▲ 160 points
back
106 comments
Arxiv link to the actual paper: https://arxiv.org/abs/2008.02373
Looking at the actual graphs, what is presented appears at first blush to be a bistable system prepared in its higher-energy stable state. (Or at least, prepared in a way that several of its constituent subsystems "fall into" that state.) Transitions to the lower-energy stable state require overcoming a barrier, and jumping over that barrier is of course faster at higher temperatures. At lower temperatures you essentially have a slow internal self-heating from the periodical jumps of subsystems through this forbidden region by thermal excitations, not unlike phosphorescence having to tunnel through an energetically forbidden region; this self-heating prevents the thing from cooling down fast enough.
Actual applicability to water is kind of harder to evaluate. I've long thought that you could indeed have an Mpemba effect caused by persistent induced convective currents in a fluid: so a hotter thing placed into a fridge might create a more dramatic internal flow as its outer boundary layer cools and changes in density, the convective current would certainly increase the slope of the hotter cooling system beyond the naive temperature scaling, and might then persist as a physical difference even after the two systems arrive at the same temperature, leading to a faster cooling speed of the convective system that "started out hotter". But I mean I have never really done experiments to confirm that sort of thing.
> Actual applicability to water is kind of harder to evaluate. I've long thought that you could indeed have an Mpemba effect caused by persistent induced convective currents in a fluid: so a hotter thing placed into a fridge might create a more dramatic internal flow as its outer boundary layer cools and changes in density, the convective current would certainly increase the slope of the hotter cooling system beyond the naive temperature scaling, and might then persist as a physical difference even after the two systems arrive at the same temperature, leading to a faster cooling speed of the convective system that "started out hotter". But I mean I have never really done experiments to confirm that sort of thing.
This sounds like what I was thinking but I'm not sure. Do you mean the actual boundary layer in the hotter fluid would have a higher convective heat transfer, due to increased turbulence/thermal mixing at the surface of hot water (versus cold water)? Thus hot water will cool faster?
You've put into scientific terms what I can't cause it's been too long since I've done any physics.
In layman terms, the molecules in hot water are bouncing around more and thus easier to form crystal structures because of the higher-energy state. Molecules at room temperature have to be forced into crystal structures which takes longer.
I'm reading this as hot water having more effective heat transfer deep into the container, potentially via some type of self-mixing due to turbulence from the sharp temperature gradient?
This would seem very easy to measure, with and without internal barriers to prevent large scale convection.
Once upon a generation ago, I worked at an ice rink. "Cleaning the ice" using an ice resurfacer involves cutting the top 1/16" of ice off and laying down new water. Hot water is preferred because it freezes faster and produces a denser sheet of ice; a denser sheet of ice is going to be better for skating.
It's nice to see more papers on it.
Hot water and is preferred because it has less gasses dissolved and this way it produces clear ice. Also cold water could freeze too quickly, if the rink is very cold, before it has time to spread evenly.
I am pretty sure any tiny time difference, if it exists, would not be worth the hassle of supplying hot water.
I've always imagined that hot water would also preferentially smooth the surface, melting any high spots, deep cuts, or granular ice that might have survived the zamboni.
I don't think it's the effect described in the article. It probably wouldn't be noticeable. It may be that by melting more of the ice below it removes all roughness of the surface where there could be air which is working as an insulation. Additionally, depending on details, by melting more ice it could probably make exchange of heat faster between the surface down below.
Thank you for getting the zamboni song stuck in my head. https://www.youtube.com/watch?v=vVkJbvv3pHg
Somehow this article provides no explanation at all of how this actually happens. Perhaps the research being reported doesn't really have an explanation either, it's just more demonstration under more controlled conditions?
Or the explanation as it were is just "abstract" and "geometrical", basically just mathematical?
Very odd phenomenon.
The fact is, you do need to get rid of more heat when using hot water.
Ergo, some sort of self-reinforcing feedback loop must appear when you use hot water, that sucks heat out of it faster. If you use cold water, the "turbocharged" mechanism does not operate for some reason, and heat is lost at a lower rate.
I've no idea what that mechanism is. It's nonlinear behavior for sure.
So does the hot water freeze faster or does it cool faster?
