Advanced Materials Processing, Synthesis, and Characterization

Reaction Synthesis
Filters for Power Plants
Making Nanocrystals
New Building, Laser System
Automotive Catalyst Atoms
Welding Process and Properties
Studying Auto Brakes

Materials—the right stuff—form the key to progress in energy efficiency; environmental protection and remediation; and energy production through fossil, nuclear, and alternative technologies. ORNL's program in materials research and development (R&D) is one of the nation's largest, as well as one of the most respected internationally. Our programs cover the full spectrum from materials synthesis, processing, and characterization through modeling and performance evaluation. A key factor in the success of these programs is the integration of basic, applied, and industrial R&D.

This multidisciplinary research focuses on several important classes of materials, including ceramics and composites, metals and alloys, superconductors, and thin films. In addition to providing scientific leadership, ORNL's materials programs are major contributors to technology.

ORNL researchers developed the Exo-Melt furnace concept for efficiently and safely melting and casting alloys of nickel and iron aluminide. Photograph by J.W. Nave. Computer photo enhanced by Mark Robbins.

In 1995 ORNL received 5 R&D 100 awards, 3 of which were for advances in the materials sciences. Our prizewinning developments were the Exo-MeltTM process, which provides a furnace-loading method for low-cost, energy-saving production of nickel and iron aluminides, which have industrial uses; gelcasting, a new ceramic-forming process for making high-quality, complex-shaped ceramic parts that is being advanced to the commercialization stage by AlliedSignal and other companies; and (in collaboration with 3M researchers) the 3M ceramic composite filter, a fiber-reinforced ceramic composite candle filter that removes particulates from hot gas streams in pressurized fluidized bed combustion systems and coal gasification plants, protection turbine blades needed for electricity generation.

Past products that evolved from our integrated programs have an estimated private-sector value of more than $300 million a year. Commercial applications of ORNL technologies include artificial hip joints, furnace fixtures for manufacturing structural steel and automobile parts, dies for making aluminum cans, filters to reduce emissions from coal-burning plants, new pressure vessel steels for power plants, and ceramic drills and cutting tools.

ORNL's materials program embraces an unparalleled infrastructure of specialized equipment, facilities, and expertise. These are central to extensive interactions and collaborations with both the private and public sectors. The High Temperature Materials Laboratory (HTML) and the Shared Research Equipment Program offer state-of-the-art instruments for characterizing microstructure and properties of materials with the aim of linking one to the other. The High Flux Isotope Reactor houses the Neutron Scattering Research Facility for characterizing the structure of materials with neutron probes. The Surface Modification and Characterization Research Center includes unique facilities for ion implantation of materials to improve their properties. And the Residual Stress User Center Combines strengths of HTML and the Neutron Scattering Research Facility to measure residual stresses in welds, automobile brake rotors, and other materials and machines.

Reaction Synthesis: Take a Powder, Make a Product

At high temperatures, nickel and iron aluminides modified at ORNL are stronger than stainless steel, exhibit ductility, and resist corrosion. But as blocks, sheets, or tubes, these heat-resistant alloys also resist being shaped into a final product because they're so brittle at room temperature. We have found a promising solution: form the alloy from metallic powders using internal heat generated during the powders' reactions. This "reaction synthesis" method is considered attractive for producing aluminide products in desired shapes, with little or no machining.

Reaction synthesis may also be used to produce highly dense aluminide products for structural uses. Aluminides have many uses now, but because they are porous, they are too weak to bear heavy weights. Using reaction synthesis we have reduced the porosity from 3% to 0.1% in a ductile nickel-aluminide product.

Reaction synthesis could lower costs in two ways. Aluminide products can be formed directly from elemental powders rather than in two steps (melting and casting), and the self-generation of heat during alloy formation reduces energy requirements.

In reaction synthesis of aluminides, powders of aluminum and the other metal are mixed. The unreacted samples are shaped into disks of a desired shape and heated in a chamber under vacuum to a temperature that initiates a self-sustaining, heat-liberating reaction, like burning logs in the fireplace. By applying a little pressure, we found we could squeeze out the material's pores, making it almost fully dense.

