## Photonics

#### 2D vs. 3D Optical Simulations for Pyramid Solar Cells

• spacelover17
Subscriber

I am trying to modify this pyramid solar cell example for my work, and I was struggling to understand why the electrical simulations could be made 2D but not the optical simulations. I know the page talks a bit about why the electrical simulations can be 2D, but it's still a bit unclear to me, and more importantly I am confused about why the optical simulations can't be 2D (or what would be the discrepancy if I DID make them 2D). I know I can change the simulation to 2D and run it to demonstrate if there is a discrepancy, but I want to better understand WHY.

• Guilin Sun
Ansys Employee
For FDTD simulation, the geometry is important, as the simplified 2D means the 3rd dimension is infinite, which is not true in this case.
CHARGE is a different story, as the current is collected by summing up the charges. This is why it can use 2D simulation: it sums up the generated charges in the 3rd dimension, and the total charge flow is counted for the current. Theoretically there is minor error doing so, for example the recombination loss might be different as it depends on the charge density and paths. However such error is very small so you can neglect it. You can compare the results between 2D and 3D in CHARGE. But in FDTD 2D and 3D are totally different.
I hope this helps.

• spacelover17
Subscriber
So the generated charges in the third dimension aren't infinite even though in FDTD the third dimension is infinite? (I guess the symmetric / anti-symmetric boundary conditions make it that way?)?
And is this idea true for all simulations, that we should not use 2D simulations for FDTD? Or are there situations where we can? I seem to remember a planar example that used a 2D FDTD simulation, so why is that okay?
• Guilin Sun
Ansys Employee
In CHARGE, the geometry is not that important compared to FDTD, it is the total charges that matter. You may find some basic information on the current and charge in textbooks.
Yes, you are right that, in general, 2D FDTD is rarely used, since objects always have limited size in 3rd dimension. Planar case is a perfect example for 2D, since it assumes periodic, eg, infinitely large size compared to limited size along the layers.

This image may show the difference between 3D and 2D:
in 2D simulation, the 3rd dimension is assumed to be infinite.