back to article Can't stand the heat? Harden up if you want COLD, DELICIOUS BEER

Why does warm water freeze faster than cold? Researchers from Singapore believe they've cracked the long-standing paradox – it's down to the way stretched hydrogen bonds store energy. The warm-to-frozen question has vexed boffins including Aristotle, Francis Bacon and Descartes, but didn't become part of accepted science until …

COMMENTS

This topic is closed for new posts.
  1. arctic_haze
    FAIL

    "The hypothesis should, The Register supposes, be relatively easy to test: different initial states could be calculated for different temperatures, and used to predict the degree of the Mpemba effect"

    Well, if The Register looked at the original paper it would notice that this is exactly the point of its Figure 1.

  2. jake Silver badge

    Total, complete & utter bullshit.

    Don't believe me? Try it for yourself. I did, when I was about six years old. I've demonstrated my results many times in the ensuing nearly five decades.

    1. Anonymous Coward
      Anonymous Coward

      Re: Total, complete & utter bullshit.

      Thank you for your comprehensive and rigourous scientific analysis

      1. jake Silver badge

        Re: Total, complete & utter bullshit.

        Again, try it for yourself.

        Sticking hot & room-temperature water into a common-or-garden freezer and observing the rate of crystallization ain't exactly rocket science.

        Please, do report back. We're all earseyes.

        1. Anonymous Coward
          Anonymous Coward

          Re: Total, complete & utter bullshit.

          Thanks, Jake, for what sounds like your expert and meticulously conducted scientific experimentation using laboratory grade materials and equipment. Did you use Evian or Perrier?

    2. Schultz
      Boffin

      Re: Total, complete & utter bullshit.

      It's not bullshit if the process is not adiabatic.

      Let me translate adiabatic: an the adiabatic system is always in its equilibrium. If the hot sample cols adiabatically (== slowly), then it will be at equilibrium when reaching, say, 20 degrees and it obviously can't cool faster than another sample that started at equilibrium at 20 degrees.

      If the cooling is not adiabatic, then all kinds of funny things can happen: One sample might go below 0 degrees without crystallizing (supercooled liquid), or a supercooled surface layer might slow down the evaporative cooling process, ... I didn't read up on it so I can't give you anything beyond generalities.

      1. jake Silver badge

        @ Schultz (was: Re: Total, complete & utter bullshit.)

        Again, try it. Report back.

        Why are folks so recalcitrant when it comes to actual experimental observation?

        1. Thecowking

          Re: @ Schultz (was: Total, complete & utter bullshit.)

          Mainly people like to read scientific evidence, not anecdotes which are a sentence of unverfiable data.

          If you actually want people to listen to you, tell us what you did, how you did it, what your results were and how many times you did it. Then you can also explain why other people might observe something different to you, or why your methodology is better than theirs.

          I can blithely say whatever I want, but unless I can back it up with some data, people are free to deride me as much as they want. They probably should.

          For what it's worth I've seen this effect with water in ice cube trays, but I'm not willing to say it's evidence of anything because the trays are different materials and shapes. Though the cubes themselves are approximately the same size. Then again, anecdotes do not an experimental dataset make.

          1. oolor

            >Then again, anecdotes do not an experimental dataset make.

            That requires some serious statistics gymnastics!

            >Thanks, Jake, for what sounds like your expert and meticulously conducted scientific experimentation using laboratory grade materials and equipment. Did you use Evian or Perrier?

            The finest artesian Napa Valley well water, expensive crystal glass, and a prototype chip cooling unit he built from a repurposed soda fountain to turn his calculator into a supercomputer.

        2. Ian Yates
          Boffin

          Re: @ Schultz (was: Total, complete & utter bullshit.)

          If you can prove it doesn't exist at all, you should really be publishing your results as the scientific community is generally in agreement that this is a real phenomenon... hence this published paper.

        3. Vociferous

          Re: @ Schultz (was: Total, complete & utter bullshit.)

          @jake: I really wish people wouldn't vote you down for something they can easily check themselves is true.

    3. Dave 126 Silver badge

      Re: Total, complete & utter bullshit.

      > Total, complete & utter bullshit.

      >Don't believe me? Try it for yourself. I did, when I was about six years old. I've demonstrated my results many times in the ensuing nearly five decades.

      jake, you didn't prove the Mpemba effect doesn't happen, you only proved that the Mpemba effect doesn't happen in the circumstances you tested.

      Variables include respective temperatures of the two (or more) liquid water samples, their size, the shape and size of the containers, the temperature of the freezer, the texture of the test containers (nucleation sites), any difference between the two samples in terms of dissolved solids or gases. It is possible to have observed what you observed, yet still consider the Mpemba effect to be plausible in other circumstances.

      There is also ambiguity in the definition of the term, as well; whether it refers to when the water starts to freeze, or to when it is frozen completely.

