A fusion startup shows how lithium is enriched for nuclear fusion fuel without the mercury process confined to Russia and China

How lithium is enriched for nuclear fusion fuel now runs through 1 mercury process

Every power plant that fuses deuterium and tritium, the 2 hydrogen isotopes used as fusion fuel, has to breed its own tritium supply inside a blanket of lithium. That means how lithium is enriched for nuclear fusion fuel is not a side detail, it decides whether a reactor can run at all. Right now, that enrichment happens through a mercury based chemical process called COLEX, run commercially only in Russia and China, according to Marathon Fusion, a fusion technology startup. The United States has no commercial lithium enrichment capability of its own, a gap rated the highest risk category in a December 2025 supply chain report from the Special Competitive Studies Project, an independent policy research group. The 1 domestic project underway, a lithium processing facility under construction at the Y 12 site of Oak Ridge National Laboratory, costs 0.96 to 1.6 billion dollars, is projected to finish in the early 2030s, and processes lithium from old stockpiles rather than enriching newly mined lithium, so it will not count as commercial capacity even once it opens.

Today the world makes less than a tonne of enriched lithium each year, with a mercury-based process now used only in Russia and China.

Per F. Peterson, distinguished professor of nuclear engineering at the University of California Berkeley and an adviser to Marathon Fusion. Source 1.

A plasma centrifuge splits the isotopes without a nuclear reaction

On August 27, 2026, Marathon Fusion said it had separated both lithium and hydrogen isotopes using a plasma centrifuge, a machine that spins a partly ionized gas, some atoms stripped of electrons and some left neutral, so that the lighter lithium 6 atoms and the heavier lithium 7 atoms sort apart by weight. That is a physical separation, the same principle a gas centrifuge uses to enrich uranium, not a nuclear reaction. Earlier attempts at this kind of plasma centrifuge struggled because lithium 6 and lithium 7 are close in mass, which demanded very high rotation speeds and generated too much heat to separate cleanly, according to Heatmap News, a climate technology news outlet. In its own press release, the company says this process, which it calls differential pumping, could also cut the flow rate of tritium and the size of the equipment that handles it by a factor of ten or more.

Marathon's results point to an intriguing way to make this system smaller and more practical.

Charles Swanson, vice president of Fusion Systems at Thea Energy, a separate fusion company. Source 3.

The reactor sized gap that remains

A single reactor at gigawatt scale needs roughly 100 tonnes of lithium 6 inside its breeding blanket, according to a 2026 study from Eindhoven University of Technology and the University of Bristol, more than 100 times the amount the whole world enriches in a year today. That comparison is this publication own arithmetic on 2 sourced figures, not a number either source states directly.

Marathon Fusion says it is now moving toward a first commercial pilot facility, targeting full scale production of tens of tons of lithium 6 a year by 2029, according to Heatmap News. That would still fall short of a single reactor blanket, but it would be the first domestic source of enriched lithium that does not depend on mercury or on Russia and China.

How much enrichment a blanket needs also depends on its design. A typical breeding blanket needs lithium 6 enriched to 60% to 90%, but a design that adds a neutron multiplier can work with as little as 30%, according to the same study.