A research team at the University of Central Florida has applied Artificial Intelligence (AI) to perovskite solar cell (PSC) research to develop a system to identify the best materials. The Organic-Inorganic halide perovskite material used in PSC helps in converting photovoltaic power into consumable energy. These perovskite solar cells can be processed in the solid or liquid state thereby offering flexibility.
The researchers reviewed more than 2000 peer-reviewed publications about perovskites and collected more than 300 data points which were then fed into a machine learning algorithm. Thereupon, the system analyzed the information and predicted which recipe for spray-on perovskite solar technology would work best.
The researchers said that the machine learning approach helped them in understanding how to optimize material composition and predicting the best design strategies and potential performance of perovskite solar cells. The machine learning predictions corresponded with the Shockley-Queisser limit. Machine learning also helped in predicting optimum frontier orbital energies between the transport layer and the perovskite layer.
Spray-on solar cells could be used to spray-paint bridges, buildings, homes, and other structures to capture light, turn it into energy and feed it into the electrical grid. It is anticipated that the formula could become the standard recipe/guide for making flexible, stable, efficient, and low-cost perovskites.
The research was published in Advanced Energy Materials (www.doi.org/10.1002/aenm.201970181).