UNSTABLE STATES OF SOLUTIONS OF SODITJM BEHENATE. 2751 CCCLXXVI1.-Unstable States of S01;zttiolns of Sodium Behenate. By MARY EVELYN Lma. PREVIOUS communications from this laboratory were devoted to describing the most stable forms of soap and its solutions. In the case of sodium behenate it was incidentally noted that the type of solution obtained could be radically affected by suitable treatment. For example on repeated rapid cooling the behenate solutions could be obtained at room temperature for a short period as limpid, mobile liquids whereas slow cooling produced the usual hard white curd. Reheating the mobile liquid of a 0*2N,-solution to about 90" produced a clear stiff isotropic jelly which quick cooling once more temporarily rendered fluid. Some of the most interesting solutions could only be kept less than half an hour.The appearance and behaviour of these solutions and their transitory states have been quantitatively examined. The task was undertaken by a team of experienced investigators, who worked simultaneously on portions of solutions prepared by Mr. G. M. Langdon. He also made the macroscopic observations. The microscopic study was undertaken by Mr. S. E. Wiltshire, Miss M. E. Kieser carried out the E.M.F. experiments Miss M. E. Ling the freezing-point determinations Misa K. M. Hay the indicator tests and Mr. W. C. Quick measured the conductivity. Sodium behenate solutions were made up in silver tubes by shaking the requisite quantities (Bunbury and Martin J. 1914, 105 417) of standardised alkali with weighed quantities of behenic acid m.p. 814-82.0" kindly made for us by Messrs. J. C. Crosfield and Sons Ltd. by the catalytic hydrogenation of Kahlbaum's erucic acid. The homogeneous soaps were transferred to thick-walled glass tubes which were sealed for optical investigation according to the method previously described (McBain and Langdon, this vol. p. 852). Parallel experiments were carried out with solutions made up in 1913 (McBain and Taylor 2. p h p i k d . Chem., 1911 76 179). The strengths used were 0.01 0.05 0.1 0.2 0.5Nw, i.e. mole per 1000 g. of water. The observations here recorded were repeated many times; e.g. a O*lN,-solution was alternately heated and cooled twenty times under diBerent conditions and the results were reproducible the only effect that varied being the number of " flocks " appearing in the mobile liquid.The stable forms of the behenate solutions at room temperature, with the exception of O-OlN, were solid white moist curds. Curd formation however could be suspended by special heat treatment 2752 LAING : as was well illustrated by the behaviour of a O*05N,-solution. On being heated to about 88" the curded O~05Nw-solution melted to a fairly mobile milky liquid containing a few white flakes. I f this solution were allowed to cool slowly more flakes separated and the whole gradually became a network of curd ; the transformation from Brownian particles in movement to short and then long fibres could be observed under the ultramicroscope. Photographs of such gradual transitions have been published elsewhere (McBain Darke, and Salmon PTW.Roy. SOC. 1921 A 98 395; Bogue .' Colloidal Behaviour," Vol. I pp. 41-29 ; Ivature 1921,107,45 ; Alexander, '' Colloid Chemistry Theoretical and Applied," Vol. I 1925). On cooling rapidly the white flakes which formed at about 75', rose to the surface leaving a clear lower layer. On heating to go", these flakes melted and the system consisted of two clear layers. As the tube cooled again curding took place in two stages the upper layer solidifying at 75" and the clear lower layer after some hours a t room temperature. The quantitative investigatian had to be carried out as quickly as possible before the homogeneous mobile liquid produced by rapid cooling set to a mass of curd fibres. The hydroxyl-ion concentration of the O-05Nw-solution at 18" was found by E.M.F.measurements with the hydrogen electrode to be 0-007N during the first hour but this value declined to O~OO08Nw when the clear liquid had solidified to form white curd. The con-centration of hydroxyl ions in the mobile liquid was shown by the indicator method using alizarin yellow G and Sorensen's buffers to be 0.007Nw a t W" and slightly higher at room temperature. The clear O~05Nw-sodium behenate a t 18" therefore is hydrolysed to the extent of 14% (= O-O07Nw-OH) and