Aluminum and aluminum alloys have long been among the preferred materials for ultrahigh vacuum (UHV) systems operating in the 10−10–10−11 torr (10−8–10−11 Pa) range. Pure aluminum and aluminum alloys have an extremely low outgassing rate, are completely nonmagnetic, lack crystal structure transitions at low temperatures, are not sources of heavy metals contamination in semiconductor processing applications, have low residual radioactivity in radiation environments, and are lightweight. Because of aluminum's high thermal conductivity and low thermal emissivity, aluminum components can tolerate high heat fluxes in spite of the relatively low melting point of aluminum.

Recently developed aluminum alloys and new surface finishing techniques allow the attainment of extremely high vacuums (XHV) in the 10−12–10−13 torr (10−10–10−11 Pa) range. XHV technology requires the use of special aluminum alloy flange/gasket/bolt, nut and washer combinations, aluminum alloy-ceramic seals, windows, bellows, right-angle and gate valves, turbomolecular pumps, sputter ion pumps and titanium sublimination pumps, Bayard-Alpert ion gauges, quadrupole mass filters, and related aluminum alloy vacuum components. New surface treatment methods and new techniques in welding and extremely sensitive helium leak testing are required. In short, a whole new technology has been developed to take advantage of the opportunities presented by these new vacuum materials. This article describes some of these newly developed fabrication technologies and vacuum materials.

The TRISTAN electron-positron collider constructed at the National Laboratory for High Energy Physics in Japan is the first all-aluminum alloy accelerator, and the first to use UHV technology.

Multilayers, as considered in the following articles, are manmade thin-film materials periodic in one dimension in composition or in composition and structure. This composition/structure variation is generated during synthesis, which is typically accomplished using atom-by-atom technologies. Individual component layers in a multilayer may vary in thickness from one atomic layer (~2 Å) to hundreds of atomic layers (~1,000 Å) of a given material.

An example of these synthetic micro-structures, a lattice image transmission electron micrograph of a cross section of a hundred period titanium (63 Å)/titanium-nickel (40 Å) multilayer microstructure fabricated using magnetron sputtering, is shown on the cover of this issue of the MRS BULLETIN. The titanium-nickel layers are amorphous as a result of the low substrate temperature (<75°C) and the very large atomic quench rates characteristic of vapor deposition (>1012 K/s). The elemental titanium layers are fiber textured with the basal plane of this hexagonal close-pack structure element in the plane of the layers. These (00.1) planes are the ones lattice imaged in this electron micrograph.

The general concept of a multilayer structure, as illustrated above, is now well accepted because the ability to synthesize such materials for scientific study and technological application has been demonstrated at many nationally and internationally based laboratories.

This is the second of two issues of the MRS BULLETIN on interface materials and is entirely devoted to their mechanical and high-temperature behavior. Mechanical properties provide a rich area for investigating the effect of the local inhomogeneities near the interfaces, and their effect on the interrelation between the structure and chemistry on one hand, and the elastic and fracture behavior on the other.

Based on much experimental work on grain-boundary fracture it seems that, with the exception of “beneficial” segregants, the embrittlement potential of most impurities is governed by their propensity for segregation to the grain boundaries, which in turn is strongly influenced by the energies of the pure boundaries. To investigate the role of the grain-boundary structure in its fracture behavior, one must therefore consider the correlations between (1) the structure (i.e., the five macroscopic degrees of freedom) and the energy of pure grain boundaries, (2) impurity segregation and the grain boundary energy, (3) structure, impurity segregation and elastic response at the interface, and finally (4) the correlation between embrittlement and segregation. In addition, the mobility of dislocations near a crack tip also plays an important role. Unfortunately, relatively little knowledge has been accumulated on most of these complex interrelations even though their unraveling is widely recognized as the ultimate goal.

Dr. D. Allan Bromley, President Bush's science adviser, is firmly committed to the administration's gaining a better understanding of materials science and engineering and then going forth with a cohesive program from there. The MRS BULLETIN interviewed him on July 17 in his office in the Old Executive Office Building. Following is an excerpt of that interview.

BULLETIN: The Department of Commerce recently issued a report that identified advanced materials as by far the leading emerging technology. Is special consideration being previded to spur this critical technology sector?

