Does Hot Water Freeze Faster Than Cold? Debate Continues Over The Mpemba Effect

Does hot water freeze faster than cold water? On its face this idea seems like it should be ridiculously simple to test, and even easier to intuit, but this question has in fact had physicists arguing for decades.

Erasto Mpemba’s observations initiated decades of research into the Mpemba effect: whether a liquid (typically water) which is initially hot can freeze faster than the same liquid which begins cold.

There’s a name for the phenomenon of something hot freezing faster than something cold: the Mpemba effect,  named for Erasto Mpemba (pictured above) who as a teenager in Tanzania witnessed something strange in high school in the 1960s. His class was making ice cream, and in a rush to secure the last available ice tray, Mpemba skipped waiting for his boiled milk-and-sugar mixture to cool to room temperature first, like everyone else had done. An hour and a half later, his mixture had frozen into ice cream whereas the other students’ samples remained a thick liquid slurry.

Puzzled by this result, Mpemba asked his physics teacher what was going on. He was told “You were confused. That cannot happen.” Mpemba wasn’t convinced by that answer, and his observations ultimately led to decades of research.

What makes this question so hard to nail down? Among many of the issues complicating exactly how to measure such a thing is that water frankly has some odd properties; it is less dense as a solid, and it is also possible for its solid and liquid phases to exist at the same temperature. Also, water in the process of freezing is not in equilibrium, and how exactly things act as they relax into equilibrium is a process for which — physics-wise — we lack a good theory. Practically speaking, it’s also a challenge how to even accurately and meaningfully measure the temperature of a system that is not in equilibrium.

But there is experimental evidence showing that the Mpemba effect can occur, at least in principle. How this can happen seems to come down to the idea that a hot system (having more energy) is able to occupy and explore more configurations, potentially triggering states that act as a kind of shortcut or bypass to a final equilibrium. In this way, something that starts further away from final equilibrium could overtake something starting from closer.

But does the Mpemba effect actually exist — for example, in water — in a meaningful way? Not everyone is convinced, but if nothing else, it has sure driven a lot of research into nonequilibrium systems.

Why not try your own hand at investigating the Mpemba effect? After all, working to prove someone wrong is a time-honored pastime of humanity, surpassed only in popularity by the tradition of dismissing others’ findings, observations, or results without lifting a finger of your own. Just remember to stick to the scientific method. After all, people have already put time and effort into seriously determining whether magnets clean clothes better than soap, so surely the Mpemba effect is worth some attention.

92 thoughts on “Does Hot Water Freeze Faster Than Cold? Debate Continues Over The Mpemba Effect

  1. I would be wondering if it had something to do with absorbed gases. Leaving it to cool naturally may let it dissolve more air back into it, less so than cold water from the faucet, but maybe enough. But having been heated (usually to boiling) the gases are purged. Therefore the quicker you bring the temp down the less chance it has to suck in more gases at the surface that impede freezing, maybe by lowering the freezing point a tad or by their insulative effect when they start forming bubbles in the slush ice, preventing heat from transferring out of the water effectively.

    It’s surprising how much we don’t know about water really. Some really strange effects from self polarisation of water ions too.

    1. I read an explanation of the effect once.

      If you set out equal amounts of hot and cold water, the hot water will lose mass due to steam. Energy loss is proportional to difference in temperature, so the hot water loses energy faster than the cold water until they reach mostly the same temperature.

      At this point the hot water has less mass and has a higher surface/volume ratio, and so freezes faster than the original cold water.

      1. I’ve learned somewhere that they use water heated up to around 40 C in ice rings in machines pouring a new layer of ice, as “it freezes faster than cold water” .
        Anyone working in an ice ring to confirm it?

        1. Rink, not ring. In my youth I spent many hours/days as a “rink rat” and hockey player. The hot water has more than one trick. At the back edge of the Zamboni hot runs water runs through a sprinkler bar to wet a mop. The hot water does create a bonding to the original ice surface, as well as smoothing scratches and divots not scraped off with the knife bar.

    2. Easy to find out by pulling a vacuum on the water to see if it has a similar effect. And also try water that has been boiled but then let to cool down to room temperature before moving into the freezer.

