Nuclear fusion fuel without toxic mercury

Nuclear fusion reactors, if and when they become practical, will need a steady supply of fuel. The source of that fuel is the uncommon lithium isotope lithium-6. And the established process to obtain Li-6 was banned in the US in 1963 because of the harmful mercury waste it generates.

Now researchers have invented a simple electrochemical route to selectively grab Li-6 from aqueous solutions of lithium salt (Chem 2025, DOI: /10.1016/j.chempr.2025.102486). The new mercury-free method works as well as the one based on the poisonous metal and could clear the path for industrial-scale production of fusion fuel, its developers say.

Lithium exists in natural sources mainly as the isotope Li-7. Just 7.5% occurs as Li-6, which is critical for nuclear fusion. Bombarding Li-6 with neutrons generates the hydrogen isotope tritium, the most common fuel for fusion.

Powering nuclear fusion reactors of the future will require hundreds of metric tons of Li-6. But separating Li-6 from Li-7 in natural sources is “really difficult because the two isotopes are so similar in properties,” says Sarbajit Banerjee, a chemist at Texas A&M University. The separation method used historically relies on the solubility of Li-6 in liquid mercury.

Banerjee and colleagues’ new process separates the isotopes based on the principle behind lithium-ion batteries, namely intercalation: the insertion and release of lithium ions from a layered cathode material. The cathode material they use is a polymorph of vanadium oxide called ζ-V₂O₅ (or zeta-V₂O₅), which has an atomic structure composed of long 1D tunnels that are a few atoms wide.

The researchers coupled the cathode with a carbon anode in an electrochemical cell. When they pumped an aqueous lithium solution through the cell while applying a voltage, lithium ions entered the tunnels of ζ-V₂O₅. But only Li-6 ions were trapped; they got stuck at four types of interstitial sites in the ζ-V₂O₅ crystal.

The researchers admit they don’t fully understand why the material selectively traps Li-6, and they want to explore the effect with modeling and experiments. But Andrew Ezazi, a postdoctoral researcher in Banerjee’s group, says their theory is that it has to do with the two isotopes’ different bonding ability and travel speeds through the tunnels. The lighter lithium-6 rushes in, makes a strong bond with the ζ-V₂O₅, and becomes trapped, while the Li-7 slowly migrates through the material.

Reversing the voltage releases the Li-6 back into the solution, boosting the concentration of the rarer isotope. One electrochemical cycle enriches the solution by 5.7%, Ezazi says. “The amount of enriched lithium-6 you need for fusion is 30–90%, depending on reactor design and fuel needs. We can achieve that in 25–40 cycles.”

This work “presents a stunning new idea that will be quite transformative,” says Partha P. Mukherjee, a mechanical engineer at Purdue University. He is also impressed by the synchrotron-based X-ray scattering and spectroscopy studies that the researchers undertook to understand the insertion and trapping of lithium ions in the ζ-V₂O₅. “Despite its apparent technological relevance, this work shows the importance of deep fundamental chemistry,” he says.

Amy Prieto, a chemist at Colorado State University, calls the work “incredibly creative and elegant. This could enable a lower-cost, scalable method to achieve efficient [Li-6] separation without the use of mercury.” Electrochemical processes are known to be scalable, so provided that ζ-V₂O₅ can be made on a large scale, this approach would be very practical for separating Li-6, she says.