Is POTCAR(PAW) included the relativistic effects?

Message

pga · #1 Post by **pga** » Tue Feb 28, 2006 4:54 pm

Dear all,
I would like to simulate an actinide metal but I don't know whether the PAW-POTCAR file is included the relativistic effects such as spin orbit coupling.
Best regards

pga · #2 Post by **pga** » Tue Feb 28, 2006 9:57 pm

If no, How should I do?

#3 Post by **admin** » Wed Mar 01, 2006 10:06 am

the pseudopotential generation package includes sclalar relativistic effects. Spin orbit coupling is is added if you set
LSORBIT=True
in INCAR.
please have a look at the online handbook
http://cms.mpi.univie.ac.at/vasp/vasp/node144.html
for a detailed description.

tsemi · #4 Post by **tsemi** » Wed Jan 20, 2010 3:40 pm

Hello,

I am not clear how VASP includes relativistic effects. In Hafner's 2008 paper, 'Ab-Initio Simulations of Materials Using VASP:
Density-Functional Theory and Beyond', he states: 'In the standard mode, VASP performs a fully relativistic calculation
for the core-electrons and treats valence electrons in a scalar
relativistic approximation.'

The 'head admin' comments indicate that pseudopotentials are generated using the scalar relativistic approximation. I believe that the pseudopotentials include the core electrons.

Could you please clarify how VASP incorporates relativistic effects? Or point me toward a reference that will explain?

Thank you,
torey semi

panda · #5 Post by **panda** » Mon Jan 25, 2010 12:14 am

This reference might explain what you are looking for:

http://prb.aps.org/abstract/PRB/v76/i2/e024309

tsemi · #6 Post by **tsemi** » Mon Jan 25, 2010 3:56 pm

Hi,
Thanks for the reference; I appreciate your help. I'm mainly trying to find out what equation VASP is solving; does it use the relativistic K-S equations when solving for the core electrons? How does VASP include 1)the scalar relativistic approximation and 2)SOC?

Are the 'regular' K-S equations solved, and the pseudopotentials (PAWs) adjust for relativistic effects?

Thanks again,
tsemi

panda · #7 Post by **panda** » Mon Jan 25, 2010 11:20 pm

Scalar relativistic effects are included in the pseudopotentials. The "core electrons" are up to you to select, there are core valence and semi-core valence pseudopotentials available to you when you purchase the software, they should be included in your pseudopotential directory like this:

V
V_pv
V_sv

for V, as an example, where V would include ONLY the core-electrons in the potential, _pv would include p-states as semi-core valence states, _sv would include s-states as semi core valence states.

Make sense? The K-S equations are solved iteratively and there is a detailed description in the manual, along with references.

tsemi · #8 Post by **tsemi** » Tue Jan 26, 2010 9:20 pm

Thank you very much. This makes sense! So if I choose V for my pseudopotential, then the corresponding core electrons are treated fully relativistically (per Hafner)- meaning that the relativistic K-S equations are solved for these core states?

panda · #9 Post by **panda** » Fri Feb 05, 2010 7:14 pm

yes, exactly, you've got it!