CAREER: Bayesian Model of Chemisorption for Adsorbate-Specific Tuning of Electrocatalysis

职业：用于电催化吸附质特异性调节的化学吸附贝叶斯模型

• 批准号: 1845531
• 负责人: Hongliang Xin
• 金额: $ 54.95万
• 依托单位: Virginia Polytechnic Institute and State University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2019
• 资助国家: 美国
• 起止时间: 2019-01-01 至 2024-03-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1845531&HistoricalAwards=false
• 关键词: CAREER; Bayesian; Model; Chemisorption; Adsorbate

Ammonia (NH3) is best known as a starting material for fertilizers, but its reaction with oxygen (called oxidation) is required in applications such as ammonia sensing, wastewater treatment, and direct ammonia fuel cells - all of which are carried out electrochemically, and usually assisted by a catalyst material called an electrocatalyst. Even with state-of-the-art platinum-based electrocatalysts, the oxidation reaction is inefficient and requires excess electrical energy. The project will investigate, through theoretical and computational means, the possibility of improving both the energy efficiency and the rate of electrochemical ammonia oxidation by combining platinum with other metals in nano-scale particles known as nano-alloys. The predicted nano-alloy compositions will help guide the design of more efficient electrocatalysts, not only for ammonia-related applications, but also for a broad range of energy and environmental technologies. The project also integrates research with educational and outreach initiatives designed to excite high-school students about STEM opportunities and train undergraduate and graduate students in the application of computer models for energy security and environmental stewardship.Electrocatalytic reactions at the core of artificial photosynthesis involve multiple proton-coupled electron transfer steps. Arguably, for a given type of catalysts, e.g., d-block metals, the scaling relations among adsorption energies of atoms and their hydrogenated species limit the efficiency of electrical/chemical energy conversion. To overcome those obstacles for the ammonia oxidation reaction, the project will utilize a Bayesian framework for advancing the orbital-level understanding of adsorbate-surface interactions and catalytic processes at the metal-electrolyte interfaces, paving the path toward adsorbate-specific tuning of electrocatalysis. The free formation energies of key reaction species will be selectively tuned via orbital-wise perturbation of chemical bonding, e.g., nano-alloying, such that the activation barrier of the rate-limiting N-N bond formation or N-H cleavage step is reduced without poisoning the surface with adsorbed N adatoms. Catalysis theory, quantum chemistry, and machine learning will be combined to unravel atomistic mechanisms of sluggish NH3 electro-oxidation kinetics and develop the Bayesian model of chemisorption with machine-learned Hamiltonians. Modulation of adsorbed species by engineering their interactions with atomically-tailored metal sites guided by the Bayesian models will further advance the theory of chemisorption and its applications in catalysis, enabling design of catalytic systems with physically interpretable insights rather than trial-and-error searches. The educational component of this CAREER plan aims to further develop the informatics for photon harvesting at nano-engineered structures, via a mobile device application, iPhanes, developed by the investigator. This effort will energize student learning using materials informatics on mobile devices, demonstrate a multidisciplinary perspective of energy issues, and stimulate the students' collaborative learning via materials design projects. This "experiment" will enhance recruitment and retention of women, minorities, and persons with disabilities in STEM fields, and will motivate the students towards lifelong learning and careers related to advanced renewable energy and environmental technologies.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

氨（NH3）是最著名的肥料原料，但它与氧的反应（称为氧化）在氨传感、废水处理和直接氨燃料电池等应用中是必需的——所有这些都是通过电化学进行的，通常由一种叫做电催化剂的催化剂材料辅助。即使使用最先进的铂基电催化剂，氧化反应也是低效的，而且需要多余的电能。该项目将通过理论和计算手段，通过将铂与其他金属结合在纳米级颗粒（即纳米合金）中，来研究提高能源效率和电化学氨氧化速率的可能性。预测的纳米合金成分将有助于指导设计更高效的电催化剂，不仅适用于氨相关应用，而且适用于广泛的能源和环境技术。该项目还将研究与教育和推广活动相结合，旨在激发高中生对STEM机会的兴趣，并培训本科生和研究生将计算机模型应用于能源安全和环境管理。人工光合作用的核心电催化反应涉及多个质子耦合电子转移步骤。可以说，对于特定类型的催化剂，例如d-嵌段金属，原子及其氢化物质的吸附能之间的标度关系限制了电能/化学能转换的效率。为了克服氨氧化反应的这些障碍，该项目将利用贝叶斯框架来推进对吸附物表面相互作用和金属-电解质界面催化过程的轨道级理解，为吸附物特异性电催化调节铺平道路。关键反应物质的自由生成能将通过化学键的轨道扰动选择性地调整，例如，纳米合金化，从而降低限速N-N键形成或N- h裂解步骤的激活势垒，而不会用吸附的N原子毒害表面。催化理论、量子化学和机器学习将结合起来，揭示缓慢NH3电氧化动力学的原子机制，并利用机器学习的哈密顿量建立化学吸附的贝叶斯模型。在贝叶斯模型的指导下，通过设计吸附物质与原子定制金属位点的相互作用来调节吸附物质，将进一步推进化学吸附理论及其在催化中的应用，使设计具有物理可解释见解的催化系统成为可能，而不是反复试验。该职业计划的教育部分旨在通过研究者开发的移动设备应用程序iphone，进一步发展纳米工程结构中光子收集的信息学。这一努力将激发学生在移动设备上使用材料信息学的学习，展示能源问题的多学科视角，并通过材料设计项目激发学生的协作学习。这项“实验”将加强STEM领域对女性、少数族裔和残疾人的招聘和保留，并将激励学生终身学习和从事与先进可再生能源和环境技术相关的职业。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

项目成果

Machine learning of lateral adsorbate interactions in surface reaction kinetics

表面反应动力学中横向吸附质相互作用的机器学习

• DOI: 10.1016/j.coche.2022.100825
• 发表时间: 2022
• 期刊: Current Opinion in Chemical Engineering
• 影响因子: 6.6
• 作者: Mou, Tianyou;Han, Xue;Zhu, Huiyuan;Xin, Hongliang
• 通讯作者: Xin, Hongliang

Interpretable Machine Learning for Catalytic Materials Design toward Sustainability

• DOI: 10.1021/accountsmr.3c00131
• 发表时间: 2023-11
• 期刊: Accounts of Materials Research
• 影响因子: 14.6
• 作者: Hongliang Xin;Tianyou Mou;H. Pillai;Shih-Han Wang;Yang Huang

Ternary PtIrNi Catalysts for Efficient Electrochemical Ammonia Oxidation

• DOI: 10.1021/acscatal.9b04670
• 发表时间: 2020-04-03
• 期刊: ACS CATALYSIS
• 影响因子: 12.9
• 作者: Li, Yi;Li, Xing;Wu, Gang
• 通讯作者: Wu, Gang

Catalyst design with machine learning

• DOI: 10.1038/s41560-022-01112-8
• 发表时间: 2022-09
• 期刊: Nature Energy
• 影响因子: 56.7
• 作者: Hongliang Xin

Identification of Active Sites for Ammonia Electrosynthesis on Ruthenium

• DOI: 10.1021/acsenergylett.2c02175
• 发表时间: 2022-11
• 期刊: ACS Energy Letters
• 影响因子: 22
• 作者: Lin Hu;H. Pillai;Corbin Feit;Kaige Shi;Zhengning Gao;P. Banerjee;Hongliang Xin;Xiaofeng Feng