I have heard of the following procedure intended to make water cleaner:

  1. Put a pot of water in the freezer.
  2. After a few hours the water will be partly frozen. Discard the ice. Keep the liquid in the freezer.
  3. After a few more hours the water will be, again, partly frozen. Discard the liquid. Then melt the remaining ice and use it for drinking.

I'm not sure what cleaner means, considering that tap water is perfectly good for drinking where I live. Although it seems reasonable that different solubles would have different concentrations in ice and in remaining liquid, possibly depending on mass or solubility. On the other hand I see no reason why this would discard more harmful substances than useful ones.

Can concentrations of solubles be significantly altered using such method? Can these changes have an effect on the quality of water?

  • Where did you hear about this, I can't seem to find this claim on internet when searching for it. Commented Jun 21, 2014 at 19:43
  • @DavidMulder, I was told that personally. Now I've found a similar procedure here. Commented Jun 21, 2014 at 19:53
  • I'm not sure if it is true in general, but freezing water will get rid of those substances which freeze at lower temperatures than water. I.e., if you freeze Coca-Cola (yes, I tried it), you'll get some ice and black extra-sweet substance, so, clearly, it will rid the water from some impurities at least in some cases.
    – sashkello
    Commented Jun 22, 2014 at 13:57
  • @sashkello, that seems reasonable. My doubts are, firstly, if the same applies to water (where concentrations of solubles are tiny) and secondly, if this discriminates harmful substances (I'm under the impression that significant portion of solubles in tap water are useful). Commented Jun 22, 2014 at 14:07
  • could be asked at chemistry.SE to know how it would work. Commented Jun 23, 2014 at 12:09

1 Answer 1


Wikipedia explains that this technique of fractional freezing can be used for desalination:

Fractional freezing can be used to desalinate sea water. In a process that naturally occurs with sea ice, frozen salt water, when partially melted, leaves behind ice that is of a much lower salt content. Because sodium chloride lowers the melting point of water, the salt in sea water tends to be forced out of pure water while freezing. Likewise, the frozen water with the highest concentration of salt melts first. Either method decreases the salinity of the frozen water left over, and with multiple runs can be drinkable.

I would not rely on it to remove arbitrary toxic chemicals from water. In survival situation (with access to freezing and energy to melt things and prevent yourself from freezing) it could be worth doing to purify water. I cannot compare it to alternatives.

  • 7
    But then, the OP says to throw away the ice and keep the water. That would increase the amount, not lower it...
    – Duralumin
    Commented Jun 23, 2014 at 13:33
  • @Duralumin, the steps I listed would remove both substances that accelerate freezing (if such exist?) as well as those that prevent it. If the goal is to remove salt, then it's sensible to skip step 2. Commented Jun 24, 2014 at 15:39
  • @Karolis I'm not aware of any solute that would raise the freezing temperature of water. Still, is my fault as i didn't read correctly point 3.
    – Duralumin
    Commented Jun 25, 2014 at 7:20
  • @KarolisJuodelė: If there are any particles in suspension when the water freezes, ice crystals may form around them. Although such particles would be surrounded by pure ice, melting the ice would put the particles back into the resulting water.
    – supercat
    Commented Sep 26, 2017 at 16:59

You must log in to answer this question.

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