SOLUBILITY IN SOAP SOLUTIONS PART 10.PHASE EQUILIBRIUM, STRUCTURAL AND DIFmJSION PHENOMENA INVOLVING THE TERNARY LIQUID CRYSTALLINE PHASE BY A. S . C. LAWRENCE Dept. of Chemistry, Sheffield University Received 3 1 st January, 1958 The structural features described here are discussed in terms of interaction between the ionized group in a soap and the polar group of the organic compound in the presence of water ; as partial miscibility in a three-component system ; and as a balance between the resultant enhanced solubility and the opposing insolubility of the hydrocarbon chains. The regularity of the structures is due, in part at any rate, to the necessity for long-chain molecules to pack into layer lattices when in a condensed state. Charge effects are markedly absent. The upper and lower limits to the existence of the ternary liquid crystal- line phase are described.With aliphatic solutes, the lower limit is found to depend upon the polar groups in both soap and solute and upon the length and shape of their hydro- carbon chains ; correlation with melting temperature of the solute is shown. It is now well recognized that organic substances insoluble or sparingly soluble in water are much more soluble in an aqueous solution of any soap ; their solubility increases with increase of soap concentration. It was pointed out by the writer in 1937 1 that, when the organic third component, called additive for brevity, contains a polar group such as -OH, -COOH, -NH2 or, as shown later, C H group, the solubility of both soap and additive are enhanced as, indeed, might be expected from simple thermodynamic considerations. Later, McBain intro- duced the unfortunate name “ solubilization ” to include solubility of both polar and non-polar substances, ignoring the differences between the two systems.It has been shown in this laboratory that the system soap + water + polar sub- stance is the simple one of 3 partially miscible components and is peculiar only in that there exists an area of ternary liquid crystalline phase; this, although novel, is analogous to the well-known cases where the ternary phase is a liquid or solid solution.2 Complete triangular equilibrium diagrams have been worked out for the system sodium dodecylsulphate + water + caproic acid and for the same soap plus water and n-octylamine; 3 Dervician 4 has published the diagram for the system potassium caprate + n-octanol + water.It should also be noted that, for aliphatic additives containing more than 5 normal carbon atoms, the region of isotropic solution included in McBain’s solubilization is a small part of the whole system. It is often convenient to work with a chosen concentration of soap and observe the changes with progressive addition of additive; with larger amounts of the latter, the overall concentration of the soap is much reduced but the soap/water ratio, which is important, is unchanged. The general result for any soap, anionic or cationic, with aliphatic homologous alcohols, carboxylates, amines, etc., in which the number of normal C atoms is 5 or more and so long as they are above their melting points is shown in fig.1.1 The initial soap concentration must be at least about 10 %; below that concentration, only two liquid layers are formed with the soap plus some water mainly in the organic-rich layer.3~ 5 5152 SOLUBILITY IN SOAP SOLUTIONS In fig. 1,1 E is the point at which separation into two layers occurs at high concentration of additive. If the additive is solid at room temperature, then it crystallizes at its T f - aAT for the other components but, around E, this AT, is not more than 2 to 3°C. It was observed, however, that at lower con- centrations, ternary liquid crystalline systems persist where AT, is much larger. An extreme case is cholesterol, m.p. 148-5, which remains in the liquid crystalline state at room temperature with cholesterol to soap ratio of 1 mole : 1 mole.This is not a metastable condition since sufficient cholesterol can be dissolved in a soap solution at 100°C, so that, on cooling, some crystallizes out and remains in equilibrium with the saturated liquid crystalline phase. The reverse procedure was therefore tried : a single crystal of either cholesterol or cholesterol hydrate (the latter are more readily prepared) is placed on a microscope slide and flooded FIG. 1.-Sodium dodecyl sulphate + water + caproic acid system at 25°C. with soap solution. It is viewed by polarized light with the Nicols at 45". As soon as the soap solution touches the crystal, a gelatinous membrane is formed; this retards attack but it gradually becomes more fluid, the soap penetrates further and surface tension pulls the ternary liquid crystalline phase into more or less spherical anisotropic lumps.At the same time, the exteriors of the liquid crystals are dissolving away into the exterior soap solution but more slowly than the in- vasion of the crystal. Fig. 2 shows the sequence of the events but without the initial membrane formation ; the sequence shown took 40 min from start to com- plete solution. This experiment was then tried on dodecanol and similar results obtained but the higher homologues formed no liquid crystal phase, but only a very slow solu- tion