Get I-V curve in DEVICE

I’m trying to find I-V curve of a graphene photo detector with a nonamer structure,I simulated it in FDTD to achieve generation rate.Since that graphene hasn’t defined in DEVICE,so I used graphene properties from this link:

And also I read some pages about simulation in DEVICE like this:

I simulated it in DEVICE (based on paper parameters) but my simulation has some errors that I can’t solve them.Would you please help me to get I-V curve like it’s paper:

here are my simulations:
nonamer with generation rate-1.fsp (280.3 KB)
nonamer.ldev (5.9 MB)


Hi. I am afraid DEVICE is not suitable for such simulations. The reason we do not have a graphene model in the material database of DEVICE is because the CHARGE solver currently does not support 2D density of states (DOS). The model that you have used is basically a 3D fitted DOS that closely mimics the charge-chemical potential relationship of graphene. However this is a physically inaccurate model (as mentioned in the example) and is only applicable in limited cases (like the modulator in the example where no transport takes place in graphene).

A few other points,

  1. In the CHARGE solver, metals can only be used as electrical contacts and they only act to provide boundary conditions to the simulation region. Therefore you will not be able to place the nanomer structure in a CHARGE simulation.

  2. In FDTD, you cannot use the generation rate analysis group for the 2D graphene sheet as the analysis group is designed for 3D materials only.

Thanks a lot for your help.
I was hoping that I could simulate it with DEVICE,but unfortunately you said I will not be able to simulate it with this structure,So how can I change my structure to get I-V curve?Is it will be possible?
For example,if I chose a semiconductor (like ZnO) instead of metal for nonamer structure,could I simulate this photo detector with (or without) graphene by DEVICE?
what is your suggestion?

Best Regards,

As I mentioned, the two limitations that you are facing are, (i) no 2D model for graphene and (ii) no transport in metal. So if you use ZnO instead of metal and remove the graphene layer, you should be able to simulate your photodetector using DEVICE.

I have a question:as you mentioned,in DEVICE I have to use metal just as a contact not particle,so how can I solve this problem with only gold not semiconductor?Could I apply a zero voltage to all particles of this cluster to solve this problem?
What’s your idea?
I must get I-V curve of this structure with gold nonamer not any material else!


To simulate this structure in DEVICE we have to make the approximation that the gold nonamer does not affect the electrical simulation. It will affect the optical absorption and FDTD will be able to account for that and the effect will be imported into DEVICE through the optical generation rate. However, since the nonamer will not be connected to any external contact or anything, we should be able to simply omit it from the simulation volume and run the DEVICE simulation (with the actual electrical contacts in place).

Thanks for help.As you mentioned:[quote=“aalam, post:6, topic:2298”]
we have to make the approximation that the gold nonamer does not affect the electrical simulation.

What do you mean?Would you please explain this?
I send you my simulations as you mentioned before,Would you please modify them as you mean?
nonamer with generation rate-2.fsp (280.6 KB)
nonamer.ldev (5.9 MB)

I will wait for the answer,

Hi @hananeh, I’m really sorry for the late reply. When I said that we can “make the approximation that the gold nonamer does not affect the electrical simulation”, I meant that in the electrical simulation we can simply omit the metallic structure in the middle. So you will have the structure with just two metal regions for the two contacts (drain, source).

NOTE: since you are interested in calculating the absorption in graphene and not in the metal, you may set the volume of your analysis group in FDTD to cover the graphene layer only and reduce the size of the saved data.