Nickel-Iron & Carbonyl Metallurgy

Ore already smelted

Asteroid Nickel-Iron Is Scrap, Not Ore — and Scrap Is the Right Comparison

The expensive half of steelmaking is stripping oxygen off iron. That step is already done out here.

Terrestrial iron comes out of the ground as oxide — hematite, magnetite — and the blast furnace exists to take the oxygen away using carbon at 1,500°C. That reduction is where most of the energy and nearly all of the carbon dioxide in primary steelmaking goes. Asteroidal nickel-iron has no oxygen on it. It condensed as metal in the solar nebula and has stayed metal ever since, so the only thing left to do is heat it until it is workable.

Melting it from belt temperature is straightforward arithmetic. Iron's mean specific heat across the range is about 600 J/kg·K, and the trip from 160 K to the 1,811 K melting point is 1,651 K:

600 × 1,651 = 991 kJ/kg  →  0.275 kWh/kg
plus latent heat of fusion, 247 kJ/kg  →  0.069 kWh/kg
total 0.344 kWh/kg ideal

Induction efficiency brings the realized figure to roughly 0.53 kWh/kg. That is parity with remelting scrap in a terrestrial electric arc furnace, which runs in the same range. It is not the 20-fold advantage sometimes claimed against the 5.6 kWh/kg figure for primary steel — that comparison puts an electrical number against a primary-energy number and counts the reduction step the asteroid never needed twice over.

Parity with scrap is the honest claim and it is still a strong one. A belt operation melting its own metal spends about what a recycling plant on Earth spends, with no mine, no furnace reduction chemistry, no slag handling and no carbon dioxide. The scrap yard is the correct mental picture for the belt’s metal supply. Modeled

What the metal is

Iron Meteorite Composition: a Stainless-Adjacent Alloy Nobody Had to Mix

90 to 95% iron, 5 to 10% nickel, with cobalt along for the ride.

Iron meteorites are the best-characterized extraterrestrial material in existence, held in tonnes rather than grams — Canyon Diablo, Gibeon, Sikhote-Alin and hundreds of others. They are fragments of the cores of planetesimals that were destroyed early, and their interlocking Widmanstätten crystal bands took millions of years to grow, which is only possible inside a body cooling slowly under insulation.

The composition is usable as it comes out. Nickel at 5 to 10% is in the range that gives iron good toughness and corrosion behavior, and cobalt adds high-temperature strength. What also rides along is less welcome: phosphorus as schreibersite and sulfur as troilite, both of which embrittle steel and both of which a real process has to remove.

These figures come from meteorites whose parent bodies no longer exist. No sample has ever been returned from a known metal-rich asteroid, so surface condition, coherence and regolith depth at an actual target remain unmeasured — see Vesta, Psyche & the Metal Bodies.

The Mond process

Carbon Monoxide Picks Metal Up as a Gas at 60°C and Puts It Down at 200°C

Ludwig Mond found this in 1890 and a refinery in Wales has been running it since 1902.

Nickel reacts with carbon monoxide at atmospheric pressure and 50 to 60°C to form nickel tetracarbonyl, a volatile liquid that boils at 43°C. Heat the vapor to 200 to 250°C and it falls apart again, depositing pure nickel and releasing the carbon monoxide intact:

Ni + 4 CO  ⇆  Ni(CO)₄
Fe + 5 CO  ⇆  Fe(CO)₅

The carbon monoxide is a catalyst, not a consumable. It goes around the loop again, and what is consumed is heat and whatever gas leaks out. That single property is what makes the process viable at a place where every kilogram of reagent would otherwise be a 15.5-month import.

The two reactions run under different conditions, which is a feature. Nickel volatilizes at one atmosphere; iron needs roughly 100 to 200 atmospheres at 150 to 200°C. Running the vessel at low pressure therefore extracts the nickel and leaves the iron behind — the separation is built into the chemistry rather than requiring a separate refining stage. This is precisely how the process separates the two metals on Earth.

None of it is speculative chemistry. Vale’s refinery at Clydach in south Wales has produced carbonyl nickel continuously since 1902. BASF has made carbonyl iron powder at Ludwigshafen for decades, and Chinese producers now supply a large share of the world’s carbonyl iron powder. The process is more than a century old, industrially routine, and has never been asked to run in vacuum on a body with no gravity. Observed

Why it fits the belt

Gas-Phase Refining Does Not Care About Gravity

Melting does. That is the whole argument for doing it this way out here.

Every liquid-phase metallurgical process on Earth relies on gravity without saying so. Slag floats because it is less dense; inclusions settle; bubbles rise and leave the melt; the pour falls into the mold. In microgravity none of that happens. A melt on a small asteroid is a blob with its impurities distributed through it, its gas trapped inside it as porosity, and no way to separate anything by density. Containerless melting by electromagnetic levitation becomes easier in microgravity, which is a genuine advantage, but the separation problem does not go away.

Carbonyl chemistry never makes a liquid pool. The metal leaves the feedstock as a gas, travels as a gas, and is deposited as a solid onto a surface. Density plays no part, so the process runs identically at 0.0289 g on Ceres, at three hundred-thousandths of a g on a rubble pile, and in free space alongside a captured body. The separation from silicate gangue is also automatic: the rock does not form a carbonyl and simply stays where it is.

The second reason the belt suits this process is the carbon. Carbonaceous bodies make up roughly three-quarters of the belt population by number and carry 2 to 5% carbon by weight, along with the water that provides the oxygen. Carbon monoxide is therefore a local product. The Moon has essentially no carbon, which is why this route is discussed for the belt and not for Shackleton, and it is one of the few genuine industrial advantages the belt holds over closer destinations.

Vacuum helps too. There is no atmosphere to exclude, so the reaction vessel does not need an inert purge, and clean metal surfaces stay clean. Modeled

What it makes

Vapor Deposition Produces Parts, Not Ingots

The output is a finished shape, and the shape is set by what the gas was allowed to land on.

Decomposition deposits metal onto any heated surface, so a form held at 250°C inside the chamber grows a metal skin conforming to it exactly. Dissolve or withdraw the form afterward and what remains is a hollow part with the inside dimensions already correct. Thin-walled pressure vessels, tubing, ducting, shells, radiator panels and foams all come out of the process directly, at 99.9% purity or better, without a mill, a press, a lathe or a mold.

That matters more than the energy figure. A belt operation's binding problem is a 30.8-month worst-case spares queue, and a route that goes from raw body to finished part inside one facility is the only thing that shortens it. Melting and casting would still require rolling, machining and joining equipment, each piece of which is itself an import with its own failure modes and its own spares. Carbonyl deposition collapses several of those stages into one vessel.

What it does not do is make everything. Deposition rates are slow, so large structural members are better cast; the process suits thin-walled and complex geometry rather than heavy sections. A realistic shop runs both — induction melting for bulk structure, carbonyl deposition for parts with internal geometry — which is the configuration described on the Machine Shop MVP. Modeled

What it does not solve

Carbonyl Metallurgy Gives You Iron and Nickel, and Nothing Else

Partial closure is the honest description, and the gaps are specific.

The carbonyl route works for nickel, iron and cobalt. It does not work for the elements that turn iron into engineering steel — chromium, molybdenum, vanadium, manganese — and it does not touch aluminum or titanium at all. A belt operation using this process can make structural iron-nickel parts indefinitely and still has to import every stainless alloy, every high-temperature superalloy, and every aluminum or titanium component. The closure is real and it is partial, and what stays on the import list is covered under Permanent Imports.

Three other constraints belong on the record.

Record

Provenance and Data Sources