      1. even easier to just buy an ice maker.

        warm water makes small cubes. the first few rounds i dump back into the reservoir to obtain larger cubes.

        this does not represent all scenarios, however, putting warm water into your freezer is a waste if energy.

    3. Is it not simply that putting something hot in the freezer momentarily melts exisinting ice underneath it forming a much better conductive path to the freezing cold surroundings, and in some designs directly to the heat exchanger where as the cold one sits on what is largely a layer of air and relys mostly on convection to cool.

  2. Here in Minnesota, when the temperature drops to -40, the TV weather reporters like to show the kids in their audience a science experiment. They bring a pan of water to a boil on the stove, (about a cup), then take it outside and throw it high in the air. The spray of water crystallizes almost instantly, and falls to the ground as snow. (I don’t know if cold water behaves the same.)

    There was one weather guy who liked to leave various things outside overnight and try to use them to drive nails. A frozen banana can pound a nail into a board, at least a little ways, before destroying itself.

    1. I believe the reason why the hot water is used for the demonstration is that hot water has very low surface tension compared to cold water (this is also why cold water sounds different from hot water) . Because the surface tension is lower, the water breaks apart into more droplets and aerosolizes more, creating more surface area for the cold air to remove heat and freeze the water.

    2. That’s nothing. Round here our water is so hard we can use it to drive nails without freezing it :P

      Joking aside, I wouldn’t be surprised if boiling our water changed the composition enough to alter how it froze.

      1. It is not about boiled versus non-boiled water at the same temperature, though that would be an interesting test as boiled water ice cubes seem to have less air bubbles and are clearer than from the cool tap.

        Rather Mpemba is about 100C water appearing to freeze faster than 15C water, which is counter-intuitive as the 100C water has more total energy in the system that needs to be removed to freeze than the 15C water.

      2. I joke that the water where I grew up (the Cotswolds, ie, limestone country) was so hard, you could make pipes by taking a sheet of lead and bending it around the stream of water as it came out of the tap ;)

    3. Hah, thet effect that is often called the Leidenfrost effect (but that’s a whole different thing, where hotter pans might boil water slower than cooler ones) or the Mpemba effect (which we don’t know is the same effect yet).

  3. Is this the same phenomenon as supercooled water not turning to ice even though it is below the freezing point (until something happens, at which point it turns to ice quickly)?

  4. I would be quicker to believe the young experimenter actually got lucky and found a place in the freezer that was colder than the others, making the loss of heat a bit quicker in his tray than the others, making it appear to freeze sooner than it should have, when the real question might be why did the other students trays take so long to freeze that day? perhaps there is some matereal that will behave oddly and thru some un discovered process freeze from the hot state faster, but not water as the effect would hapen more regular and therefore be known to happen . Freezing used to be explained as the reduction in the motion of a matereals molecules until almost stopped. pseudoscience at best water pipes freezing hot first was observed in older homes and past days sure the hot tap froze first, it had less flow and less use so it had more chace to freeze than the cold kept in motion by constant use but that only proves moving water is slower to freeze or still water is most likely to freeze first, which is why we leave a tap drip in cold days to keep itr thawed.

    1. It may be named after the young experimenter, but this is a pretty popularly known effect (or myth, not sure) at least around me. Had no idea it had a name, just thought it was a weird quirk of water that most people knew about. Water is weird.

    2. Simple. The ice cream in the shared cooler reached equilibrium, but the previously hot ice cream now has evaporated some and is in the coldest spot in the cooler.

  5. I heard this myth. I tried it. a glass of 160deg. water and a glass of room temp. Both put in the freezer at same time. The room temp water froze 30 minutes before the hotter one.

    1. I couldn’t reproduce this – couldn’t even get water to 160deg without it turning to steam.

      I guess you’re measuring in F and I’m using C? (Fairly obvious in this case, but if I’d said 100deg that’d be boiling for me and probably room temperature for some parts of the US…). Point being, comes back to Eliot’s point below – document stuff properly! You can’t have 160deg water; you can have 160F or 160C.

      1. 160F = 160 Farad
        160C = 160 Coulomb
        What do “electrical capacitance” and “electric charge” have to do with any of the mentioned effects with water?!

