杏吧原创

鈥淲ATER is very strange,鈥� says Anders Nilsson. He should know: he has been studying the stuff for most of his working life. His claim may be hard for the rest of us to swallow 鈥� after all, what could be more ordinary than water? Its behaviour is so familiar, its appearance so commonplace, that we are tricked into assuming that it is more or less the same as everything else. But water is uniquely weird. If it weren鈥檛, none of us would be here to notice.

For example, if water weren鈥檛 densest at around 4掳C rather than as ice, lakes and rivers would freeze from the bottom up, slowly killing their inhabitants. If it weren鈥檛 so spectacularly good at absorbing heat, the planet would have boiled over long ago. And if its molecules, barrelling through membranes or darting down veins, weren鈥檛 so good at sweeping other chemicals along, plants and animals would die of malnutrition.

杏吧原创s have been plumbing the depths of water鈥檚 strangeness since at least the time of Galileo, to no avail. But now, thanks to the work of Nilsson and others, we might be on the verge of understanding why it behaves the way it does. Their explanation is as strange and wonderful as the stuff itself: water isn鈥檛 one liquid, but two.

On one level, it鈥檚 no surprise that water comes in multiple forms. It exists in three phases, as a solid, liquid or gas, depending on the temperature and pressure where you find it. At sea level, water turns to steam at 100掳C, but at altitude, where the atmospheric pressure is reduced, you can get it to boil at much lower temperatures 鈥� saving you time but ruining your tea.

For the most part, liquid and gas phases are distinct. Amplify temperature and pressure, however, and that starts to change. As a gas becomes ever more compressed, it starts behaving more like a liquid. And as liquid heats up, it wants to behave like a gas. Dial the conditions up enough 鈥� past a delicate equilibrium known as the critical point 鈥� and it becomes impossible to tell which one you are looking at. Let either the temperature or pressure drop a hair, though, and you will find yourself with either one or the other.

鈥淲ater may be commonplace but it is uniquely weird鈥�

And this is where the story of two liquids starts. Ignoring pressure for the moment, almost all substances have a high-temperature critical point where their gas and liquid phases converge, but a handful of materials display a mysterious second critical point at low temperatures. For instance, if you cool liquid silicon and germanium under the right conditions, they can be turned into one of two different liquids of different densities. Both consist of the same atoms, but in alternative configurations that give them their respective properties.

Peculiar though they are, these second critical points were of little more than academic interest. If you didn鈥檛 specialise in liquid silicon, the subject probably didn鈥檛 cross your radar.

That all changed in 1992, thanks to a group of researchers at Boston University led by Peter Poole and Gene Stanley. They had been intrigued by experiments claiming that water鈥檚 density started to fluctuate at low temperatures, and that the cooler things got, the more the fluctuations increased. This ran counter to all predictions. Ordinarily, the cooler you make something, the less it fluctuates.

To probe what was happening, Poole, Stanley and their collaborators simulated water being cooled so carefully that it was able to pass its freezing point and still remain liquid, a process known as supercooling. Their computer model supported the idea that density fluctuations were indeed happening in the supercooled liquid, and got bigger as the temperature plummeted. Something funny was definitely going on.

and Stanley鈥檚 team hypothesised that this was the sign of a second critical point, where water splits into two liquids of different densities 鈥� each being a phase in its own right. Above the second critical point, as the two phases become indistinguishable, the water would start rapidly changing between the two. Given their differing properties, any instance of water metamorphosing into its alter ego would be marked by rapid changes in density, changes that peak at the critical point itself.

Their suggestion that water might have two critical points was a pretty outlandish one. 鈥淓verything that comes with water is controversial,鈥� says Nilsson. Most researchers thought that water鈥檚 properties could be explained in more conventional ways. One was that at really low temperatures, supercooled water transforms into a type of disordered solid, with none of the crystalline structure of ice. Another was that what appears to be a second critical point is actually a quirk of the water freezing.

鈥淚t took 26 years, now two teams have data 鈥� within months of each other鈥�

The team knew that if they were right about a second critical point, proving it experimentally would be fiendishly tricky, if not impossible. The region where the critical point should exist is so cold, at -45掳C, that even supercooled water should spontaneously crystallise into ice. 鈥淭he challenge is to cool water very, very, very fast,鈥� says Stanley. 鈥淪tudying it needs clever experimentalists.鈥�

Nilsson, who works at Stockholm University in Sweden, could not resist the challenge. For years, he and his team had been obsessed with the strangeness of water, and in particular with Poole and Stanley鈥檚 two-liquid picture. Nilsson鈥檚 work in the early 2000s on water at room temperature and pressure had convinced him that its molecules seemed to exist in two structural configurations: a disordered, densely packed jumble, and a regular tetrahedral layout of much lower density (see 鈥�Troubled waters鈥�). His experiments hinted that under these ambient conditions, small pockets of this low-density phase would float about in a bath of the jumbled, high-density phase. But, he says, his peers were not convinced.

