Abstract
The commercialization of perovskite solar cells is bottlenecked by inefficient, trial-and-error approaches reliant on human expertise in both material discovery and device fabrication (1-3). Here, we introduce an autonomous closed-loop framework that integrates machine learning (ML)-driven material discovery with an automated manufacturing platform. The system employs active learning and quantum modeling to rapidly identify high-performance molecules, while the platform uses Bayesian optimization and symbolic regression in a feedback loop to continuously refine the fabrication process. This integrated approach enabled the discovery of a passivation molecule, 5-(aminomethyl)nicotinonitrile hydroiodide (5ANI), which yielded 0.05 cm² solar cells with a power conversion efficiency (PCE) of 27.22% (certified maximum power point tracking (MPPT) efficiency of 27.18%) and 21.4 cm² mini-modules with a PCE of 23.49%. Moreover, the devices exhibited long-term operational stability, retaining 98.7% of their initial efficiency after 1,200 hours of continuous operation under the ISOS-L-1I protocol. Crucially, the automated platform achieved an efficiency reproducibility nearly 5 times that of manual fabrication. This work establishes an automated closed-loop system that synergizes ML-powered discovery with the high-fidelity data from automated manufacturing, setting a benchmark for autonomous discovery and manufacturing in photovoltaics and materials.
This is a preview of subscription content, access via your institution
Access options
Access Nature and 54 other Nature Portfolio journals
Get Nature+, our best-value online-access subscription
$32.99 / 30 days
cancel any time
Subscribe to this journal
Receive 51 print issues and online access
$199.00 per year
only $3.90 per issue
Rent or buy this article
Prices vary by article type
from$1.95
to$39.95
Prices may be subject to local taxes which are calculated during checkout
Additional access options:
Similar content being viewed by others
Author information
Author notes
Xianglang Sun (孙祥浪)
Present address: Hubei Key Laboratory of Material Chemistry and Service Failure, Key Laboratory for Material Chemistry of Energy Conversion and Storage, Ministry of Education, School of Chemistry and Chemical Engineering, Huazhong University of Science and Technology, Wuhan, P. R. China
These authors contributed equally: Danpeng Gao, Shuaihua Lu, Chunlei Zhang, Ning Wang, Zexin Yu, Xianglang Sun
Authors and Affiliations
Department of Chemistry, City University of Hong Kong, Kowloon, Hong Kong, China
Danpeng Gao (高丹鹏), Chunlei Zhang (张春雷), Ning Wang (王宁), Zexin Yu (余泽鑫), Xianglang Sun (孙祥浪), Francesco Vanin, Liangchen Qian (钱良辰), Bo Li (李博) & Zonglong Zhu (朱宗龙)
Department of Materials Science & Engineering, City University of Hong Kong, Kowloon, Hong Kong, China
Shuaihua Lu (陆帅华), Nan Li (李楠) & Xiao Cheng Zeng (曾晓成)
Department of Chemical Engineering and Biotechnology, University of Cambridge, Cambridge, UK
Rebecca Martin & Samuel D. Stranks
Department of Chemistry, Imperial College London; MSRH Building, White City Campus, London, UK
Francesco Vanin, Nicholas Long & Nicola Gasparini
Institute of Materials for Electronics and Energy Technology (i-MEET), Department of Materials Science and Engineering, Friedrich-Alexander University (FAU) Erlangen-Nürnberg, Erlangen, Germany
Larry Lüer & Christoph Joseph Brabec
School of Materials Science and Engineering, Central South University, Changsha, P. R. China
Bo Li (李博)
Electronic Engineering Department, The Chinese University of Hong Kong, New Territories, Hong Kong, China
Martin Stolterfoht
Department of Computer Science, City University of Hong Kong, Kowloon, Hong Kong, China
Junhui Hou (侯军辉)
Department of Applied Physics, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong, China
Jun Yin (殷骏)
Department of Electronic and Computer Engineering, The Hong Kong University of Science and Technology, Kowloon, Hong Kong, China
Yen-Hung Lin (林彥宏)
Department of Chemistry, The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, China
Haipeng Lu (吕海鹏)
Helmholtz-Institute Erlangen-Nürnberg for Renewable Energy (HI ERN), Forschungszentrum Jülich, Erlangen, Germany
Christoph Joseph Brabec
Energy Campus Nürnberg (EnCN), Fürtherstrasse 250, Nürnberg, Germany
Christoph Joseph Brabec
Institute of Materials Data Science and Informatics (IMD-3), Forschungszentrum Jülich, Jülich, Germany
Christoph Joseph Brabec
Hong Kong Institute for Clean Energy, City University of Hong Kong, Kowloon, Hong Kong, China
Zonglong Zhu (朱宗龙)
Authors
- Danpeng Gao (高丹鹏)
- Shuaihua Lu (陆帅华)
- Chunlei Zhang (张春雷)
- Ning Wang (王宁)
- Zexin Yu (余泽鑫)
- Xianglang Sun (孙祥浪)
- Rebecca Martin
- Francesco Vanin
- Liangchen Qian (钱良辰)
- Nicholas Long
- Larry Lüer
- Bo Li (李博)
- Martin Stolterfoht
- Junhui Hou (侯军辉)
- Jun Yin (殷骏)
- Yen-Hung Lin (林彥宏)
- Haipeng Lu (吕海鹏)
- Nan Li (李楠)
- Nicola Gasparini
- Christoph Joseph Brabec
- Samuel D. Stranks
- Xiao Cheng Zeng (曾晓成)
- Zonglong Zhu (朱宗龙)
Corresponding authors
Correspondence to Samuel D. Stranks, Xiao Cheng Zeng (曾晓成) or Zonglong Zhu (朱宗龙).
Supplementary information
Supplementary Information (download DOCX )
The file includes: Supplementary Notes 1 to 8, Supplementary Figs. 1 to 66, Supplementary Tables 1 to 16, and Supplementary References.
Reporting Summary (download PDF )
Supplementary Data (download XLSX )
Source Data for Supplementary Figs. 4 to 23.
Peer Review File (download PDF )
Supplementary Video 1 (download MP4 )
Demonstration of the automated manufacturing platform for perovskite solar cell processing. This video demonstrates the end-to-end, continuous manufacturing process of perovskite solar cells on an automated platform, encompassing thin-film fabrication, electrode thermal evaporation, and device performance testing. The specific steps are labeled in the top-left corner of the video as follows: (1) Perovskite solution intake: automated positioning of Pipette A and aspiration of the perovskite precursor solution. (2) Antisolvent intake: automated positioning of Pipette B and aspiration of the antisolvent (chlorobenzene, CB). (3) Spin coating: robotic gripper handling the substrate and the subsequent automated dispensing of the perovskite solution. (4) Antisolvent dropping: precise, automated dispensing of the antisolvent during the spin-coating process. (5) Film annealing: robotic gripper transferring the fabricated thin film to a hotplate for thermal annealing. (6) Edge trimming: automated mechanical cutter performing the P2 scribing process on the device. (7) Thermal evaporation: automated transfer and loading of samples into the thermal evaporation chamber for electrode deposition. (8) Performance testing: automated picking and transferring of the devices to the testing station, followed by automated data acquisition.
About this article
Cite this article
Gao, D., Lu, S., Zhang, C. et al. Autonomous closed-loop framework for reproducible perovskite solar cells. Nature (2026). https://doi.org/10.1038/s41586-026-10482-y
Received: 23 September 2025
Accepted: 01 April 2026
Published: 14 April 2026
DOI: https://doi.org/10.1038/s41586-026-10482-y