If I have a container that is full of water and I attempt to freeze the water by freezing the container, what would happen if the container is strong enough to prevent the water expansion? Could the water be in liquid state even though it is at or below freezing point?

This is probably a daft question but its something that's been bugging me.


In order to keep the density of water at 1kg/litre as you decrease its temperature below 0 centigrade you would need a container that remained rigid at many thousands of atmospheres pressure. The water would then remain liquid until something like -20 degrees centigrade.

At still lower temperatures it would become solid, but it would not form crystalline ice which always has a density lower than 1kg/litre. Instead it would form a glassy solid known as amorphous ice which has a higher density under such pressures.

There are several high density phases of amorphous ice and to get the exact details of which phase it would pass through you need a phase diagram showing density as a function of temperature and pressure extending to very high pressures and low temperatures. You can then follow the contour corresponding to a density of 1kg/litre to see how the pressure and phase would vary as the temperature decreased. I can only find enough information to give the vague answer above and it is possible that the diagram is not known well enough yet to say much more.

  • $\begingroup$ Here is a list with 14 (!) different modification of water ice. None of them is amorphous. google.de/imgres?imgurl=http://www.helmholtz-berlin.de/media/… $\endgroup$ – Georg Nov 8 '11 at 13:50
  • 1
    $\begingroup$ en.wikipedia.org/wiki/Amorphous_ice $\endgroup$ – Philip Gibbs - inactive Nov 8 '11 at 20:26
  • 1
    $\begingroup$ Your list is incomplete according to the wikipedia list en.wikipedia.org/wiki/Ice which has a number 15, subdivides type I and adds three amorphous phases. Unless one of the crystalline phases can exist at exactly 1kg/l density for some temperature and pressure they are not possible answers to this question and only the amorphous phases could work. Your table suggests that the densities of the crystalline phases are all different from 1 kg/l but this could be wrong. I am not sure. Anyone know? $\endgroup$ – Philip Gibbs - inactive Nov 8 '11 at 21:09
  • 3
    $\begingroup$ @Georg, the wikipedia article has plenty of references to papers in peer-reviewed journals that study amorphous ice. Here is another page that gives a more complete table along with the range of densities for each phase lsbu.ac.uk/water/ice.html As I indicated in my original answer there is still much unknown or uncertain in this subject but I don't think the existence of amorphous phases is so controversial $\endgroup$ – Philip Gibbs - inactive Nov 8 '11 at 21:52
  • 1
    $\begingroup$ You can have a mix of crystalline structures, chosen to make the density correct. Based on the graph provided, a mix of $I_h$ and $II$ or $III$ seems most likely $\endgroup$ – Ross Millikan Dec 18 '13 at 18:04

Yes, of course, the freezing point will decrease by the pressure developed, while part of the water freezes.

But do not underestimate the pressures! In such an experiment easily some thousand bares may be developed. (Depends on the rigidity of the vessel and the volume of water)

Here is a video showing how freezing water cracks a cast iron sphere.

(The bath is ice/salt at about -10°C)

  • $\begingroup$ Hard to tell from the final photo, but did the ice inside the grenade form a hollow sphere? $\endgroup$ – Mark Betnel May 10 '11 at 13:59
  • 1
    $\begingroup$ @Mark I see (and think) so. Heat will stream outwards of the sphere and the solidifying will start at the wall. After some ice layer has built, the resistance to heat flow will be almost solely in the ice (no convection there) so the process of heat transfer will be self-regulating for a rather uniform ice layer. $\endgroup$ – Georg May 10 '11 at 14:15
  • 5
    $\begingroup$ This does not answer the question at all. The question is about what happens if the container is strong enough to hold the pressure. Pointing out that a cast iron sphere can crack is completely irrelevant. There are plenty of things stronger than cast iron. It is a very good question that deserves a better attempt at a direct answer. $\endgroup$ – Philip Gibbs - inactive Nov 8 '11 at 21:59
  • $\begingroup$ Dear Philip, You should accept the reasoning of "Net": A video beats an image, an image beats any text (irrespective of truth) :=( $\endgroup$ – Georg Nov 9 '11 at 9:55
  • $\begingroup$ I agree with the comment of Philip Gibbs here. I'm glad this answer isn't the accepted one. $\endgroup$ – thermomagnetic condensed boson Feb 11 '18 at 13:42

I think what you are really getting at is the nature of phase changes in respect of changes in pressure, volume and temperature. This 3D diagram explains a lot though it can a while to interpret it.

And the associated article should explain in more detail. Here is also a simpler 2D version.

  • $\begingroup$ The 3D diagram is for a generic material. Water is exceptional in that specific volume increases when it freezes so the diagram shown here is not appropriate. $\endgroup$ – Philip Gibbs - inactive Nov 9 '11 at 8:45

I think that increase in pressure highers the freezing point of water from 0 degrees -ve thes the water will still freeze but at a lower temperature

  • 2
    $\begingroup$ Pressure lowers the freezing point and melts ice. This is why ice skates work. $\endgroup$ – Philip Gibbs - inactive Nov 9 '11 at 8:52

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.