(375d) Developing First-Principles Computational Tools for Clean Metal Extraction

The transition to a sustainable clean-energy economy demands a dramatic increase in the production of critical metals such as copper, nickel, cobalt, lithium, and rare earth elements. However, conventional extraction processes are energy-intensive and largely powered by fossil fuels. Developing cleaner, more sustainable alternatives, including pathways that enable metal recycling, is therefore a pressing challenge.

I will present recent progress in developing first-principles computational tools to support and accelerate the design of clean metal-extraction technologies. For redox-mediated leaching in aqueous systems, we introduced an empirically corrected density-functional theory (DFT) method for accurate predictions of mineral redox potentials and Pourbaix diagrams [1]. For electrolytic extraction in molten salts, i.e., Hall–Héroult-type processes, we combined DFT and machine learning to predict salt melting and oxide reduction temperatures [2,3]. We also demonstrate how DFT-driven models and machine-learned interatomic potentials can aid the discovery of novel ligands for selective metal separation.

[1] B. Donovan, A. C. West, and A. Urban, in revision (2025).
[2] V. Gharakhanyan et al., J. Phys. Chem. 160 (2024) 204112.
[3] J. A. Garrido Torres et al., Nat. Commun. 12 (2021) 1-9.