Simulation methods in chemical physics: Monte Carlo and ab initio studies

The exponential growth of computing power in the last few decades has enabled chemists and physicists to perform theoretical calculations of realistic systems with accurate quantum-mechanical methods. It is, however, still a long way from making precise predictions on physical properties of real systems and computer simulations remain to be rapidly expanding field in all of science as well as physics and chemistry. We make an attempt to develop some of the simulation methods used in the field of chemical physics in particular. We start with a review of modern computational chemical physics in relation to the development of supercomputers, especially parallel supercomputers in recent years. The mutual effect on the development of each field of supercomputing and computational chemical physics is elucidated. Path Integral Monte Carlo (PIMC) has been developed as a computational tool to solve Schrodinger equation using computers. Fourier path integral (FPI) formalism is one way to implement PIMC and possesses some conceptual advantages. We employ FPI formalism to form an alternative way to solve classical-mechanical problems using double-ended boundary conditions. This method is applied to a molecular model as an illustration. Further elaborating on PIMC, we utilize wavelets, which have been successful in engineering applications, to develop an alternate formulation. Wavelet Path Integral Monte Carlo (WPIMC) is initiated and applied to a few model problem calculations. The implementation of WPIMC is successful, and its efficiency is at least comparable to conventional FPI. Finally, we turn to first-principles calculations for analysis of nickel clusters. Nickel clusters of size up to tetramer are studied using a few different ab initio methods. Results are compared with previous works and experiment. Suggestions for further research and reasonable set of ab initio method factors for nickel clusters studies are presented.