Accessibility of Ring-Closing Depolymerization for Polycarbonates with Multiple Theory Validation

We present an accelerated protocol to qualitatively screen polycarbonate monomers for viability of depolymerization in solvents of varying polarities. The largest barrier in the exploration of reaction kinetics is the necessary conformational analysis to identify lowest energy paths in depolymerization, which can depend on the solvent. Tight-binding DFT allows for an accurate quantification of depolymerization barriers as compared to DFT, with an increase in speed by up to 3 orders of magnitude. By a detailed featurization of over 30 compounds we verified that depolymerization barriers are not linearly correlated to the dielectric constant of the solvent. Volume of the molecules is explicitly not correlated with the depolymerization barrier; the solvent-molecule interactions and chemistry of the molecules is extremely influential due to the over coordinated oxygens in the transition state. Though computation of the solvent-solute interaction parameter is necessary for future predictive work in kinetic barriers, even alongside the 1800 Mordred features, conformational analysis is still necessary to determine the lowest kinetic barrier across solvents. We present multiple examples of the of lowest-barrier polycarbonates in acetonitrile, tetrahydrofuran, and toluene to highlight the large opportunity in this field for development in this multi-dimensional optimization problem. For the current available simulation and predictive models, our high-throughput protocol can evaluate the ideal solvent for 6-member polycarbonate depolymerization. From this work, further investigation into solvent-molecule interactions for quick computing features will lead towards the goal of data-driven methods for polycarbonate discovery and optimization.