Aug., 1950
SOLUBILIZATION IN ALCOHOL-SOAP MICELLES
3581
Solubilization in Alcohol-Soap Micelles’ BY H. B. KLEVENS’ There have been a number of indications that when the refractive indices of soap solution and solubilized are different. there are a t least two loci of solubilization in soap material To determine solubilities in swollen micelles the followmicelles. Definite evidence, based on extensive ing procedure was adopted. To weighed amounts of soap X-ray studies of soap solution-hydrocarbon sys- solutions in ampules, a weighed amount of a long-chain tems, that the hydrocarbon center of the micelle is alcohol or other polar material was added, ampules were sealed off and shaken for twenty-four hours. Increasing one of these loci has been summarized by M ~ B a i n . ~amounts of n-heptane were then added and the vials were The very recent work on similar X-ray studies on placed in a shaker for another forty-eight hours or more soap solution-polar compounds by McBain and until constant optical density values (D) were obtained. McHan5on systems containing soap solutions and Optical densities were obtained and when plotted against moles of added hydrocarbon (if the systems below saturaadded dimethyl phthalate and on long-chain al- tion were clear) yielded two straight lines, the intercept cohol-soap solutions6 in which no increase (at of which was the solubility. One line had a constant or times a decrease) in long spacing is noted can only slightly decreasing D with added hydrocarbon below satbe explained by postulating a second loci of solu- uration in which D was essentially equal to that of the original soap solution ; the other increased markedly a t bilization. In these systems, the long-chain alco- saturation and D increased linearly with further addition of hols must take a position like that of the soap hydrocarbon. molecules with their polar groups oriented toIf the amount of n-heptanol added initially exceeds its ward the water layer and their hydrocarbon tails normal solubility in KClr solutions, and if to these solutions are added various amounts of n-heptane, curves toward the micelle center. This has previously similar to that seen in Fig. 1 are obtained. The initial been noted from the extensive X-ray studies of opacity is due to the presence of n-heptanol emulsion dropMattoon’ on soap mixtures in which it was found lets in equilibrium with mixed KClr-n-heptanol micelles. that the long X-ray spacing varies linearly with Addition of n-heptane, which enters the hydrocarbon-like of the soap-alcohol micelle, causes the micelle to mole ratio of the two soaps in the mixture. This center swell. This then allows more alcohol molecules to enter would indicate mutual orientation of each soap the micelle as is indicated by a marked decrease in D . in the mixed micelle for in soap mixtures the more The solution then becomes clear upon addition of further soluble soap can be expected to act as a solubilizer n-heptane indicating that all the excess alcohol molecules as well as added hydrocarbon have been solubilized. for the less soluble Addition of further n-heptane yields the typical two If the above hypothesis, as the two types of straight lines mentioned above in which the intercept is the solubilization, or to be more exact, two loci of limit of solubility. The limit of solubility is then this solubilization in the micelle, is correct, i t should second transition point. The presence and independence of the two loci of solubilization is definitely indicated by be possible to prepare swollen micelles with en- the results of this type of experiment. There appears hanced solubilization characteristics. Two meth- to be little or no mutual mixing of these polar and nonods of preparation are possible by essentially sat- polar compounds being solubilized. I t is also possible to urating soap solutions with (1) a polar compound obtain these end-points (limits of solubility) by reversing such as a long-chain alcohol and with (2) a hydro- the order of the addition of compounds to be solubilized. carbon. Data, indicating as much as a tenfold increase in amount of hydrocarbon solubilized by alcohol-soap micelles, have been obtained. 1.2
