Author Topic: Some thing wrong with MoTe2 calculations ?!!!!!!!!  (Read 3527 times)

0 Members and 1 Guest are viewing this topic.

Offline naomi

  • Regular QuantumATK user
  • **
  • Posts: 16
  • Country: ir
  • Reputation: 0
    • View Profile
Hello Dear,

I am trying to get band structure of MoTe2( expected to see roughly 1.07 eV bandgap) well something wrong with simulation. I am asking for some guidance after simulations of two different type of MoTe2(1 from ATK database, 1 From MaterialProject).
Back to My question,
I have chosen MoTe2 1.1.1 (GGA/PAW/PBE), cutoff 75Hartree and 41K sampling. I saw that pseudopotential of Mo by default is FHI/Ti0 (interesting) so to ease the calculation I changed Te to the same thing. Optimization part I chose : Maximum step 200 and Tolerance 0.01 e/V and 0.01 e/V^-3 .

After 1.5 days, MoTe2 simulation is in the 44 step, it has not stopped yet. when I look at MoTe2 video precisely (.tar file I mean) 2 atom of Te come much closer to each other which is strange. I had MoS2 or even MoSe2 bandstructure less than 24hrs. what is wrong with its?

If I need to provide more information please let me know.

Thanks in advanced  :)

Offline Petr Khomyakov

  • QuantumATK Staff
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 1290
  • Country: dk
  • Reputation: 25
    • View Profile
Re: Some thing wrong with MoTe2 calculations ?!!!!!!!!
« Reply #1 on: June 22, 2017, 15:27 »
Could you enclose the actual python script related to your MoTe2 calculation? What ATK version are you using?

Offline naomi

  • Regular QuantumATK user
  • **
  • Posts: 16
  • Country: ir
  • Reputation: 0
    • View Profile
Re: Some thing wrong with MoTe2 calculations ?!!!!!!!!
« Reply #2 on: June 22, 2017, 17:05 »
Dear Petr,

I attach the file I am trying to simulate.  I  also am using 2016.3 version.


Regards

Offline Petr Khomyakov

  • QuantumATK Staff
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 1290
  • Country: dk
  • Reputation: 25
    • View Profile
Re: Some thing wrong with MoTe2 calculations ?!!!!!!!!
« Reply #3 on: June 23, 2017, 10:29 »
Three things:

- You do not need to use that many k-points to sample the 2D Brillouin zone in the self-consistent calculation of bulk MoTe2. So, you may consider reducing it from 41x41x1 to, e.g., 12x12x1 Gamma-centered k-point grid. 

- I would suggest changing the pseudopotential/basis set to SG15-type/Medium basis set.

- In principle, you do not need to set the same ATK-DFT calculator to do the band structure analysis after geometry optimization.

If you make the changes mentioned, you must be able to do this calculation rather quickly, at least it worked for me, taking 6 minutes on my laptop.   

   

Offline naomi

  • Regular QuantumATK user
  • **
  • Posts: 16
  • Country: ir
  • Reputation: 0
    • View Profile
Re: Some thing wrong with MoTe2 calculations ?!!!!!!!!
« Reply #4 on: June 24, 2017, 09:38 »
Dear Petr,

You help me a lot :) :) :) :) . Thank you   :D :D :D :D :D