Diamond films of 15–20 μm thicknesses were prepared by microwave plasma enhanced chemical vapor deposition onto crystalline silicon substrates. The growth surfaces of the films were rough with polycrystalline crystallographic habits, while the substrate sides of these films were smooth and featureless as viewed by optical microscopy. A heated cast iron scaife was used to polish the rough growth surfaces, and free-standing films were removed from the silicon substrates by dissolving the silicon in an aqueous HF. Both infrared optical transmission and reflection spectra were measured over the range of 600–4000 cm−1. For polished films, near 70% transmittance was obtained over the whole range, while the transmittance for nonpolished films was much lower and varied strongly with the wave number. Absorptions due to carbon-hydrogen stretching bands as well as a silicon carbide phase were observed in the transmission spectra. The optical absorption coefficient and the refractive index were found to vary from as high as 150 to as low as 7 cm−1 and 2.41 to 2.49, respectively (depending on the film quality and the wave number). A weak signature of the two-phonon absorption band of diamond was observed. The relationship between deposition conditions and infrared optical properties of diamond films before as well as after polishing is discussed.

The effects of additions of Fe and Zn on the relative stabilities of the cubic L12 and tetragonal DO22 structures in TiAl3- and NbAl3-base alloys are evaluated using ab initio electronic band calculations. The Fe or Zn distribution on the aluminum sublattice is modeled by a periodic array which corresponds to alloying additions of 12.5 or 25 at.% Fe or Zn. Addition of 12.5 at.% Fe is sufficient to stabilize the L12 structure in both TiAl3- and NbAl3-base alloys. Interpolation of the structural energy differences suggests that the DO22 and L12 structures have the same energy at ∼4.5 at.% Fe in Ti(Al, Fe)3, in agreement with experiment,1,2 and at ∼11 at.% Fe in Nb(Al, Fe)3. The L12 stabilization effect per atom of Zn in Nb(Al, Zn)3 is approximately half as large as that of Fe. The stable structure is the one for which the Fermi energy lies in a minimum in the density-of-states (DOS) distribution, as in the binary compounds. The results are explained on the basis of a simple model involving charge transfer to alloying additions on the Al sublattice.

Detailed Fourier line shape analyses considering x-ray diffraction profiles from fault unaffected 10.0, 00.2, 11.0, 20.0, 11.2, and 00.4 reflections and fault affected 10.1, 10.2, 10.3, 20.1, 20.2, 10.4, and 20.3 reflections have been performed on three magnesium base hexagonal alloys used extensively in the aircraft industry. The first of the three alloys (Mg–Al–Mn, Alloy I) had the nominal composition in wt.% of Al-8.3, Mn-0.35, Si-0.2, Cu-0.12, Fe-0.2, and other 0.01; the second alloy (Mg–Zn–Mn, Alloy II) had the nominal composition in wt.% of Zn-4.0, Mn-0.15, Si-0.01, Cu-0.03, Fe-0.01, Zr-0.70, and rare earth elements-1.50; and the last of the alloys (Mg–Zn–Al, Alloy III) had the nominal composition in wt.% of Zn-4.3, Al-0.15, Mn-0.01, Si-0.03, Cu-0.01, Ni-0.005, Zr-0.6, and rare earth-1.4. The microstructural parameters determined in these analyses indicated the average domain size in alloys I, II, and III as 208 Å, 314 Å, and 400 Å, respectively. The deformation fault densities, α, in these alloy systems (∼54 ⊠ 10−3, 35 ⊠ 10−3, and 28 ⊠ 10−3, respectively, in alloys I, II, and III) were found to be appreciably high compared to the earlier work on pure magnesium (0.63 ⊠ 10−3). The deformation twin fault density, β, which was found to be negligible in pure magnesium (∼0.21 ⊠ 10−3), was found to be negative here, also indicating the negligible presence of twin faults in these alloys. These results establish that on cold work the solutes introduce deformation stacking faults in an appreciable quantity in magnesium which is not normally susceptible to faulting when in pure form. Of these three alloys, however, Alloy I (Mg–Al–Mn) was found to be the most prone to deformation faulting.

The structure of pulsed laser irradiated graphite surfaces has been elucidated. The pulse fluences range up to 4 J cm−2 with durations no longer than 30 ns. The region exterior to the irradiated spot is littered with ∼0.1 μm diameter carbon spheroids. The boundary region is characterized by both spheroids and torn layers 1–5 μm. in lateral extent. The central region displays carbon spheroids and surface upheavals. The carbon spheroids are attributed to hydrodynamic sputtering of carbon. The surface upheavals and torn carbon layers are attributed to constrained thermal expansion and contraction of the irradiated region. It is estimated that a nearly instantaneous 1000°C temperature change is necessary to cause the observed surface deformation. Pulse fluences in excess of 0.8 J cm−2 cause a thin layer of carbon to melt. This is proven by the fact that the irradiated layer in the solid phase has a turbostratic structure. Electron diffraction experiments and dark-field imaging experiments show that the basal plane grain size of the resolidified material varies from ∼20 Å at the melt threshold to ∼100 Å for samples irradiated with 4.0 J cm−2.

