大语言模型在电催化领域中的应用

doi:10.1016/S1872-2067(23)64612-1

催化学报 ›› 2024, Vol. 59: 7-14.DOI: 10.1016/S1872-2067(23)64612-1

大语言模型在电催化领域中的应用

张成翼^,¹, 王兴宇^,¹, 王子运^*()

奥克兰大学化学院, 奥克兰, 新西兰

收稿日期:2023-12-13 接受日期:2024-01-24 出版日期:2024-04-18 发布日期:2024-04-15
通讯作者: *电子信箱: ziyun.wang@auckland.ac.nz (王子运).
作者简介:
¹共同第一作者.

Large language model in electrocatalysis

Chengyi Zhang^,¹, Xingyu Wang^,¹, Ziyun Wang^*()

School of Chemical Sciences, University of Auckland, Auckland 1010, New Zealand

Received:2023-12-13 Accepted:2024-01-24 Online:2024-04-18 Published:2024-04-15
Contact: *E-mail: ziyun.wang@auckland.ac.nz (Z. Wang).
About author:Ziyun Wang (University of Auckland) is an emerging researcher specializing in computational chemistry and theoretical catalysis. He conducted his doctoral research in theoretical catalysis at the Queen’s University of Belfast under the guidance of Prof. Peijun Hu and Prof. Chris Hardacre. This was followed by postdoctoral research with Prof. Jens Nørskov and Prof. Thomas Bligaard at Stanford University, and with Prof. Edward Sargent at the University of Toronto. In 2021, Dr. Wang joined the School of Chemical Sciences at the University of Auckland, where he is currently a Senior Lecturer.
¹Contributed equally to this work.

摘要/Abstract

摘要：

大语言模型作为现代科技的杰出代表, 已经在社会的各个领域产生了深远影响. 这些模型由数十亿个神经元相互连接构建而成, 集成了人类至今为止的丰富知识, 具备与全球各地人们进行流畅、精确对话的卓越能力. 其类人智能特性不仅使其能够应对现代社会的各种挑战, 更在多个科学领域中展现出巨大的应用前景. 然而, 目前大语言模型广泛应用的潜力尚未被完全挖掘. 本文旨在深入探讨现代大语言模型在电催化领域的潜在应用, 以期阐明这些人工智能驱动的模型如何助力科研人员深化对催化学科的理解, 并推动催化剂的智能设计. 通过对大语言模型在电催化领域应用的系统研究, 我们期望能够为未来的科研工作提供新的思路和方法.

本文首先分析了传统的实验法和多尺度模拟方法在催化研究中的局限性, 包括资源消耗大、实验周期长、研究进展缓慢以及受人类自身能力限制等问题. 随后, 着重探讨了大语言模型在电催化研究中的显著优势, 如突破人类认知界限、理论上具备无限知识累积能力等. 尽管大语言模型展现出巨大的潜力, 但仍面临泛化能力与领域专业性之间的权衡、文本限制等挑战. 为应对这些挑战, 本文进一步介绍了多模态大语言模型的发展及其在电催化研究中的具体应用, 包括与实验者的直接互动、基于实验反馈的持续优化、预训练模型的微调以及结合视觉编码器实现多模态数据融合等. 最后, 强调了多模态大语言模型在光谱分析、实验路径设计、过渡状态搜索、分子结构设计、催化剂优化和问题诊断等方面的巨大潜力. 总之, 多模态方法在催化领域具有广泛的应用前景, 能够整合多种数据源, 为催化剂设计、反应机理研究和反应条件优化提供强大的工具和技术支持.

最后, 本文展望了大语言模型在科学研究中的作用和未来发展趋势. 大语言模型在知识积累和效率提升方面展现出了出色的能力, 但在创造全新知识方面, 或许还无法与科学先驱相提并论. 展望未来, 大语言模型应当与实验和模拟方法深度融合, 以增强其预测能力和多模态学习能力. 通过这样的结合, 大语言模型将能够更全面地辅助科学研究者, 进而推动科学研究更迅速地发展. 这一发展趋势不仅有助于提升科学研究的效率和准确性, 还将为科学领域带来更多的创新和突破.

