Aluminum dimer electric field and mesh size

TaejunTaejun Member Posts: 1


I have a question about the E-field magnitude spectrum of the aluminum dimer. I used 25 nm radius and 35 nm height Al disks (material: CRC) with 3 nm Al2O3 layer (material: Palik). I observed that the FDTD simulations with the above 0.2 nm of the mesh size showed one peak at 345 nm in the electric field spectra. However, the simulations with the below 0.1 nm of the mesh size had two peaks at 326 and 366 m. I tried until 0.078 nm, and all the FDTD results with different mesh sizes showed the different intensities for the two peaks. A smaller mesh size than 0.078 nm could not be run because of the warning message related to the total memory requirements per engine. May I please ask how I could figure out which mesh size works best for me? Thank you.


  • kjohnsonkjohnson Posts: 230Ansys Employee

    Hello @Taejun ,

    Plasmonic simulations can often require very fine mesh sizes to resolve the field near the metal-dielectric interface. This will be especially true when you have a thin 3nm coating over the metal. I would recommend using a finite element solver like DGTD for a plasmonic simulation like this if you can.

    If the results are changing as you decrease the mesh spacing it usually indicates the mesh isn't fine enough (see convergence testing). How large is your mesh override region? It should only need to cover the metals with about 10-20 nm buffer around them, the entire simulation region doesn't need to be covered by the mesh override.

Sign In or Register to comment.