Hi,
I am trying to simulate chiral media as explained in "alphard.ethz.ch/hafner/Workshop/Comsol2009.pdf". My understanding of adding the chiral expression for the electric displacement is that I add a wave equation to the EM physics and then select my domain (in the middle of the structure) and then change the equations for that domain. I found the electric displacement equation and changed it accordingly for all three E components, unlike what is shown in the presentation here (only one equation). First of all is this correct?
Secondly once I simulate it for linearly polarized light, what I find odd is that the E field before reaching the chiral domain has also been affected. What could I be doing wrong? I found the equation for the B field it was curl of E, and hence I assumed no change was required to it as it would be calculated in accordance to the changes made to D. Could this be the problem?
I do not use COMSOL in my simulations. So I can not really be helpful in this area. Nevertheless, I suggest you do your simulation with 3D-FDTD. Calculations will be longer but you'll get good results.
Hi,Affar. I have also read the corresponding paper of modeling the chiral media using COMSOL. And i have been trying to reproduce the results in the mentioned PPT file, but i haven't succeed in making it out. I have made a success in calculating the modified weak form for the chiral media, but the resulting boundary condition, i.e., adding the surface current on the interface between the chiral domain and its adjacent domain, can not been implanted in COMSOL RF module. Did you have fix this problem?
Dear Haizi,
Well we went on a different approach which was to define the material as anisotropic and implement customized permittivity tensors. We basically define the values for epsilon xy and epsilon yx as a complex variable. If you look up permittivity in wikipedia, then under the heading Tensorial permittivity, you will see a matrix. Here e2 are the off diagnol components mentioned above. Based on the ratios of the real and imaginary part you can control ellipticity, and the resultant value will control the ORD. I hope this help.s
Dear Affar,
Thank you for your response. You said that the chirality (or ellipicity) can be controlled by adjusting the ratios of the real and imaginary part of the epsilon xy and epsilon yx. But i doubt that because the anisotropic material is different from the bi-anisotropic (or chiral) material for which the electric displacement D and magnetic flux B are not only dependent on the electric field E but also the magnetic field H. Therefore, the chirality does not just mean that the material is the spatially dispersion that can be described by the permittivity tensors.
I think the approach that modifying the Helmholtz equation for chiral material is the right way, as mentioned in reference of " Klimov, V. V., et al. "Eigen oscillations of a chiral sphere and their influence on radiation of chiral molecules." Optics express 22.15 (2014): 18564-18578."
Unfortunately , i can not reproduce the results in the above paper. I appreciated very much if you can have a try using this approach and share your results with me. Please don't hesitate to contact to me if you need my semi-successful COMSOL model via the following email address: [email protected].
Dear Haizi,
Thats really interesting. These kinds of models are new to me so I appreciate the useful information. We have a student here who is looking into the matter, I shall ask him to consider this further. If we find anything useful, I shall let you know.
Dear Affar,
How about your progress about modeling the chiral media using the COMSOL software? I have succeed in replicating the Figure 7 in the paper of "alphard.ethz.ch/hafner/Workshop/Comsol2009.pdf", but i can't replicate the Figure 5. I don't know why, maybe my model is not completely correct? Besides, i found that the model geometry is not as what described by the author in the paper. when have a close view on the Fig.7 we can easily find that the modeled chiral sphere has a shell layer, however the author state in the section one that the chiral sphere is modeled as a uniform particle of a effective parameters. Though i also have some tried in modeling it with a shell layers, i have be successful in replicating the Fig.5.
Did your group get some substantial progress in the problem? it will be appreciated if you can share your valuable experience with me. And please don't hesitate contact me if you need my COMSOL model. My email address is : [email protected].
Best wishes !
Haizi Yao
Hi Haizi, afraid I have not progressed with it, I am caught up in nano fabrication. I am uncertain what paper you are talking about as the link you sent is the comsol pdf. Could you tell me exactly what paper it is? I hope to continue this work in the summer and would be happy to collaborate with you on this. I will email you with some more details on how we could proceed.
Hi Affar,
I am pretty sure that you have already solved your problem, however, for future researchers, I have prepared a text file which I hope it helps them in simulation of chiral media. I will attach it here since this question may be followed easily by a google search.
Best,
Thanks Mohammed. We have indeed and I have the details in our latest manuscript. Your document is quite similar and very useful. I shall upload or link the paper once it is published. Haizi, I believe I had emailed you earlier, but in any case you also seem to have solved the problem. Thanks for your input.
Dear Somayyeh,
The anisotropy is done by changing the tensors and the chirality is added by altering the equations. Please have a look at our latest paper (supporting info) where the procedure for altering the equations is explained..
10.1021/acsphotonics.7b01071
Article Controlling Metamaterial Transparency with Superchiral Fields
As an update to the wonderful white paper that Mohammad Albooyeh shared above, there are a few typos in his pdf worth mentioning. In equation 3 and 4 there is a cos(k*L) term in the numerator that should be deleted. Fig 1 should also be updated, and I have attached a corrected version. Note that this is amplitude, not intensity, and S31/S41 are the simulated results. This particular figure was done with a mesh of lambda/10, but I was able to achieve comparable results with a mesh as coarse as lambda/5.
- Badges
- Science topic