Machine Learning a Predictive Tool for the Analysis of NiCoS/Graphene Composites for Supercapacitor.

Nickel cobalt sulfide (NiCoS) has considerable potential electrode material for supercapacitors owing to its distinct physical and chemical characteristics. However, the practical applications of pristine NiCoS have been limited by issues such as small specific surface area, agglomeration, and volume changes during cycling, leading to low specific capacitance/capacity and cyclic stability at high rates. Several efforts have been taken to address these challenges. Among those the design and development of NiCoS-graphene-based composites have been widely investigated. This review explores the effect of NiCoS architecture and its nanocomposite with graphene on the electrochemical properties. How the various preparative parameters such as synthesis methods, precursors, experimental conditions contributed to efficiently accelerating charge transport kinetics is outlined. Finally, the effect of introduction of graphene on the electrochemical performance of NiCoS is discussed using density functional theory (DFT). Also, machine learning (ML) models are used to analyze the specific capacitance variation with respect to different synthesis parameters, morphology, energy density, and power density. ML models identify limitations and scope of work for working on NiCoS/graphene composites. It is found that most influencing parameter is annealing time that can alter specific capacitance. The review outlines future research directions, challenges, and opportunities in NiCoS/graphene-based supercapacitor.