Helium-3

The isotope

What Is Helium-3: Physics and Why It Matters

Two protons, one neutron. Rare on Earth because the atmosphere does not capture it and the crust does not produce it.

Helium-3 (³He) is a stable, non-radioactive isotope of helium with two protons and one neutron, versus the two neutrons of common helium-4. On Earth it is extraordinarily scarce — roughly 0.000137% of all helium is He-3. Virtually the entire terrestrial supply comes from the radioactive decay of tritium produced in nuclear reactors for weapons programs. The U.S. Department of Energy controls the primary domestic stockpile. Annual production is roughly 22,000–30,000 liters against estimated demand of 40,000–60,000 liters — a structural deficit that has driven prices from $40–$85 per liter before 2001 to $1,900–$2,600 per liter in bulk and as high as $10,000–$20,000 per liter for refrigerator-grade purified material. Observed

He-3 matters for three reasons: it is the working fluid in dilution refrigerators that cool quantum computers to millikelvin temperatures, it is used in neutron detectors for nuclear security and safeguards, and it is the theoretical fuel for aneutronic deuterium-helium-3 fusion. The first two markets exist and are supply-constrained today. The third does not yet exist at any scale.

He-3 on the Moon

Solar Wind Implantation: Where Lunar Helium-3 Comes From

Four billion years of unshielded solar wind, depositing He-3 atom by atom into the top few nanometers of every grain.

The Moon has no atmosphere and no magnetic field. Solar-wind ions — including He-3 at roughly 4–20 parts per billion by mass — are implanted directly into the outer surfaces of regolith grains. Over billions of years of exposure, the cumulative inventory is large: estimates range from hundreds of thousands to over a million tonnes of He-3 in the entire lunar regolith, though distributed at ppb concentrations across the full surface.

Concentration varies by terrain. Mature, fine-grained mare soils that have been exposed to the solar wind for longer periods and have more surface area per unit mass contain the highest concentrations — up to roughly 20 ppb. Young, coarse highland soils contain less. The equatorial mare regions (Oceanus Procellarum, Mare Tranquillitatis) are the richest targets, not the polar sites where Artemis and ILRS are landing. The best He-3 sites and the best water-ice sites are on opposite ends of the Moon. Observed

Extraction is thermal: heating regolith to roughly 700°C releases implanted solar-wind volatiles, including He-3, He-4, hydrogen, nitrogen, and carbon species. The He-3 must then be separated from this gas mixture — a cryogenic distillation problem at scale.

The market that exists now

Quantum Computing and Neutron Detection: The Near-Term He-3 Demand

The fusion case is speculative. The quantum computing case is not.

Every superconducting quantum computer on Earth runs inside a dilution refrigerator — a human-sized plumbing system that uses a mixture of liquid He-3 and He-4 to reach temperatures below 10 millikelvin. Bluefors (Finland) is the dominant manufacturer. A single dilution refrigerator costs roughly $600,000 and uses tens of liters of He-3. At peak pricing, the He-3 alone can add $200,000 to the cost of the system.

The quantum computing buildout is the primary demand driver. IBM, Google, Microsoft, and dozens of national laboratories are scaling qubit counts, and each system needs its own refrigerator. As quantum arrays scale to thousands and then millions of qubits, individual systems could require hundreds or thousands of liters. The existing terrestrial supply — dependent on tritium decay from nuclear weapons programs in the United States and Russia — cannot scale to meet projected demand.

Neutron detection is the second market. He-3 proportional counters are the standard technology for detecting neutrons at nuclear facilities, border crossings, and ports. After September 11, 2001, the U.S. deployed thousands of detectors, draining the strategic reserve and triggering the price spike that has never reversed.

A third market — hyperpolarized lung MRI imaging — uses He-3 for diagnostic medical imaging but at smaller volumes.

None of these markets requires fusion to work. They are real, operating, and supply-constrained today. Observed

One countervailing development: DARPA has issued an urgent call for cryogenic cooling technologies that eliminate He-3 entirely. In early 2026, a group demonstrated a continuous adiabatic demagnetization refrigerator (cADR) that reaches 20 millikelvin in single-shot mode, holds below 30 millikelvin continuously, and uses no He-3 at all. The German company kiutra has commercialized a version. If continuous ADR reaches production scale for large quantum arrays, it could reshape He-3 demand faster than any lunar harvester reaches the regolith. Building

Interlune

Interlune: The Company Building the Extraction Hardware

Founded by an Apollo 17 moonwalker and the former president of Blue Origin. Funded by NASA. Hardware in the lab.

