This isn't a tutorial, but I attach a paper below as an example of how to deal with measured elements that are common to both the thin film and the substrate.
Normally when both the thin film and the substrate *do not* contain common elements, and both thickness and composition are unknown, we can utilize multiple accelerating voltages to allow the analytical iteration in STRATAGem to converge. That is to say, by utilizing multiple accelerating voltages, e.g., 10, 15 and 20 keV, on both our bulk standards and our thin film unknowns, we can determine both the thickness and composition of our thin films. Multiple beam energies are required because at a single beam energy one can find a range of inversely varying compositions-thicknesses that fit the k-ratios from a single beam energy measurement.
So, for the situation where one is measuring an element that is present in both the thin film *and* the substrate, e.g., Ni-Si on Si or say, Ge-Si on Si, and both the thickness and the composition of the thin film are unknown, one must constrain something or the STRATAGem software will not converge on a solution.
The paper attached below (remember to login to see attachments!), describes an approach we developed at Univ of Oregon for these types of samples. Please consider that if we are measuring an element in a thin film that is also present in the substrate, we are essentially subtracting two very large numbers from each other to obtain a very small number!
john