NOTES pressure of up to about 40% will result from the increase in mass in labeling dodecanoic acid. Using the same method, it can be predicted that the corresponding increase in the heat of sublimation would be too small to be meauured by these methods.
2245 ciation constants we observed in these systems is presumably to be identified with the presence of solventsolvent and ion-solvent interactions. To clarify the former type, we therefore made a parallel study of the system MeOH-MeCN with the use of nmr spectrosCOPY.
Experimental Section Ion-Pair Association of Strong Electrolytes in Binary Mixtures of Polar Solvents : Alkaline Perchlorates in Methanol-Methyl Cyanide Mixtures
by Filippo Conti and Gianfranco Pistoia Istituto d i Chimica Fisica, Universitd d i Roma, Rome, Italy Accepted and Transmitted by The Faraday Society (October 11, 1067)
Many authors have studied the conductometric behavior of strong electrolytes in binary mixtures of solvents in order to show the effect of the ion-solvent interaction on the value of ion-pair association constants (&). While a thorough study has been made of mixtures of solvents with rather different dielectric constants (D), there is not sufficient information on the behavior of the K A in systems with a small variation of D. From measurements made on tetrabutylammonium bromide in methanol-nitrobenzene mixtures (D = 32.634.9), Sadelr and Fuossl observed a discontinuity in the curve of dissociation constants as a function of the relative concentration of the two solvents. I n a subsequent paper, Kay2 pointed out the unreliability of the values reported by these authors, both in the data and in the method used for calculating KA. More recently, Coplan and FUOSS,* studying some salts of triisoamyln-butylammonium in MeOH-MeCN (D = 36.0-32.6) , observed the presence of a maximum in the values of the Stokes hydrodynamic radii for the picrate, iodide, and tetraphenylboride anions. They connected this behavior with the strong interaction between the two solvents, confirmed both by viscosimetric measurements and by ir measurements, and they proposed an explanation based on the formation of less associated forms particularly active in the solvation of the ions. For these salts too, the data for K A as a function of the two solvents are not very reliable, since they have very low values. For this purpose, we undertook a systematic study in our laboratory of virtually isodielectric solvent mixtures, using salts which showed sufficiently high values of KA,namely, Kc104 and CsC104 in binary mixtures of MeOH-MeCN. The singular behavior of the variation of the asso-
MeCN and MeOH (C. Erba R P products) were purified by the Brown and Fuoss4 and Hartley and Raikes6 methods, respectively. The KC1o4 was a Fisher guaranteed product. The CsC104 was prepared by a double-exchange reaction from CsC1 and 60% HClO4 (both BDH Analar products). The two salts, crystallized twice by conductivity water, were dried in a high vacuum apparatus at 200 and 230", respectively. The solvent mixtures were prepared by weighing directly into the conductivity cells, care being taken to maintain an anhydrous nitrogen atmosphere. The electrical apparatus, thermostatic bath, and conductivity cell have all been described in previous papersg6 The nmr measurements were made with a Varian A-60 spectrometer, using Wilmad coaxial cells, with benzene as an external reference. The measurements were carried out in duplicate a t a temperature of 29 f 1" and were corrected for the varying susceptibilities of the solutions.
Results
Conductivity. The physical constants of the mixtures (viscosity ( T ) , density (d),and dielectric constant (D)), determined by techniques used in this laboratory,' are reported in Table I. The values obtained are in comTable I : Properties of Acetonitrile-Methanol Mixtures NO.
1, 9 2 3 4 5 6
7 8, 13 10 11 12
70 of CHsCN
D
'I
d
0.00 5.85 19.87 40.82 59.31 62.79 81.40 100.00 39.35 62.14 74.44
32.63 32.80 33.30 34.00 34.42 34.55 35.35 36.01 33.95 34.50 35.13
0.00544 0.00513 0.00450 0.00390 0.00352 0.00347 0.00333 0.00345 0.00394 0.00348 0.00336
0.78665 0.78690 0.78685 0,78535 0.78305 0.78258 0.77970 0.77679 0.78545 0,78270 0.78080
(1) H.Sadek and R. Fuoss, J. Amer. Chem. SOC.,72, 301 (1950). (2) R. Kay and F. Evans, J . Phys. Chem., 68,2748 (1964). (3) M.Coplan and R. Fuoss, ibid., 68, 1181 (1964). (4) A. Brown and R. Fuoss, ibid., 64, 1341 (1960). (5) H. Hartley and H. Raikes, J. Chem. SOC.,624 (1925). (6) F. Accascina, A. D'Aprano, and R. Fuoss, J . Amer. Chem. SOC., 81, 1068 (1959). (7) F. Accascina, S. Petrucci, and S. Schiavo, Sei. Technol., 2, 27 (1958).
Volume 7.9, Number 6 June 1968
2246
NOTES
plete agreement with those of Coplan and F u o ~ s . ~ The calculations were made with an IBM 7040 comTable I1 gives the concentrations c (equiv/l.), and the puter, using the programs worked out by I