Where Fusion’s Real Investment Case Sits: Minerals, Not Reactors

Private capital has committed more than $14.24 billion to fusion energy companies, according to figures cited in the Fusion Industry Association’s 2026 industry report. Yet the mineral supply chains those reactors will physically consume remain almost entirely unpriced by equity markets.

That gap is the story. Money is flowing toward the reactor builders, while the commodity markets sitting upstream of them stay quiet.

Fusion reactors have stopped being pure technology bets. They are industrial construction projects that will demand specific, scarce, and geographically concentrated minerals, and no current supply chain is scaled to deliver them at the volumes required.

The thesis here does not rest on knowing when commercial fusion arrives. It rests on the seven to fifteen year lead time between a mine’s discovery and its first production. That arithmetic means the window to establish resource positions is now, not when reactors approach deployment.

What this piece delivers is a commodity-by-commodity map of the fusion mineral supply chain across four critical inputs: where the production gaps sit, and which parts of the chain carry the most investable signal for a globally oriented resource investor. This is an investment map, not a fusion science lecture.

The lithium-6 problem: why fusion’s fuel cycle begins with an enrichment bottleneck

The most acute supply gap in the entire fusion mineral chain is not a mining problem at all. It is an enrichment problem, and that distinction changes where the money should go.

Naturally occurring lithium contains only 7.5% of the isotope lithium-6, the specific variant fusion reactors need. Lithium-6 sits in the reactor blanket and absorbs neutrons to breed tritium, the fuel that sustains the plasma reaction. Because the isotopic ratio in natural lithium is fixed, obtaining it is a separation process, not a matter of digging up more ore.

Here is the constraint that makes this structurally different from every other critical mineral theme. Commercial lithium-6 supply at fusion-relevant grades is effectively zero today.

The legacy technology that once produced it, the COLEX (column exchange) process operated at the US Y-12 plant during the Cold War, relied on mercury-based chemistry now banned under environmental regulations. Reviving it at fusion scale would demand mercury throughput orders of magnitude beyond current global production. No alternative method has yet been demonstrated at industrial throughput.

Now let the scale of the demand accumulate against that zero. A single commercial fusion pilot plant is estimated to require between 50 and 150 tonnes of enriched lithium at 60% to 90% Li-6 concentration (figures currently unverified). With commercial deployment projected across the 2035-2050 window, the industry faces a supply preparation lead time of 10 to 15 years.

The implication writes itself. Enrichment decisions made in 2026-2028 determine whether fusion-grade lithium exists in the late 2030s.

Enrichment decisions made in 2026-2028 determine whether fusion-grade lithium exists in the late 2030s, and nuclear-grade lithium markets are already being reshaped by bilateral supply agreements that lock in volumes years before commercial demand arrives.

Projected demand signal (unverified) An MIT Sloan thesis (2025) projects that fusion-related lithium demand alone could approach 30% of global lithium supply by 2050, and drive total lithium demand to eight times current annual production by 2100.

Emerging enrichment technologies and their commercial readiness

Several separation methods sit at pilot stage, each with a measured but limited efficiency:

• Deep eutectic solvents: custom solvent systems showing Li-6/Li-7 separation factors of approximately 1.068 (unverified), marginally better than COLEX at around 1.054, with far less toxic waste.
• Mercury-free electrochemical methods: a recent study reported roughly 5.7% enrichment per cycle (unverified), projecting to approximately 90% Li-6 across about 45 cycles.
• AVLIS (atomic vapor laser isotope separation): a laser-based approach still at experimental scale.
• Crown-ether mixer-settler systems: chemical extraction using ring-shaped molecules that selectively bind lithium isotopes.

None has been demonstrated at the throughput needed to deliver several tonnes of fusion-grade lithium per full-power year. That is the whole opportunity. The investable position here is not lithium mining but midstream processing. Whoever funds the first commercially scaled, environmentally compliant Li-6 enrichment facility captures the one chokepoint substitution cannot bypass, because the moat is technological and regulatory rather than geological.

Where beryllium and tungsten sit in the supply chain concentration risk map

Two more construction minerals show how thin the fusion supply chain really is. Neither is scarce in the ground. Both are concentrated to a degree that should concern any developer trying to source them.

Beryllium plays a dual role inside a reactor, acting as a neutron multiplier and a plasma-facing material. Its low atomic weight, high melting point, and neutron behaviour make it difficult to substitute. The market barrier is toxicity: beryllium processing demands specialised handling facilities, which keeps new producers out.

