Interconnection technology is a key factor in the continual advancement of integrated systems. The rapid increase in device density and circuit complexity through scaling demands a similar increase in the interconnection density. Traditionally, this is achieved by reducing the metal pitch as well as gradually increasing the number of interconnection levels. As the width and spacing of interconnections are scaled down to submicron dimensions at the chip level and micron dimensions at the board level, signal delay, crosstalk, electromigration, and stress-induced migration become important concerns.

Cu holds promise as an alternative metallization material to Al alloy due to its low resistivity and ability to reliably carry high-current densities. Cu has a bulk resistivity of 1.68 μΩ-cm, whereas Al has a bulk resistivity of 2.65 μΩ-cm. The only metal with a resistivity lower than Cu is Ag. Since Cu has a melting point and atomic weight both higher than Al, it is expected to have better resistance to electromigration, although properties such as grain structure and resistance to corrosion at high temperatures may also affect electromigration characteristics.

Recently, there have been a number of reports identifying technologies of strategic importance. These technologies, which reflect the full range of national critical technology needs, are: materials, manufacturing, informatics and computing, biotechnology/life sciences, aeronautics/surface transport and energy/environment. In support of these technologies there has been much discussion on their research infrastructure, e.g., instrumentation, telecommunication networks, and supercomputing facilities. With the exception of biotechnology/life sciences, however, there has been little discussion on the uses of evaluated numeric and factual databases at the research level. The uses of databases are more advanced in biotechnology and life science research than in other fields, and this has been driven by the needs of genetic research, protein engineering, and drug design, where researchers need data and models for the design of new products. The needs of databases and their manipulation tools in materials science research are also essential in developing an intelligent research infrastructure. Given the present financial constraints, there is a need to use existing funds more efficiently and effectively. One way to achieve this is to use all available experimental data from various intersecting disciplines and bind them together with knowledge which will aid in the design of new materials.

Many attempts have been made in recent centuries to investigate friction, adhesion, lubrication, and wear. Most of the experimental approaches and theories were based on macroscopic experiments, such as tensile and indentation tests. For a long time, only the bulk properties of the materials were considered.

Late in this century a new term was created combining all of the above-mentioned properties which deal with the science of interacting material interfaces in relative motion: tribology. The state of the art of science today reveals that processing in nature depends strongly on interfaces that cannot be described only by bulk properties. Tribologists realize they must study the sliding surfaces by analytical surface-science tools. With the surface force apparatus developed by J.N. Israelachvili and D. Tabor, we have a surface analysis tool that provides new insight into the field of macroscopic sliding contact of lubricated systems.

After Amontons' laws were established as a first attempt to describe sliding friction analytically, theories were advanced over the course of this century. A classic discipline was developed: contact mechanics. More quantitative treatments of friction were developed by various authors. The energy dissipation in most processes in tribology induced the theorists to consider the sliding bodies as spring models creating phonon-phonon interactions. And with modern computer facilities, they started to perform computational experiments whenever classical experiments could not provide information on the submicron scale.

Structural materials that can withstand oxidizing and aggressive environments at temperatures above 1000°C constitute an enabling materials technology for a wide range of applications in the industrial, aerospace, and automotive arenas. A few of the industrial uses for such materials are furnace elements and components, power generation components, high-temperature heat exchangers, gas burners and igniters, and high-temperature filters. Aerospace applications include turbine aircraft engine hot-section components such as blades, vanes, combustors, nozzles, and seals. Automotive applications involve components such as turbocharger rotors, valves, glow plugs, and advanced turbine engine parts.

There is increasing interest in silicide-based compounds for high-temperature structural uses under oxidizing conditions in the range of 1200–1600°C. In this temperature range, for oxidation and strength reasons, the choice of materials is limited to the silicon-based structural ceramics such as Si3N4 and SiC, and the new class of “high-temperature structural silicides.” An extensive survey of progress in the area of high-temperature structural silicides has recently been published.

Solar cells formed with thin silicon active layers (<100 μm thick) offer advantages over thick ingot-based devices. The advantages come in two forms: the first is the potential for higher conversion efficiency than that of conventional thick devices, and the second is a reduction in material requirements. The use of thin polycrystalline silicon for solar cells offers the potential of capturing the high performance of crystalline silicon while achieving the potential low cost of thin films. Experimental and theoretical studies initially uncovered the issues of grain size and thickness as limiting factors. Subsequent work added the issue of back-surface passivation. This article addresses the conditions required for the successful development of polycrystalline silicon into a high efficiency, low-cost, terrestrial product.

The prediction of new materials is, in theory, a completely straightforward problem. The fundamental equations of quantum mechanics for solids are well-established and can be numerically solved, at least approximately, by powerful first-principles methods developed over the last decade. It remains, then, only to examine the solutions to find those systems that will exhibit the properties desired.

