Since the commercial availability of SiC substrates in 1990, SiC processing technology has advanced rapidly. There have been demonstrations of monolithic digital and analogue integrated circuits, complementary metal-oxide-semiconductor (CMOS) analog integrated circuits, nonvolatile random-access memories, self-aligned polysilicon-gate metal-oxide-semiconductor field-effect transistors (MOSFETs), and buried-channel polysilicon-gate charge-coupled devices (CCDs). In this article, we review processing technologies for SiC.

Oxidation

A beneficial feature of SiC processing technology is that SiC can be thermally oxidized to form SiO2. When a thermal oxide of thickness x is grown, 0.5x of the SiC surface is consumed, and the excess carbon leaves the sample as CO. Shown in Figure 1 are the oxide thicknesses as a function of time for the Si-face and the C-face of 6H-SiC, and for Si. The oxidation rates are considerably lower for SiC than for Si. The oxidation rate of the C-face of 6H-SiC is considerably greater than that of the Si-face. Hornetz et al. have shown that the reason for the slower oxidation rate of the Si-face is due to a 1-nm Si4C4−xO2 (x < 2) layer that forms between the SiC and the SiO2 during oxidation of the Si-face. When oxidizing the Si-face, the Si atoms oxidize first, which inhibits the oxidation of the underlying C atoms that are 0.063 nm below the Si atoms. When oxidizing the C-face, the C atoms readily oxidize first to form CO, with no formation of the Si4C4−xO2 layer for temperatures above 1000°C.

For many years, it was believed that when two isovalent semiconductors are mixed, they will phase-separate (like oil and water) at low temperature, they will form a solid solution (like gin and tonic) at high temperatures, but they will never produce ordered atomic arrangements. This view was based on the analysis of the solid-liquid equilibria at high temperatures and on empirical observation of phase separation at low temperatures. These observations were further rationalized and legitimized by applying the classic (Hildebrand) solution models, which predicted just this type of behavior. These models showed that the observed behavior of the AxB1 − x alloys implied a positive excess enthalpy ΔH(x) = E(x) − xE(A) − (1 − x)E(B) (where E is the total energy) and that this positiveness (“repulsive A-B interactions”) resulted from the strain energy attendant upon packing two solids with dissimilar lattice constants. The larger the lattice mismatch, the more difficult it was to form the alloy. Common to these approaches (“regular solution theory,” “quasiregular solution theory,” “delta lattice-parameter model,” etc.) was the assumption that the enthalpy ΔH(x) of an alloy depends on its global composition x but not on the microscopic arrangement of atoms (e.g., ordered versus disordered). Thus, ordered and disordered configurations at the same compo sition x were tacitly assumed to have the same excess enthalpy ΔH(x). Clearly the option for ordering was eliminated at the outset. While these theories served to produce very useful depictions of the immiscibility of many semiconductor alloys (and continue to guide strategies of crystal growth), they also cemented the paradigm that semiconductor alloys don't order, they just phase-separate. This was true, at the time.

Electroluminescence (EL) is the emission of light generated from the radiative recombination of electrons and holes electrically injected into a luminescent semiconductor. Conventional EL devices are made of inorganic direct-bandgap semiconductors, such as GaAs and InGaAs. Recently EL devices based on conjugated organic small molecules and polymers have attracted increasing attention due to easy fabrication of large areas, unlimited choice of colors, and mechanical flexibility. Potential applications of these organic/polymeric EL devices include backlights for displays, alphanumeric displays, and high-density information displays.

Electroluminescence from an organic material was first demonstrated in the 1960s on anthracene crystals by Pope et al. at New York University. Subsequently several other groups also observed this phenomenon in organic crystals and thin films. These organic EL devices had high operating voltages and low quantum efficiency. Consequently they did not attract much attention. In 1987 a breakthrough was made by Tang and VanSlyke at Eastman Kodak who found that by using multilayers of sublimated organic molecules, the operating voltage of the organic EL devices was dramatically reduced and the quantum efficiency was significantly enhanced. This discovery touched off a flurry of research activity, especially in Japan. The Japanese researchers, as welt as the group at Kodak, have since improved the device efficiency and lifetime to meet commercial requirements. This progress is reviewed by Tsutsui in this issue.

The following is an edited version of the David Turnbull Lectureship address, given by recipient Robert E. Newnham at the 1996 MRS Fall Meeting. Newnham received the lectureship for “pioneering the field of ceramic composites for electronic and optical applications, and in recognition of a distinguished career of guiding students, lecturing, and writing.” Newnham is the Alcoa Professor of Solid State Science at The Pennsylvania State University.

