CDS&E: D3SC: Understanding and Discovering Two-Dimensional Electrocatalyts Using Grand-Canonical Density Functional Theory (DFT) and Machine Learning

可再生能源的一个主要挑战是电能和化学能的低效相互转换。需要新的电催化剂来解决这一挑战。二维(2D)材料只有一个或几个原子厚，因此，它们是理想的、高比表面积的能量转换电催化剂。在这个项目中，刘元月博士正在使用计算机模拟来提供对2D材料作为电催化剂的能力的基本了解。他的模拟使用人工智能来加速发现过程。刘博士正在对学生进行计算化学、催化和能量转换方面的培训，方法是让他们积极参与多学科研究，并将他的研究成果整合到课堂上。刘博士正在与技术学院工作室合作，制作说明电催化应用的教育视频。这些视频被用于各种推广计划，以提高K-12学生和未被充分代表的群体追求科学、技术、工程和数学(STEM)职业的兴趣。在化学系化学催化计划的资助下，德克萨斯大学奥斯汀分校的刘元月博士正在加深对2D材料(一种新兴的电催化剂)的电催化机理的理解。通过执行巨正则密度泛函理论(GC-DFT)计算，并将计算与机器学习相结合，该团队寻找性能更好的新2D电催化剂。传统的密度泛函理论计算往往忽略了电化学反应中电荷变化和恒电位的影响，然而这些参数对二维材料来说是至关重要的。GC-DFT包含了这些效应，因此可以解释2D电催化剂对放氢、放氧和二氧化碳还原等反应的机理。机器学习模型用于揭示控制催化活性和选择性的潜在因素。这些模型还预测大型2D材料库会有更好的催化剂。这一奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

A major challenge in renewable energy is the inefficient interconversion of electrical and chemical energy. New electrocatalysts are needed to solve this challenge. Two-dimensional (2D) materials are only one or a few atoms thick and as such, they serve as desirable, high surface area electrocatalysts for energy conversions. In this project, Dr. Yuanyue Liu is using computer simulations to provide a fundamental understanding of the capabilities of 2D materials as electrocatalysts. His simulations use artificial intelligence to accelerate the discovery process. Dr. Liu is training students in computational chemistry, catalysis and energy conversions by actively engaging them in multidisciplinary research and integrating his research findings into the classroom. Dr. Liu is working with the Faculty Technology Studio to create educational videos that illustrate electrocatalysis applications. These videos are being used in various outreach programs to increase the interest of K-12 students and underrepresented groups in pursuing science, technology, engineering and mathematics (STEM) careers.With funding from the Chemical Catalysis Program of the Chemistry Division, Dr. Yuanyue Liu at the University of Texas at Austin is developing an improved understanding of the electrocatalytic mechanisms of 2D materials (an emerging class of electrocatalysts). By performing grand-canonical density functional theory (gc-DFT) calculations and coupling the calculations with machine learning, the team searches for new 2D electrocatalysts with better performance. Conventional DFT calculations often neglect the effects of varying charge and constant potential in electrochemical reactions, yet these parameters are found to be critical for 2D materials. The gc-DFT includes these effects and thus can elucidate the mechanisms of 2D electrocatalysts for reactions such as hydrogen evolution, oxygen reduction/evolution, and carbon dioxide reduction. Machine learning models are used to uncover the underlying factors that govern the catalytic activity and selectivity. The models also predict better catalysts from the large 2D materials library.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.