As a starting point, what do you have to set up a calibration curve with? Do you know the percentage of each component?
Don't know what kind of samples you will be analyzing. We ran into Si:N films mostly as a mold coating, which made analysis difficult.
If you have a polymer or plastic film, you can try ATR (attenuated total reflectance) for the analysis. You may also hear this called MIR (multiple internal reflectance). Trying to do transmittance analysis could be tricky, but it is possible.
Good luck.