Abstract
The photosynthetic production of hydrogen peroxide (H2O2) from water and oxygen presents a sustainable alternative to the energy-intensive anthraquinone process. Covalent organic frameworks (COFs) have emerged as promising photocatalysts for H2O2 generation. However, most existing COF photocatalysts yield H2O2 at concentrations too low for practical applications, largely due to ongoing challenges in simultaneously optimizing photocatalytic activity and structural stability. Here, we introduce a large language model-driven design strategy for the targeted synthesis of high-performance COF photocatalysts. By analyzing a curated corpus of 355 peer-reviewed articles on COF-based photocatalysis with a language model-driven knowledge extraction pipeline, we extract and structure over 11,000 chemical relationships related to building block identity, linkage robustness, and H2O2 yield. Guided by this artificial intelligence-derived knowledge base, we identify 4,4′,4″-(1,3,5-triazine-2,4,6-triyl)trianiline and benzo[1,2-b:3,4-b′:5,6-b″]trithiophene-2,5,8-tricarbaldehyde as optimal building blocks and thiazole as the preferred linkage motif for constructing a robust, photocatalytic COF. The resulting Thz-COF achieve a high H2O2 concentration of 82.3 mM (~0.28 wt%) in aqueous solution (without using sacrificial agents), with a solar-to-chemical energy conversion efficiency of 1.39%.
Data availability
The data supporting the findings of this study are available within the paper and its Supplementary Information files. Source data are provided in this paper.
Code availability
All code developed and used in this study is available at https://doi.org/10.5281/zenodo.1821801662.
References
Ciriminna, R., Albanese, L., Meneguzzo, F. & Pagliaro, M. Hydrogen Peroxide: A key chemical for today’s sustainable development. ChemSusChem 9, 3374–3381 (2016).
Campos-Martin, J. M., Blanco-Brieva, G. & Fierro, J. L. G. Hydrogen peroxide synthesis: an outlook beyond the anthraquinone process. Angew. Chem. Int. Ed. 45, 6962–6984 (2006).
Shiraishi, Y. et al. Resorcinol-formaldehyde resins as metal-free semiconductor photocatalysts for solar-to-hydrogen peroxide energy conversion. Nat. Mater. 18, 985–993 (2019).
Qin, C. et al. Dual donor-acceptor covalent organic frameworks for hydrogen peroxide photosynthesis. Nat. Commun. 14, 5238 (2023).
Krishnaraj, C. et al. Strongly reducing (diarylamino)benzene-based covalent organic framework for metal-free visible light photocatalytic H2O2 generation. J. Am. Chem. Soc. 142, 20107–20116 (2020).
Geng, K. et al. Covalent organic frameworks: design, synthesis, and functions. Chem. Rev. 120, 8814–8933 (2020).
Wang, H. et al. Covalent organic framework photocatalysts: structures and applications. Chem. Soc. Rev. 49, 4135–4165 (2020).
Wang, X. et al. Sulfone-containing covalent organic frameworks for photocatalytic hydrogen evolution from water. Nat. Chem. 10, 1180–1189 (2018).
Chen, Y. et al. Hierarchical assembly of donor-acceptor covalent organic frameworks for photosynthesis of hydrogen peroxide from water and air. Nat. Synth. 3, 998–1010 (2024).
Shu, C. et al. Mixed-linker strategy for the construction of sulfone-containing D–A–A covalent organic frameworks for efficient photocatalytic hydrogen peroxide production. Angew. Chem. Int. Ed. 63, e202403926 (2024).
Liu, F. et al. Covalent organic frameworks for direct photosynthesis of hydrogen peroxide from water, air and sunlight. Nat. Commun. 14, 4344 (2023).
Yue, J.-Y. et al. Regulating the topology of covalent organic frameworks for boosting overall H2O2 photogeneration. Angew. Chem. Int. Ed. 63, e202405763 (2024).
Cheng, J. et al. Unlocking topological effects in covalent organic frameworks for high-performance photosynthesis of hydrogen peroxide. Adv. Mater. 37, 2410247 (2025).
Liu, Y. et al. Enhanced hydrogen peroxide photosynthesis in covalent organic frameworks through induced asymmetric electron distribution. Nat. Synth. 4, 134–141 (2025).
Ju, Y. et al. Regulating the electronic structure of covalent organic frameworks via heterocyclic isomers for highly efficient photocatalytic H2O2 generation. Nat. Commun. 16, 5658 (2025).
Sun, Z. et al. Simultaneous high-efficiency photocatalytic production of H2O2 and synthesis of 3-HBA using COFs broadened by π conjugate networks and photogenerated charges. ACS Appl. Mater. Interfaces 17, 25828–25838 (2025).
Alam, A. et al. Covalent organic frameworks for photocatalytic hydrogen peroxide generation. ACS Mater. Lett. 6, 2007–2049 (2024).
Abdelshafy, A. M., Qiannan, H., Zisheng, L., Zhaojun, B. & Li, L. Hydrogen peroxide from traditional sanitizer to promising disinfection agent in food industry. Food Rev. Int. 40, 658–690 (2024).