The number of possibilities to be considered and sorted is many orders of magnitude larger than can be managed in a case-by-case first-principles analysis. Therefore, in practice, the approach to theoretical prediction of new materials is to establish easy-to-apply rules for screening large numbers of candidate stoichiometrics (which might or might not be known compounds). These rules are based on physical understanding of the occurrence of the structure or property, inference from known examples, and general principles governing chemical trends in the structure and stability of known compounds. Data from the scientific literature on the occurrence and structure of crystalline compounds are now organized into a number of crystallographic databases, providing a unique opportunity to investigate, from a global perspective, the factors that influence structure and stability. To display, access, and extract general principles from this enormous amount of information, a graphical method such as the “quantum diagram” method is essential, and can be thought of as providing a convenient “roadmap” for navigation around the database. Useful conclusions about individual compounds or small sets of compounds can also be drawn from this type of analysis. These conclusions form the basis for rules for predicting new representatives that direct attention to special regions on the roadmap, as illustrated on the cover of this issue. Systems that satisfy these rules are the most likely to reward experimental investigation.

Highly absorbing ceramics can be heated in a standard multimode microwave oven. However, most of the microwave energy is dissipated in their ceramic surfaces and heat is transported slowly throughout the sample volume by the usual thermal conduction mechanisms. As with conventional methods, this process leads to non-uniform heating. In contrast, ceramics that are fairly transparent to microwave energy are difficult to heat in standard microwave ovens. When they can be heated, reverse temperature gradients often develop. Reverse temperature gradients can be beneficial in some processes, e.g., combustion synthesis, chemical vapor infiltration (CVI) and binder burnout, but in most ceramic processes, temperature gradients are undesirable. These characteristics have presumably been partially responsible for the ceramic industry's failure to examine microwave processing more closely.

The ability of dipoles to respond to the oscillating electric field and their subsequent inability to keep up with rapid reversals in the field result in dielectric losses that are manifested as heat. While some of the mechanisms by which electromagnetic radiation interacts with ceramics are well known, interaction mechanisms in the microwave frequency range are less well understood. Since several mechanisms, including electrical conductivity, can operate simultaneously, the total energy dissipation and heat generation depends on the effective loss factor. The effective loss factor (ε″eff) is described by

where ε′ is the dielectric constant and is a measure of the extent of polarization, and tan δ indicates how responsive the dipoles are to the oscillating field.

There is a tremendous demand for improved performance and speed in consumer electronics that is likely to continue as new applications and developments occur. This demand necessitates a reduction in the critical dimensions and an increase in the density of devices in microelectronic circuits. As a result, new materials must be considered for integration into microelectronics technology. In particular, the metal wiring or interconnects that connect different components in silicon-based semiconductor devices is a subject of great interest. As the dimensions of transistors shrink below the 0.5 μm level, their speed will become limited by the delays in the existing interconnect material, Al-Si-Cu alloy (p ~ 3 μΩ cm). Therefore, to avoid problems associated with RC (“resistance/capacitance”) time delays and voltage drops, it will be necessary to construct interconnections of materials that possess lower resistivities, resistance to electromigration and hillock formation, and resistance to diffusion into other materials (see Table I).

A number of materials are possible candidates to replace the Al-Si-Cu alloy, including W, Ag, Au, and Cu. Tungsten has excellent resistance to electromigration and hillock formation, but has higher resistivity compared with the Al-Si-Cu alloy. Thus, applications of W are likely to be found where short interconnection distances are necessary. Silver has the lowest resistivity of all metals, but is easily corroded and diffuses rapidly into many materials used in semiconductor devices. However, some specific applications for silver are viable, such as the formation of contacts on ceramic superconductors. Gold has a lower resistivity than the Al-Si-Cu alloy and is inert to chemical corrosion. As a result Au is used where device reliability is the primary concern-for example, for wiring in GaAs-based semiconductors and electrical contacts in packaging.

Smart materials combine sensors, intelligence, and actuators to allow a material to respond to its environment. Magnetostrictive materials can be used as both the sensors and actuators in such materials. High-power magnetostrictive actuators can deliver forces greater than 50 MPa with strains of up to 0.6%, while other magnetostrictive sensor materials can provide hundreds of times the sensitivity of semiconductor strain gages. Magnetoelastic materials also have adaptable elastic moduli which may be varied by external magnetic fields.

Magnetostriction is the change in any dimension of a magnetic material caused by a change in its magnetic state. In this article we concentrate on ferromagnetic materials exhibiting Joule magnetostriction, which is a change in linear dimension parallel to an applied magnetic field (see Figure 1), and the reciprocal effect in which the material changes its magnetic state under the influence of applied stress.

