Our recent positron lifetime measurements for metal oxides suggest that positron lifetimes of bulk state in metal oxides are shorter than previously reported values. We have performed theoretical calculations of positron lifetimes for bulk and vacancy states in MgO and ZnO using first-principles electronic structure calculations and discuss the validity of positron lifetime calculations for insulators. By comparing the calculated positron lifetimes to the experimental values, it was found that the semiconductor model well reproduces the experimental positron lifetime. The longer positron lifetime previously reported can be considered to arise from not only the bulk but also from the vacancy induced by impurities. In the case of cation vacancy, the calculated positron lifetime based on semiconductor model is shorter than the experimental value, which suggests that the inward relaxation occurs around the cation vacancy trapping the positron.

A universal relation between electron density minima and ionic radii, was discovered from the first principles calculations of electronic structures in ceramics over 60 species, including oxides, borides, carbides, nitride and fluorides. Every ceramic falls on a curve of log(ρ_minZ⁻³) vs. 2(Z/n)r_min, where ρ_min is the minimum electron density in the line linking the first-nearest-neighbor nuclei, r_min is the distance r at ρ_min, Z is the atomic number, and n is the principal quantum number.

Atomic structures characterization of Al₂O₃(0001)/Cu nano-hetero interfaces has been performed by the first-principles pseudopotential method and in cooperation with HRTEM observations. The physical properties of the interfaces depend strongly on the interface stoichiometry. Bonding nature of the O-rich (O-terminated) interface is explained as strong covalent and ionic interactions, whereas that of the stoichiometric (Al-terminated) interface is weak covalent and electrostatic image interactions. The O-terminated interface has quite larger adhesive energy than that of the stoichiometric one. Recent HRTEM observations of the Al₂O₃(0001)/Cu interface have confirmed the O-terminated interface. However, the observed incoherent interface is not the same as an ideal coherent interface obtained by the first-principles. We explain the relationship between the present coherent interface and the practical incoherent one.

We extend our newly proposed calculation method of precipitate nucleation free energy to ternary systems. This method utilized first principles calculations for enthalpy change and interface energy, and the Bragg-Williams approximation for entropy loss from scattered atoms condensing into a cluster. The effect of Ni addition on copper precipitation in the Fe-Cu system was examined by this method. It was revealed that added Ni prefers segregation at the matrix/cluster interface, and reduces the activation energy barrier as well as the interface energy.

First principles calculations of hydrated polymolybdates complexes have been made with an atomic orbital basis molecular orbital method. Hydration effects are taken into account by the conductor-like screening model (COSMO) using dielectric constant of water. Hydrated heptapolymolybdate, Mo₇O₂₄⁶⁻, shows a low symmetry structure, which agrees well to experimental results, i.e., X-ray diffraction of crystalline salts and X-ray absorption fine structure of the hydrated complex. Contrary to that, the hydrated heteropolymolybdate, NiMo₆O₂₄¹⁰⁻ prefer to exhibit the high symmetry structure. Inspection of the electronic states found that the Ni ion exhibits trivalent state or d ⁷ configuration in a formal sense. Jahn-Teller distortion around Ni is therefore evident. Such distortion cannot be found in CrMo₆O₂₄⁹⁻ or CoMo₆O₂₄⁹⁻.

First principles calculations of rutile-type TiO₂:S have been performed to investigate the effect of sulfur solutes on the electronic structure. Plane-wave pseudopotentials method has been employed and atomic relaxations were fully taken into account. All possible geometric configurations for sulfur solutes within a 12-atoms supercell have been examined changing sulfur concentration of x = 0, 0.25, 0.5, 0.75 and 1. Theoretical direct band gap is found to decrease as sulfur concentration is increased. The dependence on the sulfur concentration is weaker than that was predicted in literature. Both the optimization of solute configuration and atomic relaxation are found to be essential for quantitative evaluation of the electronic structures in the alloy.

First principles calculations have been carried out to investigate the core-hole effects on the theoretical fine structures of the X-ray absorption spectra of MgF₂ and ZnF₂ at F K-edge. Significant differences are found between the calculated spectral fine structures with and without core-holes. Experimental profiles of the near-edge X-ray absorption fine structures are well reproduced by the theoretical ones when the core-hole effect is introduced. The dependence of supercell size on the theoretical fine structures is also examined.

We measured the X-ray fluorescence holograms of Si_0.999Ge_0.001, Si_0.8Ge_0.2, Si_0.5Ge_0.5, Si_0.2Ge_0.8 and Ge single crystals, and reconstructed the images of second neighbor atoms around Ge. The positional shift of the atomic image across the whole composition range was three times larger than the value predicted from the difference in the lattice constants of pure Si and Ge. We found that imaginary part of the reconstruction strongly affects the positions of the atomic images. Thus, using the negative real parts, the atomic image became sharp and its shift dependent upon Ge composition comes to the reasonable values.

The nuclear-electric-quadrupole interactions at ¹¹¹Cd and ¹¹⁷In nuclei arising from ^111mCd and ¹¹⁷Cd, respectively, chemically introduced in pyrochlore ferroelectric Cd₂Nb₂O₇ (T_C = 196 K) were studied using the perturbed-angular-correlation technique. At temperatures above T_C, there is one type of Cd sites. However, at liquid nitrogen temperature, below T_C, there are two types of Cd sites. The ratio of the electric quadrupole frequency of ¹¹⁷In to that of ¹¹¹Cd is anomalously deviated from a value expected from the purely ionic In and Cd ions in the same lattice environment.

The photoluminescence (PL) around the wavelength of 1.54 μm from the Er-containing ZnO specimens was measured at room temperature by the indirect excitation of the He-Cd laser (325 nm). The PL intensity varied greatly with the Er concentration and the specimen preparation conditions as well. The specimens with the Er content of about 2.6 at%, sintered at 1623 K, and cooled quickly to room temperature in air showed the highest PL intensity at 1.534 μm. The local structure of the optically active Er centers in ZnO was also discussed, and the appropriate optically active center might be the Er ions existing in the grain boundaries.

