I'm trying to measure thickness and index of refraction (n) of a multi-layer thin film. My very first layer is gold coating with unknown thickness, my second layer is annealed ITO coating with thickness of appr. 1450A and the substrate is glass. My device is Sopra GES-5E.

i have set my incident angle to 75 degrees. i have used several different models and here are my problems: I need to enter and approximate thickness of each layer and the analysis results are very dependent on my predicted thickness, how should I resolve this for a sample with unknown thickness?

Each approach (ex: using N&K library, or Cauchy's law, Drude Law,...) gives a different result, different by 20-30A, how should I know which one is correct? should I only rely on the one with highest goodness of fit?

For a material that is not in the library of the software how should I find the constants that i need to enter into the software for different models? for example Cauchy law requires 6 constants. If the N&K data is available, constants can be found by solving linear equations but how do we get the constants if the the N&K data is not available.

Thanks in advance