After measurement, you will receive the Fine scan spectrum data of each element you collect, and maybe the relative content of these elements if the laboratory technician has calculated for you (eg. 30% C, 20% Sn, 38% O, 12% N). Then you just need to calculate the atomic ratio of O/Sn by 38/20 = 1.9, and the vacancy concernration is 10 %. However, sometimes the 38% O content may be not all contributed by your SnO2 structure, then you need to do the peak fitting with XPSPEAK41 (or CasaXPS, Unifit...), after which you can know the ratio of peak area of oxygen which belongs to SnO2 (like 80%), so the valide content of O of your SnO2 structure is 38%*80% = 30.4%, and O/Sn =1.52.
By another situation, if you just receive the spectra data, then you need to calculate all the ratios yourself (with sensitivity factor). First is the peak fitting, after that, use the eq. Na/Nb = (Ia*Sb)/(Ib*Sa), where I is the peak intensity in the unit of cps.(namely peak area), and S is the sensitive factor of the element (normally use the relative sensitive factor referred to fluorine. So now you get the atomic ratio of O/Sn again. Attention please, the S value is different for different equipments, so you may need to contact the laboratory technician.
I did the same thing for calculating the carbon vancancy before, for reference:Article Enhancing the magnetism of 2D carbide MXene Ti3C2Tx by H2 annealing
(in supplymentary file)
For another advise, XPS is a semi-quantitative and surface sensitive (
For a starter, Zhang Kaiyu gave pretty good advice. For manual data evaluation I also always recommend the Omicron manual and the references therein:
http://uhv.cheme.cmu.edu/manuals/M470101.pdf
The interesting part starts at page 67.
However, note that XPS analysis has a somewhat large error bar both due to the various available cross section sets, the user dependence of the baseline subtraction and other parametric choices. So it will be good to compare the doping degree of various samples but the absolute values will have some sort of a question mark going along with them.
in line with what Jürgen Weippert said, I would only add a consideration.
If we are speaking of "doping" or structural/chemical "defectivity" of a film or surface, we are usually facing very small fractions of the sample composition. Do not forget that techniques like XPS (but also Auger, EELS) have a minimum real sensitivity in the order of 0.5-1% atomic contribution. That is, if you expect doping or defect amounts less than that limit, you won't detect such structure in a XPS spectrum above the noise level.
The same happens when a difference between two samples is in an mount of, say, oxygen missing when this difference is expected to be lower than the said sensitivity limit.
In the case of doping or defectivity levels lower than 0.5-1at% (as in most cases of process induced oxygen vacancies in a metal or semiconductor oxide film), XPS would be practically useless... Properly designed and calibrated electrical measurments would be much more sensible to such low concentrations of defects.
XPS is a strong technique to calculate the oxygen vacancies and different surface species that includes oxygen. Please the paper on the attachment, that can help.
Zhang Kaiyu provided great information on how to calculate the vacancies but as Giuseppe Curro pointed out the calculation will not be reliable unless the "doping concentration" is higher than XPS resolution.
Plot the XPS spectra, do the Gaussian fitting and after that use area under the curve for calculating the oxygen vacancies. I did the same procedure for calculating the oxygen vacancies in TiO2 thin films.
inline with Zhang Kaiyu, if you are using CasaXPS then use Shirley background removal method. a peak around 530 eV in O1s core level spectra corresponds to oxygen vacancies and the area under this peak quantifies vacancies.