Deep learning–assisted electron microscopy for apparent Z-height mapping of atomically dispersed catalysts
Science Advances · 13.11.2026 · 12 (46)
DOI: 10.1126/sciadv.aef7673 ↗Kurz gefasst
Fachartikel in der Zeitschrift Science Advances, erschienen am 13.11.2026. Verfasst von Xiaocang Han und 14 weiteren Autor:innen. Beteiligte Einrichtung: School of Materials Science and Engineering, Peking University, Beijing 100871, China. Die Arbeit nennt 57 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: National Natural Science Foundation of China, National key R&D program of China, Beijing Natural Science Foundation.
Abstract (Originalsprache)
Atomically dispersed catalysts (ADCs) offer maximal metal utilization and local atomic/electronic structures, yet resolving their actual atomic configurations using aberration-corrected scanning transmission electron microscopy (AC-STEM) remains fundamentally constrained by depth ambiguity and complex experimental noise. Existing STEM–machine learning (ML) approaches often rely on ideal simulations or manual annotations, limiting their transferability to realistic ADC images and causing ambiguity between isolated atoms, projected dimers, and small clusters. Here, we develop a physics-informed, domain-adaptive CycleGAN-UNet framework for depth-aware ADC analysis. CycleGAN translates height-labeled simulations into experimental-like images while preserving atomic coordinates and apparent Z-height labels, and a depth-supervised UNet performs simultaneous atom localization and discrete Z-height classification. Support-specific simulations further incorporate support-dependent background contrast, including uniform amorphous carbon (am-C), am-C with local thickness variation, and crystalline supports, thereby improving the applicability of the framework to different ADC imaging conditions. Across diverse metals (Mn, Fe, Cu, Pd, and Pt) and supports (am-C, black phosphorus, and polymeric carbon nitride), the framework achieves high atom detection precision (∼95%) and reliable apparent Z-height classification at ∼2-nanometer binning. By encoding depth-dependent beam broadening features, the framework converts conventional 2D projections into quantitative, apparent Z-height atomic maps. In geminal Pt/PCN catalysts, it resolves the loading-dependent transition from isolated Pt monomers to Pt 2 sites and links these configurations to dehydrogenative aromatization activity. This framework transforms STEM images from projected atom maps into quantitative, depth-aware atomic descriptors for ADC structure–activity analysis.
Autor:innen und Institutionen
- School of Materials Science and Engineering, Peking University, Beijing 100871, China.
- Department of Chemistry, National University of Singapore, 117543, Singapore.
- DP Technology, Beijing 100080, China.
- JEOL (BEIJING) Co. Ltd., Beijing, China.
- AI for Science Institute, Beijing 100084, China.
ORCID-Profile: Xiaocang Han, Qilong Cai, Yanhui Hong, Zanlin Qiu, Jun Zhao, Fanqi Meng, Xiao Hai, Zezhong Zhang, Jiong Lu, Jin Zhang, Xiaoxu Zhao
Förderung
- National Natural Science Foundation of China
- National key R&D program of China
- Beijing Natural Science Foundation
- NRF, Prime Minister's Office, Singapore, under the Competitive Research Program Award
Zitieren
APA
Han, X., Meng, Y., Cai, Q., Hong, Y., Qiu, Z., Hu, W., Miao, S., Li, K., Zhao, J., Meng, F., Hai, X., Zhang, Z., Lu, J., Zhang, J., & Zhao, X. (2026). Deep learning–assisted electron microscopy for apparent Z-height mapping of atomically dispersed catalysts. Science Advances, 12(46). https://doi.org/10.1126/sciadv.aef7673
Vancouver
Han X, Meng Y, Cai Q, Hong Y, Qiu Z, Hu W, et al. Deep learning–assisted electron microscopy for apparent Z-height mapping of atomically dispersed catalysts. Science Advances. 2026;12(46). doi:10.1126/sciadv.aef7673
BibTeX
@article{Han2026,
author = {Han, Xiaocang and Meng, Yuan and Cai, Qilong and Hong, Yanhui and Qiu, Zanlin and Hu, Weiwei and Miao, Shu and Li, Kangshu and Zhao, Jun and Meng, Fanqi and Hai, Xiao and Zhang, Zezhong and Lu, Jiong and Zhang, Jin and Zhao, Xiaoxu},
title = {Deep learning–assisted electron microscopy for apparent Z-height mapping of atomically dispersed catalysts},
journal = {Science Advances},
year = {2026},
volume = {12},
number = {46},
doi = {10.1126/sciadv.aef7673}
}
EUVIMED