THE AVERAGE ENERGY APPROXIMATION FOR ELECTRONIC PERTURBATION $PROBLEMS^{*}$

Loading...
Thumbnail Image

Date

1963

Journal Title

Journal ISSN

Volume Title

Publisher

Ohio State University

Research Projects

Organizational Units

Journal Issue

Abstract

“Ordinary second order perturbation theory, when applied to an electronic perturbation ${P}$ [for convenience assume $(o|{P}|o) = o]$, yields an intractable sum over states $\Sigma_{i}’ < o |{P}|i > (E_{o} - E_{e})^{-1} < i |{P}| o >$. Expanding each denominator $E_{o}-E_{1}$, about an average energy $\epsilon_{av}$, leads to a power series in $(E_{o}-E_{i}-\epsilon_{nv})/\epsilon_{nv}$. The first term, after closure, is $<o|{P}^{2}|o>/\epsilon_{nv}$ which is just the average energy approximation. Setting the second term equal to zero determines an $\epsilon_{nv}$ which is identical to the one obtained by minimizing the energy with, respect to a trial wavefunction of the form $\Psi_{trial} = \Psi_{o} + \epsilon_{av}^{-1} {P} \Psi_{o}$. Certain ${P}_{s}$, however, possess singularities which would give rise to a non-allowed $\Psi_{trial}$ (actually, $\epsilon_{{ av}}^{-1}$ is forced to be zero). This suggests choosing different $\epsilon_{av}’s$ for different regions in space so that singularities can be ignored. A general method for doing this has been developed which leads, ultimately, to replacement of the sum over states by an integral over regions in spare. Furthermore, $\Psi_{trial}$ can be correctly scaled by a suitable perturbation procedure. Applications to electric and magnetic perturbation problems will be discussed.”

Description

$^{*}$This research was supported in part by a grant extended to the University of Washington by the Office of Naval Research. $^{\dag}$In absentia from Department of Chemistry, University of Washington, Seattle 5, Washington. $^{\dag\dag}$dLNational Science Foundation Cooperative Fellow 1961-63.
Author Institution: Department of Chemistry, University of California

Keywords

Citation