When doing an experiment on growing plants in sealed plastic bottles, I learned that many plastics, such as PET used for soft drinks bottles, are more permeable to CO2 than to N2 or O2 (see table 2 of this paper for relative permeability numbers: CO2 leaks through PET about 5 times as fast as O2 and about 18 times as fast as N2).

It occurred to me that a material that allows CO2 to pass through more readily than O2 could be used to extract CO2 from the Martian atmosphere into a greenhouse without using a pump. CO2 makes up about 96% of the atmosphere of Mars, at a partial pressure of about 6 mbar. Plants grow on Earth with a CO2 partial pressure of 0.4 mbar, but can tolerate higher partial pressures of CO2 than this. A greenhouse with walls made from a semipermeable barrier like PET (but capable of withstanding the higher radiation levels found on Mars - see note below) would draw in CO2 from the atmosphere as plants use it up during photosynthesis. The pressure in the greenhouse would increase because the O2 that the plants produce would leak out more slowly than the CO2 leaked in. The maximum pressure that it would reach would depend on the ratio of the rate at which O2 and CO2 permeate the barrier material. The O2 produced by the greenhouse could be used to help sustain a colony.

It seems counterintuitive that gas molecules can travel against the total pressure gradient, but it’s the partial pressure that counts: if the partial pressure of CO2 outside the greenhouse is higher than inside, CO2 molecules should leak into the greenhouse.

Note that the transparent part of the greenhouse could be made from a material resistant to radiation damage, and the semipermeable barrier could be in a soil-covered tunnel to protect it from radiation present on the surface.

Would this work? I’m hoping for answers like “no it wouldn’t work and here’s why...” or “yes it might work and here is a link to some work investigating this...”.

  • $\begingroup$ however, if I understand your premise, it would leak nitrogen. $\endgroup$
    – user20636
    Apr 28, 2019 at 10:22
  • $\begingroup$ If your problem is that you keep running out of CO2, then you don't necessarily need a pump -- you could just use a shovel. Nothing like some dry ice frost on a brisk Martian morning. $\endgroup$
    – Roger
    May 7, 2019 at 14:24

1 Answer 1


Osmosis against a total pressure gradient is absolutely a 'thing'. Plants and animals exploit this and active transport to move fluids around and they can definitely create pressure gradients.


  • We can already separate and compress gasses already.

  • The thickness of the sheet of polymer needed to make this work would not survive any meaningful pressure difference.

  • If you balanced the pressure you'd still need an enormous surface area of the stuff to move a significant quantity of gas.

Hence I don't see it being practical. That said engineering materials that have good rates of diffusion is also a thing and there may well be a way of adding a permeable support structure to this to allow some pressure gradient.

I think you would run into problems with keeping the water-vapour in though. Water is a really small molecule and will diffuse through almost anything that allow CO2 through. I am not aware of a way around this problem or anyone working on it, but never say never...

  • 1
    $\begingroup$ Water vapor is polar where carbon dioxide is non-polar. A sufficiently clever material could make use of this to separate the two. $\endgroup$
    – Mark
    May 4, 2019 at 0:55
  • $\begingroup$ @Mark Like I say: 'never say never'. With a bit more digging, If you allow non-polymer membranes, there are already such things. $\endgroup$
    – ANone
    May 7, 2019 at 10:02

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.