Prelude to molecular dynamics: Topography-driven gaussian charge models

Abstract

A new strategy to develop Gaussian charge models (GCMs) for molecules like ammonia, water, ethene, hydrogen sulfide, formaldehyde and benzene is presented. These molecular models comprising of positive point charges and negative Gaussian charge distributions (GCDs), which represent nuclei and continuous electron charge distribution, are found to correctly represent the ab initio Molecular Electrostatic Potential (MESP) and reproduce its essential topographical features of corresponding molecules. The models use optimized parameters: positive charges at nuclei, negative charges on GCDs, Gaussian exponent and centers. The Potential Energy Surface (PES) of water dimer has been explored using water GCMs. A good agreement has been found between PES obtained using GCMs and wave function. The Gaussian models correctly predict structure of benzene-water complex. It is thus recommended to use GCMs for molecular dynamic simulations.