Stochastic GW Calculations for Molecules.

Area of Science:

• Computational physics and chemistry
• Quantum mechanics
• Materials science

Background:

• Quasiparticle (QP) excitations are crucial for understanding charge transfer in various systems.
• Standard density functional theory (DFT) methods do not reliably predict QP energies.
• Many-body perturbation techniques, like the GW approximation, are essential but computationally expensive for large systems.

Purpose of the Study:

• To demonstrate the accuracy of the recently developed stochastic GW (sGW) method.
• To validate sGW for calculating vertical ionization energies.
• To assess the performance of sGW against established deterministic GW calculations.

Main Methods:

• Implementation of a stochastic formulation of the GW approximation (sGW).
• Calculation of vertical ionization energies for a set of 10 small molecules.
• Comparison of sGW results with benchmark deterministic GW calculations.

Main Results:

• The sGW approach exhibits near-linear-scaling computational complexity.
• sGW accurately predicts vertical ionization energies for small molecules.
• Mean absolute deviations of 0.05 and 0.09 eV were observed compared to benchmark GW results.

Conclusions:

• The stochastic GW (sGW) method provides a computationally efficient and accurate approach for many-body calculations.
• sGW enables accurate QP energy calculations for very large systems, overcoming limitations of traditional GW methods.
• This advance opens possibilities for studying charge transport in complex molecular and extended systems.