I am simulating a NW based MSM Photodetector with Si/ CdTe/InAs NWs (modeled in SDE)with moderate doping, and have assumed a constant generation rate in the the Physics Section of SDEVICE
1. When using the coupled statement in SDEVICE ( coupled {Poisson eQuantumPotnetial Electron Hole}) the simulation is getting diverged with inconclusive results, but if the 'Hole' is removed form the coupled statement it seems to run fine. I have even included the LineSearchDamping for the initial solution of the coupled statement to improve convergence but with no effect.
2. For the CdTe and InAs NWs, while running the SDEVICE simulation a message showing 'UNDEFINED model' is displayed in the Linux terminal for the Auger, DopingDependence etc. models which were used in the region specific Physics section.
3. The non-local tunneling at the Nanowire-contact interface and at the heavily doped source and drain regions is not giving any change of current flowing through the device.
It will be much of much help if some-one could please help me with these issues and let me know if further details/clarifications are needed.
Hoping for a positive response soon.
Thanks in anticipation !
Dear Antardipan Pal,
did you define the used materials and their properties in dedicated files ?
Regards
Hello Laurent,
Thanks for your answer. I don't think I did that and I have already figured out that it is the cause for the issue described in point 2.
Do you have any idea about the issues described in point 1 and 3.
I really appreciate your answer.
Dear Antardipan Pal ,
For point 1, did you try to perform the SDEVICE simulation step by step in "coupled"; I mean:
- First: only poisson
- next poisson, Electrons and Holes
- Next: poisson, Poisson eQuantumPotential hQuantumPotential Electron Hole
Moreover, I think, if you would consider Hole, you should include hQuantumPotential.
For your third point, in state, I don't know. If need I could ask my research team to check your SDEVICE file.
Regards
Head of radiation effects on microelectronics and materials research group
ONERA - The French Aerospace Lab - Centre de Toulouse 2, avenue Edouard Belin - BP 74025 - 31055 TOULOUSE CEDEX
Antardipan Pal
I suggest you increment the complexity of your simulations slowly.
1/ in the physics section, use only standard models at the beginning. This includes using the Drift-diffusion transport model. and solve only for poisson, electrons and holes. same goes for the math section, just add one parameter at a time.
vary the LineSearchDamping at the beginning of the solve section.
If it converges, increase the complexity one model at a time.
2/ for the solve section check the previous comment by Laurent Artola .
3/ once your simulations are running correctly, you can remove few terms in the solve section (like hQuantumPotential...), that doesn't impact the accuracy while decreasing simulation time.
4/ check your parameter file.
Hello Laurent Artola,
Thank you so much for your time in answering my questions. Your suggestions are really helpful and try them out to see if the issues are resolved. If needed I'll try to share the SDEVICE file with you as you have mentioned.
I'll get back to you if I have further issues.
Thanks and I really appreciate your help!
Hello Ali Houadef ,
Thank you so much, and I really appreciate your elaborate explanations to my questions. I'll try out with more simpler and basic Physics models and check whether the issues are still persisting.
I'll let you know if there are further issues !
Thanks again.
Hello Laurent Artola
Hello Ali Houadef
Hope you are doing well. I did add the simulation complexities step by step and seem to solve the convergence issues. and thanks for the suggestion. However there are some other issues coming up:
1. In some of the simulations (typically for compound semiconductor NWs) an error is being display , "terminate called after throwing an instance of 'std::runtime_error', what(): ERROR; FillEdgeElementInfo:
CurrentNumberOfElement(3) < NumberOfElementsAroundEdge(5)
Aborted (core dumped) ". It will be of much help if you can suggest some fix to this.
2. Also I figured out, that many physical models and parameters like, SRH Auger coefficient, SchroedingerParameters, DopingDependence, HighFieldDependence, BandGapNarrowing etc. are not defined in the parameter files for compound semiconductors, especially CdTe. How can these physical models be added so that the simulation with compound semiconductor NWs are accurate.
3. Also even with high doping say order of 10^17 - 10^18/cm3, the tunneling model doesn't seem to give any noticeable change in current in the nanowire PD.
It will be of much help if you could please give some suggestions regarding these issues!
Thanks .
Hello Antardipan Pal , sorry I did not respond early.
1/ I do not know what the mentioned error means. But, If I had to guess, it is related to the mesh. Where did this error occur, is it in the Sprocess, Structure Editor or Sdevice?!
Could be related to the meshing parameters you introduced (refineboxes).
If it occurred during the process simulation, it is very likely to the remeshing at the end of an annealing or diffusion step.
2/ To introduce fundamental parameters of new materials, I usually create a separate file (of FPS extension), write the new parameters and/or inherent parameters from similar materials. (see attached file). This is useful for Sprocess in most cases. in Sdevice However, I try to write them directly in the parameter file.
Refer to the Advanced Calibration Guide for each tool, and Read scripts from the projects that come with Sentaurus.
3/ This is also likely related to the mesh, it depends how much the delta we are talking about.
If it is -say- 5 to 10% difference you are looking for, try with a denser mesh at the interesting regions.
if you are expecting bigger delta, then I highly recommend to thoroughly revisit how you generated the device. I also recommend simulating devices from Sprocess rather than SDE as it is more realistic, thus more reliable.
TIP: use mask command to generate refine-boxes like this:
mask name= SrcDrn left= 0.2 right= 0.66
mask name= SrcDrn left= 4.34 right= 4.82
refinebox name= SrcDrn mask= SrcDrn \
extend= 0.2 extrusion.min= -0.5 extrusion.max= 0.5 \
xrefine= "0.05/$fp" yrefine= "0.05/$fp"
I hope you find my answer useful.
Thank you so much Ali Houadef , for your response. I really appreciate your time in responding to my problems.
1. The mentioned error occurred in SDEVICE.
2. Where can I find the advanced calibration guide for various tools? Is it in the MANUALS folder ?
3. Till now, I have been generating the structures and doing the meshing and all from the SDE, but from now I'll try using the SPROCESS. Thanks for the suggestion.
Thanks again !
- Badges
- Science topic
- Similar topics
- Biological Science
- Kinesiology
- Running