        Did you mean 160°F and 160°C?
        Only Kelvin comes without a degree ;-)

          1. Heheheheh, that’s a thing now isn’t it, engagement marks for college subjects based on their mailing lists, forums or canned chat apps.

            “Oh yes, a valuable observation, I totally agree with you, and also (repeats previous point using a thesaurus.)… “

  6. I actually tested this for my 8th grade science fair with a C64 and a thermistor. I wrote a simple BASIC program to record the time and temperature once every minute, and stop when the temperature reached freezing. I ran a wire from paddle (potentiometer) input pins on the C64’s joystick port to the thermistor in an ice cube tray in the freezer. I tested both cold water and hot water several times and the hot water consistently froze faster. This was 30+ years ago, so I don’t remember the numbers, but I believe the difference wasn’t huge in practical terms. It was something like 10 minutes faster over 2+ hours of freezing.

    For the record, my science teacher also said the same thing: I must have made a mistake, hot water can’t freeze faster than cold water

    I also tested warm water, which had been heated up some, but not to (near?) boiling, and it took longer than either cold or hot water to freeze. One thing that occurs to me now is that I probably used tap water. I don’t remember if I boiled it, or just brought it close to boiling, but if it was boiled, the hot water may have been more pure as a result. I wonder if that might have an effect on the cooling time.

      1. No.. they coat the bottom of the boiling vessel. One off in a pan that’s washed regularly you don’t notice the thin film, but look inside your kettle.

        1. On reflection, I’ll amend that to some of them settle, but some form particulates in suspension in the bulk, potentially leading to higher availability of nuclei which freezing proceeds from.

    1. Can you give us some idea of the shape of your containers and the depth of fill.. thinking it might have something to do with top surface area to volume ratio.

    2. This sounds like two separate issues. The hot water should cool faster, because there is a greater temperature difference. However, once the hot water gets to the same temperature range as the cold water, they should cool at identical rates (ignoring whether the Mpenga effect exists).

  7. I can’t help but notice that hot water, in order to freeze, must first become room-temperature water. Surely there are some ancillary effects that would create a minute difference between that point and water that wasn’t heated first (evaporation being the major contributor). I imagine this would be somewhat difficult to even measure, much less notice at a glance.

    1. That’s the key paradox. But that at least can easily be resolved.

      Temperatures are defined only for objects at equilibrium which the objects in question certainly are not.

      1. Indeed, it’s the old mathematical induction trap: the mathematician sees the burning trashcan on the desk, moves it to the floor, and, confident that a solution exists (as the problem has been reduced to an already solved problem), goes on their way.

  8. It has nothing to do with the rate of heat flow? As difference in temperature between water and surroundings rises so does the rate of heat flow so water will lose heat faster hence turning into ice faster.

    1. This is equivalent to taking two capacitors, charging one up to a higher voltage, and seeing how long they take to discharge through identical resistors? The higher voltage one starts off losing charge faster, right? But I presume you know the result.

      As Doug points out, the hot water has to pass through room temp in the freezer.

      But really, try the experiment yourself.

      1. The problem with that reasoning is that temperature is only properly defined for objects at equilibrium.

        Like those silly experiments when a hot lead ball is dropped on a block of ice. What’s the temperature of that object? There is no sensible answer unless you wait for it to reach equilibrium.

        If the experiment involved stirring constantly then I’m sure the effect would vanish and we could assign reasonable temperature to each step.

  9. What the exact “effect” is has never been well defined enough for my tastes. Fill an ice cube tray with hot water on one side, cold on another, maybe leave a gap in the middle. Which freezes first? I’ve done it — the cold side. (Surprise!)

    But here, for the first time, I’ve read that the original “effect” is that warmer ice cream mixture crystalizes faster than cold. Ice cream is a _lot_ less uniform than water. Hell, that’s the point: ice regions form, but they’re held apart from each other by fat/cream so you don’t just get a solid ice block. The whole churning thing keeps new water pockets exposed to the cold walls.

    If something about the heat makes for smaller water pockets, you could totally imagine them freezing faster because they have less thermal mass. (You might also expect this “faster” ice cream to be warmer than one of the slower ones on aggregate, so it might re-melt faster unless you keep taking heat out.)

