MultiLaue: A Technique to Extract d-spacings from Laue XRD

Abstract

doi:10.1017/S1431927616009764 Microsc. Microanal. 22 (Suppl 3), 2016 © Microscopy Society of America 2016 MultiLaue: A Technique to Extract d-spacings from Laue XRD Zack Gainsforth 1 , Matthew A. Marcus 2 , Nobumichi Tamura 2 and Andrew J. Westphal 1 University of California Berkeley, Space Sciences Laboratory, Berkeley, CA, 91020, USA Lawrence Berkeley Laboratory, Advanced Light Source, Berkeley, CA, 91020, USA Broad spectrum X-ray Diffraction (XRD) is named Laue after Max von Laue, and is the original XRD technique [1]. Today, monochromatic XRD is more common because Bragg's equation allows determination of d-spacings where Laue does not. Laue still remains in use for single crystal systems because it can be used to make very accurate unit cell determinations as well as for strain and orientation mapping. A Laue technique which could provide unambiguous determination of lattice spacings, a la Bragg's equation would be a huge leap forward, especially for multiphase samples such as meteorites, interplanetary dust particles and some geological specimens. We introduce a new technique we call multiLaue which allows such determination. First an unfiltered Laue pattern is acquired. Then a filter is inserted between the radiation source and the sample – in this case a 100 um thick fused silica wafer – and a new Laue pattern is acquired. A shield protects the detector from radiation scattered off the filter. Low energy X-rays are more strongly absorbed by the filter than high energy X-rays so reflections excited by low energies are more attenuated relative to those excited by high energies. We developed a computer model simulating the process, and use it to quantitatively determine the X-ray energy exciting a reflection based on the reduction of intensity across a set of three filters (100, 200 and 300 micron fused silica). We tested multiLaue on a Si chip on beamline 12.3.2 at the Advanced Light Source synchrotron. The unfiltered pattern is shown in Figure 1. We analyzed 32 reflections (Table 1) and computed best fit energies. These are compared with energies based on indexation via the XMAS software [2]. The best fit energies matched well between 11 and 23 keV, with a mean relative error of 1.9% and a standard deviation of the error of 1.4%. This means we can compute d-spacings with an accuracy of 21 keV have large errors. Reflections < 11 keV have a harmonic contribution from the spectrum above 21 keV and also show large errors. Therefore, accurate characterization of the radiation source, and optics is crucially important to achieving high accuracy quantification.