# Which freezes faster hot water or cold water?

Dear Cecil,
Logic tells me cold water will reach the freezing point sooner than hot water! I use hot water for freezing ice-cubes as they come out of the trays more easily and do not fracture!
Jackie

I think this links to a Cecil Classic http://www.straightdope.com/classics/a2_098b

Hi Jackie and welcome

As I recall, this came up in college Physics. There were early experiment which seemed to verify that hot water froze quicker but have since been shown to be faulty. Older freezers would get thick layers of frost on the freezing coils. This would hold the cold water ice trays away from the freon tubing and insulate the tray from the tubings cooling effects. The hot water trays would melt through the frost and come in direct contact with the coolent tubing and thus be directly cooled and freeze first. Modern experiments or even one in which a clean older style freezer willl show that the warmer the water the more slowly it freezes.{

Cecil’s experiment was not useful because he boiled the water before freezing it. Boiling the water drives off dissolved gases, especially carbon dioxide.

Anyone interested in exploring this should use warm water that has not recently been boiled.

The first two lines of his answer is among the funniest things I’ve ever read.

Excuse me, “are” among the funniest, etc.

Or f…fr…frice

It seems 35 [sup]o[/sup] C water is the quickest to freeze.

All of the examples brought forward to show that hot water freezes faster include confounding factors: dissolved gases, containers that melt into refrigerator walls, loss of quantity caused by evaporation, expansion and evaporation brought on by throwing a sample in the air, etc. These are altering the orginal premise, which as I understand it, is: Given a specific quantity of pure, liquid H[sub]2[/sub]O at temperature T[sub]1[/sub] compared to an identical sample at temperature T[sub]2[/sub], and applying the same quantity of heat extraction to each, which one will become frozen first?

If you say that the hot one arrives at the end temp first because of dissolved gases, you have altered the experiment. If the hot container makes better contact with the fridge walls by melting frost, you have altered the experiment. If you allow the hot sample to evaporate part of its quantity, you have altered the experiment. If both samples are truly identical in every way except for the starting temperature, at some point in the experiment both must be at the same temperature. From then on, unless you postulate that water has memory, like Jacque Benveniste, they must perform identically. So, the first one to arrive at that temperature will be the first one to freeze. And which do you think will arrive at 35C first, a sample starting at 36C or 100C?

Cecil is right. He usually is.

Yes, of course. One has to wonder how such gems of applied knowledge ever gain currency among the great unwashed masses, however - I’m sure it’s been hashed over in the previous threads but the answer is simple.

Hot (“just boiled”) water really does freeze faster, if one is making ice cubes, say. Really tiny ice cubes, though, as evaporation is accelerated. In the interests of furthering scientific knowledge at the tender age of 8 or 10 I selflessly devoted freezer space and H20 to arrive at Truth. So as a practical matter, yes, hot water freezes faster. Sorta.

Hot water including dissolved gases may freeze faster than cold water without dissolved gasses. But that was not the OP. Nothing I said in my post needs to be changed.

Read the column again. He did the experiment with boiled water, and hot water fresh from the tap; then with cold tap water on the second trial. (I tentatively conclude that Cecil owns two ice cube trays. Yes, my friends call me ‘Sherlock’.)

Both the boiled and un-boiled hot water in the first trial entered the freezer at 125 F. And while he’s not entirely clear on the relative freezing speeds of the two trays in this trial (as he was more concerned with the difference between these two trays and the trays of cold water frozen in the second trial), it appears that the difference caused by the elimination of dissolved gases was negligible. So in fact, you appear to be twice incorrect-he didn’t make the mistake you suggested, but it seems that if he had, it wouldn’t have mattered.

Having an engineering background, I originally thought that since the water in the hotter tray had to pass through the same in which the cooler tray starter, that it was logically impossible that the hotter tray to freeze in a shorter time.

However, after some thought about the whole experiment, one should remember that a freezer is an active system. (I’m assuming ice cube trays and kitchen freezers). The thermostat on the kitchen freezer should be sensitive to the temperature of the air in the freezer. The hot water tray causes the freezer to run longer than the cool water tray. In running longer the temperatures of all the other items surrounding the tray get lower for the hot tray than for the cold tray, and thus shortening the freezing time.

If this same experiment were conducted outside on a cold winter’s night (in a passive environment), the cold tray should freeze first.

And you would be right. The rest of your post changes the conditions:

So if one tray got more heat extracted by the cooling system than the other, the outcome of the experiment is affected? Of course. But scientific experiments go out of their way to alter only one parameter at a time. If the outcome is different, that one parameter was the likely cause.

You are altering at least two. That’s not a valid experiment.

In order for hot sample X of quantity Q to be reduced in temperature to the value of sample Y, it must have heat extracted. Once it reaches the value of Y, it is identical to Y. But the extra heat extracted first must be accounted for. If you say that extracting more heat than the other sample makes it cool faster, you are showing an inverse relationship. If that were true, you could freeze something faster by heating it up first!

And if that works so well, why aren’t our refrigerators equipped with heating coils to ramp up the temp of the ice cube tray first before freezing it?

This may be true, but isn’t the point not comparing what freezes faster in tightly controlled laboratory conditions, but rather in the real world by real people who intentionally or not, alter one or both samples to reduce (or increase) freezing times?

A sample of heated-to-boiling water will have an effect on dissolved gases, will be itself affected by its own evaporation, will have an effect on the freezer’s thermostat, and in an old freezer might place the contents closer to the coils.

Because of all these effects, yes, the experiment has been altered; but back to the original point, will it freeze faster than plain unboiled, room temperature tap water?

Musicat my old mate, I must disagree. From whence does your original premise derive? I’ve read Cecil’s column, and it just seems to be concerned with the real world question of whether sloshing some hot or cold water in an ice cube tray and whacking it into the freezer will result in faster or slower freezing.

You seem rather to be concerned with a pure science question, which I value but which just doesn’t seem to be the OP at all.

What you describe as “confounding factors” I would describe as “reasons why the OP maybe correct”.

As Cecil shows, it isn’t, of course, but that doesn’t alter my point.

I’m not going to argue your point on confounding factors - you’re right - they are probably the culprits, but I agree with Princhester in that they are not necessarily confounding factors, but relevant ones in the context of the real-world observations.

However, In saying that at some point in the experiment both must be at the same temperature, I believe you are oversimplifying the experiment to a pair of homogeous samples at a uniform temperature (or at least comparable distributions of temperature) throughout.
This may or may not be the actual case - it might well be (for completely hypothetical example, you understand) that the cold water cools fairly uniformly, but the hot water maintains a broader range of temperatures, perhaps separating into layers - in this case, there might not ever be a point at which the contents of the two containers could ever be described as uniformly identical, until they both were frozen solid.