Quasiparticle self-consistent $GW$ method; a basis for the independent-particle approximation

TL;DR

The paper introduces quasiparticle self-consistent $GW$ (QS$GW$), a new self-consistent computational scheme that accurately predicts electronic band structures across various materials, overcoming limitations of previous methods.

Contribution

The paper presents the development and validation of QS$GW$, a novel self-consistent $GW$ approach that improves electronic structure calculations without many drawbacks of existing theories.

Findings

QS$GW$ accurately describes energy bands for diverse materials.

QS$GW$ outperforms local-density approximation in challenging cases.

The method converges reliably and provides a basis for total energy calculations.

Abstract

We have developed a new type of self-consistent scheme within the $GW$ approximation, which we call quasiparticle self-consistent $GW$ (QS$GW$). We have shown that QS$GW$ rather well describes energy bands for a wide-range of materials, including many where the local-density approximation fails. QS$GW$ contains physical effects found in other theories such as LDA$+U$, SIC and $GW$ in a satisfactory manner without many of their drawbacks (partitioning of itinerant and localized electrons, adjustable parameters, ambiguities in double-counting, etc.). We present some theoretical discussion concerning the formulation of QS$GW$, including a prescriptino for calculating the total energy. We also address several key methodological points needed for implementation. We then show convergence checks and some representative results in a variety of materials.