Skip to Content
Uncategorized

Chemists Discover Freezing Point of Supercooled Water

Scientists have long known that water can stay liquid at temperatures well below zero. Now they’ve discovered exactly how low they can go

It’s easy to imagine that water must be one of best understood materials in science. After all, this liquid is possibly the best studied substance on Earth. But the truth is that many of its properties still mystify scientists.

One unsolved puzzle is its freezing point. Scientists have known for many years that you can cool liquid water well below zero degrees centigrade without it freezing. That’s because water needs some nucleation event to trigger the process of ice formation. Without ice nucleation, it remains liquid.

But how low can you go?

Today, we have an answer of sorts thanks to the work of Emily Moore and Valeria Molinero at the University of Utah in Salt Lake City.

Part of the problem is that experiments to measure the freezing temperature are so difficult to perform that nobody has managed them. But the evidence points to the likelihood that ice crystals begin to form anyway at temperatures of about -41 C.

Supercoooled water should freeze at around this temperature but nobody has succeeded in measuring it because it always begins to freeze earlier.

Moore and Molinero get around this problem by simulating the freezing behaviour of over 250,000 water molecules on a computer. What they find is that once the natural process of ice formation begins to occur, then water cannot stay liquid at much lower temperatures.

In fact, their simulation indicates that the natural freezing point of supercooled water is about -43 C, just below the temperature at which ice crystals form naturally. That’s as expected but the simulation also gives new insights into the way in which this freezing occurs.

In this state, water is a mixture of low density ice and water molecules that are on the verge of becoming ice, what chemists call “four co-ordinated” meaning that each molecule is linked to four others. The structure of “four co-ordinated” water seems to have important impact on the rate at which ice can form and this is what determines the freezing point.

There is an important caveat, however. The simulations require a major correction before they produce a physically realistic result. For some reason, they suggest that the natural ice formation begins to occur at about -71 C and that supercooled water freezes at about -73 C.

That’s 30 degrees lower than in the real world. To get around this, Moore and Molinero simply add 30 degrees to all their results. Just why the simulation is out by so much isn’t clear.

If the work is valid, however, it could have a major impact in other areas of science.

The temperature at which supercooled water freezes is an important factor in cloud formation. And small changes in this process, when entered into in climate change models, can have a big impact on the predictions about the future of the Earth.

Exactly how the new numbers will change climate predictions isn’t yet clear. And of course, climatologists will want better evidence than a slightly wonky computer simulation. But it’s a decent step forward and worth keeping on eye on for its influence elsewhere in science.

Ref: arxiv.org/abs/1107.1622: Structural Transformation In Supercooled Water Controls The Crystallization Rate Of Ice

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.