Then at a series of conferences in 2008, he ventured out of the shallows of ambient research to meet colleagues with supercooled water on the brain. Nilsson quickly realised that their ideas gelled. After all, says Nilsson, water is one substance, 鈥渋t couldn鈥檛 be one description of water at ambient temperatures and then another at supercooled conditions.鈥� He saw his opportunity to unite the two. 鈥淚 was fascinated that so few experimental studies had been done and it was mostly a playground for theoreticians,鈥� he says.

In his attempts to prove that water has a second critical point, Nilsson focused on one of the key predictions made by Poole and Stanley: that the density of supercooled water would start to fluctuate as it got colder. The plan was simple: try to measure the density fluctuations, then keep changing the conditions to make them grow. Follow this trail of breadcrumbs 鈥� a path Stanley christened the Widom line after a pioneering chemist 鈥� and you will reach the critical point at the end. 鈥淥nce you have a Widom line,鈥� says , a theoretical physicist at Roma Tre University in Rome, 鈥渢here must be a critical phenomenon.鈥�

Actually doing the experiment was anything but simple. For one thing, Nilsson had to fight against water鈥檚 compulsion to freeze, which it does by rapidly building ice crystals around tiny impurities. Using , in 2017, Nilsson鈥檚 team dripped ultra-pure droplets of water into a vacuum chamber and cooled them down to -45掳C. As the droplets fell in, Nilsson measured how much their volume, and therefore their density, changed as the pressure varied.

By December, Nilsson and his colleagues had results to report. They revealed a snapshot of supercooled water in the briefest of moments before it spontaneously freezes. And, the researchers say, they captured tantalising evidence of the Widom line, the trail leading to the second critical point.

Gallo, for one, thinks Nilsson is on to something, but although researchers are largely impressed by the team鈥檚 technical feats, their conclusions about the cause of their observations leave many unconvinced. According to Alan Soper at the STFC Rutherford Appleton Laboratory in Harwell, UK, there is 鈥渃lose to zero鈥� evidence for a critical point. 鈥淭he effects seen by Nilsson are both small and have multiple possible alternative explanations, only one of which is that there is a second critical point somewhere nearby.鈥�

Rich Saykally at the University of California, Berkeley, is another sceptic. 鈥淚t is going to take more evidence than even these beautiful new results to convince most experts,鈥� he says. Like Poole and Stanley鈥檚 critics, Saykally says that the results might be explained by quirks in the way water freezes at low temperatures.

But in March 2018, another line of evidence suggested there might be more to it than that. In an experiment very different to Nilsson鈥檚, Sander Woutersen and his team at the University of Amsterdam tried another way to stop water freezing as it cooled: antifreeze. Their version of driveway salt was a molecule with a similar structure to water, chosen specifically to dissolve and blend in with the surrounding liquid. As the mixture cooled, its density did seem to change abruptly. Because their solution was so structurally similar to pure water, they say the result suggests that water does have a second critical point. 鈥淭his is beautiful work,鈥� says at Princeton University. He agrees it helps shore up the idea of two coexisting phases of water, but remains cautious about proclaiming it to be the final word in the story.

It may have taken 26 years, but two teams of canny experimentalists now have hard data consistent with the predictions Poole and Stanley made 鈥� and within months of each other. 鈥淪cience is funny like that,鈥� says Stanley.

Life-giving liquid

If the theory is right, the consequences aren鈥檛 limited to the exotic forms of water that Nilsson and Woutersen have created in the lab. 鈥淎ll these discussions have implications up to room temperature, and for life,鈥� says Nilsson. Debenedetti agrees. 鈥淭he effects can be seen quite far away, not just at the critical point,鈥� he says. 鈥淲ater is the fluid it is, and the properties it has at ambient conditions do reflect the properties when supercooled.鈥�

For Martin Chaplin at London South Bank University, the two structures apparent at low temperatures are particularly good at explaining the unusual behaviour of water that we see at room temperature and pressure. The fact that liquid water below 4掳C is denser than ice, for example, comes about because heating pushes the water into a less ordered, higher-density structure. Water鈥檚 talent for absorbing heat results from the energy going into converting molecules from one arrangement to the other. And water鈥檚 ability to diffuse more easily under pressure 鈥� as in the body 鈥� arises from the increased mobility of the disordered structure that dominates at high pressure.

For Stanley, given how essential the stuff is to our existence, the wild density fluctuations of water near the second critical point might even be a factor in helping life emerge. Debenedetti is nudging towards testing this hypothesis. He is using computer models to map out how proteins behave over a wide range of temperatures and pressures, with a specific focus on the apparent second critical point. Debenedetti says he is most interested in how proteins interact with liquid water in both of its guises 鈥� high and low-density 鈥� under these extreme conditions.

First, however, the exact nature of these guises will need to be confirmed. And coming up later this month is an experiment that may definitively settle the question. Nilsson plans to return to South Korea for one final push towards the second critical point, forging further along the Widom line than anyone before. If he succeeds, the mysteries of water鈥檚 weirdness may start to be cleared up 鈥� even if the liquid itself becomes stranger than ever.

This article appeared in print under the headline 鈥淗2Oh!鈥