An electron microscopy study of the phase Al3Fe

The phase Al₃Fe (monoclinic C2/m, a = 1·549 nm, b = 0·808 nm, c = 1·248 nm, β = 107·8°) has been studied by transmission electron microscopy (TEM) and high resolution electron microscopy (HREM). Crystals were obtained from a direct chill-cast ingot of an Al-0·25 wt% Fe-0·13 wt% Si alloy. Extracted crystals were prepared by dissolving the aluminium phase in butanol and filtering off the particles. The extracted Al₃Fe crystals were mainly (100) platelets. The monclinic lattice was confirmed by tilt experiments and the mirror plane was confirmed by convergent beam electron diffraction. Experimental HREM images from the [100] and [110] projection agreed with images calculated by the multislice method. The interpretation of images in terms of a projected crystal structure is discussed. Common defects in Al₃Fe crystals were: twins on (100) and faults on (001). The (001) faults could be described by a displacement 1/2·[100] on a fault plane at z = 0·5 in the unit cell. A model for (001) faults, based on multiple twinning, is proposed.     

Modification of Mo-Si alloy microstructure by small additions of Zr

Molybdenum and its alloys are potential materials for high-temperature applications. However, molybdenum is susceptible to embrittlement because of oxygen segregation at the grain boundaries. In order to alleviate the embrittlement small amounts of zirconium were alloyed to a solid solution of Mo-1.5Si alloy. Two Mo-based alloys, namely Mo-1.5Si and Mo-1.5Si-1Zr, were investigated by the complementary high-resolution methods transmission electron microscopy and atom probe tomography. The Mo-1.5Si alloy shows a polycrystalline structure with two silicon-rich intermetallic phases Mo₅Si₃ and Mo₃Si located at the grain boundaries and within the grains. In addition, small clusters with up to 10 at% Si were found within the molybdenum solid solution. Addition of a small amount of zirconium to Mo-1.5Si leads to the formation of two intermetallic phases Mo₂Zr and MoZr₂, which are located at the grain boundaries as well as within the interior of the grain. Transmission electron microscopy shows that small spherical Mo-Zr-rich precipitates (<10 nm) decorate the grain boundaries. The stoichiometry of the small precipitates was identified as Mo₂Zr by atom probe tomography. No Si-enriched small precipitates were detected in the Mo-1.5Si-1Zr alloy. It is concluded that the presence of zirconium hinders their formation.     

Evolution of rapidly solidified NiAlCu(B) alloy microstructure

This study concerned phase transformations observed after rapid solidification and annealing at 500, 700 and 800 °C in 56.3 Ni‐39.9 Al‐3.8 Cu‐0.06 B (E1) and 59.8 Ni‐36.0 Al‐4.3 Cu‐0.06 B (E2) alloys (composition in at.%). Injection casting led to a homogeneous structure of very small, one‐phase grains (2–4 µm in size). In both alloys, the phase observed at room temperature was martensite of L1₀ structure. The process of the formation of the Ni₅Al₃ phase by atomic reordering proceeded at 285–394 °C in the case of E1 alloy and 450–550 °C in the case of E2 alloy. Further decomposition into NiAl (β) and Ni₃Al (γ′) phases, the microstructure and crystallography of the phases depended on the path of transformations, proceeding in the investigated case through the transformation of martensite crystallographic variants. This preserved precise crystallographic orientation between the subsequent phases, very stable plate‐like morphology and very small β + γ′ grains after annealing at 800 °C.     

Electron crystallography applied to the structure determination of Nb(Cu,Al,X)2 Laves phases

The presence of primary precipitates of the Laves phases considerably improves the mechanical properties and the resistance to thermal degradation of the high‐temperature shape memory Cu–Al–Nb alloys. The structure analysis of the Laves phases was carried out on particles contained in the ternary and quaternary alloys as well on synthesized compounds related to the composition of the Nb(Cu,Al,X)₂ phase, where X = Ni, Co, Cr, Ti and Zr. The precise structure determination of the Laves phases was carried out by the electron crystallography method using the crisp software.     

