We know that ice has a lower density than water despite both having the same [molecular] mass. I know that as water turns to ice, it expands. As far as I was taught, I know that it has something to do with hydrogen bonding. I would like to know in terms of hydrogen bonding, why ice actually expands?

  • 2
    $\begingroup$ Note that water is far from being the only compound that has a liquid with higher density than the solid at the melting point. Elements such as Si, Ge, Sb, and Bi do as well, and they certainly have nothing to do with hydrogen bonding. The only real certainty is that a first order phase transition must have a volume change. $\endgroup$
    – Jon Custer
    Sep 29, 2016 at 12:54
  • $\begingroup$ I'll point out that chemists use "water" in two different ways. First it just means the chemical with the formula $\ce{H2O}$. Second it can mean more specifically the liquid state of $\ce{H2O}$. Also planetary scientists refer to a number of compounds as various kinds of "ice" that would coat the surface of a planet. $\endgroup$
    – MaxW
    Sep 30, 2016 at 21:58

2 Answers 2


The most straightforward answer is that water in the form of ice I (or $\ce{I_h}$) has an open hexagonal structure (see figure) whereas liquid water does not. In ice the oxygens are precisely tetrahedrally positioned where each oxygen is hydrogen bonded by four neighbouring oxygens with an O..O distance of approx 0.28 nm. The H atoms lie very closely along the O-O axes. Two of the H bonds are short, as in liquid water, and two are long. This is expected from residual entropy calculations and also from thermal and neutron diffraction experiments.

As this ice structure has a rather open lattice this is the reason its density is less at its freezing point than that of liquid water, rather than being due to the high density of water. One could argue that the reason that water density is a maximum at 4 $\ce{^0C}$ (277 K) is that the structure of liquid water is changing from that of ice $\ce{I_h}$ to the short range structure of the liquid.

Liquid water is, of course, also highly hydrogen bonded but the extra energy it contains means that its structure is far less regular and so becomes slightly denser than ice. As the water molecules can move about more in the liquid than in ice, the hydrogen bonds will bend a little or break as molecules rotate. This will lead to a shorter O - O distance and hence slightly higher density in liquid than in the solid.

see also answers to this post Why really is ice less dense than water?


This figure shows the hexagonal ice structure; The O atoms are shown as dots and the lines are the (almost linear) O-H-O bonds.(Image from Murrell & Jenkins 'Properties of liquids & solutions')

(Ice also exhibits eight different phases depending on temperature and pressure of which $\ce{I_h}$ is the one found at normal atmospheric pressure. In these other phases, angles are distorted from being tetrahedral)

  • $\begingroup$ Could you explain the structure of water a bit more? How irregular is it? $\endgroup$
    – Ahnaf
    Sep 29, 2016 at 16:18

Straight from Wikipedia...

Density vs Temp

The density of water is about 1 gram per cubic centimetre (62 lb/cu ft): this relationship was originally used to define the gram. The density varies with temperature, but not linearly: as the temperature increases, the density rises to a peak at 3.98 °C (39.16 °F) and then decreases. This unusual negative thermal expansion below 4 °C (39 °F) is also observed in molten silica. Regular, hexagonal ice is also less dense than liquid water—upon freezing, the density of water decreases by about 9%.

These effects are due to the reduction of thermal motion with cooling, which allows water molecules to form more hydrogen bonds that prevent the molecules from coming close to each other. While below 4 °C the breakage of hydrogen bonds due to heating allows water molecules to pack closer despite the increase in the thermal motion (which tends to expand a liquid), above 4 °C water expands as the temperature increases. Water near the boiling point is about 4% less dense than water at 4 °C (39 °F).

Other substances that expand on freezing are acetic acid, silicon, gallium, germanium, antimony, bismuth, plutonium and also chemical compounds that form spacious crystal lattices with tetrahedral coordination.

  • $\begingroup$ First, this is basically a plagiarising answer as it builds entirely off its source. Second, this does not answer the question why water is more dense than ice. $\endgroup$
    – Jan
    Sep 30, 2016 at 23:09
  • $\begingroup$ @Jan - the why - "These effects are due to the reduction of thermal motion with cooling, which allows water molecules to form more hydrogen bonds that prevent the molecules from coming close to each other." // I said explicitly that I lifted the answer from Wikipedia. The whole point was that I thought the OP should do some simple research. $\endgroup$
    – MaxW
    Oct 1, 2016 at 0:35
  • 2
    $\begingroup$ @MaxW It's ultimately preferable to use the external text/graphics to illuminate your own answer. I am confident that you have more you can add to the answer than just the raw explanation. $\endgroup$
    – jonsca
    Oct 1, 2016 at 2:29

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.