The diffusion-accommodated sliding of irregularly shaped grain boundaries in two-dimensional bicrystals is considered. The following assumptions are made: the grains adjoining the boundaries are rigid, the boundaries do not support any shear stresses, sliding displacements are infinitesimal, and sliding is accommodated only by grain boundary diffusion. The solution to this problem is illustrated for a bicrystal with a grain boundary consisting of three segments. The results of calculations involving up to 35 segments agree with Raj and Ashby's theory for the sliding of periodic boundaries. The influence of boundary conditions on the normal stress distributions along grain boundaries is examined. Zero-flux conditions at the intersection of a grain boundary with a free surface, which correspond to low surface diffusivities, can lead to high normal grain boundary stresses. The stress distributions and sliding rates of boundaries containing randomly spaced equisized bumps or equispaced bumps of random size are compared to the periodic case (i.e., equispaced equisized bumps). Substantial normal stresses can build up at such nonperiodic grain boundaries.

We describe a low-temperature solid-state interdiffusion technique that allows reaction between spatially separated reacting species and its application in the Al–Ru alloy system. This technique uses a liquid-metal solvent (Bi) as a medium for the transfer of Al to the surface of Ru powder where reaction occurs with the formation of nanocrystalline AlxRu1−x product phases. X-ray diffraction measurements are used to follow the time and temperature dependence of the reaction. Cross-sectional transmission electron microscopy allows direct imaging of the growth and morphology of the AlxRu1−x product phases.

The mixing of multilayers of two pure metals or alloys A and B with strong negative heat of mixing ΔHm can result in rapid heat release provided that the mixing time is short compared to the time to dissipate ΔHm in the environment, which is the case for sufficiently thin layers. The mixing reaction sets off the heat release when the composite is heated up to temperatures near or equal to the melting temperature of the metal layer with the lower melting point. Experimental results are given for the mechanisms and for the kinetics of the exothermic reaction. Possible applications of these materials are given.

An experiment was designed to follow the formation of higher member phases from lower member phases in the Tl2Can−1Ba2CunO2n+4 series. Coarse particles of the lower member phase were directly interfaced with various Ca–Cu–O sources and the pressed pellets were heat treated in air in a gold pocket at 850 °C for a period of 2–4 h. The reacted samples were polished and examined by scanning electron microscopy with energy dispersive x-ray spectrometry analysis (EDS). For the 2021 based reaction systems, the results suggest that the 2122 phase can be formed via reactions which involve the decomposition of the 2021 phase, the formation of a complex liquid phase, and finally the reaction between this liquid phase and a Ba/Ca compound. The formation of the 2223 phase from the 2122 phase is likely to follow a similar reaction route. The involvement of a liquid phase in the formation of the higher member phases may lead to a glassy phase, and hence to a barrier to the achievement of a high critical current density.

Chalcogenide glasses of the systems Ge–Sn–Se, Ge–Se–Te, and Ge–Sn–Se–Te have been prepared. Several compositions were found suitable for drawing fibers for CO2 laser radiation (λ = 10.6 μm) transmission. The glasses were characterized by x-ray diffraction, DSC (Differential Scanning Calorimetry), SEM with EDX analysis, FTIR spectrometry, density, and microhardness measurements. The glass transition temperature and microhardness of Ge–Se–Sn and Ge–Sn–Se–Te glasses decreased with increasing Sn content, for most of the samples. The region of high IR transparency of Ge–Se–Sn, Ge–Se–Te, and Ge–Sn–Se–Te glasses was slightly expanded (1–2 μm) toward longer wavelengths, compared to Ge–Se glasses, mainly for the glasses containing 70 at.% Se. The intensity of the impurity absorption peak of Ge–O (at λ ∼ 12.8 μm), which usually appears in Ge–Se glasses, was reduced or absent in Ge–Sn–Se–Te glasses. The best fibers were produced with the glass composition Ge–0.8Sn0.2Se3.5Te0.5. An attenuation of 20 dB/m at 10.6 μm, and a transmitted maximum power density of 2.4 ⊠ 106 W/m2 were measured. The mechanical and optical characteristics of these glasses have been related to the glasses structure. Corresponding to the reduced masses of the bonds formed in the Ge–Sn–Se–Te system (in the amorphous region), it is expected that the multiphonon edge is slightly shifted. As a consequence, as was measured, the transparency region has been expanded by less than 2 μm toward longer wavelengths.

