The 27Al MAS-NMR spectra of alumina aerogels were determined. The spectra showed different profiles according to the supercritical methods used during drying. The alumina aerogel supercritically dried in a CO2 extractor (80 °C, 15.7 MPa) displayed a pentahedral-coordinated aluminum structure. On the other hand, the alumina aerogel supercritically dried in an autoclave (270 °C, 26.5 MPa) revealed no pentahedral-coordinated aluminum structure. The absence of pentahedral-coordinated aluminum was one factor for the lower temperature of transformation from the γ to the θ phase.

The microstructures of YBa2Cu3O7−δ (YBCO) thin films grown on Si with Y-stabilized ZrO2 (YSZ) and Y2O3 buffer layers were characterized by means of high-resolution electron microscopy. At the Si–YSZ interface, a 2.5 nm thick layer of regrown amorphous SiOx is present. The layer is interrupted by crystalline regions, typically 5 to 10 nm wide and 10 to 50 nm apart. Close to the crystalline regions, {111} defects are present in the Si substrate. The typical defect observed is an extrinsic stacking fault plus a perfect dislocation close to the stacking fault which terminates extra {111} planes in the upper part of the Si. These defects are probably formed by condensation of Si self-interstitials created during oxide regrowth. Precipitates are present in the Si close to the Si–YSZ interface and indicate that in-diffusion of Zr has occurred. The YSZ–Y2O3 interface is atomically sharp and essentially planar and contains no second phases. Perfect misfit dislocations with Burgers vector 1/2〈110〉 are present at this interface along with unrelaxed elastic misfit stresses. The Y2O3–YBCO interface is atomically sharp and planar, but contains steps. (001) stacking faults are present in the YBCO above these steps; the faults are, however, healed a few unit cells away from the interface. By HREM analysis of ultrathin specimen areas, the atomic layer of the YBCO closest to the Y2O3 was found to be a barium-oxygen layer.

The reaction between alumina and yttrium barium cuprate subjected to a melt-texturing heat treatment was studied. Microstructural examination of quenched, partially transformed samples revealed that at ∼1050 °C (which is above the incongruent melting temperature of YBa2Cu3O6+x) a reaction layer forms at the alumina interface. The reaction products were identified as Ba6Y2Al4O15 and a copper-rich liquid phase. On cooling below the peritectic temperature, aligned domains of YBa2Cu3O6+x (123) were observed to nucleate preferentially at the reaction layer. For samples of melt-textured 123 deliberately seeded with alumina particles, it was found that nucleation and growth of 123 occurred exclusively at the particles. A reaction sequence for the formation of the Ba6Y2Al4O15 is put forward, together with a discussion of the possible nucleation mechanisms for the 123.

Processing atmosphere and precursor effects on liquid phase formation were investigated in the BaO–CuO thin film system. Liquid phase was observed at a significantly lower temperature (∼620 °C) under N2 atmosphere with carboxylate solution precursors. In the processing of the YBCO oxide superconductors, formation of the liquid phase facilitated the decomposition of BaCO3 and crystallization of the YBCO grains. Larger grains and greater (00l) preferred orientation were also observed in the YBCO films prepared with the presence of the liquid phase.

The thin films (Y/Ba/Cu = 1.0/2.7/4.7) which have a strong intensity corresponding to the c-axis peaks of the 124 phase (YBa2Cu4O8) in x-ray diffraction (XRD) patterns were prepared. The film structure and surface morphology of these films were observed by scanning electron microscopy (SEM), energy dispersive x-ray (EDX), transmission electron microscopy (TEM), and atomic force microscopy (AFM). It was found that the film on MgO(100) clearly has grain boundaries with many a-axis oriented grains of the 123 phase (YBa2Cu3O7−x). However, the film on SrTiO3(100) had a smooth surface, and a-axis oriented grains of the 123 phase couldn't be observed. This difference could not be explained by only the difference in the lattice mismatch.

