Autonomous closed-loop platform discovers 5ANI as a passivation molecule for perovskite solar cells, simultaneously improving high efficiency and stability

1/ AI 활용 방법: Closed-loop search combining active learning and quantum modeling
2/ 적용 과제: Search for passivation molecule candidates for perovskite solar cells
3/ 보고된 결과: 5ANI was found and PCE 27.22% for small-area cells and PCE 23.49% for mini-modules were reported
4/ An autonomous closed-loop manufacturing platform was applied to the search for passivation molecules, and active learning and quantum modeling were used together to narrow candidates before linking to automated manufacturing.
5/ The research team identified 5ANI and reported 27.22% PCE in 0.05 cm² cells and 23.49% in 21.4 cm² mini-modules.
6/ Long-term stability results were also presented.
7/ From a materials research perspective, the key point of this case is that candidate molecule search, automated experiments, and performance measurement were tied into a single closed loop, enabling the discovery of a promising passivation agent with fewer experiments.
8/ In particular, using active learning to select the next experimental candidates and supplementing molecular-level information with quantum calculations directly improves the efficiency of materials design.
9/ However, these results were obtained for a specific materials system and process conditions, and they cannot be directly generalized to other perovskite compositions or processes.
10/ The reported performance depends strongly on the same automated workflow and measurement conditions, so when reproducing it, the candidate molecules, deposition conditions, device structure, and stability test conditions must be strictly matched. [1]