A US physicist believes he may have unraveled the mystery behind why warm water sometimes freezes faster than cold.
James Brownridge of Binghamton University in New York, believes it comes down to water's variable, and fragile, freezing point when water becomes ice.
His finding appear on the pre-press website arxiv.
The Mpemba effect, named after a Tanzanian school boy who noticed it occurring as ice cream froze, is used to explain why warm water sometimes freezes faster than cold.
Despite its widespread observation, no one has yet been able to explain why.
"If all conditions except [temperature] are equal, then hot water cannot cool to 0°C before cooler water. That would violate the laws of thermodynamics," says Brownridge.
Rates of cooling
He says the temperature at which ice begins to form in water is around 0°C, but can vary by a few degrees depending on the impurities in the water.
"If two water specimens have a relatively large difference in spontaneous freezing temperatures and the one with the highest spontaneous freezing temperature is the warmest when cooling begins, it will more often than not freeze first," says Brownridge.
Professor Joe Wolfe, a physicist at the University of New South Wales says those few degrees can make a big difference.
He says the rate at which water cools "decreases rapidly as the sample temperature approaches that of its surroundings", allowing warmer water to catch up to cooler water in a relatively short time.
To test whether heating water up could speed up the freezing process, Brownridge took six samples of water and tested their spontaneous freezing temperature.
He then raised their temperature to 100°C for two hours and measured their spontaneous freezing temperature again.
Brownridge found that after heating, some of the samples froze at higher temperatures, and others froze at lower temperatures.
He says the formation of ice is usually seeded by impurities in the water called nucleation sites, and believes heating the water samples may create or destroy these nucleation sites, thereby changing the spontaneous freezing temperature.
Wolfe says research into freezing, "is of enormous importance in agriculture, food preparation and preservation of blood and other organs", and people in those areas might be interested in Brownridge's results.
But he warns the paper is yet to undergo peer review and people may be "reluctant to take this paper seriously until it has been reviewed by experts."