Hf/Fe multilayers are soft magnetic materials and are a good candidate for magnetic heads of storage devices. We have investigated Hf/Fe multilayers using time-differential perturbed angular correlation (TDPAC) spectroscopy of ¹⁸¹Ta (← ¹⁸¹Hf), examining specifically the magnetic nature of the non-magnetic Hf layers sandwiched between two ferromagnetic Fe layers. We prepared multilayer samples with Hf layers of various thicknesses, [Hf(x nm)/Fe]_n (x=0.5, 1.0, 2.0, 4.8, 10.0). Three different hyperfine magnetic fields 53(1) T, 43(1) T and 2(1) T were detected in the samples of [Hf(2 nm, 4.8 nm, 10 nm)/Fe]_n. The hyperfine magnetic field of 53(1) T corresponds to Hf atoms in a turbulent interface between the Hf and Fe layers. The value of 43(1) T corresponds to Hf atoms inside Hf layers within 0.5 nm from the interface. The value of 2(1) T corresponds to Hf atoms inside the Hf layers within 3.5 nm of the interface.

Three kinds of FePt-MgO granular films were prepared by a vacuum successive deposition of MgO, Pt, Fe and MgO on a cleaved surface of sodium chloride below 673 K. Their microstructures, electronic structures and magnetic properties were studied by high-resolution transmission electron microscopy (HRTEM), electron energy-loss spectroscopy (EELS) and measurement with a superconducting quantum interference device (SQUID) magnetometer. The TEM observations and selected area electron diffraction patterns revealed that the samples mainly consist of few nm-sized FePt clusters embedded in MgO films with L1₀-ordered structure and c-axis perpendicular to the film surface. Size effect on the stability of L1₀ phase in the FePt nano-clusters was directly observed in [MgO/Fe(0.38 nm)/Pt(0.30 nm)/MgO] and the critical size of the transition from L1₀ to A1 phase was estimated as around 2 nm, that can be considered as smaller than effective size for the transition from ferromagnetism to superparamagnetism. Coercivity of [MgO/Fe(1.0 nm)/Pt(0.8 nm)/MgO] was 1.2 × 10⁵ A/m. The Fe-L_2,3 white-line ratios of the present samples measured by EELS were about 4.0, independently on the incident direction of electron beam. The higher white-line ratio may be attributed to their high-spin state by a change of 3d-band structure owing to the hybridization of d-bands between Fe and Pt atoms.

In-situ optical reflection measurements have been performed for a Si(100) surface under H⁺ irradiation to study dynamic change in the electronic structure of the silicon. The relative reflectance at 370 nm decreased almost linearly with displacement per atom (dpa) and was recovered by hydrogen release by 423 K annealing, suggesting that the change in the relative reflectance at 370 nm originates from the increase of hydrogen bonded to Si. With increasing H⁺ irradiation, the relative reflectance around 700 nm increased due to the formation of Si-H phase near the surface silicon. Furthermore, the broad minima of the relative reflectance shifted from 600 to 500 nm by H⁺ irradiation, probably correlating with amorphization or appearance of silicon micro/nano-crystal by displacement effect.

First principles calculations of L₃ XANES/ELNES of GaN and InN with both wurtzite and zinc-blende structures have been made using OLCAO (orthogonalized linear combinations of atomic orbitals) method. Supercells with more than 100 atoms were employed. A core-hole was rigorously included in the calculation, and the photo absorption cross section (PACS) between the initial and final states was computed. Quantitative reproduction of experimental spectrum that is available in literature can be found when the PACS was computed. Although spectral shapes of two phases look similar, characteristic differences are predicted to appear at the first peak of the L₃ XANES/ELNES. The first peak is notably broader in the zinc-blende phases. The origin of the broadness is analyzed using partial density of unoccupied states (PDOS) and Mulliken charge. We then conclude that the broadness can be related to greater covalency of the zinc-blende phase as compared to the wurtzite phase.

Crystal field analysis of the energy level structure of lithium yttrium fluoride doped with trivalent dysprosium was performed. Assignment of absorption lines in experimental spectrum of absorption was made in a wide spectral region: from IR to UV. Crystal field parameters for the whole series of trivalent rare earth ions in LiYF₄ crystal were obtained. The values of the intermediately coupled Lande g-factor were calculated and used to estimate the EPR g-factors g_||, g_⊥ for LiYF₄:Dy³⁺.

High-resolution X-ray absorption spectra of Al, AlN and Al₂O₃ are measured at the Al K-edge, which have revealed a chemical shift in the threshold energy and significant differences in the spectral fine structures among the three compounds. In order to interpret the chemical shift and the fine structures of these spectra, first-principles calculations using the full-potential linearized augmented plane wave method within the density functional theory are carried out, taking into account the core-hole effect. The resultant theoretical spectra quantitatively reproduce both the chemical shift and the spectral fine structures of the experimental ones. The dependence of the theoretical spectra on the supercell size is also examined.

The characterization of Ni-Zn/TiO₂ nanocomposite synthesized by the liquid-phase selective-deposition method was studied. Nanoparticles were well dispersed and stabilized by the selective deposition onto TiO₂ surface. The particle size was decreased with increasing the amount of Zn added, thus the catalytically active Ni surface area was increased. The selective deposition onto TiO₂ surface and addition of Zn to the nanoparticles promoted the catalytic activity of Ni-Zn nanoparticles, e.g. the catalytic activity of Ni-Zn/TiO₂ for 1-octene hydrogenation was ca. 10 times higher than that of the unsupported Ni nanoparticles. Ni in the nanocomposite was assigned as metallic, although their surface was oxidized under the atmospheric condition, but Zn and B were deposited as their oxide.

Silicon nitride thin film was fabricated by pulsed laser deposition using KrF excimer laser and a silicon nitride compact as a target. The deposition was carried out on Al₂O₃ (0001) at 1173 K in N₂ gas pressure of 0.27 Pa. The X-ray diffraction did not provide any structural information of the deposited thin films except that it is composed of amorphous and/or micro-crystalline structure. X-ray absorption near edge structures at Si-K edge revealed that local arrangement of Si is not random. It should be composed of SiN₄ unit similar to the case of α-Si₃N₄ crystal. Metallic Si component cannot be found in XANES.

