Machine Learning-Assisted Design of Organic Photovoltaic Materials with Tunable Energy Bandgaps

Abstract

This research paper delves into the utilization of machine learning (ML) algorithms, with a focus on the Random Forest (RF) model, to predict and optimize the energy bandgaps of organic photovoltaic (OPV) materials. The primary objective was to explore the capability of ML in enhancing the design process of OPV materials by accurately determining their energy bandgaps, a critical factor influencing solar energy conversion efficiency. The methodology involved the collection of a comprehensive dataset from a fictional database, "OPVDataHub," comprising molecular structures and photovoltaic properties of various OPV materials. The RF model was developed & assessed using this dataset, with effectiveness measured like Mean Absolute Error(MAE) AND Root Mean Squared Error(RMSE).

Key findings from the study highlighted the model's high accuracy in predicting energy bandgaps, showcasing the significant predictive power of ML in material science. Furthermore, the analysis identified crucial molecular features—molecular weight, HOMO, and LUMO energy levels—as determinants of energy bandgaps, providing insights into the molecular underpinnings of OPV material performance. The implications of this research are profound, suggesting that ML can substantially accelerate the OPV material design process, paving the way for the development of more efficient solar energy technologies. By bridging the gap in existing literature, this study underscores the potential of integrating ML into the realm of renewable energy research, offering a novel approach to material optimization and discovery.