The frame
Everything Off Earth Sits Near Cosmic Abundance
Earth concentrated its ores with liquid water, plate tectonics, and hydrothermal circulation over billions of years. No other body ran those processes, so the comparison that matters is local abundance against ore grade, not against Earth's crust.
On Earth, a copper mine exists because flowing water and heat spent eons sorting copper from everything around it. The Moon, the C-type asteroids, and Mars never had that machinery running at scale, so their elements sit close to the abundances the solar nebula left behind. The light, rock-forming elements are everywhere; the trace elements that terrestrial industry leans on are absent almost everywhere. A register of what a site can source has to score each element against the grade needed to use it, destination by destination. Observed
The register
What Each Element Is, and Where It Is
Scored by role and by whether a site can win it at a usable grade. Gold is abundant, magenta a permanent import, dashed a gap the register has not yet filled.
| Element | Role | Off-Earth abundance | Status |
|---|---|---|---|
| Oxygen | Oxidizer, breathing gas, structure | 40-45% of lunar regolith, bound in oxides | Abundant |
| Silicon | PV, glass, substrate | Rock-forming, every regolith | Abundant |
| Iron | Steel, structure | Regolith plus M-type asteroids | Abundant |
| Aluminum | Conductor, structure | ~10 wt% of lunar regolith | Abundant |
| Carbon | Reductant, fiber, polymers | Lunar ~150 ppm; C-types and Mars far richer | Scarce on the Moon |
| Nitrogen | Buffer gas, fertilizer | Lunar ~80 ppm; Mars 2.6% of air; Titan abundant | Scarce on the Moon |
| Hydrogen (water) | Propellant, life, reductant | Polar ice; present or absent by site | Site-dependent |
| Chlorine / halogens | Device etch and processing | Lunar ~20 ppm vs Mars 0.5-0.8 wt% (~300x) | Lunar wall |
| Copper | High-conductance wiring | ~8 ppm everywhere | Substitute with Al |
| Lithium | Batteries | ~1,000x below ore grade on the Moon | Substitute with Na |
| Iridium / PGMs | Electrolysis, catalysts | Trace in every regolith | Permanent import |
| Uranium (HALEU) | Fission fuel | No belt uranium; enrichment is terrestrial | Permanent import |
| Plutonium-238 | Radioisotope heat and power | Made on Earth at ~0.5 kg/yr | Permanent import |
| Helium | Cryo pressurant, purge, leak-check | Gas giants only; unsynthesizable | Permanent import |
| Mo, Nb, Ta, rare earths | Superalloys, electronics, magnets | Trace; a Raptor contains a molybdenum bushing | Register gap |
The permanent-import rows total a few hundred kilograms at an outpost, and they are the rows that gate it. The gap rows - molybdenum, niobium, tantalum, the rare earths - are load-bearing and currently unscored, so the permanent-import list understates itself.
Permanent imports
Five Stocks a Site Cannot Make
The elements with no accessible off-Earth source set the resupply clock, and together they weigh almost nothing.
Five stocks recur across every architecture as things that have to come from Earth: high-assay low-enriched uranium for fission, plutonium-238 for radioisotope power, halogens for device fabrication on the Moon, iridium and the other platinum-group catalysts, and helium as the only practical cryogenic pressurant, purge, and leak-check gas. Helium is the quiet one - the propellant circulatory system of the whole architecture is pressurized by a gas that exists only inside the gas giants, and no inner-system destination can make it. The combined mass of all five at a minimum outpost is a few hundred kilograms, which is the point: the binding imports are never the bulk materials, they are the trace ones. Building
Substitution
Substitution Beats Beneficiation
Rather than build a concentration plant for a trace element, redesign the process around an abundant one. Off-Earth industry needs the chemistry Earth abandoned when concentration became free.
| Replace | With | Why it works |
|---|---|---|
| Copper wiring | Aluminum | 2.1x better per unit mass at equal conductance; Al 10 wt% of regolith against Cu at 8 ppm |
| Lithium cells | Sodium-ion | 120-160 Wh/kg; moves the battery-versus-fuel-cell crossover from 24 to 13 kWh |
| Platinum-group electrolysis | Nickel-alkaline | Removes the iridium dependency in the reference electrolyzer |
| Carbon fiber | Mineral (basalt) fiber | Mineral matrix needs no imported carbon; carbon fiber imported only where deliberately chosen |
Aluminum for copper is a free win: more conductance per kilogram and a feedstock four orders of magnitude more abundant. The crossover cases show that a sourcing decision, not a law of physics, often sets where the line between two technologies falls.
The halogen inversion
The Halogen Wall Is Lunar, Not Martian
The element that decides whether a site can make its own electronics is ~300x more abundant on Mars than on the Moon.
Chlorine runs about 20 ppm in lunar regolith and 0.5 to 0.8 wt% in Martian soil - a factor near 300. Halogens are what etch and process semiconductors, so the halogen wall that confines the Moon to making substrates rather than finished devices simply does not exist on Mars, where the perchlorate that complicates everything else is a feedstock. The Moon's industrial disadvantage is chemical and permanent; Mars's is logistical and improves every time the cadence does. That cuts against the familiar Moon-as-industry, Mars-as-settlement framing. Projected
The scrap yard
The Landed Hull Is the Best Ore Body
The concentration thesis run in reverse: Earth already paid to refine the metals, and the launch vehicle delivered them.
A spent cargo Starship is roughly 66 tonnes of 304L stainless - the concentration thesis working the useful direction, because Earth already spent the energy to sort those metals and the vehicle carried them to the surface. By register weight the copper in a hull matters most: about 1.5 tonnes of copper at 8 ppm stands in for 187,500 tonnes of regolith that would otherwise have to be processed, against roughly 5,950 tonnes for the chromium the earlier accounts quoted - a factor of 31. The whole hull is worth on the order of 272,000 tonnes of regolith equivalent, and the non-structural metals outweigh the structural ones more than three to one. Building
Contested
Deliberately Not Claimed
The load-bearing uncertainties, tagged with the confidence they earn.
Cosmic abundances do not age. The light-element abundances and the trace-element scarcities are measured facts about each body, and they will read the same in a century.
The register has gaps that change its own conclusion. Molybdenum, niobium, tantalum, and the rare earths are absent as rows, and because each is a permanent import, the import list currently understates itself.
Device-grade closure is the open question. Whether a site can ever make finished electronics turns on halogen recovery fractions and trace-element supply that have not been measured off Earth, so the top of the ladder stays projected.
Record
Provenance and Data Sources
- Confidence
- Observed for abundances and the substitution physics - Building for the register completeness and scrap-yard figures - Projected for device-grade closure
- Status date
- Q4 2026
- Organizing thesis
- Everything off Earth sits near cosmic abundance; score local abundance against ore grade, not against Earth's crust; substitution beats beneficiation
- Permanent imports
- HALEU, plutonium-238, halogens (lunar), iridium and platinum-group catalysts, helium - a few hundred kilograms that gate an outpost
- Known gaps
- Molybdenum, niobium, tantalum, and the rare earths are unscored rows
- Related
- The Primitives Register for the hardware these elements build; Water and Energy for the two abundant-element substrates; Substrates for the frame
- Weakest link
- The unscored trace-metal rows and the unmeasured halogen recovery fraction