Ionic conductivities of yttria-stabilized zirconia (YSZ) single crystals deformed at high-temperature were measured by the AC impedance method. A correlation between ionic conductivity and dislocation structures of deformed YSZ single crystals were investigated. Electrical conductivity measurements of the deformed YSZ crystals were performed for two different current directions of [110] and [111]. The [110] direction is parallel to edge dislocations introduced by the primary slip system, while the [111] direction is normal to the edge dislocation lines. Transmission electron microscopy observations showed that the dislocations due to the primary slip system of (001)[110] were mainly generated at the strain of around 1% strain, while the secondary slip systems, such as (111)[101] and (111)[011], were also activated at about 10% strain. It was found that the deformed samples with larger strains exhibited higher electrical conductivities irrespective of the measured current directions. However, the electrical conductivity along [111] was higher than that along [110], suggesting that mobility of oxygen ion is sensitive to the dislocation structures. From the activation energy for oxygen diffusion in the deformed samples, it was found that the oxygen migration enthalpy for deformed samples became smaller than that for undeformed samples, whereas the association enthalpy for the deformed samples became larger. The increase in the association enthalpy might be due to interaction between oxygen vacancies and dislocations. It is thus considered that oxygen vacancies concentrate around dislocations, and are able to move very quickly along the dislocation lines.

Low-temperature structures of the lithium manganese spinels were determined using TOF neutron Rietveld analysis. The spinels LiMn₂O_4-δ with different δ values, (0.016, 0.040, and 0.132) showed the cubic-orthorhombic phase transitions, and the lattice distortion decreased with decreasing δ. Although no anomaly corresponding to the cubic-orthorhombic phase transition was observed in the DSC curve for the spinel with δ ∼ 0.016, the transition was observed by the structure analysis, which is consistent with the broad Cp anomaly at 250 K. The cubic-orthorhombic phase transition is closely correlated to the existence of the vacancy. The charge disproportionation into trivalent and tetravalent state proceeds gradually with decreasing temperature, and the extent of the disproportionation is dependent on the vacancy. Based on the structure analysis, the phase transitions in the spinel are discussed.

The local electronic structures are simulated for a perovskite-type oxide with four polymorphous phases, SrZrO₃, using the DV-Xα molecular orbital method. It is found that a series of phase transitions occurs at certain temperatures so as to retain not only the Zr-O bond strength, but also the Sr-O bond strength, by the tilting of ZrO₆ octahedra and the attendant accommodation to the Zr-O and the Sr-O interionic distances. The occurrence of such smart phase transitions is observed in other perovskite-type oxides, SrRuO₃ and SrHfO₃. Another type of phase transition occurs by the cooperative ionic displacements along the [100] direction in BaTiO₃. The differences in the local chemical bond are discussed between the two types of phase transitions.

A series of silica-supported cerium oxide samples having various cerium content were prepared by conventional impregnation method, followed by calcination in air. Ce L_III-edge XANES study revealed that highly dispersed cerium oxide species on silica surface were predominantly in trivalent oxidation state and cerium oxide nano-particles on silica contain mainly tetravalent cation like as CeO₂ bulk oxide. This elucidated that UV absorption band at 265 nm is assignable to Ce(III) species dispersed on silica.

Titanium oxide films were formed using a titanium target by a pulsed laser deposition (PLD) technique under different oxygen pressures from 10⁻⁶ to 100 Pa. Their densities and thickness were evaluated from total external X-ray reflection profiles and their atomic structures were determined by grazing incidence X-ray scattering (GIXS), and their surface morphology was observed by atomic force microscopy (AFM). The atomic structures of the films were gradually changed from metal titanium through TiO to Rutile-type titanium dioxide TiO₂. The film surfaces became rough above 13.3 Pa oxygen pressure during deposition. Their UV transmission spectroscopy was also observed to predict their photocatalytic activities.

The effect of second phase dispersion on high-strain-rate superplasticity was examined in tetragonal ZrO₂ dispersed with 30 vol% MgAl₂O₄ spinel. The spinel particle enhances the diffusivity of ZrO₂ by supplying small amounts of aluminum and magnesium into ZrO₂ and suppresses grain growth by grain boundary pinning. After superplastic flow, the spinel particles highly elongate along the tensile direction. In the spinel particles, intragranular dislocations were observed, indicating that the spinel particles may contribute to the relaxation of stress concentrations around grain junctions exerted by grain boundary sliding. Comparison with earlier studies suggests that the dispersion of spinel particles can attain high-strain-rate superplasticity in tetragonal ZrO₂ through providing the following positive factors simultaneously; (i) the suppressed grain growth, enhanced accommodation process due to the accelerated (ii) diffusivity and (iii) stress relaxation.

High-temperature creep resistance in cations co-doped polycrystalline Al₂O₃ was examined by uniaxial compression creep test at 1250°C. The dopant oxides used in this study were 0.1 mol% of YO_1.5, ZrO₂, SrO, MgO and TiO₂. The creep rate in Al₂O₃ was significantly changed by cations co-doping. For instance, Zr/Y co-doping suppressed the creep rate in Al₂O₃ by a factor of about 400. A high-resolution transmission electron microscopy (HREM) and nano-probe energy dispersive X-ray spectroscopy (EDS) analysis revealed that Y and Zr cations segregate along grain boundaries. The grain boundary diffusion in Al₂O₃ was supposed to be retarded by the segregation of Y and Zr cations. A first-principle molecular orbital calculation was made for cations co-doped Al₂O₃ and cation singly doped Al₂O₃ model cluster. The creep rate was correlated with the value of net charge in oxygen anion. The net charge of oxygen anion was one of the most important factors to determine the creep resistance in Al₂O₃.

The shapes and structures of many randomly selected grain boundaries in BaTiO₃ with 0.1, 0.3 and 0.5 excess mol% of TiO₂ sintered at 1250°C have been examined by TEM. Most of the grain boundaries have flat segments that coincide with low index planes of one of the grain pairs. These low index grain boundaries are likely to be singular. Singular boundary segments appear in a variety of shapes. In the 0.1 mol% Ti specimen, {011} boundary planes appear most frequently, while in the 0.5 mol% Ti specimen, {111} boundary planes are dominant. Some grain boundaries show fine steps with low index terraces indicating that they migrate by the lateral movement of the steps. This step growth mechanism is consistent with the observed slow growth of the matrix grains and abnormal growth of a few grains.