Zheng, Z. et al. Large language models for reticular chemistry. Nat. Rev. Mater. 10, 369–381 (2025).
Ramos, M. C., Collison, C. J. & White, A. D. A review of large language models and autonomous agents in chemistry. Chem. Sci. 16, 2514–2572 (2025).
Bhattacharya, D., Cassady, H. J., Hickner, M. A. & Reinhart, W. F. Large language models as molecular design engines. J. Chem. Inf. Model. 64, 7086–7096 (2024).
M Bran, A. et al. Augmenting large language models with chemistry tools. Nat. Mach. Intell. 6, 525–535 (2024).
Boiko, D. A., MacKnight, R., Kline, B. & Gomes, G. Autonomous chemical research with large language models. Nature 624, 570–578 (2023).
Darren, E. et al. From local to global: a Graph RAG approach to query-focused summarization. Preprint at https://doi.org/10.48550/arXiv.2404.16130 (2025).
Maharana, P. R., Verma, A. & Joshi, K., Retrieval augmented generation for building datasets from scientific literature. J. Phys. Mater. 8, 035006 (2025).
Na, Y. et al. Advanced scientific information mining using LLM-driven approaches in layered cathode materials for sodium-ion batteries. Mater. Adv. 6, 2543–2548 (2025).
Liu, X. et al. Design of circularly polarized phosphorescence materials guided by transfer learning. Nat. Commun. 16, 4970 (2025).
Liu, A. et al. Deepseek-v3 technical report. Preprint at https://doi.org/10.48550/arXiv.2412.19437 (2025).
Vyas, V. S. et al. A tunable azine covalent organic framework platform for visible light-induced hydrogen generation. Nat. Commun. 6, 8508 (2015).
Wei, H., Ning, J., Cao, X., Li, X. & Hao, L. Benzotrithiophene-based covalent organic frameworks: construction and structure transformation under ionothermal condition. J. Am. Chem. Soc. 140, 11618–11622 (2018).
Waller, P. J., AlFaraj, Y. S., Diercks, C. S., Jarenwattananon, N. N. & Yaghi, O. M. Conversion of imine to oxazole and thiazole linkages in covalent organic frameworks. J. Am. Chem. Soc. 140, 9099–9103 (2018).
Li, Z. et al. Three-component donor-π-acceptor covalent organic frameworks for boosting photocatalytic hydrogen evolution. J. Am. Chem. Soc. 145, 8364–8374 (2023).
Cusin, L., Peng, H., Ciesielski, A. & Samorì, P. Chemical conversion and locking of the imine linkage: enhancing the functionality of covalent organic frameworks. Angew. Chem. Int. Ed. 60, 14236–14250 (2021).
Wang, K. et al. Synthesis of stable thiazole-linked covalent organic frameworks via a multicomponent reaction. J. Am. Chem. Soc. 142, 11131–11138 (2020).
Shu, C. et al. Boosting the photocatalytic hydrogen evolution activity for D–π–A conjugated microporous polymers by statistical copolymerization. Adv. Mater. 33, 2008498 (2021).
Li, Q. et al. Shear stress triggers ultrathin-nanosheet carbon nitride assembly for photocatalytic H2O2 production coupled with selective alcohol oxidation. J. Am. Chem. Soc. 145, 20837–20848 (2023).
Wang, X.-X. et al. Reducing the exciton binding energy of covalent organic framework through π-bridges to enhance photocatalysis. Adv. Funct. Mater. 35, 2421623 (2025).
Li, X. et al. CsPbX3 quantum dots for lighting and displays: room-temperature synthesis, photoluminescence superiorities, underlying origins and white light-emitting diodes. Adv. Funct. Mater. 26, 2435–2445 (2016).
Fu, G. et al. Construction of thiadiazole-bridged sp2-carbon-conjugated covalent organic frameworks with diminished excitation binding energy toward superior photocatalysis. J. Am. Chem. Soc. 146, 1318–1325 (2024).
Xiao, Y. et al. Long-lived internal charge-separated state in two-dimensional metal–organic frameworks improving photocatalytic performance. ACS Energy Lett. 7, 2323–2330 (2022).
Yang, S. et al. Engineering imine carbon catalytic sites in covalent organic frameworks for enhanced overall H2O2 photosynthesis. J. Am. Chem. Soc. 147, 30287–30295 (2025).
Zhu, X.-G., Long, S. P. & Ort, D. R. What is the maximum efficiency with which photosynthesis can convert solar energy into biomass? Curr. Opin. Biotechnol. 19, 153–159 (2008).
Shen, M.-H. & Singh, R. K. Detoxification of aflatoxins in foods by ultraviolet irradiation, hydrogen peroxide, and their combination - a review. LWT-Food Sci. Technol. 142, 110986 (2021).
Li, Y. Biological properties of peroxide-containing tooth whiteners. Food Chem. Toxicol. 34, 887–904 (1996).
