Is lithium enrichment a problem?

[Red]

Fusion’s Lithium Bottleneck: A Critical Hurdle

• Lithium-6 is essential for sustaining the most common fusion reactions, but it comprises only 7.5% of natural lithium.

• Fusion fuel requires enrichment—boosting lithium-6 content to 50–90%. Current tech can’t scale enrichment to meet future demands.

• Enormous quantities needed: Just one demonstration fusion plant will require 10–100 tonnes of enriched lithium.

• Existing supply nearly zero: The U.S. has a Cold War-era stockpile (~442 tonnes), but it was produced with toxic mercury methods now banned and still being remediated.

• Enrichment urgency: First grid-connected fusion plant projected around 2040, but development of scalable, non-toxic enrichment methods must begin now.

[Doubting_Thomas]

@Red

Just one demonstration fusion plant will require 10–100 tonnes of enriched lithium.

DEMO is projected to contain ~1400 m^3 of plasma in a torus with a major radius R = ~9 m. Using the formula V\_{plasma}=2π^2Rr^2, the estimated minor radius r = ~2.8 m. If the breeding blanket is assumed to be one meter thick (t = 1 m), its volume becomes:

V\_{blanket} = 2π^2 R ((r+t)^2 - r^2) = 1173 m^3

Lithium is very light—about half the density of water. Still, 1173 cubic meters of lithium weighs (534 kg/m^3)(1173 m^3) = 626 metric tons. That’s six times Red’s upper-bound estimate of 100 tonnes. But, to support a tritium breeding ratio (TBR) > 1, the blanket will probably need to be more than 1 meter thick.

If you built a blanket with just 10 tonnes of lithium, it would only be 2 cm thick—basically useless.

[Banana]

@Doubting_Thomas

…626 metric tons. …But, to support a tritium breeding ratio (TBR) > 1, the blanket will probably need to be more than 1 meter thick.

MCF blanket designs are usually shaped like chocolate frosting on a donut, a geometry that wastes a lot of material. In contrast, an ICF “contact“ blanket is more like the chocolate on a chocolate covered raison. Contact blankets use the least material and occupy the minimum space when they are moved close to the point of ignition. For example, a 2.5 meter thick spherical contact blanket would require only 4/3 π (r=2.5)^3 cubic meters of blanket material. (65 m^3, 35 tonnes of lithium) Thats a tiny fraction (1/18) as much lithium as Tom’s torus requires, while boasting a 2.5x thicker blanket with better protection and increased TBR.

[Fermi]

@Red

Fusion fuel requires enrichment—boosting lithium-6 content to 50–90%.

Hogwash!

A thick blanket of natural lithium (7.5% Li-6) would see its tritium breeding ratio (TBR) increase if its lithium-6 content decreased. The reason for this is that when Li-7 is hit by an energetic neutron it produces two tritium, one instantly (through spallation) and the other when the second neutron thermalizes and is eventually captured by Li-6:

^{7}Li+n \rightarrow ^{3}H + ^{4}He + n

[BS-Fusion]

Even if lithium enrichment isn’t a problem, its no solution.

The contour plots below are based on data from 1,029 OpenMC simulations, each tracking 150,000 neutrons at 14.1 MeV, emitted from a point source at the center of a lithium sphere. All simulations were identical except for two variables: the sphere’s radius (i.e., blanket thickness) and the lithium-6 enrichment level, which I varied in 5% increments. For reference, a horizontal line marks Li-6 at 7.5%—the natural, unenriched baseline.

Notice (above), for blankets <= 50 cm thick, the contour lines slant up to the right. That means going straight up (increasing the proportion of Li6) lowers the TBR. To reach TBRs > 1, blankets need to be > 70 cm thick. In all cases, extreme enrichment (near 100% Li6) results in lower TBR. For example, consider a 100 cm blanket. Its TBR drops from 1.20 (natural 7.5% Li6) to 0.95 when fully enriched (100% Li6).

The contours in the plot above correspond to the proportion of 14.1 MeV neutron energy from the initial fusion release that is deposited in the blanket. To calculate the average energy per neutron, multiply the contour values by 14.1 MeV. Note that for blankets less than 75 cm thick, a significant amount of energy is wasted—deposited outside the blanket. The simulation only considered prompt reactions as contributors to blanket heating. Delayed reactions, such as Li‑8 → Be‑8* (t₁/₂ ≈ 0.838 s) → 2α + e⁻ + ν̄ₑ, which follows the prompt Li-7 + n → Li‑8 + ~2 MeV gamma neutron capture reaction, were not included. However, the effect of including delayed reactions would be imperceptible, since fewer than 3% of neutrons contribute to delayed reactions, and most of beta decay energy escapes the blanket, carried away by high-energy neutrinos.

Leakage refers to the fraction of neutrons that escape from the tritium-producing blanket. These high-energy neutrons pose serious risks: they can damage surrounding structural materials and potentially cause biological harm, including mutations. In an ideal design, all neutrons would be absorbed within the blanket. However, when the blanket is less than 1 meter thick, even maximum lithium enrichment cannot prevent more than 15% of the neutrons from leaking out.

All the data point the same way: thicker blankets work best. A compact blanket—a design that achieves the necessary breeding thickness within a minimized radial footprint, reducing both structural bulk and material costs—remains the most cost‑efficient solution for thicknesses beyond 2.5 meters.

The plot (above) shows how blanket volume scales with distance from the center. For a 2.25‑meter‑thick lithium blanket, placing it 5 m from the core requires ~1075 m³ of lithium, while a compact design at 0 m achieves the same thickness with only ~50 m³. This illustrates the dramatic material savings (over 20×) that can be achieved by compact blanket geometries.