Towards high-throughput many-body perturbation theory: efficient algorithms and automated workflows

TL;DR

This paper introduces efficient algorithms and automated workflows to extend high-throughput computational screening from DFT to many-body perturbation theory, enabling faster and more accurate materials property predictions.

Contribution

It presents novel algorithms and implementations for automating MBPT calculations, including a robust convergence procedure for GW and BSE, and an automatic GW band interpolation scheme.

Findings

Validated on semiconductor and metallic systems

Achieved reduced computational cost with maintained accuracy

Enabled high-throughput MBPT simulations

Abstract

The automation of ab initio simulations is essential in view of performing high-throughput (HT) computational screenings oriented to the discovery of novel materials with desired physical properties. In this work, we propose algorithms and implementations that are relevant to extend this approach beyond density functional theory (DFT), in order to automate many-body perturbation theory (MBPT) calculations. Notably, a novel algorithm pursuing the goal of an efficient and robust convergence procedure for GW and BSE simulations is provided, together with its implementation in a fully automated framework. This is accompanied by an automatic GW band interpolation scheme based on maximally-localized Wannier functions, aiming at a reduction of the computational burden of quasiparticle band structures while preserving high accuracy. The proposed developments are validated on a set of representative semiconductor and metallic systems.