Database for machine learning of hydrogen storage materials properties

<strong>Database for machine learning of hydrogen storage materials properties</strong> Matthew Witman^a, Mark Allendorf^a, Vitalie Stavila^a ^aSandia National Laboratories, Livermore, CA <strong>Description</strong> This ML-HydPARK dataset provides a csv file of metal hydride compositions, capacities, and thermodynamic values that can be used as target properties for building, training, and testing machine learning models. It has been parsed and cleaned from the DOE’s original publicly available HydPARK database according to the procedure in [1] to make it more suitable for immediate use with data-driven models. Generally, this removed duplicate entries, removed entries missing critical data, and attempted to fix various entries with obvious errors in the data. It is continuously updated under version control as new metal alloy hydrides are published in the open literature. Most entries contain data on the enthalpy and entropy of the hydriding reaction, as well the maximum hydrogen capacity, for which compositional machine learning models can be trained [1,2]. <strong>Acknowledgements</strong> The authors gratefully acknowledge research support from the U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, Fuel Cell Technologies Office through the Hydrogen Storage Materials Advanced Research Consortium (HyMARC). This work was supported by the Laboratory Directed Research and Development (LDRD) program at Sandia National Laboratories. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology &amp; Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525. This paper describes objective technical results and analysis. Any subjective views or opinions that might be expressed in the paper do not necessarily represent the views of the U.S. Department of Energy<br> or the United States Government. <strong>References</strong> Witman, M.; Ling, S.; Grant, D. M.; Walker, G. S.; Agarwal, S.; Stavila, V.; Allendorf, M. D. Extracting an Empirical Intermetallic Hydride Design Principle from Limited Data via Interpretable Machine Learning. <em>J. Phys. Chem. Lett</em>. <strong>2020</strong>, 11, 40–47. Witman, M.; Ek, G.; Ling, S.; Chames, J.; Agarwal, S.; Wong, J.; Allendorf, M. D.; Sahlberg, M.; Stavila, V. Data-Driven Discovery and Synthesis of High Entropy Alloy Hydrides with Targeted Thermodynamic Stability. <em>Chem. Mater</em>. <strong>2021</strong>, 33, 4067–4076. <strong>Contact</strong> Please email [e-mail] , [e-mail], or [e-mail] for questions or to request addition of recent data from the literature to this dataset.<strong>储氢材料性质机器学习专用数据库</strong><br>作者：马修·威特曼^a、马克·阿伦多夫^a、维塔利·斯塔维拉^a<br>^a桑迪亚国家实验室，美国加利福尼亚州利弗莫尔市<br><br><strong>数据集说明</strong><br>本ML-HydPARK数据集提供了金属氢化物成分、储氢容量以及热力学数值的CSV（Comma-Separated Values，逗号分隔值）文件，可作为构建、训练与测试机器学习模型的目标性质数据集。本数据集依据文献[1]中的流程，从美国能源部（DOE，U.S. Department of Energy）公开的原始HydPARK数据库中解析并清洗得到，更适合直接用于数据驱动模型的开发。整体而言，清洗工作移除了重复条目、缺失关键数据的条目，并修正了一批存在明显数据错误的条目。随着公开文献中新型金属合金氢化物的发表，本数据集将通过版本控制系统持续更新。绝大多数条目包含氢化反应的焓变、熵变以及最大储氢容量数据，可用于训练基于成分的机器学习模型[1,2]。<br><br><strong>致谢</strong><br>作者感谢美国能源部（U.S. Department of Energy, DOE）能源效率与可再生能源办公室燃料电池技术办公室通过储氢材料先进研究联盟（HyMARC, Hydrogen Storage Materials Advanced Research Consortium）提供的研究支持。本研究得到桑迪亚国家实验室实验室导向研发（LDRD, Laboratory Directed Research and Development）项目的资助。桑迪亚国家实验室是由霍尼韦尔国际公司全资子公司桑迪亚国家技术与工程解决方案有限责任公司（National Technology & Engineering Solutions of Sandia, LLC）为美国能源部国家核安全管理局按照DE-NA0003525号合同管理运营的多任务实验室。本文仅阐述客观的技术结果与分析，文中所表达的任何主观观点均不代表美国能源部或美国政府的立场。<br><br><strong>参考文献</strong><br>[1] Witman M, Ling S, Grant D M, Walker G S, Agarwal S, Stavila V, Allendorf M D. 基于可解释机器学习从有限数据中提取金属间氢化物经验设计原则[J]. 《物理化学快报》（The Journal of Physical Chemistry Letters）, 2020, 11: 40–47.<br>[2] Witman M, Ek G, Ling S, Chames J, Agarwal S, Wong J, Allendorf M D, Sahlberg M, Stavila V. 面向目标热力学稳定性的高熵合金氢化物数据驱动发现与合成[J]. 《材料化学》（Chemistry of Materials）, 2021, 33: 4067–4076.<br><br><strong>联系方式</strong><br>如有疑问或希望将最新公开文献中的数据添加至本数据集，请发送邮件至[e-mail]、[e-mail]或[e-mail]。