Ceramic materials are currently being used in many applications and account for about $26 B of the U.S. economy in low value-added products (glass, refactories, cement, whitewares, etc.) and nearly $4 B of high value-added components (chip carriers, waveguides, capacitors, cutting tools, etc.). There are many new applications of ceramics in high technology engineering systems which will result in significant growth of the ceramics business if the materials research triad Structure-Property-PROCESSING moves from the laboratory to the manufacturing plant. Each component can be classified by function (e.g., thermal, mechanical, electromagnetic, optical, biological, and nuclear); configuration (e.g., film, monolith, composite, porous, and single crystal); or processing route (e.g., powder processing, glass forming, and crystal growth). The opportunities are greatest for those components which have a well established processing science base or an unique fabrication method.

Thirty years of research have shown that the critical physical/chemical properties are strongly related to the structure on one or more levels: few angstroms - crystal structure and phases; 10–100 Å - boundary or grain boundary layers or phases; 1 μm - microstructure (grain size, porosity, interpenetrating phases, etc.); and macro-structure including joining, shape and dimensional tolerances. The major focus of this overview relates to issues of reliability and reproducibility in processing polycrystalline ceramics and the relationship between processing and the control of the structure and properties. Examples considered are those of high value-added components where design and functionality justify improved material characteristics.

Results are presented for cation self-diffusion in Cu2O and Cr2O3 as a function of temperature and deviation from stoichiometry. A defect model for Cu2O involving neutral and singly charged copper vacancies, electron holes, and singly charged oxygen-interstitial ions is developed and fit to the tracer-diffusion data, the electrical-conductivity data, and the stoichiometry data, yielding concentrations and mobilities of these defects. The diffusion results for Cr2O3 single crystals indicate that cation self-diffusion takes place by triply charged vacancies and that previous diffusion results for polycrystalline Cr2O3 are dominated by grain-boundary diffusion.

The rate of oxidation of metals at high temperatures is usually controlled by the transport of metal ions or oxygen through the growing film of oxide on the surface. Since many technologically useful alloys (eg steels, nickelbased alloys) are based on the 3d transition metals, the transport properties of the oxides of these elements are of particular interest. In the temperature range 500–1000°C, where these alloys are used, the important transport paths in the oxide are not those that involve diffusion in the bulk crystal lattice, but other “short-circuit” paths. Basic experiments on the oxidation of nickel and direct tracer measurements of diffusion coefficients show that rain boundaries are the most important short-circuit paths under protective oxidation conditions. Similar conclusions can be drawn for other metals including chromium. The methods used in this work have been extended to study the way in which a surface treatment with CeO2 reduces the oxidation rate of nickel, apparently by inhibiting the outward transport of metal ions along grain boundaries. Segregation of cerium ions to grain boundaries increases the diffusion coefficient of nickel; the inhibition is caused instead by particulates of CeO2 in the middle region of the scale.

Tracer diffusion coefficients of Fe, DFe* in single crystals of rutile and of 0 –2.0% Fe doped rutile were measured. The oxygen pressure dependence of DFe* in pure rutile showed complicated behavior. The values of DFe* may consist of contributions from Fe2+ ions diffusing by an interstitial mechanism and from Fe3+ ions diffusing by an interstitialcy mechanism in cooperation with tetravalent titanium interstitial ions, Tii. The value of DFe* in Fe doped rutile attains a saturation value when the Fe content reaches about 0.1%, D*Fe decreases drastically when the Fe content exceeds about 0.35%. Complex impedance measurements of electrical conductivity indicate the existence of ionic conduction for Fe doped rutile containing less than 0.35% of Fe. The small ionic conductivity relative to the values of D*Fe suggests that only a small fraction of the iron ions are highly mobile. Above 0.35% Fe, the observed drastic decrease in D*Fe may result from the formation of a shear structure In highly doped rutile.

Recent calculations on (100) coincidence-site lattice (CSL) twist boundaries in the NaCl structure have shown that without point defects these boundaries are only marginally stable. Following an earlier suggestion that point defects are the likely source for the considerable stability of these boundaries observed experimentally for Mgo and NiO, Tasker and Duffy have shown recently that the creation of a Schottky pair can, indeed, stabilize a (100) twist boundary in NiO. In this article a variety of configurations in which one or more Schottky pairs have been created in the perfect CSL or anti-CSL unit cell are investigated. It is concluded that many metastable structures may exist which differ mainly with respect to their different interfacial mass densities and the relative translation of the two halves of the bicrystal.

