We've Finally Figured Out Why Hot Water Freezes Faster Than Cold

I thought it was becuase hot water hs more convection and moving currents that continue moving as it cools because of inertia which transfers more heat more evenly across the container.

But I am not a chemist : )

Drew K

That paper contains some egregious assumptions.

I just want to know when hell is going to freeze over.....I stand to reap some heretofore unattainable benefits...

Faster than heaven, anyway!

I call BS on this one.

Too much left to chance here. Show the actual conditions under which this miracle occurs.

Brownridge heated the tap water to about 100° C, while the distilled water was cooled to 25° C or lower.

As I said; BS

Should the Mpemba effect be valid,

I vote for dissolved gasses.Hot (hotter) water having more if not all of any possible dissolved gasses driven out of solution,the "cleaner" water having better thermal transfer properties.

Otherwise,I wonder if a piezo experiment could validate the "contracting covalent bond" suggestion?

Ta

I'm with you with the dissolved gases. If both water samples are identical, I don't see how one can overtake the other on the way to the freezing point.

The samples are NOT identical.

Brownridge heated the tap water to about 100° C, while the distilled water was cooled to 25° C or lower.

The samples are NOT identical.

Agreed.

I remember doing this with tap water in 8th grade science...two samples drawn from tap into beakers, one boiled and one at room temp., placed in freezer, and checked periodically....the boiled one freezes first. I've been making my ice cubes for drinks this way ever since...clear cubes...no dissolved gas. Was curious what 'gizmodo' reported, it never loads for me. What did they work out? Who paid for this observation? Did it really require physicists?

The experiment you have described is much like the experiment many people use to validate the Mpembo effect. The problem is, as it is described, the experiment cannot distinguish dynamic system behavior from a counter-intuitive property of water.

.

If you placed both samples in the same freezer, the boiled water sample is almost certain to heat the air in the freezer significantly during the initial period. The cooler sample will cool very little if the ambient air is only slightly below the temp it is at.

.

The boiled container is also likely to melt through any frost and make contact with what may be a more heat conductive material, like a metal rack, increasing its effective surface area.

.

The boiled water is also likely to deposit a layer of frost on the container of the cooler sample, insulating the non-boiled water.

.

Convection currents in the hot water have some momentum and probably benefit the rate of heat transfer even when the hot sample has fallen below the initial temperature of the cold sample.

.

.

If the separate freezers, or the same freezer at different times is used, then the hot sample initiates refrigeration much quicker than the cold sample, so heat is actively removed faster from the freezer.

.

It would not be hard to devise an experiment that avoids introducing the errors above, but I have never seen any mention of such being done. The goal of the experiment would need to be expressed in scientific terms, likely relating to the specific heat capacity of water, instead of 'water freezing faster'.

I don't remember if the samples were equal in temperature when they went into the freezer or not....guess I'll try it again. I will bet the boiled sample will produce a clearer ice cube....no dissolved gas.

to remove more variables, boil the water for both samples then cool one and re-boil the other then freeze.

Drew K

The graph:

That graph isn't in the paper this article is about.... at least that I can find.

.

I am curious about the graph you have posted though. Several things seem strange....

.

The rate of cooling for the cold water appears to be much faster than boiled water from 20°C down to slightly less than 0°C (roughly 800 seconds for that range of cooling vs over 1000 for the boiled water.).

.

Also the graph seems to show the initially cold water supercooling to about 20 below and (if the effect is to be believed) avoids freezing.

.

And then spontaneously jumps nearly 20°C around the 4200 second mark.... it certainly looks like cold threw the fight. Could the mafia really be responsible for this weird jump, that allows boiled to steal the show?

This graph was posted in a British Newspaper. I thought I would put it on CR4 without comment first.

1. The hot water seems to do what you would expect. It undergoes a few degrees of superheating above 100°C and then cools down to 0°C with a slight supercooling below 0°C before undergoing the slow transition to ice. Once converted to ice, the temperature then falls to about -25°C, which is the temperature of the freezer.
2. The cold water cools rapidly to 0°C, with a few unexplained kinks on the way. It then goes through a very long supercooling process down to an incredible -18°C. Meanwhile the hot water has already started to become ice. Since water and ice can only coexist at 0°C, the supercooled water rises rapidly to 0°C before undergoing its long transition to ice. But where does the energy to raise it back to 0°C come from? Having risen to 0°C, it then converts to ice in a shorter time than the "hot water," but its way behind in the process.

If this graph is to be believed, the difference in time taken for both samples to be completely converted to ice is negligible - definitely not worth boiling your water first!

The whole crux of the phenomena is that the cold water supercools so much, that it delays the ice forming process, whereas the boiled water starts the process much earlier.

Some people invoke the loss of dissolved gases in the boiled water as having an effect on the processes involved and I'm inclined to believe this. As to the theory on hydrogen bonds and covalent bonds - that's all it is - theory. As to "Mafia involvement," I just can't see what's in it for them.

Thanks for posting. This is interesting info. I wish I knew more about where it came from. It doesn't appear in the paper this article is written about.

It is all about the phase change. Look at phase change in metals, some must be supercooled before crystalization occcurs, but some will change faster if you insert seed crystals into the liquid.

Look at it on a molecular level where water molecules allign into ice crystal and actually expand increasing the volume of the solid.

The energy of transformation is the key here. The temperature fluctuates with the transformation energy and the molecules settle into lower states of energy.

Drew K

Sorry, but nothing you wrote explains the phenomena. If you have a point - I don't see it. When you write "energy of transformation" do you mean enthalpy (or latent heat) of fusion?

Dunno...will ahve to check my old textbook...after I finish this lovely cup of ale :)

Drew K

They need to use samples that are different only in temperature, but not different otherwise. This was a poorly controlled experiment.

I read the article hoping to learn something...I didn't

I still say this is bogus!

In one instance they used "tap" water and in the other they used distilled water.

That, alone, disqualifies this as anything other than a parlor trick.

I can make water run uphill, too.