Martian Regolith, Halogens & Materials

What the soil is

Martian Regolith Composition: Basalt, Rust, Salt and Sulfate

Remarkably uniform across the planet, which is unusual and useful.

Martian soil is weathered basaltic material, and the striking thing about it is its consistency. Viking, Pathfinder, the Mars Exploration Rovers, Phoenix, Curiosity and Perseverance sampled sites thousands of kilometers apart, and the bulk chemistry of the fine fraction came back nearly the same each time — the result of global dust storms mixing the surface for billions of years. A process developed for one site’s soil will work at another, which is not true on the Moon and not true in the belt.

Two rows are unusual enough to drive the rest of this page. Sulfur at 6–7 wt% SO₃ is roughly 60 times Earth’s crustal average, and chlorine at 0.5–0.8 wt% is roughly 300 times lunar. Mars is not a slightly better Moon in these respects — it is chemically a different kind of place.

The halogen inversion

Chlorine at 300 Times Lunar Concentration

The same perchlorate that makes the dust toxic is the reagent supply for everything past basic metallurgy.

Lunar regolith carries chlorine at roughly 20 parts per million. Producing one tonne of chlorine there means processing about 50,000 t of regolith, which is why the lunar industrial case stops at substrates and structures and why lunar aluminum is gated as firmly as lunar chip fabrication. Martian soil carries chlorine at 0.5–0.8 wt%. One tonne of chlorine takes about 154 t of Martian soil — a factor of 325 less material moved, and the soil in question is the same soil being processed for water anyway.

Better still, the chlorine arrives pre-concentrated as soluble perchlorate and chloride salts rather than locked in silicate lattice. Leaching them out is a water-wash step, not a high-temperature reduction, and the water is recovered. The halogen supply is a byproduct of a process the site already runs.

What that buys, in order of how soon it matters:

The chemistry still has to close as a loop. Halogens are not consumed so much as circulated, and every nine of recovery fraction divides the makeup throughput by ten: at a 100 t working inventory, 90% recovery demands a large continuous makeup stream, 99% a tenth of that, 99.9% almost nothing. The real specification is a recovery fraction, not an abundance — and unlike an abundance it is measurable in a terrestrial laboratory today, with simulant, for very little money.

The device gate

Why Mars Is the Only Body Where the Semiconductor Loop Could Close

Chemically possible, energetically distant. The constraint is power, and power is a schedule.

Devices sit at the top of the closure ladder at 10³–10⁵ kWh/kg — three to five orders of magnitude above bulk material, and six to eight above moving regolith. Nothing about local halogens changes that. What local halogens change is whether the rung exists at all: on the Moon it does not, at any power level, because the reagents cannot be sourced. On Mars it does.

That distinction matters more for the settlement argument than for any near-term mission. Of the shared failure modes that tie a Mars settlement to Earth — fissile fuel, semiconductors, trace stocks, spares, people, food — semiconductors are the only one that is reducible in principle, and Mars is the only body where it is. A settlement that can eventually make its own chips is a different kind of thing from one that cannot, even if the crossover is decades past the horizon of any current program.

Nothing here implies a leading-edge fab. The relevant target is the unglamorous silicon that actually fails and needs replacing: power semiconductors, sensors, motor controllers, photovoltaic cells. Those are older process nodes with correspondingly modest requirements, and they are what a maintenance-driven economy consumes.

Sulfur

Sulfur Concrete: A Binder That Needs No Water and No Cure

Melt it, mix it, pour it, let it freeze. Break it up and melt it again.

Portland cement is a poor fit for Mars: it consumes water the site would rather drink, it cures over weeks, and the chemistry misbehaves at low pressure. Sulfur concrete does none of that. Molten sulfur at roughly 130–150 °C mixed with regolith aggregate at about 35–50% binder sets on cooling in minutes, reaches compressive strengths comparable to or better than ordinary Portland concrete, consumes no water at all, and is fully recyclable — a failed pour is feedstock for the next one.

Mars supplies the sulfur. Sulfates run 6–7 wt% of the soil, tens of times Earth’s crustal average, and the same aqueous history that deposited them also left them in accessible surface beds rather than in deep ore bodies. Extraction is thermal.

The caveat is honest and worth printing: sulfur sublimates slowly in low pressure, and the published assessments of long-term durability at 6 mbar across an 80 K daily swing disagree with each other. A sealed or buried structure is a different case from an exposed one. Treat sulfur concrete as the leading candidate for pads, berms, radiation walls and foundations, and as unproven for anything pressurized, until somebody runs a multi-year exposure test that currently nobody has scheduled.

Metals

Iron, Aluminum and Why the Smelter Is an Export Rung

Metals reduction is the most expensive thing on the ladder and the site barely needs any.

