Accurate bare susceptibilities from full-potential $\textit{ab initio}$ calculations

TL;DR

This paper introduces a reliable and efficient method for calculating electronic susceptibilities using full-potential DFT, enabling accurate analysis of magnetic and electronic properties in complex materials.

Contribution

It provides a new implementation for computing static and dynamic bare susceptibilities from first-principles calculations, assessing common approximations and their impacts.

Findings

Validated the accuracy of susceptibility calculations for various materials.

Assessed the effects of matrix element approximations.

Analyzed the impact of energy band truncation.

Abstract

Electronic susceptibilities are a very popular tool to study electronic and magnetic properties of materials, both in experiment and theory. Unfortunately, the numerical evaluation of even the bare susceptibility, which depends on the computation of matrix elements and sums over energy bands, is very work-intensive and therefore various approximations have been introduced to speed up the calculations. We present a reliable and efficient implementation to compute static as well as dynamic bare susceptibilities based on full-potential density functional theory (DFT) calculations. Based on the exact results we will assess the accuracy of replacing the matrix elements with a constant and the impact of truncating the sum over the energy bands. Results will be given for representative and topical materials, such as Cr, a classical transition metal, as well as for FeSe and LaFeAsO, examples of iron-based superconductors.