Designing of Polymers for Photovoltaics Applications and Prediction of Band Gap as a Polymers Screening Criterion

The development of efficient and sustainable energy sources has propelled research into organic photovoltaics (OPVs), where polymers play a critical role as active materials. Designing polymers for photovoltaic applications involves optimizing their molecular structure to achieve desirable electronic properties, such as a suitable band gap, which directly influences the efficiency of solar energy conversion. In this study, we present a systematic approach for the design and screening of polymers for OPV applications, focusing on the prediction and tuning of band gaps as a key criterion. Six machine learning models are tried. Random Forest is selected as the best model; it shows an r-squared value of 0.78 and root mean squared error value of 0.7 eV. A library of 20,000 polymers is curated and analyzed. Polymers are screened on the basis of predicted band gap values, and 30 polymers are selected. Selected polymers have a band gap between 1.18 and 1.97 eV. Chemical similarity analysis has indicated similar structural behavior for selected polymers.