Weak-beam Electron Microscopy Research Articles

Effects of titanium concentration and tungsten addition on the nano-mesoscopic structure of high-Cr oxide dispersion strengthened (ODS) ferritic steels

To study the effects of titanium concentration and tungsten addition on the nano-mesoscopic structure of high-Cr oxide dispersion strengthened (ODS) ferritic steels, the spatial and size distributions, shape, and coherency of the oxide particles in SOC-5 (Fe–15.95Cr–0.09Ti–0.34Y2O3) and SOC-P3 (Fe–13.32Cr–1.9W–0.16Ti–0.33Y2O3) were studied by diffraction contrast techniques, including weak beam electron microscopy. When the titanium concentration is increased from 0.09 to 0.16wt.% with 1.9wt.% W added, the grain size decreases considerably, while the number density and mean size of the oxide particles significantly increases and decreases, respectively. ∼63.5% and ∼96% of the oxide particles in SOC-5 and SOC-P3 (diameter <4.5nm), respectively, are coherent with the bcc steel matrix. In SOC-5, ∼36% of the oxide particles (4.5–10nm in diameter) are semi-coherent; misfit moiré fringes are seen across only ∼12% of the oxide particles with misfit moiré fringe spacing of 1.12nm, indicating the misfit strain is 0.11. However, only ∼4% of the oxide particles in SOC-P3 (4.5–10nm in diameter) are semi-coherent; misfit moiré fringes are seen across only ∼2% of the oxide particles with misfit moiré fringe spacing of 0.98nm, indicating the misfit strain is 0.126. The mean misfit strain of coherent oxide particles in SOC-5 is ∼0.017.

Polymorphic and coherency transition of Y–Al complex oxide particles with extrusion temperature in an Al-alloyed high-Cr oxide dispersion strengthened ferritic steel

The phase and metal/oxide interface structure of the nanometer-scale particles in an Al-alloyed high-Cr oxide dispersion strengthened ferritic steel extruded at 1150 °C and 1050 °C were characterized by high-resolution transmission electron microscopy and diffraction contrast techniques, including weak beam electron microscopy. After extrusion at 1150 °C, yttrium–aluminum–hexagonal (YAH, YAlO 3) and yttrium–aluminum–perovskite (YAP, YAlO 3) oxides (diameter ⩽10 nm) constitute ∼55% and 38% of the particles, respectively; ∼78% of the particles (4.5–10 nm in diameter), which include 40% YAH oxide and 38% YAP phase with misfit (translational) moiré fringe spacing of 2.15 nm and 1.65 nm, respectively, are semi-coherent with the matrix. After extrusion at 1050 °C, almost all the particles are YAH phase, and ∼86.5% (diameter <4.5 nm) are coherent with the matrix. The coherency of the oxides is size dependent. The crystallographic orientation correlations of the oxides and matrix were found.

DISLOCATION PINNING IN HIGH TEMPERATURE DEFORMED Ni 3(Al,Ti) SINGLE CRYSTALS CONTAINING DISORDERED γ PRECIPITATES

A transmission electron microscopy (TEM) investigation has been performed on the dislocation pinning in L1 2 -ordered Ni 3 (Al,Ti) containing disordered γ precipitates. The morphology of deformation induced dislocations in the γ′ base alloys containing fine dispersion of disordered γ was investigated by means of weak-beam electron microscopy. The superdislocations are strongly attracted to the disordered particles and dissociate on the (111) plane in the γ particles, while they dissociate on the (010) plane in the γ′ matrix. The disordered γ precipitates play an important role as a pinning point during the cross-slip of superdislocations from (111) to (010) planes in the γ′ matrix and restrain the cross-slip of superdislocations. The interaction of superdislocations with disordered particles causes the formation of superkinks, jogs and closed loops.

From dislocation cores to strength and work-hardening: a study of binary Ni 3Al

A quantitative model for the peak temperature in work-hardening in L1 2 intermetallics is proposed. It is based on the competition between the exhaustion of mobile dislocations by the Kear Wilsdorf mechanism and the yielding of incomplete locks at high stress. The model is assessed by a set of experimental data measured in binary Ni 3Al polycrystals of three different compositions. These include, in particular, the planar fault energies of the dislocation cores measured by weak-beam electron microscopy, combined with computer image simulations and macroscopic data about flow stress, work-hardening and mobile dislocation exhaustion rates. These parameters are measured as a function of alloy composition. The model also fits successfully data published for other L1 2 compounds.

