Domain‐Trained Language Model for Inverse Design and Synthesis of High‐Performance Hydrogen Storage MOFs

Z Zhimeng Liu Y Yuqiao Su (Beijing Key Laboratory of Function Materials for Molecule & Structure Construction, School of Materials Science and Engineering University of Science and Technology Beijing Beijing 100083 P.R. China) H Hao Wang (Division of Quantitative Sciences, Department of Oncology Johns Hopkins University School of Medicine Baltimore Maryland USA) T Tao Ban (State Key Laboratory of Fluorine & Nitrogen Chemicals School of Chemical Engineering and Technology Xi'an Jiaotong University Xi'an 710049 China) L Lingmeng Wang (Beijing Key Laboratory of Function Materials for Molecule & Structure Construction, School of Materials Science and Engineering University of Science and Technology Beijing Beijing 100083 P.R. China) S Shaopeng Lu (Beijing Key Laboratory of Function Materials for Molecule & Structure Construction, School of Materials Science and Engineering University of Science and Technology Beijing Beijing 100083 P.R. China) Z Zuoshuai Xi (Beijing Key Laboratory of Function Materials for Molecule & Structure Construction, School of Materials Science and Engineering University of Science and Technology Beijing Beijing 100083 P.R. China) W Wenqing Li Y Yujie Guo (CAS Key Laboratory of Molecular Nanostructure and Nanotechnology and Beijing National Laboratory for Molecular Sciences) C Changan Wang (Beijing Key Laboratory of Function Materials for Molecule & Structure Construction, School of Materials Science and Engineering University of Science and Technology Beijing Beijing 100083 P.R. China) X Xiaoqi Wang (School of History) X Xu Jin H Hongyi Gao (Department of Pathology, Guangdong Women and Children Hospital) G Ge Wang

Abstract

Abstract A domain‐specific large language model, MOFs‐LLM, is developed to accelerate the inverse design and synthesis of metal—organic frameworks (MOFs) for hydrogen storage. Trained on 210 million tokens derived from over 6 000 MOF‐related publications and 15 000 crystal structures, the model integrates chemical knowledge with structural features to improve structure–property reasoning. Compared to baseline methods, MOFs‐LLM achieves a 46.7% enhancement in capturing structure–property relationships. It enables the inverse design of 60 candidate frameworks optimized for both hydrogen storage performance and synthetic accessibility. Guided by the model, a novel MOF (Cu‐LLMs‐1) was synthesized in three experimental iterations, exhibiting a hydrogen uptake of 1.33 wt% at room temperature, ranking among the top five pure MOFs under comparable conditions. These findings highlight the potential of domain‐trained language models to bridge virtual screening and experimental realization in materials discovery.

Article Details

Volume / Issue Vol. 65, Issue 2
Published January 09, 2026
ISSN 1433-7851
Publisher Wiley

Journal Info

Angewandte Chemie International Edition

Wiley

ISSN: 1433-7851 Physical Sciences

Authors (14)

Z

Zhimeng Liu

Y

Yuqiao Su

Beijing Key Laboratory of Function Materials for Molecule & Structure Construction, School of Materials Science and Engineering University of Science and Technology Beijing Beijing 100083 P.R. China

H

Hao Wang

Division of Quantitative Sciences, Department of Oncology Johns Hopkins University School of Medicine Baltimore Maryland USA

T

Tao Ban

State Key Laboratory of Fluorine & Nitrogen Chemicals School of Chemical Engineering and Technology Xi'an Jiaotong University Xi'an 710049 China

L

Lingmeng Wang

Beijing Key Laboratory of Function Materials for Molecule & Structure Construction, School of Materials Science and Engineering University of Science and Technology Beijing Beijing 100083 P.R. China

S

Shaopeng Lu

Beijing Key Laboratory of Function Materials for Molecule & Structure Construction, School of Materials Science and Engineering University of Science and Technology Beijing Beijing 100083 P.R. China

Z

Zuoshuai Xi

Beijing Key Laboratory of Function Materials for Molecule & Structure Construction, School of Materials Science and Engineering University of Science and Technology Beijing Beijing 100083 P.R. China

W

Wenqing Li

Y

Yujie Guo

CAS Key Laboratory of Molecular Nanostructure and Nanotechnology and Beijing National Laboratory for Molecular Sciences

C

Changan Wang

Beijing Key Laboratory of Function Materials for Molecule & Structure Construction, School of Materials Science and Engineering University of Science and Technology Beijing Beijing 100083 P.R. China

X

Xiaoqi Wang

School of History

X

Xu Jin

H

Hongyi Gao

Department of Pathology, Guangdong Women and Children Hospital

G

Ge Wang