Revolutionizing polymer engineering for Photodetectors: A Machine Learning-Assisted paradigm for rapid materials discovery

Abstract

Polymers are appealing candidates for photoelectric applications because of their intrinsic characteristics include flexibility in substrate compatibility, ease of manufacturing, room temperature operating conditions, and adaptable optoelectronic properties. In present study, we have used the machine learning (ML) for property prediction and polymer designing. Multiple ML models are trained. 10,000 novel polymers are generated using automatic method. The synthetic accessibility of the chosen polymers is predicted. Structural similarity among the selected polymers is also calculated that indicated good structural similarity among the chosen polymers. By effectively identifying and optimizing novel polymers, the implemented methods substantially increase the chances of uncovering superior materials suited for advanced applications.