J
Experimental Methods Saturation of micelles with additive, either hydrocarbon or polar compound, was determined by the turbidimetric method described previously.8 This method is based on the fact that as the saturation point is exceeded, emulsification, rather than solubilization, begins and the solutions become increasingly turbid with increment addition of hydrocarbon or polar compound. This method is valid (1) Presented a t 116th American Chemical Society Meeting, Atlantic City, New Jersey, September 18-23, 1949. (2) Division of Agricultural Biochemistry, University of Minnesota, St. Paul, Minn. (3) H. B. Klevens, J . Am. Oil Chcm. S O L 26, , 456 (1949); W. D. Harkins and H. Oppenheimer, THIS J O U R N A L , T I , 808 (1949). (4) J. W. McBain, “Advances in Colloid Science,” Vol. I, Interscience Publishers, Inc., New York, A-, Y., 1942. ( 5 ) J. W. McBain and H. McHan, THISJ O U R N A L , 70, 3838 (1948); J. W. McBain and 0. A. Hoffman, J.Phys. Colloid Chcm., 68, 39 (1949). (6) W. D. Harkins, R. W. hfattoon and R. Mittelmann, J . Chcm. Phys., 16, 763 (1947). (7) W. D. Harkins, R. W. Mattoon and hl. L. Corrin, THIS J O U R N A L , 68, 220 (1946). (8) W.Heller and H. B. Klevens, J. Chcm. Phys., 14, 567 (1946)
c
d
I
I
/
I
I I I I I 0.5 0.7 0.9 Mole n-C~Hl~/lOOO g. 0.375 M KCId 0.315 M ~ - C I O H Fig. 1.-Effect of added n-heptane upon the turbidity of a soap-alcohol system in which initial alcohol concentration is greater than its solubility.
1
0.3
I
+
Potassium tetradecanoate was prepared by saponification of the corresponding methyl ester, which had been
H. B. KLEVENS
3582
fractionally distilled a t low pressures. Three recrystallizations from ethanol and from ethanol-acetone mixtures were made. To prevent hydrolysis, the soap solutions were so prepared that they contained initially an excess of about 5% potassium hydroxide. The long-chain alcohols and mercaptan were obtained from Humphrey-Wilkinson Co. (formerly Chemical Division of Connecticut Hard Rubber Co.) and from Eastman Kodak Co. The long-chain amine was supplied by the Armour Laboratories. These were purified by fractional distillation. The n-heptane was obtained from Westvaco Chlorine Co. and since it had been shown t o have only negligible amounts of unsaturated or branch-chain impurities by spectroscopic measurement^,^ it was used as obtained. All solubilization experiments were run a t 25 =t2’.
Effect of Added n-Heptanol on Solubility of n-Heptane in Potassium Tetradecanoate Solutions.-The effect of polar groups on solubilization is seen in Fig. 2 where the solubility of nheptane and n-heptanol in potassium tetradecanoate solutions (KC14) are compared. These results are essentially in agreement with those reported recently.g At concentrations above 0.380.40 M KC14,gel formation occurs upon addition of n-heptanol to KC14 solutions which interferes somewhat with completing mixing. These gels can be liquefied upon slight heating or upon the addition of a hydrocarbon such as n-heptane.
I
Mole n-C,Hl&OHin 0.375 M KC,(. 0.1 0.2 0.3 I I i I I
enhancement in solubilizing power of the shorterchain non-charged alcohol or compared with the KC14 molecule. Effect of Alcohol Chain Length on Enhancedata ment of Solubilization of n-Heptane.-The indicating the effect of added long-chain alcohols, n-heptanol through n-dodecanol, on the increased solubilizing power of 0.35 M KC14 solutions for nheptane are collected in Table I. I t is evident from these data that, with increase in chain length of the alcohol, there is an ever increased solubilization of n-heptane. This increase is more pronounced than the addition of equivalent amounts of KC14. This enhanced effect is well illustrated in Fig. 3 in which the effect of the addition of these alcohols as well as additional KC14 on the solubilizing power of 0.35 M KC14 can be seen. The solubilization of n-heptane is plotted against total moles of solubilizer (0.35 M KC14 n moles KC14 or some long-chain alcohol).