BROMLEY: First of ail, we are preparing so that next year we will work with the Office of Management and Budget to develop a coherent federal program that spans ail the agencies together to get a coherent look for the first time at federal activity in materials science and engineering. I have long thought that was an orphan area, not only in government but in many of the universities, because it doesn't fit into department or agency boundaries.

At the same time, the President's Council of Advisers on Science and Technology, chaired by Ralph Gomory, the former chief scientist at IBM and now president of the Sloan Foundation, will be focusing on advanced materials and manufacturing as two very important areas that bear on economie competitiveness.

Third, the new National Critical Materials Council will be focusing specifically on advanced materials and will be comparing the program planned for research in that area next fall.

The federal budget process is often regarded as an arcane affair, the province of a select group of Washington insiders. This includes the mysteries behind the determination of budgets for R&D on materials. The purpose of this discussion is to demysrify the subject a little. It will focus largely on the role of Congress.

A common misconception is that the White House sets the budget for the federal government. What the White House does do is send detailed recommendations to Congress each year in January on what it believes the budget should be. This covers spending, tax policy, and economie predictions (the latter reflected in the anticipated tax revenues). In January 1990, for example, the administration submitted budget recommendations for fiscal year 1991, which began on October 1 and runs until September 30, 1991. These recommendations (with subsequent negociations with the Hill) can be quite influential, but serve primarily as a point of reference, which Congress can gladly accept, or completely ignore, as circumstances and preferences dictate.

This issue of the MRS BULLETIN is devoted to high Tc superconductivity. It is the sequel to a previous series of articles on the same subject which appeared in the MRS BULLETIN in January 1989. While the articles in the January 1989 issue emphasized the families of high Tc superconducting oxides known at that rime, as well as novel processing techniques and thin films, the papers in this issue focus on the physical properties of high Tc oxide superconductors.

The quality of polycrystalline and single-crystal bulk and thin-film materials has improved to the point where researchers can now make reliable measurements of many physical properties representative of the intrinsic behavior of these materials. As a result, a broad spectrum of important issues such as the nature of the electronic structure, the type of superconducting electron pairing, the magnitude and temperature dependence of the superconducting energy gap, the behavior of fluxoids in the vortex state, etc., can be addressed meaningfully. Presently emerging is a consistent picture of the physical properties of the high Tc oxides, which will form the foundation to eventually developing an appropriate theory for the normal and superconducting states of these remarkable materials.

I am particularly pleased to receive the Von Hippel Award and to do so in the city of Boston. I spent 24 happy years here at the Lincoln Laboratory of MIT, and Prof. Von Hippel was one of those on campus who made a special effort to build a bridge of communication to those of us who were out on Route 128. I would like him to know that his personal encouragement and his welcome to campus life were very much appreciated. During that period, it was Prof. Von Hippel who coined the phrase “molecular engineering” to describe the whole process of materials science. So it is in his honor that I have chosen as the title of my talk, “The Molecular Engineering of Oxides.”

Back in the 1950s, I was part of the group at Lincoln Laboratory that developed the ferrite-core memory for the digital computer. This task involved an identification of the optimum ferrospinel composition and the development of a fabrication and testing process that would give a high yield of cores with acceptable physical properties. Its realization involved, as well, the development of a theoretical underetanding of the factors that determine the shape of a B-H hysteresis loop and of why an exquisite annealing procedure was required to develop controlled chemical inhomogeneities via the dynamic Jahn-Teller effect. This experience made it clear to me that there was an urgent need to build bridges between the engineer and the materials scientist.

The original intention was to publish one special issue on the materials science of fine particles. As a Guest Editor, I approached a number of specialists with specific interests in this field, soliciting contributions. Normally, an editor expects a number of requests to be declined. In this particular instance the response was overwhelming, which indicated both the enthusiasm of the people active in fine particle science and the timeliness of the topics. Thus, the MRS BULLETIN decided to dedicate two issues to this subject, December 1989 and January 1990.

The December 1989 issue dealt essentially with the procedures and principles for preparing finely dispersed matter consisting of particles of different chemical composition, shape, and size. This issue treats the problems of the mechanisms of formation of well-defined dispersions and offers examples of applications in two areas — ceramics and adhesion. Only the limitations in space prevented the editor from including other topics relevant to fine particle science and technology, which are both fascinating and useful.

