Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
Calculation of scattering cross section from Au metal nano particle sitting over an si substrate
Posted 11.07.2018, 09:41 GMT-4 7 Replies
Please login with a confirmed email address before reporting spam
Hi All,
Myself, Achyut Maity from Ben Gurion University, Israel. I am beginner in comsol simulations. Recently I am trying to calculate the scattering and extinction cross-section from an Au sperical particle sitting over a Si slab to mimic the dark field scattering spectrum. The modelled simulation is attached here. Her I have tried to calculate the scattering extingtion cross section of the Au particle in the upper portion of the modelling, i.e in the air domain. Here the light is incident in the particle from the positive z to negetive direction. Although the simulation is running, I can not obtain the desired spectrum. Instead of I am getting some noise like response. I know very well that such request definitely hamper your busy and tight work schedule, but considering the academic prospect of an young student I hope to get your positive response on this aspect.
thanks in advance
Attachments:
Please login with a confirmed email address before reporting spam
Achyut,
To solve this type of the problem you should use two (emw) physics interfaces: first (emw) physics solves for the background field in the air+Si substrate without particle using Full field formulation, the secon (emw) physics solves scattering problem using Scattered field formulation with background field taken from from the first (emw) physics. This approach is described in details here: https://www.comsol.com/model/scatterer-on-substrate-14699
Regards,
Sergei
Please login with a confirmed email address before reporting spam
Dear Sergei , Thank you very much for your promt and insightful reply. I was trying to build my modelling using the above mentioned link. But to run this example it requires the wave optics module what we do not have. Can we run such type of example RF module? Or is it possible to modify the example for RF module? Eagerly waiting for your reply. With best regards Achyut
Please login with a confirmed email address before reporting spam
Achyut,
You don't need to have Wave Optics module to set up this problem. Functionality of (emw) is almost the same as (ewfd). There is slight difference in the default setup: (emw) default dielectric properties are defined in terms of relative permittivity, while (ewfd) default properties are defined in terms of refractive index. You can safely follow step by step instructions to create this model in RF.
Regards,
Sergei
Please login with a confirmed email address before reporting spam
Hello Sergei, Thank you very much for your reply. I have tried to build up the the modeling using the pervious example. The new .mph file is attached here for your kind perusal. It is still not working. Now it is showing an error messege: ***Failed to find a solution. Singular matrix. For mesh-case 0 there are 10456 void equations (empty rows in matrix) for the variable comp1.E210. at coordinates: (-3.21429e-007,-4.75e-007,-3.25e-007), (-3.21429e-007,-5e-007,-3.25e-007), (-2.67857e-007,-4.75e-007,-3.25e-007), (-2.67857e-007,-5e-007,-3.25e-007), (-3.21429e-007,-4.5e-007,-3.5e-007), ... and similarly for the degrees of freedom (empty columns in matrix). Returned solution is not converged. Not all parameter steps returned.
I can not find out the issue. Can you please help me to solve the problem?
Attachments:
Please login with a confirmed email address before reporting spam
Achyut,
There are several issues with your setup: PML references wrhong physics, Port 2 has wrong refractive index, setup for Periodic Conditions is valid for normal incidence only, wavelength is not propertly resolved in the substrate. Attached model is running without errors. However, I didn't try to check results.
General advise would be to start with reproducing tutorial model, make sure you are getting consistent results and comfortable with model setup. Only after that you can move on by changing geometry and materials to get solution of your problem.
Regards,
Sergei
Attachments:
Please login with a confirmed email address before reporting spam
Dear Sergei Thanks a ton for your enormous help. It is running. But still I am not able to find the sacttering response from the sample. I think now I can solve it. But I have only a question regarding to the modelling that you have sent to me last week. actually you have defined a parameter namely, "nel=6" in the parameter section and then redefine the mesh size using this parameter. Can you please discuss what is the underlaying concept of taking this parameter? and how do you put the numerical value upto 6? It will help me lots. thank again for your continous support and help. With best regards Achyut
Attachments:
Please login with a confirmed email address before reporting spam
Achyut,
To resolve wavelength you have to have at least 5-6 second order elements per wavelength. Wavelength depends on frequency and local material properties. In your case, substrate has large refractive index (n=3.4), meaning that mesh in the substrate should have at least 6*3.4=20 elements per wavelength.
Regards,
Sergei