Study: Ice forms a perfect crystal, becomes ferroelectric

Chemists at Ohio State University and their colleagues may have settled a 70-year-old scientific debate on the fundamental nature of ice.

A new statistical analysis mechanical theory has confirmed what some scientists only suspected before: that under the right conditions, molecules of water can freeze together in just the right way to form a perfect crystal. And once frozen, that ice can be manipulated by electric fields in the same way that magnets respond to magnetic fields.

Though ice may never become a practical component in electronics, this new finding could greatly enhance scientists’ understanding of how naturally occurring particles of ice interact with the environment, explained Sherwin Singer, professor of chemistry at Ohio State.

Singer’s former doctoral student Jer-Lai Kuo presented these results in a poster session Wednesday at the meeting of the American Chemical Society in Anaheim.

The results hold broad implications, Singer explained. From the coldest depths of the ocean to the upper reaches of the atmosphere, important chemical reactions — such as those that lead to ozone depletion — occur on the surface of ice crystals.

“The way water molecules are organized at the crystal surface is absolutely critical to these reactions,” he said. “If the molecules inside the crystal are ordered in a certain way, that may affect how the molecules on surface are also ordered. So anything that helps us understand how ice forms a particular structure can help us understand these reactions.”

Scientists have struggled for decades to explain why ice freezes the way it does. In 1935, chemist Linus Pauling predicted that the water molecules lock together in a disordered fashion. Within a year, Pauling’s calculations on the disordered nature of ice were confirmed by experiments.

That means, in fact, that normal ice is not truly a crystal at all.

“When we think of a crystal, we might picture a diamond, a beautifully faceted stone from our rock collection, a snowflake, or ice cube. But to a scientist, calling a material a ‘crystal’ means that the atoms form a regular, repeating arrangement like eggs in an egg carton,” Singer said. “Under the scientific definition, the ice in a snowflake or ice cube is actually not a perfect crystal.”

Ever since Pauling’s time, scientists have argued over whether ice could ever form a perfect crystal.

“Exotic forms of ice have been created in the laboratory at extremely low temperatures and high pressures,” Singer said, “and still, the ordinary ice of snowflakes and ice cubes has remained a puzzle,” he said.

The puzzle involves calculating the energy of all the possible orientations of water molecules in the ice. Each molecule has one atom of oxygen and two of hydrogen, and while the oxygen atoms line up perfectly, the hydrogen atoms can point in any of six different directions. To predict the structure of even a small ice crystal would require calculating the energy of billions of different orientations of the water molecules — a task so cumbersome that many scientists thought it could not be done.

The key to solving the puzzle came when Singer and Kuo applied a mathematical technique to simplify the analysis. They used functions called graph invariants that relate symmetries among the molecules to other physical properties, such as energy.

They attempted to calculate the structure of one of the exotic forms of ice made in the laboratory. The ice, known as Ice-XI, or “ice-eleven,” was first created at Osaka University in Japan in the 1980s.

Even though their strategy of using graph invariants simplified their calculations, Singer and his colleagues still spent a year crunching the numbers.

To check whether their calculations were correct, they compared the predicted phase transition temperature of their crystal structure to the temperature measured in the Osaka lab. The predicted temperature was 70 Kelvin (colder than negative 330 degrees Fahrenheit), and the measured temperature was 72 Kelvin.

According to Singer’s predicted structure, the molecules in ice-eleven are perfectly ordered, with each subsection of the crystal, or crystallite, having one end with a positive electric charge and one end with a negative charge. That means the ice is ferroelectric and can be manipulated with an electric field.

Because the tiny electrical charges of water molecules can affect chemical reactions taking place on the surface of an ice crystal, one of the chemists’ next steps is to use the knowledge gained so far to understand the surface of ice particles found in the Earth’s atmosphere, the site of important chemical reactions leading to depletion of the ozone layer.

Coauthors on the presentation included Michael Klein of the University of Pennsylvania and Lars Ojamäe of Linköping University in Sweden. Kuo is now a postdoctoral researcher at the University of Pennsylvania.

The National Science Foundation and internal funds at Ohio State University supported this project. Some of the calculations were performed at the Ohio Supercomputer Center.

Contact: Sherwin Singer, (614) 292-8909; Singer.2@osu.edu
Written by Pam Frost Gorder, (614) 292-9475; Gorder.1@osu.edu