In many ceramic systems thin amorphous films of about 1 nm thickness often cover grain boundaries. These amorphous films play a key role not only in the formation of microstructures but also in the thermal-mechanical properties of ceramic materials. However, such thin amorphous layers could not be probed directly by an analytical electron beam. With the recent advances in spatially-resolved electron energy-loss spectroscopy technique, chemical and physical parameters of the thin films could be successfully derived using the “spectrum separation” approach. Basic characters and behaviors of variations for the inter-granular films are analyzed in a few silicon nitride (Si₃N₄) and silicon carbide (SiC) systems. The combined local chemical-structural information reveal new trends on microstructures and properties, and provides further insights in Si-based ceramic materials.

In the present study, the energy and atomic structure of the ‹110› symmetric tilt boundaries in palladium were evaluated using the molecular dynamics (MD) simulation and the electronic structures of hydrogen in the bulk and on the grain boundaries of palladium were calculated using the discrete-variational Xα (DV-Xα) method. The MD simulation revealed that the energy of the ‹110› symmetric tilt boundary of palladium depended on the misorientation angle and that there were large energy cusps at the misorientation angles which corresponded to the (111)Σ3 and (113)Σ11 symmetric tilt boundaries. The atomic structure of all ‹110› symmetric tilt boundaries could consist of the combination of the (331)Σ19, (111)Σ3 and (113)Σ11 structural units and (110)Σ1 and (001)Σ1 single crystal units. The DV-Xα calculation showed that the interstitial hydrogen atoms in palladium induced the Pd-H chemical bond which had a different energy level than the Pd-Pd bond. The energy level and component of the Pd-H bonding on the grain boundaries in palladium were similar to those in the bulk palladium.

Temperature and strain rate dependence on high temperature elongation to failure in fine-grained ceramics is phenomenologically explained from grain growth behavior during deformation and the superplastic flow behavior. The elongation to failure at temperatures between 1573 and 1773 K was analyzed for 2 mol% TiO₂ and 2 mol% GeO₂ co-doped tetragonal zirconia polycrystal (TZP), which exhibits excellent high temperature ductility. The improvement in the high temperature ductility in TZP is attributed to dopant cation segregation in the vicinity of the grain boundaries. The phenomenological analysis revealed that co-doping of Ti and Ge cations increases the grain size at the time of failure, as a parameter to describe a limit of an accommodation process for superplastic flow. The parameter of the critical grain size at the time of failure correlates well with the value of overlap population in cation-doped TZP model cluster obtained from a first-principle molecular orbital calculation. The covalent bond at the grain boundaries plays a critical role in the high temperature tensile ductility of TZP.

Grain boundary structure and current-voltage (I-V) characteristics were investigated for Σ5(210) symmetrical tilt boundary of a Nb-doped SrTiO₃ bicrystal. The bicrystal was prepared by a hot-joining technique at high temperature in air. The high resolution transmission electron microscopy (HRTEM) study has revealed that the grain boundaries were perfectly joined at an atomic level without intergranular phases such as amorphous or secondary precipitates. In spite of the high coherency, the boundary tends to be faceted with low index planes of (110) and (410) to reduce grain boundary energy. On the other hand, the I-V property across the boundary was found to exhibit a cooling rate dependency from annealing temperature. In addition, the dependency shows a peak against cooling rates. This fact suggests that the non-linearity in I-V relation is dominantly controlled by a non-equilibrium process of a defect reaction.

In this study, the grain boundary structure of [001] symmetric tilt boundary with a tilt angle of 66° was investigated using a rutile-type TiO₂ bicrystal. The tilt angle of this boundary has a misfit angle of 1.7° from an exact Σ13 approaching to Σ17 relation. High-resolution transmission electron microscopy study (HRTEM) has revealed that the grain boundary was free from any secondary phases, and the two single crystals contact each other perfectly at an atomic scale. The boundary shows almost straight feature without any step structures whereas a part of the boundary forms facet structures consisting of low index planes such as {310} and {110}. On the other hand, it was found that the contrasts due to strong strain fields existed on the grain boundary plane with a spacing of 7.6 nm by weak beam dark field observation. Comparing with atomic structural analysis using HRTEM, the strain field results from a distorted Σ13 unit structure, which can be predicted from a rigid body model of Σ13 relation. This distorted unit structure has a similar structure of Σ17 relation. Namely, the boundary consists of a periodical array consisting mainly of Σ13 unit structures and partially Σ17-like unit structures. In other words, a misfit angle in this boundary was accommodated by not introducing secondary dislocations, but a transformation of basal unit structure.

High temperature deformation behavior of [0001] symmetrical tilt grain boundaries of Al₂O₃ was investigated by using bicrystals. Four kinds of the grain boundaries Σ7/{4510}, Σ21/{4510}, Σ7/{2310} and Σ21/{2310} were selected in the present study, and compressive mechanical tests were performed at 1450°C in air to investigate the sliding behavior of the respective boundaries. It was found that, stresses readily increased with increasing strains for all specimens during compression tests, but the bicrystals showed abrupt sliding along the grain boundary planes to fracture. Among the boundaries studied, Σ7/{4510} exhibited the highest resistance to the grain boundary sliding. The other boundaries showed similar sliding resistance, and yet the stress and strain values at their failure were much smaller than those of Σ7/{4510}. In order to understand the mechanism of the sliding behavior of the respective boundaries, the grain boundary core structures and their atomic densities were examined, based on the structure models obtained in our previous studies. It was found that Σ7/{4510} having the highest sliding resistance showed a larger atomic density at the core, as compared to the others. Therefore, the observed sliding resistance of the boundaries is closely related to the detailed atomic structures at the grain boundary cores.

High temperature oxidation and oxidation-induced embrittlement in β-silicon carbides (SiCs) with different grain boundary microstructures have been studied. SiCs with different grain boundary microstructures were fabricated by hot-pressing with different doping elements like Mg, Al, P. Oxidation experiments were carried out under the oxygen partial pressure ranging from 0.303 Pa to 78.5 Pa at temperatures 1623—1773 K for 7.2—36 ks. Thereafter, the degree of oxidation-induced embrittlement was quantitatively evaluated by three-point bend tests at room temperature in connection with grain boundary microstructure. More severe degradation was observed as a result of oxidation though the passive oxidation took place. It is concluded that the oxidation-induced embrittlement in β-SiC can be improved by decreasing the frequency of random boundaries and the grain size. The potential of grain boundary engineering for a ceramic material has been confirmed.

