The precipitation kinetics of the δ′ (Al3Li) phase in two rapidly solidified samples and one conventionally cast sample of an Al–2.3Li–6.5Mn–0.65Zr (in wt. %) alloy are compared. Following high cooling rates, manganese is retained in solid solution in the aluminum matrix (αAl) up to 6.0 wt.%, far beyond the thermodynamic equilibrium value (0.36 wt.% at 500 °C). Extended solid solution of manganese in aluminum induces strain gradients, similar to those produced by dislocations. The effect of such gradients, the size of which is proportional to the solute atomic fraction, is to enhance lithium precipitation by lowering the activation energy, as observed, and also by affecting the rate parameter. Kinetic thermal analysis has been performed in a series of nonisothermal experiments in the heat flux differential scanning calorimetry (DSC) mode. The precipitation of the δ′ (Al3Li) phase is evidenced by an exothermic peak whose characteristics were analyzed. The rate of transformation (precipitation) is assumed to obey the Johnson–Mehl–Avrami equation. The activation energy for the precipitation process as well as the kinetic rate parameter have been evaluated for the rapidly solidified and the conventionally cast specimens. The activation energy for precipitation is lowered, from 107.0 kJ mol−1 for the conventionally cast material, down to 81.8 kJ mol−1 and 77.0 kJ mol−1 for samples that exhibit manganese solid solubility extensions of 2.10 and 6.00 wt.%, respectively. The rate parameter for the precipitation reaction, which has the generally admitted value of 1.50, for a transformation involving diffusion controlled growth, is affected by the strain gradients, too. Its value is reduced from 1.40 for the slowly cast sample to 1.32 and 1.20, respectively, for the two rapidly solidified samples, as a result of competing mechanisms, namely: growth controlled by diffusion and strain-assisted precipitation.

The stress field, image force, and shielding effect of a screw dislocation in the vicinity of a Mode III crack were formulated for both semi-infinite and finite length cracks. The results show that there is an abnormal stress component, ŝ31, on the crack plane. This leads to a nonzero image force along the axis perpendicular to the crack plane when the dislocation is located on the crack plane. However, the abnormal stress component and image force disappear for orthotropic and isotropic media. The image force along a slip plane has the same expression as in isotropic media with an effective shear modulus. Generally the shielding effects are the same as in isotropic media. The anisotropy changes only the magnitude of the shielding effects. The case of multiple dislocations is also discussed.

Reaction of Ni–Al alloys within the β-NiAl phase with CrB2 was studied at 1473 K as a function of Al concentration in the alloy. Reaction of 49–50 at. % Al alloys with CrB2 occurred by interdiffusion of Ni into CrB2 and Cr into the alloy without forming a new product phase. On the other hand, a new product phase, rich in Ni and B, formed by the reaction of alloys having Al concentrations 48 at. % or lower with CrB2. The reaction product was observed both at the CrB2/alloy interface and along the alloy grain boundaries.

Small angle neutron scattering (SANS) was applied to achieve insight into the magnetic correlations in nanostructured Fe. The results confirm the expected microstructure involving ferromagnetic grains and a nonmagnetic or weakly magnetic interface region, the interfaces occupying about half the specimen volume. The SANS measurements further reveal that in nanostructured Fe the magnetic correlations are not confined to single grains, but are extended across the interfaces and result in the alignment of the magnetization over several hundreds of grains. An external field of 1.5 kOe is not sufficient for complete magnetic alignment of the entire specimen. However, the long-range magnetic correlations are considerably disturbed by this field. Reducing the external magnetic field to zero causes the magnetic correlations to resume microstructural characteristics similar to what they had in the original state.

