High- resolution energy fluorescence on MARS

The MARS (Multi-Analyses on Radioactive Sample) beamline is the only SOLEIL beamline dedicated to the characterization of radioactive materials (⁹⁹Tc, ²³²Th, ²³⁸U, ²³⁹Pu…) by X-ray diffraction, X-ray absorption spectroscopy in the energy range 3.5 to 35 keV. Optical elements have been installed to allow for experiments on complementary analysis equipment located in the same experimental hutch (Figure 1): very high resolution and transmission X-ray diffraction, X-ray absorption spectroscopy (transmission, fluorescence and high resolution) and X-ray fluorescence spectroscopy These techniques can be used for a wide range of applications in various disciplines (solid state chemistry and physics, earth science, biology, etc.) on samples that have in common the presence of one or more unstable elements (radionuclides). Radionuclide decay is the source of specific emissions from the samples at energies that may be in the same domain as the X-ray fluorescence energy.

 
Figure 1 : Photo of the 2 analysis endstation on the MARS beamline.

Energy-resolved solid-state detectors are widely used on beamlines dedicated to X-ray absorption spectroscopy to record fluorescence spectra. Different detectors are available including ultra-pure germanium detectors cooled with liquid nitrogen and semi-conductor detectors such as silicon drift detectors (SDD) cooled using Peltier effect. These detectors are relatively easy to install and allow the probe of diluted elements. However, with this type of detector, limitations occur related to their energy resolution (150-300 eV) and saturation effects. Saturation often appears in the analysis of samples containing elements that present X-rays with energies close to the energy of the fluorescence being studied. One solution is the use of crystal analyzer spectrometers coupled to a rapid detector with high count rates. The principle of this type of spectrometer is to optically select the photons by diffraction on a single spherically bent crystal and then to focus them on a detector located above the sample. Thus, the detector never "sees" the sample directly.

The novel use of this type of spectrometer at SOLEIL is not limited to the MARS beamline; a few X-ray absorption spectroscopy beamlines at other synchrotrons (SLS, ALBA, ESRF, etc.) are also equipped with it, even though these are not confronted with the specific problems regarding emissions from the radioactive samples themselves.

Taking into account the specificities of the MARS beamline such as the lack of space around the sample and the presence of shielding, a spectrometer was designed and created by the FAME beamline similar to that installed on FAME . The spectrometer (Figure 2) consists of four measuring pathways using spherically bent silicon crystals (with a radius of curvature of 0.5 m) developed by the Institute of Mineralogy and Physics of Condensed Media (IMPMC) in Paris . Due to the small radius of curvature of the crystals, the spectrometer is a compact equipment, which therefore has the advantage that it can be installed on a non-permanent basis on a multi-analysis beamline such as MARS.

 
Figure 2 : Photo of the spectrometer on the analysis endstation during an experiment on the MARS beamline. 

¹ J.-L. Hazemann et al., Journal of Synchrotron Radiation (2009) 16, p283-292
² E. Collart et al., Journal of Synchrotron Radiation (2005) 12, p473-478

The first test experiment was carried out in July 2011 at high energy (Kα1 line of niobium at 16.6151 keV) and at "low" energy (Kα1 line of copper at 8.04778 keV).
The main advantage of this type of spectrometer is its ability to discriminate the fluorescence lines of the element being analyzed from the fluorescence of the matrix or radioactive X-ray emissions from the sample. In the case of a (non-radioactive) zirconium sample containing 1% niobium, the challenge is to probe an element of atomic number Z (niobium) that is relatively dilute, in a matrix of atomic number Z-1 (zirconium). The spectra were recorded in high resolution mode with the spectrometer and in total fluorescence (Figure 3). 

 
Figure 3 : Emission spectra (left) and absorption spectra (right) collected using high resolution and total fluorescence (classical measurement) on a Zr – 1% Nb sample.

Without a spectrometer, the fluorescence signal is completely saturated by the zirconium fluorescence. Under optimal conditions, only the niobium Kα1 fluorescence is present in the emission spectrum. From these measurements, two observations can be made: (i) high-resolution measurements give a better resolution of the threshold structure and (ii) saturation of the detector has been overcome to obtain better quality data.

Beside high-resolution measurements, new spectroscopies (resonant inelastic X-ray scattering, X-ray emission spectroscopy) can be carried out for more detailed analysis of the electronic structure. Figure 4 shows the RIXS map recorded on a solution of aqueous copper +II. The two plots (horizontal and vertical) on this graph were obtained by integrating the constant incident energy (CIE) and constant energy transfer (CET). Note that these spectra are identical, respectively, to an emission spectrum on the Kα1-Kα2 zone and to an absorption spectrum.

 
Figure 4 : RIXS map recorded on an aqueous solution of copper +II – integration of constant incident energy (CIE) and constant energy transfer (CET). The absorption spectrum is recorded in high resolution and in transmission. Insert: profile of the elastic peak.

The absorption spectrum was recorded in transmission and with the spectrometer with an energy resolution of 2.3 eV. The latter clearly showed a better resolution compared to a conventional measurement (transmission), particularly in the pre-edge area, corresponding to the transition 1s --> 3d at 8.9765 keV and the threshold with a more marked shoulder corresponding to the greater transition of the ligand orbitals to the 3d orbitals of the metal. Note, however, that this improved resolution amplifies these structures due to strong resonant effects. The energy selectivity of the crystal is determined by a spectrum around the selected energy (a measure of the elastic peak).

The spectrometer will soon be open to users of the MARS beamline. In the future, it may be coupled with KB micro-focus optics for more detailed analysis of heterogeneous samples.

This new equipment has benefited from close collaboration between the FAME beamline at ESRF, MARS at SOLEIL and members of CEA Cadarache / DEN (Philippe Martin’s group) and continues with the IMPMC group responsible for producing the four crystals.