SIMPLE PROCEDURE FOR PRECISE PEAK MAXIMUM ESTIMATION FOR ENERGY CALIBRATION IN AES AND XPS

Citation
Pj. Cumpson et al., SIMPLE PROCEDURE FOR PRECISE PEAK MAXIMUM ESTIMATION FOR ENERGY CALIBRATION IN AES AND XPS, Surface and interface analysis, 24(10), 1996, pp. 687-694
Citations number
14
Categorie Soggetti
Chemistry Physical
ISSN journal
01422421
Volume
24
Issue
10
Year of publication
1996
Pages
687 - 694
Database
ISI
SICI code
0142-2421(1996)24:10<687:SPFPPM>2.0.ZU;2-I
Abstract
We present a simple, easily-reproducible method of locating the maxima of XPS and AES peaks, to an energy uncertainty of, typically, 2-10 me V, which is particularly useful in the energy calibration of instrumen ts using spectra acquired from standard reference materials. This meth od is mathematically equivalent to performing a least-squares fit of a parabola to the top of the peak, but the calculation described here i s very simple, requires no computer and can easily be done by pocket c alculator. Although parabolic fitting to give peak energies is not a n ew method, we present it here in a particularly simple tabular form, w hich makes it easy to perform even without a computer. The statistics known to apply to XPS and AES intensities then allow us to derive a ne w equation for estimating the precision of the peak energy. We use thi s to obtain simple rules which give the most efficient 'trade-off' bet ween calibration and time spent acquiring the spectre. The XPS calibra tion spectra from a VG Escalab II spectrometer are used to verify the efficacy of the method, although the same method is readily applicable to AES. Quantitative comparisons are made with previous (and rather m ore time-consuming) peak location methods. Agreement is excellent, sho wing that the new method can be used as part of an energy calibration procedure with the confidence that no discontinuity will be introduced into the calibration history of a spectrometer when switching to the new method. Precise energy calibration becomes increasingly important with the use of numerical methods such as linear least-squares fitting principal component analysis and factor analysis.