Mixtures of Ge–Sn and Ge–Pb powders were ball-milled to form a fine dispersion. After 32 h of milling the average diameter of the hard Ge particles embedded in the Sn (or Pb) matrix was about 10 nm. As the Ge concentration was increased in each system, the melting point, TM, and the enthalpy of fusion, ΔHM, of Sn (or Pb) decreased. Only small changes in ΔTM and ΔHM were observed after heating cycles in the DSC to above the melting point. The melting endotherm measured by DSC disappeared for Ge-rich compositions (88 and 95 vol.% Ge for Ge–Sn; 93.5 vol.% Ge for Ge–Pb). It is suggested that atomic disorder/melting is nucleated at the Ge/Sn (or Ge/Pb) interfaces and the melting point and enthalpy of fusion decrease as the interfacial area increases. When the Ge volume reaches a value where essentially all the Sn (or Pb) atoms are adjacent to the Ge particle surfaces, the Sn is in a disordered–perhaps amorphous–state such that no melting transition is observed.

The angles between the zone axes of decagonal quasicrystals in two different Al–Mn alloys have been studied by means of precision tilt experiments in a transmission electron microscope. The results confirm that a tenfold symmetry axis is present in the reciprocal lattice. The angles between the principal axes in the decagonal phase do not coincide with those between the corresponding axes in the icosahedral phase. The presence of this distortion is independent of quenching rate and size of the decagonal quasicrystals.

The kinetic master equation is developed in the pair approximation to study disorder ⇉ B2 order transformations in bee binary alloys. Coupled sets of rate equations for the pair variables are obtained for atom movements by either the direct interchange mechanism or the vacancy mechanism. Numerical integrations provide the nonequilibrium relaxations of short- and long-range order (SRO and LRO) and the vacancy balances between the two sublattices. For binary alloys, disorder ⇉ order transformations were studied for all combinations of interatomic interaction strengths, activation barrier heights, and alloy stoichiometry variations. After a transient vacancy relaxation, all cases began with a quick relaxation of SRO, followed later and independently by the growth of LRO and additional SRO. There were some variations in kinetic path through SRO and LRO, moderate variations in overall kinetics, and large variations in vacancy balance. Some nonphysical aspects of kinetics in the pair approximation are discussed.

The room-temperature tensile properties, fracture mode, and grain boundary chemistry of undoped stoichiometric NiAl, as well as NiAl doped with boron, carbon, and beryllium, have been investigated, Pure, stoichiometric NiAl fractures with limited tensile ductility in a predominantly intergranular manner. Auger analyses revealed that the grain boundaries in NiAl are extremely clean and free of any segregated impurities, indicating that they are intrinsically brittle. Boron, when added to stoichiometric NiAl at a bulk level of 300 wt. ppm, segregates to the grain boundaries and suppresses intergranular fracture. However, there is no attendant improvement in tensile ductility because boron is an extremely potent solid solution strengthener in NiAl, more than doubling its yield strength. As a result, any potential benefit of improving grain boundary strength is more than offset by the increase in yield strength. Unlike boron, both carbon (300 ppm) and beryllium (500 ppm) are ineffective in suppressing intergranular fracture in NiAl, and Auger analyses of the C-doped alloy revealed that carbon did not affect the fracture mode because it did not segregate to the grain boundaries. Although neither beryllium nor carbon suppressed grain boundary fracture, their effects on the tensile ductility of NiAl were quite different: the ductility of the Be-doped alloy was higher than that of the B-doped alloy because beryllium, unlike boron, has a rather modest strengthening effect in NiAl, whereas the C-doped alloy was brittle like the B-doped alloy, because carbon is a potent solid solution strengthener, just like boron. These observations were rationalized by considering a hard-sphere model for interstitial and substitutional sites in NiAl. It was concluded that boron and carbon occupy interstitial sites, whereas beryllium dissolves substitutionally. In all the alloys that were investigated, the Ni and Al contents of the grain boundaries were not significantly different from the bulk levels, and no evidence was found for B–Ni cosegregation.

Four kinds of pitch precursor fibers possessing different mean sizes of mesophase domains in their transverse sections, but having the same chemical constituents, were prepared by controlling only the spinning conditions. The fibers were thermoset and heat-treated at various temperatures from 600°C to 3300 °C. Lc (002) or d002 vs THT curves showed that pitch precursor fibers with finer transverse microstructures have a lower degree of graphitizability. Further, these four kinds of fibers exhibited systematically different d002 vs Lc (002) curves. The resistivity at room temperature is dependent on d002, being independent of the mean size of the mesophase domains in the transverse section of the pitch precursor fiber. The transverse magnetoresistance at 77 K and 10 kG, (Δρ/ρ)t, also showed a universal dependence on d002 if (Δρ/ρ)t was positive. However, the (Δρ/ρ)t vs d002 curves seemed to be slightly split in the negative region of (Δρ/ρ)t for each of the four kinds of fibers.

