Does hot water freeze faster than cold water? The obvious answer seems to be no, because all else being equal, the hot water takes longer to cool down than cold water as it has to go through every stage the cold water does – in order to freeze. So it just doesn’t sound possible for hot water to freeze faster, right?

Well, in fact, we don’t know the answer to this question yet. If you’re curios, the experiment is pretty simple – you have got water, and most of us have a freezer, right? But believe me, it’s a lot more complicated than you might think, and this very experiment has been performed by brilliant people all over the world. But the results are not just – reproducible, because in some experiments, hot water freezes faster than cold water, and in some – it just doesn’t. So how it is possible for hot water to freeze faster in some of the experiments?

The phenomenon in which hot water freezes faster than cold water is called the Mpemba Effect. It was named after Erasto Mpemba who in 1963 was a Tanzanian student making ice cream with his classmates as a school project. The students were meant to boil a mixture of cream and sugar and then let it cool down before they put it in the freezer. But worried about not getting a spot in the freezer, Mpemba put his mixture into the freezer immediately after boiling it. After an hour and a half later, Mpemba found that his mixture had frozen while his classmates’ had not.

Intrigued by this phenomenon, he eventually enlisted the help of a physics professor Dennis Osborne to investigate the effect using cream and also water, and they managed to replicate Mpemba’s finding and published a physics paper in 1969 [PDF] showing that warm water freeze faster than cold water.

Mpemba was not the first person to observe this effect. Aristotle observed the same thing in the 4th century BCE. He noted “the fact that the water has previously been warmed contributes to its freezing quickly: for so it cools sooner. Hence many people when they want cool hot water quickly begin by putting it in the Sun.”

Sir Francis Bacon also observed the same phenomenon in his experiments. He wrote “tepid water will freeze more easily than water which is altogether cold.” Also, Rene Descartes in his essay on Meteorology states “we can also see by experiment that water which has been kept hot for a long time freezes faster than any other sort, because those of its parts which can least cease to bend evaporate while it is being heated.”

So the observation that hot water freezes faster has a long and illustrious history. So what is the physics behind this strange phenomenon? Well, Derek Muller of Sciencium explains that there are five proposed mechanisms related to this phenomenon, and they are:

1. Frost Melting: When placing two containers of water in a frosty freezer, the frost acts as a thermal insulator and so around the cold beaker, the frost won’t melt but around the warm beaker it will and this creates a conductive layer of liquid water around the base of the warm beaker dramatically increasing its cooling rate.

Derek explains that this may be the case for some observations of the Mpemba effect. However, other experimenters have claimed to observe the effect when they insulate the base of the beaker from the freezer, too.

2. Dissolved Gases: One known difference between warm and cool water is the amount of dissolved gases. There are more dissolved gases in colder water then in hot. So as warm cools down, gases are constantly dissolving into it. But that is an exothermic process so it should actually slow cooling down rather than speed it up. Researchers propose that differences in dissolved gas concentrations might play a role in cooling rates.

3. Supercooling: Water freezes at zero degree Celsius but in reality, it may become significantly colder before it hits the freezing temperature. This phenomenon is known as supercooling, and it typically happens because ice needs a nucleation site, such as an impurity or an air bubble before it can form. The idea is that the water initially warmer may experience less supercooling and so freeze before the cold water.

Although some experiments have shown that the difference in supercooling causes initially warm water to freeze faster, the reason for this in not entirely clear. The results are rather inconsistent, so it’s basically not possible to achieve same result even in the same lab at the same time. As Derek explains, freezing requires a lot of energy to be given off. Changing the state of water from liquid to solid at zero degrees Celsius takes as much as energy as cooling it down from 80 degrees Celsius to zero. So as a matter of fact, the time taken for water to cool and freeze should be dominated by the actual freezing process rather than the time taken to cool the water to zero.

4. Evaporation: It’s probably the simplest and the best one. If you start out with two beakers containing the exact same amount of water, the beaker containing the hot water will lose more water molecules through evaporation. So there’s less water left in the beaker to freeze. But the reality is that the amount of mass loss is typically less than three percent, which is not enough to account the very different cooling rates. The Mpemba Effect has been observed even in experiments using sealed containers, too.

5. Convection: Some researchers claim that the reason warm water cool faster may have something to do with convection currents.

Convection currents occur because water cools primarily from its surface and from the side of the beaker with the middle remaining warmer, so the colder water sinks and warm water rises to take its place. The beaker with warm water will contain larger convection currents and therefore experience larger temperature gradients, and this could affect cooling rates.

Back in 2012, the Royal Society of Chemistry offered the prize for the best explanation of the Mpemba effect. Although they received 22,000 entries from all over the world, none of the answers were convincing enough to draw consensus. And well, here’s Derek’s view on the Mpemba effect.

Contrary to what common sense would suggest, people have witnessed hot water freezing faster than cold water in real-world conditions for thousands of years, and a number of experiments appear to substantiate these observations. So it’d be just wrong to dismiss the Mpemba effect out of hand.

The freezing of water is a more unpredictable phenomenon than most people imagine, and frustratingly, there are so many variables. It is actually very hard to ensure everything is equal except the initial temperature of two beakers.

Also in 2016, researchers conducted their own experiments by cooling water under carefully controlled conditions. But sadly, there was evidence to support meaningful observations of the Mpemba effect, suggesting that the phenomenon of hot water freezing faster than cold water may not even exist at all. The warm water, in fact, took measurably longer than cold water to cool to zero degree Celsius. However, when they changed the placement of the thermocouple – a device for measuring temperature – by just a centimetre in the same experiments, they were able to observe a slight Mpemba effect.

Researchers reviewed all the other Mpemba studies that provided sufficient data and showed that the reported effects were within the margin of error. So there is no real evidence of the Mpemba effect. Hot water does not freeze faster than cold water.

“This is not just an esoteric finding about a quirk of nature,” Derek explains. “If the Mpemba effect really did exist, it would have significant implications for thermodynamics. It would mean that water molecules have a kind of memory of their earlier state or states and that would mean having to update thermodynamic tables to include not only what the temperature of the water is but also what it was before.”

“So maybe we should be thankful that the intended effect, at least, under carefully controlled conditions does not exist because if it did, that would make thermodynamics a lot more complicated,” he concludes.