Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.

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.

Transmission spectrum for optical waveguide

Frédéric Peyskens
Hello,

I am currently trying to extract a transmission spectrum from a Si waveguide on top of SiO2 substrate. There are some gold particles on top of the Si (this configuration is taken from an article in which they simulate it with FDTD). I tried several things to extract the transmission spectrum:

1) S21 parameter in dB
2) ratio of the Poynting flux at the input and output (tried both with fields Ex,Ey,Ez, emw.Hx, emw.Hy and emw.Hz and with fields from the Boundary mode electric/magnetic field)

I am however not able to get the right profile (for some wavelenghts I even get quit high gains which is unphysical).

It seems that also reflections are taken into account, but in an experiment you will only take the ratio of the actual input power (which is forward propagating) and the forward propagating output power.

Does anybody know how to extract a sensible transmission spectrum (which you would obtain in experiments).

Thanks in advance.

F. Peyskens

9 Replies Last Post 2013年3月20日 GMT-0400上午8:17
Alexander Kuznetsov
Posted: 5 years ago 2012年8月9日 GMT-0400下午2:53
plasmonic wire grating example is a good start. In general, the approach shown there gives the right reflection/transmission.
plasmonic wire grating example is a good start. In general, the approach shown there gives the right reflection/transmission.

Frédéric Peyskens
Posted: 5 years ago 2012年8月10日 GMT-0400上午4:44
Thank you for the answer. I tried however a similar reasoning with the waveguide: input of the waveguide is port 1 and output is port 2. Then I extract the S21, but this gives me not the right transmission spectrum. I tried it through the z-component of the Poynting vector (as the waveguide propagates waves in the + z-direction), but this also does not give me the right results. In case of this 2-port system I would expect that the transmission is defined as

(Power propagating in the + z-direction at the output port)/(Power calculated in the Boundary Mode Analysis at the input port)

Or am I seeing this wrong? At least, this is what you would measure experimentally.

Kind regards,
F. Peyskens
Thank you for the answer. I tried however a similar reasoning with the waveguide: input of the waveguide is port 1 and output is port 2. Then I extract the S21, but this gives me not the right transmission spectrum. I tried it through the z-component of the Poynting vector (as the waveguide propagates waves in the + z-direction), but this also does not give me the right results. In case of this 2-port system I would expect that the transmission is defined as (Power propagating in the + z-direction at the output port)/(Power calculated in the Boundary Mode Analysis at the input port) Or am I seeing this wrong? At least, this is what you would measure experimentally. Kind regards, F. Peyskens

Alexander Kuznetsov
Posted: 5 years ago 2012年8月10日 GMT-0400下午1:14

input of the waveguide is port 1 and output is port 2.

port 1 is input as well as output for the reflected wave.
abs(S11)^2 would be reflectivity coefficient, and abs(S21)^2 - transmission.
To validate your model you can use it to calculate reflectivity (or transmission) of SiO2 film and compare to theoretical value.
[QUOTE] input of the waveguide is port 1 and output is port 2. [/QUOTE] port 1 is input as well as output for the reflected wave. abs(S11)^2 would be reflectivity coefficient, and abs(S21)^2 - transmission. To validate your model you can use it to calculate reflectivity (or transmission) of SiO2 film and compare to theoretical value.

Frédéric Peyskens
Posted: 5 years ago 2012年8月12日 GMT-0400上午6:17
Hello,

Thank you for your fast response. I have tried what you suggested, but the problem with the S-parameters is that in transmission it gives me large gains for some wavelengths, which is unphysical. I would try to extract transmission through the Poynting vector, but I only want the forward propagating power at the output (and not the reflected power going back to the input as you mention) and I guess part of the reflected power is taken into account too. Is there a way to extract:
1) the excitation power at the input (and not taking the reflective power into account)
2) the forward propagating output power

Thanks in advance.

Kind regards,
F. Peyskens
Hello, Thank you for your fast response. I have tried what you suggested, but the problem with the S-parameters is that in transmission it gives me large gains for some wavelengths, which is unphysical. I would try to extract transmission through the Poynting vector, but I only want the forward propagating power at the output (and not the reflected power going back to the input as you mention) and I guess part of the reflected power is taken into account too. Is there a way to extract: 1) the excitation power at the input (and not taking the reflective power into account) 2) the forward propagating output power Thanks in advance. Kind regards, F. Peyskens

Posted: 5 years ago 2012年8月29日 GMT-0400下午2:45
COMSOL's sign convention for the imaginary part of dielectric constants is opposite to that of most other photonic modeling software. Your non-physical gain might be caused by this.
COMSOL's sign convention for the imaginary part of dielectric constants is opposite to that of most other photonic modeling software. Your non-physical gain might be caused by this.

Frédéric Peyskens
Posted: 5 years ago 2012年9月12日 GMT-0400上午4:29
Thank you for this response. We got the same answer from the comsol helpdesk and they now found indeed a loss instead of a gain.
Thank you for this response. We got the same answer from the comsol helpdesk and they now found indeed a loss instead of a gain.

Posted: 5 years ago 2013年1月4日 GMT-0500下午11:18

Thank you for this response. We got the same answer from the comsol helpdesk and they now found indeed a loss instead of a gain.


Hi, I have meet similar problem with you. Can you share your mph file? thank you.
[QUOTE] Thank you for this response. We got the same answer from the comsol helpdesk and they now found indeed a loss instead of a gain. [/QUOTE] Hi, I have meet similar problem with you. Can you share your mph file? thank you.

Posted: 5 years ago 2013年1月10日 GMT-0500下午10:21
Hi, thank you very much for your fast response.

How do you calculate the transmission spectrum, use S21 or output power/input power?

I attached my ring resonator and need to simulate the transmission spectrum, I try but always fail, would you please help me?
Hi, thank you very much for your fast response. How do you calculate the transmission spectrum, use S21 or output power/input power? I attached my ring resonator and need to simulate the transmission spectrum, I try but always fail, would you please help me?


Posted: 5 years ago 2013年3月20日 GMT-0400上午8:17
hi,
im also experiencing a similar problem like yours. can u pls share ur mph file so that i will come to know how u have taken the transmission spectrum.
hi, im also experiencing a similar problem like yours. can u pls share ur mph file so that i will come to know how u have taken the transmission spectrum.

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.