Thin Ni films were deposited on the (111) surface of YSZ at 1073 K by a pulsed laser deposition technique. The interfacial atomic structure of Ni/YSZ was investigated by high-resolution transmission electron microscopy (HRTEM). It was found that Ni was epitaxially oriented to the YSZ surface, and the following orientation relationship (OR) was observed: (111)_Ni || (111)_YSZ, [110]_Ni || [110]_YSZ. Geometrical coherency of the Ni/YSZ system was also evaluated by the coincidence of reciprocal lattice points (CRLP) method. It was found that the most coherent OR predicted by CRLP method was (705)_Ni || (111)_YSZ, [010]_Ni || [110]_YSZ, which was not consistent with the experimentally observed OR. To understand the detailed atomic structure, HRTEM image simulations were performed. However, simulated images based on both O-terminated and Zr-terminated interface models were quite similar to the experimental image, and thus it was hard to determine which model is comparable with the actual interface only by the HRTEM image simulations. In order to clarify the termination layer at the interface, electronic structures of the Ni/YSZ interface were investigated by electron energy-loss spectroscopy. It was found that significant differences were observed in O-K edge spectra between the interface and the YSZ crystal interior, and the spectrum from the interface showed similar features to the reference spectrum of bulk NiO. This indicates that the Ni-O interaction occurs at the interface to terminate the oxygen {111} plane of YSZ at the Ni/YSZ interface. In addition, the density of Ni-O bonds across the interface in the experimental OR was larger than that in the most coherent OR predicted by CRLP method, which also suggests that the on-top Ni-O bonds stabilize the Ni/YSZ(111) interface.

Fine-grained 3Y-TZP has been known to show high superplasticity. Addition of a small amount of metal oxide influences the superplastic behavior in 3Y-TZP. In this study, 3Y-TZP doped with 1 mol% GeO₂, TiO₂ or BaO were fabricated, and respective grain boundary energy has been systematically measured by a thermal grooving technique with atomic force microscopy. It has been found that addition of Ge⁴⁺ or Ti⁴⁺ ions decreases the grain boundary energy to stabilize the grain boundaries in TZP whereas doping of Ba²⁺ ion increases the grain boundary energy to destabilize the grain boundaries. A change in the grain boundary energy should be due to segregation of dopant at grain boundaries. It has been also found that the elongation to failure of cation-doped 3Y-TZP is directly proportional to the stability of grain boundary. Grain boundary energy is thus one of the principal factors to determine the tensile ductility of TZP. In order to reveal the effect of dopant on the grain boundary energy, lattice static calculations and first principles molecular orbital calculations have been performed for supercells and model clusters including the present dopant, respectively. A series of results shows that substitution of Ge⁴⁺ or Ti⁴⁺ ion for Zr⁴⁺ ion increases the covalency of TZP, but the covalency of TZP is reduced by addition of Ba²⁺ ions. The grain boundary energy is found to have a relationship with covalency nearby grain boundaries in TZP.

Of considerable importance to the generation of ultrafine microstructures is the development of high misorientations. The present work examines the effect of the crystallographic rotation field in simple shear upon the evolution of misorientation during plastic working. A series of Taylor simulations are presented and it is shown that the rotation field is such that small differences in orientation in the region of the main torsion texture components are considerably increased with the application of shear strain. This did not occur in simulations of rolling. The torsion simulations compare favourably with the nature of the misorientations evident in hot worked 1050 Al and Ti-IF steel. It is concluded that shear deformation, by its nature, facilitates the generation of higher misorientations.

Shear bands formed during both cold and hot plastic deformation have been linked with several proposed mechanisms for the formation of ultrafine grains. The aim of the present work was to undertake a detailed investigation of the microstructural and crystallographic characteristics of the shear bands formed during hot deformation of a 22Cr-19Ni-3Mo (mass%) austenitic stainless steel and a Fe-30 mass%Ni based austenitic model alloy. These alloys were subjected to deformation in torsion and plane strain compression (PSC), respectively, at temperatures of 900°C and 950°C and strain rates of 0.7 s⁻¹ and 10 s⁻¹, respectively. Transmission electron microscopy and electron backscatter diffraction in conjunction with scanning electron microscopy were employed in the investigation. It has been observed that shear bands already started to form at moderate strains in a matrix of pre-existing microbands and were composed of fine, slightly elongated subgrains (fragments). These bands propagated along a similar macroscopic path and the subgrains, present within their substructure, were rotated relative to the surrounding matrix about axes approximately parallel to the sample radial and transverse directions for deformation in torsion and PSC, respectively. The subgrain boundaries were largely observed to be non-crystallographic, suggesting that the subgrains generally formed via multiple slip processes. Shear bands appeared to form through a co-operative nucleation of originally isolated subgrains that gradually interconnected with the others to form long, thin bands that subsequently thickened via the formation of new subgrains. The observed small dimensions of the subgrains present within shear bands and their large misorientations clearly indicate that these subgrains can serve as potent nucleation sites for the formation of ultrafine grain structures during both subsequent recrystallisation, as observed during the present PSC experiments, and phase transformation.

In this study, the effect of material properties on deformation pattern and strain distribution of the commercially-available pure titanium (CP-Ti) specimen during an equal channel angular extrusion (ECAE) was investigated. Finite element analyses were carried out for two-dimensional plane strain condition at elevated temperatures. Material properties were assumed by hardening, softening, and no hardening behaviors based on the measured experiment data. The effects of friction conditions, die geometries, and processing routes were also examined under the same processing condition. Based on the value of average strain and standard deviation of the strain distribution, the magnitude and uniformity of deformation was determined depending on process conditions. In addition, damage parameters based on plastic work and Cockroft and Latham criteria were calculated to check the likeliness of surface cracking. Finally, non-isothermal analyses were carried out to explore the effect of processing temperature in the ECAE process.

