The pervasive impetus of modern semiconductor technology has become an accepted fact. Scientific mastering of materials and processes has increased tremendously within a short time frame. Technological control has been derived from this scientific base. An industry with more than $100 billion worth of silicon devices per annum in 1994 as well as incredibly high growth rates of production and applications is an economic reality in addition to being a matter of international industrial policy. The materials aspect of using perfected single crystals and applying local doping control provides the basis of this unusual success. Earlier usage of materials differed remarkably. Bronze and steel are used for their specific bulk properties. Shaping and connecting pieces is at the heart of iron-age or bronze-age technologies. Integration inside a regular spatial array of a host crystal is the semiconductor principle. The “Royal Road” to modern microelectronics consists of initially procuring a perfected single-crystal host, then locally establishing electrical and optical properties inside the host by specific replacements of host atoms by foreign “dopants.” The somewhat disparaging expression of defect as a generic term for all deviations from the host perfection does not really convey the power of this “doping doctrine” for semiconducting materials. The early pioneers of germanium and silicon however, placed great emphasis on the experimental verification that n-type and p-type doping by elements of the adjacent columns in the periodic table were accompanied by changes in the lattice parameter with atomic substitution of the host atoms as the guiding principle.

For many thin-film applications it is useful or necessary to accurately monitor and control the film's physical properties. Particularly important are composition and structure, which determine most of the material's properties. Examples include the semiconducting properties of SixGe1−x, the ferroelectric properties of BaTiO, and the superconducting properties of Bi2Sr2CaCu2Oy, all of which depend strongly on composition. The development of in situ analyses not only provides the possibility of feedback control during deposition, but also assists in a more efficient evaluation of the best processing conditions to achieve particular properties. Control of the film structure and composition during film growth has not been easy, and in situ analyses during film nucleation and growth has been a goal of many investigators.

Our interest in in situ monitoring of film deposition was motivated by sputter growth of Bi2Sr2CaCu2Oy from a composite target. These studies found that the film composition was a strong function of deposition parameters, particularly the substrate temperature. Investigating these complex relationships required the growth and ex situ analysis of many films. It was clear that a technique to determine the composition of films during growth could efficiently unravel the complex relationships between composition and processing parameters.

An important consideration for an in situ analysis technique is that it not interfere with the growth process either by altering the film surface or by changing the deposition dynamics. During deposition, the substrate temperature, gas flow, and deposition rate all must be maintained; any interruptions for the analysis may alter the growth irreversibly.

Although it is an integral part of any structure involving a film, the substrate is often taken for granted. The choice of substrate is, however, one of the most important materials issues in thin-film growth. This article focuses on substrates for thin films and will provide criteria for selecting the proper material needed to fill specific application requirements. As will become obvious, the ideal substrate for a given film often does not exist. Specific applications require different substrate materials that offer an acceptable compromise for the purpose at hand. Ideally, the substrate should give mechanical support but not interact with the film except to provide sufficient adhesion, and in many cases, the provision of a template for atomic ordering. In practice, however, the substrate exerts considerable influence on film characteristics. The search for viable substrate materials is an active area of research.

Formation of aggregates has been observed as a general phenomenon for a wide variety of organic molecules, especially aromatic compounds and dyes. Aggregation is most commonly encountered in crystals or in thin films. However, it has been increasingly observed in microheterogenous media or in other situations where high local concentrations occur or where specific orientation is favored. Two limiting types of aggregation have been defined based on the orientation of transition dipoles and their absorption spectral characteristics. These are the “J” aggregate, in which head-to-tail arrangements of transitiondipole moments are characterized by a sharp, intense, red-shifted transition compared to the isolated (solvated) monomer, and the “H” aggregate, where head-to-head (card-pack) orientations are characterized by a blue shift of the prominent transition compared with the monomer. Several treatments have been proposed to correlate the observed spectral shifts with the aggregate structure. For a number of compounds, the association of known x-ray-determined structures with spectral features has supported the theoretical predictions developed by Kasha and Hochstrasser or by Czikkely, Försterling, and Kuhn. The focus of the studies described here has been on aggregation occurring in Langmuir-Blodgett (LB) films and related media, such as bilayer vesicles, which are characterized by an assembly of molecules in an interfacial situation where a polar-nonpolar or water-hydrocarbon boundary should provide a strong organizing influence. In early cases where aggregates were encountered as prominent components of mixed LB films (even when relatively dilute mixtures were used), the phenomenon was usually dismissed as “microcrystallization,” which was considered an unavoidable nuisance and not really due to fundamental intermolecular interactions. More recent studies have shown that aggregation in these media is really a significant molecular phenomenon that shows dependence both on the specific molecules and the topology of the film-forming surfactant. Although some previous investigations have been carried out with different results for various substrates, we have embarked on a study to correlate aggregation behavior for a number of different chromophores incorporated into amphiphilic structures to obtain a general picture of the relative importance of different factors that can control aggregation phenomena.

