For unresolved peaks additional information is required to provide meaningful intensities. The fit quality can deteriorate markedly for fitting to a peak shape different to that used in the peak fitting algorithm (ie fitting sum function peaks with product function software or vice versa). The fitting of the individual peaks is only good for software using the same peak shapes as those of the data. One full set of spectra is constructed with Gauss–Lorentz product function peaks and a second full set with Gauss–Lorentz sum function peaks since each software was designed for fitting only one of these options. Synthetic spectra approximating the carbon 1s peaks for poly(methyl methacrylate), poly(vinyl acetate), poly(vinyl chloride) and poly(isobutylene) allow an analysis of accuracy of the fits in these tests. In addition, each data system had programming faults, some of which were later rectified. The results show that all the data systems studied accept VAMAS format files but not all would accept both of the options of direction of scan. The synthetic spectra allow tests for: (1) acceptance of VAMAS formats with kinetic energy or binding energy scales that may have positive or negative abscissa increments (2) Gaussian–Lorentzian sum or product function peaks (3) effectiveness of Shirley background subtraction (4) effectiveness of Tougaard background subtraction (5) peak synthesis with partly resolved peaks with no background and (6) peak synthesis with a Shirley background. The development of a range of VAMAS format data files containing synthetic test spectra and their use in evaluating the peak fitting routines of three commercial XPS data analysis systems A, B and C, is reported.

Xps peak analysis sofrware