direct to an isotropic state over 10 to 12 h. When, however, they were heated on a hot stage, the penetration process and liquid crystal formation set in suddenly at a temperature Tp characteristic of each substance and below its melting point.Fig. 3 shows this transition temperature for the homologous even-number fatty acids in a number of soap solutions. Tp was found not to change with concentra- tion of soap over wide limits : e.g. for myristic acid in Teepol, T, was 30°C for concentration of soap from 5 to 33 7;. It is seen that the penetration temperature lies below the m.p. of the solid ; that it is not merely a thermal opening- up of the solid is shown in that Tp is clearly also dependent upon the nature of the Fig. 4 shows T, for the even-number iz-alkanols.A. S. C . LAWRENCE 53 FIG. 3.-M.p. x and Tp, 0 for fatty acids in 15 % solution of various soaps. I6 I 18 20 , 10 I 2 14 Cn FIG. 4.-M.p. x and Tp. 0 for n-alkanols in 15 % solutions of various soaps.54 SOLUBILITY IN SOAP SOLUTIONS polar group of the soap.The chain length of the soap seems to affect the slope of the curves : e.g. the two c16 soaps as compared with the C12 ones. The Teepol curve lies surprisingly high compared with the n-C12 sulphate ; this may be due to the chain branching in the Teepol. It was also found that samples of fatty acids not quite pure and melting a few degrees below the true value also had their Tp lowered by an equivalent amount. Mixtures of stearic and palmitic acids were made up and their freezing points and Tp in 15 % Teepol observed ; fig. 5 shows the eutectic for f.p. in agreement 7c 6( 0; 5 ( 4c 3c I with the results of Francis, Collins and Piper 6 and a eutectoid graph for Tp with the minimum at the same composition as that for f.p.Since we have found in these ternary systems marked shape effects upon phase equilibria and solubility, we tried oleic and elaidic, erucic and brassidic, and stearolic and behenolic acids ; the results are shown graphically in fig. 5 ; ,Tp could not be measured for oleic acid as it was below 0" (in Teepol). A lower limit to the phenomenon is set by the Krafft point of the soap, i.e. the temperature at which its solubility in water drops to a very small value : for Teepol this is below 0" but is considerably higher for the straight-chain soaps. MYELINIC FIGURES When cholesterol is dissolved in soap solutions, many rings, usually short fat sausages, are seen ; the ring is sharply defined, strongly birefringent and con- 43 .b .4 I .2 tains isotropic solution. This form of the liquid crystalline phase is rarely seen Compori tion FIG. 5.-M.p. and rp in 15 % Teepol with aliphatic solutes where the penetra- ternary phase at T', first by formation of a membrane around the solid and then by bulging into hemispherical anisotropic forms and finally sometimes to spherul- ites. Ekwall et al.7 using carboxylate soaps, find various structures during the penetration into liquid alcohols, c6, C7 and Cg ; we also find similar features at temperatures above Tp, as was the case in Ekwall's studies. It seems doubtful whether the liquid crystalline sausages formed should be called " myelinic " ; the latter are tubes filled with liquid ; these " sausages " are birefringent across the lumen of the tube.They are, in fact, temporarily distorted liquid crystalline spherulites. In myeline formation, water is diffusing into a solute, forming the membrane of sufficient strength; further swelling must result in expansion and the tubular forms result from their smectic liquid crystalline nature ; i.e. they have one plane of easy fluid shear so that telescopic extension of the tubes results. The long axes of the molecules forming the membrane are, of course, normal to the length of the tube. In the penetration of soap solution into a homogeneous solid (or liquid) amphi- philic substance, the soap molecules, as well as water, must diffuse through the membrane first formed. From bulk diffusion experiments in which decanol was solution for mixtures of stearic and palmitic tion ofthe soap leads to liquid crystalline acids.A .S . C. LAWRENCE 55 placed on top of 15 % soap solution,8 a membrane of liquid crystalline material was formed through which soap passed readily into the decanol with a small amount of water. In two days, the decanol layer had become a paste of ternary liquid crystalline phase and it was only after 3 months that water had diffused in sufficiently to double approximately its volume. With these ternary systems, it would appear that the solid which forms myelines in water must contain two components-soap and a long-chain amphi- phile which, it will be noted, is the type of mixture which has an exceptionally large surface plasticity on aqueous solutions 9-and it usually appears to be hetero- geneous. Sodium laurate and lauric acid in equimolar proportions form myelines in water ; very slowly at room temperature but much more rapidly at 40°C (i.e.above the m.p. of lauric acid). Oleic and sodium laurate give them quickly at 80- 60 y 4 0 - a. ; - / / ?;I, , p , , , / / ( 0 C“ Y 8 10 12 14 16 18 2 0 2 2 FIG. 6.