Abstract - McCoy - 9810194 Free-radical reactions are common in industrial applications such as polymerization, depolymerization, petrochemical cracking and oligomerization, etc. The PI plans to investigate fundamental and general models that quantify the mechanisms and kinetics of these macromolecular reactions in complex mixtures. Most of the applications involve distributions of many components that can be mathematically described by distribution kinetics. The PI hypothesizes that these radical processes can be modeled as combinations of elementary reactions among semi-continuous molecular weight distributions. This hypothesis will be tested by theoretical and experimental investigations with the common feature that macromolecular distributions will be monitored, for example, by gel permeation chromatography, and compared with mathematical models. Elementary reactions, such as initiation, hydrogen abstraction, chain scission, propagation by coupling, and termination, will constitute the mechanisms. Recent results indicate that these reaction steps lead to experimentally verified rate expressions for random chain and chain-end cleavage processes for polymers when the well-known long-chain and quasi-stationary-state approximations are applied to radicals. Extending the analysis to include hydrogen donors and mixtures of macromolecular radicals is planned. Most published results in continuous distribution kinetics are based on the assumption that rate coefficient, k, are independent of molecular weight, x. This is a valid approximation for polymer decomposition experiments that last only a few hours. Longer experiments yield a larger shift in average molecular weight and need to be interpreted by a model with k(x). Polynomial or power-law expressions for k(x) will be tested to describe such experiments. The radical processes planned for experimental study are hydrogen-donor assisted cracking, polymer mixture reactions, and cellulose degradation. Theoretical investigations to de velop mathematical solutions to the fundamental balance equations for the radical mechanisms will be conducted in parallel with experiments. The PI will extend current models to include multivariable distributions, such as branching and molecular weight. Exact mathematical solutions that are related to similarity solutions will be developed to show the evolution of molecular weight distributions over a longer time range.
