[QE-users] Spin counting in hp.x with paramagnetic metals

Thu Sep 14 12:04:10 CEST 2023

Dear Nicola,

Thank you for the further advice, I'd seen in the literature that there were a number of different ways to model paramagnetic systems but I wasn't sure necessarily what the best approach was (balancing accuracy and computational cost).

In any case, for the system I'm currently interested in I thought I'd use a non-magnetic calculation for initial modelling since the computational cost is significantly smaller but for future more accurate calculations I will definitely look to use a basic AFM state (for simplicity). Fortunately, I do seem to be lucky that the non-magnetic lattice parameters and volumes with a Hubbard U seem to agree reasonable well with the experimental values.

Best wishes,

Theo

-----Original Message-----
From: users <users-bounces at lists.quantum-espresso.org> On Behalf Of Nicola Marzari via users
Sent: Tuesday, September 12, 2023 11:03 AM
To: users at lists.quantum-espresso.org
Subject: Re: [QE-users] Spin counting in hp.x with paramagnetic metals

Dear Theo,

unrelated, but it's not a super good idea to describe a paramagnetic state with a non magnetic calculation - many paramagnetic systems have still strong local moments, that are disordered. At high pressure the local moments can disappear (e.g. iron at the ~330GPa of the liquid-solid boundary inside the earth), but if you are in a phase where they exist, and you neglect them with a non-magnetic calculation, your energetics will be very much off, and the volumes will not be right.

Better to use at the very least a simple AFM state, or even better a special quasirandom structure to choose a pattern of up/down dipoles.

