Perovskite solar cells are an advanced solar cell technology, and research to improve their energy conversion efficiency is being actively pursued by many researchers. Existing perovskite solar cell development has relied on fabricating and experimentally testing devices to improve efficiency. This approach is time-consuming and costly, and the limited testing options lead to opportunity costs.

To minimize wasted costs in conventional perovskite solar cell research, the research team led by Professor Won-yeop Noh of Jeonbuk National University (JBNU), Jimmy Carter School of International Studies, School of International Engineering and Sciences has reported findings showing that using artificial intelligence to predict optimal composition ratios can improve energy conversion efficiency. The work has attracted significant interest in the academic community.

The study predicting the combination of two materials in the perovskite solar cell fabrication process via artificial intelligence is notable as the world's first report of its kind.

The team announced on the 12th that they used artificial intelligence to predict the optimal weight percentages (wt%) of nanopatterned titanium dioxide (mp-TiO2) and phenyl-C61-butyric acid methyl ester (PCBM), thereby improving the energy conversion efficiency of perovskite solar cells.

The research was published in the prestigious international journal Energy & Environmental Materials (IF: 15.0) under the title "Machine Learning-Assisted Fabrication of PCBM-Perovskite Solar Cells with Nanopatterned TiO2 Layer."

Master's researcher Siti Norhasana Binti Sanimu, PhD candidate Jibun Kundel, Master's researcher Moon Ye-chong, PhD candidate Kangasagar Sharma Gaudel, and PhD candidate Yoo Seung-ju participated in this research.

For this study, the researchers collected energy conversion efficiency data for perovskite solar cells and used artificial intelligence to identify patterns and relationships within the data and to make predictions.

After the predictions indicated a significant increase in energy conversion efficiency, the team fabricated perovskite solar cells accordingly and confirmed an increase in energy conversion efficiency of 32.5%, the researchers reported.

The Principal Investigator (PI), Professor Won-yeop Noh, stated, "This study is the first case of integrating artificial intelligence into the perovskite solar cell process, and we expect it will have a positive impact on related research and industrial fields."

Meanwhile, the PI, Professor Won-yeop Noh, who serves as director of the Energy-AI Convergence Graduate School Human Resources Training Project Group, has conducted various studies with Professor Yong-ju Kim (Department of Oriental Medicine Resource), Professor Se-jung Kim (Department of Chemical Engineering), Professor Bong-hyun Jeon (Konkuk University, Department of Systems Biotechnology), and Dr. Hwa-young Yang (UNIST) on improving solar cell energy conversion efficiency using artificial intelligence. Going forward, they plan to apply artificial intelligence to a variety of fields including future solar cells, secondary batteries, displays, nanoparticles, and biotechnology.

This research was carried out with support from the Ministry of Trade, Industry and Energy's Korea Energy Technology Evaluation and Planning (KETEP).