Trace impurities and dopants in sintered ceramics, often at levels below 100 parts per million, can significantly affect the performance of the materials. In alumina (A12O3), it is believed that calcium and silicon, which are the main background impurities, cause abnormal grain growth. The addition of magnesium to alumina produces several beneficial results, including enhanced densification and improved corrosion resistance of the ceramic. For example, LUCALUX®, which is a dense, transparent ceramic, is made by adding a small amount of MgO (<0.25%) to the alumina. Primarily because they are difficult to detect and image, it has not been easy to determine experimentally the mechanisms by which trace constituents influence the sintering of ceramics. Imaging secondary ion mass spectrometry (SIMS) is a valuable analytical technique for the submicrometer characterization of ceramics. Relevant to this study, magnesium and isotopes of silicon, which have largely eluded detection by traditional techniques, can be measured with SIMS.

Metal matrix composite (MMC) materials are promising materials because of their higher specific mechanical properties (e.g. strength-to-weight ratio) relative to conventional metallic alloys. However, during the processing of these materials, the development of an interfacial reaction zone between the reinforcing material (generally a ceramic) and the metallic matrix may occur, which typically degrades the mechanical properties of these materials if not controlled. Since these interfaces are generally quite small (from 5 to 500 nm), Transmission Electron Microscopy (TEM) is the technique of choice for characterization because of its outstanding spatial resolution. However, specimen preparation for TEM (e.g. ion milling) is quite difficult for MMC materials, due to the combination of two completely different material types, specifically a hard and brittle reinforcement phase incorporated in a relatively soft metallic matrix. Also, since there is only a small amount of the material which is transparent to electrons after specimen preparation, TEM as a materials characterization technique suffers from a lack of statistical robustness, which is a serious drawback.

To characterize materials with a spatial resolution close to that of TEM, low voltage field emission scanning electron microscopy is an option because of the small interaction volume of low energy electrons, as well as the small probe size of these microscopes (typically < 5 nm).

The atom probe field ion microscope is a powerful technique for the quantification of solute segregation at internal interfaces due to its ultrahigh spatial resolution. The segregation behavior for all elements is determined by collecting the atoms that originate in a cylinder centered on the boundary region. The method of analysis is dictated by the orientation of the boundary in the field ion specimen. The general case and the two special orientations are shown schematically in Fig. 1. The special case where the unit normal to the interface plane is at 90° to the unit vector parallel to the cylinder of analysis, ϕ = 90°, permits a large number of atoms to be collected but the spatial resolution is defined by the effective diameter of the probe aperture and is typically chosen to be between 1 to 5 nm. The special case where ϕ = 0° provides the highest spatial resolution but only a limited number of atoms are collected from the boundary in each cylinder of analysis.

Because of their high coercivity, cobalt alloy thin films are among the most popular materials used for ultra-high density longitudinal magnetic recording media. The recording and magnetic properties of the materials are related to their microstructure; in particular, depletion of Co in a grain boundary phase, and physical separation of the grains act to increase coercivity and thus to produce low noise media. We are studying a new alloy system comprising 18 nm thick Co-Cr-P-Pt films (Mr.t ≈ 0.9 memu/cm2), prepared by DC sputtering. A coercivity of 2600 Oe or higher was obtained in these films even when they were deposited without heating the substrate or applying a bias voltage. The effects of P and Pt addition were characterized by high-resolution TEM coupled with energy dispersive x-ray spectroscopy (EDS) and electron energy loss spectroscopy (EELS). A Hitachi HF-2000 field emission TEM was used to image both low P (≈ 6 at. %) and high P (≈ 12 at. %) samples, and to provide a 1 nm beam for high spatial resolution EDS and EELS.

For more than 20 years interfacial segregation has been studied by analytical electron microscopy (AEM). Two recent commercial developments should allow more widespread application of 1 nm resolution AEM studies of interfacial segregation. The first is a system for integrated acquisition that in its most sophisticated form allows spectrum imaging [a full spectrum recorded at every pixel of an image (e.g. STEM)] for simultaneous multiple spectroscopies (e.g., EDS and PEELS). A field emission gun (FEG) is necessary for sufficient sensitivity at high spatial resolution. The second commercial development has been imaging energy filters that allow the production of elemental maps from a series of energy-filtered TEM (EFTEM) images at inner shell ionization edges. Resolutions of ∼1 nm can be achieved without an FEG, but analysis is limited to the electron energy-loss signal. Although complex microstructures may require full two-dimensional mapping, for many planar interfaces a one-dimensional profile of composition or chemistry normal to the interface plane is sufficient.

