Discovering the Materials and Materials Transformations in Electron Images Using Deep Learning

Designing active and stable catalysts and catalyst layers remains a grand challenge for the sustainable energy sector. Pre- and post-mortem as well as in-situ characterization has become essential for linking device performance to changes in the architectural components of the systems. In this talk I will mainly discuss two case studies demonstrating how we develop and apply electron microscopy methodologies to visualize the complete three-dimensional structure of fuel cell catalyst layers, and to observe the real-time evolution of Cu-based electrocatalysts. I will focus on the critical role of advanced image processing methods using deep learning algorithms in extracting meaningful and quantitative information from the experimental electron characterization data.

Short Bio
Vasiliki Tileli received her PhD in materials science from the State University of New York in the USA working on development of statistical, quantitative models of electromagnetic signals in environmental conditions of electron microscopes. She then completed her Marie Curie Individual Fellowship at Imperial College London before moving to the Institute of Materials at EPFL where she is now an Associate Professor. With her group, they study on a fundamental level the nanoscale properties of functional materials using in-situ electron microscopy techniques.