Simple Peak-fit Example using O (1s) Chemical State Spectrum
Peak-fitting Process – Raw to Finished Product for a Single Peak Spectrum
Finish of Peak-fitting O (1s) Signal (Peak)
Advanced Peak-fit Example using Si (2p) Chemical State Spectrum
Peak-fitting Process – Raw to Finished Product for a Multiple Peak Spectrum
Assigning Chemical States after Peak-fitting
The final peak-fit is NOT 100% correct because the D and E peaks should have doublets.
After Peak-fitting Si (2p) spectrum is Completed, then
we look at chemical state tables that show which chemical states can be assigned to which BEs.
This Spectrum shows which chemical state belongs to each of different peaks based on BE measured
Peak-fit for Native Oxide of Silicon – fully labelled with all spin-orbit coupled pairs of peaks
This peak-fit example includes the spin-orbit coupling for the each chemical state (species)
which produces a complicated peak-fit.
The 4 chemical states include: pure Si (peaks A&B), Si2O (peaks C&D),
Si2O3 (peaks E&F), and SiO2 (peaks G&H).
To produce this peak-fit we constrained the BE difference between the 2p3 and the 2p1 peaks to be 0.6 eV which is based on the pure Si peaks. The peak-fit uses a peak area ratio of 2.0 for the four (4) different 2p3 – 2p1 theoretical peak areas. The peak-shape of all signals was constrained to 90% Gaussian – 10% Lorentzian. Asymmetry was not applied. Different FWHM were tested for the oxidized forms of Silicon. A FWHM = 1.00 was used for peaks C-H. It is possible that a better FWHM may be slightly larger. Pure elements normally have FWHM smaller than the FWHM of their chemical compounds such as SiO2. Useful FWHM values are correctly being added to this website.
Silicon peak-fitting is can be very challenging. Here is an example of Si (2p) spectrum from a Synchrotron facility. The drawing shows the relation between the Silicon atoms and their XPS peaks. In this advanced example, the scientist, for ease of viewing, has also ignored the 2p3/2 and 2p1/2 of the SiO2 peaks.
Resolving Peaks that are Overlapping – What is Possible?
- Automated peak-fitting
- Baseline averaging
- Baseline smoothing
- BE difference – fixed
- Chi-square values
- Core-hole lifetime
- Differential charging artifacts
- Gelius peak shape
- Instrument induces Gaussian
- Iterative baseline
- Linear baseline
- Lorentzian X-ray beam
- Marquardt-Levenberg fit routine
- Natural FWHM
- Number of data points/eV
- Pass energy effect on FWHM
- Peak area ratios
- Peak constraints
- Peak shapes: Gauss vs Lorentz
- Peak shoulders-low BE side
- Shirley baseline
- Step sizes
- Tougaard baseline
- Voigt peak shape