Learning Forum

[amerturan]

I am a Lumerical license user and I am trying to simulate a hexagonally structured gold sphere. Polystyrene spheres with a diameter of 260 nm are placed on a glass substrate in a hexagonal pattern using colloidal lithography. Then, the structure is covered with approximately 60 nm of gold. SEM images of the experimental results are attached.

I would like to show this structure in an FDTD simulation. I have prepared the FDTD simulation as attached, but the transmission is not accurate, especially when we coat more than 15 nm of gold on the glass, the transmission drops to zero. What are we doing wrong in the simulation?

Could you please advise us on what adjustments are needed? Your help would be greatly appreciated.

Best regards,

 

https://drive.google.com/drive/folders/1_j-GB2MDymxjj1j9LvHj0ot_yUxEIxGP?usp=sharing

[Guilin Sun]

Please note that the forum policy does not allow any of us to download your files.

As long as the coating is correct, and the simulation uses periodic BCs with plane wave, and runs sufficiently long, the result should be correct. You ca use index monitor to check the structure.

Did you use single wavelength or broadband? single wavelength can be difficult to know if the simulation is correct or wrong. I would suggest that you simulate a broadband source, even though you only need one wavelength. This is because, from the broadband , you will know if the transmission curve is smooth. You can use "wavelength start" to be the wavelength you desire, and use larger "wavelength stop" at source spectrum.

You can also use a time monitor to see if the signal decays completely, or check the log file to see if the simulation reaches its autoshutoff min. Sometimes you may need to reduce the autoshutoff min.

One more note: since the coating is only 15nm thick, and it is spherical, you will need mesh override to resolve the coating. you ca try about 3nm or finer mesh. Make sure the mesh is also an integer fraction of the periodicity, eg, the override mesh size is Period/N where N is an integer. 

[amerturan]

Hello,

I couldn't solve the problem after my last message. I think I probably made a mistake in choosing the mesh. Can you please help me with this? The structure is as follows:

1- Glass is used as substrate
2- Polystyrene spheres are placed on the glass.
3- The surface with these polysterene spheres is gold plated. In order to do this in the simulation, I added another sphere on top of the polystyrene spheres and obtained nested spheres.
4- Finally, there is a gap between the spheres on the glass. In order to fill this gap, I covered the entire surface with a golden structure on the glass.

I am attaching the side and top view of the design I mentioned as a picture. Please here;

1-Substrate
2-Polysterene spheres
3- Au sphere coatings
4- Gold plated on glass substrate

Can you tell me what should be my mesh order for structures? Also, what boundary conditions should I choose?

[Guilin Sun]

For such question specific to mesh orders for structure, it is better to have a new, dedicated post. Please do write a new post next when your question is not exactly the same as the origial post.

From what you described, the overlaps happen among the polysterene sphere, au coating sphere and a gold thin film.

Obviously the gold thin film should not have the highest mesh order.The polysterene  sphere should. The au-coating shphere should have lower mesh order since it has to keep the polysterene sphere well. Regard to the substrate, if any of the above mentioned 3 geometries penetrates into it, the substrate should have the most-highest mesh order, as the sphere does not penetrate in experiment.

In mesh order, the small the number, the higher the priority. Thus from my experience in the lab, I recommend the following:

substrate, the default mesh order 2;

polysterene  sphere, mesh order 3;

au-coating sphere, mesh order 4.

gold thin film, mesh order 4;

Depending on the process techniques, you may change them. 

Usually the polysterene sphere may not well maintain its spehre shape on the substrate by the process. It can be cut a little by the substrate, or by another residual material in the process.

 

Please use index monitor to check the resulting index profile.