I am trying to perform similar analysis of SiO2 + graphene stack using FDTD Solutions as in the article Tunable spectral and spatial filters for the mid-infrared based on hyperbolic metamaterials. Mainly I would like to achieve similar results as in Fig. 2.
I was using the 2D graphene model with 2D FDTD region as shown below. There were 20 layers of combined SiO2 (150 nm) + 2D graphene. To perform analysis I was using the S parameters group from Object Library - with plane wave (Bloch/periodic) as a source. The monitors were placed 5 um from the centre of the structure. The FDTD boundaries were set to PML (16 layers) in x direction and periodic in y direction. Additionally there was mesh override region around the structure with dx = 1 nm.
FDTD Solutions view
Although I was able to obtain satisfactory results for transmission and reflection (at least their characteristics), the results for permittivity are far from being acceptable. Below I enclosed the neff result for stack with graphene chemical potential set to 1.5 eV. The Re(eps) plot starting at around 7.1 um looks fine, while below this value is “shattered”.
Could you please give me some hints on how to fix this?
Additionally I would like to perform angular analysis of the same structure (also in the same article). I have changed the angle theta property in the source and then tried changing FDTD boundaries from periodic to Bloch or changing plane wave type to BFAST, but to no avail. Either I did not get plane wave any more (for S parameters analysis) or the simulation was diverging. Could you also help me with this issue?
I guess there is some issue with the s-parameters calculation. I have tried to repeat the same simulation without any graphene layers - only 3 um slab of SiO2 set as a dielectric with constant refractive index = 1.45. The results of neff (both real and imaginary parts) for both Bloch/periodic and BFAST plane waves are below.
Bloch/periodic plane wave
BFAST plane wave
I still did not manage to run this simulation for different angles.