Exclusive: Marathon Fusion Achieves Reactor Fuel Breakthrough

While most coverage of the buzzy fusion energy industry — including my own — tends to focus on the startups promising to build commercial reactors within the next decade, a whole host of supporting industries will also need to mature in order to make that long-held scientific dream a reality. Isotope production is one of the biggest. No matter a company’s technical approach to fusion, it likely demands hydrogen and lithium isotopes — the former to fuel reactors, and the latter to breed more of that fuel.

That’s where Marathon Fusion comes in. The San Francisco-based seed-stage startup is developing isotope separation technology for two key purposes: recycling tritium — an extremely rare hydrogen isotope — from reactor exhaust so it can be reused as fusion fuel, and enriching lithium-6, which is needed to breed new tritium. On Thursday, the company announced that it succeeded in using its plasma centrifuge technology to enrich lithium-6 and hydrogen isotopes in the lab. (It can’t yet test the tech on actual tritium, which is expensive, radioactive, and tightly regulated by the Nuclear Regulatory Commission, so Marathon is validating its separation physics using the non-radioactive proxies deuterium and protium.) Marathon now plans to raise a Series A based on the results.

“People have wondered for a very long time when fusion is going to come, and everyone’s waiting on the big scientific announcements,” Marathon’s CEO Kyle Schiller told me. But while the industry waits for those breakthroughs, he argued, it’s high time to start commercializing the infrastructure fusion will need to become an actual commercial industry. “Ultimately, what we’re doing is reactor agnostic. Everyone’s going to need it.”

In the near term at least, most fusion companies plan to use deuterium-tritium plasmas to power the fusion reaction. But the process is inherently inefficient — only a small fraction of the fuel actually fuses in the reaction, while the rest gets expelled, even though it still contains valuable, unburned tritium that can be captured and reused.

Today, neither tritium nor the lithium-6 needed to make more of it are produced at anything close to the scale even a single commercial fusion reactor would require to get up and running. And existing isotope separation technologies — largely designed for small-volume defense programs and experimental reactors — aren’t sufficient to bridge the gap.

“When you have a single fusion power plant, that’s going to need about 1,000 times more lithium than anyone is producing today in any country,” Schiller told me, referring to lithium-6. “It would be totally prohibitive to build a fusion power plant at those economics.”

And while it’s at least possible to produce enough of this isotope to supply a future fusion industry by enriching lithium mined from rock, tritium presents a more fundamental problem. Because it’s radioactive and decays relatively quickly, it doesn’t occur naturally in meaningful quantities. Today it’s produced commercially as a byproduct of some fission reactors, but that supply amounts to just a few kilograms per year. A single 1-gigawatt commercial fusion reactor, by contrast, would need an estimated 56 kilograms annually. Meeting that demand will require fusion companies to breed their own tritium inside the reactor, a process that involves fusion-generated neutrons hitting lithium-6 nuclei, splitting them into tritium and helium.

It will also necessitate recycling the substantial amount of tritium that passes through the reactor without burning up. That’s where Marathon’s plasma centrifuge comes in. Centrifuges themselves are nothing new — engineers have used them for decades to separate uranium isotopes for nuclear fuel, spinning the gas at such high speeds that isotopes with different masses separate. Plasma centrifuges work on the same principle and have been studied since the Manhattan Project, but no one has yet successfully commercialized the approach for lithium and hydrogen.

Part of the reason is that, until recently, there simply wasn’t much demand for these isotopes. But the raw materials also present a physics challenge: Lithium and hydrogen isotopes have very similar masses. Separating them thus requires spinning the plasma so rapidly that, historically, the resulting heat has undermined the separation process itself. To address this, Marathon’s proprietary centrifuge tech uses a “partially ionized” plasma, in which some atoms have been stripped of their electrons while others remain neutral. The company says this configuration allows the centrifuge to operate at lower temperatures.

The materials testing lab Covalent has certified Marathon’s lithium-6 enrichment. The company hasn’t had its hydrogen separation results independently verified, though an MIT nuclear engineering professor has reviewed the device’s design. As a participant in ARPA-E’s Vision OPEN program, which solicits and supports ambitious energy projects, Marathon has also presented its hydrogen separation methodology and results at the ARPA-E fusion programs meeting in June.

Now, Schiller told me, the challenge is scaling up the technology’s core systems. “We need bigger magnets, better cooling, bigger power systems, and so that’s a buildout that’s going to take time and more capital,” he said. “But as far as the science is concerned, we feel like it’s at the point where we’re ready to make those kinds of commitments.”

Marathon is now looking to raise capital to build its first commercial pilot facility, with the goal of reaching full-scale production by 2029. Schiller told me the company expects its first full-scale facility to produce tens of tons of lithium-6 per year — enough, he says, to fuel a new gigawatt-scale fusion plant roughly every two years. Marathon also plans to recover and repurpose about 560 kilograms of tritium annually — roughly the amount that cycles through a 1-gigawatt reactor’s fuel system each year, most of which exits in the reactor’s exhaust without ever fusing.

Once fusion reactors are operating at scale, Marathon has a few other tricks up its sleeve. The startup also plans to build an “isotope production” business, using the copious volume of high-energy neutrons generated by fusion to manufacture valuable isotopes. The company made headlines last year with its claim that fusion-generated neutrons could transmute mercury into an unstable isotope that eventually decays into gold — potentially doubling a fusion reactor’s economic output (and proving the old alchemists right). But that work is still theoretical, based on computer simulations rather than peer-reviewed or experimentally validated work.

Marathon certainly has plenty to keep it busy in the near term, though. “There is a really amazing opportunity right now to say, look, the fusion supply chain is ready to go. We can start scaling up,” Schiller told me. “The science will progress in parallel, and we really want to land this together — not wait another 10 years after scientific results come in.”