Julien Lam

Can a system as thoroughly studied as the Lennard-Jones gas still reveal intriguing phenomena? In a collaboration with the Laboratoire de Physique de la Matière Condensée published in Phys Rev Letters, we employed Molecular Dynamics (MD) simulations and rare-event sampling to investigate liquid nucleation from the vapor phase. Contrary to the predictions of Classical Nucleation Theory, our results demonstrate that critical clusters nucleate at a density significantly different from that of the macroscopic equilibrium liquid. We rationalize these numerical findings using a two-variable nucleation theory that is both accessible and highly effective.

Proud that Quentin Gromoff succesfully defended his thesis named: “Machine Learning Approaches to the Simulation of Nanoparticle Crystallization“. Massive thank you to his PhD committee including Jérôme Creuze, Christinne Mottet, Mathias Rapacioli and Emilie Gaudry. Special thank as well to my partner in crime Magali Benoit for the co-supervision.

Very excited to announce that our projet NanoPHIA is now funded by the PEPR Diadem! We will receive 905k€ to work on the formation process in physical synthesis of nanoparticles by exploiting AI to couple characterisation, in-situ measurement and simulations. It will a collaboration with CEMES, INSP and ILM and 4 postdocs will be hired soon to do both simulations and experiments.

As every year, we are hiring summer internship students. If you are interested in working around machine-learning, atomistic simulations, nanoparticle or nucleation, please feel free to send me CV and cover letter. More info here and here.

Many thanks to the ANR for supporting our project called PHASTRACK. I will have the opportunity to renew with my former colleagues and friends in Lyon namely David Amans and Karine Masenelli-Varlot. We will be working on nano-alloys melting mechanisms combining synthesis experiments, TEM measurements and of course ML-assisted molecular dynamics.

Happy to officially welcome Johannes SANDBERG in the group as a postdoctoral researcher for 1.5 year. He will be working on linear models of machine-learning interaction potentials in collaboration with Genevieve Dusson and funded by AISSAI.