To fit an XPS spectra I typically do the following steps:

1. Smooth the curve

2. Subtract the baseline (shirley background)

3. Figure out the peak positions and FWHM from the literature

4. See if the peak shows multiplet or spin-orbital splitting. The splitting energy is constant between the corresponding peaks, so incroporate those constraints in your fitting parameters.

As far as the software goes you could use any, origin will work. I use peakfit or XPSpeak never had any trouble fitting any XPS curve.

CASAXPS is a good tool to do curve fitting for XPS data, the only trick in good analysis is keeping balance between mathematical fitted peaks you obtain form software and the actual chemistry justifying those curve fits 

Casa XPS fits Voigt approximation Gaussian-Lorentzian peaks. You can set the fitting up to use either the product or sum approximation Voigt function (the latter has been shown to give a better approximation - Hesse et al Surface and Interface Analysis 39, pp.381-391 (2007)). You can use CasaXPS to specify the intensity and spin-orbit splitting of doublet peaks. E.g. if you fit a 2p peak, the ratio of their intensities is 2:1 due to degeneracy (for d, 3:2 and f, 4:3), so you can specify that one peak has half the intensity of the first, and you can also set the BE offset between them (the splitting), to be from literature values (use the NIST datbase http://srdata.nist.gov/xps/). Note that the Gaussian-Lorentzian mixing can also be adjusted, depending on whether line broadening is being more affected by the sample environment or the experimentation.

If you're new to CasaXPS I suggest you read through their manuals first. As Dr Riaz says, it's easy to create some peaks, but they should have physical meaning. See http://www.casaxps.com/ebooks/ebooks.htm

You can use any software that allows for curve fitting to do this, e.g. Origin or Matlab - simply use a least squares curve fitting algorithm, using sum-approximation Voigt functions. In all cases you allow for the intensity, width and BE position of the peak(s) to be fit. Many people use CasaXPS because of its ease of use.

Dear Mr. Ben F. Spencer,

I tried to use casaxps. It's a good software for analyzing XPS but I cannot save or export data due to my case is Demo version. Do u know how to use this as well without license! Thank in advance

You'll need a license to save and export. If you can't buy a license then maybe you could look into fitting Voigt approximation peaks in other software e.g. Igor, Matlab, but these all require licenses that will be much more expensive than CasaXPS.

Dear Anusuyajanakiraman,

You can try one of the following free versions of softwares.

http://www.uksaf.org/software.html

Manokaran Anusuyajanakiraman Deconvolution of a composite peak into its individual peaks plays an important role in the interpretation of many types of graphs including XRD, XPS, FTIR, and PL etc. In this video, I have discussed how to deconvolute simple combined peaks, composite peaks and how to correct missing data in a given peak with the help of deconvolution. In the case you want to further ask about it, please do comment on the specific video, I'll respond to it shortly. I have provided the practice files (OriginLab) here. Thanks

https://youtu.be/WX_X3XlbKRA

How to plot XPS survey graph using Origin

https://youtu.be/t2xrywy5988

How to plot XPS data in Origin: #aminulcheminnovation #AminulSir

https://youtu.be/vulZ4N8BxxA