Recycling of rare earths with ionic liquid solvents: Bridging the gap between molecular modelling and process design

While rare earth metals are used in relatively small quantities, they play a major role in cutting edge technologies, such as electronics, information technology and in automobile industries. These metals are used in the high-powered magnets used in computers, they are components of wind-turbines and electric cars, they are used in fluorescent lights and in several catalytic processes. Roughly 86% of all rare earths come from China. This has been recognised as a significant risk to be so dependent on one country. Unfortunately the UK possesses very few rare-earth containing minerals, but what it can do to become more self-sufficient is to recycle the rare-earths that are in waste-piles. Currently the UK has little activity in this area, but recent parliamentary reports draw attention to the need for protecting the supply of rare earths and one foresees a growing effort in this area.

The UK does possess, however, a strong scientific base in the reprocessing of nuclear fuel using liquid-liquid extraction. We have worked in this area, alongside the National Nuclear laboratory. The knowledge gathered from these activities can usefully be re-chanelled into designing efficient extraction methodologies for the chemically-related rare earths. This is our intention.

We will focus on the extraction of the rare-earth, Samarium, from waste high-powered magnets using ionic liquids as extractants. Our aim to to scale-up the chemical processes currently investigated by the Binnemans group in Leuven, Belgium. While we believe our general methodology can usefully be applied to many, disparate processes, our focus will be on three systems. Our proposal is firstly to study these systems at a molecular level, using molecular dynamics simulations, to understand the molecular structures that form during the extraction process. Secondly we shall use these insights to construct soundly based, reliable thermodynamic models so that we can predict system properties over a range of...