Slow-light propagation in plasmonic system


Dear engineer, Figure 1 illustrates the plasmonic waveguide composed of an insulator layer sandwiched between two different semi-infinite metal claddings (the upper layer is HMM1 material, the middle is si3n4, then the substrate is HMM2).

To investigate the time evolution of the pulse propagating through the plasmonic waveguide system, the Gaussian-pulse source located in the centre of the insulator is set to excite the TM mode, with the center wavelength of 1550nm, 50nm bandwidth. We want to get the signal results of propagation distance 80nm, 120nm, 150nm, 180nm, 200nm, similar to these articles (see the appendix). And my model is also shown in the appendix.

slow light propagation(20170525).fsp (6.4 MB)

The the Gaussian-pulse propagates in the plasmonic waveguide

Dear @liangshuhai

Thank you for reaching out and providing details.

I have a few questions for you to clarify the case a bit more. What are the P and a0 values in the second figure? And also how the TM mode profile from mode source look like? Currently it does not look like the mode is confined in Si3N4 layer. Also, it looks like you missed to provide the link for the articles.

Other than I did some minor modifications in your simulation file:

  • Since HMM layers are semi-infinite, I changes BCs to PML from periodic.
  • I increased rectangle objects span to extend through PML layers.
  • I modified mesh override region as well. I think mesh override region is only required for Si3N4 layer.

Here is the simulation file for for your review:
slow light propagation_modified.fsp (267.7 KB)

Please keep me updated with your thoughts and I will be glad to be of a help.



Thank you,engineer, and the answers are as follows
First:why are the P and a0 values in the second figure? P means the propagation distance of the signal and a0 is just length, about 0.891um

Second: how the TM mode profile from mode source look like? like this,notice it shows the magnetic field

Third: the articles:
ref 1
ref 2

and i have one quetion for you in the simulation file modified by you;
if you changes BCs to PML from periodic, the E field of signal is not Gaussian- shaped.
Thank you very much


and the mode is confined in Si3N4 layer: HMM is like metal and the SPPs mode exsits the interface between the HMM and the core


Dear @liangshuhai

Thanks for providing additional information.

Periodic boundary conditions are used when your structure is periodic. In this case, since you mentioned HMM planes are semi infinite, you need to use PML boundary conditions.

I modified the simulation and monitor locations for the new a0=891nm. Below is the mode profile I found from mode source:

Since this mode is very lossy, field decays very quickly. The plot below shows the electric field profile along the x-axis (propagation direction):

I am not still quite sure if we are injecting the proper mode? Also, can you elaborate more on what you mean by Gaussian pulse source? We have a mode source here and its profile, as is shown in the plot above, is not Gaussian. I took a look at the references and it looks like they have used a Gaussian source to excite TM modes. More elaboration on what was exactly injected into simulation region is required.

slow light propagation_modified (1).fsp (863.0 KB)



dear engineer, what do you think the Gaussian source? can i use the Gaussian source in FDTD


Dear @liangshuhai

We have Gaussian source that you can use in FDTD simulations. You can either use scalar approximation or thin lens:

To properly set the Gaussian source, we will need more information, such as beam waist or lens NA.

In your case, I do not think that you need a Gaussian source. If we know what modes were excited in the paper, we could simply use mode source. Please bare in mind that the mode profiles will not be necessarily Gaussian.

Please let me know of your thoughts and I will be glad to be of a help.