Marathon Fusion says it turned mercury into gold, and the market shrugs

Marathon Fusion says it turned mercury into gold, and the price doesn't budge

Marathon Fusion, a San Francisco-based fusion start-up, announced this Wednesday that last month it converted mercury into stable gold, beating the CERN record by a factor of ten. The news revives the alchemists' dream and the antiestéticar that lab-made gold could do to bullion what synthetic diamonds did to gemstones: send prices tumbling. The claim, however, has not been independently verified, and the available material does not reveal how much gold was produced or at what cost.

The bar it says it surpassed is a modest one. CERN's Large Hadron Collider produced about 29 picograms of gold between 2015 and 2018, trillionths of a gram that existed for a fraction of a second before being destroyed in the collisions. Multiplying that by ten is still a lab curiosity. Gold was trading this Thursday at $4,062 an ounce, more than 27% below its January record high of $5,598, and the announcement has not changed its course.

What exactly Marathon Fusion does with mercury

The company, founded in 2023 by Kyle Schiller and Adam Rutkowski, a former SpaceX engineer, develops fuel systems for future fusion power plants. Its method fires neutrons from fusion reactions at mercury, which turns into an unstable isotope and decays into gold. On paper it is well-known physics, the same kind used in reactors to produce medical isotopes.

The problem is scale and timing. No fusion plant sells electricity yet, and the first commercial facilities are years away. A study by the company itself estimated that a fusion plant could produce about 5 tonnes of gold per year per gigawatt. A hundred such plants would add 500 tonnes a year, roughly 14% of last year's mining output, which was 3,672 tonnes.

The calculation that defuses antiestéticars of a gold crash

The energy math doesn't add up. A gigawatt running all year is 8.7 TWh, and that electricity costs no less than €30 million per TWh in the most optimistic scenario; Britain's Hinkley Point nuclear plant runs around €100 million. A tonne of gold fetches more than €100 million with the market this hot. The gap between cost and value is so narrow that any deviation breaks it.

There are more drawbacks. Transmuted gold could contain radioactive isotopes and might have to be stored for up to 18 years before it could be sold. And the raw material is not free: the enriched mercury the process requires trades at around $15,000 per milligram, versus $135 per gram for gold. You load the reactor with something thousands of times more expensive than what comes out of it.

Why synthetic diamonds worked and gold has it harder

The precedent hovering over the news is the lab-grown diamond, chemically identical to the mined kind. Wholesale prices for synthetic stones fell between 90% and 95% from 2015 levels, according to analyst Paul Zimnisky, and dragged natural gems down with them. De Beers launched Lightbox in 2018 at $800 per carat and shut it down in 2025 to focus on natural diamonds, taking an underlying loss of $511 million that year.

Gold plays in a different league. Mines produced 3,672 tonnes in 2025, and some 222,600 tonnes have been extracted throughout history. Central banks hold it as a safe-haven asset and have been buying heavily for years, China and Poland among them. Unlike synthetic diamonds, transmuted gold, once its radioactivity is gone, would be indistinguishable from mined gold. That is more unsettling than the quantity: if the technique ever matures, the origin label disappears.

For now there are no public plans to produce and sell such a product. The news moves headlines and little else. The full calculation, with energy costs, raw materials and years of storage, leaves a margin so thin that even the most bearish on the metal find no reason to sell.