Optimized expanded ensembles for simulations involving molecular insertions and deletions. I. Closed systems

Area of Science:

• Computational Chemistry
• Statistical Mechanics
• Molecular Simulation

Background:

• Monte Carlo (MC) simulations with molecule insertion-deletion are crucial for studying complex systems' thermophysical properties.
• Estimating chemical potentials in closed-system ensembles often suffers from poor ergodicity with standard MC methods.

Purpose of the Study:

• To develop and present efficient expanded ensemble methods for MC simulations.
• To overcome ergodicity limitations in molecular simulations involving gradual coupling/decoupling of molecules.
• To enhance the accuracy and efficiency of thermophysical property calculations.

Main Methods:

• Utilizing an arbitrary physical parameter (Lambda) to gradually couple/decouple partial molecules.
• Employing acceptance ratio methods for robust free-energy change estimation between Lambda states.
• Implementing non-Boltzmann sampling for Lambda states to achieve flat or optimized histograms.
• Developing an approach to select intermediate Lambda stages that maximize simulation round trips.

Main Results:

• Demonstrated efficient coupling/decoupling of molecules using expanded ensemble techniques.
• Successfully estimated free-energy changes and optimized sampling of parameter space.
• Validated the methods through simulations of hard sphere solvation and block copolymer mesophase formation.

Conclusions:

• The presented expanded ensemble methods significantly improve ergodicity in molecular simulations.
• These techniques provide a robust framework for accurate thermophysical property estimation.
• The approach is effective for complex systems, including solvation and polymer mixture studies.