Interlune is a Seattle-based natural-resources company founded in 2020 by Harrison Schmitt (Apollo 17 lunar module pilot, the last geologist to walk on the Moon) and Rob Meyerson (former president of Blue Origin). It is the only company with a funded program to extract He-3 from lunar regolith.

The Prospect Moon payload is the critical near-term deliverable. It will measure volatile-gas concentrations in heated regolith on the lunar surface — the first in-situ measurement of He-3 release rates, replacing reliance on Apollo-era sample inference. Interlune prefers equatorial mare sites for maximum He-3 concentration. The company plans multiple copies of the relatively inexpensive payload on different commercial landers. Building

Interlune’s business model does not depend on fusion. Revenue from He-3 sales to quantum computing and DOE customers funds the development of broader lunar regolith-processing infrastructure. The company explicitly plans to expand into hydrogen extraction, industrial metals, rare earths, and water once regolith-processing hardware is operational. He-3 is the wedge product. The regolith-processing platform is the business.

The fusion case

Deuterium-Helium-3 Fusion: The Theoretical Fuel and Its Missing Reactor

The physics is real. The reactor is not.

The deuterium-helium-3 (D-³He) fusion reaction produces a proton and a helium-4 nucleus. Unlike the deuterium-tritium (D-T) reaction used in most fusion research, D-³He is aneutronic — it produces charged particles rather than high-energy neutrons. That means less radiation shielding, less neutron damage to reactor walls, less radioactive waste, and the possibility of direct energy conversion from charged-particle kinetic energy to electricity without a thermal cycle.

These are genuine advantages. They are why He-3 fusion has been discussed since the 1950s and advocated by lunar-return proponents since the 1980s. But D-³He requires plasma temperatures roughly six times higher than D-T — on the order of 600 million kelvin — and D-T fusion itself has not yet achieved sustained net energy gain at commercial scale. No D-³He reactor has been built. No design has passed conceptual review. The fusion case for lunar He-3 requires a reactor that does not exist to burn a fuel that is not yet extracted from a surface no one has mined.

This page does not dismiss D-³He fusion. It notes that the fusion demand case is multiple technological breakthroughs away from creating a market, and that the quantum-computing and neutron-detection demand cases already exist. Modeled

Extraction economics

Processing Scale: How Much Regolith for How Much He-3

At 20 ppb, one tonne of He-3 requires processing roughly 50 million tonnes of regolith.

The concentration arithmetic is the central challenge. At a rich mare site with 20 ppb He-3 by mass, extracting one kilogram of He-3 requires heating roughly 50,000 tonnes of regolith to ~700°C, capturing the released gases, and separating He-3 from a mixture dominated by He-4 and hydrogen. Interlune’s excavator prototype processes 100 tonnes per hour; at that rate, one kilogram of He-3 takes roughly 500 hours of continuous operation — about 21 days.

At current Earth prices ($2,000–$20,000 per liter, depending on purity grade), He-3 is one of the most valuable substances by mass on Earth. One kilogram of He-3 gas at standard conditions occupies roughly 7,400 liters. Even at the lower bulk price of $2,000 per liter, one kilogram is worth roughly $14.8 million. The value-to-mass ratio is high enough that return-to-Earth logistics may close economically even before any large-scale lunar infrastructure exists — if the extraction and Earth-return costs per kilogram fall below the sale price. This is Interlune’s thesis: He-3 is the first lunar resource valuable enough per kilogram to be worth shipping home.

The constraint is not the value. It is the processing volume. 50,000 tonnes of regolith per kilogram of He-3 requires industrial-scale excavation, continuous thermal processing, and gas handling — all on the lunar surface, in vacuum, with dust. The Vermeer partnership and NASA MMPACT program are developing exactly this class of hardware, but no system has operated at this scale off Earth. Projected

Adjudication

What Is Real, What Is Funded, and What Requires a Breakthrough

This page exists because lunar He-3 is the topic people ask about most — and the one most commonly either dismissed or oversold.

The honest summary: the extraction side is further along than most people think. The demand side is real but different from what most people think. The near-term market is quantum computing refrigeration and nuclear security, not fusion. The extraction is funded, has hardware in the lab, and is scheduled to fly to the Moon in 2028. The economic case may close on quantum-computing demand alone if the extraction and return logistics costs come down — or it may be undercut by He-3-free cooling alternatives (cADR) reaching production scale first. Fusion remains a possibility measured in decades, not years.

Record

Provenance and Data Sources