(442b) Multiscale Modeling and Characterization of Radical-Initiated Modification of Molten Polyolefins

Radical grafting mechanisms provide powerful and frequently-used methods for the functionalization of polyolefin chains which are relatively devoid of functional handles, having found applications as structural adhesives, high performance elastomers, and reactive copolymer intermediate, etc. Here, by investigating the reaction kinetics through a hybrid quantum calculation procedure, a comprehensive mechanistic view of the complex free-radical mediated synthesis of functional polyolefins is achieved, which includes grafting of vinyl silane monomers, inter-and intramolecular H-transfer as well as chain scission. Different radical species are found to be in kinetic equilibrium with their relative concentration depending on both the thermal stability and the abundance of specific carbon moiety. Our results clearly show that the overall mechanism is dominated by grafting of single vinyl silane monomer via intramolecular H-transfer to hydrocarbon substrates, rather than forming localized or oligomeric grafts, and this occurs at the expense of polymer crosslinking due to the termination of radicals via combination.

NMR+GC/MS spectroscopy of model compounds were also used to further elucidate the dominant reaction pathways for the initiation and the termination mechanisms. On the basis of the intrinsic kinetic dataset discerned from both computational chemistry and model compound studies, quantitative relationships between the product properties and the reaction conditions are therefore revealed with the intension to establish a polymer topology based kinetic model. To evaluate the quality of the resulting kinetic model for radical grafting reactions, it is compared to a factorial design of melt processing experiments on industrial grade polyolefin samples, including comparisons of the degree of grafting via FTIR, the molar mass measured via high temperature GPC, and the linear rheology of the polymer melts. The results show strong correlation between model predictions and the experimental results across a wide range of parameter space, suggesting the predictive ability of this approach for complex chemical networks.