Molecular Simulation of Liquids: Efficient, Parallelizable Monte Carlo Algorithms

M. Scott Shell

Chemical Engineering , Princeton University

Computer simulation has become an important and well-established method for evaluating the structural, dynamic, and equilibrium properties of substances. In particular atomic-level models, molecular dynamics and Monte Carlo (MC) simulations have enhanced our predictive capability in the design of new materials and provided detailed physical insight in areas where experiments are normally difficult or expensive. Motivated by the experimentally elusive low-temperature behavior of an eminent liquid, water, we have been developing MC methods for the accurate and efficient low-temperature simulation, which has traditionally been computationally difficult. In this talk, I will first briefly review the concepts behind MC molecular simulation in general, and then elaborate on our ongoing research in formulating and characterizing efficient and easily parallelizable algorithms pertinent to low-temperature simulation.