This thread is locked.Only browsing is available.
Top Page > Browsing
How to distinguish different spins in band structure by non-collinear DFT with spin-orbit coupling
Date: 2008/01/25 18:58
Name: Yun Li   <yun.li141@gmail.com>

Dear Prof. Ozaki,

In non-collinear DFT calculation with spin-orbit coupling OPENMX only outputs one banddata file. So in band structure plot one can't distinguish spin-up and spin-down levels. In the case of spin-orbit coupling calculations spin-up and spin-down states are mixed up. But in density of states one can plot spin-up and spin-down states. It means that actually spin-up and spin-down can be distinguished. At least one can know at a given K point and a given energy eigenvalue the eigenstate is spin-up or spin-down, or percentages of spin-up and spin-down. I ask you to do me a favour, how can we distinguish spin-up and spin-down with present output files by openmx. Is there a program to deal with this problem?

Thank you.

Yun Li
メンテ
Page: [1]

Re: How to distinguish different spins in band structure by non-collinear DFT with spin-orbit coupling ( No.1 )
Date: 2008/02/06 10:35
Name: T.Ozaki

Hi,

In the non-collinear calculations, since the quantization axis of spin can
vary in space, striclty speaking, saying the spin-up and the spin-down is
meaningless, while it is assumed the quantization axis to be z-axis in the
collinear calculations.

What we can see in the PDOS provided by OpenMX is the PDOS projected on the
quantization axis of angles calculated from Mulliken populations on each
atomic site (the total DOS is projected on the quantization axis calculated
by total averaged spin angles).

Since the angle can differ from one site to other sites,
the "spin-up" and "spin-down" PDOSs are just resolved along the quantization axis
on each site.

How they are calculated can be found in the technical note
http://www.openmx-square.org/tech_notes/tech2-1_0/tech2-1_0.html

Regards,

TO
メンテ

Page: [1]