Quantitative AES IX and quantitative XPS II: Auger and x-ray photoelectronintensities and sensitivity factors from spectral digital databases reanalysed using a REELS database
Mp. Seah et al., Quantitative AES IX and quantitative XPS II: Auger and x-ray photoelectronintensities and sensitivity factors from spectral digital databases reanalysed using a REELS database, SURF INT AN, 31(8), 2001, pp. 778-795
An extension has been made of previous analyses of peak area intensities fo
r elemental spectra in digital Auger and x-ray photoelectron databases. The
intensities, instead of being analysed after removal of a background deriv
ed from the Tougaard Universal cross-section, are now analysed after remova
l of the extrinsic characteristic loss background calculated by deconvolvin
g the relevant angle-averaged reflected electron energy-loss spectroscopy (
REELS) spectrum. The angle-averaged REELS data for each element are calcula
ted from a digital REELS database using a recently defined scaling of the m
easured characteristic losses from the elastic peak. The new background rem
oval procedure leads to an improvement in the correlation between experimen
t and theory for intensities in both AES and XPS. Analysis of these correla
tions shows that a systematic divergence remains for each element, which is
the same for XPS as for AES. This divergence is attributed to an inadequac
y either of the angle-averaged REELS method or of the material-to-material
dependence of the TPP-2M equation used in the calculation of the inelastic
mean free paths. Correction for this is possible in a new matrix-less quant
ification formulation using average matrix sensitivity factors. This leads
to correlations between experiment and theory with scatter factors of X/div
ided by 1.08 and x/divided by 1.11 for AES and XPS, respectively, for a wid
e range of elements and peaks. These scatter factors are much better than f
or previous correlations and underpin the choice of formulae to calculate t
he relevant theoretical intensity predictions. These calculations lead dire
ctly to values of average matrix sensitivity factors, appropriate for use w
ith spectrometers giving true spectra or with those spectrometers that may
be calibrated to provide such spectra. (C) Crown Copyright 2001. Published
by John Wiley & Sons, Ltd.