Mitch Jacoby

C&EN Chicago

In 1982, Dan Shechtman peered through the window at the viewing stage of an electron microscope and saw something that broke the rules of crystallography. His alloy specimen produced a diffraction pattern containing rings of sharp spots--indicating a well-ordered material. But the number of equally spaced spots per ring was 10. "Tenfold [symme-try]? There's no such animal," the materials engineering professor at Technion--Israel Institute of Technology, Haifa, recalls telling himself as he studied the bizarre pattern. One- two-, three-, four-, and sixfold rotational axes were firmly established in crystallography. All other symmetries were thought to be impossible.

A millimeter-sized holmium-magnesium-zinc quasicrystal prepared by Paul C. Canfield and Ian R. Fisher at Ames Laboratory using a molten-metal technique known as flux-growth reveals its icosahedral symmetry via pentagonal faces.

The type of aluminum-manganese particles that Shechtman was examining came to be known as quasicrystals. These materials derive their name from the fact that their atoms form orderly patterns, but unlike regular crystals, the patterns are not periodic--they don't repeat in three dimensions at fixed intervals.

Hundreds of alloys have since been observed to form quasicrystalline phases. And earlier this year, Shechtman won the $100,000 Wolf Prize in physics, awarded by the Wolf Foundation of Herzlia, Israel, for his pioneering work. The exact structure of these high-symmetry substances remains to be settled, but in the meantime, researchers are investigating quasicrystals' hardness, corrosion resistance, and other properties and have begun putting them to good use. A number of applications have hit the market recently, and others may follow soon.

One of the most talked about quasicrystal applications is a cookware coating that recently has become commercially available. Jean Marie Dubois, a materials scientist and research director at the National Center for Scientific Research (CNRS) in Nancy, France, says that a French manufacturer inquired about metal coatings for frying pans. An author of patents in metals hardening, Dubois knew that high cooking temperatures would cause amorphous aluminum materials to crystallize and degrade, so he suggested using related quasicrystalline alloys instead.

Dubois: alloys with useful properties

"We knew how to change the composition of our materials to modify their corrosion resistance and mechanical strength," Dubois says. The aluminum-copper-iron-chromium quasicrystal alloys that Dubois and coworkers formulated have low coefficients of friction, high hardness, and low surface energy (which may result in an unreactive surface).

"What we didn't know was how the coatings would respond to contact with food," Dubois continues. After cooking a steak on a plate coated with their alloy, the French materials scientists had their answer. Quasicrystal coatings make excellent cooking surfaces.

Why do they work so well? Dubois explains that, although they're composed of metal atoms, quasicrystals exhibit poor thermal (and electrical) conductivity. That causes quasicrystal coatings to transfer heat gradually and uniformly to the surface of foods, unlike aluminum or other good heat conductors that get hot very quickly and can burn foods.

The effect is rooted in contact temperature--the temperature "sensed" by food sitting in a pan. Particularly important in cooking applications, contact temperatures in turn are related to a layer of water that resides between a cooking surface and food. On a quasicrystal surface, the contact temperature is near 115 C, while on aluminum it measures roughly 200 C, Dubois says.

In addition to being inexpensive, the new coatings are also stick-resistant, though not so much as DuPont's famous Teflon nonstick coating, Dubois acknowledges. But unlike nonstick polymer coatings, quasicrystals are harder than ordinary kitchen utensils and can't be scratched and ruined by forks and spatulas. "For a durable material with low adhesion, high hardness, and other good mechanical properties, quasicrystals are a good compromise," Dubois stresses.

Quasicrystals also have been put to use hardening steels. Researchers at Sandvik Steel, Sandviken, Sweden, found that certain steel formulations and heat treatments combine to produce incredibly strong products. Their analysis shows that these preparation procedures cause small quasicrystal particles (precipitates) to embed themselves in the carbon steel framework.

