Probing deeper into quasicrystals.

Probing deeper into quasicrystals

Less than a year ago, the term "quasicrystal' was practically unknown. Now, hundreds of researchers throughout the world are energetically poking into what many scientists believe is a new kind of crystalline matter. This week, the topic of quasicrystals was highlighted in Washington, D.C., at a National Academy of Sciences symposium featuring significant advances in materials research.

Until recently, most crystallographers believed that atoms within crystals had to be arranged in blocks that stacked evenly to create a regularly repeating pattern. The discovery at the National Bureau of Standards (NBS) in Gaithersburg, Md., of "shechtmanite,' an aluminum-manganese alloy that shows a noncrystallographic, fivefold symmetry in electron diffraction patterns, shattered this belief (SN: 1/19/85, p. 37; 3/23/85, p. 188).

"You're talking about a kind of physics in which it's easy to do the experiments,' says Paul J. Steinhardt, a physicist at the University of Pennsylvania in Philadelphia. "Once this material was reported, there were many laboratories that were immediately able to reproduce the result.'

Recent studies reveal that this icosahedral structure turns up in dozens of alloys, says NBS materials scientist John W. Cahn, who was involved in the initial discovery. These include many aluminum alloys and unusual combinations like uranium, palladium and silicon.

Almost all of these forms are "metastable.' A touch of heat, for instance, nudges the atoms of a quasicrystal into a more stable periodic arrangement. However, evidence has emerged that sometimes, at least in the case of an alumium-lithium alloy, the icosahedral form is the stable phase at room temperature.

Researchers are also finding new ways of making quasicrystals. In the Oct. 7 PHYSICAL REVIEW LETTERS, a team from Cornell University reports that using a xenon ion beam to bombard a thin film of an aluminum-manganese alloy can produce the quasicrystalline phase. Previously, quasicrystals had been created by methods like "splat colling,' which involves rapidly freezing molten metal.

"The advantage of the ion beam technique is that we can control everything,' says Cornell's James W. Mayer. This allows researchers to do careful experiments. "The big push is to understand the structure,' he says. Cornell graduate students David A. Lilienfeld and Michael Nastasi, who did the ion beam work, are now trying the technique on other materials and exploring the range of alloy compositions that can be jostled into a quasicrystalline state.

Other studies are unveiling nonperiodic symmetries beyond the fivefold, icosahedral structure initially discovered (SN: 8/17/85, p. 102). For example, in the Sept. 30 PHYSICAL REVIEW LETTERS, Leonid Bendersky of the Johns Hopkins University in Baltimore reports the formation of a decagonal phase, which has neatly stacked layers, each showing a nonperiodic, 10-fold symmetry.

Despite the recent flood of research papers devoted to quasicrystals, the theoretical interpretation of the results as a genuinely new crystalline structure remains controversial. In the Oct. 10 NATURE, Linus C. Pauling of the Linus Pauling Institute of Science and Medicine in Palo Alto, Calif., argues that the "icosahedral' structures are really "multiple twins of a cubic crystal.'

Pauling proposes that aluminummanganese alloys, when suddenly cooled, solidify into a cubic form in which each unit contains about 1,120 atoms. About 20 crystals, made up of these cubic units and roughly tetrahedral in shape, could grow out from a central seed to produce an approximate icosahedral shape. Pauling's structure seems to account for the way X-rays diffract from powdered samples of the new materials.

"Crystallographers can now cease to worry that the validity of one of the accepted bases of their science has been questioned,' Pauling concludes.

"I'm certainly not convinced that he [Pauling] has the correct explanation for all of the experiments,' says Harvard physicist David R. Nelson. "I'm skeptical that his model will account properly for a single-crystal diffraction pattern.'

Nelson's comments are typical of the reaction among quasicrystal researchers. Although Pauling's structure seems to work for a powder, consisting of a host of tiny crystals sitting in random positions, it doesn't work, they say, for the distinctive pattern of spots seen in a single-crystal electron diffraction experiment.

"This material really is a quasicrystal,' says Steinhardt, "but one that has a lot of defects in it. We'd really like to have a more perfect sample.' This would allow researchers to check more closely proposed theories about the structure of the new materials.