A tiered approach to Monte Carlo sampling with self-consistent field potentials.

Area of Science:

• Computational Chemistry
• Quantum Mechanics
• Statistical Mechanics

Background:

• Ab initio electronic structure methods like Hartree-Fock and density functional theory (DFT) are crucial for molecular simulations.
• Monte Carlo (MC) methods are widely used for sampling nuclear configurations in molecular systems.
• Accurate and efficient sampling is essential for calculating thermal properties.

Purpose of the Study:

• To develop a novel, computationally efficient tiered approach for Monte Carlo sampling of nuclear configurations.
• To integrate this method with ab initio self-consistent field (SCF)-based potentials.
• To improve the efficiency of molecular simulations without compromising accuracy.

Main Methods:

• A tiered Monte Carlo sampling strategy is implemented, testing individual self-consistent field (SCF) cycle energies.
• The method avoids approximations and ensures detailed balance is obeyed.
• Modifications are included for challenging cases with poor initial SCF guesses.

Main Results:

• The tiered MC approach demonstrates proper detailed balance.
• A factor-of-two reduction in SCF cycles was achieved compared to standard methods.
• The method proved competitive with accelerated molecular dynamics techniques, even without using forces.

Conclusions:

• The proposed tiered MC sampling is an effective and efficient method for ab initio calculations.
• This technique offers a significant computational advantage for molecular simulations.
• It provides a viable alternative for sampling nuclear configurations in quantum chemistry.