The Al–Ni2Si reactions were studied in lateral diffusion couples containing Al islands on Ni–Si multiple layers. The samples were first in situ annealed in a transmission electron microscope at a temperature of 370°C for 5 min to form the Ni2Si phase in the multiple-layer area. Then they were in situ annealed at temperatures ranging from 498–545 °C. During the second-step anneal, a sequential formation of Al3Ni, Al3Ni2, and Ni3Si2 was observed. After the nucleation of the third phase (Ni3Si2), the three phases grew simultaneously with time. The lateral growth of Al3Ni and Al3Ni2 is a result of the Al diffusion and the Al–Ni silicide reactions; the lateral growth of Ni3Si2 is caused by the diffusion of Si atoms dissociated from the silicides.

Post-deposition rapid vacuum annealing of tantalum nitride (Ta2N) thin film resistors (TFR) was successful in improving the temperature coefficient of resistance (TCR) to ±5 ppm/°C with starting TCR at about −140 ppm/°C. A subsequent aging study revealed degradation of the nichrome (NiCr) contact interlayer. Two improved contact layers, TiW and Tamelox (Ta/NiCr), were compared. The structural grain growth induced by the annealing effect resulted in Ta2N films having 100–1000 Å polycrystals in an amorphous matrix. The corresponding current conduction mechanisms were identified with a substrate-assisted tunneling model. The frequency response predicted potential applications to 100 GHz.

Samples of CeSi1.86 which exhibit Kondo behavior are shown by neutron powder diffraction and transmission electron microscopy to consist of two closely related tetragonal phases. The primary phase is of the ThSi2 structure type with some vacancies in the silicon sublattice. The second phase presents an ordering of these vacancies. These two phases coexist at low temperature, but the abundance of the second phase increases with decreasing temperature. Neutron diffraction measurements and TEM experiments show that the phase separation occurs reversibly around 260 K, in close relation with an anomaly in the transport properties. The presence of a hysteresis indicates that we are dealing with a first order transition.

The epitaxial growth of Cu on Si(111) substrate at room temperature was achieved using the Partially Ionized Beam (PIB) deposition technique in a conventional (10−4 Pa) vacuum without prior in situ cleaning of the substrate or post-annealing of the film. The beam contained ≍2% of Cu self-ions, and a bias of 0 to 4.2 kV was applied to the substrate during deposition. X-ray diffraction studies showed the existence of a twin structure in the epitaxial Cu layer deposited at 1 kV. A mechanism of epitaxial growth of Cu(111) on Si(111) substrate via an η″—Cu3Si intermediate phase is proposed. Based on the crystal structure of η″—Cu3Si, it is demonstrated that the geometrical lattice matching concept provides a simple picture of lattice continuity at the interface in this epitaxial system.

Scanning and isothermal calorimetry, together with x-ray diffraction and high resolution transmission electron microscopy (TEM), have been used to characterize Al–Mn and Al–Mn–Si films sputtered onto substrates at 60 °C, 45 °C, and −100 °C. In the case of Al0.83Mn0.17, the monotonically decreasing isothermal calorimetric signal, characteristic of a grain growth process, has proved decisive in identifying the as-sputtered “amorphous” state as microquasicrystalline, with an average grain size of ∼ 20 Å, in agreement with an estimate of correlation range from the x-ray pattern. The TEM at 400 keV reveals well-defined atomic or lattice images in annealed films but only barely resolved grains (ordered clusters) in the as-sputtered films. The relation between the metallic glass and the microquasicrystalline state in these alloys is discussed.

The microstructure of anodic oxide films grown on Ti–6A1–4V in H2SO4 was investigated by SEM, TEM, STEM, EDX, and AES, as a complement to a recent surface spectroscopic investigation of the same oxides by XPS, AES, and SIMS. The anodic oxide films are heterogeneous and the texture reflects the duplex microstructure (α and β phases) of the underlying metal. Porous oxide regions are observed with different appearances on α-phase and mixed-phase regions. The oxide films are essentially amorphous in the investigated thickness range 60–300 nm (in contrast to as-grown anodic films on pure Ti), but crystallize to the anatase structure upon annealing. Considerable lateral variation of the V content in the oxide is observed, reflecting the corresponding variation in the underlying metal. The results are compared with a previous, similar investigation of anodic oxides on pure Ti.