A Similarity Renormalization Group Approach to Green's Function Methods.

Area of Science:

• Computational physics and chemistry
• Electronic structure theory
• Quantum many-body methods

Background:

• Green's function methods, particularly the GW approximation, are popular for electronic structure calculations due to their accuracy and cost-effectiveness in weakly correlated systems.
• Self-consistent GW (scGW) calculations face convergence challenges, recently linked to the intruder-state problem.
• Existing methods struggle with efficient and stable convergence in complex electronic systems.

Purpose of the Study:

• To address the convergence issues in self-consistent GW (scGW) calculations.
• To develop a more stable and accurate self-energy formulation for quasiparticle self-consistent GW (qsGW) methods.
• To improve the efficiency and applicability of GW-based electronic structure calculations.

Main Methods:

• A perturbative analysis of the similarity renormalization group (SRG) approach was applied to Green's function methods.
• The SRG formalism was used to derive a static, Hermitian self-energy expression from first principles.
• This SRG-derived self-energy was implemented within quasiparticle self-consistent GW (qsGW) calculations.

Main Results:

• The SRG-based regularized self-energy significantly accelerates the convergence of qsGW calculations.
• A slight improvement in the overall accuracy of the electronic structure predictions was observed.
• The new method demonstrates straightforward implementation within existing computational codes.

Conclusions:

• The SRG approach provides an effective solution to the convergence challenges in scGW methods.
• The derived static and Hermitian self-energy offers a robust alternative for qsGW calculations.
• This work paves the way for more efficient and accurate electronic structure studies using GW methods.