The structural evolution and cationic stoichiometry of Bi(Pb)–Sr–Ca–Cu–O superconductors have been studied using transmission electron microscopy and x-ray microanalysis. The nature of the incommensurate modulation changes systematically as increasing amounts of lead are added. X-ray microanalysis studies reveal that lead replaces Bi in the structure. Pb addition improves the microstructural homogeneity leading to the formation of a nearly homogeneous sample consisting of the “2223” phase. Based upon all the experimental results, it is inferred that the role of Pb substitution is related to the thermodynamics and kinetics of the formation of the n = 3 phase.

A continuous annealing, using the Joule effect, is performed during the fabrication of a tinned copper wire. During this operation, which lasts between 1/10 and 1/100 of a second, the wire is brought in air to a temperature which exceeds the melting point of tin. The influence of the continuous annealing on the nature of the surface layer of the tin coating is studied using XPS. The thickness of the tin oxide film covering the metallic tin is shown to be reduced after the annealing. This result is confirmed by measurements of the electrical contact resistance using the crossed rods method.

The progress of solid-state reaction in Ni–Zr composite wires with different elemental layer thicknesses has been studied in detail. Besides x-ray diffraction and differential scanning calorimetry, dilatometric measurements, magnetization and resistivity measurements, and cross–sectional transmission electron microscopy were used to monitor the reaction during constant-rate heating and to characterize the various reaction products. An amorphous phase initially forms at the interface between the elemental layers. As soon as the layer thickness exceeds a critical value, the intermetallic NiZr phase appears at the interface between the amorphous phase and pure Zr, as shown by TEM investigations. This is due to a reduced velocity of the reaction front caused by the longer diffusion path enabling the intermetallic phase to become stable. As shown in experiments at a constant heating rate, a second intermetallic phase forms at higher temperatures at the interface between Zr and crystalline NiZr. The amorphous phase remains unchanged up to crystallization at about 520 °C. To obtain fully amorphous material, the interdiffusion reaction must be completed (or especially the Zr layers must be completely reacted) before the intermetallic NiZr phase starts to form. A criterion for achieving completely amorphous bulk material is derived.

A theory of vacancy formation in Cu3Au-type ordered fcc alloys is presented. The present theory, which is based on the pairwise bonding model, is found to be in good agreement with the experimentally observed vacancy properties in ordered Cu3Au and Ni3Al. Various shortcomings in the previous theoretical calculations have also been identified.

Phase equilibria in the ternary system Pr–Fe–Al have been established in an isothermal section at 800 °C from room temperature x-ray powder diffraction analysis of about 50 alloys, which were melted, annealed at 800 °C, and quenched. Phase equilibria are characterized by the formation of rather extended homogeneous regions, i.e., by a random substitution of Fe/Al in Pr(Al1−xFex)2, 0 ≤ x ≤ 0.15, in Pr2(Fe1−xAlx)17, 0 ≤ x ≤ 0.65, as well as by the formation of at least four ternary compounds. Whereas the existence of PrFe4Al8 with the CeMn4Al8-type structure has been confirmed, there were no indications for a compound “PrFe6Al6” earlier claimed to crystallize with the ThMn12-type structure. Pr6(Fe1−xAlx)14, 0.16 ≤ x ≤ 0.36 with a homogeneous region parallel to the Fe–Al binary, was found to be isotypic with the La6Co11Ga3-type of structure. Pr-rich alloys are liquid at 800 °C, and all the alloys Pr2(Fe1−xAlx)17 with aluminum concentrations less than 5 at.% Al (x ∼ 0.07) enter a two-phase equilibrium with the Pr-rich liquid. At temperatures below 800 °C, alloys with compositions close to 30 at.% Pr and 5 at.% Al show a further ternary phase on solidification, whose crystal structure is related to the La6Co11Ga3-type. PrFe2Al8 is a new representative of the CeFe2Al8-type structure. The crystal structure of the ternary compound richest in Al, PrFe2Al10, has not been solved yet.

