A few questions regarding the "Scatterer On Substrate" model

Hello everyone,

I had a few questions about the "Scatterer On Substrate" model. My understanding is that the general strategy is to solve for the field in the absence of the scatterer, then put the scatterer back in and solve for the scattered field. My first question: is it possible instead just to solve for the total field with the scatterer in place, and then extract the scattered field from that?

When the scatterer is in place, PMLs are used to absorb the scattered field, so this simulation is only for a single particle on a substrate. My second question: How does one implement periodic boundary conditions for the scattered field, so that an array of particles can be simulated instead?

My third question: If, instead of a particle, I wanted to simulate the effect of a surface texture (like a 3D pyramid cut out of the surface of silicon), would the strategy above be applicable? I.E solve for the field in the absence of the pyramid, then cut the pyramid out of the surface and solve for the scattered field?

Thanks
-Kevin