Engines burning aluminium and liquid oxygen have been investigated for use on the moon because there is plenty of both elements in the soil - about 13% and 40% of highland soil, respectively. Propellant thus has the potential to be sourced from the moon directly, but it isn't practical unless a refining process can be found that uses the moon's environment to extract pure aluminum efficiently using local resources. A viable way to extract the oxygen is known already, using a pyrolysis process.

Is there a way of taking advantage of the moon's environment to do this effectively? You have at your disposal:

  • A hard vacuum
  • A month-long day that is strong sunlight for two weeks then a two week night - it is estimated that a furnace using concentrated sunlight could reach 3000 K.
  • 1/6 Earth gravity
  • No moisture - chemicals are reduced
  • Powdery soil that is loose to a depth of 10 or 15 cm
  • Some soil samples from Apollo 16 were a quarter alumina by weight

A source of fuel in space is a prerequisite for a larger ongoing human presence there. A rocket launched from Earth can't manage to be more than about 5% payload. Shipping fuel from Earth to any off-world base by rocket would be prohibitively expensive (even if the rocket is reusable). This is why there has been so much interest in sources of water ice in space - it can be electrolysed into liquid hydrogen and liquid oxygen, which makes excellent propellant. But you have to go to the asteroid belt to get it (there are indications there might be significant reserves in craters at the moon's poles in permanent shade, but this is inconclusive). Once there is a way to refuel in space, all sorts of things become possible that are beyond our reach right now. An option that can compete with the price of mining asteroids has business potential. All such undertakings are going to require vast initial outlays that won't pay off for a decade or more. The process being considered here only has to meet that standard.

This question is a partner to a question at SE's Space Exploration site - How feasible is it to use aluminum and liquid oxygen as future propellant sourced from the moon?. I realized it was largely a chemistry question and so decided to ask about it here. There are two references in the other question: one from Wickman Spacecraft & Propellant, and one from the Artemis Data Book. I'm afraid neither discusses the chemistry and I haven't found anything else. The AIAA has several papers on it in their catalogue.

  • $\begingroup$ Given the insurmountable transportation costs of the newly-mined raw material, I don't see how anyone will be willing to foot the bill. Therefore, I think the possibility for this refining to occur is essentially zero. $\endgroup$ – LordStryker Nov 18 '14 at 15:32
  • 1
    $\begingroup$ The purpose of this is to create fuel to be used in space. I should probably edit the question to reflect that. $\endgroup$ – kim holder Nov 18 '14 at 16:13
  • $\begingroup$ Okay, now i've added a paragraph on the value of propellant sourced in space. $\endgroup$ – kim holder Nov 18 '14 at 16:35
  • $\begingroup$ While my comment was meant to be tongue-in-cheek, I'm glad to see the addition of content to your post. It makes it much more interesting. $\endgroup$ – LordStryker Nov 18 '14 at 17:11
  • 1
    $\begingroup$ That's how this process works, right? Glad to get comments like this. And i'm a little thick about word-play sometimes. $\endgroup$ – kim holder Nov 18 '14 at 17:15

My theory was to use electromagnetic separation, aka the Calutrons used in the Manhattan Project. Fine powder feedstock (check), into an arc discharge to vaporize and ionize (lots of good designs for high volume ion sources out there), electrostatic acceleration, then electromagnets to mass separate. Power with a solar panel array. No moving parts (except for the powder feed). No vacuum systems needed. And you separate all of the elements at once. Now, it only works during sunlight, but during the lunar night you could do maintenance and collect the separated elements.

