Working in a warmer winter wonderland

Caltech's answer to Jack Frost is cooking up snow crystals in sunny California. Steve Farrar reports.

One day in 1947, it snowed in Pasadena. Startled children and bemused adults awoke to see this perennially warm corner of California plunged into unfamiliar wintry weather. Even the California Institute of Technology's academic community, so often focused on more esoteric matters, could not help but be impressed. A fading photograph outside Ken Libbrecht's laboratory records the scene. It is a sight that has not been repeated in 54 years. Yet, inside Libbrecht's laboratory, it snows all the time.

The delicate manmade crystals grown at Caltech evolve in patterns as beautiful and intricate as any produced by nature. But in this carefully controlled environment, under watchful electronic eyes, these crystals are beginning to betray secrets that have baffled some of science's most brilliant minds, including astronomer Johannes Kepler and philosopher René Descartes, the first to accurately describe snow crystal form. He concluded: "It is impossible for men to make anything so exact." Libbrecht, professor of physics at Caltech, knows otherwise.

Usually found investigating the gravity waves that are thought to ripple through space, or helping to develop cold atomic beams to study Bose-Einstein condensates, Libbrecht has, in his spare time, developed an obsession with snow crystals. It started five years ago after a conversation with a postdoctorate researcher about interesting problems in physics. Libbrecht found few had ventured into the field. Its pioneer was a Japanese physicist called Ukichiro Nakaya. Nakaya began studying snow crystals in the 1930s at Hokkaido University. He was the first to grow snow crystals under controlled conditions and made some brilliant observations. Yet, a host of questions remain unanswered - not least, how snow crystals form such myriad complex shapes.

Ice is an awkward substance. There are 14 ways of packing its molecules together in different crystalline structures, more than any other substance. Although most of these forms occur only at high pressures, the ways in which water molecules interact with each other and other materials remain complicated and largely unexplained. Libbrecht is undaunted. "I wanted to do it for no other reason than snow crystals are pretty and no one has done it before," he says.

Snow crystals begin life as dust particles that absorb water molecules from within clouds. These pack themselves into hexagonal crystal nuclei whose six facets attract more molecules to form six symmetrical arms or plates. Libbrecht makes his clouds inside cold chambers the size of portable television sets. He keeps the temperature steady, controls the amount of water vapour and watches things evolve. He prefers to grow the crystals on the ends of long needles of ice. He has perfected a way of using high voltage to assemble polarised water molecules into long, thin spines. These have proved ideal substitutes for the dust, holding the crystals in place without interfering with their development.

The crystals are scrutinised by technology not unrelated to what Libbrecht is using with the Laser Interferometer Gravitational-wave Observatory (Ligo), a new continent-spanning gravitation wave detector. Laser interferometry enables him to measure how the crystals grow, which can then be related to the controlled environment inside the chamber. Two dominant influences have emerged. The first is supersaturation, a measure of the amount of water vapour in the air above 100 per cent humidity. This provides the medium from which the crystal is sculpted. The second is temperature, which appears to be the true hand of the artist. Libbrecht's experiments have shown just how sensitive snow crystals are to changes in temperature. A difference of just a few degrees Celsius can affect the growth rate by a hundred fold. The fluctuating temperatures that the snow crystals encounter as they fall determine their shape. The International Commission on Snow and Ice classifies seven principal snow-crystal types, including plates, stellar crystals, needles and fern-like dendrites. Libbrecht is establishing which conditions result in a particular form. During its descent, a crystal can switch from one form to another. As each arm of the crystal experiences the same sequence of temperature and supersaturation, it mirrors the other five, giving the crystal its satisfying symmetry.

There are an estimated trillion trillion snow crystals created in the atmosphere every year, yet the chances of any two crystals passing through the same variations is negligible. But what if you can control that environment? He says: "You can design snow crystals to some extent. If you really knew how to grow them you could grow two alike."

He is not there yet but the gallery of designer crystals he has produced marks impressive progress. Among them is what, at 2.5cm across, may be the largest snow crystal grown. At the other extreme, Libbrecht is also exploring the surface physics of snow crystals. The unusual surface structure of ice, with its uppermost layer of loosely bound molecules, seems to play a key role. He was once a lone figure in this field, but his enthusiasm has spread. His website - http://snowcrystals.net - has drawn responses from a wide audience. Even his family has been drawn in. When Libbrecht returns to his native North Dakota he often ventures out with his wife to search for snow crystals. He finds it strange that others do not do the same.