Volume 94, Issue 4

The sub-Ångström probe of an aberration-corrected scanning transmission electron microscope will enable imaging and analysis of nanostructures and interfaces with unprecedented resolution and sensitivity. In conjunction with first-principles theory, new insights are anticipated into the atomistic processes of growth and the subtle link between structure and functionality. We present initial results from the aberration-corrected microscopes at Oak Ridge National Laboratory that indicate the kinds of studies that will become feasible in the near future. Examples include (1) the three-dimensional location and identification of individual dopant and impurity atoms in semiconductor interfaces, and their effect on local electronic structure; (2) the accurate reconstruction of surface atomic and electronic structure on nanocrystals, and the effect on optical properties; and (3) the ability to distinguish which configurations of catalyst atoms are active, and why.

Lattice mismatched heterointerfaces are classified by a simple five parameter configuration space which allows to quantify the following properties: gradual partial coherence, distribution and localisation of misfit dislocations, anisotropy of strain fields, elastic dissimilarity of the lattices. HREM images are digitally processed into one-dimensional strain and Fourier spectrum profiles along the interface at selected distance from the interface. The interpretation of these profiles as a Fourier expansion of displacement waves is justified through a link to continuum modelling approaches presented earlier. Limitations and microscope conditions for this simple direct image interpretation approach are listed and discussed.

A numerical model for grain growth has been developed. The governing equations for the model are derived from a variational principle that enforces the balance between energy dissipation and changes of internal energy. This approach significantly simplifies previous modelling of grain growth by replacing a curved grain boundary with several straight lines and relaxing constraints on triple point angles. Such simplifications provide a basis for a finite element method by reducing the analysis of a complex network of interconnected curved lines to the motion of interacting points. In the finite element formulation two approaches are used, namely an element based formulation and a node based formulation. In the latter case, the properties of the elements are lumped together at their common nodes. Such an approach allows the nodal velocity to be determined independently of the forces on other nodes in a lumped mobility methodology, leading to a further reduction of computer time. Several planar grain networks of Voronoi polygons have been studied. The evolution of the grain structures determined from these two approaches is found to exhibit the same trends as those obtained using other models published in the literature. Owing to its simplicity, the node based formulation is recommended.

Polycrystalline alumina substrates were dip-coated in dilute suspensions formed with dispersed, nano-silica particles. The fractional surface coverage of the alumina substrate was varied between ≈ 0.05 to ≈ 0.4 by changing concentration of particles in the silica slurry. After a heat treatment to partially sinter the particles to the surface, the surface was made hydrophobic by a reaction with a solution containing fluorosilane molecules. Wetting measurements showed that the contact angle between the surface and water droplets increased with decreasing area coverage of nano-silica spheres, consistent with a previous theory, modified here to include the pressure of the water droplet. The modification of the theory predicts that the super-hydrophobic effect disappears when the particles become too widely spaced, causing the pressure of the water droplet to spontaneously wet the substrate.

Available experimental thermodynamic and phase diagram data were critically reviewed. A thermodynamic assessment of the binary In-Se system was performed based on the evaluation of literature. The intermediate compounds taken into account are In 4 Se 3 , InSe, In 6 Se 7 , In 9 Se 11 , In 5 Se 7 , and polymorphic In 2 Se 3 (α, β, γ and δ). All these compounds were treated as stoichiometric phases. The liquid phase was described using the association model with the associate species of ‘In 2 Se 3 ’. A set of thermodynamic parameters was obtained, and the calculated phase diagram is presented.

In monotectic systems, there are thermodynamic restrictions on compositions and temperatures below the monotectic where the equilibrium solid cannot form directly from the undercooled homogeneous liquid. In this region, the solid can form only after liquid – liquid phase separation has occurred. As suggested by earlier research, the thermodynamic restrictions on the solidification process may be useful to control the microstructure of rapidly solidified materials. Thus, it could be commercially important to understand the thermodynamic limitations on the formation of the solid in monotectic systems. The liquid miscibility gaps, metastable liquidi, and spinodal curves in the binary Zn– Bi, Zn–Pb, and Zn–Tl systems were assessed by using a subregular solution model to describe the Gibbs energies of the liquid phases. The present results were compared and discussed with other experimental and assessed data available in the literature.

