Over the past decade, extensive research into single-atom catalysts (SACs) has revealed that the catalytic behavior of metal adatoms is highly dependent on how they interact with their support. A strong dependence on the local coordination environment has led to comparisons with metal-organic complexes, and there is growing excitement about the potential to fine-tune SACs by controlling the adsorption geometry. The rise of computational screening to identify the optimal support-metal combinations underscores the need for rigorous benchmarking of theoretical methods, to validate realistic geometries, mechanisms, and the impact of adsorption on stability and catalytic activity. The surface science approach is particularly well-suited for this task because it allows to precisely determine the geometry of the metal atom and interpret its catalytic behavior. Moreover, the effects of temperature and molecular adsorption on the model catalysts stability can be studied in isolation, and conclusions drawn from UHV studies tested in increasingly common near-ambient pressure and electrochemical setups. This perspective highlights recent breakthroughs and specific systems, including metal oxides, metal-organic frameworks, and carbon nitrides, where insights from surface science experiments can significantly advance understanding in this rapidly evolving field.
Autonomous computational catalysis through an agentic research system
Beijing Key Laboratory of Artificial Intelligence for Advanced Chemical Engineering Materials, State Key Laboratory of Chemical Engineering and Low- Carbon Technology, Department of Chemical Engineering, Tsinghua University（北京先进化工材料人工智能重点实验室、化学工程与低碳技术国家重点实验室、清华大学化学工程系）
Enhanced Hydrogen Evolution Catalysis of Pentlandite due to the Increases in Coordination Number and Sulfur Vacancy during Cubic-Hexagonal Phase Transition
Optimization of High Entropy Alloy Catalyst for Ammonia Decomposition and Ammonia Synthesis
高熵合金（HEA）纳米颗粒的成功合成是材料科学长期追求的目标，它开辟了材料科学的新前沿，在催化、电子、结构合金和含能材料方面具有应用。最近，由地球上丰富的元素制成的Co25 Mo 45 Fe 10 Ni 10 Cu 10 HEA显示出对氨分解具有高催化活性，其可与现有技术的钌催化剂相媲美，但昂贵得令人望而却步。使用基于第一性原理计算的计算方法，结合数据分析和机器学习，我们建立了一个模型来快速计算H，N和NHX（x= 1，3）物种在不同合金成分和原子排列的CoMoFeNiCu合金表面上的吸附能。我们表明，25/45钴/钼的比例确定实验上作为最活跃的氨分解组合物增加的可能性，表面吸附氮相当于钌，而在同一时间相互作用的中间体适度强烈。我们的研究强调了计算建模和机器学习在其近乎无限的材料设计空间中识别和优化HEA合金的重要性。