nicola

On 12/09/2023 11:41, Theo Weinberger wrote:
> Dear Iurii,
>
> Yes, to be more precise I mean that when I am looking at these
> systems, I am modelling them as being nonmagnetic and so I was
> wondering whether it was the spin degeneracy that was causing this difference.
>
> I am typically looking at pressure-induced structural transitions in
> systems with complex magnetic ground states. However, in some cases
> the exact magnetic configurations are unknown and so to simplify the
> problem (and since these structural transitions occur at high
> temperature in the paramagnetic state) I have been trying to perform
> the calculations in the nonmagnetic state. It is in these instances
> where I see this approximate doubling of U: i.e. when I allow the
> system to relax into a FM ordered state (with nspin=2) before using
> hp.x the value of U ends up being around half that of when I don't include spin effects (nspin=1).
>
> I hope this clarifies what I have been seeing.
>
> Best wishes,
>
> Theo Weinberger
>
> PhD Student
>
> University of Cambridge
>
> *From:*users <users-bounces at lists.quantum-espresso.org> *On Behalf Of
> *Iurii Timrov via users
> *Sent:* Tuesday, September 12, 2023 10:31 AM
> *To:* users at lists.quantum-espresso.org
> *Subject:* Re: [QE-users] Spin counting in hp.x with paramagnetic
> metals
>
> Dear Theo,
>
> How do you model the paramagnetic state? Please note that "paramagnetic"
> and "nonmagnetic" state is not the same thing.
>
> For nonmagnetic calculations there is a factor of 2 due to spin
> degeneracy when computing sums over electronic states (i.e. in charge
> density, occupation matrix, and other quantities).  If you take a
> nonmagnetic material (e.g. LiCoO2) and model it as a spin-polarized
> system (nspin=2), the value of U will be the same as when modeling it
> as nonmagetic (because the magnetization will be zero). So I do not
> understand why do you have a factor of 2 difference for U in your
> simulations.
>
> Greetings,
>
> Iurii
>
> --
> Dr. Iurii TIMROV
> Senior Research Scientist
>
> Theory and Simulation of Materials (THEOS)
>
> Swiss Federal Institute of Technology Lausanne (EPFL)
>
> CH-1015 Lausanne, Switzerland
> +41 21 69 34 881
>
> http://peopl/
> e.epfl.ch%2F265334&data=05%7C01%7Ctiw21%40universityofcambridgecloud.o
> nmicrosoft.com%7Cecdeb6499ed44439532808dbb3787aec%7C49a50445bdfa4b79ad
> e3547b4f3986e9%7C1%7C0%7C638301102261046448%7CUnknown%7CTWFpbGZsb3d8ey
> JWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C300
> 0%7C%7C%7C&sdata=HaR%2BSGis8q9ifWtJw7EBKtUsnc4VLVnHByqw3k6CDbg%3D&rese
> rved=0
> <http://peop/
> le.epfl.ch%2F265334&data=05%7C01%7Ctiw21%40universityofcambridgecloud.
> onmicrosoft.com%7Cecdeb6499ed44439532808dbb3787aec%7C49a50445bdfa4b79a
> de3547b4f3986e9%7C1%7C0%7C638301102261046448%7CUnknown%7CTWFpbGZsb3d8e
> yJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C30
> 00%7C%7C%7C&sdata=HaR%2BSGis8q9ifWtJw7EBKtUsnc4VLVnHByqw3k6CDbg%3D&res
> erved=0>
>
> ----------------------------------------------------------------------
> --
>
> *From:*users <users-bounces at lists.quantum-espresso.org
> <mailto:users-bounces at lists.quantum-espresso.org>> on behalf of Theo
> Weinberger <tiw21 at cam.ac.uk <mailto:tiw21 at cam.ac.uk>>
> *Sent:* Tuesday, September 12, 2023 11:21:03 AM
> *To:* users at lists.quantum-espresso.org
> <mailto:users at lists.quantum-espresso.org>
> *Subject:* [QE-users] Spin counting in hp.x with paramagnetic metals
>
> Dear Quantum Espresso Users,
>
> I have been using the hp.x code to calculate the Hubbard-U corrections
> for correlated metallic systems in both their spin-polarised and
> paramagnetic states.
>
> In several materials I have noticed that the Hubbard-U value
> determined for a material in its paramagnetic ground state is
> approximately twice that compared to when a spin-polarised ground
> state is assumed (with all other parameters kept the same).
>
> I was wondering whether anyone had any insight into how the accounting
> for spins in occupied Hubbard states works for the hp.x code and
> whether the paramagnetic implementation of hp.x perhaps counts all
> spins states twice resulting in this larger value.
>
> Thank you in advance,
>
> Theo Weinberger
>
> PhD Student
>
> University of Cambridge
>
>
> _______________________________________________
> The Quantum ESPRESSO community stands by the Ukrainian people and
> expresses its concerns about the devastating effects that the Russian
> military offensive has on their country and on the free and peaceful
> scientific, cultural, and economic cooperation amongst peoples
> _______________________________________________
> Quantum ESPRESSO is supported by MaX
> (http://www/.
> max-centre.eu%2F&data=05%7C01%7Ctiw21%40universityofcambridgecloud.onm
> icrosoft.com%7Cecdeb6499ed44439532808dbb3787aec%7C49a50445bdfa4b79ade3
> 547b4f3986e9%7C1%7C0%7C638301102261046448%7CUnknown%7CTWFpbGZsb3d8eyJW
> IjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%
> 7C%7C%7C&sdata=SirkQlzrCTGUTvPf%2B%2FIFjFSTkWR6M%2FwTMapMvf59obM%3D&re
> served=0) users mailing list users at lists.quantum-espresso.org
> https://list/
> s.quantum-espresso.org%2Fmailman%2Flistinfo%2Fusers&data=05%7C01%7Ctiw
> 21%40universityofcambridgecloud.onmicrosoft.com%7Cecdeb6499ed444395328
> 08dbb3787aec%7C49a50445bdfa4b79ade3547b4f3986e9%7C1%7C0%7C638301102261
> 046448%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJ
> BTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=H53IRO63%2FRqZy28rA2
> %2BWiKTdSEeh0HDAwO4CtRhG8eU%3D&reserved=0

--
----------------------------------------------------------------------
Prof Nicola Marzari, Chair of Theory and Simulation of Materials, EPFL Director, National Centre for Competence in Research NCCR MARVEL, SNSF Head, Laboratory for Materials Simulations, Paul Scherrer Institut Contact info and websites at http://theossrv1.epfl.ch/Main/Contact

_______________________________________________
The Quantum ESPRESSO community stands by the Ukrainian people and expresses its concerns about the devastating effects that the Russian military offensive has on their country and on the free and peaceful scientific, cultural, and economic cooperation amongst peoples _______________________________________________
Quantum ESPRESSO is supported by MaX (http://www.max-centre.eu/)
users mailing list users at lists.quantum-espresso.org
https://lists.quantum-espresso.org/mailman/listinfo/users