Machine Learning Accelerates Discovery Of Solar Cell Perovskites
AIhub 1:20 pm on May 28, 2024
On May 28, 2024, EPFL researchers developed a machine-learning method to identify optimal halide perovskites for photovoltaic applications using advanced computational techniques. They created an extensive dataset of accurate band gaps and employed hybrid functionals in their calculations, surpassing traditional Density Functional Theory (DFT). Their approach successfully discovered 14 new promising perovskite materials suitable for high-efficiency solar cells by predicting optimal band gaps.
- EPFL Research Breakthrough:
- Developed advanced computational method to identify best halide perovskites.
- Extensive Dataset Creation:
- Band gap values for 246 perovskite materials were accurately calculated using hybrid functionals.
- New Perovskite Materials Discovery:
- Machine learning model narrowed down to 14 new candidate materials with promising band gaps and stability for solar cells.
https://aihub.org/2024/05/28/machine-learning-accelerates-discovery-of-solar-cell-perovskites/