Steric and Electronic Effects Govern Electron Transfer Rates in Pyridinium Electrolytes

N-substituted pyridiniums are promising candidates for electrochemical energy storage applications due to their high solubility and low redox potentials; however, they often exhibit heterogeneous electron transfer (ET) kinetics that are insufficient for high power applications.Rational molecular design to improve ET rates remains challenging because the quantitative relationships between substituent-level structure and heterogeneous ET kinetics are under described for substituted pyridinium systems.While first-principles ET models provide mechanistic descriptors, they do not provide direct mapping of molecular substituents onto kinetic parameters.Here, we combine electrochemical characterization of a series of 2,6-dialkyl-4arylpyridinium derivatives with supervised machine learning to develop a predictive structurekinetics model for the heterogeneous ET rate constant (k 0 ).The resulting model relates k 0 to the percent buried volume of the pyridinium nitrogen atom (VB,N) and the Hammett σm parameter, which yielded a strong performance for both the training set (13 compounds, R 2 = 0.85) and external validation set (four compounds R 2 = 0.92, Q 2 = 0.84).Notably, the model accurately predicts k 0 for a 2,6-diethylpyridium derivative that lies outside the structural and kinetic space of the training set.Electrochemical data were further analyzed within the Marcus-Hush-Levich framework to extract mechanistic parameters.Interpreting the machine-learning-derived structural descriptors through this first-principles model identifies outer-sphere reorganization energy (e.g., solvent polarization and ion pairing energy) as the primary contributor to the observed structurekinetics relationships.By connecting data-driven models with ET theory, this work provides a physically grounded framework for guiding molecular design to tune heterogeneous ET kinetics in organic redox systems.