+
I
0.9
i
I
I
I
I .
t
A
IA
p’i n- CH ,,
Vol. 72
L i
/ /
in
0.1 0.3 0.5 B Moles KC14/1000 g. of solution. Fig. 2.--Solubilization of n-heptane and n-heptanol in potassium tetradecanoate (KC14) solutions (B) : enhancement of n-heptane solubility in 0.375 MKCtr by addition of n-heptanol (A). Arrow indicates limit of solubility of nheptanol in 0.375 M KC1, solutions.
A marked enhancement of solubility of n-heptane is obtained upon addition of n-heptanol to KC14 solutions. This is clearly evident as indicated in Fig. 2 in the case where varying amounts of n-heptanol are added to 0.375 M KC14 solutions before solubilization of n-heptane is attempted. A linear increase in solubilization of n-heptane with increase in n-heptanol is noted. The slope of this curve is much steeper than the solubility of n-heptane in pure KCld solutions indicating an (9) H. B. Klevens and J. R. Platt, THIS JOURNAL, 69, 3056 (1947).
t-
1
0
0.1 0.2 0.3 Mole added t o 0.35 M KClr. Fig. 3.--Enhancement of solubilization of n-heptane in 0.35 M potassium tetradecanoate solutions by the addition of long-chain alcohols. Dotted curves indicate the corresponding effect when other fatty acid soaps are the additives.
For comparison with the effect of long chain alcohols as additives, data obtained with the addition of other fatty acid soapsloare included in Fig. 3. The addition of potassium dodecanoate (KC12) and shorter chain soaps results in a decrease in solubilization below that of equivalent amounts of KC,, whereas the addition of potassium hexadecanoate (KC16) shows a marked increase in solubilization. The increase upon the (10) H. B. Klevens, unpublished data.
SOLUBILIZATION IN ALCOHOL-SOAP MICELLES
Aug., 1950
3583
TABLEI
EFFECT OF ADDEDLONGCHAIN ALCOHOLS ON SOLUBILIZATION OF n-HEmANE IN 0.35 M POTASSIUM TETRADECANOATE SOLUTIONS Moles added aIcoho1/1000 g. KC14 soln.
0 0.0411 .0812 .120 ,195 .230 .270 .318
0.0380 .0902 .158 ,201 ,249
0.0381 .m1 .126 .208 .248
0.0390 .lo7 .162 .184 .264 .322
0.0221 .0442 .0901 .131 .156
Moles n-C7H1s solubilized/ 1000 g. alcoholKCI4 soh.
0.116 ,174 ,222 ,268 ,356 ,395 .442 ,500
0.183 .262 .375 .451 ,528
0.198 ,307 .386 .558 .647
0,227 ,410 ,570 ,629 .855 .989
0.220 ,301 .472 ,645 ,737
Moles KC14 Moles n-c7Hl8 Molecules Moles n-CIHls/ in KC14-alcohol M o l e s -~ Sol.in KC14 (alcohol f KC14) (alcohol KClr) micelle
+
n-Heptanol 1.0 0.895
1.0 1.49 1.91 2.31 3.07 3.40 3.82 4.32
,810
,745 ,642 ,603 ,563 ,524
1.58 2.26 2.78 3.90 4.55
n-Octanol 0.902 ,796 .690 .636 .583
1.71 2.65 3.33 4.81 5.58
n-Nonanol 0.902 ,796 ,734 ,627 ,585
1.96 3.53 4.91 5.41 7.37 8.53
n-Decanol 0.900 .765 ,683 ,655 .570 ,521 n-Dodecanol 0.942
1.89 2.60 4.07 5.57 6.35
,888 ,795 ,727 ,692
addition of KCle is much more than could be expected if only the increase in chain length by two carbon atoms over KClr is important. The insolubility of the KCla a t the temperature a t which these experiments have been run and the equivalent insolubility of the long-chain alcohols may also be factors in addition to chain length which influence this enhancement in solubilization. The d e c t of increased temperature on these solubilization experiments will throw more light on this point. If points are taken a t a definite solubilizer con0.10 M centration, say 0.45 M (0.35 M KC14 additive) for each of the additives and if these are plotted aa functions of alcohol chain length or
+
0.331 ,445 ,515 ,570 ,652 ,695 ,714 .747
0.472 ,595 .739 .820
.a0
0.510 ,698 .811 1.01 1.08
0.583 .895 1.11 1.18 1.39 1.47
0.597 .765 1.07 1.34 1.46
22 30 34 38 44 46 48 50
31 40 49 55 59
34 47 54 67 72
39 60 74 78 92 98
40 51 72 90 97
No. KCM +molecules/ alcohol
Molecules KCi4/ micelle
micelle
66 80
Av.
