Learning Forum

[marius.crouzier]

Hello,

I am using the sript Zemax_interoperability_metalens_unit_cell and I am trying to adapt it to different nanoparticles with different substrates (gold nanoparticle on Si/SiO2 substrate). But, for some particular value of period or size of nanoparticle, the simulation diverged or the autosutoff level was oscillating.

I decided to change the boundary condition with PML and still using the symmetry.

However, I am having some issues to calculate the reflected power by a specific grating order, it differs very much with the total power reflected. In the following picture R stands for the total power reflected and R_Gn for the power reflected at the order 0.

[Guilin Sun]

For the diverging issue at certain period or nano particles, it might be mainly due to larger higher order diffraction angle. Please use a visualizer to check it farfield projection roughly, using log scale to display. You can also use grating analysis function to get the diffraction angle. If it is broadband, please only check the frequency points that have strange reflection or transmission, to save time. Another way to check is to plot the diffraction orders. If the diffraction order changes from one frequency to another, it may mean there are frequency that has diffraction angle around 90deg.
If the diffraction efficiency at glazing angle is high, it might be very challenge to remedy. In general, you can
1: use steep-angle PML with sufficient number of layers
2: increase the distance between PML and the structure.

AS for how does gratingpolar - Script command work, it just decomposes the output fields. You can begin with grating equation and one you know the diffraction angle, you can do an overlap of the outputfield with the plane waves.