Optimized effective potential method and application to static RPA correlation

TL;DR

This paper advances the optimized effective potential (OEP) method by introducing acceleration techniques, employing the KLI approximation without loss of accuracy, and reformulating it for direct RPA correlation functionals, enhancing its applicability and efficiency.

Contribution

It presents a novel acceleration scheme for OEP, applies the KLI approximation effectively, and reformulates OEP for direct RPA correlation functionals, addressing computational and theoretical challenges.

Findings

KLI approximation maintains accuracy for magnetic transition metals.

Accelerated OEP scheme reduces computational cost.

Reformulated OEP applicable to direct RPA correlation functionals.

Abstract

The optimized effective potential (OEP) method is a promising technique for calculating the ground state properties of a system within the density functional theory. However, it is not widely used as its computational cost is rather high and, also, some ambiguity remains in the theoretical framework. In order to overcome these problems, we first introduced a method that accelerates the OEP scheme in a static RPA-level correlation functional. Second, the Krieger-Li-Iafrate (KLI) approximation is exploited to solve the OEP equation. Although seemingly too crude, this approximation did not reduce the accuracy of the description of the magnetic transition metals (Fe, Co, and Ni) examined here, the magnetic properties of which are rather sensitive to correlation effects. Finally, we reformulated the OEP method to render it applicable to the direct RPA correlation functional and other, more precise, functionals. Emphasis is placed on the following three points of the discussion: i) Level-crossing at the Fermi surface is taken into account; ii) eigenvalue variations in a Kohn-Sham functional are correctly treated; and iii) the resultant OEP equation is different from those reported to date.