    The point is that the phenomenon is never really well described in the public discussion about this, and that’s really step #1 in science — figuring out exactly what you’re talking about. And that’s how something that’s possible (ice cream freezes faster) got turned into something that’s pretty much ridiculous (water freezes faster).

    That’s the _real_ lesson here, IMO. Document phenomena precisely.

    In retrospect, I feel a little bad about the person I totally mocked about this hot water thing. It wasn’t their fault that that this entered the public consciousness imprecisely.

    1. Note that in this experiment, as any sane person would expect, cold water still reaches 0° Centigrade before the hot water does.

      The Mpemba effect is about at what point it is observed the water turns into ice (freezing), not about reaching a certain temperature.

      As the freezing point of water depends on a lot of things, even in lab conditions there is already a lot of variance in exactly when water crystalizes.
      Try to take the same water, divide it in equal amounts, put in the same containers and put them in the same freezer. You will notice even these will not freeze at the same time. (You can retry the experiment to account for the location of the container in the freezer.)

      There are no theories that explain/predict, or even scientific observations that confirm the Mpemba effect.

    2. Major assumptions are leaving you ridiculing the entire premise, as well as all those who haven’t made up their minds or experimented yet. Seems a tad unprofessional.

    1. You are correct, sir! Getting hot water to freeze faster than cold water can be accomplished by engineering conditions that manipulate these factors in favor of faster heat loss by the hot water. I’m thinking open containers that promote convection to keep moving hot water to a broad surface. Lots of evaporative heat loss at the top, bonus conductive heat loss at the bottom. Add in a big surface to volume ratio of a thermally conductive container and hot water has a fighting chance to freeze first.

      If a “scientific experiment” is designed to account for all those factors (plus a few more like impurities in the water) I’m thinking the thermodynamics of heat loss and the chemistry of crystalization will be vindicated yet again.

  10. Hmm. Been a couple years since this last went big. Fourth of July heatwave in the US. It’s time.

    Here’s the weird thing: this is a reproducible effect, both with mixtures and with water, so THAT it exists under some conditions has been confirmed. It’s the how and why, and what are the sufficient conditions, that are pretty wide open. (“I tried it once at home and didn’t see it” isn’t a refutation, by the way)

    Yes, more energy must be removed from the higher temp starting case. IF the fluid was always in a state of perfectly uniform distribution of heat, and there was no possibility of differences at the interface with the environment between the cool and hot starting cases, then it is a simple Newton’s law of heating/cooling problem, and the cool initial case will freeze first every time.

    But, neither of these holds.

    One way or another, for this effect to occur, the rate of heat flow from the sample to the environment must be significantly greater for the hot initial case than the cool initial case. There are a number of ways this can, potentially, happen, any or all of which may be in play in a particular circumstance.There have been a number of hypotheses as to what is happening when this effect is observed. Mass/latent heat loss due to evaporation leads to rapid initial cooling and a lower mass tho freeze in the end, for example. Contact effects, where the hot initial case ends up with better mechanical contact with the enclosure (freezing condensate, melt and refreeze a frost layer, or any of several other possible reasons). The nonuniform distribution of heat in the fluid as it cools (maybe convection cells are set up in the hot case, but not the cool case, improving heat transfer to the boundary from the center)

    This is what can make science so much fun. Every result opens new questions, the first often being “What am I really asking here?”

      1. It seems you’re rejecting the entire premise of the article, directly challenging anyone who might have observed it or believe the report of someone else’s observations. Why so antagonistic? I’d understand if this was a journal, but this is an observable effect (*under the “right” circumstances). Your experiment not leading to any observable differences from your expectations isn’t surprising, nor is it evidence the phenomenon isn’t real.

  11. I had an argument with a roommate about this “effect” we weren’t in a lab but we tried to follow some form of scientific method by using multiple samples of tap and bottled waters in different materials containers (glass, polystyrene, metal) anyway after some debate we couldn’t reproduce the “effect” reliably, in certain cases it wasn’t clear but our freezer was certainly not a lab calibrated machine, I’d say far from it. Probably accurate enough nonetheless to “see” there was nothing significant to see…

    1. It probably depends on the amount of ice on the heat pipes. A hot container would melt the ice and get in contact with the heat pipes. Then freeze faster because of better heat transfer.

