We have developed a state-of-the-art x-ray microprobe with many unprecedented capabilities for a broad range of scientific applications. The combination of the high brilliance third-generation synchrotron radiation sources and the development of high performance x-ray focusing optics have led to the construction of a x-ray microprobe at Argonne National Laboratory with a FWHM focal size less than 0.25 μm and photon flux exceeding 5×l010 photons/cm2/sec/0.01% BW. The microprobe is very flexible and versatile and can be used simultaneously to perform x-ray diffraction microscopy, fluorescence microscopy, spectromicroscopy, and 3D tomography with submicron spatial resolution. It has been used for a broad range of applications since its commissioning in December of 1996. For example, it was used in the study of the symbiotic relation between plant root and fungi, platinum-based anticancer agents, integrated laser/modulator microdevices, electromigration in Al wires in semiconductor devices, and submicron spatial resolution threedimensional tomography of a single cell. The capabilities of the microprobe are described below, and applications using these capabilities will be presented.

X-ray absorption spectroscopy (XANES and EXAFS) applied with x-ray microprobe instrumentation (figures 1) can be used for studying the electronic structure of specific elements in complex materials in a spatially-resolved manner (figure 2). Such techniques are valuable in a wide range of studies including hydrothermal fluid processes, migration and encapsulation of toxic and radioactive wastes, and redox evolution of solar system bodies.

One of the major technical challenges in this work is the production of high flux microbeams from high power, hard x-ray synchrotron sources. Some of the microbeam technologies under development include tapered glass capillaries, zone plates, and elliptical mirrors. Each approach has advantages and disadvantages and the optimum microfocusing device depends on the particular experiment.

One of the most versatile of these devices is the dynamically bent, elliptical mirror, especially when a pair of mirrors are arranged in a Kirkpatrick-Baez (KB) geometry to provide two-dimensional focusing. This versatility derives mainly from four attributes.

Coral aragonite encodes a rich, complex record of ambient environmental conditions (including temperature, salinity, pollutants) during skeletogenesis. Annual growth banding in appropriate species provides an internal calendar, and the rapidity of skeletal growth permits temporal sampling to at least a sub-monthly basis. The longevity of reef coral colonies can provide continuous centuries-long environmental records. The presence of well-dated Pleistocene coral reefs at sites across the world ocean extends this record into the more distant past. Thus corals are a remarkable resource for the investigation of environmental conditions over perhaps 7 orders of magnitude of time.

The past few years have seen considerable use of Sr/Ca ratios in scleractinian corals to reconstruct sea surface temperatures (SSTs) to document global change. The detailed correlation between Sr/Ca ratios in living corals and measured ambient water temperature is often remarkable, as is the match in older corals between Sr/Ca ratios and such other SST temperature proxies as 18O /16O ratios.

X-ray microfluorescence (XRMF) is spatially resolved x-ray fluorescence (XRF) analysis of elemental composition. Formerly available only at synchrotron sources, recent development of commercial instruments has brought a tamer version of this technique into the size and price range of the academic laboratory. Our Kevex Omicron system features a primary beam diameter as small as 30 μm, with an energy dispersive spectrometer collecting the x-ray spectrum from Na to U. The micro-positioning stage and optical microscope permit selective spatially resolved analysis on the surfaces of objects as large as a desktop dictionary. The large sample stage, functional spatial resolution, and non-destructive analysis present significant opportunities for elemental analysis of a variety of archaeological materials. In particular, objects which for aesthetic, religious, regulatory, or economic reasons cannot be subjected to any alteration in sample preparation may still be analyzed by XRMF. Our laboratory has been conducting a number of investigations which illustrate the potential impact of XRMF technology on archaeometry.

We use the Scanning Transmission X-ray Microscope (STXM) at the National Synchrotron Light Source (NSLS) (1) to acquire images and spectra of polymers. To interpret data correctly, the effects X-ray radiation has on polymers must be understood. We have therefore started to characterize radiation damage in a variety of carbonyl containing polymers in two ways. First, we want to ascertain the critical dose for mass loss and the critical dose for the carbonyl in a variety of polymers and relate the critical dose to the polymer structure. (The critical dose is the radiation dose at which the optical density of the material is decreased by 1/e of its original value.) We also want to understand the damage mechanism. STXM acquires images and Near Edge X-ray Absorption Fine Structure (NEXAFS) spectra by using X-ray photons to excite inner shell electrons to unoccupied valence orbitals or to the continuum.

Characterization of deformed crystals containing extensive inhomogeneities and localized lattice reorientations presents a formidable challenge, particularly when the available sample volume is limited and/or irregular in configuration. The crystals analyzed in this work were originally high-symmetry single crystal tungsten rods, 6.9 mm in diameter and 102 mm long, which had been ballistically launched against steel blocks. At ordnance velocities (1.5 km/sec), the rods evert as they penetrate the block, resulting in a cylindrical cavity partially lined with a hollow tube of tungsten, with a slug resting at the bottom. These tungsten residuals contained many localizations, fractures, and assemblages of discrete sub-entities, typically 100-200 microns in size, within a macroscopic flow pattern extending over many millimeters. The measured cavity depths displayed a strong crystallographic symmetry dependence. Because the post-event samples, which were either blocks containing polished axial sections through the cavity or loose ejecta, were both massive and frangible, area-resolved X-ray diffraction, utilizing a fine-beam Laue technique with a precision specimen translator to permit correlation with micrographs, was used to map and characterize the lattice sub-structure.

The ESRF source properties offer the required coherence and brilliance to extend X-ray microscopy techniques, which have been extensively used in the soft X-ray region, to higher energies. This will enable new investigations such as the study of thicker specimens compared to the soft X-ray region, access to absorption edges of elements of major interest in the biological and materials sciences, in particular from Potassium to Chromium, the use of X-ray fluorescence for trace element mapping with high spatial resolution and the application of phase contrast.

Two microscopes are under development at the ESRF to cover a wide range of biological and material sciences applications:

A scanning transmission X-ray microscope (STXM) working in the photon energy range of 0.5-8keV. Whilst in its simplest form this might consist of taking multiple images of a single sample region at different incident energies, an interesting extension is to perform highly spatially resolved XAS scans on small regions of the sample.

X-ray imaging started over a century ago. For several decades its only form was absorption radiography, in which contrast is due to local variations in beam attenuation. About forty years ago, a new form of X-ray imagery, Bragg-diffraction imaging or X-ray topography, developed into practical use. It directly reveals crystal defects in the bulk of large single crystals, and paved the way to microelectronics by leading to the growth of large, practically perfect, crystals. The advent of third-generation synchrotron radiation sources of X-rays such as ESRF and APS is now making possible, through the coherence of the X-ray beams, a novel form of radiography, in which contrast arises from phase variations across the transmitted beam, associated with optical path length differences, through Fresnel diffraction. Phase radiography and its three-dimensional companion, X-ray phase tomography, are providing new information on the mechanics of composites as well as on biological materials.

Recent MXRF reports have demonstrated marked x-ray flux enhancements over traditional aperturecollimating systems when using a total reflection capillary optic. In the present work, several standard reference materials (SRMs) were examined with a Kevex Omicron MXRF system to compare the elemental sensitivities achieved with an aperture x-ray guide to those obtained with a monolithic polycapillary optic (X-ray Optical Systems). This study is relevant to the MXRF community using commercial aperture-based instruments where trace elemental concentrations in a sample sometimes mandate the need for a higher primary x-ray dose, but the flexibility of incorporating various sized apertures for other samples is still maintained.

By scanning the x-ray beam over a nickel/gold knife edge, the capillary focal spot diameter at the time of this work was determined to be ∼400 μm, while a 300 μm aperture provided a beam size of ∼550 μm due to beam divergence past the aperture.

Now that correctors for objective lens spherical aberration are becoming feasible, questions have been raised about the usefulness of high-resolution transmission electron microscopy at zero Cs and the possible difficulties of interpretation of such images. In general, high resolution TEM images are interpreted either by comparison with simulations from model structures or by contrast transfer functions (CTFs) to determine the weight (and sense) of spatial contributions to images from corresponding diffracted beams. At zero Cs, HREM image simulations will work, but a projected charge density theory should be used (instead of CTF theory) to interpret images. Both theories use approximations; CTF theory relies on kinematic scattering and PCD theory on zero Cs and limited defocus.

One of the most challenging tasks for high-resolution electron microscopy (HREM) is the atomistic investigation of defects and interfaces in thin films of semiconductors and ceramics. In particular, the application of these materials in electronic devices requires the understanding of the atomistic structure responsible for electronic and optical properties. The imaging of these structures requires a point resolution down to 0.1 nm. For example the smallest atomic spacing of GaAs in the (110) projection is 0.14 nm. According to Scherzer the point resolution of the TEM is proportional to spherical-aberration coefficient Cs1/4 and to the wave length λ3/4. Commercial medium-voltage microscopes up to 400 kV offer only a point resolution of 0.16 nm due to the high spherical-aberration constant Cs of the electromagnetic objective lens. Decreasing the wave length λ by increasing the electron energy improves the point resolution, but has the drawback of severe radiation damage and high costs.

Effective use of the ANL 300kV CFEG/A AEM requires a knowledge of the probe current and size distribution during operation. In most cases this can be calculated once basic parameters of the system are experimentally determined. Central to this is a determination of the source brightness of the Cold Field Emission Gun (CFEG) under standard operating conditions of the instrument. For the last ten years most probe calculations of CFEG's have been done assuming a brightness parameter of β∼×109 A/cm2/sterad2 at 100 kV, while few if any experimental values have been reported at 300 kV. Using β and the relationship that the operating probe diameter is obtained by adding in quadrature the Gaussian probe size (dg, the CFEG source contribution), and the diameters of least confusion due to spherical abberation (ds) and diffraction limits (dd) one obtains the total probe size dt as :

Although, critical performance criteria of TEM objective lenses, such as spherical and chromatic aberration and, more recently, 3-fold astigmatism, are of common concern, the important properties of TEM energy filters are less well-known and not often measured or quoted. In an effort to bring these to light, we have developed methods for automatically measuring and tuning three optical properties important to practical filter performance: non-isochromaticity, geometric distortion, and non-achromaticity.

A filter is perfectly isochromatic if the range of electron energies that contribute to an energy-filtered image is the same across all image pixels. In practice, energy filters display some degree of nonisochromaticity, i.e. the energy-loss range forming the periphery of the image differs by anywhere from a fraction of an eV to as much as few eV from that forming the center of the field of view.

We have modified a CTEM in order to perform Lorentz microscopy experiments at high magnification and resolution on magnetic materials. The modification consists of the adaptation of a second side entry goniometer (SEG) to the CTEM (JEOL 200CX STEM) column above the objective lens, in a region of the column where the measured residual magnetic fields are < 0.5 Gauss with the objective lens in a fully excited state (see Figure 1).

With the specimen positioned above the objective lens, it is necessary to increase the excitation of the lens by approximately 20% in order to bring the image into focus. This combination of a specimen positioned above the object pole piece, and the higher objective lens excitation, affects several changes in the optics. First, if a specimen is positioned within the pole piece gap and the objective lens is turned off for Lorentz microscopy purposes, the highest typical magnification achievable is approximately 1,000 X and the resolution is not much better than 0.1μm.

A major reason for performing microanalysis in the Field-Emission Gun Scanning Transmission Electron Microscope (FEG-STEM) is the very high spatial resolution of the information obtained. It has been estimated that in ideal cases, energy-dispersive x-ray analysis in such an instrument can provide 0.1 wt.% detection sensitivity for an element in a region about 1.5nm in diameter of a foil about 30nm thick. It is obvious that such performance requires that the instrument be properly adjusted, and hence that the probe-forming characteristics be fully understood.

There are three fundamental limits on the minimum size of an electron probe, these being i) the geometrical demagnification of the source, ii) diffraction at the beam-limiting aperture, and iii) spherical aberration in the probe-forming lens. In addition, misalignment of the beam-limiting aperture will result in probe aberrations.

