We present a new technique for X-ray microtomography under high pressure. By modifying an opposed-anvil high-pressure cell known as the Drickamer cell, monochromatic X-ray radiographs can be collected through the entire cell assembly and a thin-walled containment ring. We designed a rotation mechanism to rotate the Drickamer cell from 0 to 180° under hydraulic load, and examined pressure-generation efficiencies of the Drickamer cell up to 8 GPa at room temperature using the energy-dispersive technique through the containment ring, which allowed us to conduct a detailed evaluation of effects of geometric factors of the Drickamer anvils for tomography application. The maximum attainable load supported by the containment ring is proportional to the anvil diameter. Cells with larger anvil diameters are less pressure efficient, although they can reach higher pressures with much higher loads. Pressure efficiency generally increases with the tapering angle and decreases with tip diameter of the anvils. However, cells with larger tapering angles are more unstable, causing blowouts beyond a certain pressure. We evaluated the quality of X-ray images using the optical setup for conventional tomography at the GSECARS (GeoSoilEnviroCARS [Consortium for Advanced Radiation Sources]) beam line, the Advanced Photon Source. Noise level in the images depends on the material used for the containment ring. Containment rings made of either cubic boron nitride or silicon carbide allow us to better observe the images, but these materials are brittle and prone to mode-1 failure and are not suitable for high-pressure generation. The noise level of aluminum-alloy rings is somewhat higher, but the material is much more ductile, and hence it is capable of supporting higher loads. Using the aluminum-alloy containment ring, we conducted a commissioning run of tomography up to 3 GPa. We demonstrate that the high-pressure tomography setup is useful for studying internal structure of objects and density of melt and fluid under pressure.

Abstract Analytical techniques with high sensitivity and high spatial resolution are crucial for understanding the chemical properties of complex materials such as clay minerals. Several techniques are capable of trace element microanalysis, notably electron microprobe analysis (EMPA), proton-induced x-ray emission (PIXE), and secondary ion mass spectrometry (SIMS). These techniques are complementary, but none of them is suitable for all analyses and each has unique capabilities. EMPA is currently capable of mm-sized spots but minimum detection limits are no better than 50 ppm. PIXE is well-suited for analyses of relatively light elements with 10 ppm sensitivity and μm-sized spots. The x-ray fluorescence (XRF) microprobe exceeds both of these techniques in sensitivity, especially for heavy elements, and currently has comparable spatial resolution (Smith and Rivers, 1995). All three of these techniques are fluorescence-based so sensitivities are smoothly varying functions of atomic number. Elemental sensitivities for SIMS are highly variable depending on ion yield, and quantification is difficult because of matrix effects. SIMS has higher sensitivities than the other techniques for some elements and lower sensitivity for others. Quantification is comparatively straightforward for XRF because the physics of photon interactions with matter is well understood. Trace element microdistributions with the synchrotron x-ray microprobe can be determined with <10 μm resolution and <1 ppm sensitivity. Oxidation state maps can be produced with <100 μm resolution and <100 ppm sensitivity. Oxidation state maps can be produced with μ100 urn resolution and μ100 ppm sensitivity. Microtomography can provide three-dimensional images of microstructure with micrometer resolution. The purpose