Crystallized. Winner Daniel Shechtman and an atomic model of a quasicrystal (inset).

The Ames Laboratory/DOE

Discoverer of 'Impossible' Crystals Gets Chemistry Nobel—And Last Laugh

This year's Nobel Prize for chemistry goes to a scientist whose controversial discovery forced chemists to redefine the concept of a crystal. In 1982, Daniel Shechtman of the Israel Institute of Technology in Haifa discovered an alloy of aluminum and manganese that appeared to have fivefold symmetry—that is, the atoms in it formed a pattern that appeared essentially the same when rotated by a fifth of a turn, or 72 degrees. Other researchers scoffed, as such arrangement was thought to be mathematically impossible. Yet, scientists eventually realized that atoms in a solid can achieve such symmetry by arranging themselves in a pattern that almost but never quite repeats—a "quasicrystal."

"He does deserve a Nobel Prize for ushering in this new kind of phase in chemistry—crystals that are not crystals," says Oxford University mathematician Roger Penrose, who played an indirect role in explaining the materials. David Phillips, president of the U.K.'s Royal Society of Chemistry, said in a statement that quasicrystals are "quite beautiful, and have potential applications in protective alloys and coatings."

Prior to Shechtman's discovery, the definition of a crystal was a material in which atoms are arranged in a regular pattern that repeats itself. That definition puts limits on the symmetry a crystal can have, as a simple child's game shows. Suppose you want to cover a tabletop by arranging identical tiles. A repeating pattern of triangles does the trick, so it's possible to make crystals with threefold symmetry. Squares or hexagons also work, so crystals with fourfold and sixfold symmetry can also be made. But pentagons won't work; there will always be gaps between them. Thus, a repeating crystal with fivefold symmetry is impossible.

Nevertheless, Shechtman saw what he saw. On the morning of 8 April 1982, while working at the National Institute of Standards and Technology in Gaithersburg, Maryland, he took the sample of aluminum and manganese alloy, which he had cooled quickly to keep it from crystallizing and fired a beam of electrons into it. If there was an orderly arrangement of atoms in the material, the electrons would "diffract" off the various planes of atoms in it and emerge at specific angles to produce a recognizable pattern in a detector. Shechtman saw a diffraction pattern unlike any he'd seen before: concentric circles of bright dots, each circle with 10 dots in it. The tallies pointed to the impossible symmetry.

Shechtman checked and rechecked his experiment in every way he could. When he finally told colleagues about his discovery, he was met with dismissal and ridicule. His claims caused such embarrassment that his boss asked him to leave the research group. The results drew a similar hostile response when he finally published them in Physical Review Letters in November 1984. "I got an angry letter from Linus Pauling [a prominent critic]," Penrose says. But slowly other crystallographers came forward with similar results that they had earlier dismissed as evidence of twinned crystals—pairs of crystals with different orientations grown together whose boundary causes unusual diffraction patterns.

Once scientists convinced themselves the diffraction patterns were real, they had to figure out how the atoms were arranged. The answer came from mathematicians who, like children, had been thinking about patterns of tiles. Some had been puzzling over curious mosaics that have a limited number of different shaped tiles and that fitted together in patterns that never repeated themselves. Such mosaics were used by Arabic artists as early as the 13th century to decorate buildings such as the Alhambra Palace in Granada, Spain. Mathematicians in the 1960s and '70s strove to find the smallest number of tiles which could produce such an aperiodic pattern. In the mid 1970s, Penrose came up with a set of just two rhombuses that did the job. Just looking at Penrose's pattern, one sees plenty of pentagons and decagons.

That set a number of chemists thinking about whether atoms could adopt a similar pattern. Crystallographer Alan Mackay built a model with circles representing atoms at the corners of Penrose's tiles and calculated what sort of diffraction pattern it would produce. The answer: bright dots in a circle with 10-fold symmetry. Paul Steinhardt, who was then at the University of Pennsylvania, and his student Dov Levine had also been devising theoretical structures based on Penrose tiling. When a colleague showed Steinhardt a preprint of Shechtman's first paper in autumn 1984, "I leapt up in the air. The two matched by eye beautifully," he says. Steinhardt and Levine published a paper shortly after Shechtman's linking his observations to Penrose-like structures, and coined the term "quasicrystal."

Much remains a mystery about quasicrystals, such as how such complex long-range structures can form from individual atoms. "They can't be produced simply with local rules; there has to be some subtle kind of production," Penrose says. Steinhardt agrees. "The mathematical techniques we use on crystals don't work on quasicrystals," he says. "We can't predict their properties so well."

Quasicrystals have been found in nature, a mineral discovered 3 years in the Koryak Mountains in eastern Russia. They have also been found in one of the world's most durable steels, made by a company in Sweden for razor blades and surgical needles. They are beginning to find other industrial applications, such as nonstick coatings in pans, heat insulation in engines, and as thermoelectric materials to salvage waste heat. "Shechtman's quasicrystals are now widely used to improve the mechanical properties of engineering materials and are the basis of an entirely new branch of structural science," Andrew Goodwin of Oxford said in a statement. "If there is one particular lesson we are taking from his research, it is not to underestimate the imagination of nature herself."

As for the definition of a crystal, in 1992 the International Union of Crystallography changed its definition of a crystal from a regular repeating array of atoms to "any solid having an essentially discrete diffraction pattern."