关键词: 大语言模型, 电催化, 人工智能, 多模态大语言模型

Abstract:

Large language models have recently brought a massive storm on modern society in all fields. While many view them as mere search engines for specific answers or text refinement tools like a chatbot, their broader applications remain largely unexplored. These large language models, consisting of billions of interconnected neurons, derived from all knowledge of the human, possess the remarkable ability to engage in smooth and precise conversations with individuals across the globe. Human-like intelligence enables them to address modern challenges and display immense potential in various scientific domains. In this perspective, we delve into the potential applications of modern large language model and its future iterations within the field of catalysis, aiming to shed light on how these AI-driven models can contribute to a deeper understanding of catalysis science and the intelligent design of catalysts.

Key words: Large language model, Electrocatalysis, Artificial intelligence, Multimodal large language model

张成翼, 王兴宇, 王子运. 大语言模型在电催化领域中的应用[J]. 催化学报, 2024, 59: 7-14.

Chengyi Zhang, Xingyu Wang, Ziyun Wang. Large language model in electrocatalysis[J]. Chinese Journal of Catalysis, 2024, 59: 7-14.

×
扫码分享

导出引用管理器 EndNote|Ris|BibTeX

链接本文: https://www.cjcatal.com/CN/10.1016/S1872-2067(23)64612-1

https://www.cjcatal.com/CN/Y2024/V59/I4/7

图/表 2

Fig. 1. The potential applications of LLM in catalysis research. Including comprehensive guidance for designing catalysts for various catalytic reactions (OER, HER, CO2RR, and NRR), facilitating the understanding of reaction mechanisms, assisting in extracting key data from electrochemical tests, compiling experimental reports based on previous experimental results, and improving demonstration schemes based on experimental outcomes to guide experimental design.

Fig. 2. The various types of inputs and outputs in the future MLLM in catalysis. Including the combination of the graphs and language input of the spectroscopy analysis results, combinations of various figures and descriptions of previous literature for the experiments reports, and combinations of characterization data and experiments input for problem diagnosis for further improvement.

