Learning Forum

[Handon Um]

I found a issue that the RCWA result of 5 um regular pyramid array shows a peak in reflectance graph when the relatvie interfaces are more than 15.

 

But, in the case of unit cell consisted of 4 regular pyramids, the RCWA results shows reflectance spectrum although it is different with the reflectance spectrum in OPAL 2 calculator from PVlighthouse.

Can you let me know the reason why or any solution?

[Guilin Sun]

Do you mean you set 15 interfaces and the result is larger than 1? 

Please check all the settings. I tested with 30 interfaces and it is more accurate. Please try 16 or 14 interfaces. 

To verify result, you can also use FDTD simulation. 

[Handon Um]

I tried with 30 ~ 100 interfaces in RCWA but the result was always larger than 1.

To verify the result, I tested the same pyramid structure using FDTD and the result was quite accurate.

[Guilin Sun]

There are other parameters that afffect result accuracy. Please try to usemore k vectors, eg more modes. The interface is to resolve geometry only. 

[Handon Um]

You replied to me that "Please try to usemore k vectors, eg more modes.". Can you explain specifically about more modes?

I couldn't understand well.

[Guilin Sun]

This is the convergence testing. How to you justify that the number of the k vectors and the mesh you use currently are good enough to guarantee correct results? you have to do convergence testing with higher number of modes (k vectors) and mesh size.

[Handon Um]

 

Thank you for your kind reply. I resolved the issue in FDTD.

But, in RCWA, I still can’t resolve the issue that the result shows larger than 1. 

I have tested regular array of other structures (eg. tapered wire, cone, pillar, rectangle, pyramid), and only the regular pyramid arrray showed larger than 1.

Is the periodicity of regular pyramid can cause an error in RCWA?

 

[Guilin Sun]

You will need to use moe interfaces and probbaly finer mesh, more k vectors. Please follow the RCWA example and do more tests.