Using high-speed videotaping equipment, we studied the reaction behavior of iron aluminides during synthesis in air. We found that the reaction rate depends on compact composition and powder particle size; it increases with greater aluminum content, and it drops with increasing powder size.

Another ORNL group has shown that reaction synthesis, not under vacuum or pressure, can be used to melt and cast ingots. This group developed the Exo-Melt process by extending principles of reaction synthesis to the melting and casting of iron aluminide (Fe3Al) and nickel aluminide (Ni3Al). This process, which uses half as much energy as traditional processes and addresses safety concerns of the alloy preparation industry, received an R&D 100 Award in 1995.

The key to the success of the Exo-Melt process is the furnace-loading sequence. As the aluminum melts and comes in contact with the heated nickel at the top of the furnace, NiAl forms, releasing large amounts of heat. On its way down, the superheated NiAl liquid dissolves alloying elements (boron, chromium, molybdenum, and zirconium), found by ORNL to make the alloy ductile and strong at high temperatures. Additional NiAl reacts with the nickel at the bottom of the furnace to form the desired alloy—Ni3Al.

Using the Exo-Melt process, four vendors now melt and cast highly durable products from nickel aluminides for the steel, automobile, and tool industries. Products include transfer rollers for heat-treating steel plates in steel mills and "furnace furniture"—assemblies that hold automobile parts during high-temperature treatment to harden their surfaces.

As a result of these successes, a unique facility has been developed at ORNL to study reaction synthesis of intermetallic alloys. The facility will be used to study processing parameters and the feasibility of forming near-net-shape products. Someday industry may find this emerging method of making products from powders too good to resist.

This continuing intermetallics success story and example of the integration of research and technology developments highlights the value of cooperation among several programs in DOE's Office of Energy Research that are supporting this work. They are Basic Energy Sciences, Energy Efficiency, Fossil Energy, and Energy Research Laboratory Technology Applications programs.

Filters for Clean Power Plants

The most efficient power plants of the near future are likely to use hot gases from coal to drive a gas turbine and produce steam that will spin a steam turbine. A technical problem, however, has marred the performance of pilot plants of these pressurized fluidized-bed combustion (PFBC) and integrated coal gasification combined cycles. Candle filters used to remove hot-gas particulates that erode and corrode gas turbine blades survived only a short time. ORNL researchers recognized the need for a longer-lasting filter, so they developed a gas particulate filter made of a novel ceramic composite. Their invention was then licensed to the 3M Company.

ORNL-3M ceramic composite filters for future coal
power plants are a technology transfer success.

David Stinton (left) and Rod Judkins have worked on developing and commercializing the 3M hot-gas ceramic filter (shown here), which has been tested in the Westinghouse Advanced Particle Filter System at the Tidd Pressurized Fluidized-Bed Combustion Demonstration Project in Brilliant, Ohio (interior shown at right).
ORNL and 3M researchers collaborated on improving the filter to make it a commercial product. To see if it could endure real-world conditions, Westinghouse Electric Company tested the filters in a combined-cycle demonstration plant. In December 1994, ten full-size, 3M candle filters were installed in a filter vessel at the Tidd PFBC demonstration project, which was funded by the DOE Clean Coal Technology Program, at Ohio Power Company's power plant in Brilliant, Ohio. The system was operated until about March 31, 1995.

On April 26, 1995, a borescope examination of the interior of the filter vessel was performed. All ten of the filters were in place and appeared to have functioned well. The filters were removed during the next few weeks for nondestructive and destructive examination. All of the filters had performed well, and all had retained properties that suggested they could perform for times acceptable for commercial operation of a PFBC plant. Additional filters are in a second PFBC demonstration project in Karhula, Finland. The filters' versatility has been shown by their excellent performance in a coal gasification plant in Germany. As a result of the successful demonstrations, 3M has increased its production capability for the filters. ORNL's small candle filter effort has sparked the development of a business worth millions of dollars.