contains O-02lXW-acid soap expressed in terms of behenate or 0.014Nw in terms of sodium ; if this is formed as in the case of the palmitate according to the equation 3NaBe + H,O = NaOH + 2NaBe,HBe. Hence 42% of the total behenate radical is in the form of acid soap.This acid soap is colloidal and exerts no osmotic pressure (McBain Taylor, and Laing J. 1922,321 621). The osmotic data were obtained from depressions of the freezing point determined by the usual Beckmann method using inoculation : a t least four concordant readings were taken before a solution curded. The depression for the 0.06Nw-solution was 0.036". The hydroxyl ions present to the extent given above would correspond to a depression of 0.013" and together with the equivalent sodium ions would account for a depression of 0.026". The difference 0-OlO", which must be due to that part of the soap in solution as electrolyte, corresponds to 0.0027 g.-mol. of fully dissociated binary crystal UNSTABLE STATES OF SOLUTIONS OF SODIUM BEHENATE.2753 loidal electrolyte. What remains of the total concentration 0.05 -0.014 - 0.007 - 0.0027Nw = 0*0263N, must be the concentration of the undissociated neutral colloid. The constituents of this unstable but clear solution and their concentrations are very nearly those in Table I. TABLE I. Constituents of unstable O-05Nw-80dium Behenate a t 18". 1. Fr0mE.M.P. (OH') 0-007Nw l4o/ hydrolysis slka-3. Equivalent to Na,HBe O.O140N (Na-) 42% acid soap (colloid). 3. From lowering (Na.) = (Be') + (OH') 0.0097N, 4. Remainder (NaBe) O-O263N (Na) 53% neutral colloid. &ty. 5% dissociated so~p. The conductivity data are summarised in Table II. The con-ductivities of two samples of 0.05Nw-solution were measured a t intervals over a period of 20 hours and that of a O-lN,-solution at intervals during 18 hours after making up.The values recorded for the liquid are the means of those obtained within the first hour the curd values are those obtained after 18-20 hours. TABLE 11. Specific Conductivity ( K ) of Sodium Behenate Solutions at 18". Unstable liquid. Final curd. /. /- - / I Conc. (iVm). K. a%. K . a%-0.05 1-74 x low3 15.9 0-311 x 10-3 2.8 0.1 2.79 x lo- 12.8 0.42 x lC3 1-9 The degree of dissociation a calculated by dividing the actual conductivity found by the molar conductivity of sodium hydroxide at infinite dilution at 18" (V;x. 216.5) is 15.9% for the 0.05N,-solution a value slightly greater than that (14%) found by E.M.F. measurements ; the approximate agreement shows that the free hydroxide accounts for most of the conductivity.A more exact comparison may be made by calculating from the data of Table I the specific conductivity of the O-O5N,-solufion at 18". The products of the concentrations per C.C. of the various ions shown in Table I and their respective mobilities (43.6 for Na' 172.9 for OH' and 20.7 for Be') being added togebher the calculated value for the specific conductivity is found to be 1=69~lO-~. The agreement with the observed value 1.74 x is very close especially in view of the fact that different samples of unstable solutions were taken for these measurements ; one sample actually gave the value 1-69 x 10-3 reciprocal ohm. Full allowance having been made for the conductivity due to the hydroxyl ions in the final curd it is evident that the curd fibres, though not in solution (this was shown by analysis of the mother 2754 W G : liquor squeezed out from such curds) contribute to the residual conductivity.This is apparently due to the free ions of the electrical double layer of the very extensive surface of the h e fibres (compare Laing J. PhpimZ Chem. 1924,28 673; Laing and McBain J. 1920 117 1507). The data show that these super-cooled solutions are highly unstable the extent of hydrolysis being many times greater even than that for the solutions at 90". At the lower temperature after a short time the hydrolysis diminishes to a tenth of its value and the equilibrium shifts in favour of the formation of insoluble sodium behenate which separates in curd form from the solution.It seems possible that in the initial clear liquid the acid soap may peptise the neutral soap. Solutions of acid soap froth freely (McBain Taylor and Laing Zoc. cit.). The O-lN,-behenate solution in the unstable condition was similar to the O~05Nw-solution but was less hydrolysed. The concentration of hydroxyl ions due to hydrolysis