The effect of an Ni–Cr protective layer on cyclic oxidation of Ti3Al

The effect of an 80Ni?20Cr (at.%) metallic coating on the cyclic oxidation behaviour of a Ti₃Al‐based alloy with the composition Ti?25Al?11Nb (at.%) was investigated in this study. Cyclic oxidation tests were carried out in air at 600 °C and 900 °C for 120 h. For one cycle test, the specimens were held for 24 h at test temperature and then furnace‐cooled to room temperature. The oxidation rate was determined by plotting the mass gain per unit surface area of the specimen vs. exposure time. The morphology and composition of the oxidation products were characterized on the cross‐section of the specimens by scanning electron microscopy, energy‐dispersive X‐ray spectroscopy and atomic force microscopy. The oxidation scale forms during exposure at both 600 °C and 900 °C. TiO₂ is the main oxide component, whereas the Al₂O₃ layer appears only discontinuously. The remarkable improvement in oxidation resistance at 900 °C was attributed to the chemical composition and structure of the scale formed on the 80Ni?20Cr coating.     

Microstructure of a spray-formed Al/SiC composite

The microstructure of an Al–6Cu–2Mn–0·45 Mg–1(Ag, Ti, V, Zr, Cr) alloy, reinforced with 13 vol.% SiC particles, made by spray deposition has been investigated by transmission electron microscopy, high-resolution electron microscopy and electron diffraction X-ray spectroscopy. Particular attention was focused on the influence of the reinforcement on the precipitation sequence. Instead of the expected precipitation sequence due to the high Cu/Mg ratio, there is an additional σ precipitate which was previously observed in A1 alloys containing silicon. This precipitate becomes predominant at the T6 temper. The new precipitation sequence for this reinforced alloy is therefore The precipitation of σ phase is believed to be due to the presence of SiC particles, and seems to be correlated with the occurrence of large Mn-rich particles. Although expected, no S phase precipitation is found to occur in the matrix grains. At the matrix grain boundaries, small Al₂Cu (θ) and Al₂CuMg (S), as well as Mn-rich precipitates are found. At the SiC particle surfaces, preferentially orientated Ag-rich and Mg-rich intermetallic precipitates are found. They can coexist with amorphous patches containing oxygen enclosed in an irregularly shaped Al₂Cu (θ) phase remaining from large crystalline areas which did not go into solution.     

Improving the quality of electron backscatter diffraction (EBSD) patterns from nanoparticles

In this study, we investigated the relative contributions of atomic number (Z) and density (ρ) to the degradation of the electron backscatter diffraction (EBSD) pattern quality for nanoparticles < 500 nm in diameter. This was accomplished by minimizing the diffuse scattering from the conventional thick mounting substrate through the design of a sample holder that can accommodate particles mounted on thin‐film TEM substrates. With this design, the contributions of incoherently scattered electrons that result in the diffuse background are minimized. Qualitative and quantitative comparisons were made of the EBSD pattern quality obtained from Al₂O₃ particles approximately 200 nm in diameter mounted on both thick‐ and thin‐film C substrates. For the quantitative comparison we developed a ‘quality’ factor for EBSD patterns that is based on the ratio of two Hough transforms derived from a given EBSD pattern image. The calculated quality factor is directly proportional to the signal‐to‐noise ratio for the EBSD pattern. In addition to the comparison of the thick and thin mounting substrates, we also estimated the effects of Z and ρ by comparing the EBSD pattern quality from the Al₂O₃ particles mounted on thin‐film substrates with the quality of patterns obtained from Fe–Co nanoparticles approximately 120 nm in diameter. The results indicate that the increased background generated in EBSD patterns by the electrons escaping through the bottom of the small particles is the dominant reason for the poor EBSD pattern quality from nanoparticles < 500 nm in size. This was supported by the fact that we were able to obtain usable EBSD patterns from Al₂O₃ particles as small as 130 nm using the thin‐film mounting method.     

Friction and wear properties of lead-free aluminum alloy bearing material with molybdenum disulfide layer by a reciprocating test

Friction and wear behavior of Al–Sn–Si alloy with MoS₂ layer under lubricated condition was investigated by a reciprocating friction tester. It became clear that the Al–Sn–Si alloy with MoS₂ layer showed about 70% lower friction and about 1/10 lower wear depth compared to the Al–Sn–Si alloy. The worn surfaces of the Al–Sn–Si alloy with MoS₂ layer were observed and analyzed by a SEM, a TEM and an EDX. It indicated that the sliding surface of the counterface had larger area of Mo than the area of Al which was transferred from the Al–Sn–Si alloy with MoS₂ layer by sliding, resulting in low friction and high wear resistance.     

