Efficient full frequency GW for metals using a multipole approach for the dielectric screening

TL;DR

This paper introduces an efficient multipole approximation method for full frequency GW calculations in metals, accurately capturing low-energy excitations and improving computational efficiency over traditional approaches.

Contribution

The paper extends the multipole approximation to metallic systems, optimizing frequency sampling and including intra-band contributions, enabling accurate and efficient GW calculations for metals.

Findings

MPA accurately reproduces full frequency GW results for metals.

The method reduces computational cost compared to traditional full frequency approaches.

Good agreement with experimental electron energy loss spectra for Cu.

Abstract

The properties of metallic systems with important and structured excitations at low energies, such as Cu, are challenging to describe with simple models like the plasmon pole approximation (PPA), and more accurate and sometimes prohibitive full frequency approaches are usually required. In this paper we propose a numerical approach to $GW$ calculations on metals that takes into account the frequency dependence of the screening via the multipole approximation (MPA), an accurate and efficient alternative to current full-frequency methods that was recently developed and validated for semiconductors and overcomes several limitations of PPA. We now demonstrate that MPA can be successfully extended to metallic systems by optimizing the frequency sampling for this class of materials and introducing a simple method to include the $\mathbf{q}\to 0$ limit of the intra-band contributions. The good agreement between MPA and full frequency results for the calculations of quasi-particle energies, polarizability, self-energy and spectral functions in different metallic systems confirms the accuracy and computational efficiency of the method. Finally, we discuss the physical interpretation of the MPA poles through a comparison with experimental electron energy loss spectra for Cu.