参考文献

[1]	Z. W. Seh, J. Kibsgaard, C. F. Dickens, I. Chorkendorff, J. K. Norskov, T. F. Jaramillo, Science, 2017, 355, eaad4998.
[2]	W. Kohn, L. J. Sham, Phys. Rev., 1965, 140, A1133-A1138.
[3]	I. A. Hamad, M. A. Novotny, D. O. Wipf, P. A. Rikvold, Phys. Chem. Chem. Phys., 2010, 12, 2740-2743.
[4]	R. Car, M. Parrinello, Phys. Rev. Lett., 1985, 55, 2471-2474.
[5]	R. E. Rudd, J. Q. Broughton, Phys. Rev. B, 1998, 58, R5893-R5896.
[6]	T. Abe, J. Comput. Phys., 1997, 131, 241-246.
[7]	L. Börgesson, in: Developments in Geotechnical Engineering, O. Stephansson, L. Jing, C.-F. Tsang eds., Elsevier, Netherlands, 1996, 565-570
[8]	F. Liu, T. Zhu, X. Wu, B. Yang, C. You, C. Wang, L. Lu, Z. Liu, Y. Zheng, X. Sun, Y. Yang, L. Clifton, D. A. Clifton, NPJ Digit. Med., 2023, 6, 226.
[9]	L. De Angelis, F. Baglivo, G. Arzilli, G. P. Privitera, P. Ferragina, A. E. Tozzi, C. Rizzo, Front. Public Health, 2023, 11, 1166120.
[10]	R. Luo, L. Sun, Y. Xia, T. Qin, S. Zhang, H. Poon, T. Y. Liu, Brief. Bioinformatics, 2022, 23, 10.1093/bib/bbac409.
[11]	H. Touvron, T. Lavril, G. Izacard, X. Martinet, M.-A. Lachaux, T. Lacroix, B. Rozière, N. Goyal, E. Hambro, F. Azhar, A. Rodriguez, A. Joulin, E. Grave, G. Lample, arXiv preprint arXiv:2302.13971, 2023.
[12]	H. Touvron, L. Martin, K. Stone, P. Albert, A. Almahairi, Y. Babaei, N. Bashlykov, S. Batra, P. Bhargava, S. Bhosale, D. Bikel, L. Blecher, C. C. Ferrer, M. Chen, G. Cucurull, D. Esiobu, J. Fernandes, J. Fu, W. Fu, B. Fuller, C. Gao, V. Goswami, N. Goyal, A. Hartshorn, S. Hosseini, R. Hou, H. Inan, M. Kardas, V. Kerkez, M. Khabsa, I. Kloumann, A. Korenev, P. S. Koura, M.-A. Lachaux, T. Lavril, J. Lee, D. Liskovich, Y. Lu, Y. Mao, X. Martinet, T. Mihaylov, P. Mishra, I. Molybog, Y. Nie, A. Poulton, J. Reizenstein, R. Rungta, K. Saladi, A. Schelten, R. Silva, E. M. Smith, R. Subramanian, X. E. Tan, B. Tang, R. Taylor, A. Williams, J. X. Kuan, P. Xu, Z. Yan, I. Zarov, Y. Zhang, A. Fan, M. K., S. Narang, A. Rodriguez, R. Stojnic, S. Edunov, T. Scialom, arXiv preprint arXiv:2307.09288, 2023.
[13]	P. Perniss, Front. Psychol., 2018, 9, 1109-1109.
[14]	C. Stokel-Walker, R. Van Noorden, Nature, 2023, 614, 214-216.
[15]	Z. Zheng, Z. Rong, N. Rampal, C. Borgs, J. T. Chayes, O. M. Yaghi, Angew. Chem. Int. Ed., 2023, 62, e202311983.
[16]	Z. Wang, W. Liu, Q. He, X. Wu, Z. Yi, arXiv preprint arXiv:2203.00386, 2022.
[17]	O. Topsakal, T. C. Akinci, Creating large language model applications utilizing langchain: A primer on developing llm apps fast, Proceedings of the International Conference on Applied Engineering and Natural Sciences, Konya, Turkey, 2023, 1, 1050-1056.
[18]	M. Liu, A. Grinberg Dana, M. S. Johnson, M. J. Goldman, A. Jocher, A. M. Payne, C. A. Grambow, K. Han, N. W. Yee, E. J. Mazeau, K. Blondal, R. H. West, C. F. Goldsmith, W. H. Green, J. Chem. Inf. Model., 2021, 61, 2686-2696.
[19]	T. Weiss, E. Mayo Yanes, S. Chakraborty, L. Cosmo, A. M. Bronstein, R. Gershoni-Poranne, Nat. Comput. Sci., 2023, 3, 873-882.
[20]	P. K. Rubenstein, C. Asawaroengchai, D. D. Nguyen, A. Bapna, Z. Borsos, F. d. C. Quitry, P. Chen, D. E. Badawy, W. Han, E. Kharitonov, H. Muckenhirn, D. Padfield, J. Qin, D. Rozenberg, T. Sainath, J. Schalkwyk, M. Sharifi, M. T. Ramanovich, M. Tagliasacchi, A. Tudor, M. Velimirović, D. Vincent, J. Yu, Y. Wang, V. Zayats, N. Zeghidour, Y. Zhang, Z. Zhang, L. Zilka, C. Frank, arXiv preprint arXiv:2306.12925, 2023.

