Chemical oxidants such as ozone are key for eliminating waterborne diseases as well as for the abatement of micropollutants during (waste)water treatment.[1] Because a comprehensive experimental characterization of the complex reaction networks is extremely laborious and inherently incomplete, autonomous and unsupervised computational approaches are required to infer the formation of unknown and potentially toxic disinfection byproducts (DBPs) and transformation products (TPs). Here, we address this challenge using autonomous quantum based chemical reaction network (CRN) explorations driven by the novel SCINE framework.[2] To identify the appropriate quantum chemistry methodology for the exploration, we performed a benchmark study that focused on the 1,3-dipolar cycloaddition of ozone to ethene. Among the various semi-empirical, DFT and CC methods evaluated, we concluded that a composite method based on LC-PBE structures and DLPNO-CCSD(T) electronic energies, with CPCM implicit solvation, provided the optimal ratio between accuracy and computational cost. With this methodological basis, we constructed two CRNs of (i) ethene and ozone, and (ii) tetramethylethene and ozone, consisting of hundreds of compounds and thousands of reactions, respectively. Both explorations found key reported compounds such as formaldehyde, acetone, hydrogen peroxide, hydroxymethyl hydroperoxide, as well as hydroxyl and ozone radicals.[3] All the compounds of the networks can be visualized using the standalone and interactive HTML files generated with our recently developed module.[4] Utilizing graph theory, we deduced from the CRNs the identification of the well-established Criegee mechanism. Microkinetic simulations predicted the correct end-products of the two CRNs, including the lifetime of the ozonide. We also evaluated the limitations of the solvation model in the final yields for the tetramethylethene system. Overall, this first successful application of automated reaction exploration to oxidative water treatment shows great promise for elucidating the complex redox chemical space that leads to the formation of TPs and/or DBPs.[5]
Dr. Enric Petrus