Low-dielectric-constant materials (k < 3.0) have the advantage of facilitating manufacture of higher performance integrated-circuit (IC) devices with minimal increases in chip size. The reduced capacitance given by these materials permits shrinkage of spacing between metal lines to below 0.25 μm and the ability to decrease the number of levels of metal in a device. The technologies being considered for low-k applications are chemical vapor deposition (CVD) or spin-on of polymeric materials. For both types of processes, there are methods and materials capable of giving k < 3.0 dielectric stacks. This article will focus on the spin-on approach and discuss the properties of both organic and inorganic spin-on polymers.

While CVD SiO2 has been the mainstay of the industry, spin-on materials are appropriate for many dielectric applications. Polyimides have applications as electrical insulators, and traditional spin-on silicates or siloxanes (k > 3.0) have served as planarizing dielectrics during the last 15 years. The newer spin-on polymers have greatly enhanced mechanical, thermal, and chemical properties, exhibiting lower dielectric constants than the traditional materials.

The mid 1980s benefited from three major scientific announcements that altered our world view on the structure of matter and its properties. These were the discovery of quasiperiodic crystals (QCs), which are commonly referred to as quasicrystals (1984), fullerenes (1985), and high-temperature superconductivity (1986). The discovery of QC was announced to a community active in the mature science of crystallography. Crystallographers, and other scientists who studied the structure of matter and its defects, relied on a series of laws and paradigms undisputed since von Laue performed his first x-ray-diffraction experiments in 1912. One leading paradigm stated that the atomic structure of a crystal is ordered and periodic. Explanations of this paradigm, based on common sense, could be summarized as, “It is periodic because it is ordered.”

Periodicity implies a set of specific rules, among them the allowed rotational symmetries—namely one-, two-, three-, four-, and sixfold. Fivefold rotational symmetry is excluded. Past textbooks specified this, stating that fivefold rotational symmetry is impossible in periodic structures. Other books stated the impossibility of such symmetry in crystals.

Paradigms are based on experience rather than on a rigorous, scientific study process. Therefore when proven wrong, they are difficult to uproot. From 1912 to 1984, nothing shook the paradigm, “Order in crystals is periodicity.” Incommensurate structures challenged the paradigm for a while but were soon found to be a modulation of periodic crystals, saving the paradigm. However in the background, a series of articles broadened the scope of crystallography into hyperspace and prepared the mathematical platform from which the science of quasiperiodicity in crystals could take off.

In October 1986, the neutron scattering facility at Los Alamos National Laboratory became a national user facility and a formal user program was initiated in 1988. In July 1989, this facility was dedicated as the Manuel Lujan Jr. Neutron Scattering Center (Lujan Center) in honor of the long-term Congress representative from New Mexico. The Lujan Center, part of the Los Alamos Neutron Science Center (LANSCE), is a pulsed spallation neutron source equipped with time-of-flight neutron-scattering spectrometers for condensed-matter research. Neutron scattering is a powerful technique for probing the microscopic structure of condensed matter. The energies and wavelengths of thermal neutrons closely match typical excitation energies and interatomic distances in solids and liquids. Because neutrons have no charge, they penetrate bulk samples of material to give precise information on the positions and motions of individual atoms. The magnetic moment of a neutron interacts with unpaired electrons, making neutrons ideal for probing microscopic magnetic properties. Because neutron-scattering cross sections do not vary monotonically with the atomic number of the scattering nucleus, neutrons and x-rays can provide complementary structural information. This technique is particularly effective for structural problems in polymer and biological studies because hydrogen and deuterium scatter neutrons strongly but with different cross sections.

Short-wavelength-emitting devices, such as blue laser diodes (LDs) and light-emitting diodes (LEDs), are currently sought for a number of applications, including full-color electroluminescent displays, laser printers, read-write laser sources for high-density information storage on magnetic and optical media, and sources for undersea optical communications. For these purposes, II–VI materials such as ZnSe and SiC, and III–V-nitride semiconductors such as GaN have been investigated intensively for a long time. However it was impossible to obtain high-brightness (over 1 cd) blue LEDs and reliable LDs. Much progress has been achieved recently on green LEDs and LDs using II–VI-based materials. The short lifetimes prevent II–VI-based devices from commercialization at present. The short lifetime of these II-VI-based devices may be caused by the crystal defects at a density of 103/cm2 because one crystal defect would cause the propagation of other defects leading to failure of the devices. Another wide-bandgap material for blue LEDs is SiC. The brightness of SiC blue LEDs is only between 10 mcd and 20 mcd because of the indirect bandgap of this material.

