Crystal-field splitting for low symmetry systems in ab initio calculations

TL;DR

This paper introduces a new ab initio method within the LDA+U framework to calculate crystal-field excitation energies in low-symmetry systems, accounting for electronic relaxation effects, with applications to titanates.

Contribution

It presents a direct calculation approach for crystal-field excitations that includes electronic relaxation, improving comparison with spectroscopic data in low-symmetry materials.

Findings

Relaxation reduces crystal-field splitting by ~30%.

Calculated excitation energies suggest orbital liquid formation is unlikely.

Results help explain isotropic magnetism in titanates.

Abstract

In the framework of the LDA+U approximation we propose the direct way of calculation of crystal-field excitation energy and apply it to La and Y titanates. The method developed can be useful for comparison with the results of spectroscopic measurements because it takes into account fast relaxations of electronic system. For titanates these relaxation processes reduce the value of crystal-field splitting by $\sim30%$ as compared with the difference of LDA one electron energies. However, the crystal-field excitation energy in these systems is still large enough to make an orbital liquid formation rather unlikely and experimentally observed isotropic magnetism remains unexplained.