The liquidus surface of the Al –Fe –Ti system has been reexamined. 34 different alloys have been studied by differential thermal analysis, and the temperatures of invariant reactions have been determined. By examining the as-cast samples by optical microscopy, X-ray diffraction and energy-dispersive spectrometry in a scanning electron microscope, the crystal structures and chemical compositions of the primary phases and two-phase aggregates have been established. In some cases, additional investigations were carried out in the transmission electron microscope. From the combined results of these investigations the solidification paths of the alloys were established leading to a revised representation of the liquidus surface of the Al –Fe –Ti system. A reaction scheme has been set up which shows phase equilibria that are consistent with previous solid-state isothermal sections.

Taking into account new experimental measurements, the Fe –Zn – Cr ternary system is critically modified at 460 °C. A continuous solid solution between ζ-FeZn 13 and CrZn 13 compounds is shown but is shared at 460 °C by the stable CrZn 17 compound containing about 2 wt.% Fe. This ternary system is assessed with the CALPHAD method using the PARROT modulus of the Thermo-Calc software. The liquid and solid solution phases are modeled with Redlich– Kister –Muggianu equations. The intermetallic compounds ζ-(Fe,Cr)Zn 13 and CrZn 17 are treated as stoichiometric compounds in the binary systems. The experimental Fe and Cr solubilities at various temperatures modify the shape of the liquidus curve and are satisfying for industrial applications. A set of parameters consistent with most of the available experimental data on both phase diagram and solubility measurements is obtained by optimization. A comparison with previous experimental work is also presented and a reactional model between iron substrate and Zn–Cr bath is proposed. This optimization allows to interpret the growth of intermetallic layers and the formation of dross when galvanizing in Cr-added Zn bath.

The diffusion coefficients ( D ) of implanted radioisotopes – 71 Ge, 31 Si, and 11 C – in the polymer-derived amorphous (a-)Si 29 B 9 C 41 N 21 ceramic have been measured as functions of temperature ( T ) in different as-pyrolysed states ( 71 Ge) or after pre-diffusion annealing at 1600 °C for 2 h ( 71 Ge, 31 Si, and 11 C) by means of radiotracer techniques, in which serial sectioning was done by ion-beam sputtering. In the cases of 31 Si (half-life t 1/2 = 2.6 h) and 11 C ( t 1/2 = 20.4 min), implantation, diffusion annealing, and sputter-sectioning were done on beamline and in situ of a novel set-up specially constructed for diffusion studies of short-lived radiotracer atoms. In all cases, the T -dependencies of D are of Arrhenius type. Comparing the Arrhenius parameters of D ( 71 Ge) to previous data on diffusion in (B-free) a-Si 28 C 36 N 36 , it is concluded that the diffusion of 71 Ge in the a-Si 29 B 9 C 41 N 21 ceramic is controlled by vacancies in its a-Si 3 N 4 – y C y phase which become increasingly smeared out when the C-content y decreases as a result of pre-annealing. The coincidence of D ( 31 Si) and D ( 11 C) in pre-annealed a-Si 29 B 9 C 41 N 21 at all temperatures investigated is enforced by vacancy-mediated diffusion through SiC crystallites embedded in the a-Si 3 N 4 – y C y phase.

Neutron diffraction measurements were used to obtain a better understanding of the ageing behaviour in dual-phase ferrite + martensite steels. Therefore, the influence of different ageing treatments on the tempering of low-alloyed 0.4 and 0.7% C FeMnCrMo-martensite and 0.1% C FeMnCrMo dual-phase steels was investigated. On tempering at 170 and 350 °C, the sequential precipitation of g- and h-carbides, as well as the loss of tetragonality was observed in the martensite. The decrease in austenite lattice parameter after tempering is due to stress relaxation in the material. These observations offer an acceptable explanation for the large yield stress increase after ageing at 170 °C.

