This concerns ongoing work to find a plausible approach to large-scale production of clear glass on the Moon. It's for an ultra-hard science fiction project and I really want to get it right.

Olivine of good purity can be gotten from deposits of dunite in the lunar highlands. You need to get to the bedrock, but it's there. If you melt the dunite you can use fractional crystallization to improve its purity - I have asked about that on Earth Sciences but it seems to be true. If that works it would also suggest you can do the same thing again to separate the olivine into forsterite ($\ce{Mg2SiO4}$) and fayalite ($\ce{Fe2SiO4}$).

If the forsterite is pure enough, could that be used to make decent clear glass?

Could the fayalite be processed into iron, water, and silicon dioxide with something like the reaction below?

$\ce{Fe2SiO4 + 2H2 -> 2H2O + 2Fe + SiO2}$


2 Answers 2


You could proceed from either end member of the olivine solid series and yield $\ce{SiO2}$ as you suggest.

However, I'd consider mechanisms that have been researched in the course of studying so-called mineral sequestration in addition to what you've written, especially considering the energy requirements you propose: the mineral sequestration reactions are (slightly) exothermic and would thus also yield heat as a by-product which I suspect would be desirable in a cold place like the Moon. That said, the reactions are slow at standard conditions and can be accelerated at the cost of supplying energy. How that balance works in your scenario is ultimately up to you, of course.

Additionally, carbon dioxide is a reactant (but needs to be in a supercritical state, which costs energy), which you'd get from human respiration and other organic sources "for free," which relieves you of having to have a lot of hydrogen gas on-hand.

The idea with mineral sequestration is to allow supercritical carbon dioxide to react with certain minerals to yield carbonates that are stable over some long time period:

Mineral carbonation reactions are known to geologists and occur spontaneously on geological time scales. For example, the reaction of $\ce{CO2}$ with common mineral silicates to form carbonates like magnesite or calcite is exothermic and thermodynamically favored.

An example is:

$$\ce{Mg2SiO4 + 2CO2 -> 2MgCO3 + SiO2}$$

(which) illustrates the transformation of forsterite, which is the end member of the common silicate mineral olivine. One ton of olivine can dispose of approximately two-thirds of a ton of $\ce{CO2}$. Again, the reaction is exothermic and releases 90 kJ/mole of $\ce{CO2}$.

In summary: There's nothing wrong with your chemistry or geology and there is more than one way to yield silica from mafic/ultramafic minerals.

The reference I have quoted above and cited below also details the process in general terms for minerals such as olivine and serpentine, and illustrates schematics for implementing such schemes on industrial scales, which might also be of interest to you.

Reference from the National Energy Technology Laboratory within the DOE:

Goldberg, P. ,Chen, Z. Y. ,O’Connor, W. ,Walters, R., and Ziock, H. $\ce{CO2}$ Mineral Sequestration Studies in US. Technology. 1 (1): 1–10 (2000).

  • $\begingroup$ After some thought, i'm not sure that CO2 is the better fit. The minerals have to be melted anyhow to purify the forsterite and fayalite, processing with hydrogen would yield iron as well (which is highly useful), and the hydrogen can be recycled. Storing hydrogen is a pain, but the process yields water, so the hydrogen could be split from that when needed. And it should be possible to do most of the heating with sunlight, which is constant and strong during the 2 weeks of daylight each month. Also worth considering - in a place with no carbon, you avoid taking it out of the system. $\endgroup$
    – kim holder
    Commented Apr 13, 2017 at 14:49
  • $\begingroup$ I was suggesting an alternative process in addition to answering your question (see the first sentence of my reply) in the affirmative. Thank you for expanding on the other consideration that need to be accounted for in your scenario. $\endgroup$ Commented Apr 13, 2017 at 16:04
  • 2
    $\begingroup$ Fair enough. It's just that i'd accepted the answer, but decided that goes too far and undid it. The additional information is valuable, but since at this point i think it would be better to stick with the hydrogen method, i don't want to discourage people from posting more answers. $\endgroup$
    – kim holder
    Commented Apr 13, 2017 at 16:34

The previous answer is not really helpful because you would need huge amounts of CO2 which are not available on the moon. You need 2 moles of CO2 to generate 1 moles of SiO2. You simply do not have that amounts of CO2.

Furthermore, assuming you do someone get that CO2, you need to physically separate the Mg and Fe carbonates from the silica. One way would be dissolving the carbonates in acid, which is also not available on the moon. Another way would be to melt the thing and separate the two immiscible carbonate and silicate liquids by centrifuging. This will require very high pressure to keep the carbonate in, otherwise, this thing will decarbonate and just make olivine again.

If the forsterite is pure enough, could that be used to make decent clear glass?

No. Forsterite has a melting point above 1800 °C. It also has the problems I mentioned in my previous answer to you:

Can glass be made with anorthite?

The problem of quench crystals forming in your liquid once you cool the forsterite is much worse than what I described for anorthite in the answer given in the link. You will have to find a way to cool it down even faster, and I doubt you can do it. It's also probably going to crack and break when cooling down.

Then there's the question of if you can even make it pure enough, but the answer for that will be in your other question here:


Could the fayalite be processed into iron, water, and silicon dioxide...

Yes. It still doesn't solve the problem of forsterite being a terrible glass material.


There's a recently published paper:

Schleppi, J., Gibbons, J., Groetsch, A. et al. J Mater Sci (2019) 54: 3726. https://doi.org/10.1007/s10853-018-3101-y Manufacture of glass and mirrors from lunar regolith simulant

It is open access. You might find it informative.

  • $\begingroup$ The process temperatures on the Moon would have to be considerably higher than on Earth, but that is much easier to do on the Moon thanks to the sun and the vacuum. If forsterite isn't useful as a glass, it may still be useful as a crystal, under the circumstances that could be practical. Or, fluxes can be added to lower its freezing point, as you mention. Your answer as it stands to the question on Earth Sciences doesn't clarify for me if that process would work, because it doesn't seem like it deals with a molten process. $\endgroup$
    – kim holder
    Commented Apr 17, 2017 at 14:53
  • $\begingroup$ I am marking this as the accepted answer because it addresses the matter much better than the other answer, but is not likely now to be voted up enough to rise above it. However, i wish i could discuss the issues i've raised in my comment above. $\endgroup$
    – kim holder
    Commented Apr 20, 2017 at 21:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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