Related Experiment Video
Updated: Aug 6, 2026
Influence of Hybrid Perovskite Fabrication Methods on Film Formation, Electronic Structure, and Solar Cell Performance
Published on: February 27, 2017
Artificial Intelligence-Guided Cosolvent Design for High-Performance Perovskite/Silicon Tandem Solar Cells
Lu Liu1,2, Xinying Cai2, Bita Farhadi3
1State Key Laboratory of Photoelectric Conversion and Utilization of Solar Energy, Center of Materials Science and Optoelectronics Engineering, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian, 116023, People's Republic of China.
We used artificial intelligence (AI) to screen over 8000 solvents, identifying gamma-valerolactone (GVL) as a high-performance cosolvent for perovskite solar cells. This AI-guided approach enhances perovskite crystallization for efficient and stable tandem solar cells.
Area of Science:
- Materials Science
- Renewable Energy
- Artificial Intelligence
Background:
- High-performance perovskite/silicon tandem solar cells depend on controlled wide-bandgap perovskite crystallization.
- Solvent engineering is key but complex, hindering intuitive design.
- Current methods struggle with precise control over perovskite film formation.
Purpose of the Study:
- To develop an AI-guided solvent engineering strategy for perovskite crystallization.
- To identify novel, non-toxic cosolvents for high-performance perovskite solar cells.
- To improve the scalability and stability of perovskite layers in tandem devices.
Main Methods:
- Utilized a retrieval-augmented large language model to screen over 8000 solvents.
- Investigated the coordination mechanism of gamma-valerolactone (GVL) with FA+ cations.
- Analyzed the impact of GVL on perovskite crystallization kinetics, nucleation, and grain growth.
Main Results:
- Identified gamma-valerolactone (GVL) as a non-toxic, effective cosolvent.
- GVL precisely modulates crystallization, promoting oriented, large-grain perovskite films.
- Achieved 23.3% single-junction and 32.5% tandem solar cell efficiencies with enhanced stability.
Conclusions:
- Established the first AI-guided cosolvent strategy for 1-μm-thick perovskite layers in tandem architectures.
- Demonstrated GVL's role in improving crystallinity, reducing recombination, and enhancing film tolerance.
- Highlighted the potential of generative AI in advancing high-performance photovoltaic technologies.
