Machine Learning Models Screen and Design Organic Solar Cell Materials with Low Bandgaps to Boost Efficiency
-
Machine learning models are applied to design new low bandgap organic semiconductors for organic solar cells (OSCs) through data mining and property prediction.
-
Multiple databases are mined and over 20 machine learning models are developed to predict properties like bandgap to screen potential OSC materials.
-
The extra trees regressor model shows the best predictive performance for bandgap prediction based on statistical analyses.
-
Designed OSC candidate materials are evaluated by structure similarity analysis and library enumeration techniques.
-
Low bandgap organic semiconductors designed this way show potential to increase OSC efficiency for solar energy conversion.