The numbers show just how narrow Western supply is. Global beryllium mine production ran at roughly 430 metric tons in 2025, with the United States producing about 230 tons, or 53% of the world total (figures unverified). One company anchors that Western capacity.

Materion Corporation operates the Spor Mountain mine in Utah, which produced 418,000 pounds of beryllium in 2024 while running at around 65% of mill capacity (unverified). The site holds 8.4 million tons of proven and probable reserves grading approximately 0.246% beryllium, with an expected life of at least 75 years. For an investor, beryllium is a single-stock concentration risk sitting on a very long-duration reserve base.

Tungsten tells a geopolitical version of the same story. It is the leading candidate for reactor first-wall and divertor components thanks to its extreme heat tolerance and low erosion under particle bombardment. Fusion requires purities and geometries that standard industrial tungsten does not meet, adding a specialised processing layer on top of an already concentrated supply.

Global tungsten mine production reached about 85,000 metric tons in 2025, with China accounting for roughly 80%, around 67,000 tons (unverified). Western fusion programmes face domestic content requirements that make Chinese sourcing commercially or legally problematic, and no new tungsten-specific Western policy has emerged beyond broad critical-mineral frameworks.

Near-term institutional demand ITER’s tungsten divertor components alone require several hundred metric tons, a concrete demand signal already in the procurement pipeline.

The precedent for solving this exists. Japan cut its rare-earth dependency on China from roughly 90% to under 60% (unverified) through diversified sourcing and strategic reserves. The read for you is straightforward: developers who lock in non-Chinese tungsten and expand Western beryllium refining early carry a structural cost advantage, because the ore is available but fusion-grade processing capacity is thin and not currently expanding.

Helium and tritium: the operational inputs that extraction cannot fix

The construction minerals get a reactor built. Two operational inputs determine whether it can run, and they behave nothing like bulk commodities you can stockpile.

Helium is the cryogen that cools superconducting magnets to near absolute zero, confining the plasma. There is no cost-equivalent substitute at scale. It is a byproduct of natural gas extraction, lost to the atmosphere if not captured at the wellhead, which ties its supply to gas field economics that have nothing to do with fusion.

Helium supply is structurally entangled with natural gas market dynamics because wellhead capture economics determine whether helium is separated or vented, meaning any investment thesis around helium tightness must account for LNG project timelines and gas field development economics that operate on entirely separate demand drivers.

Global helium production runs at roughly 175-200 million cubic metres annually, concentrated across a handful of jurisdictions:

• United States: historically the dominant supplier, now facing structural change.
• Qatar: a major exporter, previously the source of supply shocks that pushed spot prices to record levels.
• Algeria: an established producer with steady output.
• Russia: a growing but geopolitically constrained source.

The structural shift for helium came on 27 June 2024 (unverified), when the US Bureau of Land Management closed the sale of the Federal Helium Reserve to private gas company Messer. That reserve had supplied over 20% of US domestic demand and roughly 9% of global demand. Privatisation adds a pricing risk layer for research and industrial buyers, and supply diversification depends on extended-timeline projects in Canada, Tanzania, and South Africa.

Tritium is the more time-sensitive input, and the arithmetic is what makes it urgent. It is the primary fusion fuel alongside deuterium, produced only as a byproduct of heavy-water reactors, chiefly Canadian CANDU units. Civilian access is restricted.

Tritium decays with a half-life of 12.32 years, meaning roughly 5.5% of any stockpile disappears every year. Global civilian inventory sits at approximately 25-30 kilograms, and projections suggest it will decline from the mid-2020s as legacy reactors retire, potentially falling below 5 kg by the 2050s (unverified). Commissioning a single plant may require up to 10 kg (unverified).

The commercial reality of tritium At a reported $30,000 to $40,000 per gram (unverified), tritium is among the most commercially inaccessible inputs in the entire fusion chain.

What this tells you is that tritium scarcity is a deployment sequencing problem, not merely a volume problem. Developers cannot warehouse it the way they warehouse ore, because it decays and civilian access is capped. If several reactors attempt first ignition in overlapping windows, competition for existing inventory becomes a binding constraint on how fast the industry can deploy.

Tritium breeding and the self-sufficiency question

Commercial reactors are designed to breed their own tritium from lithium-6 blankets, which is why some developers downplay the bottleneck. Commonwealth Fusion Systems states its ARC tokamak carries a maximum tritium inventory of about 90 grams (unverified), a modest startup load.

The gap sits between design intent and engineering reality. Some developers cite tritium breeding ratios (the amount of tritium produced per unit consumed) slightly above unity, around 1.01 (unverified), as sufficient. Critics argue practical self-sufficiency needs ratios of 1.15-1.17 once losses are counted, while current blanket designs achieve effective ratios of only around 1.03-1.05 (unverified).

