Massively Parallel Combined Monte Carlo and Molecular Dynamics Methods to Study the Long-Time-Scale Evolution of Particulate Matter and Molecular Structures Under Reactive Flow Conditions

摘要

An important challenge in computational modeling is the development of new computational methods and capabilities for studying molecular-scale structures over very large time-scales. In particular, there is great interest in understanding the nucleation and growth of carbon soot particles as well as their fate in the atmosphere. We have recently developed and implemented a new computational tool to time-integrate the detailed structure of atomistically resolved surfaces and nanostructures driven by chemical and physical kinetic rule-based rate expressions. Fundamental chemical and physical processes such as chemical reactions, surface adsorption and surface diffusion are performed using a non-lattice realspace kinetic Monte Carlo scheme and driven by user-defined rule-based kinetic rate expressions, while atomic structure relaxation is approached using molecular dynamics. We demonstrate the sensitivity of particle evolution to chemical and physical kinetic mechanism using a parallel implementation of the combined Monte Carlo and molecular dynamics code.