gate voltage

[SAKINA]

Hi,

i would like to  know if my steps to add metal and oxide to the structure is either right or wrong?
This is because the problem that i encounter is when the gate voltage is varied from 1V to 5V but the IV that i obtained is the same.
i attach the  structure figure using GNR and value of gate voltage below. After setting the values from the builder is sent to script generator and IVCurve is taken from the analysis the bias voltage is set till 3V, and lastly sent to job manager. Is there any steps that im missing. What should i do? 

[Jess Wellendorff]

I am not seeing anything obviously wrong with your configuration: The gates and dielectrics look ok. However, I do notice that you set up a rather strong field between the gates; the gate voltage potential differences is 6 V. Did you try to scan in the range of low gate-voltages before going to strong fields?

[Anders Blom]

Attach full script, else we can't help. My immediate guess is that you use a semi-empirical method but you are not doing it self-consistently, but again, without seeing the script, it's just a guess.

[SAKINA]

Thank you for your reply.
I changed the gate voltage potential differences to 1V, upper gate with 0.5V and below gate with -0.5V.
While for calculator i used Extended Huckel. Is this the script? Should i tick the No SCF box or untick it?

[Jess Wellendorff]

You should attach the full Python script that you use. Much better than PNGs.

[SAKINA]

Sorry for the late reply.. this is the script.. And i would like to know what is the difference between using self consistently and without consistently? and which way is more accurate?

[Jess Wellendorff]

1) without self-consistency the electron densities are not updated from the initial guess (superposition). Use self-consistency if possible (default behavior).

2) I had a look at your script. A few comments:
a) Looks like you removed one of the metallic+dielectric regions.
b) Looks like you chose the Krylov self-energy calculator for the IV curve calculations. Why not use the default Recursion method?
c) You use a potential difference of 3 Volts between the electrodes in the initial device calculation. Why not leave this as a zero-bias calculation and then do the IV curve?

I have attached a script and a png of the IV curve I got. Looks all right to me. Hope this helps.

[SAKINA]

Thank you

but i have some questions about the self-energy calculator. In tutorial it said that using Krylov method is more faster than other method. Therefore
1)What is the difference between Krylov and default method?
2) Which method is more suitable to use?
3) Does the result obtain differ using different method ?

[Jess Wellendorff]

The Krylov method can be much faster than the Direct and Recursion methods. However, sometimes (in special cases) it is less accurate. If doing many calculations with the Krylov method we usually recommend to do a single one with Direct or Recursion method, just to make sure that Krylov works well for those particular calculations.

[SAKINA]

Now i understand.  Thank you for your help