Hi Mark,

Thank you for reading the article!

You are correct. I think I didn't write it clearly. How do you think about the following version? :D

===============================

The main difference between RCWA code and native model about handling of the electric field is as bleow:

(1) Native model rotate the electric field to ensure that k.E==0 keeps for the diffracted ray, where k is the propagation vector (L,M,N) , and E is the electrid field (Ex,Ey,Ez).

(2) RCWA code, on the other hand, comprehensively gives the diffracted ray's electric field, including ampltitude and polarization state.

================================

Best regards,

Michael

Hi Michael,

Sorry for the delay in replying, I must have missed your response somehow.

I'm still not sure what the difference is, however. What is 'comprehensive' about the RCWA result? Any ray with k.E=0 has amplitude and polarization. What specifically is different with the RCWA result?

- Mark

I think I have worked it out. In the built-in diffraction model, the user must define the diffraction efficiency directly, like so:

whereas with the RCWA model, the amplitude of diffraction is explicitly calculated. Is that it? And therefore, in the limit where you enter RCWA data into the dialog above, the two rays are identical?

- Mark

Hello Mark,

Yes! If we compare to table split, the advantages of RCWA are like below.

1. Table split only considers the amplitude, but RCWA directly gives data of (Ex, Ey, Ez) based on solve Maxwell equations for the given structures.

2. Table split always gives same amplitude no matter what incident angle and wavelength is with the incidnet ray. RCWA correctly considers incident rays' angle, wavelengths and electric field, and give corresponded diffracted electric fields.

Hope that clarifies, but please let me know if I can explain more!

Michael

Perfect, I think I've got it!

Thanks,

- Mark