Negative losses dielectric medium


I make simulations in a material KYW the structure looks like as shown below

where the upper part is air with refractive index 1, red part KYW with refractive index= 2,
the green/yellow part an adhesive with refractive index of 1.56 and the lower green part SiO2
with refractive index of 1.44.

I do my simulations at a wavelength of 1.55 micrometer and obtain the results as shown in
the figure. There are negative losses which I do not expect to obtain because all the materials
are dielectric.

My simulation region is 12 micrometer wide and 8 micrometer high all with PML boundary conditions.
The gridsize is 20 x 20 nanometer

Can someone give an explanation why the losses are negative?



There can be some artificial loss or gain introduced when using PML boundaries with the FDE solver. For this reason we typically recommend starting with metal boundary conditions instead of PML. There’s a good discussion on this here:

I hope this helps!