      EDIT: Link added http://phys.org/news188801988.html

    4. James Micallef Silver badge

      Re: Total, complete & utter bullshit.

      @jake - I wouldn't go so far as to say "complete and utter bullshit", since clearly the people writing the "Cool?" paper seemed quite convinced with their results. As another commenter posted, there are maybe some situations where the effect works and others where it doesn't.

      Anyway, this IS bloody curious, so I've decided to try myself. There's plenty of space in the office freezer, and 2 identical glasses just went into it, one with room temp (approx 20C) water and teh other with water from the coffee machine (approx 80C). Will let you know how it goes

      1. Benjol

        Re: Total, complete & utter bullshit.

        At the same time? Couldn't that screw things up a bit?

    5. Martin Budden Silver badge

      Re: Total, complete & utter bullshit.

      To paraphrase the student's original teacher:

      Well, all I can say is that that is Jake's physics and not the universal physics.

    6. PastorB

      Re: Total, complete & utter bullshit.

      When I was in school this issue was used as an illustration of the problems with the way Aristotle did science. It was said that he wasn't an experimentalist but rather just thought out his experiments and decided that his thoughts were right. Well, he "thought" that hot water cooled faster than cool water so he said that hot water would freeze faster than cold water. Sounds reasonable...until you really think it through.

      BUT...

      Then several (ten?) years ago I read an article in either Science News or Chemical and Engineering News that showed, when conditions are as they would have been if Aristotle was doing the experiment, that Aristotle was right. Hot (boiling) water DOES freeze faster than cold water.

      Here are the conditions:

      What would Aristotle have used to hold the water? Likely a wooden bucket, certainly not a glass vessel and probably not a clay pot (though that's possible if you heat the water in the pot), and probably not a metal container either (you'd probably use two identical wooden buckets and pour hot water into one and cold water to the same level in another - or use a third pot to get really identical amounts).

      So, boiling water in a wooden bucket. What happens? You lose a considerable amount of the water through evaporation. The wood insulates the water and keeps it warm which keeps it evaporating. By the time the water reaches the temperature of the cool water it has a LOT less mass than the water that initially was cool. From there out the "warm" water gets colder and freezes faster because it has less heat to lose (due to its lesser mass).

      So...Aristotle was right - if we use his same equipment.

      But if we start with glass or metal containers they transmit the heat much faster and the hot water doesn't have enough time to evaporate much, which means that it takes longer to reach the original temperature of the cold water, which means that it will never catch up. Using our methods cold water does freeze faster than the same amount of hot water.

      THAT'S why reproducible experimental methods are so important in science.

      I thought that was a pretty cool article! What do you think?

      --Michael

  3. Paul

    a simple thought experiment

    Take a litre of water at room temp, 20C ish and two identical vessels.

    Take two identical freezers and add internal camera.

    Share the water two ways.warm one up to 60C. If any evaporation took place, add a small amount from the other to balance them again.

    Put the vessel one into each freezer. Immediately press the boost button (did I mention that?) To override any cycling off the thermostat.

    Surely the vessel that started at 60C will have to cool to 20C and meanwhile the one which was at 20C will now be colder. Logically the one which started hotter will always be chasing the cooler one.

    1. jake Silver badge

      Re: a simple thought experiment

      You don't even have to go that far. One ice cube tray with boiling water, one with room temperature tap water, side-by-side on the same shelf of the freezer.

      The tap water tray freezes faster every time. Seriously, try it.

      1. Anonymous Coward
        Anonymous Coward

        Re: a simple thought experiment

        "The tap water tray freezes faster every time."

        What effect does the heated tray have on the experiment? It's not just the water you're freezing now.

      2. David Dawson

        Re: a simple thought experiment

        You've missed the point of this a little.

        The experiment is like this.

        Sample 1, starts at 20c. Put it in a freezer, time how long it takes to freeze, that's result A.

        Sample 2, starts at 60, put it in a freezer. Time from the moment it hits 20c until it freezes, that's result B.

        You would expect them to be the same, being the time taken to freeze water from 20c, but it isn't. B < A.

        Water that starts warmer will take a shorter time to freeze from a given temperature than water that starts at that same given temperature.

        1. jake Silver badge

          @David Dawson (was: Re: a simple thought experiment)

          Your equipment also has delta-h ...

          Again, try it for yourself. It ain't exactly rocket science.

        2. monkeyfish

          Re: a simple thought experiment

          Your 60°C sample is coming from the hot water tap? Your heating system puts a great deal crap from the pipes into the water (which is why you shouldn't drink it), what affect does that have? You'd have to filter and boil both samples from cold in order to be sure you had the same mineral content.

