Design of organic electronic materials with lower exciton binding energy: machine learning analysis and high-throughput screening

The potential use of organic electronic materials in various optoelectronic devices has drawn considerable attention in recent years. For organic electronic devices to operate more effectively, the exciton binding energy must be reduced. Using high-throughput screening methods and machine learning analyses, a thorough framework for creating organic electronic materials with lower exciton binding energies has been described in the current research. Our approach combines computer simulations, databases of materials, and data-driven algorithms to find viable compounds for materials with low exciton binding energies. Python software has been used to analyze various ML models. Breaking retrosynthetically interesting chemical substructures methodology has been used to form new compounds. The characteristics of compounds has been represented through molecular descriptors. Random forest regressor model, gradient boosting regressor model, K Neighbors regressor model and extra tree regressor model have been used to analyze the performance parameters. Twenty organic semi-conductors have been selected with low Eb values. The synthetic accessibility score indicated easy synthesis of selected semi-conductors. Similarity analysis has indicated structural similarity between selected semi-conductors.

Machine learning is helping the potential use of organic electronic materials in various optoelectronic devices, such as organic photovoltaics, organic light-emitting diodes, and organic field-effect transistors.