The solidification processes for the compositions Bi2Sr2CaCu2Oy (2212) and Bi2Sr1.75Ca0.25CuOy (2201) were determined as a function of oxygen partial pressure. During solidification in argon, the superconducting phases were generally not observed to form for either composition. In both cases, the solidus is lowered to approximately 750 °C. Solidification of Bi2Sr1.75Ca0.25CuOy in Ar resulted in a divorced eutectic structure of Bi2Sr2−xCaxOy (22x) and Cu2O while solidification of Bi2Sr2CaCu2Oy in Ar resulted in a divorced eutectic structure of Bi2Sr3−xCaxOy (23x) and Cu2O. Solidification of Bi2Sr1.75Ca0.25CuOy in O2 resulted in large grains of 2201 interspersed with small regions containing the eutectic structure of 22x and CuO/Cu2O. Solidification of Bi2Sr2CaCu2Oy in partial pressures of 1%, 20%, and 100% oxygen resulted in multiphase samples consisting of 2212, 2201, some alkaline-earth cuprates, and both divorced eutectic structures found during solidification in Ar. For both compositions, these latter structures can be attributed to oxygen deficiencies present in the melt regardless of the overpressure of oxygen. These eutectic structures are unstable and convert into the superconducting phases during subsequent anneals in oxygen. The formation process of the 2212 phase during solidification from the melt was determined to proceed through an intermediate state involving the 2201 phase.

Superconducting Bi–Sr–Ca–Cu–O thin films were obtained from post-annealing partially crystallized and amorphous films grown on MgO(100) by pulsed laser deposition. The substrate temperature investigated was in the range of 350–750 °C, over a range of pressure 0.1 to 100 mTorr. The as-deposited films were annealed in 7.5 vol.% O2/Ar or in air at 800–865 °C from several minutes to a few hours. Unlike the pure Bi2Sr2CaCu2O8+δ (2212) phase (Tc = 80 K) which is easily formed after a long continuous period of post-annealing at a temperature below 830 °C, the formation of (Bi, Pb)2Sr2Ca2Cu3O10+δ (2223) phase from the as-deposited amorphous films requires repetitive annealing cycles of short duration in air at 850 °C to simultaneously achieve good crystal quality, small surface roughness, and sharp diamagnetic transition (Tc = 110 K). After annealing, the temperature is lowered down to ∼650 °C by quenching in air and then a slow-cooling step is employed. This procedure was found to enhance the volume fraction of the 2223 phase as compared with a direct slow-cooling process. The trade-off between annealing temperature and time was observed to affect the phase formation and the smoothness of the annealed films. To optimize the post-annealing conditions, Rutherford backscattering spectrometry, x-ray diffraction, and scanning electron microscopy were systematically used to examine the composition, structure, and morphology of the films, respectively.

The 2212 phase formation during annealing of melt textured Bi–2212 (Bi2Sr2CaCu2Ox) was investigated using differential thermal analysis, thermal gravimetric analysis, x-ray diffraction, scanning electron microscopy, energy dispersive x-ray analysis, and high resolution transmission electron microscopy. After zone melting, the material is multiphase consisting of 2212, 2201, Sr1−xCaxCuO2, and the eutectic. The 2212 phase formed is highly perfect with less than 5% intergrowths of 2201 layers; the 2201 phase shows no intergrowth of 2212 at all. In the first period of the annealing, remelting of the eutectic leads to fast oxygen diffusion and a high 2212 formation rate. The 2201 → 2212 transformation proceeds via intermediate states of high defect density. The 2212 grains contain up to 30–70% 2201 intergrowths. Further heat treatments lead to an annihilation of the great majority of intergrown 2201 layers. We propose a model for the formation of 2212 grains with a low planar defect density, based on frequent stacking faults, that allows diffusion of Ca- and Cu-atoms over a short distance. The model provides a schematic description of this solid-state process and correlates it to the characteristic microstructural features of melt-processed Bi–2212.

Ac susceptibility measurements have been used to monitor the changes in the superconductive properties of sintered and uniaxially pressed samples of Pb-stabilized 2223-phase bismuth cuprate ceramic as the structural damage was annealed in air in a sequence of steps in temperature between 500 and 850 °C. It is concluded that below 600 °C a relaxation of residual stresses is responsible for a 2% shrinkage in sample volume and a small improvement in bulk superconductive transition temperature, Tc. Above 700 °C, a recovery of the original properties occurs through grain regrowth governed by an activation energy of ∼200 kJ/mol. However, in the region between 600 and 700 °C, a decrease in Tc of ∼40 K appears to be the result of plastic flow and amorphization associated with local decomposition of 2223 which reduces the effective grain size and weakens the intergranular superconductive coupling.