  12. The effect is easily explained through the use of the probability drive on the heart of gold spaceship. You see the odds of the hot water freezing first are 10 to the power of -16 and falling. So long and thanks for all the fish. ;)

  13. Normally your water heater boils out the minerals of the tap water. Water with less mineral content will freeze faster. So a true experiment of hot versus cold must have the same purity to test the theory that hot freezes faster than cold water.

    1. Such confidence in that answer.
      ‘Come and see the Dunning-Kruger inherent in the system!’

      If your water heater boils the water, you should have it checked. Seriously, that’s generally bad.

  14. I did the experiment and was unable to get the hot water to freeze faster.

    Experimental setup:
    Deep freeze (temp well below freezing)
    Two aluminum cups (about 1/4 full)
    Purified Water from a water cooler (like you have in an office)
    One cup filled from the hot (almost boiling) side, one from the cold (refrigerated) side.

    Tried to keep the conditions as similar as possible, ran the experiment twice to see if freezer placement was an issue. The samples did get agitated whenever I checked them, so, that might have an effect.

    I didn’t record the exact temperatures or timing since I don’t feel the results were significant. Too many variables are hard to control/uncontrollable in my home for me to list them here.

    1. Gonna have to figure a non-disruptive method of determining freezing vs zero C for this. Maybe weak light beam and change of refractive index (beam shot in from above surface moves off target sensor, but if surface skins first it’s not reliable to determine bulk freezing.)

      1. That itself is an interesting question, how to reliably detect freezing, though a commenter above suggested they were able to observe just the temperature change with a c64 and see the effect.

        1. As the water goes through the state of change the temperature will plateu. You could measure the length of the plateu and determine when the state change actually occurs. 1 B.T.U. will raise 1 pound of water 1 degree F. It takes 144 B.T.U. to bring water from 211F to 212F. The commenter with the c64 setup would have seen this plateu occur. I do not know how many B.T.U. it takes to change state on the cold side.

  15. Conductivity. When you boil water it evaporates and increases the conductivity of the water which causes it to cool faster. Try it with pure water. The cold water will freeze faster because it has a lower btu value

  16. Sometimes in science, the observation of something that it turns out you had erroneous data on, leads to questions which bear further investigation, even if the original premise may be deemed false because it was not what you thought it was. In this spirit then, I say that Mpemba may have opened a valuable line of inquiry, whether it leads to new science or greater rigour in our methods. But… the conditions of the original observation may have been other than perceived….

    The classic science class ice cream making experiment that I have seen used, is not really to demonstrate the making of ice cream but more to illustrate how you can use the energy of solution required for salt in water to suck heat out of a slurry and drop temperature below freezing point to around -14C

    Therefore the ice/salt/water bath used may have been subject to variation. I offer for consideration the postulate that a large part of the class had cooking experience, and when directed to add salt to the ice bath, sprinkled a little “to taste” as if they were preparing edible food, whereas Mpemba just dumped a load of salt in there, as is appropriate to the task at hand. Therefore his icecream froze first because he had made a more effective ice bath.

  17. We had this discussion in chemistry class back in 75, warm water freezes slower under normal conditions. I have noticed that ice cubes made with water from the hot spigot (not necessarily hot) seem to get stuck in plastic trays less. Where’s my Nobel prize?

    1. If the air is cold enough, the water first forms a sheet of ice on the surface that encloses the fluid underneath. The freezing water expands and presses a droplet out of the weakest or last open spot in that sheet, which freezes on top. That releases some heat which ensures this spot continues to be the weakest in the growing ice sheet, and some more water gets pressed out. Does not work if the air is too warm, or if the water has other options to expand.

  18. well as hot waters inner mechanisms works faster its no surprise reactions occur faster, the fact is hot waters bubles/particles are seperated they dont get extra heat, so if total negative heat is enough to freeze hot water chemical processes will occur faster. probably relative to toching surface and its heat transmission capability, but on equal transmission surface, the opwrations will mostlikely happen faster as far as there is enough negative energy to supply system. So shortly heat transfer probably is faster, as time accelerates effectively by increase 9n heat ;)

    1. “Exploring more configurations” is exactly what I posited as an explanation when I first heard of this effect, along with the idea that ice has a crystalline structure, and crystal structures are generally good thermal conductors. This means heat will quickly be conducted *away* from the core of any crystals that do happen to form, maybe increasing the chance they will persist.

