Anomalous behavior of water

Does the anomalous behavior of liquid water between 0 and 4 Celsius degrees (i. e. expands while cooled) have something to do with the other anomalous behavior of ice (i.e. floats in water)? Are those behaviors correlated ?

DaveC426913
Gold Member
Does the anomalous behavior of liquid water between 0 and 4 Celsius degrees (i. e. expands while cooled) have something to do with the other anomalous behavior of ice (i.e. floats in water)? Are those behaviors correlated ?
What do you think?

Does an expanding volume - such as, say, an inflatable life preserver - affect its buoyancy?

What do you think?

Does an expanding volume - such as, say, an inflatable life preserver - affect its buoyancy?
Of course ice will float in water because it is less dense. But for me, the fact that water at 4°C has a maximum density has nothing to do with the fact that ice is less dense than water, as I am comparing 2 different phases. I want to understand if there is any microscopic correlation between these 2 facts, i.e. if any substance that presents a liquid phase denser than the solid phase would also present a maximum density point (different from its melting point, off course)

DaveC426913
Gold Member
Of course ice will float in water because it is less dense. But for me, the fact that water at 4°C has a maximum density has nothing to do with the fact that ice is less dense than water, as I am comparing 2 different phases.
I don't understand why you think they have nothing to do with each other; they're the same thing, it seems to me.

(Of course, the reason ice is less dense is because its solid crystalline matrix occupies more space per unit molecule than the formlessness of its liquid, but you know that.)

I want to understand if there is any microscopic correlation between these 2 facts, i.e. if any substance that presents a liquid phase denser than the solid phase would also present a maximum density point (different from its melting point, off course)
How can this not be?
If a substance has a liquid phase denser than its solid phase, then it must have a maximum density that is liquid (unless you posit a substance that continue to get less dense the colder it gets.)

But my comments are rhetorical; I assume I am missing the crux of your question - and thus merely muddying the waters - so I'll let someone else step in.

Last edited:
hutchphd
(unless you posit a substance that continue to get less dense the colder it gets.)
And even more unlikely it would need to get continually more dense with increasing T to deny a minimum

DaveC426913
If a substance has a liquid phase denser than its solid phase, then it must have a maximum density that is liquid (unless you posit a substance that continue to get less dense the colder it gets.)
I think I am probably the one who is not getting it right. I will try to illustrate, this is the density diagram for water:

My question is, why couldn't it be this way (ice still floats in water, but liquid water does not have a local maximum anymore)?

Or this way (ice is now denser than water, but water still have a maximum local density):

Staff Emeritus
2019 Award
I liked that message, because it's now clear enough to start to answer.

Diagram 2 asks the question "why is water's coefficient of thermal expansion negative?" because both liquid and solid have positive coefficient of thermal expansions in that sketch Do you want an answer for that?

Diagram 3 has the density go up once the liquid is cooled past its phase transition - i.e. the molecules get closer together. Do you want an answer for that?

Does the anomalous behavior of liquid water between 0 and 4 Celsius degrees (i. e. expands while cooled) have something to do with the other anomalous behavior of ice (i.e. floats in water)? Are those behaviors correlated ?
It is actually something different. The anomaly of liquid water results from a transition from an irregular "structure" to a local tetrahedral structure with higher specific volume. The further reduction of density during cristallisation results from a transition from the local tetrahedral structure to a hexagonal structure with an even higher specific volume.

Last edited:
There are multiple elements that also expand on freezing: Ga, Si, Ge, Sb, Bi, Pu.
Do any of these display a density maximum within a single phase, as water does?