-M.p. x - x , and Tp 0- -0 for saturated, cis and trans ethylenic and - - - - acetylenic fatty acids ; in 15 % Teepol. room temperature. When, however, the K soap and oleic mixture was tried, no myelines were formed ; only liquid crystalline bulges which fioated free as liquid crystalline spherulites after the manner of the penetration solubility process. PHASE EQUILIBRIA IN SOAP $- WATER $- AMPHIPHILE SYSTEMS The general picture of the ternary phase diagram at room temperature is now well established but little work has been reported upon temperature changes and equilibrium at higher temperatures and nothing upon lower temperature limits of existence.In this laboratory we have accumulated data on parts of several systems over the temperature range from 210°C to below 0” and can now form a general picture. if we keep the soaplwater ratio constant and add various amounts of amphiphile, we can plot TIC diagrams as for binary systems by using the Krafft point as the “freezing point” for soap + water as one component and the orthodox melting point of the additive as the other. Fig. 7(a) shows the picture diagrammatically for systems in which the m.p. of the organic amphiphilic com- ponent is below the Krafit point ; 7(b) shows a reverse case in which it is above.Fig. 7(a) includes three classes in which the three broken lines represent room temperature : (i) is that in which no liquid crystalline ternary phase is formed ; EX is a saturation point at which the single liquid phase becomes saturated and forms two liquid phases. This is found with low molecular weight amphiphiles,56 SOLUBILITY I N SOAP SOLUTIONS usually with a small solubility in water; e.g. butanols, eresols, aniline, benzyl alcohol, etc. Case (ii) is that shown in fig. 1. It is found generally for all amphi- philes containing 5 or more n-C atoms; of the two isotropic solutions beyond E, one may be a gel which with some soaps, e.g. carboxylates, only becomes fluid at about 130°C. The point B has always been well above both m.p. of amphiphile and Krafft point of soap ; it is usually above 100°C.With cholesterol and sodium dodecyl sulphate in equimolar amount and the soap in 20 % of aqueous solution, B was 195°C. EX frequently slopes backwards as shown. For (iii) we have two possibilities. With normal soaps whose Krafft point is above 0", the first solid component to separate is pure soap. If the Krafft pt. is itself below zero (as with Teepol) and m.p.A still lower, then ice is the first solid to appear on cooling. All of these cases have been found. The separation of soap from solutions of X K .Pt I P'* 1.0 soap so1n I compoittion X \ I ~otroplc iolution \ 2 Irottoplc 1 Soap Sol" I.( composition FIG. 7.-Temp./comp. phase equilibrium. (a) m.p. of A below Krafft point.(b) m.p. of A above Krafft point. C120S03Na and C16Me3NBr in mixtures of water and lower amines has been observed : e.g. with NEt3, NPr3, N-isoPr3, N-BuzH, N-BuzMe, N-isoBu2, N- sec.Bu2, etc. Also from lauric acid in C120S03Na solutions at the water-rich corner of the triangular diagram. Fig. 7(b) in which m.p.A is above the Krafft pt. it is clear that the amphiphile freezes out first. Since, however, stable liquid crystalline systems persist at room temperature over part of the concentration range, there must be some sort of eutectoid graph. With C120S03Na + water + lauric acid, the freezing point drops rather slowly at first from pure lauric acid and dips more steeply later. The graph exaggerates the dip. This type of behaviour is that of all soaps and all fatty amphiphiles with m.p.above room temperature, i.e. roughly from Cl2 upwards. SOLID SOLUTION FORMATION During the examination of the dodecanol in sodium dodecyl sulphate system in this laboratory, Mr. K. Hume observed the existence of a large amount of solid persisting above the m.p. of the alcohol and the Krafft pt. of the soap, 23-8"C and 22-23' respectively. The material, with slow cooling, formed large crystals of a perfection never seen in soap crystals; in place of the usual mosaic of im- perfect flakes, uniform six-sided single plates existed: these appeared to be rhombic plates with two opposite corners truncated. On warming on a hot stage in the polarizing microscope, they were seen to melt uniformly and sharply, at(b) FIG. 9.-Crystals of solid solution of sodium dodecyl sulphate + water + dodecanol.Crossed Nicols. x 60 [To face page 57A . S. C . LAWRENCE 57 temperatures falling from 32" to 29°C with increasing additions of dodecanol. On further heating the isotropic solution becomes cloudy by formation of the liquid crystalline phase and clears at a higher temperature. Tetradecanol be- haves similarly but does not clear at 100". Analysis showed these crystals to contain water which was determined by distillation in xylene (Dean and Stark method). Another sample was dissolved in 50 % ethanol in water and the dodecanol extracted by petroleum ether ; the soap was determined by the methylene blue method. It was then realized that, after removal of all water, the soap should remain in the boiler insoluble in xylene while the dodecanol will be in solution in it; the soap filtered easily and was washed, dried and weighed and the xylene removed from the dodecanol which was also weighed.No similar solid phase was observed in the ternary system in which dodecoic acid