Calculations of 5dN-16s systems using orthogonal operators: do orthogonal operators survive configuration interaction?

Abstract

The merits and pitfalls of using orthogonal operators in systems with strong configuration interaction are illustrated by a study of a number of (5d+6s)N configurations (N=8, 9) ranging from Ir to Pb. Both the perturbational and the full diagonalization approach are applied. Criteria are given to indicate the regions where either operators based on perturbation theory or explicit configuration interactions are dominant. In both regions, orthogonal operators describe the (ground and) excited states well, provided they are extended by interconfiguration operators in the case of lower ionization, where full diagonalization of the three (5d+6s)N configurations is required. Ab initio calculations using second order perturbation theory show satisfactory agreement with the results from fits to experimental energy levels, except again for the lowest ionizations, where perturbation theory breaks down. It turns out to be a good approximation to exclude interactions treated by complete diagonalization from the perturbation expressions; both approaches can then be combined to describe excited state interactions.