Optical storage of digital information has reached the consumer market in the form of the compact audio disk. In this technology, information is stored in the form of shallow pits embossed in a polymer surface. The surface is coated with a reflective thin metallic film, and the digital information, represented by the position and length of the pits, is read out optically with a focused, low-power (5 mW) laser beam. When used for information storage for a computer this device is called a CD-ROM, a Compact Digital-Read Only Memory. The user can only extract information (digital data) from the disk without changing or adding any data. That is, it is possible to “read” but not to “write” or “erase” information.

While it is an advantage to have permanently stored information in some cases — for example when listening to Beethoven's Ninth Symphony—in other situations the read-only feature is not appropriate. For most computer applications, it is essential that the user be able to store information on the disk and read it back at will. For example, in a word processing task—such as typing this article — it is often necessary to store the document on a disk. Optical data storage for this purpose is available in the form of a Write Once Read Many times (WORM) optical disk drive. The operating principle in a WORM drive is to use a focused laser beam (20 – 40 mW) to make a permanent mark on a thin film on a disk. The information is then read out as a change in the optical properties of the disk, e.g., reflectivity or absorbance. These changes can take various forms: “hole burning” is the removal of material (typically a thin film of tellurium) by evaporation, melting or spalling — sometime s referred to as laser ablation; bubble or pit formation involves deformation of the surface, usually of a polymer overcoat on a metal reflector.

Some of the most important properties of materials in high-technology applications are strongly influenced or even controlled by the presence of solid interfaces. For example, interfaces are the critical element in fiber-reinforced structural ceramics with mechanical properties not even imagined a decade or two ago. The entire electronics industry is based on the fascinating electrical properties of semiconductor interfaces, with ceramic-semiconductor, metal-semiconductor and metal-ceramic interfaces playing critical roles as well. Other examples are surface modification techniques, designed to enhance the corrosion resistance of materials in hostile environments or tailored for tribological or catalytic applications. In contrast to their enormous technological importance, our basic understanding of even the simplest interfaces, such as free surfaces and grain boundaries, is rudimentary at best. It is increasingly recognized, however, that truly significant technological advances can be achieved by a better understanding and control of interfacial processes.

To draw attention to the tremendous opportunites that lie ahead in this lively area of materials research, some recent promising developments in the atomic-level understanding of solid interfaces are highlighted in this issue and in the October issue of the MRS BULLETIN. The theme common to ail the articles in the September and October issues lies in the atomic-level insights that have been gained. While this issue of the BULLETIN focuses on structure, chemistry and some electronic properties of crystalline interfaces, the October issue will be devoted to their mechanical and high-temperature behavior.

An interface can be defined as a surface that serves as a common boundary between two phases. Examples include the boundaries between two solids, two immiscible liquids, a solid and a liquid, a solid and a gas, and a liquid and a gas. Interfaces have been studied for decades by scientists of many different disciplines. One reason for this interest is that the atomic structure and the chemical composition at the interface can differ from that of the bulk material on either side of it. Consequently, the properties of the interface can differ greatly from those of either bulk phase, and chemical reactions can occur more readily at the interface than in the bulk.

All the interfaces listed in the previous paragraph are of interest to materials scientists. However, this article will only consider the grain boundary because it has received the most attention by researchers in materials science. Furthermore, we will only consider grain boundaries in metals; nonmetallic systems will be covered in other articles in this issue.

A grain boundary is an interface that exists where two single crystals are joined in such a way that their crystallographic orientations are not completely matched. Thus, any polycrystalline material contains many grain boundaries. They occur wherever the individual grains meet one another and can usually be observed by etching a polished cross section of the surface as shown in Figure 1. Grain boundaries first form in a metal as a result of the multiple nucleation sites that occur during solidification.

This article is dedicated to a class of specialty materials that have come to be known as low-density microcellular materials, or LDMMs. LDMMs have been developed and studied over the last decade at several U.S. Department of Energy (DOE) National Laboratories. LDMMs are used in various high energy physics programs where they meet unusual and demanding specifications (see Table 1.1).

To set the stage for introducing LDMMs, it is instructive to contrast them with more familiar cellular materials. Examples of commercially available low-density materials are plastic “blown cell” foams, such as polyurethane cushions and polystyrene coffee cups. Asymmetric membranes and filters, on the other hand, are representative of commercially available microcellular materials. The processes used to make these products generally are not suitable for making LDMMs, however, because they are limited by a tradeoff between density and cell size. That is, they yield low densities only at the expense of increased cell size, or small cell size at the expense of increased density. The challenge to DOE materials scientists, therefore, has been to develop new methods for making materials that simultaneously have low density and microcellular morphology, and also meet other application requirements (e.g., composition, homogeneity, size, and strength).

