By Matt Busel, who built the apps, crates and game he writes about · every number links to its source and every dataset is downloadable (methods) · corrections · featured in Kirino, gamebiz, This Week in Rust, New Mobile Life, By-Press
Mercury to gold: Marathon Fusion beat CERN 10x. The US buries mercury
On this page · 7 sections
Short answer: Yes, mercury can be turned into real, stable gold, and on Oct 21 2026 the San Francisco startup Marathon Fusion said it did it at 10 times the rate of CERN's 2025 record. Physics is no longer the limit. Our numbers say the limits now are fusion power plants that do not exist yet and a US policy that treats the feedstock as waste: the government is building a storage site for up to 7,000 tons of mercury, 33 years of world mine output.
TL;DR: Marathon's own 2025 paper models about 2 tons of gold per gigawatt of fusion heat per year, which at $4,017 an ounce on Oct 27 2026 is $258 million a year of gold per gigawatt, on top of the electricity. A kilo of gold is worth 2,583 kilos of mercury at today's prices, and gold is up 126% in five years. Matching every gold mine on Earth (3,300 tons in 2025) would take about 1,650 gigawatts of fusion heat, so the first people this hits are fusion investors, not jewelers.
Key numbers
| What | Number | Source, checked Oct 27 2026 |
|---|---|---|
| Marathon's claimed rate vs CERN's 2025 record | 10x | Marathon Fusion on X, Oct 21 2026 |
| Gold made at CERN's LHC, 2015 to 2018 | 86 billion nuclei, 29 picograms | CERN |
| Modeled gold output | about 2 t per GW of fusion heat per year | arXiv 2507.13461 |
| Value of that gold per GW-year (our math) | $258 million | Yahoo GC=F at $4,017/oz |
| Gold vs mercury, per kilo | $129,149 vs $50 (2,583x) | USGS MCS 2026 |
| World gold mine output, 2025 | 3,300 t | USGS MCS 2026 |
| World mercury mine output, 2025 | 210 t | USGS MCS 2026 |
| DOE long-term mercury storage contract | up to 7,000 t | USGS MCS 2026 |
| Marathon seed round (2024) | $5.9M, led by 1517 Fund and Anglo American | Pulse 2.0 |
Alchemy is now an engineering problem, and Marathon says it out loud
"We've reduced alchemy to a series of difficult but tractable engineering problems." That is Marathon Fusion's own line from its Oct 21 post, which passed 1.28 million views by Oct 27.
The trick is simple to say. A fast neutron from a deuterium-tritium fusion reaction knocks one neutron out of mercury-198, which becomes mercury-197, which decays into stable gold-197 within days. The 2025 paper calls this an (n, 2n) reaction: one neutron in, two out.
The detail that matters for builders: the same reaction multiplies neutrons, which fusion plants need anyway to breed tritium fuel. The paper's claim is that the gold comes without costing the plant any electricity.
CERN got there first by accident. Its ALICE team counted about 86 billion gold nuclei made in lead collisions during 2015 to 2018, a total of 29 picograms, and each one lived for a fraction of a second before smashing into the beam pipe. Marathon says its gold is stable and its rate is 10 times that record. Nobody outside the company has published a check of that number yet. Our bet: the reaction is real physics, and the fight will be over the rate.
Follow the money: a mining giant funded the alchemists
Here is the part the "lab-grown gold will crash the market" crowd skips. Marathon's $5.9 million seed round in 2024 was led by 1517 Fund and Anglo American, one of the biggest mining companies on the planet. A miner paying for the thing that could make mining look old is either a hedge or a bet. Either way it is not a blockade.
The US government is in too. ARPA-E, the Energy Department's moonshot office, gave Marathon a CREATE award that took its 2024 total to $6.9 million, and NEI Magazine reports a $3.63 million award from ARPA-E's VISION OPEN program. We checked USAspending for "Marathon Fusion" and got zero matching rows, so those grants likely sit under a lab or a prime contractor name. That is a transparency gap, not a scandal.
The incentive is getting bigger every month. Gold closed at $1,773.60 an ounce in November 2021 and $4,017 on Oct 27 2026, up 126%. At that price, Marathon's modeled 2 tons per gigawatt-year is $258 million of metal. Its December 2025 paper argues gold co-production can cut the plasma gain a fusion plant needs to pay off from somewhere between 10 and 100 down to 3 to 5.
Who blocks it: we looked, and it is not the gold lobby
We went in expecting the usual villain. The data points somewhere duller and more fixable.
Blocker one: there is no commercial fusion plant. Every ton in this story is modeled. Marathon's 2 tons is per gigawatt of fusion heat, and the world has roughly zero commercial gigawatts today. To match the 3,300 tons that mines dug up in 2025, you would need about 1,650 gigawatts of fusion heat running gold blankets.
Blocker two: America treats the feedstock as trash. Since Jan 1 2013 the Mercury Export Ban Act has blocked most elemental mercury exports, and on Nov 21 2024 the Energy Department hired a Texas company to build a long-term storage site for up to 7,000 tons of it. That is 33 years of the whole world's mercury mine output, headed for a tomb.
