Geometry building issue in 3D simulation for PD

device

#1

Dear
I have run the Germanium-on-Silicon Lateral Photodetector example from the knowledge Base.
I have some trouble when i run it in 3D simulation.vpd_electrical.ldev (4.6 MB)

The simulation area contains Germanium,cathode ,anode and silicon wavegudie
The error messasege is No area for degenerate linear planar polygon.
I have read the similar topic from KX ,but the solution didn’t work.


#2

Hi @zhoude, You are correct. This does look like a geometry building issue. In our latest release the geometry builder in DEVICE has been updated considerably and in most cases the new geometry builder should be able to handle complex 3D geometries. Therefore, I am guessing that you are probably still using an older version? If that that is the case then can you please upgrade to the latest version and try to run your file again? You can get the latest version of DEVICE from here: https://www.lumerical.com/downloads/customer.html.


#3

Thank you .! I have got the latest version of DEVICE. Now I have another question.
I want to simulate the MSM photodetector. How can I set the schottky contact?
and Can this version simulate the MSM photodetector?


#4

Another question ,sir!
in 3D simulation, I get the data in 3D model ,is there way that I can get one across section data,just like the optical generation analyst in FDTD .


#5

Hi @zhoude, You can model Schottky contacts at the metal semiconductor interface by setting the “force ohmic” option to false in the properties of the electrical contact.

By default the “force ohmic” option is set to true and the solver adjusts the work function of the metal internally to match with the work function of the semiconductor in contact to form and ohmic contact. However, if you choose to set this to false then the solver will calculate the band bending at the interface depending on the work function of the metal and the semiconductor and therefore you will be able to model the effect of the Schottky contact.

One thing to note here is that the CHARGE solver does not perform a quantum mechanical calculation of the tunneling current at the Schottky barrier but uses an “effective barrier lowering” model to adjust the thermionic current and thereby account for the tunneling current. This approach is only valid as along as the contribution of tunneling current is much smaller compared to the thermionic current. However this assumption should be valid for MSM photodetectors (in my understanding) and therefore you should be able to simulate them with the CHARGE solver.