Dislocation interactions in the ferrite matrix of a dual-phase steel have been studied as a function of martensite volume fraction (11 – 25%) and tensile straining (0 –5 %) by strain-induced amplitude-dependent (strain amplitude varying from 10 – 7 to 10 –4 ) internal friction measurements at room temperature, using a high-frequency composite oscillator (40 kHz). Results showed that a higher volume fraction of martensite increases the internal stress in the softer ferrite matrix and restricted the dislocation motions, giving rise to a lower amplitude-dependent internal friction. Application of tensile straining further shortened the effective mobile dislocation length which was revealed by the internal friction measurements and this resulted in a high initial strain hardening rate of the high-martensite samples. Reduction in the martensite volume fraction and the mobile dislocations in the ferrite matrix through process control resulted in an increase in the amplitude-dependent internal friction, indicating a longer effective mobile length of dislocations which, in turn, delayed early necking of the dual-phase samples during tensile straining.

The morphology of discontinuous precipitation of a liquid has been studied in Cu-9.5 at.% In and Cu-7.5 at.% Sn alloys. The transformation of a supersaturated solid solution (α o ) was carried out in the two-phase (α + liquid) field. A lamellar structure, consisting of alternate solute-depleted (α) and liquid phases, has been observed to grow from grain boundaries of the supersaturated α o . The nucleation of the liquid droplets occurred at grain boundaries as two hemispherical caps abutted at the original grain boundary positions. A thin liquid film replaced the grain boundary and started migrating as the transformation progressed. The lamellae multiplication occurred by nucleation of new lamellae at the migrating front. The lamellae multiplication also occurred by nucleation of new lamellae at mid-lines between the liquid lamellae where the interlamellar spacing was large. A cylindrical loop of liquid formed behind a migrating liquid film, resulting into double-seam morphology and an irregular lamellar structure on one side of the original grain boundary. The liquid film also migrated in the shape of sine wave when the liquid film between neighbouring droplets migrated in opposite directions near the original grain boundary. This has resulted in double-seam morphology with a regular lamellar structure.

The order – disorder transitions in AuCu alloys have been studied by differential scanning calorimetry and resistometric methods. We have collected data with different sets of samples of equiatomic composition, the structure of which was determined by X-ray diffraction. Grain growth occurs mostly below 300 °C either in rolled and nanocrystalline samples. The occurrence of ordering does not depend on the grain size, but ordering is favoured by high defect contents in the alloy. The shape of heat flow traces provide evidence that ordering and disordering transitions are diffusive at low temperature (below 385 °C) being exothermic and endothermic, respectively, in wide temperature ranges. The transition from AuCuI to AuCuII changes the mechanism above 385 °C. The transition from AuCuII to face-centered cubic (fcc) solid solution is of first order. The reverse transition from the fcc structure to AuCuI occurs by nucleation and growth on continuous cooling.

The growth kinetics of intermetallic compound layers formed between Sn–3Ag–8Bi – 5In solder and (Cu, electroless Ni – P/Cu) substrate were investigated at temperatures between 70 and 170 °C for 1 to 60 days. The layer growth of the intermetallic compounds in the Sn–Ag– Bi – In/Cu and Sn–Ag –Bi – In/electroless Ni –P systems satisfied the parabolic time law. As a whole, because the values of time exponent have approximately 0.5, the layer growth is mainly controlled by diffusion. The apparent activation energies for the growth of the Cu 6 Sn 5 and Ni 3 Sn 4 intermetallic compounds are 81.4 and 67.1 kJ/mol, respectively.

An experimental investigation was conducted to examine the wear behaviour of cross-ply carbon– carbon – silicon carbide composites (C/CαSiC) under dry sliding against a chromium steel disc at various speeds and loads. Results indicated that under identical conditions, C/C– SiC composites with weak fibre resin interfaces at all times had a larger weight loss than those with a strong bond. In the first few minutes of wear, weight losses were approximately the same for both types of composites. Also, the wear rates of both the composites decreased sharply with sliding distance up to 7 km, and then gradually tended to become constant. Scanning electron microscopy showed that the graphite film was transferred on the test pin side, which stabilizes the increase in test pin temperature after the formation of a thick film of graphite debris.