Name Surname
Machine Learning Assisted Atomistic Simulations to Unravel Element Partitioning of Critical Raw Elements in Magmatic Systems
Main Supervisor: Prof. Maurizio Petrelli (University of Perugia)
Co-Supervisors: Razvan Caracas (Institut de physique du globe de Paris), Anouk Borst (Katholieke Universiteit Leuven)
Objectives
The project aims at performing focused ab initio simulations, to generate ML interatomic potentials (e.g., using packages like DPgen). These potentials will be used to study element partitioning and the kinetics of the chemical reactions taking place during magma crystallization. The results will be validated by the development of selected petrologic experiments aimed at reproducing the same conditions of ab initio MD simulations. The overall objective is to improve our knowledge of the behavior of element partitioning and the kinetics of chemical reactions during magma evolution in systems that are relevant to the economy of critical raw materials like the Rapasaari lithium deposit in Finland.
Epected Results
- The development of new ad initio MD simulations for magmatic systems that are relevant to the economy of critical raw materials;
- Improve our knowledge of the behaviour of element partitioning and the kinetics of chemical reactions during magma evolution (e.g., Li in pegmatitic systems).