Amount Awarded: $19,200
Classical molecular simulations are typically required to study thermodynamic and transport properties of adsorbates in nanoporous materials due to the current computational cost of ab initio molecular dynamics or ab initio Monte Carlo methods. Standard, off-the-shelf force fields such as the Universal Force Field (UFF) or Dreiding force field have found significant use in the study of nanoporous materials, yet these force fields are very general, and far from perfect in describing guest-host interactions in a variety of chemically interacting systems. Hence significant efforts have been directed towards parameterizing classical force fields to describe adsorbate-host interactions for situations in which off-the-shelf force fields fail. In this project we take advantage of state of the art methods currently used, and extend these to be more general, reduce the computational cost, as well as minimize the amount of human interaction in force field development. In this project we merge methods such as the use of approach-paths1 to minimize the required number of electronic structure calculations and the use of genetic algorithms as a method of global parameter optimization2, with new approaches. We introduce the use of second order Møller-Plesset perturbation theory (MP2) cluster calculations that scale O(n) rather than O(n5) to obtain more accurate forces and binding energies to parameterize our force field in host clusters whose size are much larger than the cut off radius of a classical force field. Our global optimization technique uses genetic algorithms which can parameterize multiple pairwise potential functional forms to best describe the underlying potential energy landscape. These concepts have been implemented in a modular Python framework to fully automate force field fitting, thus creating a tool of significant use to the molecular simulations community. The reduction in time required to generate a force field for a particular adsorbate-host combination will reduce the barrier to obtaining accurate thermodynamic and transport properties in complex interacting nanoporous material systems.