The research and technology transfer efforts were supported by DOE Office of Fossil Energy's Advanced Research and Technology Development Materials Program and the Morgantown Energy Technology Center.

Making Nanocrystals with Ion Beams

As their semiconductor "brains" get smaller, computers get smarter and faster. In the past few decades, researchers have shrunk many materials devices to increase computer performance. When sizes are decreased to billionths of a meter (nanometers), materials are reduced to small clusters of only hundreds of atoms. Such nanoclusters, or nanocrystals, often exhibit fascinating physical properties very different from those of the bulk material. For example, electronic energy levels and optical properties such as absorption and emission of light can be shifted dramatically when dimensions of crystallites approach dimensions of fundamental excitons (electrically neutral excited states of insulators and semiconductors).

Ion beams synthesize nanocrystals, forming light-emitting materials and compound semiconductors.

Striking colors arise from gold nanocrystals formed in sapphire and fused silica by ion implantation. Encapsulated gold nanocrystals produce purple or red.

Nanocrystal properties have been exploited throughout the centuries. For example, since the Middle Ages, glaziers experienced in the art of making stained glass have incorporated small metal precipitates in molten glass to produce vibrant colors. However, researchers are only now beginning to understand enough about such phenomena to predict and intentionally control the behavior of small clusters of atoms.

At ORNL, high-energy accelerators are used to implant metal ions, such as gold, into silica (SiO2), one common type of glass. The surface region of gold-implanted silica is much more supersaturated with metal atoms than silica treated by conventional techniques. During high-temperature annealing, gold nanocrystals approximately 10 nm in diameter are formed by precipitation. In addition to their striking colors (shown in the photograph below), such ion-implanted samples reveal many other useful optical properties, such as a refractive index that depends strongly on the intensity of light striking the material.

Scientific and technological interest in nanocrystals extends far beyond stained glass. ORNL researchers have shown that ion implantation can be used in a novel way to create high densities of a wide range of nanocrystals in a number of technologically important materials, including silica, sapphire (Al2O3), and even crystalline silicon. In this approach, high-dose ion implantation of the near-surface region creates a solid solution that is supersaturated with impurity ions. When the sample is heated, the highly concentrated impurity ions precipitate out, forming nanocrystals that usually cannot be synthesized by conventional methods.

The objective of this research is to understand and control the size, structure, and optical and physical properties of nanocrystalline composites produced by high-dose ion implantation. In addition to the metallic nanocrystals shown in the photograph, ORNL researchers have demonstrated that elemental semiconductor nanocrystals of silicon and germanium can be synthesized in SiO2 and Al2O3.

We have observed a strong red light from SiO2 samples containing silicon nanocrystals. The light intensity is comparable with that from porous silicon, and the wavelength (i.e., color) can be tuned by changing nanocrystal size. Because size determines the wavelength of emitted light, full-color panel displays for computers may someday be made of appropriately sized semiconductor nanocrystals formed by ion implantation.

The researchers have also synthesized more complex compound semiconductors and alloy nanocrystals (e.g., silicon germanium, gallium arsenide, cadmium selenide, and gallium nitride) by implanting combinations of different ions into silica and sapphire hosts. By controlling implantation and annealing conditions, they discovered that it is possible to control orientation and, in some cases, crystal structure of nanocrystals.

The scientists also incorporated compound semiconductor nanocrystals such as gallium arsenide (GaAs) into a silicon matrix. The optical properties of this nanocrystalline composite are expected to be considerably different from those of current silicon-based devices because energetic electrons can be directly converted into visible light in GaAs, but not in silicon. It may be possible to produce buried continuous layers of GaAs by extending ion implantation to higher doses. This approach could provide a way to combine silicon-integrated circuits with high-speed GaAs layers to achieve fully integrated optoelectronic devices.

This research provides a new way to form encapsulated nanocrystals in materials to investigate the nanocrystals' size-dependent properties. The synthesis of a wide range of nanocrystals of various sizes in technologically important host materials should give rise to even smarter computers, as well as new electronic and optical devices never dreamed of by makers of stained glass windows.

