Photonic crystal bandgap

Hi all,

I am new to the photonic crystal study. I am trying to simulate the band structure given in the paper “A proposal for optical WDM using embedded photonic crystal ring resonator with distributed coupling”. I am using 2D simulation (as the paper), period 552 nm and crystal radius=110 nm.

I have few questions.

  • Although the band gap generated from my file seems to have the same bandgap like fig. 1 (a) on the paper, the states looks very different. Below freq value of 0.3 I have only two states, while in the paper there are about 13 states. Why this is happening?

  • How can I generate the single mode waveguide (red line in fig 1 (a))

  • I can say that this is a bandgap for TM polarization as I am using vertical electric dipoles. Is there any way else I can use to determine if this is a TM or TE states (may be using mode monitors or other thing).

My project file is attached for one of my points.
band1_sweep1.fsp (7.5 MB)

My band gap structure for TM is

I took a quick look at figure 1 at the paper, looks like it may be referring the bandstructure plot of some structures with a defect. The simulation you attached does not seem to include any defect in it. Can you clarify that?

If you need to simulate bandstructure of devices that have defect, you can refer to this application example.

Similarly, I think you should be able to return the bandsturcture plot of a single mode waveguide - draw the single mode waveguide in the simulation

In this 2D configuration, I believe only TM modes can be excited. In 3D simulations, you can use boundary conditions to filter out some modes but I think this technique does not apply to your 2D simulation.

Dear @kchow,

Thank you for your reply. I got the idea now. Previously, I thought the band structure is not related to the defect. That is why I did not include the defect in the simulation. Obviously, this is wrong. I made a new file with the defect and made the simulation region as 2D rectangle that spans 6 rows below and above the waveguide width. the file is band1_sweep1.fsp (485.2 KB)

but still the band structure looks very wired. just straight lines!.

Can you help me with this.
I swept kx and ky value to follow the path Gamma-X-M.image

Sorry for the slow reply, we are very busy recently.

I have edited your file. I think your file is missing some settings to calculate bandstructure:

  • Sweeping the bloch vector: You will need to set up a sweep to sweep kx in the FDTD region, this is a key in bandstructure calculations as mentioned in the simulation methodologies

  • you should span the dipole sources and the monitors to cover some area of the simulation so to make sure all modes are excited, and include enough monitors to record signals

  • looks like some simulations are diverging, therefore I have used stabilized type PML and it seems to fix the issue. We have more info here if you would like to refer to ways to debug a diverging simulation.

The edited files are attached, results shown below. I think the plot below shows some similarities with fig 1c in the paper.


band1_sweep.lsf (2.7 KB)
band1_sweep.fsp (975.6 KB)


Dear @kchow ,

Thank you very much. It was of a great help indeed. My sweeping was the problem.

Another question please. The plot of the band structure in the following figures, how can I plot them in MATLAB. When I open it in MATLAB, all dots are connected, making noisy figure.

This topic was automatically closed 24 hours after the last reply. New replies are no longer allowed.