Ice, Hydrogen bond, and Ice rules Wikipedia articles seem to imply that the covalently bonded hydrogen atoms are fixed to their given oxygen atom and do not change positions as long as the ice remains solid. This would be in contrast to liquid water that self ionizes.

enter image description here

Is this actually true in reality? For example if you were to put a pure H2O ice crystal into a cup of water containing an increased concentration of D2O (or HDO), would the deuterium atoms diffuse into the ice crystal structure after some period of time? (presuming the temperatures were perfect such that the ice crystal did not grow or shrink) Or would the makeup of the ice crystal remain unchanged if you removed it and tested it say a month later?

  • 5
    $\begingroup$ Self-ionization actually still occurs in solid water. In the "standard" phase, $\mathrm{I_h}$, the autodissociation constant is $k_w = 2 \times 10^{-20}$ at -4 °C (ref). I'm guessing a Grotthuss-type mechanism to shift the hydrogens around still takes place. $\endgroup$ Jan 28 at 10:26
  • $\begingroup$ chemistry.stackexchange.com/questions/50704/… $\endgroup$
    – Mithoron
    Jan 28 at 14:17
  • $\begingroup$ chemistry.stackexchange.com/questions/64719/… $\endgroup$
    – Mithoron
    Jan 28 at 14:21
  • 2
    $\begingroup$ This example isn't good - crystal would melt and reform in a dynamic equilibrium. Over time you'd get pretty much a new one, except it wouldn't be via diffusion, well, not only diffusion. $\endgroup$
    – Mithoron
    Jan 28 at 16:23
  • $\begingroup$ @Mithoron Your comment made me consider the inverse experiment: a cube of solid $\ce{D2O}$ in liquid $\ce{H2O}$ between 0 °C and 3.8 °C (under ambient pressure). Naively, one might imagine that if H-D exchange throughout the crystal did not happen, the $\ce{D2O}$ would never melt, and if it melted, it would be evidence of such exchange. However, of course this completely disregards the surface of the $\ce{D2O}$, treating it only as a bulk solid - in all likelihood it would melt regardless. But perhaps very careful study of the melting rate could tease out the effects of exchange in the bulk. $\endgroup$ Jan 29 at 12:48

1 Answer 1


Just as electrons exist in a "cloud", whole atoms and molecules do as well. The classic double-slit interference experiment has been performed with deuterium atoms and even molecules massing more than 25 kDa, such as tetraphenylporphyrins with fluoroalkylsulfanyl chains.

Then given that position and velocity of a particular hydrogen cannot be determined precisely, they could be interchanging continually -- though calculating the probability of such interchange is decidedly (no pun intended) difficult.

As a similar phenomenon, I find amazing the simple, inexpensive, Esaki or tunnel diode, in which electrons cross a barrier (well, disappear on on side and appear on the other) by quantum tunneling. Is that not happening to the hydrogens in different positions in ice? See this question on proton tunneling.


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.