Electron microscopy of dislocations introduced into GaN by plastic deformation

Dislocations introduced by plastic deformation into GaN with the wurtzite structure have mostly an a -type Burgers vector. Dislocations on prismatic planes are not dissociated as observed by weak-beam electron microscopy. Highresolution electron microscopy of basal dislocations has revealed that both the dissociated and undissociated states are stable. From the widths of the dissociated basal dislocations observed by weak-beam electron microscopy, the energy of the intrinsic stacking fault has been estimated to be 22 4 mJm-2. This value is consistent with the correlation previously found between the reduced stackingfault energy and the c/a ratio of the wurtzite crystals.

Yield stress anomalies in single crystals of Ti-54.5 at.% Al II. ½[112](111) slip

Abstract A yield stress anomaly has been observed for ½[112](111) slip in single crystals of γ-Ti-54.5 at.% Al. Weak-beam electron microscopy studies have been carried out to elucidate the origin of this anomaly. At room temperature, screw and 60° dislocations are dissociated in the glide plane, and 30° dislocations are dissociated by glide on intersecting {111} planes, forming weak locks. At high temperatures, ½[112] dislocation loops are elongated along their edge direction, and these edge dislocations are dissociated by climb and glide into sessile configurations consisting of Frank and Shockley partials on different intersecting {111} planes, with a stair-rod dislocation along the line of intersection. There is some evidence for further dissociation of the Frank partial. These strong locks are responsible for the yield stress anomaly. The driving force for the anomaly is the reduction in dislocation energy; the mechanism involves climb, which becomes easier with increasing temperature.

Misfit Strain Relaxation by Dislocations in InAs Islands and Layers Epitaxially Grown on (001)GaAs Substrates by MOVPE

The misfit dislocation configurations in InAs islands as well as in more or less continuous layers grown on (001) oriented GaAs substrates were studied by weak-beam and high-resolution electron microscopy. The islands are confined by {101} and {111} facets where the aspect ratio (height/lateral extension) can be affected by the growth conditions. It is possible to grow well-defined islands as well as relatively continuous layers by MOVPE under As-stabilized conditions. At constant deposition parameters the growth is characterized by islands of different sizes (but with constant aspect ratio) in various strain states depending on their dislocation content. Coherently strained islands without any dislocation can be observed for heights up to 23 ML InAs, or otherwise, up to a maximal island extension of about 12 nm (for the particular aspect ratio ≈0.585). With further increase of island height and lateral extension, the introduction of dislocations becomes favourable. Independent of the island size, the layer thickness and the dislocation density, a residual elastic strain of about er = —0.8% remains after relaxation. This means, about 88% of the total misfit strain of e = —6.686 × 10—2 were compensated by Lomer dislocations. These sessile Lomer dislocations lie in the island interior only, where single 60° dislocations were observed exclusively in their near-edge regions. With increasing island size and/or layer thickness some close-spaced 60° dislocations occur additionally within the interfacial region. The Lomer dislocations that are always located 4 monolayers (ML) above the InAs/GaAs interfacial plane result from the well-known fusion of two 60° slip dislocations. These 60° dislocations have been nucleated 7 … 8 ML above the interface at surface steps on the {111} facets confining the islands. Based on our experimental observations a new mechanism is proposed that explains the origin of these 60° dislocations. Their further fusion to sessile Lomer dislocations that compensate the misfit strain most efficiently occurs in the way as commonly accepted.

Characterisation of dislocations, nanopipes and inversion domains in GaN by transmission electron microscopy

Transmission electron microscopy is used to analyse a range of defects observed in hexagonal GaN films grown on sapphire and GaN substrates by metalorganic chemical vapour deposition. Large angle convergent beam electron diffraction is used to analyse the Burgers vectors of dislocations and to show that hollow tubes, or nanopipes, are associated with screw dislocations having Burgers vectors±c. Weak-beam electron microscopy shows that dislocations are dissociated into partials in the (0001) basal plane, but that threading segments are generally undissociated. The presence of high densities of inversion domains in GaN/sapphire films is confirmed using convergent beam electron diffraction and the atomic structure of the { 10 1 ̄ 0 } inversion domain boundary is determined by an analysis of displacement fringes seen in inclined domains.

