面向催化应用的金属团簇(Au_(n)、Ag_(n),和Cu_(n))物性的第一性原理研究  

Study of the first-principles of physicalproperties of coin metal clusters for catalytic applications

作　　者：耿兴泽 孟祥颖 CENG Xingze;MENG Xiangying(College of Sciences,Northeastern University,Shenyang 110819,China;Institute of Materials Intelligent Technology,Liaoning Academy of Materials,Shenyang 110004,China)

机构地区：[1]东北大学理学院,沈阳110819 [2]辽宁材料实验室,材料智能技术研究所,沈阳110004

出　　处：《中国体视学与图像分析》2023年第2期178-193,共16页Chinese Journal of Stereology and Image Analysis

基　　金：辽宁省自然科学基金应用基础研究计划项目(2023JH2/101300146)。

摘　　要：金属团簇因其独特的光学、磁学、电子性质和拓扑结构,已经成为一类新型高效、有前途的催化材料。本文基于密度泛函理论(DFT),系统研究了三类金属团簇(Au_(n)、Ag_(n)和Cu_(n),n=3~80)拓扑和电子结构随团簇规模的演变规律。进一步采用机器学习(集成学习)方法,挖掘出基于拓扑结构的描述符用于合理预测三类金属簇的费米能级,揭示了费米能级与拓扑结构之间定量的相关性。综上,本研究从原子水平上提供了三类金属团簇的拓扑/能量/电子信息,研究成果有助于理解金属团簇催化性能的本源,为特定催化反应筛选稳定、高催化活性的团簇类型催化剂提供了基础数据和候选材料。Metal clusters have become new efficient and promising catalytic materials due to their unique topology and optical,magnetic,and electronic properties.This paper systematically investigates the evolution of structural stability and electronic trends of three types of metal clusters(Au_(n),Ag_(n),Cu_(n),n=3~80)based on density functional theory(DFT).Further,a machine learning(integrated learning)approach is used to mine topology-based descriptors for reasonable prediction of Fermi energy levels of precious metal clusters and reveal quantitative correlations between Fermi energy levels and topology.This study provides topological,energy,and electronic information of the three types of metal clusters at an atomic level.The research results help to understand the nature of the catalytic performance of metal clusters.It provides fundamental data and candidate materials for the screening of cluster-based catalysts with stable and high catalytic activity for specific catalytic reactions.

关 键 词：金属团簇 电子结构 催化性能 第一性原理计算 机器学习 

分 类 号：O482.5[理学—固体物理]