Calculations with SOC

[Cyrille]

Hi all

I have a question regarding calculations with spin-orbit coupling.
I am used to do such calculations with quantum-espresso and one needs a special relativistic pseudo to include SOC.
With ATK I have noticed that it is possible with the same openmx pseudo to perform both collinear (i.e. without SOC) and SOC calculations.
Does-it mean it is possible to "switch off" the spin-orbit coupling using a relativistic pseudo?
Or am I doing something wrong?

thanks in advance

Cyrille

========================
Cyrille Barreteau
CEA Saclay, IRAMIS, SPEC Bat. 771
91191 Gif sur Yvette Cedex, FRANCE
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
DTU Nanotech
Ørsteds Plads, building 345E
DK-2800 Kgs. Lyngby,  DENMARK
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+33 1 69 08 29 51 /+33  6 47 53 66 52  (mobile)   (Fr)
+45    45 25 63 12/ +45     28 72 55 18  (mobile)  (Dk)
email:     cyrille.barreteau@cea.fr  /cyrbar@nanotech.dtu.dk
Web:     http://iramis.cea.fr/Pisp/cyrille.barreteau/
========================

[Anders Blom]

Yes. While it's a requirement to use the relativistic pseudopotentials to use spin-orbit coupling, you can use them in a collinear approximation just fine.

[Cyrille]

OK thanks that's a very good point.
It means that you can perform an initial collinear calculation to get a good starting point.

BTW is it possible to perform some tests calculations where SOC is multiplied by a given factor?

Cyrille

[Anders Blom]

It means that you can perform an initial collinear calculation to get a good starting point.

That is, in fact, precisely what you should do!

BTW is it possible to perform some tests calculations where SOC is multiplied by a given factor?

With DFT, the soc is computed from first principles, so no. I wasn't aware this was relevant, is it a technique used elsewhere? Refs?

[Cyrille]

By the way I have tried to start a SOC calculation from a collinear one using OMX pseudo and the 2014 version.
I get the following error:
Traceback (most recent call last):
  File "/tmp/7130187185888656.py", line 69, in <module>
    initial_spin=initial_spin, 
  File "./zipdir/NL/CommonConcepts/Configurations/AtomicConfiguration.py", line 1157, in setCalculator
NL.ComputerScienceUtilities.Exceptions.NLValueError: The initial state must have the same spin type as the new configuration.

Am I doing something wrong or is this option not available in 2014 version of ATK?

Cyrille

[Anders Blom]

Only available in 2015

[Cyrille]

OK thanks... We should get the 2015 licence soon in principle:-)