Learning Forum

[Sungmin Yoon]

Hello,

I'm running simulation with a stack consists of metal-semiconductors-metal to see charge behavior. 

Applied bias from 0 V to -20 V on the one metal(gold) end, and GND(i.e., 0 V) to the other metal end. 

It was reverse bias, so I set the solver type as 'gummel'.  

Voltage step was -0.1 V, range backtracking is enabled.

And it took 2 days to get - 3.75 V with error message as below:

=========================================================================================

Solver terminated at step 42 of 40.  Error  (charge): Drift-diffusion Newton solver failed to converge after 40 iterations.
Device solution update error ratio norm is 1.92713e+11
Device residual norm is 2.93284e-51.
Residual ratio (last iteration/first iteration) is 7.43227e-11.

Warning: The charge transport solver failed to converge (Charge Convergence Failure). A partial result will be saved. For instructions on how to troubleshoot this issue, please refer to this: online reference
The program terminated due to an error: The simulation diverged and the result is incomplete.
Error: there was an unknown parallel error. The error code is 9002, the process number is 0

=========================================================================================

From CHARGE monitor, I was looking at 'n' attribute, there's fast depletion of electron density at the early stage of bias increase, as expected and attached images.

 

 

 

 

 

But my question is..

Whether the solver diverged at -3.75V since it reached extreme depletion and there's nothing solver can do about it anymore? 

.. or it is due to solver setting problem, like one in this article?  

From your professional point of view, I'd be really glad if you give me a hint. 

 

Thanks,

Sungmin 

 

 

[Sungmin Yoon]

btw, it is CHARGE in Lumerical. 

[Dev]

Hello, Aplologies for late response. We missed to see this post in forum. 

It is bit hard to predict in this case. Can depend on lot of parameters. 
Can you please incease the number of global iterations and run the simulations again? 40 seems not enough here. 

-3.75V seems reaching extream depetion region and solver can't do anything about it from my understanding. 

You can find more information here: https://optics.ansys.com/hc/en-us/articles/8355464963987-Troubleshooting-convergence-errors-in-CHARGE

Thanks and regards

Devika