Preliminary results were obtained from thick sections of biological specimens using 200 keV electron microscopes with an in-column “Omega” energy filter. Images of selectively stained specimens prepared by osmium impregnation of frog spinal ganglia, with thickness ranging from 0.5 μm to 2.5 μm, show drastic improvement in image quality with the use of the energy filter.

Two instruments with Omega filters were used. One was a non-commercial research prototype at JEOL Ltd. equipped with a LaB6 gun and the other was a commercial instrument (JEM-2010FEF) with a field-emission gun, installed at Kyushu University. We found that, although the field-emission gun offered a brightness 2 to 3 orders of magnitude higher, the LaB6 gun was more suitable for studying thick specimens, primarily due to the higher beam current of the LaB6 gun.

In the field of biological and materials sciences, the importance of energy filter transmission electron microscope(EF-TEM) is increasing. Because it is a powerful instrument for contrast enhancement and obtaining elemental mapping images. We have developed a 200kV EF-TEM equipped with a fieldemission gun and in-column spectrometer. The new EF-TEM JEM-2010FEF inherits the performance in high resolution imaging and analysis from field emission TEM. The outer view is shown in Fig.l.

Figure 2 shows the lens configuration of JEM-2010FEF. An in-column Q-type spectrometer is introduced within the imaging lens system. It was designed to have image distortion less than 1% and dispersion power 1.2p.m/eV for 200keV electrons. There is no need of compensating procedure of distortion. Imaging lens system consists of two objective lenses, three intermediate lenses and three projector lenses. The 8-stage imaging lens system enables wide range of imaging modes equal to conventional TEMs; energy spectroscopic image of magnification from ×200 to × 1,500,000, energy spectroscopic diffraction of camera length from 200mm to 2,000mm.

A new system for energy-filtered electron microscopy (EFTEM) has been developed, adapted for the study of life science specimens. Special attention has been given to integration, ease of use and the typical problems encountered while investigating unstained and cryo-specimens. Low-dose, cryocapabilities and a high-contrast objective lens have been combined with an imaging energy filter, digital image recording with phosphor scintillator and advanced image-processing capabilities. The CM120-BioFilter is an integration of the CM120-BioTWIN or the CM120TWIN with the Gatan postcolumn imaging filter GIF100. The BioTWIN objective lens (Cs=6.2 mm, f=5.9 mm) is optimized for high-contrast imaging. The TWIN lens(Cs=2mm,f=2.8mm) is optimized for high resolution in structural biology. The GIF 100 is a computer-controlled, second-order corrected energy filter with a cooled multiscan CCD for digital image recording. Improvement of the contrast can be achieved if inelastically scattered electrons are removed.

In order to reduce beam damage, biological TEM specimens are often observed at temperatures close to the boiling point of liquid nitrogen (77 K). Recently, encouraging results on single particles as well as on 2D crystals have appeared, derived from images taken near liquid helium temperature (4 K), in dedicated TEMs. At these temperatures the high resolution frequencies are much better preserved, increasing the allowable dose and thus the signal to noise ratio.4 Here we present the design of a new dedicated Philips He-TEM which combines the full functionality of a CM300 TWIN with a vacuum transfer system and a liquid helium cooled specimen holder.

A schematic overview of the Cryo-TEM is shown in figure 1. The key differences compared to a standard CM microscope are: 1) The tip of the specimen rod is cooled below 10 K and the rod itself cannot be taken out of the goniometer (CompuStage). 2) The specimen enters the column on the opposite side.

Three-dimensional electron microscopic imaging, or electron tomography, involves the reconstruction of individual objects from projection data collected over a large range of specimen tilts. Practical limitations of tomography are set by the range of specimen tilt available to generate projection data. In terms of tilt range, current data sets typically cover a range of up to +/-700° of tilt.

When used for applications requiring high tilt such as tomography, problems with specimen holder designs which use circular 3mm diameter grids include restricted field of view created by the standard mesh pattern of commercially available specimen grids. The standard square or hexagonally patterned support grids provide minimal area between grid bars to view a sample at high tilt. The width of the specimen clamping device and holder tip blade also limit the useable grid area to the center of the specimen when the holder is at high tilt positions.