Block copolymer melts with a near-symmetric composition can microphase-separate to form a lamellar morphology where unfavorable monomeric interactions are reduced by an antiparallel layering of the polymer chains (see Figure 2, discussed later). The symmetry of such a block copolymer is the same as for small-molecule, smectic-A liquid crystals, which also exhibit (parallel or antiparallel) layering. Because of their shared symmetry, their quasistatic mechanical properties are of the same form. To lowest order, the energy of distortion of the lamellar pattern can be expressed as a sum of a compressional/dilational and a bend energy (see, for example, Reference 2):

where n is a unit vector normal to the layers, and ε and ∇ n are the layer dilational and bending strain, respectively. B and K are the corresponding moduli. These have been measured only recently for block copolymers. The ratio of these moduli yields a characteristic microscopic length:

which is calculated to be a fraction of the layer spacing d0.

Similarities are also found in the dynamic properties of polymeric and smallmolecule smectics. These similarities are found at low strain rate and amplitude where the rheological response is controlled by their similar layered structures and associated defects. During either steady or dynamic shear, certain defect and domain motions are induced such that a steady-state alignment of the director has been observed. In this report, we discuss defect structure and how the much more viscous nature of these polymeric layered materials influences the distribution and relative population of each defect type. We also discuss how these defects can control material stress-relaxation behavior.

Bamboo and certain other plants have excellent characteristics based on functionally gradient structures. Figure 1 shows a cross section of bamboo, corn, and barley culms. The collection of solid dots in bamboo is called the bundle sheath, which acts as fibers for the plant. These bundle sheaths are shown in the magnified photograph of corn and barley. These two plants are made of composite materials reinforced by fibers of the bundle sheath; the fiber strength is about 10 times that of the matrix. Furthermore, the fiber distribution is dense in the outer region and sparse in the inner region. It can be said that the fiber distribution forms a gradient structure. The fibers in barley are sparse in the narrow cross section and are different from the fibers in corn and bamboo. The number of honey-comb-shaped cells in barley is higher on the outside than the inside region. Therefore, barley also has a gradient structure. These relations of the microscopic gradient structure are displayed graphically by the cross-sectional photographs. Figures 1a and 1b present the fiber contents of bamboo and corn, respectively, with respect to radius measured from the outer radius. Figure 1c gives the gradient distribution of the cell number for barley with respect to radius. These gradient features probably relate to the bending-stress distribution caused by wind, rain, and snow—the so-called environmental loads.

Application of silicon carbide (SiC) as a structural material has been limited thus far by its low fracture toughness, even though, in comparison to other ceramic materials, SiC has superior high-temperature strength and creep, wear, corrosion, and oxidation resistance. For automotive applications, a higher fracture toughness is required. For example, the brittleness and catastrophic fracture behavior of SiC materials have resulted in limited use in automobile exhaust-valve systems and turbocharger rotors. High-density SiC bodies can be produced by a pressureless sintering process. However, the sintered bodies often include flaws which are related to processing, primarily, the presence of agglomerates and crystallographic defects in the starting powders. The importance of grain size and shape refinement in the improvement of mechanical properties has been recognized, and thus, processing procedures and sintering aid compositions have been examined extensively. However, one of the key factors is the “as-received” powder characterization (distribution of grain sizes, polytypes, and impurities) for producing sintered bodies of SiC with consistent physical properties.

A complexity in SiC materials is that SiC can form various crystal structures having essentially the same chemical composition but a differing number of stacking layers in the unit cell. This is commonly called a polytype. There is only one crystal structure with cubic symmetry, which is identified as 3C or the β-phase. At high temperature, the β-phase transforms to α-phases with hexagonal or rhombohedral symmetry, with 4H, 15R, and 6H (Ramsdell notation) being the major polytypes observed in SiC materials. Preference of the polytype selection during the β- to α-phase transformation is dependent on the chemistry of the sintering aids and metallic impurities in the grain boundaries.