replaced dodecanol as the third component. Crystals of the solid solution are shown in fig. 9. DISCUSSION In the absence of water, ionic soaps of all kinds are generally poorly soluble in organic liquids; with water present, high solubilities are found with lower molecular weight substances containing a polar group; e.g. ethanol, di- and tri-propylamines and butylamines, di-ethyl ketone, phenols, etc. With higher homologues (C5n and above) ternary liquid crystalline phase is formed; this is of the smectic type and is therefore a layer lattice.In the presence of caproic acid in excess, 7 molecules of water per molecule of sodium dodecyl sulphate is the minimum amount to convert all the soap to liquid crystalline phase at room temperature; i.e. to prevent any solid soap crystallizing out. The maximum which the smectic phase can hold is 110 molecules of water/molecule of soap- that is about 10 % soap in water; with more water the system breaks down to two liquid phases. Similar figures are obtained with the same soap in n-octylamine.3 The system is visualized as one in which the ionized head of the soap molecule and the polar group of the added amphiphilic substance are forming a local solu- tion while the attached hydrocarbon chains are insoluble. The closeness of the packing is attested by the Ay produced by the additives, by the very large surface plasticity of the binary layer adsorbed on the water, and by the - A V on mixing the components; Mr.B. Boffey has shown in this laboratory that the partial specific volume of soap and water decreases on addition of additive. The smectic layer arrangement is a sandwich of which the centre is a layer of water whose thickness is from 8 to 13Ow for 7 to 110 moles water/mole of soap. The water can be regarded as held together by its own high internal pressure and the liquid crystalline nature by the terminal polar groups being in local solution in the water. Crystallization in 3 dimensions require them to crystallize from the water. The high temperatures of the transition from smectic to isotropic solution do not require large specific forces in the planes-of the polar heads because increase of temperature does not increase the solubility of the hydrocarbon tails.Given that water separates, it cannot form a bulk phase because the soap and amphiphile polar heads are dissolved in it ; it must form either an emulsion or the smectic liquid crystals and we cannot preclude lateral attraction between polar groups. On the other hand, the water evaporates readily from the additive-rich liquid crystals and measurements of vapour pressures of ternary systems by Mr. Boffey show large positive deviations from Raoult's law. In the soaps and amphiphiles, we have the necessary condition for this abnormally low attachment of A to B since the shape and amphiphilic nature of the molecules reduce markedly the extent to which complete homogeneity on a molecular scale can be achieved.The aggregation of soaps in water is a similar case which, in the formation of micelles, is an incipient separation into two phases: i.e. the extreme case of positive deviation from Raoult's law. The structural properties all follow from the smectic layer lattice but it is clear that the plane of fluid flow is the water58 SOLUBILITY I N SOAP SOLUTIONS inside the sandwich whereas in classical smectic melts it is the plane of terminal methyl groups. Shearing is therefore a process which breaks up the unit of the sandwich structure; this may be the explanation of the silky thread texture in the very high viscosity region in which any reformation of larger units is extremely slow ; in the organic components-rich region, where viscosity is low, reformation to classical textures of focal conics and bgtonnets is rapid.In a myeline sheath we have the smectic layer lattice curved into a closed cylinder; there is therefore a longitudinal channel of water along which con- ductivity and diffusion can take place but not at right-angles to this direction (fig. 8). u 0 c z I i I I m y e I i n e I 1 sheath I I FIG. &-Soap + amphiphile + water myeline. It is sometimes suggested that addition of a wetting agent to water in contact with a lipid membrane initiates semipermeability by its wetting action but this is surely incorrect. The wetting agent can alter the permeability only by entering into the membrane and altering its nature; such action is particularly likely to occur when the membrane is composed of substances adsorbed from and in equilibrium with the solutions on either side of it. The diffusion results mentioned here are examples; they also illustrate the wide control, available by varying concentration or temperature, over the physical condition of such membranes. 1 Lawrence, Trans. Faraday SOC., 1937, 33, 325. 2 Faraday Soc. Discussions, 1954, 18, 239. 3 in course of publication. 4 Proc. 2nd Int. Congr. Surface Activity, 1957, 1, 327. 5 Lawrence and Stenson, 2nd Int. Congr. Surface Activity, 1957, 1, 388. 6Proc. Roy. Soc. A , 1937, 158,691. 7 Ekwall, Salonen, Krokfors and Danielsson, Acta Chem. Scand., 1956, 10, 1146. 8 Lawrence, 2nd Int. Congr. Surface Activity, 1957, 1, 475. 9 Blakey and Lawrence, Faraduy Soc. Discussions, 1954, 18, 268.