新闻与活动 活动信息

Chemistry Colloquium | Zhipan Liu: AI-Driven Chemistry Exploration: Simulation, Database and Experiment

时间

2023年9月7日(周四)
下午16:00-17:30

地点

西湖大学云谷校区E10-201

主持

Dr. Tao Wang, PI of CAP for Solar Fuels and School of Science, Westlake University

受众

全体师生

分类

学术与研究

Chemistry Colloquium | Zhipan Liu: AI-Driven Chemistry Exploration: Simulation, Database and Experiment

Time:16:00-17:30, Thursday, September 7th, 2023

Host: Dr. Tao Wang, PI of CAP for Solar Fuels and School of Science, Westlake University

Venue: Room E10-201, Yungu Campus

Prof. Zhipan Liu

Fudan University, Shanghai

Email: zpliu@fudan.edu.cn


主讲人/Speaker

Prof. Zhipan Liu got Ph.D in 2003 from Queens Univ Belfast under the supervision of Prof. Peijun Hu, and then did PostDoc with Professor David King in University of Cambridge. He returned to China in 2005 and has been a full professor in the Department of Chemistry, Fudan University since then. He has published more than 200 research papers with H-index 65. He was appointed as Executive Editor for J. Phys. Chem. A/B/C since 2017. Zhipan Liu's research focuses on the reactivity prediction of chemical systems for energy storage and conversion. His group developed the first machine-learning global optimization program, LASP (large-scale atomic simulation program with neural network potential) starting from 2016, which incorporates a series of theoretical methods developed in the group, such as the stochastic surface walking global optimization (SSW) method and global neural network method. LASP is now utilized by thousands of researchers worldwide for finding structures of material and identifying low-energy pathways.


讲座摘要/Abstract:

This talk will introduce our recent practice of chemistry exploration using LASP software, an AI-based total energy calculation package. LASP software is developed by our group from 2016, when we proposed a “Global-to-Global” approach for material discovery by combining the SSW (stochastic surface walking) global optimization method with neural network (NN) techniques. To date, LASP software (www.lasphub.com) has been utilized by many groups in the world for chemistry, material and physics applications. We will introduce the basic framework of the software and provide a few examples to demonstrate the key ingredients of the machine-learning atomic simulation using the global neural network potential. The main part of the talk will focus on how we utilize LASP simulations to resolve the reaction mechanism of Fe-based Fisher-Tropsch synthesis, to build database for specific chemical systems and to predict new chemical routes for experiment.


讲座联系人/Contact:

理学院,刘畅,邮箱:liuchang@westlake.edu.cn

School of Science, Chang Liu, Email: liuchang@westlake.edu.cn