The point defect reaction which governs the magnetic and electrical properties of transition metal oxide spinels is the distribution of cations between tetrahedral and octahedral sublattices. Two types of redistribution occur -- ionic and electronic. The thermodynamics and kinetics of these processes can be studied via the electrical properties, thermoelectric coefficient and conductivity.

Failure processes in brittle solids have been described. Particular emphasis has been placed upon a distinction of the various flaw populations that dictate failure and their relation with the dominant manufacturing (or environmental) variables. Micromechanics models of failure from the more significant flaws, and considerations of their relative roles in the failure process, have been cursorily presented. The fracture toughness, its significance in failure, and its dependence upon microstructure, have been examined, in conjunction with considerations of crack shielding and enhanced toughness. Finally, the role of statistical processes has been outlined, within the context of the important flaw populations.

The displacive transformations in complex oxides of the type LaNb1-xVxO4 has been studied by x-ray diffraction and Raman scattering for 0 < x < 0.3. X-ray diffraction results indicate that the transformation from the tetragonal high temperature structure (C4h6) to the monoclinic low-temperature structure (C2h6) is higher than first order and that the transformation temperature Tc is depressed significantly by V substitution. Raman scattering results show that the force constant between the nearest (Nb, V)O4 tetrahedral units behave uniquely compared to others. It softens at Tc as a function of composition and it also softens as a function of temperature as Tc is approached from above.

The structure of low-angle tilt boundaries in magnesium-aluminate spinel has been studied by transmission electron microscopy. Special low-angle boundaries have been prepared from bicrystals but also boundaries in polycrystalline spinels were studied. The dislocations were all dissociated into partials. Their dissociation widths were found to increase with deviation from stoichiometry. Observations at lattice dislocations indicate that the composition at the boundaries may be different from the nominal composition.

The stress-strain behavior of <100> single crystals of Mn1-δQ has been measured for the first time. The experiments were performed at 1200°C varying the oxygen partial pressure PO2 from 10-12 to 10-5 atms which causes the deviation from stoichiometry δ to vary between 5×10-2 to 2 percent. The upper yield stress and the stress achieved at a zero work-hardening rate vary by a factor of 2 as a function of δ and exhibit a maximum at pO2∼10-9 atms. The present results suggest that the diffusion of oxygen interstitials control deformation for large δ while oxygen vacancies control deformation for small δ, an interpretation used to explain the deformation Co1-δQ. These results illustrate the intimate relationship between defect properties and deformation in nonstoichiometric oxides.

A method, based upon subtractive coloration or high-temperature reduction, was developed to produce metal precipitates or colloids in oxides. The structure, symmetry, morphology and composition of precipitates were analyzed by analytical electron microscopy and neutron scattering technioues. In the MgO:Ni system, the colloids were coherent nickel precipitates having a fcc structure (a0 = 3.52 Å), and occasionally a bcc structure (a0 = 2.88 Å). The coherent precipitates were analyzed to have either <100>m,∣∣ <111>p or <l00>m ∣∣<100>p matrix-precipitate orientation relationship depending upon the high-temperature reduction treatment. The optical absorption spectra of these materials were characterized by broad optical bands centered near 2.2 and 5.0 eV. By changing the size distribution of the precipitates, it was possible to obtain optical spectra of these materials suited for solar selective absorber applications. The dislocations and sub-boundaries provided nucleation sites for the formation of precipitates, and decreased the temperature of reduction. By providing sources of dislocations, the precipitates enhance the ductility and inhibit the propagation of cracks.

The introduction of a dispersion of small (≃ 1μ1m) insulating particles into an ionic conductor results in an increase in the ionic conductivity. This behavior is not in accord with classical theory. A number of different examples such as a dispersion of a metallic conductor in a semiconductor, a dispersion of an insulator in a semicondutor, and a dispersion of a metallic conductor in an ionic conductor all yield unusual transport behaviors when the surface area to volume ratio of the dispersoid is large. Under such conditions, a space charge layer or double layer adjacent to the dispersoid must be considered. Practical applications such as solid electrolytes, electrodes, corrosion systems and geological systems were discussed.

