ITR: Perturbation Theoretic Approach to the Electron Correlation Cusp Problem

Suny At Buffalo, Amherst NY

Investigators

Abstract

The goal of this research is the production of an improved computer code for molecular electronic structure computation that permits the practical computation of the energy of a molecule with an accuracy at least ten times better than that currently possible. The project will apply this code to the computation of energy barriers and potential energy surfaces for elementary gas phase chemical reactions. Once it becomes aailable to the general scientific community, the code will find application to a wide range of other problems in chemistry and materials science. The computed potential energy surfaces will have immediate application for reactive scattering computations. In this way the project contributes significantly to the infrastructure that will eventually enable the computation of chemical rate constants from first principles. These constants are widely applicable. For example, they are essential input for computer models of atmospheric chemistry and combustion. The new feature of this method is the use of Rayleigh-Schroedinger perturbation theory to treat the electron correlation cusp problem. Existing highly-developed orbital-based methods are well suited to description of the global behavior of the wave function. However, they are not well suited to the description of the cusp, a secondary feature that eventually limits the accuracy of large scale electronic structure computations. The new method solves the first-order wave function by use of explicitly correlated geminal basis functions.