This research was supported by DOE, Office of Energy Research, Basic Energy Sciences.

New Research Building and Laser MBE Film Growth

Brick by brick, a new building for ORNL's Solid State Division was constructed in 1995. Now, researchers in Building 3150 are using a pair of new laser molecular beam epitaxy (MBE) systems to build novel materials, atomic layer by atomic layer. Dedicated September 15, 1995, Building 3150's second floor houses the laser MBE systems that are used to synthesize new crystalline thin-film materials for electronic and photonic uses, as well as facilities for semiconductor processing, supersonic-jet film growth, and photolithography.

The Solid State Division's Building 3150 under construction in 1995.

The two laser MBE systems developed at ORNL are being used to grow thin films on single-crystal substrates in a controlled way near the atomic level. As in conventional pulsed laser ablation film growth, laser MBE films are grown in a vacuum or at very low pressures. A pulsed excimer ultraviolet laser beam vaporizes pressed polycrystalline targets containing mixtures of different elements, such as yttrium, barium, strontium, and copper. The vapors deposit as a film on a heated crystalline substrate that acts as an atomic-scale template to align the film's crystal structure.

ORNL's new laser MBE systems offer two advantages over conventional laser ablation equipment: better control of growth at the atomic level and a cleaner environment for growing highly pure crystalline films. Improved control is achieved through new capabilities. Now, it's possible to "tune" energies of incident atoms and ions, control substrate temperature from below room temperature to 900°C, and rotate the substrate to make film thickness and composition more uniform. ORNL is developing a new type of ellipsometer that will monitor film thickness and optical properties during film growth in gaseous environments, where conventional electron beam monitoring cannot be used. The laser MBE system also provides auxiliary evaporation, ion beam, and plasma sources to assist film growth and doping. The clean, ultrahigh-vacuum-growth environment ensures that very few water vapor or other unwanted gas molecules will be present to contaminate the product.

New laser molecular beam epitaxy systems in a new
research building should produce better
semiconducting and superconducting
materials and greater understanding
of their properties.

ORNL Director Alvin Trivelpiece (center), Associate Director Bill Appleton (right), and researcher Doug Lowndes admire new laser molecular beam epitaxy equipment in the Solid State Division's new building. Photograph by Curtis Boles.

One laser MBE system will be used to grow semiconducting films that contain several different chemical elements (e.g., zinc selenide telluride alloys or copper indium gallium diselenide) and to dope them with deliberately added impurity atoms. The goal: improve the films' electrical, light-emitting, and light-absorbing properties and better understand the physics underlying the improved properties. Applications for such materials include photovoltaic cells for converting solar energy to electricity, flat panel displays to replace bulky computer monitors, higher-density data storage for compact disc players and computers, and perhaps optical computers.

Another laser MBE system is being used to build and study thin films of artificially layered high-temperature superconducting materials. The goal of these experiments is to provide basic data that will allow researchers to relate systematic variations in film structure and composition to changes in their superconducting properties.

While the new laser MBE growth systems were being developed, ORNL researchers used conventional laser ablation growth chambers to simulate and investigate several capabilities of laser MBE. For example, in 1994 they grew "superlattice" structures consisting of alternating strontium-copper-oxide and barium-copper-oxide layers to form two new families of superconductors that do not exist as bulk materials. These new, artificial thin-film superconductors, built of sheets of copper oxide, were shown to be high-temperature superconductors (with superconductivity observed at temperatures as high as 70 Kelvin). The work was reported in Science magazine. In 1995, the growth of highly doped (electrically conducting) zinc telluride semiconductor films was achieved by another group of ORNL researchers and reported in Applied Physics Letters.

The improved capabilities of the new laser MBE systems should make possible even better semiconducting and superconducting materials and greater understanding of their properties. Consequently, ORNL's new research building and laser MBE facilities are expected to help our solid-state physicists build on their previous successes, bit by bit.

The building and new instruments were made possible with funding from DOE, Office of Energy Research, Basic Energy Sciences.

