Machine-learning approaches for nanoparticle simulations

The laboratory "Unité Matériaux Et Transformations" was created in January 2010 following the merger of four former laboratories of the Lille campus. The UMET laboratory now hosts a large portion of the research in Materials Science of the the Université de Lille. The laboratory includes approximately 80 professors and assistant professors, and CNRS researchers, 40 technical and administration staff, 60 PhD students, and 15 temporary contracts researchers or emeritus faculty.

Research overview While nanocrystals in material science are ubiquitous, the mechanisms of their formation which spans from nucleation to crystal growth remain one of the most intriguing process in nature. From the fundamental point of view, the main challenge is related to the stochastic nature of the process and the very small number of atoms involved. Altogether, it leads to numerous technical difficulties that have hindered the ability to systemically study nucleation in the most complex systems. In this context, numerical simulations involving both statistical mechanics and molecular quantum mechanics have been pivotal for providing an atomistic view of the underlying processes. Recently, machine-learning approaches have enabled for more precise and more extensive use of these numerical simulations which are now increasingly converging with experimental measurements.

Project The student will have the opportunity to pursue numerous research avenues depending on their preferences. Indeed, further numerical developments based on machine-learning approaches can be envisaged for statistically sampling free energy barriers associated to nucleation or for better modeling the interactions between atoms fitted on electronic structure calculations. Meanwhile, it will also be possible to focus on a specific material and study different approaches leading to the nanocrystal formation including gas phase condensation, solvent mediated synthesis or deposition mechanisms.

Organization The internship will be carried out at the "Unité Matériaux et Transformations" which is located at the Université de Lille. The student will benefit from the supervision of Julien Lam who is a CNRS researcher expert in numerical simulations for atomistic simulations. Potential students are not required prior knowledge of computational materials science as they will be trained in a large number of research domains including molecular quantum mechanics, statistical physics, machine-learning, material science and computer programming.

- Master in physics, chemistry or materials science
- Good knowledge of statistical mechanics and computational physics/chemistry
- Previous experience in C++/Python/Bash would also be a plus