The scaling of integrated-circuit device dimensions in the horizontal direction has caused an increase in aspect ratios of contact holes and vias without a corresponding scaledown in vertical dimensions. Conventional sputtering has become unreliable for handling higher aspect-ratio via/contact holes because of its poor step coverage. Several studies have attempted to overcome this problem by using W-CVD and reflow technology. The W-CVD is used for practical device fabrications. However, this technique has several problems such as poor adhesion to SiO2, poor W surface morphology, greater resistivity than Al, and the need of an etch-back process.

Al reflow technology using a conventional DC magnetron sputtering system can simplify device-fabrication processes and achieve high reliability without Al/W interfaces. In particular, the Al reflow technology is profitable for multi-level interconnections in combination with a damascene process by using Al chemical mechanical polishing (CMP). These interconnections are necessary for miniaturized and high-speed devices because they provide lower resistivity than W and simplify fabrication processes, resulting in lower cost.

This article describes recent Al reflow sputtering technologies as well as application of via and interconnect metallization.

Over the past 30 years we have seen a strong but uncoordinated effort to both increase the availability of numeric materials-property data in electronic media and to make the resultant mass of data more readily accessible and searchable for the end-user engineer. The end user is best able to formulate the question and to judge the utility of the answer for numeric property data inquiries, in contrast to textual or bibliographic data for which information specialists can expeditiously carry out searches.

Despite the best efforts of several major programs, there remains a shortfall with respect to comprehensiveness and a gap between the goal of easy access to all the world's numeric databases and what can presently be achieved. The task has proven thornier and therefore much more costly than anyone envisioned, and computer access to data for materials scientists and engineers is still inadequate compared, for example, to the situation for molecular biologists or astronomers. However, progress has been made. More than 100 materials databases are listed and categorized by Wawrousek et al. that address several types of applications including: fundamental research, materials selection, component design, process control, materials identification and equivalency, expert systems, and education. Standardization is improving and access has been made more easy.

In the discussion that follows, we will examine several characteristics of available information and delivery systems to assess their impact on the successes and limitations of the available products. The discussion will include the types and uses of the data, issues around data reliability and quality, the various formats in which data need to be accessed, and the various media available for delivery. Then we will focus on the state of the art by giving examples of the three major media through which broad electronic access to numeric properties has emerged: on-line systems, workstations, and disks, both floppy and CD-ROM. We will also cite some resources of where to look for numeric property data.

Handbells have become very popular musical instruments, especially in schools and churches; about 40,000 hand-bell choirs have been reported in the United States alone. Tuned handbells are generally made of cast bronze, which has been the traditional bell material for many centuries.

Demand for handbells of lower and lower pitches has led to the development of bass bells as low as G0 (fundamental frequency of 24.5 Hz). Unfortunately, these large bass bells radiate inefficiently, especially the bells made of bronze. This is because the speed of bending waves in these bells is considerably lower than the speed of sound in air, a condition known as “being below the coincidence frequency.”

In order to obtain a higher radiation efficiency and thereby enhance the sound of bass bells, the Malmark Company has created a new bell design using aluminum rather than bronze. These bells are larger in diameter, and they have lower coincidence frequencies, both of which lead to more efficient radiation of bass notes. In addition, they are considerably lighter in weight, and thus are much more easily handled by bell ringers.

In this article, the acoustical properties of two G1 handbells, one of aluminum and one of bronze, are compared. The aluminum handbell has a diameter of 48.5 cm, a height of 34 cm, and a wall thickness from 4 to 5 mm. The bronze bell has a diameter of 38.5 cm, a height of 28.5 cm, and a wall thickness from 3 to 4 mm.

Magneto-optic effects were first discovered in 1845 by Michael Faraday, but these effects continue to make a major impact on the materials community to this day. The need for new magnetic information-storage media has stimulated new approaches and opened new scientific opportunities in the exploration of thin-film and surface magnetism. This article provides background to some of these developments and highlights examples of contemporary issues that provide a focus for the field. In the Faraday effect, the polarization plane of linearly polarized light rotates when a magnetic field is applied in the propagation direction. The analogous phenomenon was subsequently discovered by the Rev. John Kerr in 1877 for light reflected from opaque materials. The works of Faraday and Kerr serve as cornerstones for our present understanding of magneto-optic effects in magnetic materials. Magnetooptics is presently described in the context of either microscopic quantum theory or macroscopic dielectric theory. Microscopically, the coupling between the electric field of the propagating light and the electron spin in a magnetic medium occurs through the spin-orbit interaction. Macroscopically, magneto-optic effects arise from the antisymmetric, off-diagonal elements in the dielectric tensor, as discussed in the next section.