Silver doping into (Bi, Pb)2Sr2Ca2Cu3O10 superconducting composite tapes was found to accelerate the formation process of high-Tc (2223) phase owing to lowering the partial melting point of the samples. The differential thermal analysis (DTA) results showed that the partial melting temperature of the sample was lowered by about 10 °C from 850 °C to 840 °C by silver doping. However, with sufficient sintering both the silver-doped and undoped samples can reach a very high level of high-Tc phase fraction, suggesting that the silver doping only speeds up the rate of high-Tc phase formation, but does not change the final phase assemblage of the materials. The reaction kinetics was analyzed by using the Avrami equation, and the results indicated that the conversion process of low-Tc (2212) phase to high-Tc (2223) phase was a diffusion-controlled, two-dimensional reaction. The correlation of the phase evolution with electrical property inside the superconducting tape during the process of heat treatment was also discussed.

This paper, the second of three parts, presents a detailed analysis of the crystallographic texture observed in lead lanthanum titanate (PLT) thin films deposited by multi-ion-beam reactive sputtering. Since films were deposited at a substrate temperature of 100 °C, they exhibit an amorphous structure that can be described by the structure zone model. The as-deposited microstructure is transformed via crystallization of the perovskite phase and PbO evaporation during postdeposition annealing. Transmission electron microscopy was used to determine that phase pure PLT films develop textured clusters (as large as 700 nm in diameter) consisting of 〈100〉 aligned 10 nm nanocrystals. As excess PbO is added to the PLT film, the textured cluster size decreases until only isolated PLT nanocrystals exist. Below a critical quantity of excess PbO in the as-deposited film (approximately 15 mole%), the textured cluster structure produces a 〈100〉 texture through the film thickness, which generates x-ray diffraction patterns typical of textured microstructures. At high PbO contents, the excess PbO causes the formation of clusters that are smaller than the film thickness, resulting in a polycrystalline-type x-ray diffraction pattern. A qualitative model describing microstructure development is presented.

This paper, the third and final of a three part series, presents the electrical properties of postdeposition annealed, lead lanthanum titanate (PLT) thin films deposited by multi-ion-beam reactive sputtering (MIBERS). Also, a model is presented that explains the relations among composition, crystallographic structure, microstructure, and electrical properties of the PLT thin films. Thin films of PLT consisting of the perovskite phase exhibit 〈100〉 textured microstructures. Addition of a critical quantity of excess PbO results in the loss of this 〈100〉 texture, and continuity of the perovskite phase is disrupted while both excess PbO and porosity phases become continuous due to a percolation effect. Films with textured microstructures consisting of a continuous perovskite phase exhibit relatively high dc resistivities, high dielectric permittivities, and high remanent polarizations. At the transition between textured and nontextured microstructures, a discontinuous drop in the electrical properties occurs due to the ensuing continuity of the excess PbO and porosity. These composition-induced changes in the electrical properties were quantitatively modeled by applying a simple mixing rule model to the microstructure model developed in Part II of this series.

A resonant frequency mechanical vibration method was used to nondestructively detect fatigue crack initiation in notched 304 stainless steel samples prepared by powder metallurgy. This method allowed the determination of an effective elastic modulus and the direct measurement of internal friction. Changes in the modulus and internal friction were found to correlate well with the presence of 50 μm long fatigue cracks. The length of the through cracks initiated at the notch was measured using surface replicas, which were examined in a scanning electron microscope. Small crack growth rate data were also obtained and plotted versus the stress intensity factor. The crack growth rates were compared with long crack data performed on compact tension samples of this material. The short cracks grew at intensity factors below the long crack threshold stress intensity factor.

Earlier it was shown that some liquid metals may facilitate the cutting of hard materials, being strongly surface-active media with respect to these solids (decreasing their surface energy), e.g., Zn for steels, Cd for titanium alloys, etc. Recently, a similar physico-chemical principle has been proposed and realized in the course of active metals (Zn, Cd) depositing on the surface of a treated sample (ground tempered steel, 50 HRC) by means of reduction of their cations on a sample used as a cathode in a corresponding electrochemical cell, at room temperature, with a very small amount of active substance, and at very low current densities (∼50 mA/cm2); under such conditions, grinding is 1.5–2 times more effective than in the same aqueous solutions without current.

Mild steel plates were welded under stationary and dynamic (longitudinal and transverse vibration) conditions. The frequency and amplitude ranges selected were 0 Hz to 300 Hz and 0 μm to 30 μm, respectively. It has been observed on the basis of experimental findings that vibration, in general, reduces the grain size of the weldment, and this reduction is appreciable at a lower amplitude of vibration. This may be due to the lower energy required for nucleation of the solid phase. Fast removal of latent heat of solidification from the solid liquid interface plays a significant part in the grain refinement under vibration.