Mars has the richest iron fraction in the inner system at 16–18 wt% oxide, and three plausible reduction routes: carbothermal reduction using carbon monoxide that the propellant plant already produces as a waste stream, hydrogen reduction using electrolysis hydrogen, and molten regolith electrolysis, which needs no reagent at all and a great deal of power. Aluminum is harder and needs a halide salt bath, which Mars can supply and the Moon cannot.

None of that makes a smelter a good first investment. Metals reduction at 100 t/yr draws 228–457 kW, which implies 84–167 t of power-plant mass by itself, against a site that consumes roughly 400 kg of structural metal a year and has tonnes of finished alloy sitting in the scrap yard. Regolith-to-metal is an export rung, not a closure rung — something a mature site builds when it has a customer, not something a young one builds to become self-sufficient. The popular account of off-Earth industry has this exactly inverted.

The scrap yard

Landed Hulls Are a Parts Bin, Not an Ore Body

Earth already paid for the beneficiation. The launch vehicle is the delivery mechanism.

A cargo vehicle that lands and stays is the highest-grade material on the planet by a wide margin. Salvage sits at 0.04 kWh/kg to liberate material carrying 10–15 kWh/kg of embodied energy — a ratio of roughly 300:1, which puts it off the closure ladder entirely rather than on a rung of it.

The usual framing counts the steel and stops. A 120 t dry hull is roughly 66 t of stainless, which is real, but the steel was never the binding import: the site consumes about 400 kg a year of it. What binds is components — the roughly 31% of failures a machine shop cannot address — and a landed vehicle carries about 9 t of exactly those classes: engines, fluid systems, valves, avionics, harness, motors, bearings. Salvage does not close the import rungs. It defers them, by 3 to 17 years per hull depending on how much of the vehicle was designed to come apart.

By avoided processing, the ranking also inverts. The roughly 1.5 t of copper in a hull represents about 187,500 t of regolith not processed at Martian copper abundances, against 5,950 t for the 11.9 t of chromium in its steel — a factor of 31, and the chromium figure is the one usually quoted. Non-structural material beats structural material roughly 3.5 to 1.

The Mars scrap yard is richer in kind than the lunar one, and the reason is entry. A vehicle that lands on Mars arrives with ceramic thermal protection tiles, which are furnace refractory and cryogenic tank insulation, and with four aerodynamic flaps carrying high-power electric actuators — motors, gearboxes, bearings and encoders, the exact component classes the shop cannot make. A stripped lunar lander carries neither. That partly offsets the 25.6-month resupply queue.

One recommendation follows and costs no launch mass: design cargo vehicles for disassembly. Tank internals that cut out cleanly, pre-planned penetrations, separable skin seams, common fasteners and bus voltages, residue that does not complicate habitation. It has to be decided before the vehicle is built, and at present the reusable fraction of a landed hull is zero by design.

Composites

Mineral Fiber and Matrix: Mars Can Make Both Halves

The Moon can make the fiber and not the matrix. Titan the matrix and not the fiber. Mars makes both.

Basalt-class mineral fiber is drawn from melted regolith with no reagents and no purification — a furnace, a bushing and power. The matrix is the hard part elsewhere: the Moon has carbon at roughly 150 ppm, so lunar structural composite runs through an all-mineral route or not at all. Mars has carbon in unlimited supply in the atmosphere and the polymer chemistry to go with it.

Two rules carry over from the wider materials work. No thermosets anywhere — a two-part resin system has a shelf life measured against a 25.6-month resupply window, and the arithmetic does not survive contact with that. Thermoplastic or mineral matrices only, both of which are reheatable, reformable and indefinitely storable. And carbon fiber is imported deliberately: the high-performance grades need precursor chemistry and process control far beyond an early settlement, the mass involved is small, and it belongs on the permanent-import list beside HALEU, plutonium-238, iridium and helium rather than being treated as a closure target.

Ore bodies

The Only Body Outside Earth Where Concentration Actually Ran

Terrestrial ore bodies were made by water, heat and time. Mars had all three, briefly.

Earth’s ore deposits exist because liquid water, hydrothermal circulation and plate tectonics spent billions of years sorting a well-mixed crust into concentrated seams. Nothing else in the solar system ran those processes at scale, which is why off-Earth abundances sit close to cosmic values and why the permanent imports are always the trace elements rather than the bulk ones. The general rule that follows — substitution beats beneficiation — holds everywhere in this series.

Mars is the partial exception. It had surface water, hydrothermal systems, evaporite basins and volcanic differentiation, all of which stopped long ago but all of which ran. The observed sulfate and chloride beds are themselves evaporite concentration products. Crustal uranium and thorium sit in the fractions of a part per million — one to two orders above chondritic material and within about one order of Earth’s crustal average — which makes indigenous fissile material geologically plausible here and nowhere else in the inner system.

Nobody has prospected for any of it. Every orbital and surface instrument sent to Mars was designed to answer questions about habitability and geological history, not to find ore. That is a named gap rather than a finding, and it is the kind of gap that a single purpose-built instrument on an existing orbiter could start to close.

Record

Provenance and Data Sources