Magneto-optic characterizations of surface magnetism began only a decade ago. The first surface magneto-optic Kerr-effect study, better known by its acronym SMOKE, concerned the magnetichysteresis loops for ultrathin Fe films grown epitaxially on Au(100). Since then, SMOKE has emerged as a premier surface-magnetism technique. SMOKE has been applied to various topics in low-dimensional magnetism, ranging from the detection of magnetic order to the characterization of critical behavior, magnetic surface anisotropies, and the oscillatory antiferromagnetic coupling exhibited by giant-magnetoresistance

heterostructures. Additional interest in SMOKE has been generated by the recent commercialization of high-density, magneto-optic information-storage media, and especially by the next-generation candidate material based on Co/Pt superlattices.

Second to silicon (Si), the most highly developed technology for semiconductor processing exists for gallium arsenide (GaAs). Unfortunately, GaAs processing is more complex than that of Si, mainly because GaAs is a compound semiconductor. Additionally, the lack of a stable native GaAS oxide and other disadvantages relative to Si have prevented this material from expanding beyond the small niche of applications where its high intrinsic electron mobility, superior radiation hardness, and direct bandgap are essential. Adequate understanding and modeling of the process physics are important for extending the “process window” available to GaAs manufacturers and for increasing the appeal of this material. This article deals with one of the most important process events: dopant diffusion.

In the next section we briefly describe device-fabrication technology and show the importance of diffusion modeling in the prediction of device characteristics. We then review some elementary diffusion mechanisms and outline the dopants that are important in GaAs-processing technology as well as the methods by which these dopants are introduced into the substrate. In subsequent sections we review the research community's current understanding of diffusion mechanisms as well as model parameters for specific dopants. Much work has been done in this field, at Stanford and by other groups, since the publication of a major review of the subject by Tan et al. in 1991. In this article, we focus on these recent contributions.

When a discontinuity in material properties exists across a bonded interface, stresses are generated as a result of any thermal or mechanical loading. These stresses significantly affect strength and failure characteristics and may be large enough to prevent successful fabrication of a reliable joint. The use of an interlayer material to successfully reduce mismatch stresses, thereby preventing joint failure or improving joint strength and reliability, requires knowledge of failure mechanisms and of the effects of interlayer properties on the critical stress components.

The origin of residual stresses developed during cooling of a ceramic-metal joint from an elevated fabrication temperature is illustrated qualitatively in Figure 1. Away from edges, the in-plane (parallel to interface) stresses are typically compressive in the ceramic and tensile in the metal. These stresses can cause cracking perpendicular to the interface, leading to spalling or delamination failures. Such failures are frequently observed in thin-film and coating geometries. Where the interface intersects a free edge, large shear and axial (perpendicular to the interface) stresses are generated. The edge stresses are typically tensile within the ceramic and tend to promote crack propagation within the ceramic parallel and adjacent to the interface. This is the most commonly observed failure mode in bonded structural components.

The evolution of the steel pan during the last half century represents the emergence of a truly new musical instrument. Originating in the folk culture of the Caribbean, it has become popular throughout North and South America, as well as in Europe. The haunting sounds of steel drums, generally called “steel pans” by players and tuners, are now enjoyed worldwide. Pans are usually crafted from commercial 55-gallon steel oil drums. The top with its various openings is discarded, and only the bottom, along with the side, is used for the pan. The length of the side skirt varies depending on the pitch range of the completed instrument. Thus, a bass pan will utilize almost the entire length of the original drum, whereas a lead pan, or a single tenor, will be cut to retain only a very short skirt. The pitch of each note section is primarily determined by the size and geometric shape of the section as bounded by a grooved seam. Figure 1 shows representative layouts of several pans. The note arrangement in some pans is such that several pans are required to play a melody. For example, a double-second instrument consists of two pans, while a bass includes as many as six. In addition to the section size, the mass density, stiffness, and curvature contribute significantly to the pitch of each note. The overall timbre of the tone produced depends to a great extent on the spectral content of the note. This, in turn, depends strongly on the normal modes of the excited note section, and to some lesser extent, on the normal modes of all other sections. This latter dependence is caused by the fact that all note sections are imbedded in the same solid matrix of the drumhead. Therefore, vibrations in one note section cause all note sections to vibrate. The nature of the coupling between sections has been the subject of considerable interest to researchers in the field. Inasmuch as the entire pan represents a highly nonlinear system, the tone quality also varies significantly with excitation amplitude.

