I have a number of nonlinear waveguides, say six. I don't want coupling between specific two waveguides. Other than keeping them large distance away, what are the other possibilities to ensure no coupling between the waveguides? Will using a specific material composition of the waveguides help?
Thanks.
See the attached file. Here, you can see that the coupling between the WG1 and WG2, and WG3 and WG4 are not required, we need other coupling, say, between WG1 and WG3, and WG1 to WG4 etc.
Then, in this configuration, how to ensure no coupling between WG1 and WG2, and WG3 and WG4?
One way to reduce coupling is to have groove in between the waveguides......so that waveguides are having high contrast in the coupling side, and guided modes are having small or almost no evanescent tail to couple another wg through air.
For tightly spaced waveguides, we made trenches between them. It also helped with a polarization compensation. See the link to a patent below:
https://www.google.com/patents/US20040234223?dq=ininventor:%22Serge+Grabtchak%22&ei=VY_OU4eLHsfesASXrIBQ
@ Pranabendu Ganguly
I have limited knowledge about groove. Can you elaborate a bit on what this groove is
and how it can block coupling between waveguides?
Thanks.
Then you might think of getting high optical confinement to ensure that the mode overlap between the two is low, or design properly so that the propagation constants of the two waveguides differ
HIgh contrast waveguides have better confined modes. The trenches and grooves sort of make them higher contrast. Another thing you could try is to make each waveguide a little different so that their propagation constants are different. This will weaken the phase matching between waveguides and reduce coupling.
As I understand it, you have an array of six waveguides. You want them basically uncoupled, but you do want to couple between some of them, but not between adjacent waveguides? If this is the case, I have never done that, but let me brainstorm. You could make adjacent waveguides dissimilar to destroy the phase matching, but make the waveguides you want to couple similar. Now you have the problem of greater distance between the non-adjacent waveguides that also weakens the coupling. Perhaps there is some 3D arrangements of the waveguides that could make all this work, but that will be more difficult to fabricate. Maybe you will be forced to make them low contrast waveguides to get larger evanescent fields, but I would imagine that would also increase coupling where you do not want it. Maybe all of your waveguides are dissimilar, but you make clever use of gratings to strengthen the coupling where you want it. Given the propagation constants of two waveguides (b1 and b1), you can calculate the period of the grating that it would take to couple these waveguides as follows...
b1 - b1 = 2*pi / a
where a is the period of the grating.
Sorry I don't have a specific answer for you. Hopefully my random thoughts will spur more ideas for you.
Good luck!
Using high power femtosecond laser, you can write a hollow channel locally at the laser focal point. Writing this channel between two waveguides will reduce coupling, similar to grooves. I guess there would be a way to make those 3D channels around a waveguide you don’t want to be coupled. Note that those hollow channels create higher refractive index around. I guess this could be used for better coupling to the second next waveguide. I joined a quick draw. Advantages of this technique: would keep you sample surface flat and clean, and retain strength of your sample.
There are several options for you indeed. It all depends on availability, cost etc. Trenches worked for us in industrial environment so redesigning a mask on III-V wafer was the best reliable and cost-saving way for chip production. We had about 60 identical (some of them curved) waveguides in AWG fan out so decoupling all of them was of the primary interest.
Jerome Lapointe
Thanks for your valuable feedback for this.
Can you provide me some reference papers for your methods of blocking coupling between desired waveguides.
Thanks all for your feedbacks.
Can anyone give me some reference papers in this regard of blocking coupling between disired waveguides.
Search coupled mode theory (CMT) on google.
Per CMT, the energy will transfer 100% back and fourth between two coupled waveguides with the same propagation constant.
The energy does not transfer 100% if the two waveguides have different propagation constants. You can change the propagation constant by index of the material or waveguide geometry.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Optics