help for metal-dielectric structure

These days I'm using frequency-dependent FDTD to study the behavior of surface plasmon polaritons. I first consider a very simple case, that is, a 2D air hole perforated in a thin metal film. The radius is about 20 percentages of periodicity. The grid size is about lambda/20 and time-step half of the grid size. Two power monitors were put on both sides of metal film (far away) to measure transmission and reflection spectra. A downward planewave source is normally incident on the metal surface. If the metal thickness is about 100 nm, two monitors can detect the outgoing energy. But If I increase the metal thickness to 150 nm, the results seems scattering. Do anyone meet such problems and how to resolve it? Thanks a lot.

May be 150nm thick metal is large than the plasmon decay deepness, I want to simulate the surface plasmon too, may I ask you some question about how to simulate the surface plamon?