Turning Mercury Into Gold, Raising Questions About the Metal’s Future Value
US Startup Claims Breakthrough in Turning Mercury Into Gold, Raising Questions About the Metal’s Future Value
Marathon Fusion says it has produced stable gold using nuclear fusion technology, but questions remain over production costs, scalability and the potential impact on global gold prices.
A San Francisco-based American startup has claimed a breakthrough that could challenge one of the oldest assumptions in economics: that gold’s scarcity is an enduring source of its value.
Marathon Fusion announced on 21 October 2026 that it had successfully converted mercury into stable gold using neutrons generated by nuclear fusion. The company said its process achieved a production rate ten times higher than a benchmark reported by CERN, the European particle physics research organisation, in 2025.
The announcement has revived interest in the possibility of manufacturing gold rather than extracting it from mines. However, the company has not disclosed the quantity of gold produced in its experiment, and its claim has not yet been independently verified.
The distinction is important. Although nuclear reactions can transform one element into another, producing gold in commercially significant quantities at a competitive cost remains a substantial engineering and economic challenge.
How can mercury be turned into gold?
Marathon Fusion’s technology relies on nuclear transmutation, a process that changes the nucleus of an atom to create a different element or isotope.
According to the company’s proposed method, high-energy neutrons generated by deuterium-tritium fusion reactions interact with mercury-198, a naturally occurring isotope of mercury. The process produces radioactive mercury-197, which subsequently decays into gold-197, the stable isotope that accounts for virtually all naturally occurring gold.
Unlike chemical reactions, which rearrange atoms without changing their elemental identity, nuclear reactions can alter an atom’s nucleus. The resulting gold is therefore genuine gold at the atomic level, rather than a gold-coloured substitute or a chemically similar material.
Gold-197 is also the isotope found in natural gold, meaning that material successfully produced and purified through this process would be chemically indistinguishable from conventionally mined gold.
The challenge lies not in establishing whether such a transformation is physically possible, but in achieving it efficiently, safely and economically at industrial scale.
From a scientific experiment to a potential gold industry
Marathon Fusion outlined its proposed gold-production process in a research paper published in 2025. The company estimated that a future fusion power plant could produce approximately two tonnes of gold annually for every gigawatt of thermal power, under its original model.
The company has also discussed higher production estimates based on different reactor configurations and power assumptions. These figures are projections rather than demonstrated commercial output.
The startup’s October announcement represents a separate claim of experimental progress. However, the amount of gold recovered, the process’s efficiency and its production costs have not been publicly established.
The technology would also depend on the development of commercially viable fusion power plants. Despite decades of research and significant private investment, fusion has not yet become an established source of commercial electricity generation.
If the technical and commercial obstacles can eventually be overcome, producing gold as a by-product of fusion power could offer reactor operators an additional revenue stream alongside electricity generation.
CERN has already demonstrated that gold can be created artificially
The concept is not entirely new. In 2025, CERN researchers reported results from experiments at the Large Hadron Collider showing that lead nuclei could be transformed into gold nuclei during high-energy collisions.
Between 2015 and 2018, the collider’s experiments produced approximately 86 billion gold nuclei, equivalent to around 29 picograms of gold. That quantity was far too small to have any commercial significance, and the gold nuclei existed only briefly before being destroyed in further interactions.
Marathon Fusion claims that its process has achieved a production rate ten times higher than CERN’s reported benchmark. However, the two approaches use different nuclear processes, and the startup has not published sufficient production data to establish the precise scale or commercial significance of the comparison.
The potential importance of Marathon’s approach is that it aims to integrate gold production into a future fusion power plant rather than rely on a large particle accelerator built primarily for scientific research.
Could manufactured gold undermine the metal’s value?
The prospect raises an important question for investors: if gold can eventually be manufactured in large quantities, will it lose some of the scarcity that has helped sustain its value for centuries?
Gold occupies a distinctive position in the global economy. It is used in jewellery, electronics and other industrial applications, while central banks and investors also hold it as a reserve asset and a potential hedge against economic uncertainty.
Its value is influenced by several factors, including interest rates, inflation expectations, geopolitical tensions, central-bank purchases, currency movements and investor demand. Limited mine supply is another important consideration.
