This research is supported by the NSF theoretical and computational chemisty program. Improved computational techniques will be implemented in the General Atomic and Molecular Electronic Structure System (GAMESS) program. This includes the development and implementation of methods for performing electronic structure computations for large molecules using massively parallel computers, and computational methods for treating solvent effects. Electronic structure theory is combined with dynamics to obtain a qualitative and quantitative understanding of both thermodynamic and kinetic properties. The theory is applied to: (a) the elucidation of single and multiple bonds between main group and transition metals, (b) theoretical analysis of strained rings and hypervalent bonding, and (c) the computation of potential energy surfaces for and the prediction of reaction mechanisms of main group organometallic reactions. Specific reactions include: (a) hydrosilation and coupling reactions, (b) unimolecular decompositions and isomerations, and (c) reactions of multiply bonded compounds. Electronic structure theory treats a molecule, or a molecular assembly, as a mechanical system of strongly interacting electrons and nuclei. In principle, all chemical behavior can be understood by solving the quantum mechanical equations of motion of this many-particle system, but experience shows that very heavy computations are required to achieve the accuracy required to make reliable predictions of the chemical properties of moderately large molecules. With the advent of modern supercomputers, particularly, of massively parallel computers, this has become an active field of theoretical chemical research. This project is an example of such research. It emphasizes the application of the theory to compounds of Carbon, Silicon, and Titanium and the bonding of these elements to organic molecules containing transition-metal atoms such as iron. Of particular i nterest are moderately small molecules with unusual chemical bond structures such as strained rings or double bonds such as C=Si or Si=Si. Detailed study of the geometries and energetics of such unusual molecules contributes to our ability to design much larger molecules which possess these same bonding features or perhaps are formed by chemical reactions involving transient species which possess them.
