TOPIC #9
Harnessing Edge Computing and Deep Learning for Transformative TEM in Battery Materials Analysis (EDGE-TEM)
Research area: Advanced Methods
Keywords: Transmission Electron Microscopy; 4D-STEM; Edge Computing; Deep Learning; Battery Materials; In situ/Operando Characterization
Supervising team: Arnaud DEMORTIERE (LRCS) & Eric LARQUET (SOLEIL Synchrotron)
Abstract
This PhD project aims to harness edge computing and deep learning to transform transmission electron microscopy (TEM) for the analysis of advanced battery materials. Building on the installation of a next-generation TEM (Spectra 200, Thermo Fisher Scientific) at LRCS lab, equipped with hybrid pixel detectors and advanced techniques such as 4D-STEM, iDPC, and electron ptychography, the candidate will design computational workflows enabling real-time treatment of large, multimodal datasets. The research will couple edge computing architectures with machine learning (PCA, NMF, clustering) and deep learning (CNN, VAE, GAN, diffusion models) to deliver rapid and reliable structural and chemical quantification during in situ and operando electrochemical experiments. Particular emphasis will be placed on metadata automation, noise reduction, dynamic process analysis, and beam-damage mitigation through adaptive feedback loops. The student will work in close collaboration with European research centers and industrial partners, gaining training in advanced microscopy, data science, and AI-driven instrumentation. By accelerating the link between data acquisition and interpretation, this project will provide unprecedented insights into degradation mechanisms and transformation pathways in Li-ion and solid-state systems, supporting the design of durable next-generation energy storage technologies.
Interest for the student
Expected mobility: The PhD will involve secondments at industry partners (at least 3 months), training workshops within DESTINY2, and participation in international conferences (e.g., MRS, Microscopy & Microanalysis, IBA). Short visits to partner labs (CIC-enegigune, Forschungszentrum Jülich) are also planned.
Career opportunities: This PhD offers strong prospects for careers in academia (materials science, microscopy, AI for materials) and industry (battery manufacturing, instrumentation companies, AI/data analytics for microscopy). The acquired expertise in AI-driven TEM workflows is highly transferable and in growing demand.
Contacts
IMPORTANT: you may contact the potential supervisors to have more information about the topic, however, sending them your application directly is not permitted.