Impurity segregation at grain boundaries (GB) can be detected by EDS in a dedicated STEM. Quantification of the segregation requires not only quantification of the spectra but also consideration of the geometry of the experiment. Our aim was to obtain a value which characterises only the segregation of the impurity and is independent of experimental parameters. The problem is that the specimen composition at the GB is extremely inhomogeneous on an atomic scale in the case of Bi segregation at GBs in Cu. The analysed volume which is defined by the irradiated area and the beam broadening of the primary electron beam inside the specimen contains the interfacial plane as well as neighbouring bulk Cu. One approach is to put the focussed primary electron beam on the interface which is aligned edge on and acquire a spectrum. Both the primary beam diameter and the beam broadening inside the specimen have to be known.

A typical microprocessor chip has several million devices patterned into it and, as fabrication techniques improve, device dimensions continue to shrink. This miniaturization results in a corresponding decrease in the dimensions of the interconnects between devices, causing an increased susceptibility to failure by current-induced mass transport, or electromigration (EM). It is well known that the addition of a small amount of Cu to Al interconnects increases their median time to failure profoundly. Since EM in these films occurs almost exclusively at grain boundaries, the simplest explanation for the role of Cu is that it segregates to the grain boundaries and reduces their diffusivity. Analytical electron microscopy studies have shown that Cu segregates to grain boundaries in Al under the proper conditions, and its concentration was measured using X-ray energy dispersive spectroscopy (EDS), but only a small number of boundaries were measured.In order to determine whether the distribution of Cu affects the resistance of a material to EM damage, it is first necessary to characterize the Cu content of the material as a whole, which requires the analysis of large numbers of boundaries rather than just a few.

Equilibrium segregation to grain boundaries has been studied for many years, but only recently have the techniques to quantify such segregation at high spatial resolution become widely available. These include, for example, the Auger surface analysis of fractured samples, or the study of boundaries edge-on in a field emission gun STEM equipped with energy dispersive X-ray analysis or electron energy loss spectroscopy. The key question, however, is whether these different techniques yield the same results when applied to a specific set of samples, and this paper reviews some of the successes and problems associated with such studies.

For STEM analysis of grain boundary segregation, several instruments now routinely give probe sizes of ∼1nm. Hence line profiles across grain boundaries at this level of resolution are quite feasible. However a larger probe with greater probe current and better X-ray counting statistics is often more useful, especially if sample drift is a potential problem.

A quantitative comparison has been made of intergranular segregation in four 9%-Cr ferritic steel alloys, S1-S4 doped, respectively, with 560, 300, 120 and 25ppm P, by both Auger Electron Spectroscopy (AES) and EDX in the analytical electron microscope (FEG-STEM). The strengths, weaknesses, relative accuracies and precision of the two methods were investigated in view of the different equipment, sample preparation, characteristic signals and unrelated assumptions in quantifying the results. No such investigation of ferritic steels has previously been reported, even though they are widely used in engineering components and can be susceptible to embrittlement. Comparative studies have been reported for austenitic alloys.

The steels were solution-treated, tempered at 750°C and aged at 500°C for 100 hours. In-situ impact at low temperature in the VG MA500 AES apparatus caused mixed cleavage and intergranular fracture, the latter increasing with P content. Spectra from up to 80 individual intergranular facets were recorded from each alloy using a 5nA-10KeV beam from an LaB6 electron source under conditions which had been optimised by prior investigation.

The fact that there is a region at the epoxy/adherend interface known as the interphase whose chemistry and structure are different from that of bulk epoxy is well established (1-2). The specific nature of the interphase depends on the resin and curing agent, the adherend surface-chemical properties, and the adherend sizing treatment if any, among other possible variables. The interphase is critically important, since it is responsible for stress transfer between the matrix and reinforcing fibers. While it has been studied by a variety of analytical methods (eg. 3-5), none have had sufficient spatial or spectral resolution to establish the physical extent of the interphase or its compositional deviation from bulk epoxy. This research exploits high-spatial-resolution analysis in a field-emission STEM to study the interphase in epoxy/adherend systems using high-angle electron scattering (HAADF-STEM) and energy-loss spectroscopy (PEELS).

Thin specimens were cut by ultramicrotomy using a diamond knife from single-fiber tensile bars consisting of anodized aluminum wire embedded in an epoxy matrix (Fig 1.)

