Temperature and its control in molecular dynamics simulations

TL;DR

This paper reviews recent deterministic algorithms for controlling temperature in molecular dynamics simulations, emphasizing their physical basis, advantages, disadvantages, and numerical integration methods.

Contribution

It provides a comprehensive overview of recent developments in thermostatted dynamics, highlighting the physical principles and numerical techniques involved.

Findings

Different thermostat algorithms have unique advantages and limitations.

Numerical methods for integrating thermostatted equations are crucial for simulation accuracy.

Open questions remain in the development of thermostatted molecular dynamics.

Abstract

The earliest molecular dynamics simulations relied on solving the Newtonian or equivalently the Hamiltonian equations of motion for a system. While pedagogically very important as the total energy is preserved in these simulations, they lack any relationship with real-life experiments, as most of these tests are performed in a constant temperature environment that allows energy exchanges. So, within the framework of molecular dynamics, the Newtonian evolution equations need to be modified to enable energy exchange between the system and the surroundings. The prime motive behind allowing energy exchange is to control the temperature of the system. Depending on the temperature being controlled and the modifications made to the equations of motion, different evolution equations, or thermostat algorithms, can be obtained. This work reviews the recent developments in controlling temperature through deterministic algorithms. We highlight the physical basis behind the algorithms, their advantages, and disadvantages, along with the numerical methods to integrate the equations of motion. The review ends with a brief discussion on open-ended questions related to thermostatted dynamics.