During the past decade, Fourier transform infrared (FT-IR) microspectroscopy has been used successfully in many studies to differentiate normal and diseased tissue samples obtained from a variety of organs, including colon, cervix, prostate, and breast. IR images were constructed by collecting spectra point-by-point using a mapping stage on a FT-IR microscope equipped with a single-element detector. Five years ago, in collaboration with NIH scientists Dr. Neil Lewis and Dr. Ira Levin, Procter and Gamble researchers developed a technique for performing vibrational spectroscopic imaging microscopy using a liquid-nitrogen-cooled focal-plane array (FPA) detector and a step-scanning FT-IR spectrometer coupled to a refractive microscope. With this configuration, equipped with a new 64-pixel x 64-pixel Mercury-Cadmium-Telluride (MCT) FPA detector, we are able to image an 800-μm x 800-μm area of a specimen without moving the sample.

For all of these IR imaging studies, the tissue samples were prepared in a manner that was special and different from that commonly used in a pathology laboratory, i.e.,the histopathology sample was not identical to the spectroscopic sample.

The field of infrared imaging for analytical chemical analysis has developed over the last few years into a technique that has clearly demonstrated its utility in a variety of applications(l-6). This imaging system is based upon the coupling of an infrared array detector to an infrared spectrometer and microscope, where the use of an array detector allows both spatial and spectral information to be generated simultaneously. Because of the high spatial content and data acquisition speed, researchers have been able to apply this technique to samples and experiments where traditional infrared microscopy has not proven successful.

Previously, in order to obtain an infrared image, the researcher was required to undertake a step and collect mapping experiment. In the mapping experiment, an aperture is set to obtain the desired spatial resolution at the sample. If high spatial resolution is required, most of the energy is discarded and long data collection times are needed to obtain a reasonable signal-to-noise

Castaing’s first approximation assumes that the characteristic radiation emitted by each element in a compound is unaffected by the presence of the other elements, and that the intensity of the radiation emitted by each element in the compound is proportional to its mass concentration. From Castaing’s first approximation based on mass fractions, the initial estimate of the concentration of an element in a compound is

where C and C° are mass fractions in the sample and standard, and I and I° are x-ray emission intensities of a specific transition measured in the sample and the standard. Surprisingly, the mass fraction approach works well in practice, even though it lacks any obvious physical basis. This can be understood by considering that for typical energies involved in EPMA (50 keV or less), the size of the nucleus is several orders of magnitude smaller than the average incident beam electron wavelength.

Both precision and accuracy are central to quantitative microanalysis. While precision may be evaluated from x-ray counting statistics and replicate measurement, the determination of analytical accuracy requires well characterized standards of which there are few that span a wide range of compositions in binary and ternary systems. The accuracy of silicate mineral analysis has been previously studied via measurement of α factors at multiple accelerating potential and the subsequent evaluation of correction algorithms and mass absorption coefficient (mac) data sets. This approach has been extended in this study to the In2O3-Ga2O3 and HgTe-CdTe systems.

Single crystals of ln2O3, Ga2O3, and an InGa-oxide of unknown composition were used to evaluate accuracy in the In2O3-Ga2O3 binary, using the GaKα, GaLα, and InLα x-ray lines, with WDS measurements performed at 15, 20, and 25KV relative to the ln2O3 and Ga2O3 standards (see Table I). The Ga Kα line exhibits minimal absorption, has no fluorescence correction in this system and is not critically dependent on the correction algorithm or mac data set used.

The ϕ(ρz) method is commonly used in quantitative X-ray microanalysis to calculate the proper correction for absorption and atomic number. Several models are used to reproduce generated X-ray intensity as a function of mass thickness. Those models rely on equations which has been adjusted to fit experimental measurement of the ϕ(ρz) curves. The classical tracer method for the experimental measurement of ϕ(ρz) curves proposed by Castaing (1951). The limitations of this method are related to the fabrication steps of the sample. The successive layers must be deposited in thin uniform layers of the same mass thickness and must be the same composition as the substrate.

The method described here, first proposed by J. Cazaux, uses the effect of self-absorption in a bulk sample to determine the generated intensity. The basic idea is to produce a set of equations for several absorption conditions while maintaining a constant geometry for the x-ray generation.

