Numerical determination of a true absorption spectrum from grazing-incidence fluorescence EXAFS data

Abstract

The fluorescence EXAFS spectra, measured at a grazing incidence angle larger than the critical angle for total reflection, were examined to test a numerical conversion from the distorted fluorescence data to the 'true' absorption spectra, achieving a reliable transferability between the transmission and the fluorescence data. Under various entrance and exit angles, full numerical corrections were made in terms of the sample's selfabsorption, the measurement geometry, and the energy dependent efficiency function of the I,, ionization chamber detector. Particularly, a numerical code was developed for the x-ray fluorescence geometry configured without a scattering discrimination filter.For an illustration of glancing-angle fluorescence EXAFS spectra, we performed an in-situ measurement for the thick porous nickel specimens undergoing a heated-air surface oxidation process.Keywords: Grazing Incidence; Fluorescence EXAFS; Numerical Code. Description of Fluorescence EXAFS SignalIn the fluorescence EXAFS mode, the flux of incident x-ray at the monochromatic energy (EA is monitored by the front semitransparent ionization chamber (L,), and a combin~ition of fluorescence photon energy (EF) and the scattered radiation (EF<E.,,.<Ei) are measured by the rear fluorescence detector (IF).The scattered radiation can be largely removed by a combination of Z-I filter and slit mask, as normally employed in the Lytle detector. (Stern et al, 1979;Lytle et al, 1984). In an advanced technique, a suitable energy dispersive detector (such as Si/Li) can discriminate the background signal caused by scattering.The non-zero background is present when the fluorescence signal is detected without a scattering discrimination filter. As pointed out in other study (Troger et al, 1992), the IV/I,, spectra, modeled under a self-absorption geometry for a thick sample, can J(E,) = IF(E~) / L,(E,) p,(Ei) = + J~,,,(E~) IaT(E,) + p,(Er)sin(O)/sin(qb)be expressed as:Eq (1) Here, 0 is the entrance graze angle of the incident x-rays, and q~ is the exit graze angle of the detected fluorescence radiation with respect to the sample surface.

Au Clusters on TiO2(110) (1 × 1) and (1 × 2) Surfaces Examined by Polarization-Dependent Total Reflection Fluorescence XAFS

“…The Au Lα fluorescence was detected by a 19-element Ge solid-state detector (GL0110S, Canberra, USA) at room temperature. Self-absorption was negligible because the coverage was sufficiently small, i.e., 1.5 × 10 13 atoms/cm 2 …”

“…Self-absorption was negligible because the coverage was sufficiently small, i.e., 1.5 × 10 13 atoms/cm 2 . 26 EXAFS analysis was carried out by REX 2000 (Rigaku Co., Japan). 11,27 The EXAFS oscillations were extracted using a spline smoothing method with a Cook−Sayers criterion 28 and normalized by an edge height with its energy dependence expressed in a Victoreen equation.…”

“…35 It should also be noted that bonding interactions between the Au clusters and the substrate constituents were not detected. Further analyses were carried out using a FEFF code 26 and a real space model. 13 From a geometrical consideration of Au clusters based on preliminary XAFS results and other reports, 32,33,42 we constructed icosahedral (I h ) and cuboctahedral structures (O h ) that have energetic stability at the magic numbers of 13, 55, and 147 atoms (Figure 4).…”

Origin of Self-Regulated Cluster Growth on the TiO2(110) Surface Studied Using Polarization-Dependent Total Reflection Fluorescence XAFS

“…Polarization-dependent XAFS is measured at the coverage of 4 × 10 14 and 1 × 10 14 atoms/cm 2 . The self-absorption effect is negligible in our case because the coverage was sufficiently small . We use the notation of (111) fcc , (100) fcc and so on to show the direction of Ni fcc plane to distinguish the notation of the TiO 2 (110) surface.…”