Radiation-induced segregation (RIS) is the spatial redistribution of elements at defect sinks such as grain boundaries and free surfaces during irradiation. This phenomenon has been studied in a wide variety of alloys and has been linked to irradiation-assisted stress corrosion cracking (IASCC) of nuclear reactor core components. Therefore, accurate determination of the grain boundary composition is important in understanding its effects on environmental cracking. Radiation-induced segregation profiles are routinely measured by scanning-transmission electron microscopy using energy-dispersive X-ray spectroscopy (STEM-EDS) and Auger electron spectroscopy (AES). Because of the narrow width of the segregation profile (typically less than 10 nm full width at half-maximum), the accuracy of grain boundary concentration measurements using STEM/EDS depends on the characteristics of the analyzing instrument, specifically, the excited volume in which x-rays are generated. This excited volume is determined by both electron beam diameter and the primary electron beam energy. Increasing the primary beam energy in STEM/EDS produces greater measured grain boundary segregation, as the reduced electron beam broadening a smaller excited volume.

The corrosion behavior of Al-Cu-Mg alloys, specifically 2024 alloy (nominally, in weight %, 4.4 Cu, 1.5 Mg, 0.6 Mn), is thought to depend on heterogeneous Cu and Mg distribution through the existence of segregation-dependent local electrochemical cells at the corrosion interface. Few nanospectroscopy measurements of segregation have been made for this or similar alloys. These alloys are precipitation hardenable. The primary precipitating phases are S and the well known Θ(CuAl2) and their metastable intermediates. TEM analysis of aged alloys in this subgroup showed that the orthorhombic S phase (a=4.0Å, b=9.25Å, c=7.15Å) occurred as a thin plate type variant, called S´, within matrix grains and as larger monolithic particles on grain boundaries. Intragranular pricipitate particle densities were heterogeneous particularly near grain boundaries, indicating that strong segregation was present that would result in local electrochemical cells where grain boundaries and large precipitates intersected the alloy surface.

HRTEM and nanospectroscopy are used to analyze the structure and chemistry of heterophase interfaces and grain boundaries.

Y or Ca stabilized tetragonal ZrO2 (TZP) exhibits superplasticity at high temperature, and can also be used as solid electrolytes. Those properties are dictated by structure and chemistry of grain boundaries, which can be controlled by segregation of impurities or additives. The grain boundaries were found either covered by amorphous films or free of the film. Co-segragation of additives and stabilizers has also been observed. To fully understand the correlation between segregation and grain boundary structure, a dedicated STEM (VG HB601) capable of EDX/EELS analysis and phase/Z-contrast imaging is employed to study 3Y-TZP doped with 0.3 and 0.9 mol% SiO2.

Although Y-L lines arc dominated by overlapping Zr-L lines in EDX, Y excess at grain boundaries can still be measured by “spatial difference” which removes Zr signal with a spectrum from the bulk. The co-segregation of Si and Y is also observed (Fig. 1) at many boundaries. Their average excesses arc 5±2 nm−2and 25±10 run−2 respectively, close to 1 monolayer each of SiO2 and Y2O3.

Alloy X-750 is a γ´-strengthened Ni-base alloy that is often used in nuclear power systems for its excellent corrosion resistance and good mechanical properties. However, radiation-induced segregation (RIS) and irradiation-assisted stress corrosion cracking (IASCC) can occur under neutron irradiation. The present study examines the microstructure and composition profiles in a heat of Alloy X-750 before and after neutron irradiation.

The material was in the HTH commercial heat treatment (lh at 1121°C + air cooling + 20h at 704°C). Specimens were prepared from both control material and material irradiated at 360°C to a fluence of 2.3 × 1020 n/cm2 (E > 1 MeV), which was estimated to produce -0.4 displacements per atom. High spatial resolution analytical electron microscopy was performed in a Philips EM400T/FEG operated in the STEM mode with a probe diameter of ∼1.4 nm (FWHM). Composition profiles were acquired in two different ways: all energy dispersive X-ray spectrometry (EDS) and some parallel-detection electron energy-loss spectrometry (PEELS) profiles were acquired as individual spot analyses, whereas the majority of the PEELS profiles were acquired as spectrum lines with ∼10 s acquisition times per point. The quoted compositions are as measured (wt%); no deconvolution for the effect of the excited volume was performed.

The role of various alloying elements on the kinetics of austenite decomposition in steels is well documented. One of the least understood aspects is the ability of some elements like Mo to drastically reduce the kinetics of ferrite (α) growth in austenite (γ).Within specific ranges of transformation temperatures and alloy compositions, the transformation of austenite can cease entirely for extended periods of time (transformation stasis). The phenomenon is quite pronounced in Fe-C-Mo alloys and is clearly evidenced by a horizontal region in the plot of the fraction transformed vs. isothermal reaction time. Based on microstructural and kinetic evidence, the occurrence of the transformation stasis phenomenon has been ascribed to a "solute drag like effect" caused by the segregation of Mo atoms to α : γ boundaries. The proposal that interface chemistry alters the growth kinetics of ferrite is quite difficult to verify as the segregation is expected to be confined to a boundary region less than a nanometer in width.

