Abstract:Single-electron reduction half potentials of 95 octahedral fourth-row transition metal complexes binding a diverse set of ligands have been calculated at the unrestricted pseudospectral B3LYP/LACV3P level of theory in a continuum solvent. Through systematic comparison of experimental and calculated potentials, it is determined that B3LYP strongly overbinds the d-manifold when the metal coordinates strongly interacting ligands and strongly underbinds the d-manifold when the metal coordinates weakly interacting ligands. These error patterns give rise to an extension of the localized orbital correction (LOC) scheme previously developed for organic molecules and which was recently extended to the spin-splitting properties of organometallic complexes. Mean unsigned errors in B3LYP redox potentials are reduced from 0.40 ± 0.20 V (0.88 V max error) to 0.12 ± 0.09 V (0.34 V max error) using a simple seven-parameter model. Although the focus of this article is on redox properties of transition metal complexes, we have found that applying our previous spin-splitting LOC model to an independent test set of oxidized and reduced complexes that are also spin-crossover complexes correctly reverses the ordering of spin states obtained with B3LYP. Interesting connections are made between redox and spin-splitting parameters with regard to the spectrochemical series and in their combined predictive power for properly closing the thermodynamic cycle of d-electron transitions in a transition metal complex. Results obtained from our large and diverse databases of spin-splitting and redox properties suggest that, while the error introduced by single reference B3LYP for simple multireference systems, like mononuclear transition metal complexes, remains significant, at around 2-5 kcal/mol, the dominant error, at around 10-20 kcal/mol, is in B3LYP's prediction of metal-ligand binding. Application of the LOC scheme to the rate-determining hydrogen atom transfer step in substrate hydroxylation by cytochrome P450 shows that this approach is able to correct the B3LYP barriers in comparison to recent kinetics experiments.

Machine learning-aided first-principles calculations of redox potentials

Machine Learning-Aided First-Principles Calculations of Redox Potentials

Absolute standard hydrogen electrode potential and redox potentials of atoms and molecules: machine learning aided first principles calculations

Predicting Redox Potentials by Graph-Based Machine Learning Methods

Predicting redox potentials by graph‐based machine learning methods

Machine Learning Potential for Electrochemical Interfaces with Hybrid Representation of Dielectric Response

Machine learning potentials for redox chemistry in solution

Computational Prediction for Oxidation and Reduction Potentials of Organic Molecules Used in Organic Light-Emitting Diodes

Ab Initio Calculations of the Redox Potentials of Additives for Lithium-Ion Batteries and Their Prediction through Machine Learning

Prediction of Redox Power for Photocatalysts: Synergistic Combination of DFT and Machine Learning

Towards Predicting Metal Extraction from Oxides with Machine-Learning Augmented First-Principles Calculations

Redox potentials and acidity constants from density functional theory based molecular dynamics.

Teaching oxidation states to neural networks

Augmenting zero-Kelvin quantum mechanics with machine learning for the prediction of chemical reactions at high temperatures

Development of Accurate DFT Methods for Computing Redox Potentials of Transition Metal Complexes: Results for Model Complexes and Application to Cytochrome P450

Redox‐active ligands as a challenge for electronic structure methods

ChemInform Abstract: Redox Potentials and Acidity Constants from Density Functional Theory Based Molecular Dynamics

Predicting the Redox Potentials of Phenazine Derivatives Using DFT-Assisted Machine Learning

Density Functional Theory and Machine Learning for Electrochemical Square-Scheme Prediction: An Application to Quinone-type Molecules Relevant to Redox Flow Batteries

Machine Learning Approach to Vertical Energy Gap in Redox Processes

Machine learning surrogate models for particle insertions and element substitutions