Persistence of oscillations at higher wavelengths

We are analyzing the absorption enhancement of an a-Si PV cell due to the presence of Ag nanoparticles embedded periodically within the cell. The resultant absorption graphs obtained exhibit significant oscillations at higher wavelengths. We previously asked for a solution to this on KX, and the solution provided was to place the source within the Si cell. However, we do not find this a feasible solution as it does not replicate the original conditions. We assumed that the problem was due to the coherence length of the incident radiation being larger than the cell thickness, and thus increased the thickness. However, the oscillations still exist, regardless of what we do. We would like them to be as non-existent as possible.

Here is the fsp file: plasmonic_in.fsp (268.7 KB)

Here is one of the obtained absorption curves:


The oscillations seem to be a physical effect due to reflections between the silicon surface and the array of nanospheres. As you mentioned, moving the source into Si will remove these oscillations but then you are not simulating the reflections of the silicon surface. My suggestion if you want to simulate the entire thing is to take into account the coherence length of sunlight, which is definitely less than the 10 microns it has to propagate inside silicon before reaching the nanospheres. This can be done as a post-processing steps as explained in this post:

Hope this helps!