Automotive Catalyst Atoms Observed

Precious metals on a ceramic are helping cars clean up their act. A sprinkling of silvery-white platinum or rhodium atoms on gamma alumina "supports" in catalytic converters can increase the rate of important reactions: oxidizing the major pollutants of auto exhaust#151;nitrogen oxides, carbon monoxide, and hydrocarbons$#151;to harmless nitrogen, carbon dioxide, and water.

No one understands precisely how platinum and rhodium work as catalysts in cars, but use of ORNL-developed Z-contrast microscopy (see images on p. 60) is providing the first atomic-scale glimpse of these important materials. Now, we can see where and how metal atoms sit on the ceramic surface and relate this information to peaking or fading of catalytic action, which chemists measure by passing test gases over catalytic converter materials.

Z-contrast image (above) of a platinum catalyst on gamma alumina (Al2O3) showing individual atoms arranged as trimers. From such data the likely adsorption sites of platinum atoms (three purple balls) on the alumina surface can be determined, as seen in the simulation.

Using a 300-kilovolt scanning transmission electron microscope, we can easily distinguish the heavier catalytic metal atoms from our substrate's lighter atoms of aluminum and oxygen (atoms of higher mass appear much brighter in the image). We observed the structure of small clusters of metal atoms that sit on top of each other like a cheerleader pyramid and reconstructed their three-dimensional form. For platinum, we imaged a mixture of single metal atoms and scattered triplets. For rhodium, we observed numerous 1-atom-thick "rafts," each about 6 to 10 atoms wide (like a cluster of islands in a bay).

Rhodium is now being used more than platinum as an automotive catalyst because it can promote the oxidation of three major pollutants. Unfortunately, it is less stable, and the mechanism responsible for its degradation is not well understood. We have observed an atomic-scale mechanism that can explain the catalyst's degradation: clusters of rhodium atoms diffuse into vacancies that are naturally present in the gamma alumina lattice.

Thanks to Z-contrast microscopy, we can explain degradation in an automotive catalyst at the atomic level.

Images of a rhodium catalyst supported on gamma alumina taken with the 300 kilovolt scanning transmission electron microscope. In the phase contrast high resolution image (left), the type of image normally taken with conventional microscopes, the rhodium is invisible. However, the rhodium atoms are seen clearly in the Z-contrast image (center) and in pink in the colorized form (right). The arrangement of the rhodium atoms corresponds to no known form of metal or oxide, but does correspond to normally vacant atomic sites in the alumina. The rhodium is thought to have dissolved into these normally empty sites, where it will no longer be catalytically active. This image reveals clearly for the first time how a catalyst can lose its activity through aging, limiting its automotive applications.

If funding is available through collaboration with the automobile industry, we hope to conduct more research to identify the active sites of the catalyst metal atoms and the chemical reactions they promote. Because rhodium is so expensive, it would be desirable to maximize its efficiency and minimize its degradation, to make catalytic converters effective for the long haul. Alternatively, a less expensive metal could perhaps be made to work as well. Our studies could help the auto industry and government nail down this information as they act together to design a clean, efficient, and affordable car.

The research was supported by ORNL's Laboratory Directed Research and Development Fund.

Linking Welding Process with Weld Properties

ORNL's welding and joining researchers have been making connections—between the welding process and weld microstructure and properties and with scientific and industrial experts. As a result, our group is setting trends in welding research and development. We are internationally known as the leader in understanding and predicting relationships between weld microstructure and properties.

Ripples, or surface undulations, are formed during the stationary arc welding of a single crystal of austenitic stainless steel, as shown in this interference-contrast micrograph.

A stainless-steel weld has a microstructure that gives rise to properties such as strength and toughness. When a weld is exposed to high temperatures over a long time, its microstructure changes, often making it brittle. Brittle welds have forced costly shutdowns of petrochemical and power plants.

We have provided basic information on the aspects of microstructure that control fracture toughness in austenitic stainless-steel welds at low temperatures. Our research goals are to determine desirable microstructures for different weld materials, predict properties of weld microstructures transformed by heat over time, and develop intelligent automated techniques for controlling the welding process, to obtain welds that give the best properties.

