Renormalized Singles Green's Function for Quasi-Particle Calculations beyond the G0W0 Approximation

Renormalized Singles Green's Function for Quasi-Particle Calculations beyond the G0W0 Approximation
复制标题

DOI:
10.1021/acs.jpclett.8b03337
复制
发表时间:
2019-02-07
影响因子:
5.7
通讯作者:
Yang, Weitao
Yang, Weitao
中科院分区:
化学2区
文献类型:
--
作者:
Jin, Ye;Su, Neil Qang;Yang, Weitao

文献摘要

被引文献

相似文献

准粒子能量和带隙特别是研究新材料的关键。常用的密度泛函近似(DFA)系统地低估了带隙,GW近似是选择良好的准确性和可靠性的既定方法。然而,G(0)W(0)对DFA具有一些不期望的依赖性,而自一致GW(scGW)是昂贵的并且在精度改进方面不一致。本文提出了一种简单有效的G(Rs)W(0)方法:利用DFA密度矩阵对Hartree-Fock(HF)哈密顿量进行子空间对角化,得到了新的参考绿色函数GRs,它包含了所有单激发对自能的贡献,从而基本上消除了起始点的依赖性。对分子和大带隙固体的计算表明,与G(0)W(0)相比,该方法有显著的改进,并且大大降低了对初始DFA的依赖性。GR(0)W(0)方法也改善了其他块的结果,但程度较低,这可能是由于当前块实现的限制。结果表明,要达到良好的精度,它是没有必要使用混合DFA,这是昂贵的散装。这项工作应该是非常重要的,使GW一个更强大的方法。
Quasi-particle energies and band gaps in particular are critical for investigating novel materials. Commonly used density functional approximations (DFAs) systematically underestimate band gaps, and GW approximation is the established method of choice for good accuracy and reliability. However, G(0)W(0) has some undesired dependence on the DFA, while self-consistent GW (scGW) is expensive and not consistent in accuracy improvement. Here a simple and efficient G(Rs)W(0) approach has been developed: a subspace diagonalization of the Hartree-Fock (HF) Hamiltonian with the DFA density matrix provides the new reference Green's function GRs that incorporates the effect of all single excitation contributions to the self-energy, thereby essentially eliminating the starting-point dependence. Calculations for molecules and large band gap solids demonstrate the significant improvement over G(0)W(0) and greatly reduced dependence on the initial DFA. GR(0)W(0) approach also improve results for other bulks over G(0)W(0), but to a lesser extent, which could be due to the limitations in current implementation for bulks. The results demonstrate that to achieve good accuracy, it is not necessary to use hybrid DFA, which is expensive for bulks. This work should be greatly significant in making GW a more robust approach.