If you only need to analyze C, XPS might not be a suitable technique. Contamination carbon will always be present in the spectra no matter the sample, due to adventitious carbon.
You will be able to determine all elements and to distinguish their different species. However, C1s peak will might not that trustful due to the adventitious contamination and your substrate. But for B, N and O you could analyze them easily.
However, I would recommend you to run a survey spectrum of your single substrate to discard any B, N or O contamination.
How can we remove this discrepancy of C1s signals? Because it is really important for me to know these signals. Maybe change the substrates to silicon wafer or it doesn’t matter?
The adventitious Carbon signal will be present in all your samples unless they were not in contact with the air, no matter the substrate. But that does not mean that you won't be able to analyze the C1s signal.
Adventitious carbon typical consists of a high contribution at 284.8 eV, which corresponds to C-C bonds. I assume that your particles are g-C3N4 doped with B or O and thus, regarding the C1s region, you should see C-N and N=C-N which you might be able to identify because theirs binding energies are different compared to the C-C.
What you won't be able to do is to establish a surface stoichiometry for your samples.
Thank you for your response. But maybe FTIR can help also? They will concur with each other no matter what right? XPS is another way to know the functional groups present.