Knowing if it freezes faster but doesn’t reach 0 degrees faster would help point to the cause. Especially if it’s known whether the temp is uniform. Is the room temp water super cooling?
Someone needs to do an experiment measuring temperature distribution over time in both hot and room temp water as it freezes.
I have a vivid memory of being taught this in 2nd grade, probably because it seemed (and still seems) extremely counterintuitive. I think the intuition given was that hot water is less dense than cold water, so the coldness can more easily permeate through the liquid (I imagine a vein-like pattern). Something like that.
At least according to Wikipedia, it seems that it’s a bit more complex: https://en.m.wikipedia.org/wiki/Mpemba_effect
Anyways, I’ve been repeating this fact for 2.5 decades, so shout out to my teacher for blowing my young mind.
Cold is not something that can penetrate.
Cold is merely a relatively low amount of heat energy per unit of mass.
When something cools heat energy is conducted/convected/radiated away from an object causing it to have a lower amount of heat energy per unit mass.
In a similar way, I find the idea of cold permeating a liquid counterintuitive, because I can picture heat going into cold, but not cold into heat. The energy has to go somewhere!
All the talk in the comments about dissolved oxygen brings back a few memories of student lab work that went wrong. I was given a freeze-pump-thaw protocol to follow in order to remove oxygen from a copolymer solution. Basically we used liquid nitrogen to freeze the mixture, then used a vacuum pump to remove the gases above the mixture. When we thawed the mixture, the dissolved oxygen would escape, in order to equilibrate with the amount above the surface. I repeated this for a few cycles. It was all done with sealed glassware.
Unfortunately, the glassware kept cracking during the thaw stage. This went on for a couple of months, till finally my supervisor reviewed the procedure, and realised that, whether cracking occurred, depended upon the relative concentrations of the two polymers. If I recall correctly, the problem occurred when the polymer with the lower freezing point (which, thereby, thawed first) was present at a lower concentration. It was basically trapped inside the other still-frozen polymer, and expanding, as it thawed ... till crack!
The solution was surprisingly simple. Change the method. Bubble nitrogen through the mixture to drive out (sparge) the oxygen. Replacing the oxygen with nitrogen was beneficial, as nitrogen didn't interfere with the copolymerisation reaction.
The main problem was that I had lost two months of experimental time, and this was just the start of a series of experiments and instrumental analyses. It was now Easter and the deadline to hand in my student thesis was early December.
Please don't believe anyone who says students have an easy life.
I read the whole article and couldn’t understand how or why this happens. Is the reason behind this unknown?
My high school chemistry teacher simplified it down to the greater "rate of change" of temperature dropping, for the hot water, causes it to end up freezing faster than the cold water.
The best explanation I
got was that covalent bonds are weaker in hot water than cold.
It’s apparently not true for very pure water.
Just a few days short of 35 years ago, this came up in “The Straight Dope” [1].
[1] https://www.straightdope.com/columns/read/422/which-freezes-...
> Jearl suggested that the most likely explanation for this was evaporation: when water cools down from near boiling to the freezing point, as much as 16 percent evaporates away, compared to 7 percent for water at 160 degrees.
Random tangent: TIL that The Straight Dope predates the web and was originally a newspaper column.
This was known by many Antarctic scientists before 1963. I remember my dad telling me you could throw boiling water into the air and it would freeze into millions of tiny ice crystals before reaching the ground, but that the same wasn't true of cold water. I think he said it was due to thermal inertia.
"Thermal inertia" is more formally known as specific heat capacity or volumetric heat capacity, which is a substance's resistance to change in temperature.
It's similar to (but different from) mechanical inertia, which is an object's resistance to change in velocity.
The latter includes the concept of momentum, that an object with nonzero velocity resists changing it's velocity. There is no such thing as thermal momentum, remove the temperature differential and a thermally uniform substance that's changing temperature at 10 degrees per second will stop getting colder instantly. For related reasons, there's no overshoot when you reach equilibrium, either.
That sounds like a totally DIFFERENT effect, almost certainly caused by droplet size.
Many Canadians have known about this for a while, here is what it looks like...
I'm not 100% sure it is the same phenomenon but growing up in a very cold climate where I had to park my car outside in high school I remember many times getting into the car where a half full Poland Spring plastic water bottle would sit with water. If I tapped the bottle - the entire thing would freeze in under a second. It looked amazing.