Chandra, S. et al. Chemically stable multilayered covalent organic nanosheets from covalent organic frameworks via mechanical delamination. J. Am. Chem. Soc. 135, 17853–17861 (2013).
Nosaka, Y. & Nosaka, A. Y. Generation and detection of reactive oxygen species in photocatalysis. Chem. Rev. 117, 11302–11336 (2017).
Sun, J. et al. Pyrene-based covalent organic frameworks for photocatalytic hydrogen peroxide production. Angew. Chem. Int. Ed. 62, e202216719 (2023).
Kosco, J., Moruzzi, F., Willner, B. & McCulloch, I. Photocatalysts based on organic semiconductors with tunable energy levels for solar fuel applications. Adv. Energy Mater. 10, 2001935 (2020).
Liu, R. et al. Linkage-engineered donor–acceptor covalent organic frameworks for optimal photosynthesis of hydrogen peroxide from water and air. Nat. Catal. 7, 195–206 (2024).
Wu, C. et al. Polarization engineering of covalent triazine frameworks for highly efficient photosynthesis of hydrogen peroxide from molecular oxygen and water. Adv. Mater. 34, 2110266 (2022).
Zhi, Q. et al. Piperazine-linked metalphthalocyanine frameworks for highly efficient visible-light-driven H2O2 photosynthesis. J. Am. Chem. Soc. 144, 21328–21336 (2022).
Yu, H. et al. Vinyl-groups-anchored covalent organic framework for promoted photocatalytic hydrogen peroxide generation. Angew. Chem. Int. Ed. 63, e202402297 (2024).
Chi, W. et al. A photocatalytic redox cycle over a polyimide catalyst drives efficient solar-to-H2O2 conversion. Nat. Commun. 15, 5316 (2024).
Chen, L. et al. Acetylene and diacetylene functionalized covalent triazine frameworks as metal-free photocatalysts for hydrogen peroxide production: a new two-electron water oxidation pathway. Adv. Mater. 32, 1904433 (2020).
Zhang, Y. et al. H2O2 generation from O2 and H2O on a near-infrared absorbing porphyrin supramolecular photocatalyst. Nat. Energy 8, 361–371 (2023).
Dai, T. et al. Autonomous mobile robots for exploratory synthetic chemistry. Nature 635, 890–897 (2024).
Meng, Y. et al. Preserving and combining knowledge in robotic lifelong reinforcement learning. Nat. Mach. Intell. 7, 256–269 (2025).
Yang, X. et al. Mechanistic insights into C-C coupling in electrochemical CO reduction using gold superlattices. Nat. Commun. 15, 720 (2024).
Zhang, P. et al. Inter-site structural heterogeneity induction of single atom Fe catalysts for robust oxygen reduction. Nat. Commun. 15, 2062 (2024).
Kormann, C., Bahnemann, D. W. & Hoffmann, M. R. Photocatalytic production of hydrogen peroxides and organic peroxides in aqueous suspensions of titanium dioxide, zinc oxide, and desert sand. Environ. Sci. Technol. 22, 798–806 (1988).
Hehre, W. J., Ditchfield, R. & Pople, J. A. Self-Consistent Molecular Orbital Methods. XII. Further Extensions of Gaussian-Type Basis Sets for Use in Molecular Orbital Studies of Organic Molecules. J. Chem. Phys. 56, 2257–2261 (1972).
Shu, C. et al. Synthesis of covalent organic frameworks for photocatalytic hydrogen peroxide production guided by large language models. GitHub. https://doi.org/10.5281/zenodo.18218016 (2026).
Acknowledgements
We acknowledge the assistance of Dr. Lirong Zheng for XAFS analysis. Shu acknowledges the assistance of Dr. Zhentao Fu for the theoretical calculation. This work was financially supported by the National Natural Science Foundation of China (Grant No. 52203259; Grant No. 22575094; Grant No. 22475076; Grant. No. 22204054), National Key R&D Program of China (Grant No. 2023YFB4004800), and Fundamental Research Funds for the Central Universities, HUST (Grant No. 2024JYCXJJ041). The AI-driven experiments, simulations and model training were performed on the robotic AI-Scientist platform of the Chinese Academy of Sciences. L.C. acknowledges the University of Science and Technology of China Startup Program (KY9990000207) for funding. The icons have been designed using resources from Flaticon.com.
Author information
Authors and Affiliations
Contributions
Y.X.W., L.C., and B.T. conceived the project. C.S. performed the materials synthesis, characterization and photocatalytic experiments. L.W. performed the LLM experiments. X.J.Y. and X.Y. performed the ATR-SEIRAS experiments. X.W. and J.R. conducted the sterilization experiments. W.X. and P.X. helped with the electronic experiments. K.W. provided helpful discussions. All the authors participated in discussions of the research.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Communications thanks Bolong Huang, Cuijuan Wang and the other anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Source data
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Shu, C., Wang, L., Yang, X. et al. Synthesis of covalent organic frameworks for photocatalytic hydrogen peroxide production guided by large language models. Nat Commun (2026). https://doi.org/10.1038/s41467-026-69549-z
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41467-026-69549-z
