The bandstructure of gold from many-body perturbation theory

TL;DR

This paper uses many-body perturbation theory to calculate gold's bandstructure, analyzing various GW approximations and comparing results with hybrid functional methods, achieving closer agreement with experimental data.

Contribution

It provides a detailed GW-based analysis of gold's electronic structure, including quasiparticle self-consistency effects and comparison with hybrid functional approaches.

Findings

G0W0 shifts unoccupied bands by ~0.2 eV

Self-consistency lowers 5d bands by 0.35 eV

Interband gap increases to 0.35-0.75 eV, aligning with experiments

Abstract

The bandstructure of gold is calculated using many-body perturbation theory (MBPT). Different approximations within the GW approach are considered. Standard single shot G0W0 corrections shift the unoccupied bands up by ~0.2 eV and the first sp-like occupied band down by ~0.4 eV, while leaving unchanged the 5d occupied bands. Beyond G0W0, quasiparticle self-consistency on the wavefunctions lowers the occupied 5d bands by 0.35 eV. Globally, many-body effects achieve an opening of the interband gap (5d-6sp gap) of 0.35 to 0.75 eV approaching the experimental results. Finally, the quasiparticle bandstructure is compared to the one obtained by the widely used HSE (Heyd, Scuseria, and Ernzerhof) hybrid functional.