Double Excitation Energies from Quantum Monte Carlo Using State-Specific Energy Optimization.

Area of Science:

• Quantum chemistry
• Computational physics
• Molecular modeling

Background:

• Accurate calculation of electronic excitations is crucial in chemistry and physics.
• Existing methods like coupled-cluster theory face challenges with complex molecular systems.
• Quantum Monte Carlo (QMC) methods have shown promise for single excitations.

Purpose of the Study:

• To evaluate the efficacy of recently developed quantum Monte Carlo methods for treating double excitations.
• To assess the performance of QMC for medium-sized molecules, including challenging cases.
• To compare QMC predictions with existing benchmarks and provide new predictions where data is scarce.

Main Methods:

• Utilizing fixed-node diffusion Monte Carlo (FN-DMC) methods.
• Applying these methods to calculate vertical transition energies for double excitations.
• Comparing results with high-level coupled-cluster calculations and experimental data where available.

Main Results:

• FN-DMC methods successfully treat double excitations, extending their applicability beyond single excitations.
• Calculated excitation energies show very good agreement with reliable benchmarks.
• Accurate predictions were obtained for systems that are difficult for traditional high-level computational chemistry methods.

Conclusions:

• Quantum Monte Carlo methods, specifically FN-DMC, are highly effective for calculating double excitation energies.
• These methods provide a reliable and accurate approach for studying electronic transitions in complex molecular systems.
• QMC offers a valuable tool for predicting excitation energies in systems lacking experimental or computational reference data.