Accelerating ionic liquid research in perovskite solar cells through machine learning:Opportunities and challenges

Abstract

In recent years, there have been continuous and remarkable efforts from both academic and industry to improve the efficiency and stability of perovskite solar cells (PSCs). Among all the efforts, Ionic liquids (IL), a class of compounds with asymmetric organic cations and various anions, stand out as one of the most promising additives and interface modification layer for realizing high performance PSCs due to their unique physicochemical properties. Nonetheless, due to the variety of ionic liquids, searching an effective and optimum IL passivation materials for PSCs requires a huge amount of time and efforts in conventional trial-and-error experiments. In this context, machine learning (ML) offers powerful capabilities to handle complex, nonlinear problems, potentially accelerating the discovery and optimization of IL for PSCs applications. This review provides a comprehensive overview of the current applications of IL in PSCs, and summarizes the opportunities and key challenges in combining ML methods for IL research in PSCs. With the proposed ML frameworks, it is expected that a more predictive ML piloted research process would accelerate the discovery and optimization of IL in PSCs.