Calculated initial vacancy ratios (IL = I_{{ m KL}^{1}}:I_{{ m KL}^{0}}), extrapolated x-ray intensity relative to the main line (RL = X_{{ m KL}^{1}}:X_{{ m KL}^{0}}) and to the total intensity (RT = X_{{ m KL}^{1}}:X_{{ m KL}^{0}+{ m KL}^{1}}), average differences in X_{{ m KL}^{1}} intensity between oxides and pure elements extracted from the literature (ΔRL), and the product effect on the total x-ray yield
Table 1. Calculated initial vacancy ratios (IL = I_{{\rm KL}^{1}}:I_{{\rm KL}^{0}}), extrapolated x-ray intensity relative to the main line (RL = X_{{\rm KL}^{1}}:X_{{\rm KL}^{0}}) and to the total intensity (RT = X_{{\rm KL}^{1}}:X_{{\rm KL}^{0}+{\rm KL}^{1}}), average differences in X_{{\rm KL}^{1}} intensity between oxides and pure elements extracted from the literature (ΔRL), and the product effect on the total x-ray yield. Absolute uncertainties are listed in parentheses.
Abstract
Proton-induced x-ray emission (PIXE) was used to assess the accuracy of the National Institute of Standards and Technology XCOM and FFAST photo-ionization cross-section databases in the low energy region (1–2 keV) for light elements. Characteristic x-ray yields generated in thick samples of Mg, Al and Si in elemental and oxide form, were compared to fundamental parameters computations of the expected x-ray yields; the database for this computation included XCOM attenuation coefficients. The resultant PIXE instrumental efficiency constant was found to differ by 4–6% between each element and its oxide. This discrepancy was traced to use of the XCOM Hartree–Slater photo-electric cross-sections. Substitution of the FFAST Hartree–Slater cross-sections reduced the effect. This suggests that for 1–2 keV x-rays in light element absorbers, the FFAST predictions of the photo-electric cross-sections are more accurate than the XCOM values.