The second-order rate constant /capp for the reaction between (C2H5)3N and CH3I has been determined at 30.0 °C in cyclohexane and in mixtures of cyclohexane with variable amounts (up to ca. 0.3 M) of the following cosolvents (S); 7V,/V-dimethylacetamide, cyclohexanecarbonitrile, cyclohexyl methyl ketone, butanone, tetrahydrofuran, ethyl acetate, cyclohexyl chloride, methyl cyclohexanecarboxylate, cyclohexyl methyl ether, dibutyl ether, nitrobenzene, 7V,(V-dimethylbenzamide, benzonitrile, acetophenone, benzophenone, methyl benzoate, diphenylmethane, anisole, chlorobenzene, fluorobenzene, 1,2-diphenylethane, triphenylmethane, benzene, 1,4-dichlorobenzene, toluene, and p-xylene. Within the limits of experimental error, the following have been found: (1) The activity coefficients of the reagents are practically unaffected by the added cosolvent. (2) For low cosolvent concentrations, &app is related to the concentration of S by the equation km = kQ + kc[S], where k0 is the reactioh rate in pure cyclohexane. (3) In all cases, the catalytic effect of S is quite large: it amounts to ca. 50% of the lowering of the activation free energy of the reaction, observed on going from pure cyclohexane to pure S. (4) This major effect is largely underestimated by the Onsager-Kirkwood model. (5) The catalytic efficiency of aromatic cosolvents of low or zero dipolarity is over 1 order of magnitude larger than predicted by this model. This is traced to electrostatic interactions involving higher multipoles and to London forces. (6) Although a fraction of the catalytic effect can be attributed to a "general dielectric" contribution, these and other results strongly suggest that this reaction can proceed through a true termolecular channel.
