Electrochemical Properties of Quinone Derivatives: First-Principles Density Functional Theory Modeling and Machine Learning Approach

Abstract

Lithium-ion battery has received intensive attention as a candidate for efficient and portable electrochemical energy storage systems (EES) to be used in many applications ranging from cell phones to electric vehicles. Lithium-ion batteries not only have the high performance but also cyclic stability making them the optimal choice for recharging-discharging process. Despite the high potential, the slow diffusivity of lithium ions, which is mainly due to the intrinsic properties of the conventional cathode materials, which are mainly transition metal oxides, resulting in poor power density. Organic electrode materials have been considered as promising substitute for metallic oxide cathodes due to safety, low cost, and low density. Among various organic materials, the ones with the carbonyl functional group have been recognized to achieve high-performance redox couples with stable and reversible redox reactions. Finding cathode materials with high redox potentials is critical for organic molecules as they exhibit lower potential compared to the conventional electrode materials. The lower potential exhibited by the organic molecules compared to the metallic oxides are believed to be due to the low ionic conductivity of organic materials, despite having good electron conductivity. In this study, we characterize the electrochemical potentials of quinone-based structures in as a function of molecular structures. Quinone-based organic materials have well-defined active sites and fast electron and ionic transfer kinetics compared to other organic molecules, which is suitable for battery application. We also investigate what factors may contribute to the electrochemical properties using machine learning technique, especially artificial neural network.