Microstructure and creep characteristics of experimental SiCf–YMAS composites

A unidirectional SiC_f –YMAS glass–ceramic composite has been developed by Céramiques-Composites (Bazet) and ONERA (Establishment of Palaiseau) in France. The matrix is totally crystalline and consists essentially of two main phases, cordierite and yttrium disilicate, with some minor phases, mullite, spinel, zirconium and titanium oxides. Image analysis methods have been used to characterize the homogeneity of the composite plates and to obtain granulometric information on the different matrix phases. Different interphase layers formed during the process by reaction between the matrix and the Nicalon NLM 202 fibres have been studied by using HREM and EDX. Their chemical composition has been determined by stepping the probe (8 nm) across the fibre–matrix interface. Two distinct nanoscale sublayers have been imaged. The sublayer on the matrix side has a light contrast in the TEM. The microstructure of this layer (≈ 80 nm) is typical of a turbostratic carbon. The carbon layer also contains Al, O, Mg and Si. The silicon content is low in the carbon layer. The sublayer on the fibre side (≈ 100 nm thick) has a dark contrast in the TEM. Profiles have been taken across this sublayer also. Tensile creep tests in air have been performed to investigate the tensile creep behaviour at 1223 K. They have been conducted in the 50–200 MPa stress range. Tensile creep results indicate that creep rates are of the same order of magnitude as for other glass–ceramic composites. Optical micrographs and SEM observations have revealed the damage in the composite. Changes occurring in the interface region have been studied at a finer scale by TEM and HREM at the surface of the sample and in the core. These observations enable us to explain the mechanical behaviour of the composite observed on a macroscopic scale.     

Microstructure and joining mechanism of Ti/Al dissimilar joint by pulsed gas metal arc welding

Butt joining of titanium alloy Ti–2Al–Mn to aluminum 1060 using AlSi5 filler wire was conducted using pulsed gas metal arc welding. Joining mechanism of Ti–2Al–Mn/Al 1060 dissimilar joint with different welding heat input was investigated. Formations of precipitation and Ti/Al interface were discussed in detail. Fusion zone near aluminum is composed of α-Al dendrites and Al–Si hypoeutectic structures. A few TiAl₃ precipitations appear in the weld metal owing to metallurgical reactions of Al with dissolved Ti. When the welding heat input was in the range of 1.87–2.10 kJ/cm, titanium alloy Ti–2Al–Mn and Al 1060 were joined together by the formation of a complex Ti/Al interface. With a low welding heat input, a serrate TiAl₃ interfacial reaction layer was formed near Ti/Al interface. With the increasing of the welding heat input, α-Ti, Ti₇Al₅Si₁₂, and TiAl₃ layers were formed orderly from Ti–2Al–Mn to weld metal.     

Tribological behavior and wear mechanisms of MoSi2-base composites sliding against AA6063 alloy at elevated temperature

Intermetallic Mo(Si,Al)₂, Mo(Si,Al)₂/Al₂O₃, Mo(Si,Al)₂/SiC and Mo(Si,Al)₂/ZrO₂ composites produced by spark plasma sintering of mechanically alloyed powders were tested on a block-on-cylinder apparatus, sliding against an AA6063 alloy cylinder at elevated temperature. Abrasion, micro-fracture and surface tribochemical reactions were found to be the operative wear mechanisms, producing severe wear in the investigated alloys. Abrasive wear by pull-out of Al₂O₃ and micro-fracture of Mo(Si,Al)₂ particles promotes severe wear in the Mo(Si,Al)₂/Al₂O₃ composite. In the Mo(Si,Al)₂/SiC composites, hard SiC inclusions suppressed the abrasive wear, but a tribochemical reaction was found to be the dominant wear mechanism. A combination of abrasion by pull-out of Al₂O₃ particles and a tribochemical reaction was revealed to be the main wear mechanism in the Mo(Si,Al)₂/ZrO₂ materials. The brittleness index B = H/K_1C was applicable for prediction of the relative wear resistance. In agreement with the suggested model, the lowest wear rate, corresponding to B = 5.5–6.5 μm^−1/2, was found in the Mo(Si,Al)₂/30 vol.% SiC and Mo(Si,Al)₂/30 vol.% ZrO₂ composites.     

Dislocations in nanostructured two‐phase Fe30Ni20Mn20Al30

In a previous study, the dislocations in Fe₃₀Ni₂₀Mn₂₅Al₂₅ (at. %), which consist of 50 nm wide alternating b.c.c. and B2 phases, were shown to have a/2<111> Burgers vectors after room temperature deformation. The dislocations were found to glide in pairs on both {110} and {112} slip planes and were relatively widely separated in the b.c.c. phase, where the dislocations were uncoupled, and closely spaced in the B2 phase, where the dislocations were connected by an anti‐phase boundary. In this article, we analyze the dislocations in the two ～5 nm‐wide B2 phases in a related two‐phase alloy Fe₃₀Ni₂₀Mn₂₀Al₃₀, with compositions Fe‐23Ni‐21Mn‐24Al and Fe‐39Ni‐12Mn‐34Al, compressed to ～3% strain at a strain rate 5 × 10^?4 s^?1 at 873 K (the lowest temperature at which substantial plastic flow was observed). It is shown that slip occursby the glide of a<100> dislocations. A review of the literature suggests that the differences in the observed slip vector between these B2 phases could be due to the differences in composition, differences in deformation temperature, or possibly both. Microsc. Res. Tech. 76:263–267, 2013. © 2013 Wiley Periodicals, Inc.     

