Page 1 of 1
Spin Orbit Coupling Calculation in VASP 6.4.3
Posted: Mon Jan 13, 2025 7:41 am
by trongnhan_duong
Dear VASP team,
I have a question about SOC calculation,
I want to optimize a structure of material with SOC mode and I have set
parameters in INCAR file as follow:
ISMEAR=0
SIGMA=0.05
RWIGS=1.217
ISIF=3
PREC=Accurate
ISPIN=2
ISTART=0
ICHAG=11
MAGMOM=2*1
LSORBIT=.TRUE.
LNONCOLLINEAR=.TRUE.
SAXIS=0 1 0
ISYM = 0
NSW=0
IBRION=-1
LWAVE=.FALSE.
LREAL=.FALSE.
EDIFF=1E-8
EDIFFG=-0.01
But when I submit this job, I received an error notification in OUTCAR file
that: "non collinear calculations require that VASP is compiled without the
flag -DNGXhalf and -DNGZhalf".
How can I fix this error?
Thank you in advance.
Best wishes,
Nhan.
Re: Spin Orbit Coupling Calculation in VASP 6.4.3
Posted: Mon Jan 13, 2025 10:45 am
by ahampel
Dear Nhan,
thank you for contacting us via the forum.
The error message you received is most likely due to the fact that you tried to run a non-collinear calculation using the vasp_std
executable. However, for non-collinear calculation you have to use the "vasp_ncl" executable, which can be found along the vasp_std
and vasp_gam
binary in the VASP installation directory. Please double check that you did start the calculation / job with this executable. See also wiki/index.php/LNONCOLLINEAR for details.
Let me know if this solves your problem.
Best,
Alex H.
Re: Spin Orbit Coupling Calculation in VASP 6.4.3
Posted: Wed Jan 15, 2025 5:35 am
by hszhao.cn@gmail.com
See here for the related discussion.
Regards,
Zhao
Re: Spin Orbit Coupling Calculation in VASP 6.4.3
Posted: Wed Jan 15, 2025 8:05 am
by ahampel
Thank you Zhao for linking this post. Indeed if you are using VASP 6.5.0 there would be a conflict in your input tags. However, I do believe you are using VASP 6.4.3 correct?
Re: Spin Orbit Coupling Calculation in VASP 6.4.3
Posted: Wed Jan 15, 2025 10:59 am
by hszhao.cn@gmail.com
I use VASP 6.5.0 as I have pointed out in the linked example above. But as we all know, this is a same problem in nature, which is why I linked to the above example in this issue.
Regards,
Zhao