Studies of x-ray emission from thin films on substrates using an electron probe microanalyzer (EPMA) provide useful information on the characteristics of x-ray generation by electron beams. In this study, EPMA measurements of multilayered samples were performed in order to test and improve analytical and numerical models used for quantitative EPMA. These models provide relatively accurate results for samples consisting of layers with similar average atomic numbers, because of their similar properties regarding electron transport and x-ray generation. On the contrary, these models find difficulties to describe the process when the various layers have very different atomic numbers. In a previous work, we studied the surface ionization of thin copper films of various thicknesses deposited on substrates with very different atomic numbers. In the present communication, the study is extended to the case of multilayered specimens.

The studied specimens consisted of thin copper films deposited on a carbon layer which, in turn, was placed on a variety of single-element substrates, ranging from Be to Bi.

Monte Carlo simulation of electron scattering in solids has proven valuable to electron microscopists for many years. The electron trajectories, x-ray generation volumes, and scattered electron signals produced by these simulations are used in quantitative x-ray microanalysis, image interpretation, experimental design, and hypothesis testing. Unfortunately, these simulations are often computationally expensive, especially when used to simulate an image or survey a multidimensional region of parameter space.

Here we present techniques for performing Monte Carlo simulations in parallel on a cluster of existing desktop computers. The simulation of multiple, independent electron trajectories in a sample and the collateral calculation of detected x-ray and electron signals falls into a class of computational problems termed “embarrassingly parallel”, since no information needs to be exchanged between parallel threads of execution during the calculation. Such problems are ideally suited to parallel multicomputers, where a manager process distributes the computational burden over a large number of nodes.

SeSAME, a Monte Carlo simulation program which can handle totally arbitrarily shaped twodimensional geometries, has been extended for the modeling of whole x-ray spectra in absolute units. The so called single scattering model has been used for the simulations presented here. Elastic scattering processes are modeled using a partial wave approach according to a relativistic expression derived by Mott. For the calculation of the electron energy loss the stopping power formula from Joy and Luo has been applied.

Two different double differential cross sections for continuous x-ray production have been implemented into SeSAME. The first one is an analytical model derived by Kirkpatrick and Wiedmann and is based on a non-relativistic theory developed by Sommerfield and Elwert. The second one uses tabulated values for the shape function for atomic-field Bremsstrahlung and scaled Bremsstrahlung cross sections differential only in photon energy according to Kissel at al.

Typical modern histories of the discovery of the chemical elements report that element 43 was discovered in Italy in 1937 by Perrier and Segre. According to these accounts, the discovery was made by chemical separation in a sample of deuteron-bombarded molybdenum, supplied by E. Lawrence, and thus represent the first discovery of an element produced artificially and is the reason for the name “technetium” (Tc) chosen by the investigators. Some accounts might note in passing that the existence of element 43 was predicted in the early part of the 20th century on the basis of the periodic table, and that the element was reported as discovered in natural samples by x-ray emission spectroscopy in 1925 and called “masurium” (Ma). These reports, however, invariably call this discovery erroneous and sometimes refer to the investigators (typically left unnamed) as “deluded”.

The original reported discovery of “masurium” was made in an exhaustive investigation of some thousand mineral separates of Nb-Mo-U-rich phases involving early elegant use of wavelengthdispersive x-ray microanalysis.

The analytical expressions used in ZAF and ϕ(ρz) calculations give single values for the composition of each element for a single set of intensity measurements from samples and standards. Confidence intervals in composition are established by looking at the variability of repeated measurements. They are usually attributed to x-ray counting statistics or experimental reproducibility factors such as sample repositioning. Uncertainty in the equations themselves or the parameters that go into them are rarely considered. The derivations of ZAF and ϕ(ρz) equations are primarily based on the case where flat single-phase regions, relative to the x-ray excitation volume, are examined using normal electron beam incidence. Use of these equations has been extended to non-normal electron beam incidence as well as the quantitative analysis of layered structures, but usually with less theoretical justification. Finally, special experimental cases including porous structures, rough surfaces, vertical interfaces and small particles are very difficult or impossible to model by the single application of a set of analytical equations to convert measured x-ray intensities to elemental composition.

For quantitative thin-film microanalysis using X-ray energy dispersive spectrometry (XEDS) in analytical electron microscopy (AEM) the Cliff-Lorimer ratio technique is widely applied. One of the critical steps in the ratio technique is to determine the aiff-Lorimer k factor relating the characteristic X-ray intensities to the compositions. The k factor can be determined experimentally with the relative error of ˜ ± 1%. So, the most accurate quantification can be performed using the experimental k factors. To determine the k factors experimentally, standard thin-filrns with known composition are required. Unfortunately, this is not always possible because such the standards may not be available. Even if all standards are available, many k factors should be determined in multi-component system. This is time consuming. These limitations in k-factor determination make thin-film quantification harder. This paper presents a new quantitative procedure for thin specimens to overcome these limitations.

For a number of years we have used in our laboratory a simple extension to light elements of the classical ClifF-Lorimer approach using stoichiometric compound standards containing pairs of heavier elements to determine independently the EDS detector sensitivity contribution to quantitative analysis in the AEM. It is now also extended to direct light element LVEDS analysis in the SEM at low voltage (<5kV). This approach may be more practical, accessible and easily reproducible for routine use than the elegant custom thin film fabrications reported by Egerton, but the latter method has other substantial advantages, including well defined thicknesses, in the most sophisticated applications. As the types of advanced materials have expanded to include ceramics and polymers the requirements for analysis have also changed. In many cases the light elements are now the majority species in advanced materials and the corrections based on older standards are huge; and therefore inevitably rather limited in accuracy.

The II-VI semiconductor mercury cadmium telluride, composed of 80 mole percent HgTe and 20 mole percent CdTe, has been grown on Earth and aboard two NASA Space Shuttle flights in order to study the solidification of a solid solution semiconductor having a wide separation between liquidus and solidus. Segregation during solidification, high volatility of mercury, and strain fields associated with large temperature gradients make the growth of homogeneous, high-quality, bulk crystals extremely difficult. The traditional growth methods yield crystals with axial and radial compositional variations and secondary precipitates of tellurium, probably caused by gravityinduced, thermosolutally-driven flows just ahead of the growth interface. The purpose of growing these crystals in microgravity is to reduce gravity-induced convection. This should be advantageous for minimizing the density of crystal defects while also minimizing compositional variation transverse to the crystal growth direction.

The ground based experiments consisted of growing crystals in several different configurations of heat pipe furnaces and NASA’s Advanced Automated Directional Solidification Furnace (AADSF).

The electrical properties of SrTiO3 (strontium titanate) ceramics are strongly influenced or even dictated by grain boundary segregation of charged point defects, such as dopant atoms, impurities, vacancies, or self-interstitials. The atomistic structure of the grain boundaries, their energy, and the segregation of point defects mutually depend on each other. Grain boundary segregation of charged point defects induces the formation of space charge layers in the adjoining crystals. In order to investigate the relation between grain boundary structure and composition, grain boundaries in Fedoped SrTiO3 bicrystals and in SrTiO3 ceramics were studied by HRTEM and by AEM with subnanometer resolution.

Quantitative HRTEM served to investigate the atomistic structure of Σ=3, (111) grain boundaries in Fe-doped SrTiO3 bicrystals with a doping level of Fe/Ti= 0.04at% (Fig. 1). Analysis of the translation state revealed that the Σ=3, (111) grain boundary has an excess volume: normal to the boundary plane, the spacing between the two crystals exceeds what one would expect from a coincidence site lattice model by (0.06 ±0.01 )nm.

The small probe size and high spatial resolution of TEM offer unique capabilities for studying interfaces of materials. Compared with neutron and x-ray diffraction, TEM is often considered nonquantitative. In HREM, especially, atomic images can alter significantly with imaging parameters that are often difficult to determine experimentally. The weakness of the HREM can be circumvented by a multiprobe approach, i.e., the combined use of a wide range of TEM techniques, such as B-F/D-F imaging, nano-probe EDS and EELS, electron diffraction, and strain-field characterization. Quantitative analysis, based on the calculations and experimental TEM data, is often necessary to correctly derive the crystal-, chemical-, and electronic-structure of an interface.

We review our multiprobe approach in studying interfaces in complex systems of superconductors, starting with the simplest interface, twin boundary, in YBa2Cu3O7-δ. Although HREM suggested that there are two types of twin boundaries, with and without a ˜d110 lattice translation at the boundaries, the structural characteristics of the interfaces were determined by analyses of the interfacial δ-α fringe and the associated diffuse scattering, while EELS revealed their association with the local oxygen/hole concentration.

The interaction of a liquid with a solid substrate has been the subject of intense research since the early 1800s. The subject is relevant and very important even today from a basic science as well as technological standpoint. Liquid-phase-sintered oxide ceramics often contain a siliceous glassy layer in the grain boundaries which affect the properties of the ceramic. The wetting of the grain boundaries by the liquid has a strong dependence on the crystallography. While some boundaries are preferably wet, other special boundaries appear to be completely ‘dry’. The presence of the liquid film also affects the faceting behavior of the boundary plane. While several studies have been carried out on these intergranular glass films, many of the basic issues still remain unclear. The development of modern microscopy techniques facilitate the characterization of structure and chemistry at a high spatial resolution. This paper deals with the microstructural examination of two of the important issues involved with these intergranular films, i.e., faceting of grain boundaries in the presence of a glassy phase and the behavior of interfaces between two glassy phases of different compositions.

A large number of bulk and thin-film electroceramic systems contain electrically active interfaces which dictate the various useful electronic properties of these devices. The electrical activity of these interfaces stems from a complex interplay among various interfacial attribute but often involves formation of some form of electrostatic potential at the interfaces which is modified under applied bias of current and/or voltage.

Figure la schematically shows charge distribution at a model grain boundary (GB), while Figure 1(b) shows its corresponding potential distribution. As shown in Figure 1(c), the energy band structure bends opposite to this built-in potential, causing a downward shift at the grain boundary . The difficulty with evaluating the Schottky barrier model, which is often invoked to explain GB electrical activity, is that the charge density distribution and therefore the band bending is expected to dynamically alter as bias is applied across the grain boundary. This variation adds another level of complexity to theoretical descriptions of the barrier behavior

Some properties of transition metal oxides are related to the presence of elements with mixed valences. In electron energy-loss spectroscopy (EELS), the L or M ionization edges of transition-metal, rare-earth and actinide elements usually display sharp threshold peaks known as white-lines. EELS experiments have shown that a change in cation valence state introduces a significant change in the White-line intensity ratio [1]. With the use of valence state information provided by the white lines, an experimental approach is demonstrated here to map the valence state distributions of Mn and Co using an energy-filtered transmission electron microscope (TEM). A spatial resolution of ˜ 2 nm has been achieved. This technique should be particularly useful in studying valence states of cations in magnetic oxides.

To map the distribution of ionization states, an energy window of ˜ 10 eV in width is required to isolate the L3 from L2 white lines (Figure 1).

Bismuth titanate (Bi4Ti3O12) belongs to a family of layered perovskites (Aurivillius phases) that are important for nonvolatile ferroelectric memory applications. Recently epitaxial (001) Bi4Ti3O12 thin films have been grown on (001) SrTiO3 by reactive molecular beam epitaxy (MBE) in an adsorptioncontrolled growth regime. Bi4Ti3O12 has a monoclinic layered perovskite structure at room temperature which consists of alternating Bi2O2 and Bi4Ti3O10 perovskite layers (see the upper part of Fig. 1(b)). Due to the significant difference in crystal structure between film and substrate, the surface termination of the SrTiO3 substrate and the way in which growth is initiated might be expected to have a significant influence on the interfacial structure. In this work, we present high-resolution transmission electron microscope (HRTEM) studies of the Bi4Ti3O12 / SrTiO3 interfacial structure for Bi4Ti3O12 films grown on SrTiO3 substrates terminated with either the TiO2- or SrO-layers.

Bi4Ti3O12 thin films were grown on SrO- or TiO2-layer terminated homoepitaxial SrTiO3 buffer layers grown on SrTiO3 (001) substrates by reactive MBE.

The ability to obtain chemical information with (near) atomic resolution has recently become possible by a combined approach of Z-contrast imaging with electron energy-loss spectroscopy (EELS) in scanning transmission electron microscopes. This method is particularly interesting for the characterization of structure-property relationships in novel multicomponent oxides, which possess added functionality due to their high nonstoichiometry.

In this paper we demonstrate the capabilities of this method in analyzing the microstructural mechanisms of accommodation of non-stoichiometry, using two example systems: (Ba,Sr)TiO3thin films for DRAM applications, grown by MOCVD with different amounts of excess titanium, and an oxygen-deficient perovskite ceramic, SrCoOx. The experiments were performed in a JEOL JEM 201 OF field emission transmission electron microscope, operating at 200 kV, equipped with an annular dark-field detector, scanning unit and a post-column imaging filter (Gatan GIF 200). This microscope has been demonstrated to achieve probe sizes of under 1.5 Å .