Applications of quantum chemistry to polymerization reactions and biological activity.

摘要

This research includes primarily computational chemistry modeling to develop a composite for future dental restorative materials. The studies include characterization of the reactivity of potential monomers for a low stress/low shrinkage material. One molecule contains a backbone structure which has shown potential to expand upon polymerization, and when combined with a durable monomer may result in such a material. Without a fundamental understanding of the simpler expanding monomer, modification of this compound to reach a final product with desirable properties would be more difficult without many more trials and fortuity.;The primary intent of this dissertation was to use quantum mechanical calculations to predict thermodynamic properties for the homopolymerization and copolymerization of study molecules to understand the reaction mechanisms. A vast quantity of research has been published in the literature supporting different theoretical methods to reliably predict experimental work, and the present research includes modeling reactions involving molecules similar to those which have published experimental data. Comparison of the theoretical work to this data further validated the ability of semiempirical methods to predict thermodynamic properties with unpublished data.;Following validation of the method for the study reactions, theoretical characterization of potential mechanisms was completed to predict the experimental polymer structure. The polymer was analyzed with various analytical techniques for structure identification, and analysis of the experimental data provided information which corroborates the plausible copolymerization product although the actual structure could not be determined. The results of semiempirical calculations were compared with the analytical data and enabled prediction of competing reactions for probable mechanisms which may not otherwise have been recognized.;A second study in this dissertation included the development of the first published model using quantum mechanical descriptors to predict the sensitization potential for structures not tested experimentally, which corresponded to the understood mechanism. Statistical analysis of the model followed by testing with an external test set reinforced the predictive ability of this model. This was further confirmed by predicting the sensitization potential of candidate structures using the model and comparing to experimental tests performed utilizing the prediction for efficiently selecting test concentrations and sensitization potential.