Reaction Coordinates and the Transition-Vector Approximation to the IRC

Abstract

The appearance of a reaction profile or potential energy surface (PES) associated with the reaction path (defined as the path of steepest descent from the saddle point) depends on the choice of reaction coordinate onto which the intrinsic reaction coordinate is projected. This provides one with the freedom, but also the problem, of choosing the optimal perspective (i.e., the optimal reaction coordinate) for revealing what is essential for understanding the reaction. Here, we address this issue by analyzing a number of different reaction coordinates for the same set of model reactions, namely, prototypical oxidative addition reactions of C-X bonds to palladium. We show how different choices affect the appearance of the PES, and we discuss which qualities make a particular reaction coordinate most suitable for comparing and analyzing the reactions. Furthermore, we show how the transition vector (i.e., the normal mode associated with a negative force constant that leads from the saddle point to the steepest descent paths) can serve as a useful and computationally much more efficient approximation (designated TV-IRC) for full IRC computations, in the decisive region around the transition state.