Theoretical Studies Of Group IVA And Group IVB Chemistry

Several new developments in the effective fragment potential EFP and the fragment molecular orbital FMO methods have been accomplished. A new EFP charge transfer CT interaction energy was derived and implemented in GAMESS. The new CT approach, based on quasi-atomic molecular bond orbitals, cuts the time required for this term by a factor of two, with little loss in accuracy. The analytic gradient was derived and implemented as well, so one can optimize geometries and determine reaction paths. The R-7 dispersion terms was also derived and coded. While most workers ignore the odd power dispersion terms, these terms can be very important in anisotropic situations, such as highly viscous liquids and on solid surfaces. We have also developed an integrated FMO-EFP method called the effective fragment molecular orbital EFMO method. We have shown that the EFMO method is both faster and more accurate than FMO. Fully analytic EFMO energy gradients have been partly completed. In collaboration with the Rappe group, we used the EFP and second order perturbation theory MP2 to help explain the experimental results of the Fayer group on phenol migration dynamics. An exhaustive computational study of hydrazine decomposition was performed using both MP2 and coupled cluster methods. This study, which was prompted by the use of hydrazine as a high energy fuel, completely mapped out the competing hydrazine decomposition pathways. This work was a true tour de force. An invited review on fragmentation methods is in press in the very prestigious journal Accounts of Chemical Research. Also in press is a report on the solvated electron. This is a very important work, because it demonstrates that local orbital methods can be very effective in studying the solvated electron problem.