Bringing water down to a new critical point.

For a liquid so common and intensively studied, water can still spring a surprise or two. Indeed, further evidence of the astonishing quirkiness that distinguishes water from virtually all other liquids has now surfaced in computer simulations of the behavior of water molecules cooled to temperatures well below water's normal freezing point.

These simulations indicate that supercooled water exhibits trends in behavior that suggest the existence of a previously unsuspected critical point. Such a point corresponds to a specific combination of temperature, pressure, and density at which two forms of a substance in equilibrium with each other become identical, forming a single phase.

For water, one critical point occurs at a temperature of 647 kelvins (374 [degrees] C) and a pressure of 22.1 megapascals (218 atmospheres). At higher temperatures and pressures, water's liquid and vapor phases become indistinguishable.

"What we have found in our computer model of water is a second critical point," says physicist H. Eugene Stanley of Boston University. This one appears to mark the temperature-pressure combination at which two distinct forms of ice blend into a single phase. Stanley, Peter H. Poole, and co-workers describe their findings in the Nov. 26 NATURE.

What makes the results even more startling is that the two types of ice involved are amorphous -- ordinary water solidified into disordered, noncrystalline forms. The new findings suggest a link between these two types -- high-density and low-density amorphous ice--and supercooled water.

"I have yet to hear of another material in which two solid phases could have a critical point -- another plus for water," remarks Christopher M. Sorensen of Kansas State University in Manhattan.

Poole and his colleagues simulated the behavior of 216 particles representing water molecules enclosed in a cubic box at temperatures below water's usual freezing point. They monitored the strength of hydrogen bonds, created by the attractive force between a hydrogen atom in one water molecule and the oxygen atom of another (SN: 4/14/90, p.231; 11/30/91, p.359).

These simulations allowed the researchers to explore the behavior of supercooled water over a temperature range impossible for experiments to reach. This also meant there is no direct way of experimentally checking their results.

Nonetheless, Sorensen notes, "although never attainable in a real experiment, this distant critical point would send out its influence into the attainable regimes, perturbing the thermodynamic landscape and creating a most remarkable liquid."