Hi,
Here’s a pretty basic example.
I use a SF10 glass prism to couple into a 50 nm thick layer of gold. The analyte is water at 1.33 and the wavelength is 633 nm.
You’ll notice that the coupling conditions are not perfectly met, probably due to a slight change in refractive index of gold as compared to the source I used to determine the angle (57 degrees incidence from the vertical).
What is important for this type of simulations is to mesh fine enough to resolve correctly the metallic layer (here I go down to 5 nm resolution, but it’s probably overkill).
A note: performances will be horrible compared to simple analytical formulations (Fresnel reflection coefficients and Rouard method for example). It only makes sense if you geometry is not a simple uniform film.