Evaluation of Nonlinear Ethanol-Trifluoroethanol Correlations for Mustard Chlorohydrin and Other Anchimerically Assisted Alkyl Substrates


In an effort to understand why mustard chlorohydrin (1) gives a nonlinear plot of log k versus YCIfor solvolysis in aqueous ethanol (EtOH) and aqueous trifluoroethanol (TFE), we have constructed similar plots for several analogues. Each of the new substrates has been demonstrated to react with neighboring-group participation and to be insensitive to solvent nucleophilicity. Despite these mechanistic similarities, both linear and nonlinear plots are observed. 2-(Methylthio)ethyl chloride (2) is similar in structure to 1, without the hydroxyl group, yet both 1 and 2 give nonlinear TFE-EtOH plots; therefore, the hydroxyl group in 1 is not responsible for the nonlinear TFE-EtOH plot. The other chlorides give essentially linear correlations, although the data point for the rate in pure TFE for 2-(phenylthio)ethyl tosylate falls significantly above the line and the TFE data give a negative slope. Substrate dimerization is shown not to be a factor in the nonlinear plot of 2. Also, there is no evidence to support an earlier suggestion that neighboring-group desolvation is a factor in the nonlinear plots for 1 and 2. The nonlinear correlation with substrates 1 and 2 may be attributed to a difference in the return of free ions or ion pairs when comparing the substrates and 1-adamantyl chloride, the model for YCI. Our previous suggestion that the TFE-EtOH correlation failure for 1 results from 1 and 1-adamantyl chloride having different sensitivities to electrophilicity is consistent with the explanation put forth here. Although we have dealt with correlation of substrates in aqueous TFE, aqueous acetone, and aqueous EtOH, these results are significant in relation to the general use of the Grunwald-Winstein equation for correlation of solvolytic rates.

Document Type




Publication Date


Journal Title

Journal of Organic Chemistry