Artificial Intelligence in Quantum Intramolecular Dynamics with Application to a Heavy Central Mass Problem

Citation

Lederman, Steven Mark (1988) Artificial Intelligence in Quantum Intramolecular Dynamics with Application to a Heavy Central Mass Problem. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/8333-3M02. https://resolver.caltech.edu/CaltechTHESIS:03082013-103249804

Abstract

This dissertation consists in effect, of three parts, each involving some aspect of intramolecular vibrational relaxation. The first section contains simple approximate statistical formulas for the density of vibrational and rovibrational states by symmetry type for non-linear molecules. A modified Whitten-Rabinovitch estimate of the density of states by symmetry type for linear molecules is also derived. Sample calculations are given, which serve to demonstrate the accuracy of all formulas given. In the second section, a 4-coordinate model is presented and is used to treat the vibrational energy redistribution in a molecule with a heavy central metal atom. Local group modes are identified using perturbation theory, and their dynamical separation and importance in analyzing energy redistribution is noted. A comparison of classical and quantum calculations on the model system is also given. In the third section, artificial intelligence methods are used to treat the time-evolution of intramolecular quantum dynamics. Comparison is made of several AI search algorithms and of evaluation functions, proposed here, in an application to the study of quantum intramolecular vibrational relaxation. The methods developed are applied to an 11-coordinate heavy central mass problem and are used to treat both vibrational quantum beats and "dissipative" intramolecular energy transfer.

Thesis Files