As @mmahpeykar mentioned, please post your follow up questions directly into the forum so that it is searchable.
Regarding your inquires:
The “power” sweep for the FEEM simulations does not seem to be set properly. The index values are set to zero after a few points and it results in sharp drop of effective index. Here is the plot when the problem is fixed:
Here is the plot of effective index when the nk import is disabled:
The first mode has effective index of ~3.93 which matches with the power sweep results. This mode interferes with copper and you may try to select other modes of interest for your purpose.
I also recommend you to lock the mesh for both the HEAT and FEEM simulations. In HEAT this guarantees that sweep results have the same dimensions. In FEEM simulations, this guarantees that the change in effective index is due to perturbation in index rather than due to grid sizes.
Hope this helps.