Multiple Time-Step Dual-Hamiltonian Hybrid Molecular Dynamics - Monte Carlo Canonical Propagation Algorithm

Area of Science:

• Computational Chemistry
• Statistical Mechanics
• Physical Chemistry

Background:

• Efficiently sampling systems in the canonical ensemble is crucial for molecular simulations.
• Traditional molecular dynamics (MD) and Monte Carlo (MC) methods can be computationally expensive.
• Multiple time-step integrators aim to accelerate simulations by using different time steps for different parts of the system.

Purpose of the Study:

• To propose a novel multiple time-step integrator for canonical ensemble sampling.
• To introduce a hybrid molecular dynamics-Monte Carlo (MD-MC) approach within the Dual Hamiltonian Multiple Time-Step (DHMTS) framework.
• To ensure the accuracy and consistency of simulations with the Boltzmann distribution.

Main Methods:

• Development of the Dual Hamiltonian Multiple Time-Step (DHMTS) algorithm, utilizing two similar Hamiltonians with differing computational costs.
• Integration of a hybrid MD-MC method with a Metropolis acceptance criterion to enforce detailed balance.
• Reformulation of nonlinear differential equations into a recursive root finding problem, analogous to RESPA.

Main Results:

• The DHMTS algorithm effectively preconditions nonlinear differential equations for dynamics.
• The hybrid MD-MC version enforces detailed balance and suppresses discretization errors.
• Illustrative tests demonstrate the method's effectiveness in sampling canonical ensemble systems.

Conclusions:

• The proposed DHMTS algorithm offers an efficient and accurate method for canonical ensemble simulations.
• The hybrid MD-MC approach successfully addresses discretization errors and ensures thermodynamic consistency.
• This method provides a valuable tool for advancing molecular simulation capabilities.