MSE Seminar – Dr. Wenbo XIE (16 April 2026)
16 Apr 2026 (Thu) | 10:30 AM - 11:30 AM
MSE Seminar – Dr. Wenbo XIE (16 April 2026)
| Title: | AI4M: Leveraging Data-Driven Methods for Predictive Materials Modeling |
| Speaker: | Dr. Wenbo XIE School of Physical Science and Technology ShanghaiTech University |
| Date: | 16 April 2026 (Thursday) |
| Time: | 10:30 am |
| Venue: | Hong Kong Institute for Clean Energy, 6/F, Mong Man Wai Building |
| Abstract: | Artificial intelligence and advanced computation are transforming materials research, creating new opportunities to accelerate the discovery and design of functional materials. Yet many materials cannot be understood as static structures alone: their behavior is governed by atomic-scale transport, reaction, and structural evolution, often coupled across multiple length and time scales, making them difficult to probe experimentally and challenging to describe theoretically. Capturing such complexity remains a central challenge for modern materials science and calls for computational frameworks that combine first-principles rigor with far greater reach. To address this challenge, Dr. Xie’s research combines fundamental materials physics, efficient computational modeling, and machine learning. A central thrust of his work is the development of general, reactive machine-learning potentials that extend atomistic simulation to larger length and time scales while preserving mechanistic insight. Beyond this, he explores foundation-style and generative approaches, including diffusion methods, together with new atomistic representations, to broaden exploration in chemical and configurational space and to build AI models for materials that are more transferable, expressive, and physically grounded. Using heterogeneous catalytic materials as a primary example, he will show how these approaches enable dynamic, data-driven descriptions of materials under operating conditions and support a broader vision of predictive materials discovery and design. |
| Enquiries: | mse@cityu.edu.hk |