New VAMAS Project A6: Evaluation of Uncertainties in XPS Peak Intensities Associated with Different Techniques and Procedures for Background Subtraction

It is planned to produce simulated XPS spectra based on the preliminary design described in (1) below at the US National Institute of Standards and Technology (NIST). These spectra will be distributed to analysts who agree to participate in the project (2). The analysts will then send their results to NIST for analysis (3). Scientists who are interested in participating in this project are requested to contact either Joseph Conny ( or Cedric Powell ( Any comments or questions on the project can be sent to the same scientists.

(1)Design and preparation of simulated spectra with known peak intensities

(a)Synthesize simple XPS lineshapes using the Doniach-Sunjic equation with two values of the asymmetry parameter (0 and ~0.1) and two values of the width parameter (FWHM ~ 0.2 and 0.8 eV) (4 combinations); alternatively, asymmetry could be eliminated for the broader peak (3 combinations)
(b)Include or not include shakeup intensity (2 combinations)
(c)Convolve spectra from (a) and (b) with X-ray lineshapes (Mg, Al, monochromatic Al) (3 combinations)
(d)Simulate effects of single and multiple inelastic scattering, either by repeated convolutions of a selected cross section or by use of the GenerateˇÕ function of the Tougaard QUASES software. Three possible cross sections are proposed, one for a free-electron-like solid (e.g., Si), one for a transition metal, and one for an insulator (3 combinations)
(e)Convolve spectra from (c) with Gaussian function to represent analyzer broadening (FWHMs of ~0.2 eV and ~0.5 eV); smaller FWHM would be used for smaller linewidth from (a) and larger FWHM for larger linewidth from (a)
(f)Add a constant background to represent photoemission by bremsstrahlung (for unmonochromatized Mg and Al X-ray sources) and multiple inelastic scattering of photoelectrons at initially higher energies (for all sources)
(g) Add random noise (~1% at peak maximum)

A total of 72 (or maybe 54) spectra will be generated from this design, but only 24 (maybe 18) are likely to be analyzed by a single analyst (corresponding to a particular X-ray source in his/her laboratory).

(2)Distribution of spectra (on disc) to analysts. Analysts will be asked to document:

  • background-subtraction technique (e.g., straight-line, Shirley, Tougaard, or other)
  • background-subtraction procedure (e.g., selection of endpoints)
  • curve-fitting equation used
  • measure of goodness of fit
  • peak intensity (area) for each spectrum
  • peak energy
  • estimate of uncertainties of intensities and energies

(3)Analysis of results. The reported intensities will be compared with true intensities known from the design. Differences will be correlated with the chosen background-subtraction technique and with analyst choices in the data-analysis procedure.

SASJ Top Page.
Copyright (C) 2000-2002 Surface Analysis Society of Japan. All rights reserved.