Discussion Forum

Hello all,

I'm new to COMSOL and FEM and I very much in need of help.
I need to simulate the 3D fields distribution in a ~250nm triangular hole in a silver/gold layer illuminated by a plane-wave or a gaussian beam.
I ran over some of the tutorials, but I don't understand a few things. I hope that someone here can enlighten me.

1) When do I need to use PML and when transparent scattering boundary conditions?
To my best understanding PMLs should "kill" the fields very quickly without reflecting. So why not use them all the time?

2) If I want to solve a scattering problem from a hole - can I use the "Solve fore: Scattered field" feature, or do I need to use a port? I know that for scattering from a particle scattering filed work pretty fine. Is that not the case for holes too?

3) As I said before, I want to simulate a plane-wave going through the ~250nm triangular single and/or double hole in a silver layer 200nm thick.
So far, I tried to run two models:
a. using a scattered field feature with plane-wave background and PMLs.
b. using a port and transparent boundary conditions.
In both models I used permittivity interpolation based on Johnson-Christy data, ot tried using the Drude model.
Both models showed suspicious results as can be seen in the figure attached.
1. I get light coming from the PMLs.
2. Fringes on the top and bottom metal surfaces. They are polarization dependent. I think they are surface-plasmons created on the edge of the metal slab, but we want to simulated an "infinite" metal layer.

4) Last question - I used a domain max probe inside the hole to record the maximum enhancement in the hole and got very unstable results when going to wavelengths smaller than 600nm. Any idea why?

If some can offer even partial answers to my questions it would be a great help.
Thanks.