big difference of transmission in TE mode and TM mode


#1

dear
When I learn the example of the 2D grating coupler on your website,I found there is a big difference of transmission when the light source is set TE mode and TM mode,which as show below.

it seems that when the light source is set TM mode,the result is right.But I dont know why the transmission of monitor(“T”) is so low when the light source of the fiber analysis is set TE mode.If I want to do simulation about other grating coupler of different structure,must I set the light source to TM mode?
1.fsp (505.7 KB)


#2

Dear @jbwei

It looks like the device scatters most of the injected TE-like light from the fiber. This means that the device is not optimized for TE mode, and you have a very low coupling to TE waveguide modes. You can visualize this by adding a movie monitor. Please also note that you have to adjust the Mode Expansion Monitor to account for the mode of interest (TE mode in this case) from the General tab of mode expansion monitor.

My understanding is that you can optimize your grating for TE or TM mode. So even this device is optimized for TM mode, you can come up with other geometries that demonstrate better coupling for TE modes.

I hope I could answer your question.


#3

Hi @jbwei
For the TE mode we can see that the field exists in the center of the waveguide. For the TM mode the field exists in the edges of the waveguide and there is no field in the center.


#4

dear bkhanaliloo

In reality,the light source coupled from fiber to grating coupler is TE mode in my experiment and design, but now I will do the simulation of the coupler and get the relative performance results by 2D FDTD just using the similar model like the example on your website(it means that I just modify some value of the parameters and other setup(include the geometries) is still the same) .Is that the geometries and the 2D FDTD simulation cause that we have to use the TM mode in the 2D FDTD simulation? Should I still set the light source to be TM mode in the 2D FDTD simulation no matter what kind of the light mode in reality because the result is wrong if I set the light source to be TE mode, just as show above?

dear konslekk
how to see the field whether or not exists in the center of the waveguide? Do you mean that I must use TM mode if I use the same geometries like the example on your website in the 2D FDTD simulation.

Hope to get your feedback.Thanks very much.:slight_smile:


#5

Hi @jbwei
You have to know that grating couplers are intrinsically polarization sensitive, since the waveguide effective index has very different values for TE and TM polarized modes. It is possible that the TE-designed grating will suppress light in the opposite polarization.


#6

Dear @jbwei

The TM and TE-like modes of the fiber are defined based on the polarization direction. Your field propagates along the y direction and your dominant field can be on the z- (TM polarized) or x-(TE polarized) direction. Here are the screenshots for these polarization obtained from software:

Please note that your grating is considered to be infinite in the z-direction (2D simulaiton). This might help you to compare the polarization in your setup with the ones in simulations.

So, simulation wise there is no limit on defining TE or TM modes. As @konslekk mentioned, grating couplers can be optimized for TE or TM polarized light, although there are designs that support both polarizations.

Our design is compared with the result of the papers (link to them is provided in the same direct). You might find it useful to take a look at them and get some information about TE polarization behavior for this particular design.


#7

dear konslekk
When I set the TE and TM mode as you said above,now it can be work right in the port 1(source in the fiber).But if the field propagates along the x direction (just the port 2 light source(waveguide source),how to set or define the TM polarized and TE polarized? I have tried to set the port2 light source(waveguide source) ,but I find the TE mode has two peak while the TM mode has just only one peak,just as shown below.


I am not sure whether what I set is right or not.
1.Can you help me explain why the TE mode has two peak in the waveguide source while the TM mode just has one.
2.If the field propagate along x direction,how the set the TM polarized and TE polarized.
The two question are very important for my own grating coupler simulation.Hope to get your feedback as soon as possible.Thanks a lot.

1.fsp (505.7 KB)
TM.fsp (487.4 KB)


#8

dear bkhanaliloo
After I run the grating coupler simulation again,I find the result is still not right.The relative parameter value in my simulation is used from the papers ,which are just the same papers as the grating coupler example on your website quoted.(If you have read the three papers the example on your website quoted ,you will find that the three papers are all about TE mode.

I dont know why the example on your website is setting in the TM mode or can you give me the paper ,of which the results you compare with the simulation. Two of the three paper has the same parameter value: FF=0.5,period=630nm,etch_depth=70nm (for TE mode).

But when I do the simulation with the same condition just as the paper’s by using TE mode light source,the result I got from the papers are huge different from the paper’s,which as shown below.

But when I do the simulation again with the same condition by using TM mode,the result seems normal.

I have to doubt that whether the FDTD software have some bugs or something is still set wrong in my simulation.But what I want to say is that the problem now doesn’t have matter with the condition setting(such as FF,period,etch_depth) because the same setting has been used by others.

finally can I use the TM mode light source in my simulation although in reality the source is TE mode?

At the end I will uoload the two programm of my simulation and where the two papers come from.

TE.fsp (480.3 KB)
TM.fsp (480.4 KB)

the two papers

1.Grating Couplers for Coupling between Optical Fibers and Nanophotonic Waveguides
2.Highly Efficient Grating Coupler between Optical Fiber and Silicon Photonic Circuit

the three papers you quote on your website about the grating coupler.(the two papers above are also contained in them.)

It is very important for me.Hope to get your feedback as soon as possible. Thanks a lot.


#9

Dear @jbwei

Software defines TE and TM modes based on the electric field component defined (inside the box see figure below). Thus, the definition of TE and TM modes depends on the geometry and some care is required when you are interpreting the results.

In this example, fiber and gratings are defined in the xy plane. So, for fiber when light is TM polarized we have Ex=0 and E=Ez.

For the grating this is slightly confusing. Based on software coordinates, light is TM polarized if Ey=0 and E=Ez. This is different from the papers and the way they define TE polarized light.

I guess it is a good practice to visualize the electric field components and pay attention how we have defined the geometry.

I hope this clarifies the problem and hopefully your new simulations match the papers.