Current batteries suffer from a trade-off between high-power and high energy density. In the context of the bilateral French/Germany project HIPOBAT, we aim to develop high-power batteries (Li, Na-based), which would enable fast charging and a long life, and at the same time guarantee sufficient energy density. To do so, we need to understand the underlying physicochemistry of redox reactions and mass transfer phenomena at the microscopic scale for these systems. Molecular modelling allows to reach this level of detail. Machine learning force fields[1] are now recognized as a good compromise between accuracy and simulation cost, allowing to model redox reactivity and having large simulation cells at the same time.
[1] X. Lian, M. Salanne., J. Chem. Phys., 159, 144705 (2023)
In this project, the postdoctoral researcher will develop a machine learning augmented potential for highpower batteries electrolytes selected in agreement with our experimental collaborators. Machine learning potentials can achieve ab initio quality at short distances, but incorporating long-range physics is a more difficult task. Classical force fields can model long-range interactions with reasonable accuracy. Thus, we plan to use on-the-fly uncertainty estimation techniques and learn the correction between a baseline classical force field and ab initio level energies and forces.
The postdoctoral researcher will contribute to different aspects of the project such as:
• the development of new theoretical approaches to model electrolytes
• carrying out molecular simulations of the molecular dynamics type
• the development of codes to analyse these simulations (using existing tools in the laboratory)
• collaboration with other project participants and with experimental groups
• writing publications and participating in conferences The postdoctoral researcher recruited will work as part of a team of researchers in the PHENIX laboratory at Sorbonne University, as part of the HIPOBAT project funded by PEPR (France) and BMBF (Germany) research programs.
Candidate profile: We are looking for a motivated candidate holding a PhD in physical chemistry or physics at the start of the project, and having research skills related to the theme and activities of the project. Good experience in molecular simulation and/or machine- learning is necessary, as well as a good knowledge of the associated theoretical tools (statistical physics of liquids, etc.; programming experience among: Python, Fortran, C, C++, etc.). .). Good level of spoken and written English is required.
Period: September 2024 – August 2026. The position will take place in the PHENIX laboratory, located on the Pierre et Marie Curie Campus of Sorbonne Université. The post-doctoral fellow will be part of the Electrochemistry and Ionic Liquids team and will carry out their research as part of the HIPOBAT project. He/she will benefit from access to the laboratory and university computing resources.
How to apply: https://emploi.cnrs.fr/Offres/CDD/UMR8234-MATSAL-007/Default.aspx
Contact: mathieu.salanne@sorbonne-universite.fr, alessandra.serva@sorbonne-universite.fr rocio.semino@sorbonne-universite.fr
