非均相催化体系在复杂反应条件下表现出高度的动态效应，催化剂的动态变化也对催化体系的性能有着显著影响。然后，由于缺乏高效的原位表征手段，人们对动态催化的微观认识较为肤浅。另一方面，传统非均相催化的理论计算通常只是对催化剂进行简单的静态结构优化，忽略反应条件下催化剂本身的动态特性，对催化的理解也往往局限于静态固体表面的化学吸附等电子结构效应。本项目拟发展基于密度泛函理论分子动力学模拟的方法研究亚纳米尺度的金属团簇的动态结构，并提出统计分析方法定量描述其结构的动态特征，通过自由能计算方法模拟典型催化反应的动态反应路径，计算反应自由能和能垒。从“液体”的视角审视催化体系的动态过程，理解动态催化的微观本质，并且发展动态催化体系的普适性原理，为实验构筑动态催化体系提供理论支持。

In heterogeneous catalysis, catalyst systems are known to show highly dynamic behavior under complex reaction conditions, and dynamics of catalysts has pronounced impact on the performance of catalysts. Microscopic understanding of catalyst dynamics is however missing due to the lack of efficient in-situ measurement techniques. On the other hand, the traditional theoretical studies of heterogeneous catalysis largely rely on simple, static geometry optimization methods, omitting the dynamic behavior of catalysts under reaction conditions. The common theoretical understanding is often limited to electronic structure theories that describe chemisorption on static solid surfaces. This proposal attempts to develop density functional theory based molecular dynamics method for investigating dynamical structures of sub-nanometer metal clusters, and propose statistical analytic methods for quantitative description of their dynamic features. Free energy calculation methods will be used to calculate the dynamic reaction paths of some model catalytic reactions, and determine the free energy changes and barriers. We will scrutinize the dynamical catalytic processes from a different viewpoint based on liquid theories, which we hope will help understand the microscopic nature of dynamical catalysis. The ultimate goal is to unravel the fundamental principle behind dynamical catalysis, which can help guide experimentalists to develop efficient dynamical catalytic systems.