A New Form of Water: Melting ice turns oddly dense

It’s common knowledge that liquid water expands when it forms ice. Conversely, frozen water compacts as it melts. Now, a team of European researchers has made an ultrathin film of supercooled water that’s much denser than normal water.

The experimenters suspect that they have created the first sample of a previously hypothesized form—or phase—of water known as high-density liquid water, or HDL.

“We do not have complete proof that we’ve found this phase, but we think that it looks very promising,” says team member Simon Engemann of the Max Planck Institute for Metals Research in Stuttgart, Germany.

The results are “the biggest breakthrough in experimental measurements of water in a long time,” comments physicist H. Eugene Stanley of Boston University. It’s “the first discovery of water that’s significantly denser than ordinary water. That’s remarkable, whatever it means,” he adds.

Ordinary water is known to have bizarre traits, of which the typical modest rise in density when ice melts is just one. For more than a decade, Stanley and other scientists have supported a theory of water’s microstructure that they say explains many of water’s odd properties (SN: 1/24/04, p. 58: Available to subscribers at Wet ‘n’ Wild).

According to that theory, the molecules in liquid water are perpetually rearranging the liquid’s microstructure. In a frenetic dance, transient patches of HDL mingle with equally transient patches of another phase of water known as low-density water, or LDL. In picoseconds, any particular patch of water may bounce between HDL and LDL structures, says Harald Reichert, who led the new experiment and is also at the Max Planck Institute in Stuttgart.

The team first crystallized ice onto the oxidized surface of a silicon block about the size of a thin matchbox. While slowly heating the structure, starting from –25°C, the scientists bounced high-energy X rays off the ice-oxide boundary.

By detecting interference patterns in the reflected radiation, the team determined the molecular spacing in the ice, silicon-dioxide, and silicon layers.

The team reports that around –17°C, a dense layer of water starts to form at the ice-oxide interface. By the time the sample reaches just below 0°C, the water layer is more than 5 nanometers thick. As the layer forms, the once-firmly-attached ice can slide off the silicon block, indicating that the film is liquid, Reichert notes.

The researchers have compared the new layer’s measured density—17 percent greater than that of ordinary liquid water at similar temperature and pressure—to the densities that have been experimentally determined for various phases of water and ice, Engemann says. The layer’s density best matches that of a disorderly form of ice considered to be the proposed HDL’s close cousin, he adds. However, the team intends further experiments to determine whether the film has HDL’s proposed structure, Reichert says.

Reichert, Engemann, and their colleagues describe the new work in an upcoming Physical Review Letters.

Besides illuminating the nature of water, the new work deepens scientists’ understanding of melting, comments Johannes Friso van der Veen of the Paul Scherrer Institute in Villigen, Switzerland. Until now, experiments have failed to reveal whether a substrate on which ice rests—for instance, rock or soil—can instigate subzero melting at the ice-substrate boundary. Such melting might affect, for instance, how glaciers move. The new experiment offers convincing evidence that such melting can occur, van der Veen says.

More Stories from Science News on Physics

From the Nature Index

Paid Content