A reliable, low-cost method of producing additional gold could eventually alter the supply side of this market. But the impact would depend on how much manufactured gold reached the market, how quickly production expanded and whether the process could compete with conventional mining and recycling.
A laboratory demonstration alone would not be sufficient to trigger a sustained decline in gold prices.
The diamond industry offers a warning but not a direct comparison
The expansion of lab-grown diamonds provides a useful example of how technology can disrupt a precious-material market.
Advances in manufacturing allowed lab-grown diamonds to become widely available at prices substantially below those of comparable natural stones. Their growing supply contributed to sharp declines in wholesale prices for laboratory-produced diamonds and increased competitive pressure on the natural-diamond industry.
However, gold is different in several important respects.
Diamonds are valued partly for their geological origin, rarity, appearance and role in luxury jewellery. Gold, by contrast, is a globally traded commodity with deep financial markets and a substantial role in official reserves.
Manufactured gold would have the same elemental identity as mined gold. Once it met applicable purity and safety standards, its origin alone would not necessarily prevent it from entering the bullion market.
Nevertheless, the diamond experience does not establish that manufactured gold would follow the same price trajectory. The economics of nuclear transmutation, the scale of production and the structure of the gold market are fundamentally different.
Two tonnes a year would not immediately transform the market
Global gold production is measured in thousands of tonnes annually. According to estimates published by the World Gold Council, mine production reached approximately 3,672 tonnes in 2025.
Against that background, a single reactor producing two tonnes annually would represent only a small addition to global mine supply.
The longer-term implications could be more significant if the technology became commercially viable and was adopted by many reactors. A fleet of 100 plants, each producing two tonnes annually, would theoretically add 200 tonnes of gold a year — equivalent to roughly 5.4% of 2025 mine production.
That scenario is illustrative, not a forecast. It assumes that the projected output can be achieved consistently, that enough suitable reactors are built and that the resulting gold can be processed and sold economically.
Even at that scale, the impact on prices would depend on demand, recycling, investment flows and the decisions of central banks and other major gold holders.
Safety and production costs remain major obstacles
One of the most significant challenges is radioactive contamination. Although gold-197 is stable, the irradiation process could create other isotopes or leave radioactive contaminants in the resulting material.
The company’s earlier proposal indicated that some material might require extended storage before it could be considered suitable for commercial use. The duration would depend on the isotopes present, their concentrations and the effectiveness of purification.
Additional costs could arise from mercury handling, radiation shielding, chemical separation, waste management, quality testing and regulatory compliance.
The economics would therefore depend on more than the theoretical amount of gold generated. Investors would need to know the cost per recoverable gram, the amount of electricity required, the efficiency of the process and the time needed before the material could be sold.
Until these questions are answered through independently verifiable results, claims of commercially competitive production should be treated cautiously.
What does this mean for gold investors?
For now, the announcement does not provide a compelling reason to assume that gold prices are about to collapse.
The technology remains at an early stage, the experimental output has not been fully disclosed, and commercial-scale fusion power generation has yet to become established. There is no demonstrated pipeline of manufactured gold large enough to displace a meaningful share of existing supply.
In the near term, gold prices are likely to remain far more sensitive to monetary policy, geopolitical developments, central-bank demand and investor sentiment than to the possibility of future fusion-based production.
Over the longer term, however, the development deserves attention. If nuclear transmutation becomes an economical industrial process, it could weaken one of the assumptions underpinning gold’s traditional scarcity: that meaningful new supplies must ultimately come from mines or recycled material.
That possibility remains conditional on technological and economic progress rather than an immediate market reality.
A breakthrough in physics, but not yet a revolution in gold prices
Marathon Fusion’s announcement highlights the growing ambition to extract more economic value from future fusion reactors. If the company can independently demonstrate its results, scale up production and establish a competitive cost structure, manufactured gold could become a new source of supply.
For now, the most important question is not whether science can turn mercury into gold. It is whether the process can produce enough of it, cheaply and safely enough, to change the economics of the global gold market.
Until that threshold is reached, gold’s established role as a financial asset and store of value is unlikely to be overturned by this announcement alone.
Reporting note: The production figures and technical claims attributed to Marathon Fusion are company estimates or statements unless otherwise specified. The claimed experimental breakthrough has not yet been independently verified.