The phenomenon of magnetostriction has been known for well over a century, since Joule discovered in 1847 the change in length of an iron rod when magnetized. The modern era of magnetostrictive materials began in 1963 with the measurement of nearly 1% magnetostrictive strains at low temperatures in the basal planes of Dy and Tb. A search for magnetostrictive materials with high magnetostriction at room temperature led to the alloying of rare earths with transition metals, culminating in the discovery in 1971 of giant room-temperature magnetostriction in the Laves phase compound TbFe2.

A modern integrated circuit (IC) is composed of 106 or more electronic devices. They are connected to form a circuit through the use of metallic films patterned into strips which function as wires to interconnect devices. These wires are usually simply referred to as interconnects. In an IC occupying the surface of a 1 cm2 Si chip, there can be 10 m of total interconnect length. This length is in the form of more than 106 line segments contacting pairs of devices and different segments of the circuit. This enormous number of wires is made possible by their small widths. Interconnect widths as small as 0.55 μm are currently used in commercial circuits, and circuits and processes leading to smaller and smaller widths are continuously in development.

During operation of an IC, interconnects carry current densities as high as 4 × 105 A/cm2. This should be compared with a current density of 102 A/cm2, the maximum allowed for house wiring. Thin-film conductors can carry these high current densities only because of the relatively good heat sinking provided by the Si substrate.

Professor David Kingery, author of the classic text, Introduction to Ceramics, said it well: “Corrosion, like death and taxes in our existence, is inevitable for all materials.” Metals, ceramics, and polymers all undergo change when exposed to the environment. Metals can exhibit electrochemical corrosion, while ceramics and plastics experience primarily chemical corrosion. Although “corrosion” often is used to describe the environmental stability (or instability) of metals, a host of other definitions have been used to characterize the changes that occur in ceramics and polymers.

Definitions can become an issue when scientists from various disciplines work together to solve a problem of national/international importance. One such example involves the description of the surface changes that occur when nuclear waste forms are exposed to an aqueous environment (as might occur in a geological repository). The observed surface changes have been referred to as corrosion, leaching, weathering, degradation, and alteration, to name a few. In early symposia on nuclear waste management, sponsored by the Materials Research Society and the American Ceramic Society, definitions and technical issues were discussed extensively by materials scientists, chemists, physicists, geologists, nuclear engineers, and other professionals. Because of this interdisciplinary interaction, significant advances were made (in spite of disagreements over terminology) in understanding the aqueous alteration of these materials. Much of this knowledge will be useful in other applications where environmental stability is important.

Although aqueous alteration is an important concept in materials corrosion, the topic of environmental stability is much broader. Other important subtopics are erosion, thermal stability, oxidation and radiation damage, and molten metal, glass, and salt corrosion. Finally, environmental instability need not always be bad. Indeed, there are numerous processes and products that work only because of the material's instability in specific environments.

Understanding the atomistic mechanisms, energetics, structure, and dynamics underlying the interactions and physical processes that occur when two materials are brought together, separated, or rub against each other (hence the term tribology, from the Greek tribos, meaning to rub) is fundamentally important to many basic and applied problems. Examples include adhesion, capillarity, contact formation, surface deformation, elastic and plastic response characteristics, hardness, micro- and nanoindentation, friction, lubrication, wear, fracture, atomic-scale probing, and modifications and manipulations of materials surfaces. These considerations have for over a century motivated extensive theoretical and experimental research into the above phenomena and their technological consequences.

Explorations of materials systems and phenomena in the nanoscale regime often require experimental probes and theoretical and computational methods that allow investigations with refined spatial, as well as temporal, resolution. Consequently, until recently most theoretical approaches to the above issues, with a few exceptions, have been anchored in continuum elasticity and contact mechanics. Experimental observations and measurements of surface forces and the consequent materials response to such interactions have been macroscopic in nature.

It has been well-recognized for many years that the particle-size distributions of the cement and the grading of the aggregates play an important role in determining the properties and characteristics of cement and concrete products. DSP (densified with small particles) type cements and concretes, to a certain extent, MDF (macro-defect-free) cements, and optimized concretes are recently recognized outstanding examples of the application of this principle. The preset characteristics of the cementitious slurry are also strongly influenced by these factors. Both the workability of the fresh material, and the microstructure development are controlled to a considerable extent by these geometric parameters.

Two seminal works in the areas of continuous particle size distributions and particle packing are those of Andreason and Furnas, respectively. Furnas deals mainly with discrete systems and Andreason with continuous distributions. As early as 1907, the concept of idealized particle packing was being used to optimize cements and concretes. Figure 1a shows an idealized cross section of a simple cubic packing of monodispersed spheres. This system has a maximum packing density of 0.65%. In an ideally packed system of discrete size ranges, the size of the next smallest particles would be such that they just fit in the gaps between the largest size particles, and so on for subsequent particle sizes; this system is represented schematically in Figure 1b. Not only the sizes but also the relative numbers of particles are important; Figures 1c and 1d show systems where some fraction of the smaller and larger particle sizes, respectively, are missing. Figure 1e shows a system where the size of the second largest particles is too large to fit into the gaps between the largest particles, resulting in a lower packing efficiency. Thus, both the particle size and fractions are important when considering packing efficiency.

