[QE-users] ?==?utf-8?q? Anisotropic exchange calculation

[BARRETEAU Cyrille]

OK

OK I did not understand precisely your initial message.
Indeed my message concerned the isotropic exchange.

If you need to evaluate some kind of anisotropy you indeed must include spin-orbit.
If the evaluation of your magnetic parameters involves only collinear configurations (up, dn etc..) then you can use the force theorem I think like for the calculation of the MCA.
It should be possible also to use the FT in case of non-collinear situation but to the best of my knlowledge it has not been implemented yet.
In that case you would need to rotate the charge density differently on each atom (and therefore project on atoms). However presently for the MCA it is only a global rotation rotation that is allowed.

This type of procedure is probably easier in a code using a local basis set.

The other strategy is to do a scf calculation including SOC but of course that's more time consuming...

cyrille


========================
Cyrille Barreteau
CEA Saclay, IRAMIS, SPEC Bat. 771
91191 Gif sur Yvette Cedex, FRANCE
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+33 1 69 08 38 56 /+33  6 47 53 66 52  (mobile)
email:     cyrille.barreteau at cea.fr
Web:     http://iramis.cea.fr/Pisp/cyrille.barreteau/
========================

________________________________________
De : users [users-bounces at lists.quantum-espresso.org] de la part de Dorye Esteras Cordoba [dorye.esteras at uv.es]
Envoyé : lundi 16 novembre 2020 19:47
À : Quantum ESPRESSO users Forum
Objet : Re: [QE-users] ?==?utf-8?q?  Anisotropic exchange calculation

Dear Cyrille,

Thank you for your prompt response. About your first comment, the procedure you describe using collinear  states, please correct me if I am wrong, corresponds to the exchange calculations in the isotropic Heisenberg Hamiltonian. Or does it mean instead that could I obtain the anisotropic exchange parameters using noncolin=.true. without spin-orbit?.

Being more explicit, my plan is to make a 2x2 supercell in order to calculate the first neighbors exchange. I would perform a nspin=2 scf calculation with starting_magnetization=3 in the Cr atoms, using scalar relativistic pseudopotentials. Then perform noncollinear nscf calculations with SOC (and fully relativistic pseudopotentials) using angles to put the spins in the configurations suggested in the equation A17 in https://doi.org/10.1039/C2DT31662E (section D in the appendix). That would mean to maintain 2 spins (which represent the first neighbors) collinear and the rest of the spins perpendicular to them.

Thank you again!
Dorye


> Dear Dorye
>
> Could you be more explicit?
> If you want to estimate J exchange parameters from energy differences of  collinear states (using a super cell and various UP and DN configurations) you do not need any non-collinear calculation.
> I you want to use a non-collinear approach with small deviation from a given stable (or metastable state) I would suggest to use a penalization technique (atomic theta..) and map it onto an Heisenberg Hamiltonian by fitting an E(theta) for example.
>
> Cyrille
>
>
> ========================
> Cyrille Barreteau
> CEA Saclay, IRAMIS, SPEC Bat. 771
> 91191 Gif sur Yvette Cedex, FRANCE
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> +33 1 69 08 38 56 /+33  6 47 53 66 52  (mobile)
> email:     cyrille.barreteau at cea.fr
> Web:     http://iramis.cea.fr/Pisp/cyrille.barreteau/
> ========================
>
> ________________________________________
> De : users [users-bounces at lists.quantum-espresso.org] de la part de Dorye Esteras Cordoba [dorye.esteras at uv.es]
> Envoyé : lundi 16 novembre 2020 17:00
> À : users at lists.quantum-espresso.org
> Objet : [QE-users] Anisotropic exchange calculation
>
> Dear community,
>
> I am interested in calculating the CrBr3 J exchange parameters of an anisotropic Heisenberg Hamiltonian as it is shown in this paper:
> https://doi.org/10.1039/C2DT31662E (Equations A16 and A17 on section D in the appendix)
>  My idea would be to perform a collinear spin-polarized scf calculation (i.e. using scalar relativistic pseudopotentials), and then perform a non collinear nscf calculation for each of the spin configurations, following a similar procedure as in the MAE example https://gitlab.com/QEF/q-e/-/tree/f184591e9f34cfcc7767505a23977a92286e8ba6/PP/examples/ForceTheorem_example
> Would this be reasonable or should I do a non collinear scf calculation for each scenario and extract Jxx,Jyy,Jzz from the total energies? Has someone experience with this kind of calculations?
>
> Thank you in advance
> Dorye L. Esteras
> Predoctoral researcher
> University of Valencia
>
>
>
>
>
>
>
> _______________________________________________
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu)
> users mailing list users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users
> _______________________________________________
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu)
> users mailing list users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users





_______________________________________________
Quantum ESPRESSO is supported by MaX (www.max-centre.eu)
users mailing list users at lists.quantum-espresso.org
https://lists.quantum-espresso.org/mailman/listinfo/users