LDMMs are used by the DOE laboratories because their low density and high homogeneity (implied from their microcellular morphology) make them ideal precursors or substrates for forming high-density gases or expanded plasmas.

It is well known that thin-film technology relies increasingly on multilayered structures. As dimensions become smaller, the interfacial or contact region assumes a larger and often dominant role in the performance or properties. Many examples come readily to mind. In magnetic hard disks, the active cobaltalloy layer, itself only about 50 nm thick, is grown either on a crystalline chromium thin film or directly onto amorphous nickel-phosphorous, and capped with a protective carbon or chromium-carbon coating (see Figure 1). The recording head “flies” at 90 mph and about 0.1 ü above this combination, which is expected to be mechanically durable and magnetically reliable for thousands of recordings. Atomic-scale multilayers are being investigated to provide the ability to “tune” the magnetic properties of the active recording layer or head materials. Exchange coupled magneto-optical media consisting of a few tens of angstroms of cobalt or nickel layers on amorphous TbFeCo alloys are showing promise for improving magneto-optical coupling while maintaining perpendicular anisotropy. In microelectronic circuits, aluminum or silicide contacts to silicon are essential to any device, and multilevel integration involving a series of metal, alloy, silicon (amorphous, poly- or monocrystalline) and dielectric layers (some of which might be 1-10 nm thick) are increasingly required to achieve large-scale integration. Metal-metalloid (e.g., MoSi, W-C) multilayers are used for x-ray optical elements. Artificially produced metallic superlattices and multilayers are being used to probe the fundamental magnetic, electronic, mechanical, and structural properties of metal-metal interfaces.

For the past 10 years, reactively deposited films of titanium nitride, TiN, have been applied to cutting tools such as drills, hob cutters, and endmills. A nominal film thickness of 2–4 μm has been shown to give excellent resistance to abrasion and corrosion and to extend tool life three times or more. This is attributable to the physical properties of TiN, which include microhardness of 1,800 kg/mm2 and surface friction approximately one-third that of high-speed tool steel. Corrosion resistance is realized from the dense, fine-grain equiaxed structure of the inert TiN film. Additional applications range from decorative use based on its goldlike appearance to use as a diffusion barrier in semiconductor devices.

More recently, TiN has found application as a high quality coating for components used in ultrahigh vacuum (UHV and XHV) system apparatus and especially in high energy particle accelerators. This article discusses the application of TiN coatings to ultrahigh vacuum systems and high energy particle accelerators.

The native oxides which form on stainless steel and aluminum tend to be porous and trap large amounts of water vapor and other gases. These trapped gases can be partially removed by vacuum baking, although for particle beam devices in which beam-induced desorption is at least as important as the thermal outgassing rate, an extensive beam-conditioning process is required to get rid of the final vestiges of trapped gas. The oxide surfaces have low sticking coefficients for the adsorption of incident gas molecules, but the oxides have much higher secondary electron yields than the clean metals and consequently have very high beam-induced desorption rates.

Magnetic recording can store digital information at an areal density well beyond the fundamental density-limit of optical systems. The central problems in achieving high trans-optical information density with magnetic recording reside primarily in the materials area. In addition to improved magnetic properties, such as head materials with high saturation magnetization and recording media with low noise and high coercivity, tribological considerations impose constraints on the ultimate density that will be attained. The exponential loss of signal due to spacing between the head and recording medium is inherent in all magnetic recording, necessitating the development of durable quasi-contact interfaces.

Magnetic recording is the universal technology for electronic information mass storage. Its presence is ubiquitous as audio tapes, VCRs, floppy disks, computer hard disks, credit cards, etc. Sales of magnetic recording products exceed $50 billion annually, with strong growth projected into the foreseeable future. A steady progression in storage density and corresponding reduction in cost has characterized all phases of magnetic recording; the literature is replete with historical and projective curves showing cost and density numbers. Computer disk file memories have doubled in areal density every 2.5 years for about the past 30 years. Although many competing technologies have been proposed through the years, none has been able to displace or even significantly impact magnetic recording. The first challenge was thermoplastic recording, followed by at least three resurrections of magneto-optic recording, bubbles, semiconductors, and various nonmagnetic optical storage devices.