On green LEDs, the external quantum efficiency of conventional, green GaP LEDs is only 0.1% due to the indirect bandgap of this material. The peak wavelength is 555 nm (yellowish green). As another material for green emission devices, AlInGaP has been used. The present performance of green AlInGaP LEDs is an emission wavelength of 570 nm (yellowish green) and maximum external quantum efficiency of 1%.

Organic electroluminescence (EL) is moving from a simple curiosity in laboratories to the reality of commercial use. The day will soon arrive when high-quality green EL displays find practical usage. Charge-injection-type EL involves the combination of positive and negative charge carriers injected from electrodes in contact with an organic thin film. The occurrence of EL through bipolar charge injection into organic solids was clarified for single crystals of anthracene and related compounds in the 1960s. Few new developments in physics exist today for organic EL. However worldwide enthusiasm for charge-injection-type EL, which started in the mid-1980s, has been increasing rapidly. The motivation for this renewed interest is straightforward. High-efficiency surface emission across the whole visible spectral range can be obtained easily, and prospects now exist for full-color flat-panel-display technology.

Since the demonstration of high-performance EL devices made of multilayers of vacuum-sublimed dye films by Tang and VanSlyke, much progress has occurred in the research and development of EL devices made from molecular materials. A variety of molecular materials such as vacuum-sublimed dye films, fully π-conjugated polymers, polymers with chromophores orr skeletal chains, or side chains, and polymer-dispersed dye films can be used for EL devices.

Among a variety of EL devices, multilayer-structure versions made of vacuum-sublimed dye films exhibit the best performance. Application-oriented research in the development of high-quality flat-panel displays has been performed during the past 10 years, mainly in Japan's private-sector laboratories.

Powder pressing, either uniaxially or isostatically, is the most common method used for high-volume production of ceramic components. The object of a pressing process is to form a net-shaped, homogeneously dense powder compact that is nominally free of defects. A typical pressing operation has three basic steps: (1) filling the mold or die with powder, (2) compacting the powder to a specific size and shape, and (3) ejecting the compact from the die. To optimize a pressing operation, experienced press operators generally understand and control parameters such as die-fill density, die-wall friction, packing density, and expansion on ejection.

Die filling/uniformity influences compaction density, which ultimately determines the size, shape, microstructure, and properties of the final sintered product. To optimize die filling and packing uniformity, free-flowing granulated powders are generally used. Spherical granules (i.e., agglomerates or clusters of finer particles) range in size from ~44 to 400 μm with the average size being ~100–200 μm. They are typically produced from 0.5 to 10-μm median particle-size powders by spray drying a ceramic powder slurry. To produce processable powders, various organic additives are typically added to the slurry prior to spray drying. These include binder(s) for strength, plasticizers that produce deformable granules, and lubricants that mitigate frictional effects. Consistent batching and dispersion of the granulated feed are critical for reproducible and uniform die filling. Granule densities that are 45–55% of the theoretical density (TD), and bulk-powder and die-fill densities of 25–35% TD are typical for ceramic powders.

The past decade has seen exciting advances in the discovery, improved synthesis and processing, and molecular level engineering of new inorganic materials having specialized electronic, ceramic, and structural applications. Many such materials share two common characteristics: They are complex in structure and composition (think for example of oxide superconductors), and they must be prepared by a series of steps under carefully controlled conditions (consider the intricacies of zeolite synthesis for example). The use of low-temperature aqueous synthesis conditions, with appropriate attention to pH, inorganic and organic structure-directing agents, and subsequent drying and calcination protocols has led to a wealth of new and often metastable crystalline polymorphs, to amorphous materials, and to fine powders with particles of nanoscale dimensions. Methods such as sol-gel synthesis, chimie douce (soft chemistry), hydrothermal synthesis, chemical vapor deposition, and various beam-deposition and epitaxy techniques produce a wealth of materials not constrained to be in chemical equilibrium with their surroundings and not representing the state of lowest free energy. Modern materials chemists almost have their pet Maxwell Demon to select atoms at will and cause them to assemble in a structure of controllable dimensions. The wealth of possible new structures formed begins to mimic the riches of organic chemistry. In this field, the fact that all complex organic and biochemical molecules are metastable under ambient conditions with respect to a mixture of carbon dioxide, water, and other simple gases is irrelevant except in a conflagration.

