Marathon Fusion said in an Oct. 21 post that it turned mercury into stable gold in September and outpaced CERN’s record by a factor of 10, but the company disclosed no yield and did not define the comparison. The post also gave no purity, cost or detailed experimental results.
CERN has reported several measures of gold production at the Large Hadron Collider. Across Run 2, from 2015 to 2018, the collider’s four major experiments produced an estimated 86 billion gold nuclei with a combined mass of 29 picograms. At ALICE, CERN recorded a peak rate of about 89,000 nuclei per second. Marathon did not specify which CERN figure its comparison used.
Marathon’s 2025 paper describes a modeled process in which 14 MeV neutrons from deuterium-tritium fusion drive reactions that convert mercury-198 into gold-197. The simulation projects about two tonnes of gold annually for each gigawatt of thermal power; it is not a production result from the claimed demonstration.
The company’s website lists a reactor under construction with one kilowatt of fusion power. It says the device fires deuterium ions at a lithium-tritide target to produce fusion neutrons. Marathon presents gold as a possible co-product for larger plants, estimating annual revenue of $482 million from electricity and gold at gigawatt scale, compared with $223 million from electricity alone. Those are company projections.
The scale of that modeled output can be compared with mined supply, though the measures describe different production methods. The World Gold Council’s initial estimate for mined gold production in 2025 was 3,671.6 tonnes, and the council said the figure could be revised.
A separate market precedent comes from lab-grown diamonds: diamond analyst Paul Zimnisky told JCK in March 2025 that their wholesale prices had fallen 90% to 95% from 2015 levels. That price history does not show how fusion-produced gold would affect bullion prices.
