# Q & A: ice cubes sticking

Q:
When I put two ice cubes in a glass and pour in some liquid (water, juice), I see that the ice cubes stick together. This condition does not last too long. Here is my attempt at explaining: Since melting is an endothermic reaction, at the interface between the cubes, enough heat is taken out to momentarily re-freeze the cubes? Or are these some non-Newtonian ice cubes I got hold of?
- todd (age old)
A:
I can make a guess, along the lines you're considering. The ice cubes start to melt all around, either because the liquid is a bit warmer than 0°C or because it has some solutes in it, or both. As you say, that cools things down.

Overall that process is going to stop when the surface of the cubes cools to the melting temperature. If there are solutes in the liquid, it has purer water since melted ice is coming in from both sides. Heat can diffuse around, equalizing the temperature, faster than solutes diffuse. So if most of the ice cube surface is at its melting point, the region between the cubes is below its melting point because it has a lower solute concentration. Thus it will freeze, connecting the cubes.

There's a different reason why it could happen in pure water. If you have a crystal with an uneven surface, it can lower its free energy by smoothing out the bumps and cracks. That's just a way of reducing the liquid-solid interface area, since the interface has higher free energy than either of the phases.  In other words, the freezing temperature in a crack is actually somewhat higher than on a typical surface. Likewise if there's a little point sticking out somewhere, its freezing temperature is somewhat lower. Now if the two cubes happen to touch, the little wedge of liquid leading up to that contact can lower its surface area by filling in the crack with ice as ice turns to liquid elsewhere.

Mike W.

(published on 04/06/2013)

## Follow-Up #1: freezing rough surfaces

Q:
Mike W, thanks. My second question was going to be the freezing cubes in pure water. To recap the pure water instance: crystal with an uneven surface has a higher free energy, lower melting/freezing point--just as if an antifreeze protein molecule were adsobed to the surface, creating a higher surface area, higher free energy condition--hence, lower metling/freezng point. When two of these surface come together, the interface also has a higher surface free energy. You then say that ". . . the freezing temperature in a crack is actually somewhat higher than on a typical surface." Would it not be lower? I understand why it is lower "if there's a little point sticking out somewhere. . . " In either case, as the interfaces come together (which has a lower melting/freezing point), they freeze momentarily, reducing the surface area, interfacial free energy?
- todd (age old)
AK
A:
In the crack, freezing reduces the interface area. For the protrusion, it increases the interface area. That's why the geometry helps the crack freeze and the protrusion melt. You clearly got that idea but maybe just got a little twist at some point in thinking about the crack.

Mike W.

(published on 04/08/2013)