      But as others have explained, there’s no formal definition of “freezing” that can make this effect robust and fully repeatable at this time, but these sorts of problems always start vague so I expect this will get nailed down in time.

  19. When I first heard of this, the only thing that made sense to me is the elementary school explanation of matter.
    ie:

    All molecules are moving. Cold molecules move less than hot ones, and freezing is taking un-arranged molecules in a liquid and arranging them into a regular pattern.

    I assumed that freezing hot water would just be faster-moving molecules arranging into the regular pattern much more efficiently than slower moving ones, similar to the writeup saying that the high-energy matter is able to occupy and explore more configurations.

      1. It takes longer to radiate energy if you assume the energy is distributed and radiated evenly and everything else is static, but that’s not necessarily the case.

        For instance, ice is crystalline, and crystalline structures are excellent thermal conductors in certain directions. If hot water can explore more configurations along the boundary, some of which might be crystalline, this could violate assumptions of evenness.

        There also convection and potentially evaporation happening, ie. hot water may evaporate its highest energy mass thus leaving a *smaller* mass behind to cool compared to the cooler water, thus violating assumptions of equivalent mass being cooled.

  20. One other weird thing about water is that it is sort of like a polymer. The water molecules stack up with their polar ends and form little snakes. The longer the water is still in the presence of infrared light, the longer the snakes get. This can interfere with mixing.
    Gerald Pollock at the University of Washington has published a lot of books on this. Some of it goes too far. But there are many true and interesting applications of structured water. And I think a few fake and crazy ones.

    It is this polymer state that accounts for the fact that water is denser as a liquid than solid.

    Cells seem to exploit this in order to transport things. The water inside a cell seems to be more structured. It was observed for a century that water inside a cell is denser than water in a beaker.
    I suspect that boiling the water breaks all these snakes up and when the water is then cooled, the surface tension is less, and the convection is faster because the water is less viscous.

  21. I would like to hereby declare the opposite effect the “CRAIG EFFECT.” namely, cold water boils faster than already hot water. If this Mpemba effect is real (spoiler: it really, really is not) then certainly the reverse must also be true.

    Now I’ll be clear, “Mpemba effect” is a real effect- it is just a psychology effect and not a physics or science effect. Science that isn’t repeatable is not science. Reading all 70+ comments, not a single person has linked to a credible controlled experiment. Lots of super fun speculation and detailed knowledge of chemistry and physics, but no practical proof of concept. And it should be easy enough given the constraints of the effect. Concepts like supercooled or superheated water without freezing (or boiling) are well established, reproducible at home with minimal effort and easy to verify. Such fringe cases do not “prove” anything.

    The real problem with this whole disaster (besides wasting so much time and money) is that it is so poorly defined as to be impossible to test. Defining an answerable question is a rare valuable skill. This was addressed above many times. One example, “Freezing” being a relatively lay-definition is a soft endpoint. I’m not aware of an accepted standard for this, but some examples off the cuff are “DT/dt less than 95% of maximum” or “liquid water <5% of starting by mass" or literally anything measurable. Using the MK IV eyeball to determine something like this is just dumb and any result is de facto invalid. Look up n-rays for great validation of even well meaning scientists being fooled by their own experiments.

    I also wildly speculate (not really, look up crystal clear ice at home… my work) that dissolved gasses or other solutes may have some tiny effect on freezing rates but on a macro scale are totally negligible.

    Annnndddd now I've wasted my time on this. haha

  22. If I were gonna do it, I would make the assumption that the only thermal momentum involved was macroscopic physical momentum, i.e. convection currents.

    It is known that water has something of what may be called a “memory” in that it preserves small eddies and bulk motion for quite a long period, such that it will swirl the same way out of a container, prompted by miniscule amounts of motion in that direction, same sense, clock or anticlock as it was swirled into it, provided no other factors are present, like stirring it the opposite way first, or having a spout where the topology favors one or the other direction.