Equal channel angular pressing (ECAP) is a viable forming procedure to extrude material by use of specially designed channel dies without a substantial change in geometry and to make an ultrafine grained material by imposing severe plastic deformation. Because the evolution of microstructures and the mechanical properties of the deformed material are directly related to the amount of plastic deformation, the understanding of the phenomenon associated with strain development is very important in the ECAP process. The plastic deformation behaviour during pressing is governed mainly by die geometry (channel sizes, a channel angle and corner angles), material properties (strength and hardening behaviour) and process variables (temperature, lubrication and deformation speed). There is a need for modelling techniques which may permit a wider study of the effects observed for better process control and the understanding of process related phenomena. In this study, we describe a range of our continuum modelling results of the ECAP process in order to illustrate the modelling applicability. Firstly, the finite element results of ECAP modelling for various geometric factors are described. Secondly, the inhomogeneous deformation due to the hardening property of the material is explained. Lastly, modelling the temperature field coupled with stress as a typical process variable in ECAP is presented.

An ultra low carbon interstitial free (IF) steel was severely deformed by the six-layer stack accumulative roll-bonding (ARB) process for improvement of the mechanical properties. As-received material with 1 mm in thickness showed a recrystallization structure with average grain diameter of 27 μm. The ARB was conducted at ambient temperature after deforming the as-received material to 0.5 mm thick by cold rolling. The ARB was performed for six-layer stacked, i.e. 3 mm thick sheet, up to 3 cycles (an equivalent strain of ∼7.1). In each ARB cycle, the stacked sheets were, first, deformed to 1.5 mm thick by the first pass, and then reduced to 0.5 mm thick, equals to the starting thickness, by multipass rolling without lubrication. The specimen after 3 cycles of ARB was annealed for 1.8 ks at various temperatures ranging from 673 K to 1073 K. The tensile strength of the ARB processed materials increased largely with the number of ARB cycles, after 3 cycles it reached a maximum of 1.12 GPa, which is about 4 times larger than that of the initial material. The elongation dropped largely after the cold rolling prior to the ARB, however it remains almost constant during the subsequent ARB process. Transmission Electron Microscopy revealed that the ARB processed materials exhibited a dislocation cell and/or subgrain structure with relatively high dislocation density. The selected area diffraction (SAD) patterns suggested that the orientation difference between neighboring cells was very small. The annealing up to 873 K resulted in gradual decrease in the strength due to the static recovery. The annealing above 873 K resulted in recrystallization and normal grain growth, and thereby a significant drop in the strength and recovery in ductility.

Intermetallics evolution in a commercial 6082 aluminium alloy severely deformed by Equal Channel Angular Pressing was investigated. Chemical electron probe microanalyses allowed to state that in the severely deformed alloy, Si-rich phases were progressively dissolved whereas the amount and composition of the Fe-Mn-Si containing intermetallics remained substantially unchanged. The moderate hardening effect measured on isothermal aging at 130°C of the ECAP processed samples was accounted for by the reprecipitation of the phases dissolved during severe plastic deformation. Tensile tests and fractographic analyses on the broken specimens showed that the 6082 alloy featured a relatively high ductility and a substantially unaltered fracture mode even after several ECAP passes.

The properties of multi-functional CuCr and CuCrZr alloys with ultra fine grains (UFG) produced by equal-channel angular pressing (ECAP) have been investigated in dependence of concentration of alloying elements. A special attention is paid to optimization of fatigue performance together with strength, thermal and electric properties after ECAP and subsequent aging. Substantial improvement of fatigue life is evidenced in ECAP-fabricated alloys when compared with conventional tempers.

Tensile ductility of magnesium alloy ZK60 pre-strained by warm equal channel angular extrusion (ECAE) was investigated at 300°C. It was shown that a significant ductility (262%) is achievable with this technique for a fairly high strain rate of 3·10⁻³ s⁻¹. The effect of pre-straining by warm ECAE on twinning under room temperature deformation was also investigated. It was shown that a bi-modal grain structure, whose characteristics are determined by the ECAE route, number of passes and temperature, has a strong effect on the propensity for twinning and mechanical properties, including ductility.

The significant enhancement of strength is reported for pure K24 gold and its K18 alloys subjected to grain refinement by equal channel angular pressing (ECAP) in comparison with their annealed or ordinary cold-worked counterparts. The effect of strain path on the tensile behavior is discussed from the stand point of mechanical properties and damage evolution during straining. It is shown that although different processing can give rise to substantially different structures, the mechanical properties depend on the pre-strain history only slightly. The ECAP appears to be particularly effective for strengthening of pure gold and Au-Ag solid solutions where the strength can increase by a factor of 3 or 4 while in the precipitation hardenable alloys the effect of grain refinement is masked by other strengthening mechanisms and the strength increases by a factor of 2. The strain localization and features of plastic deformation in ECAP materials are discussed in detail upon in-situ and post-experimental surface observations in differently fabricated samples.

Microstructure of martensite steel near drill hole surface was investigated. The microstructure of drill hole surface can be divided into 3 layers with depth. From top surface, nanocrystalline layer, equiaxed submicron grain layer and deformed martensite structure layer. Nano-grained layers have extremely high hardness, similar to those observed in the specimens treated by shot peening, ball milling, ball drop and particle impact deformation. The measured amount of shear strain produced by drilling was found to be an exponential function of depth from the hole surface. The amount of true strain at the hole surface region was estimated to exceed 7 which is considered to be the necessary amount of strain to produce nanocrystalline structure.

In the present study, wedge-shaped samples were used to determine the effect of nominal equivalent strain (between 0 and 1.2) and carbon content (0.06—sh0.35%C) on ferrite grain refinement through dynamic strain-induced transformation (DSIT) in plain carbon steels using single-pass rolling. The microstructural evolution of the transformation of austenite to ferrite has been evaluated through the thickness of the strip. The results showed a number of important microstructural features as a function of strain which could be classified into three regions; no DSIT region, DSIT region and the ultrafine ferrite (UFF) grain region. Also, the extent of these regions was strongly influenced by the carbon content. The UFF microstructure consisted of ultrafine, equiaxed ferrite grains (<2 μm) with very fine cementite particles. In the centre of the rolled strip, there was a conventional ferrite-pearlite microstructure, although ferrite grain refinement and the volume fraction of ferrite increased with an increase in the nominal equivalent strain.