Solid state reactions between Ni and Sn at two compositions, Ni75Sn25 and Ni60Sn40, have been induced by means of near room temperature cold rolling and mechanical alloying. The reaction steps have been monitored by x-ray diffraction and differential scanning calorimetry. At both compositions, the first effect of plastic deformation is the formation of two metastable phases which, by further milling or low temperature thermal treatment, transform into the Ni3Sn4 compound. The chemical composition of the metastable phases has been determined to be close to that of Ni3Sn4 and the crystal structure of one of them appears to be related to that of β–Sn. Differential scanning calorimetry and thermal treatment of samples containing the metastable phases have shown that these phases transform into Ni3Sn4 at about 150 °C and that no other reaction takes place up to this temperature. Upon prolonged milling, a different behavior has been observed for the two compositions. While the Ni60Sn40 mixture eventually forms the Ni3Sn2 compound in agreement with previous results, the final product of mechanically alloying the Ni75Sn25 mixture is a phase whose structure, rather than amorphous as previously hypothesized, in our case can be described as based on that of the disordered high temperature form of the Ni3Sn compound. Differential scanning calorimetry and x-ray diffraction analysis of this sample have shown the formation, at 380 °C, of ordered Ni3Sn with an associated heat release of about 10 kJ/mole.

Continuous-fiber reinforced metal-matrix composites consisting of Ni3Al alloys and Saphikon Al2O3 single crystal fibers were fabricated by hot-pressing of fiber-foil lay-ups. Two matrix compositions were employed, namely, IC50 (Ni–22.5Al–0.5Zr–0.1B, at. %) and IC396M (Ni–15.9Al–8.0Cr–0.5Zr–1.7Mo–0.02B, at. %). Etching of the foils in aqueous FeCl3 solution prior to lay-up and hot-pressing tended to improve fiber-matrix bonding and the density-normalized room temperature yield stress. Whereas strength improvements for the IC50 matrix were only moderate, significant improvements were found for an IC396M composite reinforced with 10 vol. % of Saphikon fibers.

A simple analytical model for the effect of mechanical strength and line width (for the case of narrow lines) on the electromigration failure of metallic interconnects is presented. Because the line width/grain size ratio and the diffusivity enter differently in the model, application of the resulting failure time equation to published data can provide insight into the mechanisms of enhancement of electromigration resistance by grain structure optimization and alloying.

Tungsten disilicide (WSi2) was formed by annealing tungsten (W) films of 330 nm and 750 nm prepared by low pressure chemical vapor deposition (LPCVD) from tungsten hexafluoride (WF6) on Czochralski 〈100〉-Si substrates. The silicide was found to grow continuously from the WSi2/W interface. The thickness of the formed WSi2 was observed by Rutherford backscattering measurements (RBS) to increase parabolically with the annealing time, with an activation energy of 2.6 eV/atom. The crystal structure of the formed WSi2 and the unreacted W films was analyzed using x-ray diffraction (XRD) technique. The thermal history of the samples was found to play an important role for the film texture of the unreacted W and formed WSi2, indicating that the fast and inexpensive method, XRD, applied as a thickness monitor for kinetics studies of WSi2 growth will undoubtedly introduce large errors. The as-deposited W (on Si) and the unreacted W (on WSi2) were found to be under a tensile stress, as observed by means of the XRD technique.

Cu films with a thickness around 3.5 μm have been deposited on rough Al2O3 ceramic substrates by the partially ionized beam deposition technique. While the ion bombardment parameters are similar for all depositions, the substrate temperature during deposition is varied from 50 °C to 300 °C. Physical structure of the films is studied by SEM in cross-sectional and surface geometries. X-ray diffraction 2θ scan is performed to obtain information on the microstructure of the films, such as (111) fiber texture and average crystallite size. Resistivity of the films is also measured. It has been found that the physical structure of the films varies from a typical columnar structure at 50 °C to a completely noncolumnar structure at 300 °C. The XRD results show that the films are polycrystalline and have different degrees of 〈111〉 preferred orientation, depending on the substrate temperature. The average crystallite size increases with the increase of substrate temperature. No correlation between the physical structure and microstructure of the films is observed. The resistivity of the films was also seen to change as a function of the substrate temperature. This can be explained by quantitative models in which grain and columnar boundaries in the films are responsible for the decrease in electric conductivity of the films.