Structure studies of ball-milled ZrCuAl, NiTiZrCu and melt-spun ZrNiTiCuAl alloys

ZrNiTiCu and ZrNiTiCuAl alloys were amorphized using either a melt‐spinning or ball‐milling process in a high‐energy planetary mill. The elemental powders were initially blended to the desired composition (in at.%) of Zr, 65; Cu, 27.5; Al, 7.5 and of Ti, 25; Zr, 17; Cu, 29; Ni, 29, respectively. The composition of alloys was chosen to be the same as for the bulk amorphous ZrCuAl and easy glass‐forming ZrNiTiCu alloys. An almost fully amorphous structure was obtained after 80 h of milling in the case of both compositions. Transmission electron microscopy studies of ball‐milled powders revealed the presence of nano‐crystallites [2–5 nm for ZrCuAl and smaller (1–3 nm) for the ZrTiNiCu alloy]. High‐resolution transmission electron microscopy of melt‐spun ZrNiTiCuAl ribbons provided evidence of the amorphous structure.     

Spatially resolved electron energy-loss studies of metal–ceramic interfaces in transition metal/alumina cermets

Composites consisting of an alumina matrix and 20 vol.% transition metal (Ni or Fe) particles, prepared by hot pressing powder blends, have been studied using spatially resolved transmission electron energy-loss spectroscopy (EELS), and, to a lesser extent, by high-resolution electron microscopy (HREM). Particular attention was paid to the elucidation of the chemical bonding mechanisms at the metal-ceramic interface; EELS spectra from interfacial regions being obtained via a spatial difference technique. From both qualitative and quantitative interpretation of EELS near-edge structures, as well as observed HREM images, the data appear to be consistent with the presence of an Al-terminated alumina at the interface and the formation of direct transition metal – aluminium bonds in Al(O₃M) (M = Ni or Fe) tetrahedral units, possibly as a result of the dissolution and interfacial reprecipitation of Al during processing. These results correlate well with similar model studies on diffusion-bonded Nb/Al₂O₃ interfaces and may, in the light of recent theoretical electronic structure calculations, have implications for the resultant interfacial bond strength in such materials.     

Chemical reactions and morphological stability at the Cu/Al2O3 interface

The microstructures of diffusion‐bonded Cu/(0001)Al₂O₃ bicrystals annealed at 1000 °C at oxygen partial pressures of 0.02 or 32 Pa have been studied with various microscopy techniques ranging from optical microscopy to high‐resolution transmission electron microscopy. The studies revealed that for both oxygen partial pressures a 20–35 nm thick interfacial CuAlO₂ layer formed, which crystallises in the rhombohedral structure. However, the CuAlO₂ layer is not continuous, but interrupted by many pores. In the samples annealed in the higher oxygen partial pressure an additional reaction phase with a needle‐like structure was observed. The needles are several millimetres long, ～10 µm wide and ～1 µm thick. They consist of CuAlO₂ with alternating rhombohedral and hexagonal structures. Solid‐state contact angle measurements were performed to derive values for the work of adhesion. The results show that the adhesion is twice as good for the annealed specimen compared to the as‐bonded sample.     

Effect of titanium carbonitride (Ti(C,N)) decomposition on failure mechanisms in inconel 617 alloy

Titanium Carbonitride (Ti(C,N)) decomposition in Inconel 617 alloy creep‐exposed at 650°C for 574 hours is reported using analytical electron microscopy techniques. Cr‐enriched M₂₃C₆‐type carbides enveloped in fine gamma prime particles thought to be precipitated from the decomposition reaction are observed in the alloy. The morphology of the M₂₃C₆ carbides is irregular and blocky and the particle size up to 5μm, whereas the morphology of gamma prime particles is mostly spherical and up to 30 nm in size. Intergranular carbides are mostly secondary precipitates of the M₂₃C_c type (M predominantly Cr) and these respond to solution heat treatment and precipitate on the grain boundaries as a result of ageing. The ability of intragranular MX to decompose is sensitive to the N content, high N resists decomposition. Decomposed intragranular MX provides an excess source of C which can react locally with Cr to form heat treatable intragranular fine Cr₂₃C₆ precipitates. M₆C can segregate in interdendritic locations during melting which may be the reason for high content of Mo in M₂₃C₆. These precipitates are generally very small and contribute to an additional hardening effect and are the reason for the onset of voiding and cracking along the grain boundaries that ultimately lead to a reduced creep rupture life. Microsc. Res. Tech. 78:336–342, 2015. © 2015 Wiley Periodicals, Inc.     

