Author Topic: Device DOS  (Read 3412 times)

0 Members and 1 Guest are viewing this topic.

Offline jdgayles16

  • QuantumATK Guru
  • ****
  • Posts: 108
  • Reputation: 0
    • View Profile
Device DOS
« on: July 15, 2010, 04:40 »
Two questions: 1. Angular momenta: what numbers correspond to what orbitals.
Code
[dos.evaluate(projection_list = ProjectionList([....], angular_momenta=[2]))]
I used numbers like 0,1,2, thinking they correspond to  s,p,d respectively but my values do not add up in the end to the regular pdos. What is my mistake here? 2. Contributions: could some explain more on this argument what it means to be left, right and all? I tried to find more information but I kept find information on LDOS

Offline zh

  • QuantumATK Support
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 1141
  • Reputation: 24
    • View Profile
Re: Device DOS
« Reply #1 on: July 15, 2010, 06:11 »
The properties you want may be the partial DOS of each atom (i.e., site-projected DOS):
ProjectionList(atoms=[....])

In additional, since the s, p, d-projected DOS of atom are obtained, the partial DOS of atom can be summed up easily with additional coding.


Offline jdgayles16

  • QuantumATK Guru
  • ****
  • Posts: 108
  • Reputation: 0
    • View Profile
Re: Device DOS
« Reply #2 on: July 15, 2010, 23:41 »
so how do you use the command to get p-projected DOS  for any given atoms


and also could you elaborate on the contributions command?

Thanks

Offline Anders Blom

  • QuantumATK Staff
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 5428
  • Country: dk
  • Reputation: 89
    • View Profile
    • QuantumATK at Synopsys
Re: Device DOS
« Reply #3 on: July 15, 2010, 23:52 »
1. Angular momenta: what numbers correspond to what orbitals.
Code
[dos.evaluate(projection_list = ProjectionList([....], angular_momenta=[2]))]
I used numbers like 0,1,2, thinking they correspond to  s,p,d respectively but my values do not add up in the end to the regular pdos. What is my mistake here?
Hard to say without seeing the details, but s,p,d are for sure 0,1,2. Post/send more details so we can double-check it (but we will not until August, bcs of holidays).
2. Contributions: could some explain more on this argument what it means to be left, right and all? I tried to find more information but I kept find information on LDOS
The two-probe DOS can be divided into contribution from the left and right leads, which sum to the total. I don't have the exact equation references handy, but check the usual refs, PRB 65, 165401 (2002) and Datta's book "Electronic Transport in Mesoscopic Systems" (Cambridge Univ. Press., New York, 1996).

Offline jdgayles16

  • QuantumATK Guru
  • ****
  • Posts: 108
  • Reputation: 0
    • View Profile
Re: Device DOS
« Reply #4 on: July 16, 2010, 23:26 »
Oh I see, the DOS arising from the left and right spectral functions. Thanks


I will play around with the angular momenta