Towards fully automatized GW band structure calculations: What we can learn from 60.000 self-energy evaluations

TL;DR

This study analyzes a large dataset of 2D material GW band structures to identify error sources, evaluate approximation reliability, and propose methods for automating high-throughput G0W0 calculations.

Contribution

It provides a comprehensive analysis of 370 GW calculations, assesses approximation errors, and explores basis set extrapolation and scissors operator validity, advancing automated high-throughput workflows.

Findings

Linear QP approximation introduces significant errors.

Basis set extrapolation via 1/N_PW is reliable.

Scissors operator is generally invalid for typical error tolerances.

Abstract

We analyze a data set comprising 370 GW band structures composed of 61716 quasiparticle (QP) energies of two-dimensional (2D) materials spanning 14 crystal structures and 52 elements. The data results from PAW plane wave based one-shot G$_0$W$_0$@PBE calculations with full frequency integration. We investigate the distribution of key quantities like the QP self-energy corrections and renormalization factor $Z$ and explore their dependence on chemical composition and magnetic state. The linear QP approximation is identified as a significant error source and propose schemes for controlling and drastically reducing this error at low computational cost. We analyze the reliability of the $1/N_\text{PW}$ basis set extrapolation and find that is well-founded with narrow distributions of $r^2$ peaked very close to 1. Finally, we explore the validity of the scissors operator approximation concluding that it is generally not valid for reasonable error tolerances. Our work represents a step towards the development of automatized workflows for high-throughput G$_0$W$_0$ band structure calculations for solids.