J. Phys. IV France
Volume 04, Numéro C9, Novembre 1994
Proceedings of the European Symposium on Frontiers in Science and Technology with Synchrotron Radiation
Page(s) C9-35 - C9-39
Proceedings of the European Symposium on Frontiers in Science and Technology with Synchrotron Radiation

J. Phys. IV France 04 (1994) C9-35-C9-39

DOI: 10.1051/jp4:1994905

X-ray spectrometer for anomalous scattering experiments : tests of a graphite mosaïc and a germanium monocrystal as analyzers

M. Bessière1, S. Lefebvre1, M. Lemonnier1, F. Bley2 and S. Lequien3

1  LURE (CNRS/CEA/MEN), Bat 209d, 91405 Orsay cedex, France
2  LTPCM, ENSEEG, UA 29 du CNRS, BP. 75, 38402 Saint Martin d'Hères, France
3  ESRF, BP. 220, 38043 Grenoble cedex, France

Anomalous X-ray scattering provides a useful way for improving the contrast between different atoms present in multi-element samples. To take full advantage of this technique, the detector has to have a good energy resolution in the diffracted beam to separate the elastic scattering signal from the inelastic one due to fluorescence or resonant Raman scattering (RRS). Solid state detectors are usually used, but the energy resolution is limited to about 150 eV which does not allow the separation RRS from elastic scattering. Ice and Sparks [ l ] described a mosaic crystal spectrometer which resolves this problem. We have adapted their setup on our beam line D23A at LURE-DCI but we have encountered a loss of resolution when the Bragg planes of the sample are inclined. The shape of the diffracted beam incoming on the position-sensitive detector depends strongly on the inclination of the diffracting planes ; this explains the loss of resolution which is increased in our case by the large horizontal incoming beam size. Alternatively, we recommend the use of a perfect crystal as an analyzer coupled with a 0d detector which gives a very good resolution (≈10eV) without modification of resolution when Bragg planes of the sample are inclined in X. Furthermore this kind of spectrometer allows a high counting rate so that it is possible to measure close to Bragg peaks (clustering studies, for example) and even Bragg peaks. The loss of efficiency is not dramatic considering the high brightness of third generation synchrotron radiation and is more than compensated by the improved signal to noise ratio.

© EDP Sciences 1994