  • $\begingroup$ Alright, in a few hours i'll be hunting for material online about all this - Calutrons, electrostatic acceleration, electromagnetic mass separation, vaporization by arc discharges. This gives me lots to go on, i appreciate it. Can you suggest good sources of information? It needs to be available online, but i'll buy something if it has what i want to know in it. $\endgroup$ – kim holder Nov 18 '14 at 19:58
  • 1
    $\begingroup$ Wikipedia has a write-up on the Calutrons. Separation of uranium isotopes is really hard compared with refining ores (the difference in U-235 vs 238 and others is a small mass fraction). Other useful info might also be found for AMS (atomic mass spectroscopy), used for, e.g. C-14 dating. Again, they want a very precise, single pass measurement. Cascades of Calutrons might be more efficient. Van der Graaf or Cockroft-Walton accelerators are also on Wikipedia. Ion source are probably best found in the technical literature which you may not have easy access to. $\endgroup$ – Jon Custer Nov 18 '14 at 20:26
  • $\begingroup$ ...and wouldn't understand anyhow. I've gotten used to just letting what i'm reading wash over me and absorbing what i can. My theory is that eventually that can add up to enough working knowledge if i round it out by learning what i really need to know. Once i have looked at that stuff i may return with more questions. :) $\endgroup$ – kim holder Nov 18 '14 at 21:16
  • $\begingroup$ @kimholder Meteoric impacts on The Moon should have vaporized lots of anorthite, but did it separate into corundum and other minerals ? And ionization of anorthite will not separate it into minerals either , will it ? So why the mass separation ? I think only high enough temperature (and pressure) will do the job. www.minsocam.org/ammin/AM65/AM65_272.pdf $\endgroup$ – Conelisinspace Jun 29 at 16:09
  • $\begingroup$ The ion source, designed properly, would rip it all apart. $\endgroup$ – Jon Custer Jun 29 at 16:37

I don't think this is going to be a good idea.

Notice that the Al and O existing in the Moon's soil are not there are free elements, but are tightly bound in minerals.

On Earth, Al is extracted from bauxite - this is naturally occurring Al2O3. This extraction process is extremely energy intensive. In fact, they transport the bauxite mined elsewhere to Iceland because the electricity in there is so cheap, instead of doing it on the spot of the mine, or nearby.

The Al and O that occur on the Moon are mostly in the form of the mineral anorthite plagioclase: CaAl2Si2O8. Extracting the Al from there would be even harder! You first have to extract the Al2O3 out of the CaAl2Si2O8, and only then you can get the Al out. This mineral is not unique to the Moon. In fact, anorthite, albite (NaAlSi3O8) and all minerals in between are the most common minerals in Earth's crust, and we still don't use them as a source for Al.

That said, burning Al as a fuel would not be a good idea because the end result is a solid, not a gas.

  • $\begingroup$ The point about solid reaction product was made before, but i never answered because i wasn't able to get a more professional opinion than mine. Basically, there has to be a lot of extra oxygen in the mix. The heat of the Al + O2 reaction rapidly converts it from liquid to gas and provides thrust. As for the difficulty of refining ore, there is no easy solution to fuel on the moon. With great luck there will be lots of easily accessible water ice in polar craters, barring that, this could well be the best solution. $\endgroup$ – kim holder Dec 13 '14 at 21:33
  • $\begingroup$ You could also burn aluminum with ammonium perchlorate, making a solid fuel mixture whose combustion does give off gases (nitrogen, water vapor, maybe volatilized aluminum chloride). Now all you need is to get said ammonium perchlorate on or to the Moon. Oops. $\endgroup$ – Oscar Lanzi Apr 15 at 2:06

The standard process for electrolysing $\ce{Al2O3}$ to $\ce{Al}$ should work fine on the moon. Note that on Earth, a consumable carbon anode is used to reduce the voltage required, giving the reaction (Hall–Héroult process)

$$\ce{Al2O3 + 1.5C → 2 Al + 1.5 CO2}$$

An alternative, non-consumable anode would be used on the moon, to avoid the need for anode material and to give oxygen instead of $\ce{CO2}$.

The process for purifying $\ce{Al2O3}$ would be rather more difficult. On Earth, this is done by dissolving bauxite in sodium hydroxide solution, filtering out insoluble iron salts, cooling, precipitating out $\ce{Al2O3}$, and discarding the liquor containing silicates.

This process is highly intensive in water and sodium hydroxide. Major advances would be needed to make it less intensive in these materials, and to make the electrolysis step more tolerant of impurities.

I think we are better off looking for water to use as feedstock for our propellant factory. It's far easier to purify (all you have to do is distill it).


Ionic liquid refining of aluminum is a promising solution which uses 3 kWh of electricity per kg instead of around 15 kWh for existing methods.

  • 7
    $\begingroup$ The Gibbs free energy for the decomposition of 1 mol of Al2O3 is about 1.5MJ (for 54g of Al.) That makes 27.8MJ/kg = 7.7kWh (though a consumable carbon anode can reduce this by about a third due to the reaction Al2O3 + C --> 2Al + 1.5CO2). I fail to see how a new method can do anything to change basic thermodynamics and electrochemistry. Your linked paper is not about production of bulk aluminium, but rather a special high purity grade, in which the Aluminium seems to be produced from AlCl3 rather than Al2O3. $\endgroup$ – Level River St Sep 7 '15 at 1:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.