Method to Predict Reagents in Iridium-Based Photoredox Catalysis

Abstract

Visible-light photoredox catalysts with oxidizing excited states have been broadly applied in organic synthesis. Following photon absorption by the photocatalyst, electron transfer from an organic reagent is the most common mechanistic outcome for this class of reaction. Reduction potentials for organic reagents are therefore useful to predict reactivity and DFT proved to be useful as a predictive tool in this regard. Due to the complex mechanisms that follow electron transfer, kinetics play a crucial role in the success of photoredox reactions. We extend the predictive tools of DFT to estimate the electron transfer <i>rates</i> between an excited photocatalyst and various organic substrates. To calibrate our model, 49 electron transfer rate constants were experimentally measured in acetonitrile for the catalyst Ir[dF(CF<sub>3</sub>)ppy]<sub>2</sub>(dtbpy)<sup>+</sup>. The rate constants, k<i><sub>q</sub></i>, gave a clear predictive trend when compared to calculated ionization energies in “frozen solvent”, which was a better predictor than standard reduction potentials in our case. The calculated k<i><sub>q</sub></i> gave an average error of 17% for log(k<sub>q</sub>) values between 4 and 11. This simple method can predict the reactivity of hundreds of reagents <i>in silico</i>. Notably, the calculations offered unexpected insight that we could translate into success for the C-H activation of acetylacetone as a proof-of-concept.