Empirical molecular force field models from molecular orbital calculations

At SDSC a small research project has been initiated to explore the feasibility of determining empirical model potential functions from an ''experimental database'' consisting of numerical molecular orbital second derivative matrices from calculations on model compounds. Newly developed analytical methods for empirical force field determination make it possible to fill in the ''gaps'' which are always present in the existing experimental data. Also, there is never an experimental uncertainty, or an ambiguity of vibrational frequency assignment to deal with. And, since so much ''experimental data'' can be generated, it is now possible to test various formulations of model force fields and directly study the importance of various cross terms which are included by some investigators and ignored by others. Preliminary results for simple classes of molecules are presented. All calculations presented were performed with the aid of the SDSC SCS-40 superminicomputer. Features of this new supermini are discussed