Manufacturers of needles for surgery and acupuncture and makers of dental instruments have been taking advantage of Sandvik's quasicrystal-strengthened steels for a few years. And recently, one company began marketing a water- and lather-compatible electric shaver that uses Sandvik steel in the shaver heads.

Describing the mechanism by which precipitates increase steel's hardness, Technion's Shechtman notes that metals become deformed and bent by moving dislocations (certain lattice imperfections) through the crystal structure. "Little quasicrystal particles in Sandvik's steel serve as obstacles to the motion of dislocations," he notes. The dislocations become trapped, resulting in a much harder material. Shechtman points out, however, that most strengthened metals use precipitates that are not quasicrystalline.

Plasma arc spray gun deposits droplets of molten Al₆₅Cu₂₃Fe₁₂ on a rotating stainless steel cylinder forming protective quasicrystalline coatings. [Courtesy of Daniel Sordelet]

As commercial uses for quasicrystals begin to emerge, basic and applied research on the oddly ordered materials continues. Surface chemists, materials scientists, solid-state physicists, mathematicians, and others are investigating quasicrystal surface properties, preparation methods, atomic structure, and other aspects of this new class of materials.

Owing to their hardness and resistance to wear, quasicrystal materials offer promise as protective coatings for engine parts and other applications where sliding motions can cause abrasion. At Carnegie Mellon University, Pittsburgh, chemical engineering professor Andrew J. Gellman investigates aluminum-palladium-manganese quasicrystal tribology (the basics of wear and friction). Gellman and graduate student Jeff S. Ko recently have shown that atmospheric contaminants play a key role in the reportedly low friction coefficients measured on some quasicrystal surfaces [Surf. Sci., 423, 243 (1999)].

Examining air-exposed specimens, the Carnegie Mellon researchers measure friction coefficients between pairs of quasicrystals of roughly 0.1--the same low coefficient that others have reported. But after cleaning the surfaces of contaminants such as carbon, oxygen, and sulfur (under vacuum), the group finds that the friction value jumps to 0.6. That high value drops to 0.3 after minimal oxidation causes a thin layer of alumina (Al₂O₃) to coat the surfaces. They conclude that the material's low friction results from its hardness and the presence of oxide films.

Taking advantage of quasicrystals' hardness in wear-resistant applications has proven difficult because the alloys are often brittle and easily fractured. But material scientists have known for some while that by blending a brittle material with a more ductile component, they can make products that are less susceptible to chipping and breaking. Applying that concept to quasicrystals, Daniel J. Sordelet, a scientist in the ceramics and metallurgy program at Ames Laboratory, Ames, Iowa, uses iron aluminide to improve characteristics of quasicrystalline Al₆₅Cu₂₃Fe₁₂ coatings.

To determine an optimum concentration of the ductile additive, Sordelet prepared quasicrystalline-iron aluminide blends in concentrations ranging from 1 to 20% by volume and applied them to test substrates using a plasma arc spray gun--a device that propels molten droplets at a target. The coated samples were evaluated by a standard abrasion test.

Surprisingly, the Ames researcher found that the 1% iron aluminide blend was the most wear resistant [Mater. Sci. Eng. A, A255, 54 (1998)]. According to Sordelet, the low-concentration coating's toughness comes from two factors. The first is that the additive's presence changes the mechanics of the wear process from fracture to a type of deformation in which coating material flows out of the way of abrasive particles. And the second is that using a minimum amount of the softer component maintains the quasicrystal's hardness.

Patricia A. Thiel, a researcher at Ames Laboratory and a chemistry professor at Iowa State University, Ames, studies quasicrystal surfaces and their chemistry using vacuum-based surface probes. "It's been exciting to be able to apply the tools of a relatively mature field like surface science to a mostly unexplored class of materials," Thiel says. But because of the limited number of good samples and proven preparation methods, doing so has not been so straightforward.

Before examining chemical reactions on surfaces, researchers typically clean or pretreat specimens using procedures that render a surface in a reproducible condition; for example, contaminant-free and ordered in a particular way. But many of the common treatment methods that work well for metallic surfaces, such as high-temperature annealing or ion sputtering (atomic-scale sandblasting), can change the stoichiometry and ruin the structure of quasicrystal surfaces.

