Saddle point model for atom transfer reactions in solution

Abstract

A simple classical model for a collinear atom transfer (exchange) reaction in solution is investigated. The reaction is analyzed in terms of reactive and nonreactive modes in a saddle point region. Dynamical solvent effects are treated at the generalized Langevin equation level. The reaction rate constant k is determined and compared to the transition state theory prediction kTST. It is found that, for typical reactions governed by sharp potential barriers, the solvent is not very effective in the saddle region in reducing k much below kTST. In addition, solvent-induced coupling of the reactive mode to nonreactive modes of motion typically has only a small influence on k. Some limitations of the model are discussed.