Predicting Performance of Photochemical Transformations for Scaling Up in Different Platforms by Combining High-Throughput Experimentation with Computational Modeling

Using light to drive a chemical transformation introduces challenges for ensuring the robust transferability of photochemical reactions across different platforms and scales. We demonstrate a modeling tool to predict the performance of a photochemical reaction as a function of the reactor geometry, concentration of the photoactive species, irradiance of the light source, and residence time. High-throughput experimentation is utilized to optimize reaction conditions and to determine kinetic parameters and quantum yield. Optical characterization of the photoactive reaction species and the reactor is performed to determine the photon absorption rate. The experimental data is combined with computational modeling to predict photochemical conversion for different vial or flow reactors across multiple scales for a [2 + 2] photocycloaddition reaction and a photoredox-mediated decarboxylative intramolecular arene alkylation reaction. The method developed in this work facilitates the transferability of the photochemical processes between different photoreactors without the need for an intensive experimental optimization for each and enables a robust and efficient scale-up.