Stochastic multi-reference perturbation theory with application to the linearized coupled cluster method.

Area of Science:

• Quantum Chemistry
• Computational Chemistry
• Theoretical Chemistry

Background:

• Accurate electronic structure calculations are crucial for understanding molecular properties.
• Treating both static and dynamic electron correlation remains a challenge in quantum chemistry.
• Multireference methods are needed for systems with complex electronic structures.

Purpose of the Study:

• To perform a stochastic evaluation of the multireference linearized coupled cluster (MLCC) theory.
• To develop a fully stochastic approach for calculating electronic energies, including static and dynamic correlation.
• To apply the stochastic MLCC method to benchmark chemical systems.

Main Methods:

• Stochastic sampling of zeroth-order and first-order wavefunctions using signed walkers.
• Augmentation of the Full Configuration Interaction Quantum Monte Carlo (FCIQMC) algorithm with a source term.
• Stochastic computation of second-order energy without additional overhead.
• Application to systems like the carbon dimer and aromatic molecules (benzene, m-xylylene).

Main Results:

• The stochastic MLCC method successfully samples wavefunctions and computes energies.
• The method accounts for both static correlation (large active spaces) and dynamic correlation (perturbation theory).
• Calculated singlet-triplet gaps for benzene and m-xylylene show good agreement with experimental values.
• The method provides accurate results for m-xylylene, a system challenging for other methods.

Conclusions:

• The developed fully stochastic MLCC theory is a promising approach for accurate electronic structure calculations.
• This method enables the simultaneous treatment of static and dynamic electron correlation.
• The accuracy demonstrated on benchmark systems suggests broad applicability in computational chemistry.