Efficient treatment of molecular excitations in the liquid phase environment via stochastic many-body theory

TL;DR

This paper introduces a stochastic GW method that efficiently predicts charge excitation energies of molecules in liquid environments, accurately capturing environmental effects with reduced computational cost.

Contribution

The authors develop a stochastic GW approach using localized states and sampling to accurately model molecular excitations in condensed phases, improving efficiency over traditional methods.

Findings

Results agree with high-level calculations and experimental data.

Environmental coupling accounts for about 40% of correlation energy.

Solvent effects cause up to 0.6 eV destabilization of QP energies.

Abstract

Accurate predictions of charge excitation energies of molecules in the disordered condensed phase are central to the chemical reactivity, stability, and optoelectronic properties of molecules and critically depend on the specific environment. Herein, we develop a stochastic GW method for calculating these charge excitation energies. The approach employs maximally localized electronic states to define the electronic subspace of a molecule and the rest of the system, both of which are randomly sampled. We test the method on three solute-solvent systems: phenol, thymine, and phenylalanine in water. The results are in excellent agreement with the previous high-level calculations and available experimental data. The stochastic calculations for supercells representing the solvated systems are inexpensive and require $\sim 1000$ CPU$\cdot$hrs. We find that the coupling with the environment accounts for $\sim40\%$ of the total correlation energy. The solvent-to-solute feedback mechanism incorporated in the molecular correlation term causes up to 0.6 eV destabilization of the QP energy. Simulated photo-emission spectra exhibit red shifts, state-degeneracy lifting, and lifetime shortening. Our method provides an efficient approach for an accurate study of excitations of large molecules in realistic condensed phase environments.