Mixed Stochastic-Deterministic Approach for Many-Body Perturbation Theory Calculations

Area of Science:

• Computational Physics
• Materials Science
• Quantum Chemistry

Background:

• Accurate calculation of quasiparticle energies is crucial for understanding material properties.
• Traditional GW methods face computational challenges with large systems.
• Approximating high-energy contributions is key to improving efficiency.

Purpose of the Study:

• To present a novel GW approach for quasiparticle energy calculations.
• To achieve quasiquadratic scaling for improved computational efficiency.
• To enable accurate electronic structure analysis of complex materials.

Main Methods:

• Approximating high-energy contributions to the Green's function using effective stochastic vectors.
• Utilizing the Lehmann representation for Green's function analysis.
• Implementing the method without altering existing GW codes.

Main Results:

• Achieved quasiquadratic scaling for GW calculations of quasiparticle energies.
• Demonstrated rapid convergence with stochastic parameters.
• Successfully applied the method to a twisted MoS2 bilayer.

Conclusions:

• The developed method offers an efficient and accurate way to compute quasiparticle energies.
• It is versatile, applicable to systems of various dimensionalities and screening responses.
• Large-scale GW methods can simultaneously account for geometry relaxations and electronic correlations in complex materials.