was shown by the hydrogen electrode to be 0*0098N or 9.8% (afterwards falling to 04012N, when the clear liquid had set t o a white curd). The depression of the freezing point was 0.035". This would be fully accounted for by the free sodium hydroxide but as is seen from Table 11 there is still a small amount of conductivity to be accounted for as in the case of the 0*05N,-solution. The macroscopic and the microscopic behaviour of the 0*1N,-solution are just like those of the O~O5N~-solution described above, except that the instability increases with concentration.The 0.2AT,-solution was too unstable to permit of freezing-point determinations. This solution is very viscous and difKcult to prepare. It exhibited however a remarkable change in viscosity. The melted curd was a clear isotropic jelly at go" but on cooling, the gel " melted " to a mobile cloudy liquid. We have observed this curious phenomenon of gelation at a relatively high temperature and fluidity at a low temperature in only one other instance namely, in a solution of nitro-cotton in alcohol ; this is a clear elastic jelly at room temperature and a mobile fluid at the temperature of liquid air (observation by Mr. L. E.Smith in this laboratory).* This observa-tion can be readily explained on the basis of the theory of neutral colloids put forward by McBain (Trans. Faraday SOC. 1924,20,22). The hydroxyl-ion concentration of the O.Zfl,-solution is 0.005NW, indicating a hydrolysis alkalinity of only 2%. This value corre-sponds exactly with that found by the indicator method. Most of the soap of the O.ZN,-solution must be present as insoluble undis-sociated neutral colloid. * Compare Szegvari (Kohid-Ztg. 1924 34 34) who however used a more complicated mixture of solvent and non-solvent UNSTABLE STATES OF SOLUTIONS OF SODIUM BEHENATE. 2755 Whether cooled rapidly or slowly the O*SN,-solution of sodium behenate forms curd a t room temperature. It is a highly viscous gel at 90".It appears that ease of curding increases with rise of Viscosity, which takes place whenever separated matter (shown by streakiness) is present or upon lowering of temperature. Curd also forms more rapidly after the solution has been heated in glass. A sodium behenate solution separates into two layers on addition of 0*4iV,-sodium chlbride; whereas in the case of sodium palmitate of the same concentration twice this quantity of salt is required t o produce the same change. The salted-out system 0.05 N,-NaBe + 0*42N,-NaCl and also O.OSN,-NaBe + 0-59NW-Nac1 becomes homogeneous a t 100". This temperature is above that 72-75' at which the corresponding palmitate-sodium chloride system becomes homogeneous. Both facts are in accordance with the insolubility of sodium behenate at low temperatures.Neutral behenate solutions in the absence of sodium chloride, were never observed to separate spontaneously into two layers on standmg a t high temperatures as is the case with acid sodium palmitate (0-4N,-NaP O-lN,-HP). In appearance the acid salts of behenic and palmitic acids are very similar. On reviewing the behaviour of these colloidal behenate systems, one is struck by their analogy with certain crystdoidal system. When for example barium sulphate is produced rapidly from a mixture of barium chloride and a sulphate a highly supersaturated solution may be obtained which deposits matter in the colloidal state. On sudden cooling of the behenate solutions supersaturation with respect to neutral colloid sets in and similar amorphous, flocculent material is formed. Thus on cooling of either system, colloid or crystalloidal one gets separation of (crystalline) curd fibres on the one hand and of true crystals on the other. On standing in contact with their respective mother-liquors colloidal barium sulphate crystallises and colloidal sodium behenate curds. Summary. (1) Although the stable form of 0~05-0~5Nw-sodium behenate at room temperature is a hard white curd these solutions can be obtained temporarily as clear very mobile liquids which are hydro-lysed to an abnormally large extent. (2) The largest constituent of the unstable mobile solutions is neutral undissociated colloid the next largest being colloidal acid sodium soap with the equivalent quantity of free sodium hydroxide. There is only a very small proportion of dissociated soap. UNIVERSITY OF BEISTOL. [Received July 24th 1925.1 5 a