Accurate energy bands calculated by the hybrid quasiparticle self-consistent GW method implemented in the ecalj package

TL;DR

This paper introduces a stable and accurate implementation of the hybrid quasiparticle self-consistent GW method in the ecalj package, effectively predicting energy band properties across various semiconductors and insulators.

Contribution

The paper presents a new, numerically stable version of the QSGW method in ecalj and demonstrates its effectiveness with a hybrid approach for accurate band structure calculations.

Findings

Hybrid QSGW with 80% QSGW and 20% LDA matches experimental data well.

The method accurately predicts band gaps and effective masses.

Implementation is user-friendly and reliable.

Abstract

We have recently implemented a new version of the quasiparticle self-consistent GW (QSGW) method in the ecalj package released at http://github.com/tkotani/ecalj. Since the new version of the ecalj is numerically stable and accurate compared to the previous versions, we can perform calculations easily without being bothered with setting input parameters. Here we examine its ability to describe energy band properties, e.g., band-gap energy, eigenvalues at special points and effective mass, for variety of semiconductors and insulators. We treat C, Si, Ge, Sn, SiC (in 2H, 3C, and 4H structures), (Al, Ga, N)x(N, P, As, Pb), (Zn, Cd, Mg)x(O, S, Se, Te), SiO2, HfO2, ZrO2, SrTiO3, PbS, PbTe, MnO, NiO, and HgO. We propose that a hybrid QSGW method, where we mix 80 percent of QSGW and 20 percent of LDA, gives universally good agreement with experiments for these materials.