Liberation of ceramic science from the tyranny of high-temperature equilibrium is thus leading to new materials synthesized more quickly, at lower cost, and under environmentally more friendly conditions. There is of course a price to pay. First the synthetic procedures are more complex than traditional “mix, grind, fire, and repeat” ceramic processing. Second and more importantly, very little is known about the long-term stability of the materials formed, about their degradation during use, and about materials compatibility. Two examples of such problems are the potential corrosion of high Tc YBCO superconductors by ambient H2O and CO2, and the collapse to inactive phases of complex zeolitic and mesoporous catalysts under operating conditions. Chemical reactions in metastable materials are governed by an intertwined combination of thermodynamic driving forces and kinetic rates. For this rich landscape of new materials, neither the depths of the valleys nor the heights of the mountains are known. Often one cannot even tell which way is energetically downhill.

Scanning-probe microscopies (SPMs) have experienced an impressive development in the last 10 years, such that obtaining atomically resolved images of surfaces is now routine. The scanning tunneling and atomic force microscopes (STM and AFM) are the most widely known representatives of the growing family of SPMs. With these two instruments, which were invented more than 15 years ago, the atomic structure of the surfaces of solid materials can be imaged in real space and, very importantly, in virtually any environment: vacuum, air, reactive gas atmospheres, and under liquid. These two circumstances (atomic resolution and lack of environmental constraints) make SPM unique among microscopies. One frontier remains unconquered however: the imaging of liquid surfaces. Even though liquids constitute a very large fraction of all materials, the study of their microscopic morphology has been limited to the diffraction limit of visible optical microscopy, a situation that has remained unchanged for decades. Techniques capable of breaking this resolution limit are welcomed. One such technique was recently developed in our laboratory, which we call scanning-polarization-force microscopy (SPFM). It did not result from the discovery of any new physical principle but from the simple application of well-known methods and techniques, in combination with the AFM, to the field of liquid surface structures. In this article, we will first summarize the working principle of SPFM. Then we will show how we have applied it to the investigation of liquid films and droplets of nanometer dimensions and their interactions with solid supports.

Semiconductor alloys have become increasingly useful during the last four decades because, through the use of alloys, the properties of semiconductors can be tailored by varying the composition to precisely match the requirements for specific electronic and photonic devices. In addition the use of alloys allows the production of special structures, such as quantum wells, that require rapid changes in bandgap energy during growth. This has led to so-called “bandgap engineering,” in which device designers and epitaxial growers are working together to produce structures having virtually atomic-scale dimensions.

As feature sizes in integrated circuits approach 0.18 μm, problems with interconnect resistance-capacitance (RC) delay, power consumption, and crosstalk become more urgent. Integration of low-dielectric-constant (k) materials will partially mitigate these problems, but each candidate with k significantly lower than that of dense silica (k ∼ 4) suffers disadvantages. Current low-k commercialization emphasizes spin-on glasses (SOGs) and fluorinated SiO2 with k > 3, and a number of polymers are under development with k in the range of 2–3. These suffer from potential problems including thermal stability, mechanical properties, low thermal conductivity, and reliability. For some low-k materials, a protective liner covering the conductor is necessary. Although the material k is often cited, the value of practical concern is the effective k, which may be quite different because of this protective liner. As feature sizes shrink, the presence of the liner becomes more problematic and necessitates even lower k materials.

Another approach employs nanoporous silica with k of ∼1–4. Porous silica has been classified as an aerogel (dried supercritically) or as a xerogel (dried by solvent evaporation). We use the term nanoporous silica since it captures the key material properties that may be independent of how the films are processed. The ultralow dielectric constant results from porosity incorporation. For a porous material, the dielectric constant is a combination of that of air (∼1) and of the solid phase. The variation of k with porosity (volume fraction of pores) appears in Figure 1.

Ceramics are inorganic, nonmetallic materials, a class that includes solids with both crystalline and glassy (i.e., aperiodic) atomic structures. This article will emphasize radiation effects in conventional refractory oxide ceramics and glasses that display high-temperature strength, chemical inertness, and high electrical resistivity. The introduction to this issue cites various radiation environments to which materials may be exposed; a number of these environments receive mention in this article. However emphasis here is on giving the reader an overview of damage mechanisms and of the changes in key physical properties (optical, thermal, electrical, and mechanical) that can affect the behavior of these materials. The broad subject of radiation-damage phenomena in ceramics has recently been reviewed by Hobbs et al.

