Molecular Structure–Redox Potential Relationship for Organic Electrode Materials—Density Functional Theory–Machine Learning Approach

In this study we implement density functional theory and machine learning to investigate the structure-electrochemical performance relationships of organic electrode moieties in Li ion batteries. Namely, DFT is used to gauge the electrochemical activity of a variety of organic moieties via the computation of redox potential. However, despite its ability to provide valuable insight regarding the electrochemical properties of novel organic molecules, high efficacy DFT modeling can still require significant computational time and thus is not ideal for the vast screening of candidate materials. Therefore, we implement machine learning as a pathway for the accelerated discovery of novel organic materials, and more importantly as a method for assessing the various structure-electrochemical relationships which can provide a more general guideline for the design of organic electrode materials. We employed three different learning models, namely artificial neural networks (ANN), kernel ridge regression (KRR), and gradient-boosting regression (GBR), via three different pipelines with varying sophistication to generate an advanced ML scheme for the accurate prediction and analysis of electrochemical activity.