Hello Vidya. First you have to enable the models of the mechanisms that give rise to gate leakage. Depending on the technology you are simulating (i.e. tox, gate stack, etc) you should pay attention to which models have to be included in order to get representative results.
Generally speaking, most complete tunneling models (Non-Local Tunneling) are included in sdevice like this:
** In the Physics section include tunneling models for electrons and/or (in this case band to band is not considered) **
eBarrierTunneling "NLM" (Band2Band=None)
hBarrierTunneling "NLM" (Band2Band=None)
** In the Math section, a non local mesh named "NLM" has to be created on the barrier where tunneling will be computed. In this case the barrier consists of the region "R.Gateox" of the device's structure **
Nonlocal "NLM" (
Barrier(Region="R.Gateox")
)
It is necessary to define on your parameter file, tunneling coefficients for carriers:
** In the materials involved in the tunneling effects, define tunneling masses for electrons and holes
Material="SiO2" {
#includeext "ParFileDir/sdevice_Oxide.par"
BarrierTunneling "NLM" { mt = 0.5, 0.5 }
}
Refer to Chapter 24 of sdevice Manual for complete references to the models and syntax.
Hope this helps.
Hello Vidya. First you have to enable the models of the mechanisms that give rise to gate leakage. Depending on the technology you are simulating (i.e. tox, gate stack, etc) you should pay attention to which models have to be included in order to get representative results.
Generally speaking, most complete tunneling models (Non-Local Tunneling) are included in sdevice like this:
** In the Physics section include tunneling models for electrons and/or (in this case band to band is not considered) **
eBarrierTunneling "NLM" (Band2Band=None)
hBarrierTunneling "NLM" (Band2Band=None)
** In the Math section, a non local mesh named "NLM" has to be created on the barrier where tunneling will be computed. In this case the barrier consists of the region "R.Gateox" of the device's structure **
Nonlocal "NLM" (
Barrier(Region="R.Gateox")
)
It is necessary to define on your parameter file, tunneling coefficients for carriers:
** In the materials involved in the tunneling effects, define tunneling masses for electrons and holes
Material="SiO2" {
#includeext "ParFileDir/sdevice_Oxide.par"
BarrierTunneling "NLM" { mt = 0.5, 0.5 }
}
Refer to Chapter 24 of sdevice Manual for complete references to the models and syntax.
Hope this helps.
Hello Vidya Sir and Sebastian Sir....
I have simulated a MOS cap without using Nonlocal meshing at interfaces. I use directly
Band2Band( Model=Hurkx
DensityCorrection=Local) in Physics Section.
And used GateCurrent(DirectTunneling | Foewler) in Interface specific Physics and Math section but very small leakage current is obtained in order of 10-25.
Yes I obtained a bit improved current by using your suggested method by using Nonlocal mesh.
But when i replaced Oxide (SiO2) other high-k dielectric (Specifically HfO2) not a single value changes in characteristics.
Do i have to apply some another or different approach for High-K capacitor? or am i missing something here?
Please suggest.
Hello Vidya Sir, Vibhu Sir and Sebastian Sir....
I am also getting the same problem as is faced by Vibhu sir. Please suggest us what changes we should do to get the gate tunneling current when we are using gate stack technology. Also please tell me, How do I plot gate leakage current vs gate voltage for MOS device using TCAD sentaurus?
Hello Vibhu, Shubham, I'm sure this topic is long overdue, I'm really sorry but it is easy to loose track of questions in research gate if you don't have time to address them immediately.
In any case, Vibhu did you include the tunneling parameters for material HfO2 once you changed your structure? The output log is usually helpful to track down the problem.
Shubham, in order to visualize the tunneling current you have to include it among your your outputs, but you can always plot the total, electron and hole current through the gate contact in svisual.
Well, you have probably solved these problems by now anyway, hope your work is moving forward successfully!
Hello @Sebastian sir....
Thanks for your kind reply and useful suggestion.
Yes, i have done a lot of changes and improvement since then and getting better results here. But still need suggestions in future.
Thank you
Hello Hassan Sir....
As Sebastian Sir told about the tunneling parameters, i tried different types of tunneling parameters in HfO2.
I fetch the parameter file of the HfO2 as well as the SiO2, compare both, simulated step by step for each and every parameters because you will find a lot of difference between both files, added missing parameters in the format from different literature resources.
but still in dilemma that, is this a right way to carry out this? And are the results trustworthy? (It may be foolish)
- Badges
- Science topic