That shortfall matters for your timeline read. Breeding efficiency below target extends the period during which reactors need external tritium, which makes the lithium-6 enrichment infrastructure more urgent, not less.

The midstream gap: where the investable thesis actually sits

Step back from the four commodity stories and one thread connects all of them. The constraint is never the mine. It is the processing layer that converts raw material into fusion-grade product.

Two precedents predict how the fusion chain behaves if that midstream layer stays underinvested. Cobalt shows the downside: over 50% of global mining sits in the Democratic Republic of Congo and nearly 50% of refining in China (unverified), exposing the whole chain to cascading breakdowns from a single localised shock. Rare earths show the fix, with Japan cutting Chinese dependency from around 90% to under 60% (unverified) through deliberate diversification.

The distinction for you is between mining exposure and processing exposure. Greenfield mining carries the longest lead times and the highest geological risk. Processing capacity carries technological moats, direct policy tailwinds, and the ability to serve several demand sectors at once.

Midstream processing investment in critical mineral chains has historically delivered higher risk-adjusted returns than greenfield mining exposure, because processing assets combine technological moats with the ability to serve multiple downstream demand sectors simultaneously.

A 2026 report from Resources for the Future (unverified) reaches the same conclusion, identifying midstream processing as the layer where market concentration, and therefore resilience risk, is highest. It names four chokepoints as the key vulnerabilities.

Ranked as investment priorities, the four midstream categories the report names look like this:

1. Lithium-6 enrichment: the hardest gap to close, with no commercial supply and a technological moat that substitution cannot cross.
2. Beryllium refining: a near-monopoly position with a decades-long reserve base behind it.
3. Tungsten APT conversion: the clearest Western sourcing problem, unsolved and directly policy-supported.
4. Helium separation: tied to unrelated gas economics, harder to position around directly.

The warning worth noting The Fusion Industry Association has cautioned that fusion risks repeating the supply vulnerabilities of the solar and battery sectors if it relies on unstable critical mineral suppliers, urging early investment in diversified supply and midstream processing.

The window is defined by lead times. Mine development runs seven to fifteen years, so any supply response to late-2030s fusion demand must originate from capital decisions made across 2026-2030. Government instruments including the US Inflation Reduction Act, the CHIPS Act, and the EU Critical Raw Materials Act are lowering the cost of those decisions for qualifying projects. The highest risk-adjusted return sits in funding processing infrastructure, not in scanning for new ore bodies.

Government instruments including the US Inflation Reduction Act and the EU Critical Raw Materials Act are expressions of resource security policy frameworks that have shifted rapidly since 2022, redirecting capital toward qualifying projects through tax credits, offtake guarantees, and concessional finance that reduce the cost of midstream investment.

What the supply chain map tells investors who are paying attention now

The core asymmetry is now visible. Equity markets have priced fusion risk into the technology developers, the reactor builders that attracted that $14.24 billion in cumulative private capital. The upstream and midstream resource companies with fusion-relevant exposure remain largely unpriced. That valuation gap is what makes this a commercial opportunity rather than a speculative bet.

The Fusion Industry Association’s 2026 funding report confirms the $14.24 billion cumulative private capital figure, with record annual inflows of $4.48 billion underscoring how rapidly institutional conviction behind reactor development has accelerated.

Three characteristics separate the most defensible fusion mineral positions from the rest:

• Multi-sector demand overlap: exposure that also serves aerospace and defence (beryllium), industrial gases (helium), or the battery supply chain (lithium) reduces single-use demand risk.
• Midstream processing capability: enrichment, refining, and conversion assets carry technological and regulatory moats that greenfield mining does not.
• Western jurisdiction alignment: projects positioned within US and EU policy frameworks can access government de-risking instruments and meet domestic content requirements.

Why the window is now Mine development takes seven to fifteen years from discovery to production. Supply for the 2030s must be funded across the 2024-2030 window, which is why timing matters more than certainty about fusion itself.

The calibrated point is this. The thesis does not require fusion to succeed on any particular timeline. The USGS already designates beryllium, helium, and lithium as critical minerals, and demand from ITER procurement, defence programmes, and clean energy applications exists independently of commercial fusion. Exposure to these supply chains delivers return potential across multiple demand vectors even if fusion deployment slips by a decade.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions. Past performance does not guarantee future results. Financial projections are subject to market conditions and various risk factors, and forward-looking statements are speculative and subject to change based on market developments.