The use of single-crystal cuvettes with the properties of an optical shutter in X-ray diffractometers

G. E. Abrosimova,I. M. Shmytko
DOI: https://doi.org/10.26896/1028-6861-2018-84-6-34-37
2018-08-01
Abstract:The atomic structure of substances can be studied using X-ray diffraction methods. X-ray diffractometers contains X-ray source and goniometer with a detector of scattered radiation. A sample holder (a cuvette) with the material under study is placed in the center of the goniometer. The diffraction spectrum which represents the structure of the sample under study is recorded upon angular scanning of the sample and detector. Study of crystalline powders, amorphous substances, nanocrystalline and partially crystalline objects is often based on the Bragg-Brentano scheme (θ – 2θ-scanning) with divergent X-ray beams irradiating the entire sample surface. Scattered radiation from the cuvette can also enter the detector and affect the recorded diffraction spectrum. The aforementioned distortion is rather weak for polycrystalline samples due to strong intensity of crystalline reflections. However, when studying amorphous substances, nanocrystalline and partially crystalline objects, contribution of the scattered radiation can be compared with the radiation from the sample and thus interfere with an unambiguous determination of the sample structure. The results of using standard cuvettes for X-ray diffractometers D500, D5000 (Siemens), D8 ADVANCE (Bruker) and DRON are analyzed with a special attention paid to distortion of X-ray spectra attributed to the use of cuvettes made of plastic or amorphous quartz when studying amorphous, nanocrystalline and partially crystalline samples. Special cuvettes, being a kind of optical shutters for background scattering, which do not distort the diffraction spectra of the objects under study are developed: single crystal plates of a special orientation with cylindrical cavities for the samples. The advantages of a single-crystal silicon cuvette compared to plastic cuvette usually supplied for D500 diffractometer are estimated when studying an amorphous sample.
What problem does this paper attempt to address?