Friday, 19 April 2013

Why does hot water freeze faster than cold water?

This article is by our intern Freya Leask, who is also in her second year at the University of Bradford studying chemistry.

For most people, making ice-cream doesn't lead to the discovery of an unsolved mystery…unless you're Erasto Mpemba. In 1963, the then school boy stumbled across the phenomenon where initially-hot liquids sometimes freeze faster than initially-cold ones. Although this had been observed by Aristotle, this effect wasn't proved experimentally until 1969, and still isn't very well understood. Just what is the Mpemba effect, and why does it happen?

Many papers have been written about the Mpemba effect, though scientists can't even decide what that name refers to - is it the time taken to form a homogenous block of ice, or the time taken to reach 0°C? Both situations have been studied under various combinations of conditions, but either way, it seems simple enough on face value to explain. Evaporation takes in heat energy and the warmer liquid's higher rate of evaporation reduces the mass to be frozen. The initially warmer liquid also has a lower density, as the water molecules have more energy to move around more. This means more heat is released, and the liquid cools faster.

Graph of freezing rates of cooler and warmer water samples
Against all expectations, warmer water (red) can freeze before cooler water (blue) does.
Image credit: Pico Technology

Heating a liquid can also change its composition, which can affect its cooling time. For example, when salty water is heated, the water evaporates away, leaving a higher concentration of salt which lowers its freezing point and makes it take longer to cool1. However, none of these things can account for a big enough effect on the rate of cooling to completely explain the Mpemba effect.

There are also some more complex explanations that have been hypothesised - many involving "supercooling"2. Supercooling happens when something is cooled to below its freezing point but doesn't freeze because there is nothing for the ice crystals to form around. In 1994, it was found that, experimentally, initially-hot water supercooled less than the initially-colder water, meaning the initially-hot water froze first and at a slightly warmer temperature3. Unfortunately, this explanation simply raises more questions than answers, like why the initially-hot water supercools less.

Making ice-cream in a pan
The Mpemba effect was discovered while making ice-cream. Image credit: Antonis Achilleos
Some theories have been put forward to explain why the initially hot water super-cools less than the cold water, although none have been proven. Scientists at the University of Houston have researched the possibility that this effect might be caused by dissolved gas bubbles in the initially-hot water lowering the ability of the water4 to supercool. However, gas dissolves more readily in cold water, so why this should affect the initially-hot water only is unknown. Another group of scientists based in the University of Electronic Science and Technology of China5 have looked into how the concentration of dissolved impurities affects how water can cluster together and form a solid (ice).

Nevertheless, the Mpemba effect is a great example of how a seemingly simple phenomenon is actually harder to explain once you delve into it. So next time you tuck into some ice-cream, or enjoy an iced drink, spare a thought for the hard-working scientists trying to find out how it got that way!

References
why don't all these papers have links? 

1. Katz, Jonathan (April 2006). When hot water freezes before cold. arXiv:physics/0604224 [physics.chem-ph].
2. Auerbach, David (1995). Supercooling and the Mpemba effect: when hot water freezes quicker than cold. American Journal of Physics 63 (10): 882-885. doi:10.1119/1.18059.
3. Jeng, Monwhea (November 1998) Can hot water freeze faster than cold water. FAQ, Department of Physics, University of California
4. A.F. Heneghan, A.D.J. Haymet, Liquid-to-crystal heterogeneous nucleation: bubble accelerated nucleation of pure supercooled water, Chemical Physics Letters 368 (2003) 177–182
5. X. Pang and B. Deng, Infrared absorption spectra of pure and magnetized water at elevated temperatures, Europhys. Lett. 92 (2010) 65001

No comments:

Post a Comment