Working with Cynthia J. Jenks, Thomas A. Lograsso and other Ames researchers, Thiel and graduate student Zhouxin Shen worked out a sequence of annealing and sputtering steps to reliably prepare aluminum-palladium-manganese samples. Depending on recipe details, the group can make compositionally similar products with crystalline or quasicrystalline surfaces. (Quasicrystal products have the nominal composition Al₇₀Pd₂₁Mn₉, while for the crystalline phase the formula is Al₆₀Pd₂₅Mn₁₅.) Thiel points out that the ability to prepare samples of similar elemental composition in crystalline and quasicrystalline forms should help the team elucidate the role of structure in surface chemistry.

Not your average ordinary crystal The unusually ordered structures of a family of mainly aluminum-based alloys--now known as quasicrystals--presented the scientists who first studied them with a classification difficulty. The problem? Although the compounds’ atoms line up neatly over long distances, their building blocks don’t repeat periodically in three dimensions and they’re often endowed with five- and 10-fold rotational symmetry. That means they don’t fit into any of the 14 Bravais lattices or the 230 space groups. Their discoverer, Dan Shechtman, a materials engineering professor at Technion, Haifa, Israel, tells of the struggle to convince the scientific world that these structures were real and not the result of crystallography artifacts (twinning) or misinterpretation of microscopy results. Shechtman says some colleagues handed him textbooks and told him that if studied he’d realize his claims were impossible. Others, most notably two-time Nobel-prize winner Linus Pauling, denounced the quasicrystal concept in scientific forums. But in time many quasicrystal observations were reported and eventually a consensus formed in support of the legitimacy of the odd structures. Unlike crystalline materials, many quasicrystals exhibit icosahedral symmetry. Like the 20-sided icosahedron and the pentagonal dodecahedron (which bare the same relationship to each other that the cube and octahedron share), these structures possess six five-fold rotational axes, 10 three-fold axes, and 15 two-fold axes. They also have 15 symmetry planes. Elemental boron and the B12H122- ion adopt icosahedral symmetry but that geometry remains very unusual for molecules. For ordinary crystals, five-fold symmetry, or for that matter anything other than two-, three-, four- and six-fold symmetry, is considered impossible because a three-dimensional periodic lattice cannot be made from objects with those symmetries. But those sacred rules don’t hold for quasiperiodic materials. “The discovery of these materials created a revolution in our understanding of the structure of matter,” says Shechtman. From the days of X-ray crystallographer von Laue in the early 1900s until quasicrystal’s discovery in the 1980s, nothing shook the paradigm “Order in crystals is periodicity,” he says. But things are different now. The International Union of Crystallography recently changed it’s periodicity-based crystal definition to “any solid having an essentially discreet diffraction diagram.”

Armed with procedures for making reproducible test samples, the Ames researchers next turned to investigating surface structure. Specifically, the team set out to determine the relationship between the structures of a quasicrystal surface and its interior (bulk). Those studies, based on a low-energy electron-diffraction technique, were conducted with Iowa State professor of physics and astronomy Alan I. Goldman and Michel A. Van Hove of Lawrence Berkeley National Laboratory, Berkeley, Calif. [Phys. Rev. Lett., 78, 467 (1997)].

Ames research group includes (seated from right) Jenks, Lograsso, Thiel, and other Ames Laboratory and Iowa State University quasicrystal investigators.

Just as the structure of crystal surfaces often differs from that of a crystal's interior, so too atoms in quasicrystal surfaces may be arranged differently than their bulk counterparts. In the Ames aluminum-palladium-manganese alloys, like many quasicrystal materials, the building blocks of the interior belong to the icosahedral family--a family whose members have fivefold symmetry in several orientations.