The ban made sense when mercury's main market was poisoning rivers in small-scale gold mining, which the UN puts at 838 tons of emissions a year. But Marathon's April 2026 paper flips the frame: a fusion blanket that burns mercury into gold removes it from the environment for good. Their paper values the world's extractable mercury stock at about $200 trillion. That is their estimate, not ours, and it assumes physics that only high-flux fusion reaches. Still, a pollutant that pays you to destroy it should not be buried.
What it does to the gold price (less than X thinks)
Multicoin Capital's Tushar Jain asked the obvious question: if lab-grown diamonds crushed diamond prices, what does lab-grown gold do to gold? The honest answer from the numbers is: nothing for a long time, then maybe a lot.
Gold's buyers are not mostly jewelers who care where atoms came from. USGS puts 2025 use at 40% jewelry, 24% bars and 21% central banks. A gigawatt plant adds about 2 tons a year to a 3,300 ton mine supply, which is 0.06%. The diamond comparison breaks because a lab diamond costs less to make than a mined one; fusion gold only gets cheap once fusion electricity does.
The bigger shift is the one nobody is pricing: fusion stops being a science project that needs a breakthrough to earn its first dollar. If isotope revenue pays the bills at low plasma gain, the first plants can make money years earlier. Scarcity of energy and scarcity of metals are the same fight, and this is the first company to put both on one invoice.
What this means for you
If you build or invest in fusion, read the isotope production paper before your next model review: transmutation revenue changes the breakeven math, and medical isotopes like molybdenum-99 pay at even lower power. If you work on policy or state energy offices, the cheapest move on the board is a carve-out that lets licensed fusion operators take stored mercury as feedstock instead of paying to bury it. Download our data below, check our math, and send it to whoever writes your district's energy bill.
More from our energy and materials beat: the government's fusion equity stakes, helium as a critical material, every piece on Frontiers, Economics and the news desk.
Want this tracked for you as it moves? Company Watch watches any company's federal contracts and grants, SEC filings and lobbying, and emails you when Marathon, Anglo American or a fusion rival files something new. The free answer here is true today; the paid loop keeps it true. We keep fixing it at no extra cost until it does what it promises.
Not financial advice. Our estimates (value per gigawatt-year, gigawatts to match mine output) are our own math on the sources linked here.
Sources
- Marathon Fusion, post on X, Oct 21 2026
- Marathon Fusion, Scalable Chrysopoeia via (n, 2n) Reactions, arXiv 2507.13461, Jul 17 2025
- Isotope Production in Fusion Systems, arXiv 2512.09242, Dec 10 2025
- Self-subsidizing Mercury Remediation with Fusion Reactors, arXiv 2604.02590, Apr 3 2026
- CERN, ALICE detects the conversion of lead into gold at the LHC, 2025
- USGS Mineral Commodity Summaries 2026, Gold, Feb 2026
- USGS Mineral Commodity Summaries 2026, Mercury, Feb 2026
- Yahoo Finance, gold futures GC=F, pulled Oct 27 2026
- Pulse 2.0, Marathon Fusion closes $5.9 million, Jul 2024
- NEI Magazine, Marathon in fusion fuel cycle breakthrough
- Business Today, Marathon Fusion claims 10x CERN record, Oct 22 2026
First published here on tensorust.com on Oct 27, 2026. Related: .
Up next
More on FrontiersHelium is a critical material now. The US exports 4.75x its importsDOE just called helium critical. USGS data: the US made 43% of world supply in 2025 and exported 38M m3 vs 8M imported. The real limit is permits.Read it next →Cite this
Free to cite and reuse under CC BY 4.0 with a link back. How it was collected: methods. Spotted an error? Corrections.
Did this answer your question?
Get the next one by email.
What my AI agents tried this week, with the real numbers: what worked, what flopped, what made money. One email when there is something worth reading.
No spam, never sold. Every email has a one-click unsubscribe.
What are people saying about this? Read and post on The Floor →
More: the author, related reading, share this, a free game
Try it yourself · Want the numbers to feel real?
See any amount as a pile of cash
Type an amount, see it as $100 bills next to a person: weight, volume and height.
In The Manual: on this topic
Explainers answer one question, short answer first. Guides walk you through it step by step.
- Explainer · Money and the economyWhat is an IPO lock-up expiration, and why do stocks drop around it?
- Explainer · Money and the economyWhat is enhanced geothermal, and how is it different from regular geothermal?
- Explainer · Money and the economyWhat is a stock warrant? The AMD and OpenAI deal, explained
- Explainer · Money and the economyWhat is an at-the-market offering, and why is Meta planning a $60 billion one?
Read next
- Economics · Oct 27, 2026 · 6 min readHelium is a critical material now. The US exports 4.75x its imports
- Economics · Oct 27, 2026 · 6 min readPerpetua tungsten: Army bets $40M on a metal US mines last dug in 2015
- Story chapters · Oct 27, 2026 · 6 min readSolar passed nuclear in 2026: 1,564 TWh vs 1,326 in six months
.jpg){kind=link}