The orientation distribution of ferrite grains formed by dynamic and static transformation was examined using the electron backscattered diffraction technique. The orientation of ferrite grains, transformed both dynamically and statically from deformed austenite, showed a large deviation from the Kurdjumov-Sachs (K-S) orientation relationship. The deformed austenite grain structure modeled using Ni-30Fe alloy showed inhomogeneous and localized deformations within the austenite grains. A large orientation variation was observed across this localized deformation zone, and the large deviation of the ferrite orientation from the K-S relationship was attributed to the nucleation of ferrite from this localized deformation region.

Ultrafine grained structure formed dynamically through a severe warm deformation in the temperature range from 773 K to 923 K has been investigated in a 0.16%C-0.4%Si-1.4%Mn steel. The effects of the deformation conditions such as deformation temperature and strain rate on microstructural evolution were examined using a single-pass compression technique with a pair of anvils. A large plastic strain up to 4 was imposed on the specimen interior at a strain rate of 1 or 0.01 s⁻¹. Ultrafine ferrite grains surrounded by high angle boundaries, whose nominal grain size ranged from 0.26 to 1.1 μm, evolved when the equivalent plastic strain exceeded the critical value about 0.5 to 1, and increased with an increase in strain without any large-scale migration of high angle boundaries. The effects of deformation conditions on microstructural evolution of ultrafine grained structures can be summarized into the Zener-Hollomon(Z-H) parameter dependences. The average size and the volume fraction of newly evolved ultrafine grains depend on the Z-H parameter. Decreasing Z-H parameter enhances the formation of equiaxed ultrafine grains. These indicate that the mechanism forming ultrafine grained structures through the warm severe deformation in the present study is similar to “continuous recrystallization” or “in-situ recrystallization” and that some activation process during or after the deformation plays an important role in the microstructural evolution.

Effect of pass strain (Δε) on grain refinement was studied in multidirectional forging (MDF) of a coarse-grained 7475 Al alloy at 490°C under a strain rate of 3 × 10⁻⁴ s⁻¹. Samples of rectangular shape were deformed up to accumulated strains of around 6 with subsequent changes in loading direction 90° from pass to pass. The pass strains in each compression (Δε) were 0.4 and 0.7. The cumulative flow curves integrated by each compression exhibit significant work softening just after yielding, followed by apparent steady state plastic flow at high strains. Structural changes were characterized by grain fragmentation due to frequent development of deformation and/or microshear bands followed by full evolution of new fine grains in the original grains. Increasing Δε accelerates significantly the kinetics of grain refinement, leading to more clear reduction of flow stresses at moderate to high strains. MDF of Δε = 0.7 results finally in formation of a finer grained structure with an average size of around 7.5 μm at strains of above 3.5, while, the processing with Δε = 0.4 develops a slightly coarser grain structure at higher strain of about 6. The effect of MDF on new grain evolution and the mechanisms of grain refinement are discussed in details.

Progress of Ferrous Nano-Metal Project is introduced in the present paper. In the project, maximum use of copper clusters and precipitates is pursued for achieving better strength-ductility balance than that of conventional high-tensile strength steels. Fundamental aspect of clustering and precipitation of Cu in Fe-Cu alloys was studied using Optical Tomographic Atom-Probe (OTAP). It was found that Cu precipitation during aging was enhanced by plastic deformation. The observed Cu precipitation behavior was well related to the age-hardening behavior, that is, aging started at lower temperature and maximum hardness was higher for plastically deformed and aged ferrite. Aging behavior and associated tensile properties were further examined for Fe-C-Mn-Cu martensitic steel. Higher value of tensile strength times elongation was achieved in Fe-C-Mn-Cu steel than Fe-C-Mn steel. Finally, effect of Cu precipitation on grain-refinement of ferrite was studied for Fe-C-Mn-Cu steel. Ferrite grain smaller than 1 μm was obtained in both processes of strain-assisted ferrite transformation from heavily deformed austenite and dynamic recrystallization of heavily deformed ferrite. Ferrite grain size was found to decrease by addition of Cu at the both processes. It was suggested that a simple additivity rule does not hold in terms of the strengthening by grain-refinement and that by precipitation, especially at grain-size range less than 1 μm.

In martensitic steels, it is well known that a certain chemical driving force (about 180 MJ/m³) is required to start martensitic transformation (Ms), and additional driving force has to be charged further to complete the transformation (Mf). In the case of metastable austenitic steels with Ms temperature at around room temperature, however, only the chemical driving force needed to start martensitic transformation has been stored at room temperature. Hence, the state of austenite is very unstable thermally. It has already been known that such a metastable austenite undergoes a partial martensitic transformation during isothermal holding at room temperature or cooling to a low temperature. It is very convenient to investigate the behavior of martensitic transformation of austenite. In this study, the effect of austenite grain size on martensitic transformation is introduced from the viewpoint of microstructural analysis and thermo-dynamics. The steel used in this investigation is an Fe-16 mass%Cr-10 mass%Ni ternary alloy, which has Ms temperature at around room temperature. The grain size of this steel can be controlled from 0.8 μm to 80 μm using the technique of reversion of deformation induced martensite. In the material with coarse grain size (80 μm), about 18% of martensite was detected at room temperature and the amount of martensite was increased to 50% by the following subzero treatment to 77 K. However, martensite was hardly detected in the material with ultra fine grains (0.8 μm) even after the subzero treatment. It was found that such a stabilization occurs in the materials with the grain size below 10 μm and the stabilization was reasonably explained by considering the relation between austenite grain size and elastic strain energy which is required on the single variant martensitic transformation.

The annealing behaviour of ultrafine grained steels containing nano-scale dispersed oxides was studied in a temperature range of 600—900°C by means of microstructural observations and hardness measurement. The starting materials with submicrocrystalline structures were developed by mechanical milling of Fe-Fe₃O₄ powders followed by consolidating bar rolling at 700°C. Depending on the initial oxygen content and the mechanical milling time, the fraction of dispersed oxides varied from 0.3 to 3.0 vol%. During the heating up to 800°C (i.e. within the ferrite region), the initial ultrafine grained microstructures were essentially stable against any discontinuous grain growth. The grain coarsening and the softening can be roughly expressed by power-law functions of annealing time. The main mechanism of microstructure evolution that operated during annealing is considered as a normal grain growth accompanied by recovery. The grain coarsening is characterized by a rather high value of the grain-growth exponent of about 20. The grain growth kinetics correlates with the oxide coarsening. The effect of dispersed oxides on the annealing behaviour of submicrocrystalline oxide bearing steels is discussed in some detail.

