Taming the Sign Problem in Auxiliary-Field Quantum Monte Carlo Using Accurate Wave Functions

Area of Science:

• Quantum Monte Carlo methods
• Computational quantum chemistry
• Electronic structure theory

Background:

• Accurate quantum chemical calculations are crucial for understanding molecular properties.
• Free projection auxiliary-field quantum Monte Carlo (fp-AFQMC) is a powerful method for electronic structure.
• Incorporating accurate trial wave functions can significantly improve fp-AFQMC performance.

Purpose of the Study:

• To explore and implement various accurate trial wave functions within the fp-AFQMC framework.
• To enhance the efficiency and applicability of fp-AFQMC for larger quantum chemical systems.
• To validate the improved fp-AFQMC approach using a benchmark system.

Main Methods:

• Integration of coupled-cluster singles and doubles, multi-Slater, and symmetry-projected mean-field wave functions into fp-AFQMC.
• Adaptation of a fast multi-Slater local energy evaluation algorithm for fp-AFQMC.
• Application of selected configuration interaction expansions for constructing trial wave functions.

Main Results:

• Demonstrated mitigation of the infamous sign problem in quantum Monte Carlo simulations.
• Observed accelerated convergence of fp-AFQMC calculations with improved trial wave functions.
• Enabled the application of fp-AFQMC to significantly larger and more complex chemical systems.
• Calculations on the Cu2O22+ system revealed substantial discrepancies with previously reported isomerization energies.

Conclusions:

• The use of accurate trial wave functions is a key strategy for advancing fp-AFQMC.
• The implemented methods allow for more reliable and efficient quantum chemical simulations.
• fp-AFQMC with advanced wave functions provides near-exact results for challenging chemical problems.