And what about the surface? The group says that the topmost layers resemble a mixture of "relaxed" interior icosahedral layers with changed interlayer spacings. The top layer is aluminum-rich, they say, while the layer beneath it is a 50-50 blend of aluminum and palladium.

According to Thiel, the main conclusion of the structure analysis is that of the multitude of conceivable arrangements, nature chooses densely packed, aluminum-rich surfaces for quasicrystals. Those findings are consistent with the norms of crystalline surfaces, she says, where atomic packing density and chemical composition tend to be optimized in a way that leaves a surface in its most stable or lowest energy configuration.

Seeking to fit another piece in the quasicrystal puzzle, Thiel and coworkers compared the chemical reactivity of quasicrystal surfaces to crystalline surfaces. Conventional wisdom in this field holds that quasicrystal surfaces ought to be especially nonreactive. Chemical inertness is predicted because of the smaller number of electronic states available to electrons of a certain energy in quasicrystals relative to metallic crystals.

But so far, that pattern isn't what the Ames group observes. "We find that the surfaces behave much like one would expect for an aluminum-rich alloy," Thiel says. In one set of experiments, the researchers investigated surface oxidation of aluminum-copper-iron samples. The three compositionally similar alloys used in those tests (two crystalline and one quasicrystalline) exhibit-ed the same oxidation characteristics [Philosophical Mag. B, 79, 91 (1999)]. The researchers reached the same conclusion about the oxidation of aluminum-palladium-manganese surfaces.

Recently, Jenks, Lograsso, and Thiel examined reactions of hydrogen, carbon monoxide, methanol, and iodoalkanes on an aluminum-palladium-manganese quasicrystal [ J. Am. Chem. Soc., 120, 12668 (1998)]. Once again, the team reports that aluminum-palladium-manganese quasicrystal surface chemistry essentially emulates aluminum's surface chemistry.

"How general are these results?" Thiel asks, referring to the lack of evidence for unique quasicrystal chemistry. "No one knows," she says. "What we've learned about quasicrystal chemistry so far comes from very few experiments." Most studies have been limited to aluminum-palladium-manganese samples (like Al₇₀Pd₂₁Mn₉) and to only one type of that alloy's surfaces--those with fivefold symmetry. "We don't really know how the surfaces of other alloys behave or how surfaces with other symmetries behave." Thiel adds that by building a database that extends to other quasicrystals, researchers will be able to start answering these questions.

But even before moving on to systems with other elements--and there are many--researchers acknowledge that questions remain unanswered about the best way to prepare aluminum-palladium-manganese sample surfaces and what effect preparation may have on surface chemistry. Rather than sputtering and annealing the surfaces as the Ames team does, scientists at the Institute of Solid State Research at Jülich Research Center in Jülich, Germany, cleave their quasicrystals in vacuum. The two groups observe different results.

Ebert: cluster-subcluster model

Jülich staff scientist Philipp Ebert says scanning tunneling microscopy reveals that cleaved surfaces are constructed of aggregates of little clusters [Phys. Rev. Lett., 77, 3827 (1996)]. The small clusters are the building blocks of the aluminum-palladium-manganese quasicrystal lattice. The cluster-subcluster surface differs from the Ames step-terrace surface, which resembles the floor plan of a bilevel house in which the living room and dining room, for example, are separated in height by a step. And not only do the two preparation methods lead to different-looking surfaces, but a recent study by the Jülich group shows that heat treatments used after cleavage cause the surface structure and chemical composition to evolve [Phys. Rev. B, 57, 2821 (1998)].

Acutely aware that much remains unknown in the study of quasicrystals, Ames researchers Thiel and Jenks acknowledged in a recent review article that their paper asks many questions but offers few answers [ Langmuir, 14, 1392 (1998)]. "And in so doing," they write, "it aptly reflects the current status of our understanding of quasicrystalline surfaces. The topic is important, exciting, and rapidly moving, and more answers are sure to emerge soon."

[Previous Story][Next Story]

Chemical & Engineering News
Copyright © 1999 American Chemical Society