Sorry to bother you. I had a couple of questions about modelling dipole emission in 2D materials using lumerical. I have been playing around a little with some simple dipole simulations - but I have been struggling a little to understand how to set things up so that I am convinced that lumerical is even noticing the very thin layer of 2D material - (ie so the mesh is small enough around it) - without making the simulation time so long as to make simulating the more complicated structures I actually want to model difficult.
I guess that the actual material properties of the TMDCs might cause a problem as I go down to small sizes - but for the moment I’m happy with fudging the material database.
I am currently attempting to model a hexagonal boron nitride layer on various materials eg SiO2 on Si- and trying to calculate the collection efficiency - and the quantum efficiency/Purcell factor. Since I’m aiming simulate structures later which will have Purcell enhancement - and since I’m going to be overiding the mesh size to force lumerical to see the hBN I guess I may need to use transmission boxes to calculate the dipole output (rather than dipolepower as discussed in various other posts) - Purcell factor: dipolepower vs. transmission box
I have arbitrarily settled on a hBN thickness of 4nm (which I think (?) in terms of the material is roughly 10 layers) - though probably being able to vary the thickness would be a good idea.
My question(s) at the moment are:
If I want to create a transmission box - does it have to be all inside the hBN? - If so that would mean that the mesh size should be at most 0.2 nm to give 10 mesh boxes inside the Tbox. I guess if I can allow the T box to extend outside the HBN then it would allow me to use a coarser mesh.
Does the T box need to be a cube? ie. Can I get away with a larger TBox in X and Y than in Z - and then also be allowed to have larger mesh overide size in the X/Y directions.
I noticed that there were some layered stack analytical tools which might be helpful as well for this project - perhaps doing FDTD on such thin layers isn’t ultimately a very sensible way of modelling things - though I think they are not on the standard licence. I was wondering whether it was possible to arrange to test them out (and maybe get an idea how much extra they are) - I think our current set of licences are up for renewal in about a month.