furthermore, it is also related to the surface effects/bonding. if you look at Sio2 spectra, in addition to the Si peak, many instances can resolve the +1, +2, +3 +4 state. Mostly the Oxide peak (SiO2)eventually dominates due to full oxidation. If you do a controlled oxidation of Si surface, you can manage to see these changes without even curve fitting the spectra. Many publications have such spectra..
Lior's answer is solid, but I'd say the main reason is instrumental broadening, if charging isn't an issue. If decreasing the pass energy of the analyzer doesn't show any improvement, you might consider the other three more strongly. Uncertainty in the photon energy can also be an issue, depending on whether you use a monochromator and the radiation source.
Remember that in the instrumental broadening is included the line-width of your X-ray source. If you need high resolution, the source must be monochromatized.