This paper presents a geometrical description of aperiodic grain boundaries within the framework of a quasiperiodic lattice. Experimental evidence is given in support of a structural unit model of quasiperiodic [100]45°twist and [100]45°twist plus tilt grain boundaries in aluminum.

Stress relaxation in aluminum films of several thicknesses was characterized by using both continuous indentation and x-ray diffraction techniques. Results of the indentation and x-ray stress measurements compare closely for films of small thicknesses. Indentation data from thicker films do not compare well to the x-ray data due to the presence of a residual stress distribution.

Green's function for a generalized vector Hilbert problem is calculated which can be used to solve the Hilbert problem with any integrable inhomogeneity. The Green's function is obtained by using a complex transform defined by eigenfunctions of the homogeneous Hilbert problem. This method should be particularly convenient for the stress analysis of anisotropic composite materials containing cracks and free surfaces. The method is illustrated by applying it to calculate the stress field in an anisotropic bimaterial composite containing an interfacial crack. It is found that, in agreement with the earlier published work on isotropic composites, the stress field is oscillatory but, except very near the crack tip, the oscillations are negligible. Numerical results are presented for a stress field in a uniformly loaded (shear and compressive) Cu/Ni layered composite containing an interfacial crack as well as cracked homogeneous copper.

The yield strength and interfacial bonding are properties of interest for understanding void formation in thin film interconnect and subsequent failure of VLSI devices. A method is presented to examine the mechanical properties of thin polycrystalline films attached to substrates by measuring the change in thermal residual stress, due to the difference in coefficient of expansion between the film and substrate, as a function of decreasing temperature of the sample. The yield strengths of passivated 0.5, 1.0, and 2.0 μm thin films of Al–2% Cu on oxidized Si wafer substrates have been determined with this method to be 325, 170, and 120 MPa, respectively. Unpassivated films of the same thicknesses were also examined, but yielding did not occur for these films even though the residual stress reached a value of over 400 MPa. The lack of yielding in the unpassivated samples and the thickness dependence of the passivated samples is attributed to the grain size of these materials, which is less than the film thickness for the unpassivated case and greater than the film thickness after passivation. Debonding occurred in the 2 μm unpassivated sample but in none of the others, indicating a thickness dependence of the energy for delamination.

The fracture toughness for first matrix cracking of a uniaxially reinforced C-fiber/C-matrix composite is investigated using a modified controlled surface flaw method. The theoretical models for first matrix cracking of brittle matrix composites including the stress intensity and the potential energy approaches are reviewed in the light of the experimental results. The sharing of the applied load between the reinforcing fibers and the brittle matrix along with extensive crack front debonding enhance the fracture toughness for first matrix cracking.

The cohesive properties and electronic structures versus the structural stability of transition-metal trialuminides YAl3, ZrAl3, and NbAl3 in their cubic L12, tetragonal DO22, and naturally stable forms (i.e., the DO19 structure for YAl3 and the DO23 structure for ZrAl3) have been investigated using a total energy local-density approach. The variation of structural stability with transition-metal constituent can be simply understood in terms of the bandfilling of the bonding states in the rigid band sense, with the valence electrons gradually filling the bonding states on going from YAl3, ZrAl3 to NbAl3. This leads to a phase transition from the cubic L12 structure (for YAl3) to the tetragonal DO22 structure (for NbAl3). It is argued that this criterion may also apply to explain the variation of the structural stability of other transition-metal compounds (such as transition-metal carbides, nitrides, silicides, etc.) that are dominated by covalent interactions between the transition-metal d and the metalloid p states.