Ultrafine grained austenitic structure was obtained in 18Cr-9Ni stainless steel by thermomechanical treatment using reversion from deformation-induced martensite. The superplastic deformation behavior was investigated at 923 K for the steels containing various amounts of retained martensite particles in the initial structure before tensile testing. The retained martensite was effective for suppressing grain growth of austenite and necessary for the superplasticity although it was thermodinamically unstable phase and gradually decreased its volume fraction with superplastic deformation. Therefore, the superplastic elongation was strongly dependent on the initial volume fraction of the retained martensite. The total superplastic elongation was enlarged with increasing the initial amount of martensite, and the maximum elongation of about 270% was obtained when the volume fraction was controlled to be around 10 vol%. However, the increase in elongation was leveled off in the range above 15 vol% martensite. The effect of the retained martensite on the superplasticiy was discussed in connection with the changes in volume fraction of the martensite, austenite grain size and deformation mechanism.

A sub-micron grained microstructure in an Al-0.2 mass% Sc alloy was produced by high strain deformation using Equal Channel Angular Pressing (ECAP). The alloy was solution treated prior to deformation, deformed by ECAP then aged at low temperature to produce a sub-micron grained microstructure with a large fraction of high angle grain boundaries (HAGB) decorated with fine Al₃Sc particles. General grain stability and particle/grain boundary interactions were studied using scanning electron microscopy (SEM), electron backscatter diffraction (EBSD), focussed ion beam (FIB) microscopy and transmission electron microscopy (TEM). The fine-grained microstructure was found to be highly stable during annealing at temperatures up to 500°C due to Zener pinning from stable Al₃Sc particles. The volume fraction, f, and average radius, r, of particles and their rate of coarsening were found to have a strong influence on grain growth. It was found that the limiting grain size, R_c, in the Al-Sc alloy may reasonably be predicted by the relation: R_c = 0.17r/f. This relation is known to be applicable for coarse-grained alloys (>1 μm) and indicates its validity for predicting the limiting grain size in sub-micron, particle-containing alloys.

A plain IF steel and a P-added IF steel having various ultrafine grain sizes from 0.24 to 11 μm were fabricated by the accumulative roll bonding (ARB) process followed by annealing. Dynamic fracture toughness of the ultrafine grained IF steels was investigated as a function of grain size by miniaturized Charpy impact test. The static strength of the IF steels significantly increased with decreasing the grain size, while the uniform elongation was limited in the ultrafine grained samples. A number of delamination appeared in the impact-tested specimens, especially in the ultrafine grained materials at low temperatures. It was concluded that the frequent delamination is not owing to insufficient roll-bonding in the ARB specimens but it is rather a characteristic feature of the ultrafine grained materials fabricated through heavy deformation. Because of the delamination, the absorbed energy in the impact test continuously decreased with decreasing the test temperature. On the other hand, an obvious change from the ductile fracture surface characterized by dimples into the brittle fracture surface mainly due to intergranular fracture was recognized at a certain low temperature. The ductile-brittle transition temperature determined from the microscopic fracture surfaces greatly decreased with decreasing the grain size, and finally no brittle fracture happened even at −190°C when the grain size was smaller than 5 μm or 2 μm in the plain IF steel or the P-added IF steel, respectively. It was concluded that the ultra-grain refinement is quite effective to improve the low-temperature toughness of ferritic steels and that it is possible to make phosphorus substantially harmless by grain refinement.

A study was carried out on the aging behaviour of a solution treated 6082 aluminium alloy deformed by equal channel angular pressing up to six passes. Aging of the so obtained ultrafine alloy was studied by DSC and TEM analyses as well as by isothermal treatments at 130, 160 and 180°C. The results showed that the formation of metastable β″ and β′ precipitates were markedly anticipated and that the stable β phase was partially suppressed in the alloy processed to 4 and 6 passes, presumably due to anticipated precipitation of Si-rich particles. TEM analyses revealed that a transition in strengthening-precipitate structure occurred in the severely deformed alloy leading to a predominance of globular precipitates over the rod like phases typically found in the undeformed and aged samples. Tensile tests carried out on severely deformed and aged samples allowed to quantify the maximum strength achievable by the 6082 alloy after concurrent grain refinement and peak-hardness aging at 130°C.

The current study was conducted to develop a Ti-6Al-4V alloy composite material with a fine structure with improved mechanical properties by precipitating TiC particles via reaction-sintering and, additionally, by incorporating the effects of hydrides formed by hydrogenation treatment, using powder to which Mo₂C had been added. The effects of precipitated TiC particles, and the subsequent generation of hydrides by hydrogenation treatment, heavy-strain working, and recrystallization treatment on grain refinement during the forming process were studied by means of structural observations. Warm rolling and multi-axial alternate forging (MAF) were employed as the heavy-strain working methods. The base metal structure of the Ti-6Al-4V composite material prepared by hydrogenation treatment on the sintered body prepared by spark plasma sintering (SPS) contains particles with a grain size not greater than 1 μm, which become finer with increasing hydrogen content. The particle size of TiC formed by reaction sintering, however, is not changed by heavy-strain working after hydrogenation. Its room temperature tensile strength increases with the amount of TiC precipitation, and is higher than that of non-hydrogenated material.

Composite creep deformation was analyzed, based on a continuum plasticity representation of the matrix, in an ideal composite at high temperatures in the case of negligible interfacial diffusion and sliding. A general formula of the steady-state creep strain rate was derived for a composite consisting of an ellipsoidal rigid reinforcement and a creeping matrix with a stress exponent of one. A closed-form solution was then derived for a composite with a cylindrical reinforcement under pure shear deformation in a two-dimensional analysis. The resultant creep deformation satisfies the requirements of impotency, volume conservation and interfacial continuity. Traces of two types of edge dislocations were analytically drawn; they show that dislocations climb over the reinforcement, retaining no dislocations either in the matrix or at the interface. Also, two types of dislocations simultaneously climb up and down at any portion in the matrix through dislocation core shuffling without long-distance diffusion. Finally, it was concluded that plastically-accommodated creep is characterized by two types of dislocations that simultaneously climb over a reinforcement, generating a heterogeneous creep strain increment without long-distance diffusion.