H29)oC.The patch number n, 158 & 1 is also inter- mediate between the estimates of n for the pure system, 168 and 140,respectively. However, the enthalpy change, 3.7 kcal (mol monomer)-' is lower than the enthalpy changes re- corded for the two pure surfactants, 3.9 and 5.6 kcal (mol monomer)-', respectively.For both simple and mixed systems a higher patch number is linked with a more co- operative melting process. Further analysis of the scans for the mixed systems shows that the envelope is satisfactorily accounted for in terms of three independent two-state trans- itions; Fig. 6.We assign the lowest calculated T, at 39.4"C to a melting of domains which are richer in the surfactant with the shorter hydrocarbon tail, (C,,H,,O),PO,-Na+. Simi-larly, we assign the extremum at the highest calculated tem- perature T,, 41.9"C, to the melting of domains richer in the surfactant, (Cl,H290)2P02-Na+. In other words, in a single vesicle we envisage three different types of patches which undergo gel to liquid crystal transitions at slightly different temperatures.A similar pattern emerges for the (C14H290)2P02 -Na+ and (C,,H330)2PO2-Na+ surfactants, the T, at 57.6"C for the mixture being intermediate between that for the single surfactants for the mixture, whereas the patch number 120 000 80 000 r I Y-m--. 2 40 000 I0 30 40 50 Fig. 6 Dependences on temperature of the molar heat capacity of an aqueous solution containing vesicles formed from equimolar 14.2 x lo-, (mol monomer) dm-3] of (C,,H,,O),PO,-Na+ and (C,,H2,0),P02-Na+; (a)full line is obtained from the scan in Fig. 1A; (b)dotted lines show calculated contributions (Origin software, MicroCal Ltd.) from three independent two-state transitions where T, = 39.4 & 0.1, 39.6 & 0.01 and 41.9 k0.1 "C.2708 137 & 7 is close to that for the C,, surfactant, 140 rather than that for the c16 surfactant, 91. The overall enthalpy change is lower at 5.1 kcal (mol monomer)-' relative to 5.6 and 7.9 kcal (mol monomer)- ', respectively for the pure sur- factant solutions. However, the shape of the scan envelope is accounted for in terms of two independent transitions. Two resolved extrema are, in fact, recorded for aqueous solutions containing an equimolar mixture of (C ,H,,O),PO, -Na + and (C,,H,,O),PO,-Na' surfactants; Fig. 1C. Thus with increasing chain length, the T,s of solutions containing single surfactants move to higher temperature^'.^ but their differ- ence decreases when the difference in chain length is C,H,. At the same time, the scan pattern for equimolar mixtures evolves into two extrema; Fig.1C and 2. The scan envelope for this mixture is satisfactorily accounted for in terms of three independent transitions of the form, X(aq) eY(aq) each transition having the same patch number, 149 & 15 which is larger than the patch number of the solutions containing single surfactants, 91 and 45, respectively; Fig. 7. The calori- metric (integrated) enthalpy change over the whole envelope is 8.4 kcal (mol monomer)- which is between that calculated for the two pure systems, 7.9 and 9.1 kcal (mol monomer)-', respectively. The patterns can be accounted for in terms of vesicles in which there are three types of patches which are in turn rich in C,,-surfactant, rich in C,,-surfactant and con- taining both c16- and C,,-surfactants.It is noteworthy that all patch numbers for the mixed system are ca. 150 irrespective of chain length of the sur- factant chains, again with the proviso that they differ by only two CH, units. The similarity in C,,,, at the maximum also suggests that the range of compositions in the mixed patches is small, a conclusion again supported by the similarity in the patch numbers for the mixed systems. 200 000 c I Y-0" 100 000--. 0" Q 0 T/"C Fig. 7 Dependences on temperature of the molar heat capacity of an aqueous solution containing vesicles formed from equimolar C4.2 x (mol monomer) dm-'] of (C,,H,,O),PO,-Naf and (C,BH,,0)2P02-Na+; (a) recorded scan; (b) dotted lines show cal- culated contributions (Origin software, MicroCal Ltd.) from three independent two-state transitions where T, = 67.7 f 0.1, 74.9 f 0.3 and 76.7 & 0.2"C.J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 The complicated envelopes produced by mixtures of sur- factants where the number of alkyl groups differed by more than two CH, units meant that the shapes could only be reproduced using equations based on eqn. (1) but involving many such contributions. Indeed, when we had recorded the first scan the complexity prompted us to conclude that each subsequent scan would have slightly different shapes on the grounds that having melted the chains would, on recooling, adopt different conformations.We attributed the complex- ities in the scan to the problem of repacking chains of differ-ent lengths in the vesicles. This conclusion is supported by the complexity in the scans recorded for mixtures of saturated and unsaturated surfactants. There the scan pattern for the mixture is close to that predicted by combining the scans recorded for the solutions containing the single surfactants ; Fig. 5. Nevertheless, the reproducibility of many key features in successive scans shows that the domains have intrinsically well defined structure which maintain their integrity over many heat-cool cycles. For the scans produced by mixing K+ and Na' salts of the same di-n-alkylphosphates, the increase in T,, on going to the mixed system is indicative of an increase in chain packing of the gel phase although the enthalpy change associated with the melting decreases; 3.3 kcal (mol monomer)-' for the equimolar mixture of (C12H250)2P02-Na+ and (Cl,H,,O),PO,-Kf compared to 3.9 and 4.4 kcal (mol monomer)-', respectively, for the solutions containing the two pure surfactants.Nevertheless, this change is a striking demonstration of the importance of the counter cations in determining the properties of vesicular bilayers in aqueous solutions. The patch number for the mixture, 438 f13, is much larger than the patch number for the separate solu- tions. Therefore, the melting is much more cooperative in the mixed vesicles, suggesting that the mixed vesicles are smaller and that the accompanying change in entropy is important in determining T,, . We thank the SERC for their support through the Molecular Recognition Initiative and the British Council for an award to A.K. References T. Kunitake, Angew. Chem., Int. Ed. Engl., 1992, 13, 709. M. J. Blandamer, B. Briggs, P. M. Cullis, J. B. F. N. Engberts, A. Wagenaar, E. Smits, D. Hoekstra and A. Kacperska, J. Chem. SOC., Faraday Trans., 1994,90,2709. A. Wagenaar, L. A. M. Rupert, J. B. F. N. Engberts and D. Hoekstra, J. Org. Chem., 1983,54,2638. T. Kunitake and Y. Okahata, J. Am. Chem. SOC., 1977,99,3860. J. H. Fendler, Membrane Mimetic Chemistry, Wiley, New York, 1982. T. A. A. Fonteijn, J. B. F. N. Engberts and D. Hoekstra, Cell and Model Membrane Interaction, ed. S. Ohki, Plenum Press, New York, 1991. M. J. Blandamer, B. Briggs, J. Burgess, P. M. Cullis and G. Eaton, J. Chem. SOC., Faraday Trans., 1992,88,2874. D. Marsh, Biochim, Biophys. Acta, 1991, 1062, 1. H-N- Lin, Z-Q. Wang and C-H. Huang, Biochim. Biophys. Acta, 1991, 1067, 17. Paper 4/01 93 1H ;Received 30th March, 1994