Magnesium is an important and common solid solution strengthening agent for aluminum alloys. It has been known for many years that Mg segregation to grain boundaries (GBs) and other structural inhomogeneities can lead to precipitation of the ² (Al3Mg2) phase during low-temperature aging. This phase is anodic with respect to the matrix so its presence degrades the stress corrosion cracking (SCC) resistance of the alloy. The details of P-phase precipitation have been studied in detail in binary Al-Mg alloys [3-5] but there is some indication that ternary elements can alter the precipitation kinetics and spatial distribution which in turn will affect SCC. Therefore we are studying the segregation and precipitation of Mg in a commercially important Al-Mg alloy which contains Mn and Cr as grain refining particles.

Specimens were prepared from a commercial 5083 heat with the composition 4.47% Mg, 0.61% Mn, 0.08% Cr, 0.09% Si, 0.21% Fe, 0.05% Cu, bal. Al (all compositions in weight percent). Following cold rolling they were given a recrystallization anneal of 10 min. @ 500°C and air cooled.

A thorough and complete assessment of microstructure of materials requires a wide variety of analytical techniques, which should be sensitive to at all length scales - from mms to atomic scale. While there have been rapid advances in imaging and spectroscopy techniques for microstructural analysis - at all length scales, techniques for crystallographic analysis of microstructure have primarily relied on bulk x-ray/neutron diffraction and TEM. X-ray and neutron diffraction techniques, though very powerful, are primarily bulk techniques and extraction of local crystallography is formidable if not impossible. On the other hand, TEM diffraction techniques provide precise crystallographic information, but at a much smaller length-scale and suffer from poor statistics and tedious specimen preparation procedures. With the advent of commercially available electron backscattered diffraction (EBSD) and orientational imaging (OIM) systems for SEM, and sophisticated pattern recognition procedures, it is now possible to bridge the length-scale gap between bulk and TEM diffraction techniques.

Traditionally, the structure and orientation relationship of individual catalyst particles on various oxide substrates have been studied by transmission electron microscopy. However, the combination of high resolution scanning electron microscopes (HRSEM) equipped with Schottky field emission sources with CCD cameras for recording electron backscatter diffraction (EBSD) paterns, it is now possible to obtain both morpholgy and orientation of individual particles with high spatial resolution. In this paper, we present results on the application of combined EBSD with HRSEM to determine the epitaxial orientation relationship of 80 nm Au particles on TiO2 (110). An evaluation of the spatial resolution limit of EBSD using Monte Carlo simulation of backscattered electron trajectories is also presented.

The TiO2 (110) single crystal surfaces used in this study were prepared in UHV using surface science tools followed by in-situ metallization. After deposition of 15 nm Au at 300K followed by annealing at 800K, the samples were transferred in air to the Field Emission Scanning Electron microscope (LEO 982 Gemini) for high resolution imaging.

Image analysis techniques coupled with crystallography computer codes have been used to index electron backscatter diffraction patterns (EBSPs). The ability to automatically obtain the crystallographic orientation from EBSPs coupled with computer control of the electron beam (or stage) in a scanning electron microscope (SEM) provides a much more complete description of the spatial distribution of crystallographic orientation in polycrystalline materials than has been previously attainable using conventional metallography techniques. Orientation data obtained using this technique can be used to form images reflecting the spatial arrangement of crystallographic orientation in a microstructure. Such images enable the topological features of a microstructure to be linked with the orientation characteristics. The formation of these images, as well as the data collection technique, is sometimes termed Orientation Imaging Microscopy (OIM). The utility of this technique for exploring the property/structure relationship in polycrystalline material has been demonstrated by numerous researchers. However, as yet, this technique has almost exclusively been applied to single phase materials.

Crystal orientation mapping (COM), which is also referred to as orientation imaging microscopy (OIM), is a powerful tool which opens up enormous possibilities for investigation of materials. The principle of COM is that the microstructure is displayed or mapped according to the orientation of sampled volumes of crystal. These data are obtained in the scanning electron microscope by moving either the electron beam or the specimen stage through predetermined steps and collecting an electron back-scatter diffraction (EBSD) pattern. Typically, a null orientation is represented by a black pixel and colours are used to depict orientations, thus allowing discrete orientation changes such as grain boundaries to be plotted directly in a map format. This is exemplified in figure 1 which shows an orientation map generated from pure aluminium which has undergone 5% cold rolling. The diffiiseness of EBSD patterns further permits strain changes to be mapped.