Bragg and Barkla Polarization in Edxrf

The use of a combination target consisting of an HOPG Bragg crystal and an Al203 Barkla scatterer for the excitation of the elements Na to U represents further progress in the development of using polarized excitation in EDXRP. The high, integrated reflectivity of the HOPG crystal improves detection limits and sensitivities for Na, Mg, Al, Si, P and S-traces in light and heavy matrices in comparison with direct and secondary target excitation. The A1203 target has proven itself to be a useful Barkla u-sltterer for the exritsrtinn nf dementa with 7 > ,2(j in whirh the nrhipvd rbtertinn limits Y1..1W__ _--a 1--W W--W-l-l-v-v~_~_~_~_~1Y . . -1-B d ,. ---.a-W--W IV____. W_ -vIIII-v-______1Y are in the range >O.l ugg. The combination of the Al203 target and the HOPG crystal in a multilayer, results in a polarization filter for the whole primary spectrum of an x-ray tube, thus enabling overview analyses with measuring times I 100 s. Detection limits between 1 50 pg/g are achieved for the elements Na U. INTRODUCTION One of the most important advantages of IDXRZ is the simuitaneous analysis of eiements from Na up to U. The primary radiation scattered at the sample is recorded simultaneously with the fluorescence signal. In practice, the sample is excited several times. The excitation conditions are optimized for a small range of elements [ 11. The sensitivities and detection limits obtained with direct excitation could be improved up to an order of magnitude by using polarized radiation for the excitation of traces in light matrices [2-51. The polarization effect of x-ray radiation can be observed in the processes of Rayleigh and Compton scattering, Bragg reflection, Raman scattering and Borrmann transmission. The phenomenon of polarization, first examined by Barkla [6], has been used and continuously improved since the beginning of the 70’s [4-131. Copyright 0 JCPDS-International Centre for Diffraction Data 1997 Copyright (C) JCPDS-International Centre for Diffraction Data 1997 In contrast to radiation from secondary targets, the polarized x-rays produced by Barkla scattering are polychromatic. Thus, enabling excitation of fluorescence lines for an energy range higher than 8 keV [ 14, 151. The lower limit is determined by the ratio cross-section for scattering to mass absorption coefficient. The use of single crystals for polarization of x-rays is limited for a certain energy range. This depends on the anode material used and the lack of single crystals for energies higher than 15 keV [16]. The scattering efficiency of single crystals can be up to four orders of magnitude higher than for Barkla scatterers [ 111. Like for secondary targets, the Bragg reflected radiation is monochromatic. The constant simultaneous excitation of the elements Na to U is neither possible with only Barkla nor with only Bragg targets. It would be desirable to use a combination of both for an optimal excitation of all elements. The aim of this study is to present a suitable excitation device which is realized by a combination target consisting of an HOPG (highly oriented pyrolytic graphite) crystal and a Al203 Barkla scatterer. The performance of the excitation device is demonstrated for different applications. RESULTS AND DISCUSSION The use of a Cartesian geometry for the arrangement of tube, target, sample and detector allows a great flexibility for the excitation conditions. Barkla scatterers, Bragg reflectors or secondary targets can be used as targets. Each serves as polarizer for the incident radiation from the tube. The optimum excitation conditions are achieved with intense monochromatic radiation. The energy for which is slightly higher than the absorption edge of the element of interest. This can easily be realized with secondary targets, but only for a small group of elements per target. The Bragg reflected x-rays are also monochromatic but more intense. This is valid for energies < 5 keV. Due to their polarization and assuming that the Bragg crystal contains no elements which produce fluorescence radiation, it is possible to achieve detection limits lower than for secondary targets. Copyright 0 JCPDS-International Centre for Diffraction Data 1997 Copyright (C) JCPDS-International Centre for Diffraction Data 1997 The conditions for Bragg reflection are given in Bragg’s law: n/Z = 2dsin(9) where n represents the order of reflection, h the wavelength of the incident radiation, d the distance of lattice planes and 9 the incidence angle. The wavelength of the incident radiation is defined through the angle 9 as 45’ given by the Cartesian geometry and the lattice distance of the single crystal. The crystal used for reflection should have a highly integrated reflectivity, lattice planes with low Miller indices, no impurities, no characteristic radiation in the range 1 10 keV, thermal stability and should be commercially available. Different investigations show that an HOPG crystal possesses one of the highest integrated reflectivities in comparison with other crystals [ 171. Also it f%lfils the other requirements mentioned above. A further condition for using Bragg crystals for low energies is a highly intense primary radiation. This can be realized with an end-window tube equipped with a thin Be window. Because of the shorter distance between anode and target, high intensities can be achieved between 2 and 6 keV. The combination of HOPG ((002) plane, 2d = 6.71& with Rh La-radiation results in 9 = 43.2”. The Rh end-window tube used is eqipped with a 75 urn Be window. The integrated reflectivity for the 002 plane is 0.0032 rad with a degree of polarization P=O.99 [18]. This corresponds to results found by Chabot [ 191. Figure 1 shows the excitation spectra of an HOPG (highly oriented pyrolytic graphite) crystal in comparison with normal graphite. It can easily be seen that second and third order reflection can also serve for excitation. Excellent excitation is achieved for the elements Na to Ti for various applications. Copyright 0 JCPDS-International Centre for Diffraction Data 1997 Copyright (C) JCPDS-International Centre for Diffraction Data 1997