主讲人：李湃，中科院上海微系统所副研究员
时 间：2025年10月16日15:00
地 点：物电学院A栋316
联系人：夏圣轩
讲座摘要：We present an innovative approach to studying surface superstructures through the use of machine learning (ML) force fields. Currently, the exploration of surface superstructures is limited by the lack of efficient and highly accurate computational methods. This paper addresses this gap by training ML force fields on data derived from first-principles calculations. We demonstrate the effectiveness of this approach by successfully simulating the herringbone reconstruction of gold surfaces and the zigzag morphology of silicon surface steps. These findings are crucial for advancing our understanding of surface superstructure phenomena. The application of ML force fields enables more efficient and precise studies of surface superstructures, offering new insights and methodologies for investigating their properties and potential applications.
主讲人简介：李湃，中国科学院上海微系统与信息技术研究所副研究员，2013年本科毕业于湖南大学，2018年于中国科学技术大学获得博士学位，之后在中国科技大学和韩国基础科学院从事博士后研究，2024年入职上海微系统所。主要从事材料的生长与缺陷调控的理论研究，多尺度计算方法的开发和应用。在Science Advances等学术期刊发表论文40余篇，他引2400余次。入选中科院BR计划、上海市海外高层人才计划。