          1. James Micallef Silver badge

            Re: a simple thought experiment

            @monkeyfish - I don't really have the time / equipment to do a more rigorous experiment so surely there are many small details that can be improved. However the cold water is coming from a filtration system and the hot water is coming from the coffee machine that uses the same filtration system as source, so no crap in the hot sample. Coffee Machine is cleaned very regularly as well so there should be the same mineral content in both samples, though I really don't have the equipment to test this

    2. Anonymous Coward
      Anonymous Coward

      Re: a simple thought experiment

      Uhm, right.

      Now, do you see that rabbit and that turtle at the starting line over there?

    3. James Micallef Silver badge

      Re: a simple thought experiment

      "Take two identical freezers and add internal camera" - you forgot to add a flashlight :)

      "Surely the vessel that started at 60C will have to cool to 20C and meanwhile the one which was at 20C will now be colder. Logically the one which started hotter will always be chasing the cooler one."

      If you read the 'Cool?' paper, they suggest a possible mechanism - The water loses most energy from the top surface, and the rate of loss depends on temp difference between water surface and freezer. The hot water is not at a uniform X degrees because there's a convection current that cycles the water. It's possible that 2 beakers with water both have the same average temp of 20 degrees, but the one that went into the freezer at 20 degrees is uniform (so 20 degrees at surface) while the one that started at 60 has a surface temperature of 40 and is mostly 15 elsewhere (so with 40 degrees at surface loses temperature more quickly)

      Anyway, my experiment is currently in the freezer, will see how it goes

      1. James Micallef Silver badge

        Re: a simple real experiment - results

        Interesting stuff. After about 1 hr the 'hot' sample had already formed a film of crystallization on the surface, the 'cold' sample hadn't. However after 1.5 hours the 'cold' sample had frozen more than the 'hot' one.

        One possibility is that I used very small amounts (about 1/4 glass) and maybe in this case the heat loss through the sides is a larger component than surface heat loss (same could be said of jake's ice cubes)

        The experiments described in 'Cool?' paper used wider beakers that are more fully filled, so maybe I'll give it another go with full glasses.

        Is anyone else trying this??

        1. chr0m4t1c

          Re: a simple real experiment - results

          >Is anyone else trying this??

          I just did a quick calculation and I think my life's too short.

        2. BlueGreen

          Re: a simple real experiment - results

          This time I'm behind jake. Asking people to do it themselves is entirely fair, even if you can't get lab controlled conditions.

          Good stuff James Micallef, thanks.

      2. arctic_haze

        Re: a simple thought experiment

        I also believed the convection explanation. However if the paper is right that the surface is warmer than the bottom (Fig. 1c) then convection in the vessel is impossible.

    4. Ken Hagan Gold badge

      Re: a simple thought experiment

      "Logically the one which started hotter will always be chasing the cooler one."

      As I understand it (and I'm only reading the article, not the paper) their hypothesis is that hotter water is effectively a different material than colder water (hysteresis at work). So if the rate of cooling both samples is faster than the rate at which the hotter material "relaxes" into the cooler material, the hotter is not the same stuff even when it reaches the starting temperature of the cooler sample and therefore is not obliged to follow the same cooling curve.

  4. John Smith 19 Gold badge
    Unhappy

    Using logic and not checking against observation was the medieval way to do physics.

    It took several hundred years to break people of that habit.

    BTW Crystallization experiments for this have to be done very carefully. The slightest difference between the surface conditions on the 2 containers that the water contacts will probably invalidate the results.

    1. oolor

      Re: Using logic and not checking against observation was the medieval way to do physics.

      Or a drop of sweat.

      http://en.wikipedia.org/wiki/Polywater

    2. Nial

      Re: Using logic and not checking against observation was the medieval way to do physics.

      " Using logic and not checking against observation was the medieval way to do physics.

      It took several hundred years to break people of that habit."

      There's still a whole "field" of science working that way.

      :(

  5. J.G.Harston Silver badge
    Boffin

    People have repeatedly *demonstrated* the effect, this paper attempts to *explain* the effect.

    1. James Hughes 1

      But why do they bother? Jake has already proven is a load of bollocks, they just need to listen to him. Well, someone needs to listen to him...someone...please listen.....please....

  6. Anonymous Coward
    Anonymous Coward

    My A-level chemistry teacher always said "If you don't know the answer to a chemistry question, just write 'It happens because of Hydrogen bonding' and you'll probably be right."

  7. jackcaesar5

    Tests with San Mig prove the TC&UBS theory

    Maybe it's the tropics, maybe my thirst is exponentially proportional to the time to get a coldie down my throat, but then again maybe it does work for water...

  8. Alan Brown Silver badge

    Nothing to do with

    Hot water containing almost no dissolved air, so being a better thermal conductor?

