Temperature control in molecular dynamic simulations of non-equilibrium processes

Area of Science:

• Condensed matter physics
• Computational materials science
• Surface science

Background:

• Thermostats are crucial for simulating energy transfer in condensed matter systems.
• Equilibrium NVT ensemble methods are commonly used in molecular dynamics simulations.
• Non-equilibrium processes require careful consideration of thermostat applicability.

Purpose of the Study:

• To critically evaluate the suitability of equilibrium thermostats for non-equilibrium molecular dynamics simulations.
• To highlight the importance of the generalized Langevin equation (GLE) for accurate temperature control.
• To introduce and validate stochastic boundary conditions (SBC) as an efficient and physically appropriate thermostatting method.

Main Methods:

• Critical discussion of equilibrium thermostat approaches.
• Application of generalized Langevin equation (GLE) based methods, specifically stochastic boundary conditions (SBC).
• Comparative analysis of SBC and equilibrium thermostats in two distinct simulation scenarios.

Main Results:

• Equilibrium thermostat approaches are often inadequate for non-equilibrium condensed matter problems.
• Stochastic boundary conditions (SBC), derived from GLE, provide a computationally efficient and physically sound NVT thermostat.
• The choice of thermostat can lead to qualitatively different physical behavior in simulated systems.

Conclusions:

• Accurate temperature control in non-equilibrium molecular dynamics simulations necessitates methods based on the generalized Langevin equation (GLE).
• Stochastic boundary conditions (SBC) offer a practical and effective solution for thermostatting in solid-state and surface simulations.
• Equilibrium thermostats can misrepresent the physical behavior of systems undergoing non-equilibrium processes.