A particular type of grain morphology observed in commercial polycrystalline compacts of α-alumina is discussed along with special features which have been found to be associated with this morphology. The particular feature of interest is the occurrence of grains which are referred to in this paper as being lath-like in appearance. The important features of these lath-like grains include the occurrence of faceted grain boundaries which are often associated with pores and the special geometric relationships between the grains and the surrounding matrix.

Cathodoluminescence (CL) emission from as-grown and indented LiNbO3 single crystals has been studied. The spectrum is strongly affected by deformation of the surface region. Three emission bands at about 410, 525 and 580nm have been observed. Irradiation with the electron microscope beam at very high current intensities causes a severe damage on the irradiated area. This effect can also be observed on the indented areas where the indentation pyramid can not be recognized after prolonged irradiation.

Photoemission from evaporated Polyethylene films was measured in the spectral region between 6 and 11 eV. A significant increase in photoemission yield was observed after the films were exposed to O2 and to air containing ozone and ions from a diffuse glow discharge at reduced pressures. The enhancement can be attributed to oxygen and ozone reacting with the defects in the evaporated polyethylene films, which are unsaturated carbon double bonds in the polyethylene chain. For the ozonized polymer films deposited on gold substrates, the photoelectron spectra show a relatively high density of occupied states at 6.8 ± 0.3 eV below the vacuum level of the polymer and a photoemission threshold of 3.8 ± 0.3 eV. The implication of the results to contact charging of polyethylene is discussed.

Defect and impurity valence changes induced by γ irradiation of nickel- doped MgO crystals are studied using magnetic resonance and optical techniques.

Alkali ions, which compensate for substitional Al3+, play an important role in the frequency stability of α-quartz crystals. In this work, low temperature dielectricloss measurements (between 2.9 and 300 K) are carried out on crystals that have been “swept” so as to introduce either Li+ or Na+. High quality synthetic crystals as well as natural crystals are employed. The well known loss peaks due to Al-Na pairs are further explored and similar peaks due to Al-Li are sought after but not found. It is concluded that the Al-Li pair is oriented along the C2 -axis of the A104 distorted tetrahedron. After irradiation, large peaks are observed at very low temperatures both in Li+- and Na+-containing crystals. These peaks, which are distorted below ∼6 K due to the onset of quantum effects, may originate in alkali centers produced when alkali ions are liberated by the irradiation.

A time delay has been observed in HgCdTe in the photoplastic response of these crystals. Both the magnitude of the time delay and the total change in stress, Δσ, are dependent on the wavelength of the light. These observations provide a basis for establishing that the photoplastic process is a dynamic one, in contrast to prior treatments of the problem as a static process. Solution of the kinetics associated with this time delay allow us to show the dynamical character of photoplasticity.

Physical and structural property changes resulting from ion implantation and thermal annealing of α-A12O3 are reviewed. Emphasis is placed on damage production during implantation, damage recovery during thermal annealing, and impurity incorporation during thermal annealing. Physical and structural property changes caused by ion implantation and annealing are correlated with changes in the mechanical properties.

Single crystal yttria stabilized zirconia was implanted with 100 keV Ca+, Al+, and O2+ ions at fluences of 1015 to 6 × 1016 ions/cm2; . Blistering was observed at doses of 3 × 1016; O2;+ cm−2; and 6 × 1016; Al+ cm−2; but none was evident with Ca+. Knoop microhardness with a shallow indenter penetration depth peaked at a dose of 1016; ions/cm−2; for both Al+ and O2;+ but Ca+ produced no effect on microhardness. Vicker's microhardness with a much greater indenter penetration depth was not changed detectably by implantation but fracture toughness measurements from the same Vicker's indentations exhibited 10–23% increases at the highest O2+ doses and 20–25% increases at high Al+ doses. Annealing the highest implant doses at 1200° reduced the fracture toughness to pre-implant levels. Reflection electron diffraction showed that the surface had not been made amorphous by the 6 × 1016; Al+ dose as a well crystallized diffraction pattern was obtained.