| The full nature of theoretical chemistry and physics cannot be captured by static calculations alone. The underlying goal of this thesis is the utilization of dynamics methods, appropriate to the size and type of chemical system under consideration, as well as the type of desired data obtained from the calculation. A small, potentially high-energy molecule, FN5, was studied with high level, "on-the-fly" ab initio (AI) methods in order study isomerization and decomposition pathways, and ultimately predict the existence (lifetime) of the species. Experimentalists confirmed these calculations. In order to study solvation processes, large numbers of molecules, as well as dynamics methods, are important. The Effective Fragment Potential (EFP) method for solvation, a method based on quantum mechanics calculations, was parallelized to facilitate these calculations within the quantum chemistry program GAMESS. The parallel algorithm employs both atom decomposition of data, as well as non-blocking communication. Speedup and scalability of the code was achieved. The EFP method has been shown to provide excellent results for small water clusters in previous calculations. In order to test the bulk behavior of the EFP method, EFP molecular dynamics simulations were performed. The resulting radial distribution functions for water are in good agreement with experimental data. Finally, one of the most fundamental aspects of a chemical reaction was investigated: the molecular potential energy surface (PES). This involved the interface of the Grow and GAMESS programs. Grow builds a PES as an interpolation of AI data, and thus requires AI calculations of energy and derivatives from GAMESS. Classical or quantum dynamics can be performed on the resulting surface. The interface also includes the capability to build multi-reference PESs; these types of calculations are applicable to a wide array of problems, including photochemistry and photobiology. |
