[QE-users] Magnetism of holmium

Nicola Marzari nicola.marzari at epfl.ch
Tue Oct 2 00:01:49 CEST 2018


On 01/10/2018 21:21, Lorenzo Paulatto wrote:
> I've recently been using the spdf pseudopotentials for Ce and Gd from 
> the pslibrary. I can say that:
> 1. They require a monster cutoff (200 Ry for Ce, and 100 for Gd)
> 2. They require +U, especially Ce
> 3. Results were quite good
> 


Lorenzo - sure, but why? Wentzcovitch pseudopotential are more accurate,
and they require roughly ~40Ry.

			nicola

> -- 
> Lorenzo Paulatto
> Written on a virtual keyboard with real fingers
> 
> On Mon, 1 Oct 2018, 18:28 Nicola Marzari, <nicola.marzari at epfl.ch 
> <mailto:nicola.marzari at epfl.ch>> wrote:
> 
> 
> 
>     Dear Martin,
> 
>     we tested a couple of pseudo libraries with the f frozen, but apart
>     from
>     La they do not work that well in general. As mentioned by Christopher,
>     you want the f.
> 
>     Here are a few tests:
>     https://www.materialscloud.org/discover/sssp/plot/efficiency/Ho
> 
>     and if you eg brose the provennce tree for the Ho US pslibrary 1.0
>     until you get to a PWscf calculation:
> 
>     https://www.materialscloud.org/explore/sssp/details/f2383960-43b6-4bf6-b080-b49e43c6ce12?nodeType=CALCULATION
> 
>     you can see that the pseudo tested is
>     Ho.pbe-spdfn-rrkjus_psl.1.0.0.UPF
> 
>     I would tend to recommend the Wentzocvitch series, but it's true that
>     all rare earths are rarely tested. In addition, they have strong
>     multiconfigurational character, so standard LDA/GGA/GGA+U/HSE
>     DFT doesn't do that well.
> 
>                                      nicola
> 
> 
> 
> 
>     On 01/10/2018 10:30, Ing. Martin Matas wrote:
>      > Dear experts,
>      >
>      > I appreciate that many new ultrasoft pseudopotentials for rare
>     earths by Andrea Dal Corso have recently appeared on QE homepage. I
>     performed calculations with a 2-atom primitive cell of metal holmium
>     using Ho.pbe-spdn-rrkjus_psl.1.0.0.UPF, with both ferromagnetic and
>     antiferromagnetic starting configurations.
>      >
>      > The resulting magnetization is always zero, not depending on
>     calculation parameters (like starting_magnetization, occupation type
>     and degauss, or energy cutoffs). However, exactly the same
>     calculations using a norm-conserving holmium pseudopotential give
>     results with a non-zero magnetization what I consider correct since
>     holmium at zero temperature is ferromagnetic.
>      >
>      > Did anyone test these new ultrasoft pseudopotentials in this way
>     and is there a way how to get magnetically correct results using them?
>      > Thanks a lot.
>      >
>      > Martin Matas
>      > PhD student
>      > University of West Bohemia
>      > Czech Republic
>      >
>      > _______________________________________________
>      > users mailing list
>      > users at lists.quantum-espresso.org
>     <mailto:users at lists.quantum-espresso.org>
>      > https://lists.quantum-espresso.org/mailman/listinfo/users
>      >
> 
> 
>     -- 
>     ----------------------------------------------------------------------
>     Prof Nicola Marzari, Chair of Theory and Simulation of Materials, EPFL
>     Director, National Centre for Competence in Research NCCR MARVEL, EPFL
>     http://theossrv1.epfl.ch/Main/Contact http://nccr-marvel.ch/en/project
>     _______________________________________________
>     users mailing list
>     users at lists.quantum-espresso.org
>     <mailto:users at lists.quantum-espresso.org>
>     https://lists.quantum-espresso.org/mailman/listinfo/users
> 
> 
> 
> _______________________________________________
> users mailing list
> users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users
> 


-- 
----------------------------------------------------------------------
Prof Nicola Marzari, Chair of Theory and Simulation of Materials, EPFL
Director, National Centre for Competence in Research NCCR MARVEL, EPFL
http://theossrv1.epfl.ch/Main/Contact http://nccr-marvel.ch/en/project


More information about the users mailing list