I have a spin polarized system for which I need a separate band-structure for spin up and spin down system. I have used the band-prog routine (PWSCF band plotting routine) developed by J. Yates. However, there is a discrepancy of calculation of K-Points done by the said routine and the PW of quantum espresso because of which it fails to draw the separate bands for the given structure.
Can anyone suggest a way to draw the separate band-structures for a spin polarized system? Thanks in advance!
N.B. The normal way (using plotband.x in QE) of bandplotting works. This question is specifically for the band-structure of spin polarized system. Please let me know if anyone has done this already &/ has an idea about this.
Hi,
Give a look on the below link
http://web.mit.edu/espresso_v6.1/i386_linux26/qe-6.1/Doc/INPUT_BANDS.html#idm140066863768848
spin_component = INTEGER
In the lsda case select: 1 = spin-up 2 = spin-down
Run bands.x with spin_component = 1, so that you will get bands.dat file for spin-up and similarly with spin_component = 2 for down spin
Thank you.
I have one more question.
After following the procedure you have written, the .dat file which is created by bands.x executable, I am getting only 8 nks. Ideally, I should be getting the nks same as the number I provided in bands calculation. I am attaching the output of that file. This is for the spin component = 1 (i.e. for spin-up)
Any idea why is that the case?
Thanks.
For such case you have to run bands.x twice once for up spin (nspin=1) and other for spin down (nspin =2)
I have attached the sample input files to make things easily understandable.
So upbands.in will generate a file uphsl.dat and dwnbands.in will generate a file dwnhsl.dat.
You will finds the file with .gnu format which can be directly plotted using xmgrace.
The files uphsl.dat.gnu and dwnhsl.dat.gnu.
