Activity coefficient measurements of LiCI in dimethyl sulfoxide have been made by the emf method and the cryoscopic method. The emf measurements are in good agreement with similar emf measurements in the literature, but not with cryoscopic measurements in the literature. By applying an empirical correction for temperature gradients resulting from heat transfer in the steady state, agreement is obtained between our cryoscopic measurements and the emf measurements. The systematic errors from heat transfer phenomena in DMSO solutions appear to have been too large thus far for accurate cryoscopie measurements.The measurement of activity coefficients in nonaqueous solutions is of fundamental importance to a wide variety of investigations. These include the establishment of standard potentials, equilibrium constants, and other thermodynamic data; kinetic studies; and an understanding of ionic solvation and association phenomena. Solutions of lithium chloride in the dipolar aprotic solvent dimethyl sulfoxide (DMSO) have received more attention than any other similar nonaqueous system.To evaluate the results of our studies (1) of the kinetics of the lithium amalgam-lithium ion electrode in DMSO we required values of the activity coefficient of LiC1 over a wide range of concentrations. The values available in the literature at the time (2, 3) were not consistent with each other and did not cover a wide enough range of concentration. After we began our own measurements, two other studies were published (4,5). Discrepancies between the results of three different studies by the cryoscopic method (2, 3, 5) and by the emf method (4) are much larger than their precision and no critical comparison of all these various data has been made. The LiC1-DMSO system is the only electrolyte solution in an aprotic organic solvent for which sufficient data have been obtained to allow such a critical examination. This paper reports our measurements of the mean activity coefficient of LiC1 in DMSO by both the emf and cryoscopic methods and examines critically the results of other workers.
ExperimentalMaterials.--Solutions were prepared from anhydrous lithium chloride (Anderson Physics Laboratories, 99.999%) and dimethyl sulfoxide (Matheson, Coleman and Bell, "Spectroquality"). The solvent was dried using molecular sieves (Linde 5-A, washed with tripledistilled water, and dried for 24 hr at 375~ in a stream of dry argon) and filtered through a 10-20~ pore-size glass flit. Analysis was performed by gas chromatography using a column of Porapak Q (Waters Associates) and thermal conductivity detection (6). Water content of the solutions was less than 25 ppm (1.5 mM) and organic impurities were less than 10 ppm.Amalgams were prepared by weight from triple-distilled mercury (Doe and Ingalls), lithium metal (Foote Mineral Company, 99.96 % ), and thallium metal (American Smelting and Refining Company, 99.999%). All experiments were performed in a glove-box (Vacuum Atmospheres Corporation) using an atmosphere of argon which was recirculated through a ...