Hot water makes ice cubes freeze quicker?

Someone please confirm that this is false so I can get my friend to believe it!

My understanding is that the hot water evaporates quickly in the dry air of the freezer. That leaves a smaller amount of water that freezes more quickly. You can get the same results even faster results if you start with less cold water.


But according to Cecil (who cites a Scientific American article), this is really only an issue at near-boiling temperatures.

For regular hot (tap) water and cold (tap) water, the cold will freeze considerably more quickly.

Sorry for the duplicate post. It’s an old wives tale.

First of all, it’s not true that hot water freezes faster than cold water – as others have pointed out. Ignoring miniscule evaporation changes, at some point the hot water will become the same temperature as the cold water started at, and this won’t take zero time.

But I have to ask: why not just do the experiment? It’s not like this is complicated science that requires a multimillion-dollar lab or require a statistician to interpret the results. You could have your answer in an hour; three or four if you want to completely eliminate concerns about conditions like position of the water in the freezer.

There IS a reason some people make ice cubes from hot water. It has nothing to do with them freezing faster, however.

Hot water holds less dissolved oxygen. Ice cubes made from hot water are clear, lacking the whiteness from internal bubbles.

Restaurants and caterers will sometimes use warm or hot water to make “prettier” ice cubes. Perhaps this contributes to the persistence of the belief?


Here’s where you lose me. At every faucet that I have noticed, the hot water seems to have more “air” in it than cold. The hot water has always been cloudy from many more small air bubbles than the cold water.

Water that has been boiled and cooled to room or tap water temperature will freeze faster!

Water that has been boiled and cooled to tap water temperature **will freeze faster/sooner.

Uhh… care to explain this one? Water that has been boiled and cooled to room temperature, except for possibly having less dissolved gasses in it, is AFAIK exactly the same as plain ol’ water fresh from the tap.

My understanding is that it’s not that the hw has less dissolved gasses–hot water is going to allow more gasses to disolve, not less–but that it will tend to freeze more slowly, thus giving the bubbles a chance to escape, leading to a clearer, ‘pretty’ ice cube.

It’s probably just a peculiarity of the plumbing, that water from the heater doesn’t completely fill the pipes or something. The solubility of gases in liquid decreases with increasing temperature so hot water has less dissolved air in it than cold. You can see this for yourself by boiling a pot of water and watching air bubbles materialise at the bottom from what seems like nowhere.

It’s the “less dissolved gases” that lets it boil faster, IIRC.

Actually, hot water has a lower capacity for dissolved gases than cold water. The oceans have a lot more life in the very cold places like the arctic than they do in the equatorial regions, because the water has lots more oxygen dissolved in it.

If you’ve ever compared a carbonated drink – a can of Pepsi, say – at room temperature to one that’s just above freezing, you’ll see this in action. The ice-cold Pepsi feels almost like tangy fizzy water, while the warm Pepsi is a huge mess of foam and bubbles immediately, to the point that it’s harder to drink since it’s like drinking soap suds. Not surprisingly, most people prefer their carbonated beverages cold.

This is my fav explanation:

Salts, which lower freezing temperature, will precipitae out if water is bolied. Their granularity will assist crystal formation and if FlamingRamenMonster looks real-l close s/he will descry the “air bubbles” masquerading as steam.

Not such a hot idea to use hot tap water, though. Hot water tends to contain higher concentrations of potentially dangerous minerals and elements that gather as sedimentation in your water heater. At least, that’s what a plumber friend told me when I commented about loving to drink shower water. There, I said it. I drink shower water.

Is the OP asking about pure H[sub]2[/sub]O, as in a laboratory condition, or ice cubes in a consumer fridge that may have other matter dissolved in it and be subject to convection and other confounding factors?

In other words, are we arguing physics or semantics?

FYI: A previous thread exists.

I am assuming we are talking about pure water, in identical amounts, with no confounding factors like convection, movement, light, container differences, etc. – if not, all bets are off.

If by “faster”, you mean the rate of change (slope of a curver or line) is steeper, so that with any two identical samples, the one starting at the higher temp will cool faster (the slope is steeper).

But if by “faster”, you mean how long it takes for two samples to lose enough heat to reach temperature X, then there is no way that a hotter starting sample will cool faster than a cooler starting sample.

Consider this: If two identical samples differing only in their initial temperature are cooled by the same amount of heat extraction for each and are started at the same time, at some point in time the originally hotter sample must exactly match the current temp of the originally cooler sample. At that point, unless you postulate that the substance has some kind of memory about past events, they become identical in every way. If the originally hotter one subsequently reaches freezing sooner than the other, something paranormal is going on here.

I agree that this is critical. The “hot water freezes faster than cold” as generally used by the public is, to my mind, generally a plumbers myth as they propound this “fact” for anything from hot water pipes freezing faster than cold water pipes to ice cubes in a refrigerator.

I have always said the same thing, however, alterego’s cite did point out that there could be initial starting conditions such that when the average temperature of the hot water had cooled down to the temperature of the starting cool water the originally hot water could have a steep temperature gradient such that some of the water is very cold and some still very hot. In that case, we would not expect it to behave the same as the originally cold water.

So it’s back to semantics. As a general statement, cold water freezes before hot water. But like all things in life, there could be exceptional cases where the hot water could freeze faster. (Hot water ice cube trays placed on hard packed snow such that the hot tray melts into the snow and increases heat transfer while the cold sits on top, open containers that have water evaporate off the top and lower the mass of the water while also losing extra heat due to evaporation, funky shaped containers which only have heat loss surfaces at specific areas such that they set up convection currents at certain temperatures, hot water pipes that are used less frequently than cold water pipes, etc.)

Say what?

Why didn’t I do the experiment? I have common sense. I never believed this wives tale

Please propose a test where, given two samples of pure H[sub]2[/sub]O, at exactly the same temperature, you can tell which one was hotter a few minutes ago from merely testing the samples.

If “some of the water is very cold and some still very hot” then you don’t have a uniform sample and you have altered the question to “does hot water cool faster than cold water if non-uniform cooling or convection is allowed?”