66 71 71 71 70 65 63 61 67
Av.
73 78 77 77 74 76
Av.
79 86 87 86 84 84
Av.
87 98 101 100 97 94 96
Av.
92 102 109 122 122 109
88 94 105 108 111 114
82 98 112 120 126
87 109 118
138 144
96 128 147 153 170 179
98 115 139 168 177
number of carbon atoms in the chain, curves similar to those in Fig. 4 are obtained. The increase in slope of these curves is indicative of the effect of chain length in a homologous series on additive properties. It should be noted that, for any particular concentration, there is no linear relationship between solubilization and chain length. If these curves are extended, using the equivalent decrease in slope found a t longer chain lengths, they will be found to intercept the equivalent KC14 concentrations (indicated by bars in Fig. 4) a t points which would indicate that KCl4 would correspond to an alcohol of chain length of about five carbon atoms. Actually, the addition of npentanol to KCM solutions has resulted in in-
H. B. KLEVENS
3584
Vol. 72
also be used but since the effect of addition of long-chain alcohols on micellar size is not defiSOLUBl LIZ E R nitely known, the smaller value is used in these calCONCENTRATION culations. A constant factor can be used to correct these figures if a more definite value is obtained. By use of a value of 66 molecules per micelle, it is seen (column 6, Table I) that, with increasing alcohol concentration, the number of n-heptane molecules solubilized per micelle increases more than fourfold. Over the same concentration range, from 0.35 to 0.60 M , the number of n-heptane molecules ( M ) solubilized per micelle when KC14 is the additive only changes from 22 to 25, whereas, with n-decanol as additive, the increase is from 22 to 90. By assuming that the density of n-heptane in all these swollen micelles is the same, it is possible to calculate directly, using the value of 22 n-heptane molecules per micelle, the number of alcohol and soap molecules in a micelle necessary to accommodate this increased amount of hydrocarbon. The values of N , the total number of mole0.1 cules, alcohol plus soap, in this micelle are collected in column 7 of Table I. With increase in KC14 7 8 9 10 11 12 added alcohol, N is seen to approach a constant value. Xumber C atoms in alcohol. Since the initial amounts of soap and added alFig. 4.-Effect of chain-length of alcohol additive for cohols are known, it is possible, from this initial constant solubilizer concentration (0 35 M KC14+ n moles additive) upon solubilization of n-heptane. Bars indicate mole ratio and from the calculated number of soap and alcohol molecules per micelle, to calculate the effect of KClr as additive. number of KC14 molecules per micelle. These are seen in the last column in Table I and creases in solubilization above this extrapolated values are found to be fairly constant for any one added value due possibly to the solubility of n-heptane alcohol. The averaged value for any added alcoin the C6 alcohol. hol is seen to increase with increase in chain length The ratio indicating the increased solubilization of n-heptane in alcohol-soap micelles is indicated of the alcohol. The entrance of the alcohol into the palisade layer of the soap micelle will bring in column 3 of Table I. No attempts have been about a change in the curvature of the micelle made to extend this enhancement further. The surface which will be further enhanced by an insolubilities of the longer-chain alcohols are excrease in the length of the hydrocarbon tail of the tremely low and their effect as additives has not long-chain alcohol. been determined. The addition of further The Effect of Change in the Polar Group on amounts of long chain alcohols beyond equivalent Solubilization.