    (BTW this is why water-cooled PC systems should be filled with water which has been boiled, then cooled - no dissolved air to bubble out in awkward parts of the loop)

    1. M Gale

      Re: Nothing to do with

      (BTW this is why water-cooled PC systems should be filled with water which has been boiled, then cooled - no dissolved air to bubble out in awkward parts of the loop)

      Dunno about yourself, but I use ethylene glycol. Antifreeze, basically. Cheap, widely available, nowhere near as corrosive as water, and tends to already come coloured in a nice shade of blue. You could use food-grade propylene glycol if for some reason you think someone might try to drink the coolant. The term "water cooled" is a misnomer when applied to PC liquid cooling systems. Using water (even the deionized sort) inamongst copper, aluminium and whatever differing metals are in a liquid cooling system is asking for trouble. Or perhaps a practical demonstration of the "sacrificial anode" concept.

      See also this link. It may save you a lot of bother and no insignificant amount of money!

      1. Swarthy

        Re: Nothing to do with

        Not to mention that ethylene glycol is usually rendered black-light active, and UV-A LEDs are not expensive.. could make your liquid-cooled PC quite visually stimulating.

  9. davcefai
    Pint

    The original story

    I am old enough to remember when this started. (That's not a brag, more of a rueful comment.)

    New Scientist carried the story in the early 70's following a letter from the teacher involved.

    Mpemba was late to class, when they were making ice cream. The process involved boiling the milk. After the cooling down period Mpemba's batch was noticably hotter than that of his classmates but it got put in the freezer anyway.

    To everybody's surprise it was the first to harden. The beauty of this phenomenon is that every time the subject is raised the reactions are similar to these. My mother's comment when I mentioned the Mpemba effect was "Everybody knows that it happens that way"

    Sadly I have never got around the testing the phenomenon in practice.

  10. David Pollard

    Jokers!

    Water doesn't have a "relaxation time" of the order of ten to one hundred minutes as these jokers claim. If it did then then measurements of specific heat would be all over the place, depending on the recent history of the test sample. In fact it's known very accurately.

    http://www.kayelaby.npl.co.uk/general_physics/2_3/2_3_6.html

    The experimental realisation of the "relaxation time" to which they refer is more likely to be achieved in an appropriately warm bath, with salts and perfumes added according to personal taste.

  11. harmjschoonhoven
    Boffin

    Explanation of the Mpemba effect

    There is an alternative explanation for the Mpemba effect.

    The thermal conductivity of water ice is two orders of magnitude smaller than that of a typical metal. So if the hot sample is capable of melting a frost layer on the cooling surface of even a few hundred microns (which is almost always there) and the cold sample remains isolated by a frost layer, the result may be the Mpemba effect.

    1. Tom 13

      Re: Explanation of the Mpemba effect

      Nope, they covered that when the second year university students did the experiment. They put a thermal insulator between the beakers and the bottom of the freezer so the hot beaker would still be isolated from the metallic bottom of the freezer. Effect was the same.

      As for me, I'm believing the second year uni students over jake.

  12. Johnny Canuck

    Haven't read the paper but a plausible explanation occurs to me. When you place two containers of differing temperature liquids in the freezer, the cooler one will gain heat energy from the energy the hotter one is losing, and may actually rise in temperature ( or at least cool at a slower rate). This condition will persist until the two containers reach an equilibrium. At that point the one that started out hot will cool faster because it has lost more mass due to a higher evaporation rate.

  13. Malc

    Does this mean?

    That frozen meals would heat faster?

  14. Jtom

    What this shows is that even the simplest sounding experiment is fraught with possible sources of error, and can be done many different ways. For the first run, I would take a glass beaker, pour distiiled, room temperature water in it, place an RF thermometer in it, record the temp, put it in the freezer and time how long it took for the thermometer to signal it was 0 degrees C. For the second run, I would then let the water thaw, microwave it for one minute, put the thermometer back in, record the temp, then put it in the freezer and time how long it took to reach the initial temp of the first run, and how long it took from that point to another zero reading.

    So I would be using the same water, container, thermometer, and freezer for both runs. This would eliminate any variability you might otherwise get using two separate water samples and containers, and determining when the ice started forming.

    No doubt you can find a loose variable in my methodology as well, but controlling all the variables is very different, and why different results are often obtained for any given reasearch.

  15. Osgard Leach

    Isn't this just wonderful, wonderful stuff?

    Here we are knocking on the door of quantum computing, contained fusion, Human cloning, the colonisation of Mars....

    But Ice Cubes? That's a tricky one.

    Almost makes you nostalgic for religion; at least you knew where you stood.

    1. Michael Wojcik Silver badge

      But Ice Cubes? That's a tricky one.

      Right up there with bicycles. (And magnets, if you're a Juggalo.)

      1. Vociferous

        > And magnets, if you're a Juggalo.

        Even if you're not. Quantum theory of magnetism is pretty hairy stuff. Virtual photons, anyone?

This topic is closed for new posts.

Other stories you might like