According to the Karlsruhe Institute of Technology, solar cells made from perovskite semiconductor layers are already highly efficient and can be produced cost-effectively. This technology can also be made thin and flexible. "Perovskite photovoltaics is on the threshold of commercialization. However, challenges remain in terms of long-term stability and upscaling to large areas," says Prof. Ulrich Wilhelm Paetzold, who conducts research at the Institute of Microstructure Technology and the Light Technology Institute (LTI) at KIT. "In our study, we show that machine learning is crucial for improving the monitoring of perovskite thin-film formation required for industrial production," explains the physicist. With the help of deep learning - a method from the field of machine learning (ML) that uses neural networks - the researchers were able to quickly and accurately predict the material properties and efficiency of solar cells even beyond the laboratory scale.
Step towards industrial applicability
"Based on measurement data collected during production, machine learning can be used to identify process errors before the solar cells are finished. Additional testing methods are not necessary," says Felix Laufer, research associate at the LTI and lead author of the study. "The speed and performance of this method significantly improves data analysis. It can be used to solve tasks that would otherwise be difficult to accomplish."
The investigation of a new type of data set that documents the formation of perovskite thin films enables the precise assignment of process data to target variables such as energy conversion efficiency with the help of deep learning.
"Perovskite photovoltaics has the potential to revolutionize the photovoltaics market," says Paetzold, who heads the Next Generation Photovoltaics department at the LTI. "We show how process fluctuations can be quantitatively analysed by extending characterization methods with machine learning techniques. This makes it possible to ensure high material quality and layer homogeneity across large areas and many batches. This is a decisive step towards industrial applicability," explains the scientist.