We recently received the following email:
Is it true that hot water freezes faster than cold water? It seems as if this could only be true in a small amount of circumstances where the water temps were already pretty similar to begin with and the volumes of water were pretty small. The only reason I can come up with that would make this true is that the faster the water molecules are moving, the faster they can get into their ice crystal formation. Any insights? Thanks Skeptics!
D.J Rhode Island
Thanks for the question D.J.
As with many topics in science, this one got more and more complicated the more I delved into it.
My knee-jerk answer to this question is of course, no, how can hot water freeze faster than cold? The hot water takes time to cool down to the cold water’s initial temperature. In that time, the cold water would of course get even colder. This process would continue until the colder water turned to ice well ahead of the hot water…..right???
Actually, there have been experiments carried out that show that under certain conditions warm water can freeze before cold water. That seems undeniable. This effect even has a name…The Mpemba effect.
It was apparently discovered by a student many years ago during an ice-cream-making project in school (I wish I had projects like that in school). The project instructions called for a waiting period before putting the warm milky recipe in the freezer. Mpemba was falling behind and was anxious to catch-up so he placed his warm proto-ice-cream in the freezer before it had cooled down. To his surprise, it became ice-cream before anyone else’s even though their recipes had cooled down and were already in the freezers.
You might think that if experiments show that hotter water can freeze faster than cold, then we’re done here right?
Well…kind of. The question had been answered, yes, but it would be nice to reproduce this effect reliably and know exactly why this happens. That’s where we run into problems.
This is because it is such hard thing to test when you think about it.
First off…define hot water and cold water. Different experiments use different temperatures and that can make a difference.
When does freezing occur? Is it when the first ice crystals form or when it’s all solid ice or somewhere in between?
The real difficulty lies in the fact that so many parameters affect how soon ice can form:
evaporation rate, water volume, initial temperature, shape of container, environment around the container, type of water, gas content of the water, convection within the water, random nucleation sites like dust…the list goes on and on
Evaporation is a good example. The hotter the water, the more it evaporates and the less water you need to freeze. If enough of the water evaporates away, then the remaining water can freeze faster than colder water because there’s so much less water to freeze. There’s also the idea that the act of evaporation pulls heats from the water which is another way to accelerate the cooling process.
Evaporation is only a contributor though and some experiments have shown that it’s not even required for this Mpemba effect to happen.
To really test all variables I’ve mentioned to determine their relative contributions and how they interact would require a huge multi-dimensional array of experiments that I’m sure would frighten even the most experienced experimentalists.
The bottom line then appears to be that attempts to definitively clarify the Mpemba effect would demand too much effort for what is likely to be very little return.
Return on Investment can be an annoying concept sometimes.