Embedding theory contributions to average atom models for warm dense matter

TL;DR

This paper enhances average atom models for warm dense matter by incorporating non-additive kinetic potentials from DFT embedding theories, improving the accuracy of electronic interaction modeling while maintaining computational efficiency.

Contribution

The paper introduces a novel approach to include non-additive kinetic potentials in average atom models, improving their treatment of electronic overlaps in warm dense matter.

Findings

Inclusion of $v^{ m nadd}$ affects electron densities and energy levels.

Model shows improved accuracy in mean ionization calculations.

Application to hydrogen demonstrates the method's effectiveness.

Abstract

Accurate modeling in the warm dense matter regime is a persistent challenge with the most detailed models such as quantum molecular dynamics and path integral Monte Carlo being immensely computationally expensive. Density functional theory (DFT)-based average atom models (AAM) offer significant speed-ups in calculation times while still retaining fair accuracy in evaluating equations of state, mean ionizations, and more. Despite their success, AAMs struggle to precisely account for electronic interactions -- in particular, they do not account for effects on the kinetic energy arising from overlaps in neighboring atom densities. We aim to enhance these models by including such interactions via the non-additive kinetic potential $v^{\rm nadd}$ as in DFT embedding theories. $v^{\rm nadd}$ can be computed using Thomas-Fermi, von Weizs\"acker, or more sophisticated kinetic energy functionals. The proposed model introduces $v^{\rm nadd}$ as a novel interaction term in existing ion-correlation models, which include interactions beyond the central atom. We have applied this model to hydrogen at 5 eV and densities ranging 0.008 to 0.8 g/cm$^3$, and investigated the effects of $v^{\rm nadd}$ on electron densities, Kohn-Sham energy level shifts, mean ionization, and total energies.