I try to simulate a periodic structure in IR frequency range by " electromagnetic waves, frequency domain " physics of COMSOL but recently I was find that the Vacuum thickness has effect on transmittance result, significantly. according to the image below, maybe this means that distance between Ports has cause resonanse. But there is a Vacuum (n=1) between the Port to the Sample surface (Common Polymer, n=1.5), and so we expect the distance to the Port not affect on the transmission reduction. What's wrong with this? What is your opinion?
I am neither familiar with COMSOL nor the problem that you are trying to solve, but your description of the observed anomaly (dependence of transmittance on the vacuum thickness) suggests that it could have something to do with the boundary conditions (there are reflections where there should not).
Is it possible?
Hello Elham,
I agree with Daniel. Can you tell us something more about what it means for a Port to be ON or OFF? In other words, when a Port is OFF does that mean that it acts like a mirror? And if a Port is ON does that mean it is transparent with a n=1? If that is the case, then the IR light will experience constructive or destructive inference as it goes from the top Port through the vacuum-polymer-vacuum sandwich and is reflected back through the sandwich by the bottom Port.
A clear explanation of how the Ports behave would be very helpful to anyone trying to troubleshoot your problem.
Regards,
Tom Cuff
Dear Daniel Poitras and Thomas Cuff
I appreciate for your answer.
Now, I'm going to describe problem:
Do you know about FTIR Spectroscopy? In this device IR beam is radiating from source to the IR transparent sample and then the detector is collecting transmitted beam. I want to simulate this process.
When a Port is ON, the electromagnetic wave excitation at it is ON. Also we can define input power and electric mode field amplitude (for example: x, z= 0 , y=1 V/m) for it. But in other port wave excitation is OFF, No power definition but electric mode field same as before. According to COMSOL Documentation, this port is necessary to S12 definition and calculation. But not any evidence about Port's material and behavior. I just know it doesn't consider as PEC or PMC in boundary conditions. I think your questions are my question too and the answer of them will can solve the problem.
Best Regards,
Elham Tavakkol
Thank you for the more detailed description.
In addition to the boundary conditions, I wonder: In your simulation, do you want planewaves propagating along the Z-axis? If yes, are you sure that the parameters for the wave excitation give a planar wave?
If for some reasons, the Port ON gives waves that are radially propagating (i.e. a point source on Port ON), then I suppose that the thickness of the vacuum layer should affect the transmittance observed at the Port OFF plane. Just an idea.
Hello Elham,
Thank you for the description. It helps because I am not a COMSOL user.
According to p. 42 of the COMSOL RF Module Manual, "... all ports are assumed connected to matched loads." Therefore, there are no reflections at ports. I thought the Ports were the problem, but they are not. Your input Port (the one that is ON) defines the applied electric field and input power. Your output Port (the one that is OFF) receives the transmitted EM energy passing through the vacuum-polymer-vacuum sandwich, which allows you to calculate the S12 scattering parameter, i.e., the voltage transmission coefficient from the input Port to the output Port. But the two vacuum-polymer interfaces will cause reflections and phases changes. I believe it is these two interfaces that are responsible for the result you are seeing.
I will have to look at some of my optics books, but I believe it is the interfaces that are the problem - as first mentioned by Daniel - not the Ports. I will send you some citations to some books that might help, as soon as I can find them.
Regards,
Tom Cuff
Hello Elham,
Here are two references that have examples that support the idea that the interfaces between materials of different index of refraction (n) cause reflections and so decrease the transmitted radiation above and beyond what is lost due to absorption in your polymer layer:
[1] L. Holland; Vacuum Deposition of Thin Films; John Wiley & Sons, Inc.; 1961; pp. 270-297, Chapter 9 - Optical Interference Films.
[2] O. S Heavens; Optical Properties of Thin Solid Films; Dover Publications, Inc.; 1955; pp. 46-95, Chapter 4 - Thin Film Optics.
Regards,
Tom Cuff
Hello Elham,
Consider the following simple example: a layer of glass sandwiched on either side by air. Each air-glass interface reflects about 4% of the incident perpendicular light for a total of approximately 8% reflectivity. Air has approximately the same index of refraction as vacuum. In any event, the index of refraction of glass is significantly higher than that of air just as the value of n for the polymer is significantly higher than that of vacuum, which is my point.
[1] Heinz G. Pfaender, Huber Schroeder; Schott Guide to Glass; Van Nostrand Reinhold Company; 1983; p. 65.
Note, the link to the COMSOL RF Module Manual is as follows: https://is.muni.cz/el/1431/podzim2013/F7061/um/RFModuleUsersGuide.pdf
Regards,
Tom Cuff
Thank you Dr. Daniel Poitras
Yes I want; and I was sure before you tell that. If you think the parameters are incorrect, how can they be corrected?
Maybe you are right. I'll share your idea with my professor.
And another point: ports are consider as periodic port and periodic structure as floquet periodicity that k-vector define from periodic port.
- Badges
- Science topic