Author Topic: How to set the spins of the right and left electrode atoms respectively?  (Read 6669 times)

0 Members and 1 Guest are viewing this topic.

Offline albatross

  • Regular QuantumATK user
  • **
  • Posts: 12
  • Reputation: 0
    • View Profile
  In atk 11.20, I want to set the spins of the right electrode atoms up, and the spins of the left electrode atoms down. In atk 2008.10, I can realize this trough revising the spins in the script file, but I can only revise the initial scaled spins in the script file of the new version, and the initial scaled spins only contain the left electrode atoms. How can I realize this? Thanks a lot.

Offline Anders Blom

  • QuantumATK Staff
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 5576
  • Country: dk
  • Reputation: 96
    • View Profile
    • QuantumATK at Synopsys
Actually you don't set the spin of the electrodes in 11.2, you set it on the electrode copy regions in the central region, and ATK copies these initial spins to the electrodes.

The tutorial on FeMgO should explain it all!
« Last Edit: May 11, 2011, 05:42 by Anders Blom »

Offline albatross

  • Regular QuantumATK user
  • **
  • Posts: 12
  • Reputation: 0
    • View Profile
OK, thank you very much.

Offline baizq

  • QuantumATK Guru
  • ****
  • Posts: 100
  • Country: us
  • Reputation: 3
    • View Profile
Actually you don't set the spin of the electrodes in 11.2, you set it on the electrode copy regions in the central region, and ATK copies these initial spins to the electrodes.

The tutorial on FeMgO should explain it all!


Hi Anders,

Why are there electrode copy regions in the central region in ATK 11.2?

Suppose I want to set up a Fe-MgO-Fe MTJ device with 6 MgO barrier layers and 4 Fe electrode layers, as well as 2 Fe surface layers, as what is recommended in the tutorial. Why does the appearance in device mode looks like we have another 4 Fe layers at each side of the device? They are called electrode copy regions as you mentioned,right? Then when we want to calculate transport properties of the system, how many Fe layers are really taken into consideration by the algorithm?

thanks!
baizq

Offline Anders Blom

  • QuantumATK Staff
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 5576
  • Country: dk
  • Reputation: 96
    • View Profile
    • QuantumATK at Synopsys
This requires a rather long explanation, however maybe you can just have a look in the upgrade guide. I hope that answers your question :)

Offline baizq

  • QuantumATK Guru
  • ****
  • Posts: 100
  • Country: us
  • Reputation: 3
    • View Profile
This requires a rather long explanation, however maybe you can just have a look in the upgrade guide. I hope that answers your question :)


good answer...thanks....

Offline camelluxin

  • Regular QuantumATK user
  • **
  • Posts: 14
  • Country: sg
  • Reputation: 0
    • View Profile
Actually, I'm confused too.
What is the difference between electrode and electrode copy and how can I change one without affecting another?

Offline Anders Blom

  • QuantumATK Staff
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 5576
  • Country: dk
  • Reputation: 96
    • View Profile
    • QuantumATK at Synopsys
First of all, you can't and you shouldn't set the initial spin separate in the electrode and the electrode copies.

The electrode copies are the first and last NL and NR atoms in the central region, where NL and NR are the number of atoms in the Left and Right electrodes, respectively. These atoms in the central region must be exact copies of the electrode atoms, by position (relative the edge of the cell), element, and order. This is a fundamental requirement in ATK 10.x and higher, and is related to how the boundary conditions as well as initial matrix elements are set up. This is also why it's not necessary to set the initial spin of the electrode atoms, you just set them on the electrode copy atoms, and the setting is then automatically copied over to the electrode.