Synthesis and characterization of permalloy-reinforced Al2O3 nanocomposite powders by mechanical alloying

Intermetallics of Fe and Ni, which are known as permalloy, are under attention due to their excellent magnetic performance. Besides, mechanical properties of the materials can be improved by decreasing crystallite size of FeNi intermetallics or by reinforcing them with hard secondary phases such as Al₂O₃. In this study, FeNi–Al₂O₃ nanocomposite powders with three different compositions were successfully synthesized through mechanical alloying of Fe₂O₃, Ni, and Al powders mixture. Characterization of the samples was accomplished by scanning electron microscopy, transmission electron microscopy (TEM), energy dispersive X-ray spectroscopy and X-ray diffraction. Effects of various parameters such as chemical composition of received materials, milling time, and annealing on the phase evolution, morphology, and microhardness of samples were investigated. It was found that by the addition of Fe as diluent, the required milling time for formation of FeNi intermetallic increased. By increasing milling time, mean crystallite size of FeNi decreased and reach to about 28 nm for FeNi-30 wt% Al₂O₃ nanocomposite powder sample. TEM observations also showed that in situ-formed Al₂O₃ particles, with particle size of about 65 nm, were uniformly dispersed within FeNi matrix.     

Tribological Properties of Al<Subscript>2</Subscript>O<Subscript>3</Subscript>/CuCrZr Composites

Al₂O₃ particles reinforced Cu–Cr–Zr alloy matrix composite was fabricated through a powder metallurgy plus hot extrusion process by using the water atomization Cu–Cr–Zr powder as raw material. The effect of aging treatment on the tribological behavior of the composite was investigated. Experimental results show that tiny coherent precipitated phases were formed in the matrix after proper aging treatment and therefore good combination properties could be obtained. The wear rates of the Al₂O₃/CuCrZr composite and its matrix alloy were obviously influenced by the aging treatment, wherein the best wear resistance was reached at the aging temperature corresponding to the highest Vickers hardness. The major reason was that the depth of plastic deformation in the subsurface region was dramatically decreased due to the improvement of mechanical properties of the matrix, and therefore adhesion induced surface materials loss could be markedly alleviated. By comparing with the SiC20 vol%/Cu composite, it is indicated that the Al₂O₃/CuCrZr composite exhibited much better wear resistance as well as higher electrical conductivity.     

On the detection of MX-precipitates in microalloyed steels using energy-filtering TEM

The properties of microalloyed steels and HSLA steels depend predominantly on the size and composition of nanometre-sized MX precipitates consisting of Nb, Ti, V, C and N. In this work we present a rapid and powerful method for visualizing these small precipitates directly in the steel matrix. This can be achieved by energy-filtering TEM in a 200-kV microscope. We have recorded elemental maps (three-window method) and jump-ratio images (two-window method) using the inner-shell ionization edges; e.g. Fe-M₂₃, Fe-L₂₃, V-M₂₃, V-L₂₃, Nb-N₂₃, Nb-M₄₅, C-K and N-K edges. We have found that precipitates in a steel matrix can be best visualized by recording the jump-ratio image of the matrix element. If the specimens are thinner than 30 nm and high resolution in the images and low detection limits are essential, low energy-loss edges should be used preferentially; in the case of steels the Fe-M₂₃ edge should be used instead of the Fe-L₂₃ edge. If the jump-ratio images are recorded with the Fe-M₂₃ edge, precipitates about 2 nm in diameter may be detected. With thicker samples and if high resolution is not as essential, higher energy-loss edges are preferable, because diffraction effects are less possible. The chemical composition of the larger precipitates (10 nm) could be determined by recording EDX spectra using an STEM probe. We found Nb and V with an atomic ratio of about 3, but we could not detect carbon or nitrogen in the spectra. The Nb and V distribution could be best visualized with a jump-ratio image including both the Nb-N₂₃ and the V-M₂₃ edges. However, it was not possible to image the carbon or nitrogen content of the precipitates unequivocally due to overlapping of the C-K and N-K edges with the Nb-M₄₅ edge.