2248
log KA 1.6
NOTES greater ion solvation since, in this solvent mixture range, there occur the greatest number of species of high solvation efficiency.
.
-
Conclusion From the considerations made, the effect observed in the K A values should be independent of the particular electrolyte used and should prove t o be of a completely general character for solvent mixtures with similar dielectric and solvation properties. For this purpose we have undertaken a systematic study of solvent mixtures of the types n/leOH-CsH6N02, MeOH-MeN02, and the like using various electrolytes.
1.0
0.5
280
2.70
2\90
.A
3.00
3.10
I””
D
Figure 2. Dependence of association constants on dielectric constant: 0, KClOd; 0, CsClO+
mixtures formed of two solvents with rather different values of D is not absolutely verified in this case. The curve for CsC104 shows a clear minimum for 1/D h.‘ 2.9 X whereas the curve for KC104 shows, for about the same value of 1/D, a discontinuity zone. This thoroughly singular behavior is probably connected with the reciprocal modification of solvent structures. On increasing the concentration of MeCN in MeOH, there occurs a depolymerization of the alcohol chains and a concurrent formation of mixed associates of the MeCN with the terminal hydroxyls and with those of the MeOH monomers (in ever increasing quantities because of the high D of the MeCK). Thus there occurs an increase in the forms capable of solvating the ions, with a respective decrease in the value of K A as compared with that observed in pure R4eOH. I n the same way, it is possible to explain the decrease of K A values which occur by adding MeOH to pure illIeCN. This solvent, which in the pure state is markedly a ~ s o c i a t e d , ~would * ~ ~ ~undergo a process of dissociation through the addition of MeOH, with increasing formation of species capable of solvating the ions. The solvating power of MeCN is clearly shown by the alkali perchlorate KA values in this solvent (Table IV). Therefore, the lower K A values in the would occur because of range of 1/D ‘v 2.9 X Table IV : K A Values for Alkaline Perchlorates in MeCN Perchlorate
KA
Li” Nab
4&2 5 f l 1714 23 4 5
K
cs
F. Accascina, G. Pistoia, and S. Schiavo, Ric. Sci., 36, 560 (1966). G. Pistoia, A. hf. Polcaro, and S. Schiavo, ibid., 37, 309 (1967). T h e Journal of Phyaical Chemistry
(18) E. Zhukova, Opt. i Spectroskopiya, 4, 750 (1968), (19) A. Saum, J. Polym. Sei., 42, 57 (1960).
The Solubility of Fluorocarbon Gases in Cyclohexane
by Keith W. Miller Department of Chemistry, University of California, Berkeley, California 94720 (Received December 7 , 1967)
This study is a sequel to a recent one by Hildebrandl on the solubility of gases in normal liquids based largely on accurate measurements of the solubility of gases in cyclohexane made by Dymond.2 Hildebrand correlated gas solubilities by using as governing parameters the energies of vaporization of the gases at their boiling points in place of their uncertain “force constants” derived from gas imperfections.* The divergence of the point for CF4 in cyclohexane from a close correlation shown by many gases suggested the desirability of obtaining figures for other fluorocarbon gases. This writer here reports measurements of the solubility in cyclohexane of C2F6, C3F8,c - C ~ FCClF3, ~, and C3H8. Spectroquality cyclohexane, supplied by Matheson Coleman and Bell, was used without further purification. The CzFe, C3F8, c - C ~ F ~ and , CClF3 were supplied by Matheson Co. ; the C3Hs was research grade from Phillips Petroleum Co. The gases were dried before being passed into the apparatus. Solubilities were measured in the apparatus described by Dymond and Hildebrar~d.~ Observed results are given in Table I, together with figures for the entropy of solution of gas at 25” and 1 atm. When these values for entropy of solution are (1) J. H. Hildebrand, Proc. Nat. Acad. Sci. U.S., 57, 642 (1967). (2) J. H. Dymond, J . Phys. Chem., 71, 1829 (1967). (3) J. H. Hildebrand and R. L. Scott, “Regular Solutions,” PrenticeHall Inc., Englewood Cliffs, N. J., 1962. (4) J. H. Dymond and J. H. Hildebrand, I n d . Eng. Chem., Fundam. 6, 130 (1967); see also ref 2.
NOTES
2249
Table I : Solubility of Gases in c-CaH~za t 1 Atm Partial Pressure; xz 3 Mole Fraction; Entropy of Solution (cal deg-1 mole-1); Energy of Vaporization of the Gases at Their Boiling Points (kcal)
CzF,
t, "C 1042~
CaFe
t, "C
104xz c-C~FS
t , "C
104xa CClFa
t, "C 104x2
CaHs
t, "C 1042~
7.0 25.37
12.8 25.56
18.7 25.00
25.0 24.51
32.3 23.87
-2.2
3.47
7.0 72.15
12.8 67.86
18.7 64,07
25.0 60.71
33.2 57.11
-4.7
4.22
-9.0
5.00
7.1 291.5
8.8 282.1
16.0 248.6
25.0 215.7
34.0 190.2
7.0 122.9
12.8 113.8
18.7 106.2
25.0 99.6
28.5 95.5
32.1 92.7
-6.1
3.33
7.0 1318
12.8 1155
18.7 1016
25.0 896
28.5 833
32.1 776
-12.2
4.03
added to the plot of entropy vs. - R In x2, which is Figure 2 in the paper by Dymond,2the new points fall very slightly to the right of the line, like those for CFd and SFG, shown in that plot. The results shown in Table I are plotted in Figure 1 as - RT In x2 a t 25" vs. AEbV, including the data for
5
4
3 A E ~ 2
1
0 Figure 1.
other gases plotted in Figure 4 of the paper by Hildebrand.l The values of AEb" in the last column of the table were derived from heats of vaporization given in the Matheson Gas Data Book. The point for CF, is plotted from its solubility as measured by Archer and Hildebrand6and confirmed by Dymond;2 its heat of
vaporization from the American Institute of Physics Handbook yields AEbV= 2.72. We note, first, that the point for C3Hs falls on the line connecting the other alkanes, indicating that the new parameter allows for the difference between a normal and a cyclic alkane. Next, the points for the fluorocarbon gases fall on a new line strongly displaced in the direction of smaller solubility, in accord with the now well-known failure of geometrical means to evaluate correctly the potential energy between certain pairs of molecules of different species. I n Figure 1 there are three distinct lines, as well as a moderately divergent point for N2, all in the same solvent. This situation empha,sizes once again the fact that we will not have a fully satisfactory theory for solutions of nonelectrolytes until we have a valid theory for unlike pair-potentiah6 Thirdly, although SF6sublimes a t 1 atm, a value for AEbV can be obtained from its heat of vaporization extrapolated to a calculated boiling point, 205"K, by Klemm and Henkel' from its heat of fusion and their own measurements of sublimation pressures. This gives A&" = 4.08. Combining this with its solubility in cyclohexane, lO4x2 = 53.9 at 25", determined by Archer and Hildebrand,6 gives a point on the line for the four perfluorocarbon gases. Evidently the parameter 4.08 should serve well for predicting the solubility of SFs in other solvents. Finally, the point for CCIFa is strongly displaced to the left, as expected, since alkyl chlorides interact normally with most other substances. Acknowledgment. This research was supported by a grant from the National Science Foundation under a contract administered by J. H. Hildebrand. ( 5 ) G. Archer and J. H. Hildebrand, J . Phus. Chem., 67, 1830 (1963). (6) Cf.ref 3, p 164.
(7) W. K. Klemm and P. Henkel, 2.Anorg. Chem., 207, 73 (1932). Volume 78,Number 6 June 1968
