AI-guided metal recovery from spent lithium batteries

Ligand/material discovery has been carried out through laborious trial-and-error approaches.

Recent advances in high throughput computational chemistry and AI/Machine Learning provided us with a more efficient, in silico approach to develop new ligands for functional materials and catalysis.

We aim to extend the Big Data/high throughput DFT methodology in Nguyen group to carry out ligand discovery in silico for Li, Mn, Co and Ni recovery from spent lithium batteries.

This will take advantage of experience in Nguyen group on using AI and Machine Learning for ligand discovery in catalysis.

The student will analyse literature ligands for Li, Mn, Co and Ni using cheminformatics and data science techniques to identify the required features of successful ligands (i.e. high binding strength).

These insights will inform an exhaustive search of the Cambridge Structural Database (CDS, 1.5M structures) for potential ligands and their performance will be evaluated with high throughput DFT calculations.

The workflow will be automated with Python code developed in Nguyen group, in collaboration with DiLabio group at University of British Columbia, which will significantly speed up our workflow to minutes per ligands. The most successful ligands identified in silico will be prepared and validated experimentally.

Further refinement of the lead ligands through rational design, computational evaluation, and experimental validation will be carried out.

The successful and novel ligands will be developed into new commercial products for recovering these critical metals from spent lithium batteries.