大语言模型在电催化领域中的应用

Large language model in electrocatalysis

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

扫码分享

图/表 2

参考文献

相关文章 15

编辑推荐

Metrics

[1]	聂翼飞, 颜红萍, 鹿苏微, 张宏伟, 齐婷婷, 梁诗景, 江莉龙. 理论指导构建Cu-O-Ti-O_v活性位点及其高效电催化还原硝酸根研究[J]. 催化学报, 2024, 59(4): 293-302.
[2]	王超琛, 葛旺鑫, 唐雷, 齐宴宾, 董磊, 江宏亮, 沈建华, 朱以华, 李春忠. 单原子修饰原子簇的多配位Cu基催化剂上CO₂高选择性电还原CO的研究[J]. 催化学报, 2024, 59(4): 324-333.
[3]	宋宁, 江吉周, 洪士欢, 王赟, 李春梅, 董红军. 以金属有机骨架为源制备单原子电催化剂用于能量转换的最新进展[J]. 催化学报, 2024, 59(4): 38-81.
[4]	谢逸, 徐湛友, 卢千, 王莹. 构建高效稳定的低温反向偏置双极膜电解槽用于二氧化碳还原[J]. 催化学报, 2024, 59(4): 82-96.
[5]	蒋远, 杨级, 李沐霖, 王雪佳, 杨娜, 陈伟平, 董金超, 李剑锋. 揭示明确定义的金属-N₄位点在电催化硝酸盐还原中的活性趋势[J]. 催化学报, 2024, 59(4): 195-203.
[6]	廖伟, 周乾, 龙瑾, 吴臣中, 王彬, 彭琼, 曹建新, 王青梅. 碳载体空位工程助力原位合成的Pt纳米枝晶促进电催化氧气还原[J]. 催化学报, 2024, 59(4): 260-271.
[7]	陈丽丽, 郝彦衡, 褚健意, 刘松, 白凤华, 罗文豪. 电催化硝酸根还原合成氨: 关于铁/铜基催化剂的研究展望[J]. 催化学报, 2024, 58(3): 25-36.
[8]	罗健颖, 傅捷, 叶丰铭, 梁汉锋, 王伟俊. 释放多孔纳米结构的活力: 可持续电催化水分解[J]. 催化学报, 2024, 58(3): 37-85.
[9]	屠振涛, 何晓洋, 刘璇, 熊登科, 左娟, 吴德礼, 汪建营, 陈作锋. 通过钨对镍活性位点的电子改性实现苯甲胺选择性氧化耦合制氢[J]. 催化学报, 2024, 58(3): 146-156.
[10]	郭一博, 薛圆媛, 周震. 铠甲催化剂助力锌-空气电池: 氮掺杂石墨烯上超薄碳封装铁镍合金增强氧电催化[J]. 催化学报, 2024, 58(3): 206-215.
[11]	李奇, 李杰浩, 白惠敏, 李发堂. 催化剂晶面工程在光/电催化中的研究进展[J]. 催化学报, 2024, 58(3): 86-104.
[12]	刘佳, 王式彬, 蔡锦福, 武立振, 刘云, 贺佳辉, 许在祥, 彭小革, 钟兴, 安亮, 王建国. 高活性方形氧化铅与可视化电解槽协同促进电催化臭氧生产[J]. 催化学报, 2024, 57(2): 80-95.
[13]	别泉泉, 尹海波, 王云龙, 苏海伟, 彭悦, 李俊华. Cu单原子与氧空位协同作用增强电催化CO₂还原中C₂液相产物活性[J]. 催化学报, 2024, 57(2): 96-104.
[14]	蔡豪, 陈芳, 胡程, 葛伟怡, 李彤, 张晓磊, 黄洪伟. 富含氧空位的Bi₄O₅Br₂超薄纳米片用于高效压电催化生成过氧化氢[J]. 催化学报, 2024, 57(2): 123-132.
[15]	李伊萍, 王谭源, 姚璋懿, 陈麒安, 李箐. 非金属掺杂提升铂族金属电催化剂性能[J]. 催化学报, 2024, 56(1): 51-73.