Standardless calibration for micro-XRF with capillary optics

SOSA, C.; LEANI J.; SÁNCHEZ H.J.; PEREZ, R.D.

Carlos Paz

Seminario; XIV Seminario Latinoamericano de análisis por técnicas de rayos x-SARX2014; 2014

The XRF emission from a bulk sample has not a linear dependence with elemental concentrations because of the so called matrix effects. The most versatile approach to correct matrix effects consists of the application of the Fundamental Parameter method (FP) which requires few calibration standards for the quantification of a wide range of matrices. For FP, the physical processes in the sample leading to the XRF emission are described by a set of analytical equations. It requires a database of fundamental parameters and the spectral distribution of the excitation radiation as well. For the Microanalysis by XRF (micro-XRF), the high intensity and anisotropy of the photon flux at the output of the focalization lens difficult a direct determination of the emergent spectrum. Different techniques have been used in the past to estimate the excitation spectrum using indirect measurements, such as attenuation of the primary intensity, x-ray scattering and induced x-ray fluorescence in targets. The first two methods are based on approximated theoretical models which produces limitations in accuracy. A recent induced XRF methodology consists of the measurement of the Ka x-ray fluorescence intensity from several thin targets covering the emission energy interval of the x-ray source. It employs the mean value theorem to obtain the transmission of the lens for micro-XRF with x-ray tube excitation assuming an empirical energy dependence of the transmission function. The method requires the input of the x-ray tube spectrum which means the dismounting of the spectrometer at least one time for an experimental determination or the employment of a theoretical model. The reported accuracy of this method is less than 10%. In the present work this last methodology is expanded to obtain the excitation spectrum for micro-XRF by means of the employment of robust thick targets and avoiding assumptions on the physical characteristics of the x-ray source. The proposed approach is described in detail and applied to characterize the x-ray source of our laboratory. Validation by means of quantitative analysis of several standard materials is shown as well.