(I always thought it was a pressure thing - but not positive if that is what the article is describing or something more nuanced)
there is a similar inverse process for hot water as well... if you take a very very clean and perfectly smooth surfaced (on the inside) cup/glass and microwave it to just before the point of boiling, then drop anything into it (sugar, coffee, a teaspoon... pretty much anything) it will very energetically, instantly boil.. often vertically... be careful performing this experiment.
Edit: here this will help explain it https://www.animations.physics.unsw.edu.au/jw/superheating.h...
Not sure if it's related to the article but this is called superfreezing or supercooling. You can do it with distilled water at home too I believe.
Supercooled - https://van.physics.illinois.edu/qa/listing.php?id=1618
If you put warm water in your ice try, you can get clearer ice near the surface, because the ice will freeze from the top surface instead of from all sides at once trapping bubbles of air with no where to go :D
That's interesting. But don't drink the hot water from the hot water tank. It can be dissolved with unhealthy minerals and chemicals. A few years ago, I had to replace a couple of hot water tanks. When I drained them, the last few gallons at the bottom of the tanks looked like yellow vomit, it was seriously disgusting.
There has been some controversy about this for years, see for example:
https://www.chemistryworld.com/news/mpemba-effect-in-hot-wat...
https://www.repository.cam.ac.uk/handle/1810/263847
This new paper may prove to be conclusive in supporting its existence, but as I'm not an expert I'll give it some time in peer review before coming to any conclusion.
This might be a better link: https://www.sciencenews.org/article/physics-new-experiment-h...
Punchline seems to be that “hot” but out-of-equilibrium material cannot be summarized by a “temperature”, providing many paths towards freezing (on the landscape of microstates), some of which could (efficiently) circumvent the “low temp” state.
A marginally related question for the experts here: I love extremely cold water, juices and drinks, so I often keep them in the freezer in plastic bottles. Now, if I take out a bottle of something that is just about to freeze, as soon as I open the bottle, it almost immediately freezes although I am certainly not removing heat from it but actually doing the opposite by keeping it with my hands, not to mention the warmer environment in which it is now.
I couldn't explain it in any other way than with the different pressure, that is, the water expands when it freezes, so the almost freezing liquid is already under higher internal pressure compared to the moment it was closed, but as soon as I open it the pressure drops abruptly, which would be enough to compensate for the higher temperature, de facto freezing the water.
Is this correct?
This is one my favorite science stories. A tale of how truth eventually surfaces and holds in front of ridicule.
https://en.m.wikipedia.org/wiki/Mpemba_effect See the section called "Mpemba's Observation"
Perhaps it has to do with the increased turbulence in the hotter fluid? This would be a result of the higher saturated vapor pressure of the liquid leading to greater evaporation/boiling. Even as it cools down the higher temperature difference between the hot fluid and room temperature should produce a more turbulent boundary layer along the surface of the water.
The turbulence will reduce the thermal resistance associated with laminar surface air films on colder surfaces, which would normally be insulating surface, so you'd get a faster heat loss then with a more stagnant layer of cold water.
Sometimes when we are so familiar with something, we forget how special it is.
Water is really amazing. It expands when it freezes. Hot water freezes faster than cold water. It dissolves almost anything, but not too fast.
Suggested explanations
https://en.wikipedia.org/wiki/Mpemba_effect#Suggested_explan...
We tried to replicate this for one of the kid's elementary school science fair project. We were unsuccessful, but the judges really liked it anyway.
yes faster. but not earlier.
No. That's Newton's theory of cooling. The phenomenon is hot water freezing earlier than cool water.
They do mean earlier, don't they?
came here to say this. the headline for this finding is almost always misleading.
Emphasis on “sometimes”.
I wonder if this is a right analogy. Imagine two cars one going fast, one slow. If driver of the fast car steps on the break harder than the driver of the slow car, he will stop faster. The temperature difference is similar to how fast you step on the break.
> Hot Water Sometimes Freezes Faster Than Cold Water
No it doesn't.
Unless you want to word play between English and really specific small technical processes, perhaps.
What's interesting is it's an old urban legend the predates Usenet.
Usenet did kill the glass is a liquid legend though.