Comparison of the Deformation Microstructures at Room Temperature in O and B2 Phases of a Ti2AlNb Alloy

The dislocation microstructures in both O and B2 single phases of a Ti 2 AlNb alloy, deformed at room temperature, are reported. Slip systems activated during the deformation of each phase are determined using weak-beam electron microscopy and the correspondence between the deformation processes is made. It is shown that the better ductility of the B2 phase compared to the O phase is related to the number of activated slip systems.

Weak-beam imaging of dissociated dislocations in HVEM-irradiated Fe-Ni-Cr alloys

We report here on studies by weak-beam electron microscopy of the evolution of microstructures at and near preexisting line dislocations in a number of Fe-Ni-Cr alloys under electron irradiation in a high-voltage electron microscope (HVEM). The detailed observations are discussed in terms of dislocation climb mechanisms in these materials and a model based on interstitial pipe diffusion.

Comment on ‘Transmission electron microscopy of dislocation structures in iron-doped Al3Ti with the L12structure’

Abstract In deformed L12 Al3Ti, a transition of the dissociation of ⟨110⟩ superdislocations from two superpartials with ⅓⟨112⟩ Burgers vectors (mode II) at a low temperature to two superpartials with collinear ½⟨110⟩ Burgers vectors (mode I) at a high temperature, has been reported by Inui et al. in 1992. The weak-beam electron microscopy observations presented by these authors are re-examined. It is shown that the contrasts observed in the microstructure of deformation at a low temperature are fairly consistent with a narrow splitting into mode I rather than a dissociation according to mode II.

Estimation of the core cut-off parameter for misfit dislocations at a SixGe1−x/Si interface

The misfit dislocation structure at a partially relaxed Si/SixGe1−x interface has been investigated using weak-beam electron microscopy in order to probe features near the dislocation core. The dislocations were found to be dissociated, allowing for the direct measurement of the stacking fault energy at the epitaxial interface. This has been supplemented with the measurement of the width of symmetrical threefold stacking fault nodes at the same interface. These independent measurements have been combined to estimate the dislocation core cut-off parameter using isotropic approximations of anisotropic elasticity theories. The size of the core of the misfit dislocation is consistent with both geometric and atomistic models predicting core delocalization at coherent interfaces.

Dislocations and their dissociation in SixGe1 - xalloys

Abstract The dissociation of dislocations in undeformed SixGe1 - x crystals grown from the melt with low (x = 0.05) and high (x = 0.5) Si contents was investigated using the weak-beam electron microscopy technique. Dislocations with total Burgers vectors b=½<110> and of various orientations were found to be dissociated into Shockley partials, separated by an intrinsic stacking fault on a {111} glide plane. Extended dislocation nodes were also observed. For x = 0.05 the dissociation widths of single dislocations ranged from Δ = 3.5 ± 1 nm for screw orientations to Δ = 6.2 ± 1 nm for edge orientations. The corresponding values for x = 0.5 were higher, with separations between Δ = 3.9 ± 1 nm and Δ = 6.9 ± 1 nm. Stacking-fault energies γ = 56 ± 8 mJ m−2 for x = 0.05 and γ = 53 ± 8 mJ m−2 for x = 0.5 were deduced from measurements of the dissociation widths as a function of dislocation orientation, using anisotropic elasticity theory. The results are compared with those of previous investigations on elemental Si and Ge.

On interface and twin boundaries in a forged alloy

This paper discusses XD (exothermic dispersion) processed TiAl/TiB{sub 2} alloys which are a new class of intermetallic composites with improved toughness, creep resistance and oxidation resistance at high temperatures. They generally have a two-phase {alpha}{sub 2} + {gamma} matrix (Ti{sub 3}Al (DO{sub 19}) and TiAl (L1{sub 0} structures, respectively), which contains monocrystalline TiB{sub 2} particles. Although the deformation mechanisms of these alloys are essentially the same as those of conventional TiAl alloys, dislocation substructures, particularly dislocations at {alpha}{sub 2}/{gamma} interfaces, remain to be unambiguously identified. In a recent paper on a Ti-45Al/TiB{sub 2} alloy, these interfacial dislocations were assumed to be misfit dislocations with 1/6 {l angle}112{r angle}{sub {gamma}}. Burgers vectors and their presence was attributed to boron interstitials in the {gamma} matrix, However, the {alpha}{sub 2} phase has been shown to getter interstitial impurities from the {gamma} phase, thus increasing the ductility of the two phase alloy. Moreover, 1/6 {l angle}112{r angle}{sub {gamma}} partial dislocations generally result from the dissociation of glissile {gamma} dislocations during deformation and cannot be resolved by conventional weak beam electron microscopy because of their small separation distances. Their assumed presence is also unusual, in that lattice mismatch is not normally accommodated by partial dislocations.

