Our colony has landed on a planet which is a complete barren desert wasteland, the surface as dead as the moon but also extremely cold, being quite far from the local star. We have settled at the only source of heat - a very deep crater with a hot geothermal spring at the bottom.

There is water in this spring, alas, the water is mixed with other toxic chemicals and so can't be drunk.

However, the heat from the spring causes a lot of the liquid to evaporate. As the vapour rises, (let's just hand-wave that they can breathe for now), it cools as it gets further and further from the heat at the bottom. Eventually, it condenses, and some of it lands on the sides of the crater, and as it comes in contact with the freezing surface, some of it freezes. This has been going on for thousands of years - or however long it needs to have

Now each component of the mixture has a different condensation/freezing point, and the temperature changes with altitude.

So this has the fortunate effect of separating out the components, and at a certain altitude, the walls will be covered in (relatively) pure ice water which can be chipped off, melted and drunk safely. The other chemicals will have condensed and frozen at either higher or lower levels.

Is this at all plausible?

  • $\begingroup$ you seems to forget water is an Universal solvent, so in this case distillation is your best bet. $\endgroup$
    – user6760
    Nov 3, 2017 at 2:54
  • 7
    $\begingroup$ Freeze distillation works, though on Earth the aim is usually to discard the water and keep the "impurities". See e.g. homebrew.stackexchange.com/questions/11019/… $\endgroup$
    – jamesqf
    Nov 3, 2017 at 3:54
  • 2
    $\begingroup$ Just going to leave this here ahem whistles innocently $\endgroup$ Nov 3, 2017 at 16:02
  • $\begingroup$ Will absolutely work, but likely not god enough in a single passthrough. Each separation event will remove a high percentage but not all impurities. $\endgroup$
    – user79911
    Nov 4, 2020 at 11:35

3 Answers 3


If you assume that there is a temperature gradient around the crater walls then yes this should actually happen, the exact formation being dependant on the compounds present, the temperature gradient and the amount of materials being discharged. Obviously salts would be left behind in the boiling liquid and many gases would not condense at all. In fact I suspect water would be the number one condensate.

But other materials present might well contaminate the ice. Gases might get trapped between the condensing water crystals in a vaguely similar way to air bubbles being trapped in compressed snow. Wind changes and temperature variation would also probably act to mix things up a bit. But depending on the exact nature of the impurities it should be relatively easy to purify the water by fractional distillation

Especially as they have plenty of hot water and plenty of cooling available. Very difficult to give a definitive answer as it depends on the impurities and their concentrations.

One potential problem is that some mixtures freeze as mixtures - If you ever freeze a can of beer you will see relatively clear ice followed by progressively darker ices and finally a very dark brown liquid at the bottom that contains most of the alcohol and a lot of the impurities.

  • $\begingroup$ Just adding to the answer instead of posting a new one: different liquids evaporate at different rates. Normally the liquid in excess quantity evaporates at a greater rate than the rest. If your spring is fed from fresh water you can easily extract water through fractional distillation as you and the OP describe. The mixture would remain inside the liquid part not the evaporated part, which would be mostly pure individual components, assuming they freeze at different temperatures. $\endgroup$
    – nzaman
    Nov 3, 2017 at 6:48
  • 3
    $\begingroup$ Yet another problem (possibly hinted by "it depends on the impurities"): the water and chemicals may form an Azeotrope $\endgroup$
    – abukaj
    Nov 3, 2017 at 10:04

What you suggest for water is a commonly used technique to purify silicon ingots for electronic applications. A molten zone is repeatedly swept in the same direction along the ingot, concentrating the impurities in the liquid phase and then in the extremity of the ingot, which is then discarded.

It works as long as the impurities are more soluble in the liquid than in the solid phase.

  • $\begingroup$ As a side note, this technique has not been used for quite some time now. It did not scale up well and was completely gone by 200mm. All silicon wafers for integrated circuits are now Czochralski-grown. These wafers have the added benefit of enough carbon "contamination" to make them strong enough to make it through various fab processes. $\endgroup$
    – Jon Custer
    Nov 3, 2017 at 14:07

Short answer: Yes, as explained much better than I could in Slarty's answer.

To give a historical example: During the Viking age, people in Britain would make weak beer stronger (as, with the methods and resources available to them at the time it was difficult to brew something stronger than a modern lager) by leaving it in open barrels overnight in winter, then removing the layer of ice (clean water) that formed on top, leaving behind a stronger mix with a higher percentage of alcohol. This process was often repeated several times for a stronger brew.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .