ION-PAIR INTERMEDIATES IN TERTIARY ARALKYL DERIVATIVES

The purpose of this research was to seek evidence for the ion-pair mechanism in tertiary substrates with no S(,N)2\u27 or elimination option available. Using techniques described by Sneen and Larsen*, the existence of ion-pair intermediates has been established in reactions of 1-bromo-1,1-diphenyl-2,2,2-trifluoroethane (DPTCBr) in the presence of added azide ion in 70% and 60% aqueous acetone. Furthermore, unlike the unsubstituted compound, the p-Me derivative (p-Me-DPTCBr) showed no rate enhancement when solvolyzed in 60% aqueous acetone with added sodium azide; this compound thus fits kinetically the S(,N)1 criterion. Application of the Hammett relationship to this system in 60% acetone resulted in a (rho)(\u27+) value of -5.18. Substitution of the CF(,3) group in DPTCBr directly at the reactive site produced a rate retardation of 2.6 x 10(\u276) compared with the corresponding hydrogen substituted compound. This factor conforms well with the value of 2.3 x 10(\u276) determined in poorly nucleophilic solvents by Koshy and Tidwell.** The competition factor m = k(,N)/k(,s), a measure of the substrate\u27s ability to distinguish between different nucleophiles, is the slope of a plot of the product ratio, {RN}/{ROS} vs. {N}. The ion-pair mechanism leads to the prediction (confirmed for DPTCBr) that m should decrease with an increase in solvent polarity. This decrease in m is not due to diminished nucleophilicity of azide ion in 60% acetone compared with 70% indicated by the fact that the bimolecular component of the observed reaction rate is actually 2.4-fold greater in the better ionizing solvent. Application of the Grunwald-Winstein equation to this rate increase led to an m(,G) value of 0.57, or ca. 70% of that of the solvolysis process. Solvolysis of a series of four diaryltrifluoromethylcarbinyl tosylates in 80% acetone, which varied in reactivity (k(,t)) by a factor of 2,200, were found to differ in selectivity (m) by only 3.8. We believe this is evidence that the immediate precursor of product for these substrates was of a common type (presumably ion pair) rather than of distinct mechanistic types (e.g., carbonium ion for S(,N)1 or covalent substrate for S(,N)2). Beyond a reasonable doubt the p-H tosylate was shown to be essentially zero-order while the m,m-(CF(,3))(,2) derivative was shown to be first-order in added azide ion. Thus we feel we have found a series of compounds spanning the traditional mechanistic range (apparent S(,N)2 (---\u3e) borderline (---\u3e) apparent S(,N)1) yet displaying essentially constant selectivity toward added nucleophile and solvent water. The mechanistic consequences of some stereochemical studies are also discussed. Dr. G. R. Felt*** found that solvolysis of 1-(m-chlorophenyl)-1-((beta)-naphthyl)-2,2,2-trifluoroethyl tosylate in 80% acetone containing 0.100 M NaN(,3) produced solvolysis product, ROH, of excess retained configuration, and azide substitution product, N(,3)R, with a sign of rotation opposite to that of the starting tosylate. It is assumed that the sign change in the azide product corresponds to inversion of configuration. Arguments are presented that the substitution product derives from attack at the intimate ion-pair stage while the solvolysis product derives from attack at both the intimate and solvent-separated stages. Both the stereochemical and kinetic data appear consistent with such a proposal. *R. A. Sneen and J. W. Larsen, J. Am. Chem. Soc., 91, 362 (1969). **K. M. Koshy and T. T. Tidwell, J. Am. Chem. Soc., 102, 1216 (1980). ***Postdoctoral research conducted at Purdue University, 1971-1973