Spin-polarized low-energy electron microscopy (SPLEEM) is a new approach to magnetic-surface imaging. Historically, it is an outgrowth of LEEM (low-energy electron microscopy), which has been successfully applied to the high-speed in situ imaging of microscopic surface structures under clean ultrahigh-vacuum (UHV) conditions. In terms of basic instrument design, in SPLEEM a spin-polarized electron source of the Pierce type replaces the LaB6 field-emission electron source used in all dedicated LEEM systems. A new, fully electrostatic version of LEEM has recently been developed, resulting in a more compact, flange-on-type instrument component, which can be integrated with standard UHV surface analytical systems. The main advantage of the new SPLEEM technique is an advanced electron-source design that allows the spin direction to be arbitrarily adjusted with respect to the sample surface (see the following).

The engineering of advanced micro-electronic circuits, optoelectronic devices, and integrated optical circuits requires precise control of the lateral dimensions and thicknesses of device features and of the stoichiometry and doping of epitaxial semiconductor regions. This is preferably achieved by real-time monitoring and control of the individual deposition and etching processes that constitute the processing sequence. The use of optical probe techniques for the real-time monitoring of etching and/or growth processes is favored because of their nondestructive character and their potential use in real-time feedback control. Some of these methods are ideal in monitoring the overall growth process and/or substrate temperature in industrial applications, requiring low cost and maintenance. For example, in situ reflectance-spectroscopy methods, such as dynamic optical reflectivity (DOS), spectral-resolved normal incidence reflectance spectroscopy (MRS), or pyrometric interferometry (PI), are successfully applied to various deposition processes and provide information on both the growth rate and the composition of the deposits. However, small changes in the reflectance (because of chemical interactions at the surface of the films with the reactants supplied from the vapor phase) are of the order of 10−3 to 10−4 and are hardly observable with normal-incidence reflectance techniques because of the high reflectivity of substrate/film interface, which is typically of the order of 40%–60% for many semiconductors.

In order to increase the sensitivity to surface- and interface-related growth properties, alternative optical-observation methods such as reflectance difference spectroscopy (RDS), surface photoabsorption (SPA), and spectral ellipsometry (SE) have been developed.

Understanding the influence of local variations in symmetry (“defects”) on the macroscopic properties of polymers in the condensed state is an ongoing experimental and theoretical challenge. Studies of defects in solids require the most information-intensive description of microstructure since it is not possible to describe a “defect” without understanding the morphology of the majority phase as well.

The nature of defects in polymers has been discussed elsewhere, including other articles in this issue of the MRS Bulletin. The structure, properties, and mobility of defects in polymers are all profoundly influenced by the covalently bonded chain backbone. In polymers, there are unique defects such as chain folds and twists that have no obvious analogue in materials of small molar mass. Here, we examine a particular type of defect that is present in all polymer systems with finite molecular weight: chain ends. Our interest will focus on chain ends in polymers that are essentially fully extended parallel to a certain preferred orientation axis.

The extended-chain microstructure was originally envisioned by Staudinger as a “continuous crystal” in which high-molecular-weight polymers would be perfectly oriented and close-packed together laterally. Extended-chain polymer fibers such as poly(paraphenylene terephthalamide) (PPTA or Kevlar®), gelspun polyethylene (Spectra®), and the rigid-rod polymers poly(paraphenylene benzobisthiazole) (PBZT) and poly(paraphenylene benzobisoxazole (PBZO or PBO) (Structure 1) closely approach this conceptual limit. The outstanding tensile moduli (100–400 GPa) and tensile strengths (2–4 GPa or higher) of these fibers have generated considerable interest for lightweight structural applications. Extendedchain polymers can also be prepared by solid-state polymerizations of appropriate monomer precursors. Perhaps the most familiar of this latter class of materials are the polydi-acetylenes, first developed by Wegner.