E1438

SIMULTANEOUS MICRO XRF/XRD ANALYSIS OF HIGHLY INHOMOGENEOUS SAMPLES. Anders Rindby1, Per Engström2 and Koen Janssens3. 1Chalmers University of Technology, Göteborg Sweden; 2ESRF Grenoble, France; 3Antwerpen University (UIA), Antwerp, Belgium

With the opening of the first really "third generation" synchrotron source in Grenoble, at fall 1994, x-ray sources of unprecedented brilliance's and qualities became available to the scientific community. Different x-ray analytical technique could now be applied on a level that was "out of imagination" only a decade ago. Here we present some preliminary results from a micro-diffraction/fluorescent experiment applied at a um level carried out at one of the most powerful synchrotron microbeam available in the world right now, the Bl 1 at ESRF. This beamline can now provide a 2 um beam with a flux density of 1010 photons/um2 at an energy of 13 keV and with a bandwidth of 10-4.

In this experiment x-ray diffraction and x-ray fluorescence has been combined in order to perform a kind of "micro-chemical" reconstruction of heavily inhomogeneous fly-ash particles which normally are inaccessible to conventional micro analysis. The experiment was done in a scanning mode and 2D images of different analytical information were reconstructed from the data recorded during the scan.

From the diffraction data achieved by this combined experiment, the main minerals in u-sized fly-ash particles have been identified and the size distributions of these minerals have also been determined. The distribution of heavy elements, recorded from their fluorescent intensity, is presented and the impact of sample surface topology and sample attenuation is discussed. Estimates of the concentration of the these heavy elements are given and correlation analysis has been performed indicating that most of these elements seem to appear at the surface of the fly-ash particles.

The major features and limitations of this micro-analytical technique will be outlined and different examples on how the analytical information can be used for generating 2D images of the sample will be demonstrated and discussed.