We find and predict relationships between the welding process and weld microstructure and properties.

Using data from our experiments with single-crystal stainless-steel welds, we have developed analytical and computational models that describe relationships between the welding process and microstructure of welds. Applying this knowledge, we have been working with Westinghouse Electric Corporation on ways to repair gas-turbine-engine components made of single-crystal, nickel-based superalloys. These components are critical to the operation of turbine engines in jet aircraft and power plants.

We are also studying the role of oxide inclusions in improving or degrading properties of steel welds. Oxygen that has been dissolved in molten steel forms oxide inclusions with the steel's residual elements, such as aluminum, manganese, silicon, and titanium. As the steel cools and solidifies, these oxide inclusions are frozen into the microstructure. If a steel weld has the "right" number and distribution of oxides and the "right" overall composition, the weld will contain good microstructural constituents that would make the weld tough. If not, it may have poor mechanical properties. We have developed a model to predict a weld's inclusion characteristics and microstructure based on oxide inclusion composition, number density, and size distribution.

We hope our knowledge of welding can be applied to the design of lightweight, efficient cars built from aluminum parts for the Partnership for a New Generation of Vehicles. In the meantime, our group has received new recognition: The Institute of Materials in London, England, has founded the first scholarly journal in the fieldScience and Technology of Welding and Joiningand named ORNL's Stan David the inaugural editor-in-chief. To advance welding research, we hope to continue to make the right multidisciplinary connections.

The research is sponsored by DOE, Office of Energy Research, Division of Materials Sciences.

Improving Auto Brakes With Neutrons

You're driving a Ford Taurus when a car veers into your lane. You slam on the brakes, stopping in time and avoiding a collision. As usual, the brakes are reliable, but because they vibrated that time, you take the car to the dealer to have them checked.

You're told these vibrations are more annoying than serious. But because service personnel frequently must deal with the problem, one of the Ford Motor Company's goals is to reduce the costs of servicing brakes—to the company and the owner—over the lifetime of each vehicle.

The Ford Motor Company analyzes brake rotors
at ORNL's Residual Stress User Center.

To help accomplish this goal, Ford researchers conducted studies at ORNL's Residual Stress User Center. This center provides neutrons and X rays to probe materials for residual stresses—internal stresses that develop and remain in a material as a result of manufacturing and other forces. The Ford researchers wanted to better understand the effect of residual stresses on disc brake rotors. These discs spin when the car moves until brakes are applied, hydraulically pressing pads against the discs to slow or stop them—and the car.

Bill Donlan, a Ford Motor Company researcher, prepares to use neutrons from ORNL's High Flux Isotope Reactor to map residual stresses in automobile brake discs. Photograph by J.W. Nave.

It was believed that when disc brake rotors overheated, changes in residual stresses in the discs caused them to distort, prompting vibrations during braking. The researchers wanted to know if heat treatment of rotors would reduce distortions and the probability of vibrations.

The Ford researchers brought to ORNL a standard rotor for measurement. They then heat treated it and brought it back to ORNL for comparison measurements. The goal was to determine the relief of stresses by heat treatment. The residual stresses in the rotors were mapped using neutron diffraction at our HFIR. Residual stresses cause distances between planes of atoms in a crystalline material to shrink or stretch compared with normal lattice spacings in stress-free areas. Because varying distances between atomic planes affect the angles at which neutrons are diffracted or scattered, lattice strains resulting from residual stresses can be precisely located and measured.

Neutron stress mapping studies showed that standard rotors have significant stresses that can be reduced by heat treatment. The data should help Ford modify the brake manufacturing process. Ford may soon be putting a stop to stresses that could affect brakes — and drivers.

The research is sponsored by DOE's Office of Energy Efficiency and Renewable Energy, Office of Transportation Technologies, as part of the High Temperature Materials Laboratory User Program. The High Flux Isotope Reactor is sponsored by DOE's Office of Energy Research.

Next article