Calculation of Natural Orbitals and Wavefunctions by Perturbation Theory

Abstract

A computationally convenient formulation of perturbation theory is developed. In this formulation either the SCF wavefunction or the SCF function plus important corrections may be used as a trial function. Corrections to the trial function are expanded in powers of the residual error. The reduced density matrices may be expanded in a series in this error. If the SCF function is used as the trial function, parts of the density matrix are needed to second order in the error to determine the natural orbitals and geminals to zeroth order.