I had a chance to look at your file and I it is not clear what you are hoping to achieve. Could you please share more information on this topic? The FDE solver is looking for guided modes at it seems that you are using a 2D rectangle based on sampled data. This would have been calculated based on transmission through a thin film? What sort of guiding behavior are you expecting? Do you have any sense of what the effective index might be? Is the orientation correct? What substrate are you using?
The modes that the FDE solver is finding are non-physical in the metal case although they are valid solutions to the eigensolver. The simulation is not seeing the 2D structure, you can check this by disabling the WS2 rectangle and running the solver again for either case.