To main content

Automated workflow for accurate high-throughput GW calculations using plane waves

Abstract

The GW approximation represents the state-of-the-art ab-initio method for computing excited-state properties. Its execution requires control over a larger number of parameters, and therefore, its application in high-throughput studies is hindered by the complex and time-consuming convergence process across a multidimensional parameter space. To address these challenges, we develop a fully-automated open-source workflow for G0W0 calculations within the AiiDA framework and the projector augmented wave (PAW) method. The workflow is based on an efficient estimation of the errors in the quasi-particle (QP) energies due to basis-set truncation and ultra-soft PAW potentials norm violation, which allows a reduction in the dimensionality of the parameter space and avoids the need for multidimensional convergence searches. Protocol validation is conducted through a systematic comparison against established experimental and state-of-the-art GW data. To demonstrate the effectiveness of the approach, we construct a database of QP energies for a dataset of over 320 bulk structures.

Category

Academic article

Language

English

Author(s)

  • Lorenzo Varrassi
  • Florian Ellinger
  • Espen Flage-Larsen
  • Michael Wolloch
  • Georg Kresse
  • Nicola Marzari
  • Cesare Franchini

Affiliation

  • SINTEF Industry / Sustainable Energy Technology
  • University of Bologna
  • Swiss Federal Institute of Technology of Lausanne
  • University of Vienna
  • University of Oslo

Year

2025

Published in

npj Computational Materials

Volume

11

Page(s)

1 - 14

View this publication at Norwegian Research Information Repository