Name Surname
AI supported hypothesis formulations on metal-rich granites: Implications for critical metal prospectivity
Main Supervisor: Fabrice Gaillard (Centre National de la Recherche Scientifique)
Co-Supervisors: Anouk Borst (Katholieke Universiteit Leuven), Jérémie Melleton (BRGM)
Objectives
Rare metals granites are small magmatic bodies that contain critical metals such as Li, Nb, Ta, W, and Sn at concentration levels well exceeding conventional granites. In Europe, granites and the related pegmatite rocks, constitute the main resources for these elements. Yet, no model can explain their genesis, why they are so rare and why they form such small magmatic bodies. With this project, we plan to use available solid-liquid-vapour partitioning data for these critical elements to simulate melting – crystallization – degassing processes operating in granitic magmas and train AI models under a large range of P-T-chemical conditions. The obtained numerical simulations will be compared with a large database of granite bulk rock compositions (major and trace elements and isotopes), which have been established by the associated partner BRGM, to provide hypothesis on the range of thermodynamic conditions that can potentially produce rare metal-rich granites.
Epected Results
By harvesting a large number of melting-crystallization-degassing data and simulations, we plan to:
- Provide unpreceded comprehensive constraints on the genesis of these magmatic geochemical anomalies;
- Propose a new generation of predictive models based on AI that could be used to search for new deposits. The results will be applied to French rare metal granites including the Beauvoir rare metal granite (Imerys kaolinite and Li deposit).