Efficient stochastic thermostatting of path integral molecular dynamics

Area of Science:

• Computational Chemistry
• Statistical Mechanics
• Quantum Mechanics

Background:

• Path Integral Molecular Dynamics (PIMD) is crucial for calculating quantum mechanical properties of condensed systems.
• PIMD faces challenges in efficiently sampling a wide range of system frequencies.
• Thermostatting is essential for managing these frequency ranges in PIMD.

Purpose of the Study:

• Introduce and evaluate novel stochastic thermostats for PIMD.
• Compare the performance of new thermostats against conventional methods.
• Assess the computational efficiency and sampling accuracy of the proposed thermostats.

Main Methods:

• Developed a stochastic Path Integral Langevin Equation (PILE) thermostat.
• Applied a Generalized Langevin Equation (GLE) thermostat with colored noise.
• Compared PILE and GLE with the Nosé-Hoover Chain (NHC) thermostat.
• Evaluated thermostats for liquid water and hydrogen-in-palladium systems.

Main Results:

• The PILE thermostat demonstrated comparable performance to NHC with improved computational efficiency.
• The GLE thermostat showed robust and near-optimum sampling efficiency across all tested systems.
• Both new thermostats effectively addressed the frequency sampling challenge in PIMD.

Conclusions:

• The PILE and GLE thermostats offer efficient and robust alternatives for PIMD simulations.
• These novel thermostats are expected to be valuable tools for future condensed phase quantum mechanical studies.
• The findings suggest wider applicability of these stochastic thermostats in computational chemistry.