In single-crystal organic salts, we find a keen competition between superconducting, magnetic, insulating, and metallic states. The physics of these materials is further enriched by the sensitivity of these states to pressure, temperature, chemical formulation, and magnetic field. A growing international community of scientists have turned their attention to these materials, and are applying the techniques and theories of metal and semiconductor physics to probe these new systems. In this article we will explore these materials. We will discover that these materials have given us many new things: a renaissance in fermiology, new high-magnetic-field states of matter, a bulk quantum Hall effect, new challenges in the calculation of energy bands on a small energy scale, and elusive behavior which seems one step away from our present understanding of physics in low dimensions. Electron correlations probably play an important role in determining the phenomena, and should be considered in more microscopic theoretical treatments of these systems.

A big barrier impeding the expansion of large-scale power generation by photovoltaic (PV) systems was the high price of solar cell modules, which was more than $50/Wp (peak watts) by 1974. Therefore, cost reduction of solar cells is of prime importance. To achieve this objective, tremendous R&D efforts have been made over the past ten years in a wide variety of technical fields, from solar cell materials, cell structure, and mass production processes to photovoltaic systems. As a result, more than an order of magnitude in cost reduction has been achieved, and the module cost has come down to less than $5/Wp in a firm bid for the large-scale market. Two phases of technological innovation can be identified. The first innovation in progress is based on low-cost polycrystalline technologies applicable to well-developed single-crystalline silicon solar cell fabrication processes. The second remarkable innovation is a-Si:H (hydrogenated amorphous silicon) technology, which we will discuss.

We open our discussion with a brief overview of the present status of a-Si solar cell R&D efforts, with some new insights in device physics. Next, we discuss some new approaches and key technologies for improving solar cell efficiency with stabilized performance using new materials such as a-SiC:H (amorphous silicon carbide), μc-SiC:H (microcrystalline silicon carbide), and a-SiGe:H (amorphous silicon germanium). Also, the progress of conversion efficiency in various types of amorphous silicon solar cells is surveyed and summarized.

Piezoelectric and electrostrictive actuators, capable of moving something electromechanically, are forming a new field between electronic and structural ceramics. Application fields are classified into three categories: positioners, motors, and vibration suppressors. The manufacturing precision of optical instruments such as lasers and cameras, and the positioning accuracy for fabricating semiconductor chips, which must be adjusted using solidstate actuators, is of the order of 0.1 μm. Regarding conventional electromagnetic motors, tiny motors smaller than 1 cm3 are often required in office or factory automation equipment and are rather difficult to produce with sufficient energy efficiency. Ultrasonic motors whose efficiency is insensitive to size are superior in the minimotor area. Vibration suppression in space structures and military vehicles using piezoelectric actuators is also a promising technology.

New solid-state displacement transducers controlled by temperature (shape memory alloy) or magnetic field (amorphous magnetostrictive alloy) have been proposed, but are generally inferior to the piezoelectric/electrostrictive actuators because of technological trends aimed at reduced driving power and miniaturization.

Atomic force microscopy (AFM) was invented in 1986 by Binnig, Quate, and Gerber as “a new type of microscope capable of investigating surfaces of insulators on an atomic scale.” Stemming from developments in scanning tunneling microscopy (STM), it became possible to image insulators, organic and biological molecules, salts, glasses, and metal oxides — some under a variety of conditions, e.g., ambient pressure, in aqueous or cryogenic liquids, etc. In 1987, Mate and co-workers introduced a new application for AFM where atomic-scale frictional forces could be measured. Likewise, in 1989, Burnham and Colton used the AFM to measure the surface forces and nano-mechanical properties of materials. Today, there are many examples of using AFM as a high-resolution profilometer, surface force probe, and nanoindentor. Several new imaging techniques have been introduced; each depending on the type of force measured, e.g., magnetic, electrostatic, and capacitative. Because of the diverse nature of the field and instrumentation, the names “scanned probe microscopy” and “XFM” (where X stands for the force being measured, e.g., MFM is magnetic force microscopy) have been adopted.

The problem of stress-induced voiding in the metal lines used for VLSI interconnections appeared about 10 years ago and was first reported in the literature in 1984. Some typical voids are shown in Figure 1. It is not entirely coincidental that the problem is of about the same age as the personal computer. The continuing advances in process technology that make possible the ever lower costs and improved performance of VLSI devices often have unexpected and troublesome side effects. Continuous attention to potential effects of process changes on the behavior of the materials comprising a VLSI chip is essential to avoid passing on problems to customers. The history of stress-induced voiding provides a good illustration.