Development of a multi‐step screening procedure for redox active molecules in organic radical polymer anodes and as redox flow anolytes

Andreas J. Achazi,Xhesilda Fataj,Philip Rohland,Martin D. Hager,Ulrich S. Schubert,Doreen Mollenhauer

DOI: https://doi.org/10.1002/jcc.27299

2024-01-24

Journal of Computational Chemistry

Abstract:A benchmark study of benzo[d]‐X‐zolyl‐pyridinyl (1‐X‐R)‐type radicals is presented. The accuracy of semi‐empirical methods and density functional theory in predicting the redox potential is determined. The results are compared with experimental and high‐level wave function‐based benchmarks. A three‐step screening procedure has been developed to find 1‐X‐R‐type radicals with a suitable redox potential for batteries. First a semi‐empirical method, then DFT orbital energies, and finally DFT Gibbs energies are used. Benzo[d]‐X‐zolyl‐pyridinyl (XO, S, NH) radicals represent a promising class of redox‐active molecules for organic batteries. We present a multistep screening procedure to identify the most promising radical candidates. Experimental investigations and highly correlated wave function‐based calculations are performed to determine benchmark redox potentials. Based on these, the accuracies of different methods (semi‐empirical, density functional theory, wave function‐based), solvent models, dispersion corrections, and basis sets are evaluated. The developed screening procedure consists of three steps: First, a conformer search is performed with CREST. The molecules are selected based on the redox potentials calculated using GFN2‐xTB. Second, HOMO energies calculated with reparametrized B3LYP‐D3(BJ) and the def2‐SVP basis set are used as selection criteria. The final molecules are selected based on the redox potentials calculated from Gibbs energies using BP86‐D3(BJ)/def2‐TZVP. With this multistep screening approach, promising molecules can be suggested for synthesis, and structure–property relationships can be derived.

chemistry, multidisciplinary