Even at low doses of energetic radiation, whether particulate or electromagnetic, isolated atomic-level defects may be generated and can offer the potential to induce changes in physical properties. However for some applications such as fission reactors, fusion reactors (now in the conceptual design stages), and high-energy accelerators, one must deal with high fluences capable of generating relatively large damage aggregates such as dislocation loops and voids. Thus one must understand the evolution of damage microstructures from point defects to aggregates in both model materials and engineering ceramics. Even when induced changes are only electronic in nature, the effects on some properties (e.g., electrical conductivity) can be dramatic. In the following sections, evolution of the damage state and consequent changes in key properties are described, with some attention to the often critical role of impurities.

In recent years, SiC has received a great deal of attention as a nearly ideal material for the fabrication of high-speed, high-power transistors. The high electric breakdown field of 3.8 × 106 V/cm, high saturated electron drift velocity of 2 × 107 cm/s, and high thermal conductivity of 4.9 W/cm K indicate SiC's potential for high-power, high-frequency operation. A wide bandgap should also allow SiC field-effect transistors (FETs) to have high radio-frequency (rf) output power at high temperatures.

These material qualities have been verified through outstanding device performance. Recent results for SiC metal-semiconductor field-effect transistors (MESFETs) have included superior frequency and power performance, with power gain at frequencies as high as 40 GHz and power densities as high as 3.3 W/mm. This represents significantly higher operating frequencies and power densities than current Si rf power FET technology, and nearly three times the power density of GaAs MESFETs, which are currently used in many commercial rf power applications. Similarly, SiC static induction transistors (SITs) have much higher power densities than their Si counterparts and have recently been demonstrated in modules with as much as 470-W total pulsed output power. This article describes microwave device operation, discusses material properties needed for rf power generation, and summarizes state-of-the-art SiC high-frequency device performance. Emphasis is placed on MESFETs and SITs since they are currently the most mature SiC-based device technologies.

Hundreds of new alloys have been observed to form quasicrystalline phases. Now it is clear that quasicrystals are not at all a rarity but in fact are quite common. Many quasicrystalline phases are metastable, but some are stable. The discoveries of stable quasicrystalline alloys have led to a breakthrough in the understanding of quasicrystalline structure and have established a new group of quasicrystalline materials. In this article, alloy systems that form quasicrystals and their preparation techniques are reviewed.

Polymers are routinely placed between solid walls to provide lubrication or adhesion. Their function in these roles depends critically on the degree of dissipation within the polymer film and at the film/wall interface as the film shears or ruptures. Good lubrication is achieved by minimizing frictional dissipation while dissipation increases the strength of adhesive bonds.

Gent and Schultz suggested a direct link between frictional losses and the adhesive performance of polymers. This correspondence has been supported by recent experiments and by some of the molecular-dynamics simulations to be described. However we find that the correspondence breaks down when the molecular motion producing dissipation occurs at different locations during shear and rupture. In the following sections, we discuss the types of rate-dependent dissipation observed in thin films and the different factors that control whether dissipation occurs within the polymer or at the wall/film interface. The results suggest an origin for interesting memory effects observed in surface-force-apparatus (SFA) experiments on thin films and expose the atomic-scale processes that produce dissipation during internal rupture of a thin film.

Polyethylene has been manufactured commercially for over sixty years. However recently a high-modulus variant has become available. This article describes the background to the development of high-modulus polyethylene and describes the story behind the sequence of scientific discoveries and technological developments that occurred to enable this advance.

Polyethylene was discovered by accident in the 1930s when two scientists working for the ICI chemical company carried out a series of experiments to study the effect of pressure on the reaction kinetics of certain organic liquids and gases. During these experiments, the scientists noted a waxy deposit in their small reaction vessel which marked the birth of one of the world's largest commodity polymers. The crystallographic structure of polyethylene was determined by another ICI industrial scientist who at the time was pioneering x-ray techniques. The crystallography of polyethylene, shown in Figure 1, is relatively simple in that it shows the long CH2-CH2 polymer chains packed in the all “trans” configuration to form an orthorhombic cell.

Polyethylene rapidly developed as a commodity thermoplastic finding application in injection molding, extrusion, and packaging. Its prime material characteristics are its toughness, flexibility, and ease of processing. The stiffness of the polymer for most applications was found to be of the order of 1 GPa, which was similar to other thermoplastics but less than metals such as aluminum at 60 GPa and steel at 210 GPa.