-Data, similar to those obtained concentrations of alcohol and soap has not been showing the enhancement of solubility of nattempted so no answer as to the minimum num- C,H16 by the addition of long-chain alcohols, have ber of soap molecules necessary for micelle formaobtained with added amines and mercaptans tion in the presence of relatively large amounts been as seen in Table 11. I n contrast to the effect of of long-chain alcohol molecules can be advanced. added n-octanol, substitution in the additive moleHowever, when these data are obtained, much cule of -NH2 and S H for the -OH group results more information can be surmised as to the role of in a 1.6- and 1.9-fold increase respectively over both the charged hydrophilic head and the hydro- that of the added alcohol. The n-octylamine is phobic long-chain tail on orientation of soap seen to be about equivalent to n-decanol in the molecules in solution and on their aggregation as increase of the solubilizing power of KC14, and the micelles. From the mole ratio of hydrocarbon to solubil- n-octylmercaptan falls midway between n-decanol alcohol), i t is possible to calculate and n-dodecanol, approximately where one would izer (soap the number of n-heptane molecules per soap mi- expect the data for n-undecanol to fall. This incelle if a value of 66 soap molecules per micelle is crease found when -NH2 and -SH are substituted used. This is about an average of the values for a for -OH groups in an alkyl chain of the same soap of this chain length." A value of about 150 length would indicate that more than just the molecules per hydrocarbon swollen micelle12might length of the alkyl chain is important in micelle formation and in solubilization, Thus the ef(11) P. Dehye, J . Phys. Colloid Chcm., US, 1 (1949). fect of added -SH>-NH2>-OH for equal chain (12) H.B. Elevenr, ik'd., in pres..
4
t-
-I
I
+
Aug., 1950
SOLUBILIZATION IN ALCOHOL-SOAP MICELLES
3.555
traction energy is much larger in the case where the additive is a long chain alcohol than where i t is the same or another soap. Further, as the Discussion chain length of the alcohol is increased, there is The addition of long-chain alcohols, amines and a corresponding increase in the total attraction. mercaptans to soap solutions has been shown to This would result in some increase in micelle size, greatly enhance the solubilization of n-heptane. and coupled with this, an increase in solubilizing This increase must depend on a t least two factors : power. This is consistent with the results on the (1) alkyl chain length as indicated by increase in increase in solubilization of n-heptane with innumber of carbon atoms in the n-alkyl alcohols crease in chain length of added alcohol and with - the observation that increase in chain length of and (2) type of polar group for -SH >-"*-OH> COO- for equal chain lengths in the effectiveness the soap will increase the size of the micelle'' and of these alkyl derivatives as additives. will increase the solubilization of a hydrocarbon.8 Coupled with this decrease in CMC upon addiTABLE I1 tion of long-chain alcohols is the fact that there is EFFECTOF CHANGE IN POLARGROUPOF ADDITIVE UPON an increase in the size of the micelle. Thus, SOLUBILIZATION OF %-HEPTANE IN 0.35 M POTASSIUM based on a modified long spacing X-ray band, TETRADECANOATE SOLUTIONS Mattoon has calculated that, in a 0.57 M KCI4 Moles solution, there are about 75 soap molecules per No. n-heptane Moles solubilKC14 + micelle and in this solution saturated with nized/ Sol. in Kc14 addiKO!MoleMoles Mole- tive Moles dodecanol the number increased to 1S0.6 This is additive cules additive/ cules mole(addiabout 2.4 times the original value compared to an 1000 g. additive Sol. in tive + n-CTHls/ cules/ KCl4/ micelle micelle micelle KCic solo. soln. KCM soln. KCiS increase of 1.8-1.9 times as is indicated by the n-Octanol solubilization data in Table I in which a value of 66 soap molecules per micelle is taken as a unit. 0 0.116 1.0 1.0 22 66 66 The series -COO-
