NextFin news, On August 18, 2025, Terence Tarnowsky, a physicist at Los Alamos National Laboratory (LANL) in the United States, unveiled a proposal to use nuclear waste as a source to produce tritium, a rare hydrogen isotope critical for fueling future nuclear fusion reactors. Tarnowsky presented his findings at the American Chemical Society's Fall 2025 meeting held from August 17 to 21 in the US.
Current nuclear power plants generate electricity through nuclear fission, which splits uranium or plutonium atoms, releasing energy but also producing long-lived radioactive waste. In contrast, nuclear fusion reactors, which are still in development, would generate energy by fusing atomic nuclei, specifically isotopes of hydrogen called deuterium and tritium. Fusion promises vast energy output with minimal radioactive waste.
While deuterium is abundant, tritium is extremely rare and expensive, with commercial tritium valued at approximately $33 million per kilogram. The US currently lacks domestic capability to produce tritium and relies on limited supplies, including those from Canadian fission reactors. The total global tritium inventory is estimated at about 25 kilograms, enough to power over 500,000 homes for six months.
Tarnowsky's proposal involves using particle accelerators to initiate atom-splitting reactions within the radioactive nuclear waste, which contains uranium and plutonium. These reactions would release neutrons and trigger nuclear transitions that produce tritium. The nuclear waste would be surrounded by molten lithium salt to provide cooling and radiation shielding, enhancing safety and reducing risks of weaponization.
Computer simulations conducted by Tarnowsky suggest that a reactor system operating at 1 gigawatt thermal power—equivalent to the annual energy needs of 800,000 US homes—could produce about 2 kilograms of tritium annually, matching the total yearly output of all Canadian reactors. The design is projected to yield over ten times the tritium production efficiency compared to fusion reactors at the same thermal power.
This approach could also help mitigate the growing challenge of managing thousands of tons of nuclear waste in the US, which currently requires costly and long-term storage due to its radioactivity and environmental risks.
Tarnowsky emphasized that while the concept builds on earlier theories from the 1990s and 2000s, recent technological advances make it more feasible. He plans to refine reactor simulations and calculate production costs to support decision-making for potential implementation.
He noted the importance of long-term planning for tritium supply, as tritium decays with a half-life of 12.3 years and cannot be stockpiled for extended periods. The proposal calls for collaboration between public and private sectors to develop this technology as part of the transition to clean energy.
References for this report include the American Chemical Society's presentation at ACS Fall 2025, and articles from Tech Xplore, Gizmodo, and Interesting Engineering published on August 18, 2025.
Explore more exclusive insights at nextfin.ai.
