Adding ellipticity to a wave plate



Hello everybody,
at the moment, I would like to design a polarization rotator in Interconnect based on wave plates. In real life waveguide polarization rotators have an elliptical eigenvector. The simplest way to include this information in spherical coordinates would be to give Chi a certain value of the building block.
I think Psi is related to the circularity and the retardation is related to the orientation of the wave plate. But is there a way to give Chi a non-zero value? Maybe you can recommend a more suitable element to create such realistic polarization rotators.

Thank you for your answer!
Axel Schönau


Hi @axel.schoenau,

Sorry for the late reply. Do you want to actually design the polarization rotator or just want to use it to rotate the light polarization in circuit? There is a Polarization Rotator element in the library and you may want to use it instead. The WavePlate element is based on an analytical model and the only parameters you can define are the orientation and circularity.


Hi @gwang,
Actually, I’d like to have an element, where I can vary delta, Psi and Chi independently. At the beginning of our simulations, we took the element which already covered the most of the three variables. In this case, it was the waveplate (delta and Psi).
Our monolithically integrated waveguide structure can be used as a classical polarization rotator but gives also additional ellipticity due to the two different waveguide sections which form our polarization rotator.
That is why I am actually looking for an element, where a can consider such effects in the way of giving Chi a certain value.
I keep you updated when I get more concrete information.

Thank you and have a nice week,
Axel Schönau


Our wave guide structure behaves like a circular retarder. Waveplates are actually linear retarders, so only a certain combination of two wave plates leeds to behavior like elliptical or even circular retarder. At the moment, I did not manage to combine two waveplates to get a circular retarder. It is not as easy as explained :grinning: .

It would be much easier to assign an element a certain mueller-matrix, which includes all information about retardation, diattenuation and depolarization. Is there any possibility to fill an element with a mueller-matrix like s-parameters for MMIs?

Thank you for your help!
Axel Schönau


Hi @axel.schoenau,

Sorry about the late reply. I will file a feature request on this and let you keep you in the loop :slight_smile:


Hi @gwang,
At the moment, I found the solution with two wave plates to create a circular retarder. My calculations were based on mueller-matrices and thus a appreciate your feature request.

Thank you and have a nice week!


Hi @axel.schoenau,

Sorry for the late reply. I’ve filed this feature request for you and it is under review right now. I will keep you in the loop on this. :slight_smile: