QuantumATK Forum

QuantumATK => General Questions and Answers => Topic started by: newbiee on April 22, 2014, 19:26

Title: MOS2 Slater Koster
Post by: newbiee on April 22, 2014, 19:26

Hi Quantumwise Staff,

I was following this paper http://arxiv.org/pdf/1401.4553.pdf

The authors have used ATK to get DOS and they claim the bandgap turns out to be 1.8eV (see first line under results and discussions).This bandgap is in agreement with published literature.

I was trying to reproduce the DOS, my DOS agrees qualitatively (as in the shape) but the bandgap I end up getting is ~2.2eV.

Can someone look at my attached script and let me know if I have missed something. I don't want to go for more rigorous DFT or Huckel if Slater-Koster(as claimed in paper) works

Regards!

Title: Re: MOS2 Slater Koster
Post by: Shinji Usui on April 23, 2014, 07:31

Your system is periodic in the directions B and C (Y and Z) so please increase ky and kz in SCF and DOS calculation
 
like
numerical_accuracy_parameters = NumericalAccuracyParameters(
    k_point_sampling=(1, 27, 27),
    )

(I am not sure k value 27 is the best one)

and

density_of_states = DensityOfStates(
    configuration=bulk_configuration,
    kpoints=MonkhorstPackGrid(1,101,101),
    energy_zero_parameter=FermiLevel,
    bands_above_fermi_level=None,
    )

Usually we need lager values of ks in DOS calculation than SCF's ones.

Title: Re: MOS2 Slater Koster
Post by: newbiee on April 23, 2014, 19:56

Thanks Shinji!

I tried both your suggestions, still I get Bandgap of 2.2eV :( instead of 1.8eV.

I used DFT and I have the correct bandgap. But I see something weird. On application of gate voltage. The transmission spectra doesn't change. From what I believe transmission must shift up or down depending on polarity of gate voltage. I tried both single gate and dual gate (Vdd and 0 volt) but can't see any shift only the peaks of transmission shifts downwards.

As such when I calculate I-Vgs I would get a flat curve instead of typical saturating curve. Any suggestion is appreciated.

Thanks !

Title: Re: MOS2 Slater Koster
Post by: newbiee on April 27, 2014, 05:07

Dear Quantumwise Staff,

Any suggestion on why gate voltage doesn't seemto have any effect?

Thanks!

Title: Re: MOS2 Slater Koster
Post by: Anders Blom on April 28, 2014, 00:47

Maybe wrong boundary conditions? But without any details provided, one can only guess...

Title: Re: MOS2 Slater Koster
Post by: newbiee on April 28, 2014, 02:12

Dear Dr. Anders,

Attached is the script. Kindly have a look.

Following are some trial I made based on your previous comments on forum:

1. Periodic(since unit cell has enough vacuum), Periodic , Dirichlet boundary
2. Neumann, Periodic, Dirichlet Boundary
3. Single gate vs dual gate

 I was able to get correct bandgap, but wasn't able to see effect of gate voltage.
I am stuck any help is appreciated!

Title: Re: MOS2 Slater Koster
Post by: newbiee on April 29, 2014, 20:03

Anyone !

Title: Re: MOS2 Slater Koster
Post by: Anders Blom on April 29, 2014, 20:25

One of your metallic gates is outside the system. It will not be included in the calculation.

Title: Re: MOS2 Slater Koster
Post by: newbiee on April 30, 2014, 22:17

Dr Anders,

I tried your suggestion by including the gate within the unit cell. Still no change in results, attached is my script.
Can you please look into it.

Thanks!

Title: Re: MOS2 Slater Koster
Post by: Anders Blom on May 1, 2014, 09:35

Short of redoing the work, it is difficult to answer a question like this. The only immediate thing I would comment on is that you only have 1 k-point in the periodic B-direction, that's at least inaccurate.
Also, the central region is very short compared to the electrode lengths, it should probably be doubled to make real sense.

Title: Re: MOS2 Slater Koster
Post by: newbiee on May 1, 2014, 17:45

I appreciate your patience !

I will try re-simulating a bigger device with increased k - points. Let's hope that works :)

Thanks!