Dislocations in plastically deformed L12 compounds based on Al3Ti

The fine structure of dislocations in lightly deformed samples of several cubic ordered alloys with composition based on Al3Ti has been examined by weak beam electron microscopy. For all the materials examined the dislocations tend to dissociate into two 1/2(110) partials separated by APB. Dislocation dissociation is not complete at very small strains and the strain required to dissociate, as well as the dissociation distance, varies from one alloy to another. Improvements in ductility achieved by alloying are directly related to the ease and extent of this dissociation.

Dislocation structures at Cu[sbnd]MgO and Pd[sbnd]MgO interfaces

The dislocation structures at Cu and PdMgO interfaces have been studied by high-resolution and weak beam electron microscopy. The metal-oxide interfaces were formed by internal oxidation of Cu-1wt%Mg and Pd-1wt%Mg alloys. The observations have been made at (111), (100) cube-on-cube and (111) twin interfaces. All interfaces have been found to be partially coherent despite the large misfit of 14.2 and 7.6% for Cu- and PdMgO systems, respectively, The dislocation configuration, network periodicity and dislocation type are presented and discussed in terms of the geometrical DSC-CSL lattice model.

Dislocation glide at a (100) SixGe1−x/Si interface

An unusual fourfold extended dislocation node at a SixGe1−x/Si strained layer interface has been revealed using weak-beam electron microscopy. A detailed contrast analysis shows that this node structure is the result of the constriction of Lomer–Cottrell dislocations formed by intersecting slip dislocations on {100} and {111} type planes. This mechanism necessitates the glissile behavior of dislocation nodes at the (100) epitaxial interface.

Climb of extended dislocations in silicon caused by oxygen precipitation

Morphology of dislocations during climbing caused by excess interstitials associated with oxygen precipitation in Czochralskigrown silicon is investigated by means of weak beam electron microscopy. Dislocations induced by plastic deformation are extended in any stage of climbing. The climbing process of an extended dislocation in a diamond-type crystal is described with a model based on an idea originally presented by Thomson and Balluffi for a fcc crystal.

On the structure of faulted interfaces in aluminium nitride ceramics

Abstract An unusual two-dimensional structure in hot-pressed wurtzite A1N, composed of a mixture of both curved and flat elements, is analysed by weak-beam and high-resolution electron microscopy. It is demonstrated that the structure, in its elementary form, is a ‘dome’-like faulted defect. This defect is characterized by a Frank-type dislocation loop containing a metal-vacancy-induced stacking fault in the basal plane with displacement Rp = (1/3) [1100] + 0·15[0001] and a curved faulted interface with displacement vector Rc = (1/3)[1100] + (1/2)[0001]. The elementary dome is found always pointing in the same direction, reflecting the lack of symmetry in the tetrahedral unit structure of A1N with respect to the basal plane. Condensation of supersaturated metal vacancies on both interfaces, combined with the segregation of impurities, mainly oxygen, only on the planar interface, is considered to be the driving force for the nucleation and growth of the structure.

Mechanisms of defect formation and growth during thermal ramping and annealing in oxygen implanted silicon-on-insulator material

The development of defects in high-dose oxygen implanted silicon-on-insulator (SIMOX) material during thermal ramping and annealing was directly studied by weak beam and high resolution electron microscopy. The ramping and annealing cycle was simulated by annealing a series of samples for 2 h at temperatures from 700 to 1250°C. Short stacking faults, 10–30 nm lenght, are present in as-implanted material in the superficial silicon layer with some extending to the wafer surface from small, near-surface precipitates. During annealing from 900 to 1100°C the near-surface precipitates grow and generate additional stacking faults, some of which grow downward through the superficial silicon layer toward the buried oxide. At higher annealing temperatures the near-surface precipitates dissolve, but stacking faults are stabilized by the wafer free surface and by precipitates near the buried oxide. Lateral dislocation segments also from between precipitates near the buried oxide during thermal ramping. At the latter stages of high temperature annealing, precipitates and lateral pinned dislocations are incorporated into the buried oxide, but defects running through the superficial silicon layer are stabilized by the wafer surface and by the buried oxide, resulting in a high defect density of 10 9 cm −2. Techniques for processing SIMOX for defect density reduction are discussed in terms of the role of the defect formation mechanisms.