    So, momentum sploit.. kill all momentum in the cold sample, enhance momentum in the hot sample.

    The containers I might select (Out of common super easy to find containers) are 2lb ~900g margarine tubs. I want a container where one can fill it to a depth equal to the radius, therefore we fit two circles of diameter equal to the radius across it, enabling a circulating toroid to form in convection currents, with minimal resistance.

    To minimise factors involving conduction through the base, like the samples seeking thermal equilibrium primarily through the metal freezer rack they are resting on etc, I would stand each on it’s own square of styrofoam. half inch thick should be enough, but best available.

    So experiment would proceed thus.. set up on a tray, for convenient transfer to freezer, two styrofoam pads, with tubs on top, fill one to radius depth with tap water and leave it standing there for hours, want all momentum in it to die and for it to come to equilibrium with room temperature. Now boil water, and due to margarine tubs being prone to melting, let it drop to 90C and fill second tub to radius level.

    Now, moving extremely carefully, put it all in the freezer…

    What I kinda hope will happen, is that robust convection currents form in the hot tub, which when established will persist for a couple of hours. Versus weak if any currents established in the cool tub. (Also the closer it gets to 4c, the harder it is to maintain currents in it, due to water properties changing, which may dampen weak currents but allow a strong established current to keep on motoring, or at least die out less soon)

    If that doesn’t work, then I doubt much else will, unless it’s an area thing driven by evaporation.

    1. I like the styrofoam idea. I was thinking of a clear cube of acrylic or epoxy with a cylindrical core which would be the vessel to hold the water. Perhaps having two of these to run the experiment concurrently. A probe is mounted to the side of the cube to measure the water temperature. Perhaps a bit of foam over the probe to insulate it from the cold air. A lid could be placed on top of the cube to isolate the water from the air in the freezing container. Multiple runs could be done (lid on vs lid off). The vessel should be either flame or solvent polished and allowed to off-gas first to remove any potential nucleation sites.

      The freezer could be made of clear plastic panels into a cube. A peltier might make a suitable heat pump for cooling the freezer enclosure. The hot side of the peltier could be water cooled with a large volume loop to even out any room temperature variations that might affect cooling performance of the petiler element. Hot and cold side temperature, and voltage and current could all be monitored to see how much energy is removed from the system and how much energy and power was required to do so. A camera could be placed so as to watch the water sample for ice crystals to form. Recording the video would allow measurement of time to freeze via frame counting and timestamps.

      All water samples would be boiled first and allowed to return to room temperature before heating to target temperature or insertion into the freezer.

      This is the best I can think of at almost 1am

      1. > All water samples would be boiled first and allowed to return to room temperature before heating to target temperature or insertion into the freezer.

        Heating a second time could reduce the total mass of water that needs cooling (evaporation) which would skew the results. Then again, evaporation from the heat could be the cause of this effect, so weighing everything carefully could be one way to ferret this out.

        1. I probably should have mentioned that the water should be measured after boiling rather than before. As I mentioned before, a lid could be put on the block/vessel to test if evaporation makes any difference.

  23. The answers to this is simple, agitation. Warmer particles move faster, increasing the likelihood that they will find the fixed position required to form a crystal as the surface cools.

  24. Are you all serious? In theory or whatever, it’s bullshit old wives tale regurgitation. None of you who believes hot water freezes faster can provide proof, with simple or complex experiments. Say whatever you want, in theory is not proven fact. None of you idiots can make it happen. I thought this was where the smarter kids were supposed to be hanging out.

  25. “But there is experimental evidence showing that the Mpemba effect can occur, at least in principle.”

    There really isn’t. There’s a review from a number of years ago collecting data from people who’ve tried to replicate Mpemba’s original result. None of them could. All of their data lie on a very consistent line and Mpemba’s results are way off it.

    It’s just bad data. It’s over 50 year old data at this point. Happens.

  26. That’s really informative! Perhaps, maybe it deals with the space created between the hot water molecules? So the vapour freezes faster, causing a chain reaction, making water freeze faster. Who knows? I’ll research not about this and write in my blog too! Great post btw

Leave a Reply to CanuckfireCancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.