Author Topic: Transmission spectra vs k-points  (Read 350 times)

0 Members and 1 Guest are viewing this topic.

Online Jahanzaib

  • Regular ATK user
  • **
  • Posts: 18
  • Country: gb
  • Reputation: 0
    • View Profile
Transmission spectra vs k-points
« on: August 30, 2022, 12:28 »
I am doing research in 1-D stanene nanoribbon for device calculation and interested to optimize the Pd on it.
I want to know what is the role of k-point in transmission spectra of 1-D stanene nanoribbons, I have chosen 1×1 and 9×9 k-points but couldn't get the desired results for transmission spectra.

Offline Anders Blom

  • QuantumATK Staff
  • Supreme ATK Wizard
  • *****
  • Posts: 5126
  • Country: dk
  • Reputation: 85
    • View Profile
    • QuantumATK at Synopsys
Re: Transmission spectra vs k-points
« Reply #1 on: August 30, 2022, 21:47 »
You don't need k-points the the transverse directions if it's a 1D structure.

Online Jahanzaib

  • Regular ATK user
  • **
  • Posts: 18
  • Country: gb
  • Reputation: 0
    • View Profile
Re: Transmission spectra vs k-points
« Reply #2 on: August 31, 2022, 00:31 »
Thank you for your guidance
As mentioned earlier, for 1×1 k point in transmission spectrum there is broad peak at fermi level. What is the signifance of this peak.For 1-D, transmission spectrum are independent of k point in the tranverse direction then why there is broad peak at fermi level.
 In manual, it is mentioned that for a perfect 1D system, the transmission exhibits a sequence of steps with integer transmission. I couldn't understand this.
For 2-D, transmission spectrum dependent on k point? Why

Can you please explain me?



« Last Edit: August 31, 2022, 15:29 by Jahanzaib »

Offline Anders Blom

  • QuantumATK Staff
  • Supreme ATK Wizard
  • *****
  • Posts: 5126
  • Country: dk
  • Reputation: 85
    • View Profile
    • QuantumATK at Synopsys
Re: Transmission spectra vs k-points
« Reply #3 on: September 1, 2022, 00:15 »
The peak at the Fermi level means there is a lot of transmission at that energy. Why or if it's correct or not, one cannot say with a proper analysis of your actual system.

The transmission is step-like (integer values) for 1D systems simply because you have 1 transmission channel per band, so the number of bands crossing the energy level is how many units of transmission you get.

For a simple physical picture of the different between 1D and 2D, in 1D all electrons will enter and exit the system parallel to the Z axis (transport direction) so we just need to count how many. In 2D, i.e. with periodic boundary conditions in one direction perpendicular to C) the electrons arrive at some angle (determind by the k-vector) and we have to average or integrate over all incoming angles. And of course in 3D we also average over the other periodic direction.

Online Jahanzaib

  • Regular ATK user
  • **
  • Posts: 18
  • Country: gb
  • Reputation: 0
    • View Profile
Re: Transmission spectra vs k-points
« Reply #4 on: September 2, 2022, 01:43 »
Thank you very much. Now, It's clear for me.
I have attached the PDOS vs Transmission spectrum for Pd adsorbed on 1-D stanene nanoribbon for device calculations.

Offline Anders Blom

  • QuantumATK Staff
  • Supreme ATK Wizard
  • *****
  • Posts: 5126
  • Country: dk
  • Reputation: 85
    • View Profile
    • QuantumATK at Synopsys
Re: Transmission spectra vs k-points
« Reply #5 on: September 6, 2022, 22:58 »
Right, the transmission peak at the Fermi level corresponds well with the DOS there. You might want a few more points on the transmission curve to get a more accurate picture of the peak.

Online Jahanzaib

  • Regular ATK user
  • **
  • Posts: 18
  • Country: gb
  • Reputation: 0
    • View Profile
Re: Transmission spectra vs k-points
« Reply #6 on: September 15, 2022, 01:20 »
Thank you so much for your response.
I tried alot to figure out why there is peak at fermi level, but I couldn't find the suitable analysis to understand this.
As mentioned in the above messages, my system is 1-D stanene nanoribbons and Pd is adsorbed on the central region but I found "with and without Pd" there is peak at fermi level. What analysis should I do to figure out this peak?
Secondly, how to find out that this lenght of central region would be for analysis, or I should randomly changed the central region lenght and then figure it out.

Thank you