Perovskite Solar Cell Design Revolution: A Machine Learning Perspective

Amid the present situation in the photovoltaic community, Perovskite Solar Cells (PSCs) are the nexus of interest due to their rising power conversion efficiency (PCE), cost-effectiveness, and the feasibility of their solution-processable manufacture. However, experimental research on PSC performance is very sophisticated, requiring immense effort and extensive resource expenditure. To further this technology, deployment of computational techniques becomes vital in shaping PSC performance in the material science field. In the adoption of data-driven machine learning (ML), the material science domain is facing a marked upturn. Computational capabilities and extensive databases cause this upturn. To predict the features of perovskite materials, imparting key insights into their physical properties and expediting development in PSCs ML techniques are deployed. This study embarks on fundamental steps and techniques deployed in the domain of ML as an introductory overview. Subsequently, considering their application in predicting properties and sorting of potential materials for PSCs, emphasis is placed on recent strides in ML. The study looks further into the deployment of ML in PSC designing, optimization and manufacturing. Eventually, this work ends with expert viewpoints, bringing attention to existing challenges as well as the bright future in this ever-evolving field.