Tracer diffusion of 57Co in amorphous Fe100−xDyx (x = 20–40), Fe75Tb25, Fe67Ce33, and Fe80Si6B14 alloys prepared by dc sputtering has been studied at temperatures of 523 and 573 K. In the Fe–Dy alloys the diffusion coefficient of 57Co shows a maximum at 33 at.% Dy. The magnitude of the diffusion coefficient of 57Co in Fe75Tb25 is nearly equal to that in Fe75Dy25, while those in Fe67Ce33 and Fe80Si6B14 are about one order of magnitude less than the values in Fe67Dy33 and Fe80Dy20. This suggests that the atomic size of the diffusant and the density of the matrix are dominant in the diffusion. Temperature dependence of the diffusion coefficient D of 57Co in the amorphous Fe75Dy25 alloy has been determined in the range from 493–673 K. It shows a linear Arrhenius relationship expressed by D = 5.7 × 10−2 exp(−199 kJ mol−1/RT) m2 s−1. The magnitudes of the pre-exponential factor and the activation energy suggest that the cobalt tracer atoms in the amorphous Fe75Dy25 alloy diffuse by an interstitial-like mechanism.

Using Chemical Vapor Deposition (CVD), TiN was deposited on sharp tungsten needles. The reactant gases were TiCl4, N2, and H2. A Transmission Electron Microscopy (TEM) investigation revealed that the first nuclei of the CVD–TiN coating on tungsten did not consist of δ–TiN, but were a mixture of α–TiN and δ–TiN. These results were also verified with x-ray measurements. From these experimental results a possible mechanism for the initial growth of TiN on tungsten is suggested. It may be that the change in relative concentrations of the different titanium nitrides suggested as mechanism of the initial growth of CVD–TiN can be applied in general for all TiCl4/H2/N2/metal systems where the original substrate surface material partly or completely consists of a metal with catalytic properties.

Diamond films were grown by microwave plasma-assisted chemical vapor deposition using mixtures of 13CH4 and CO. Mass spectrometry was used to identify CO, CH4, and C2H2 as the stable gaseous products in the reactor exhaust gas. By comparing gaseous 13C compositions with that of the diamond films, the efficiency of diamond growth from methane (possibly via the methyl radical) is found to be about two orders of magnitude higher than that for carbon monoxide. Most of the diamond that is formed from the CO reagent results from the conversion of CO to hydrocarbons. The conversion of CO to hydrocarbons is attributed to activation of CO by high-energy electrons in the plasma.

Diamond growth experiments were performed in a microwave plasma ball reactor on silicon wafers or on a molybdenum sheet provided with cones (stamped into the sheet with a punch). All substrates had been treated by scratching with diamond powder in advance. The gas mixture used was CH4/H2, sometimes with the addition of CO. Substrate temperatures ranged from 953 to 1428 K, pressures from 100 to 400 mbar, and microwave powers from 250 to 700 W. A strong preference of diamond growth was observed on the cones in the molybdenum substrates. This is interpreted as being caused by gas transport hindrance. The resulting deposition coefficient of the “active” species is about 0.1 under all conditions investigated. The deposition experiments on silicon substrates are numerically modeled in two steps. In the first step, temperature fields and electron density and energy distributions in pure hydrogen are calculated following the method described previously. The output of this first simulation step is taken as input data for the second step. The condition is applied that chemical reaction rates due to thermal or electronic activation and diffusional flows compensate each other at every point of the reactor. In this way stationary concentrations of the 13 species in 29 elementary reactions are computed and, from these, the expected deposition profile of diamond on the silicon substrate, assuming one of the carbon-containing species to be the “active” one. When the experimental deposition profiles are compared with the calculated ones, C2H as the “active” species gives the best match to all the experimental results. CH3 and C2H2 (and perhaps others) might contribute to the diamond growth to a limited extent only.

We have deposited hydrogen-free diamond-like amorphous carbon films by ArF (193 nm) pulsed laser ablation of graphite. The deposition process is performed with a laser power density of only 5 × 108 W/cm2 at room temperature without any auxiliary energy source incorporation. The resulting films possess remarkable physical, optical, and mechanical properties that are close to those of diamond and distinct from the graphite target used. The films have a mechanical hardness up to 38 GPa, an optical energy band gap of 2.6 eV, and excellent thermal stability. Analysis of electron energy loss spectroscopy reveals the domination of diamond-type tetrahedral bonding structure in the films with the sp3 bond fraction over 95%. Compared with other reported results of pulsed-laser-deposited diamond-like carbon films, our experimental results confirm that the laser wavelength or photon energy plays a crucial role in controlling the properties of the pulsed-laser-deposited diamond-like carbon films.