Hot water can sometimes cool faster than warm water – SFU experiment confirms

By Diane Mar-Nicolle

Scientists have long been puzzled about why hot water can sometimes freeze more quickly than warm water. After all, we know that a hot object, left to its own, will cool and eventually reach the temperature of its surroundings. So logically, it would seem that the hotter the object is at the beginning, the longer it would take to cool. And it seems this should also apply to hot water.   

But this is not always true says SFU physics professor John Bechhoefer.

In a world-wide first, Bechhoefer and grad student Avinash Kumar created and controlled the phenomenon in the lab, leading to a better understanding of why it happens.

The peculiarity that hot water can sometimes cool and start to freeze faster than warm water is called the Mpemba effect, named after Tanzanian teen Erasto Mpemba who performed the first systematic, scientific experiments on this effect in the 1960s.

Until now, the effect has been difficult to prove, partly because of the long time it takes to cool large volumes of water and because water is a much more complicated system than it might at first seem.

But Bechhoefer and Kumar devised a way to speed up the cooling process by inserting a microscopic glass bead in water and then subjecting it to carefully designed forces as well as random thermal forces from surrounding water molecules. 

Bechhoefer credits their system of using the microscopic bead, which cools in less than 1/10th of a second, for allowing them to conduct the large number of trials needed to understand the effect.

The pair determined that, based on the variation of forces on the bead in the water, some things can cool much faster than normal if the “landscape” of these forces is “properly” designed.

These design principles include shaping the free-energy landscape so that hot systems have a more direct path to “cooler” states.

Bechhoefer says this work does not explain all the particulars of why hot water placed in a freezer can start to freeze more quickly than cold, but it does provide important insight into this curious phenomenon and suggests that analogues could exist in many other settings and materials.

The paper is published in Nature.