[Wien] LSDA+U on Yb compounds

Wed Jul 26 21:38:12 CEST 2006

Dear Stefaan,

to get Yb3+ from Yb2+ you could superpose a small repulsive potential Vf
to the LDA+U potential. Vf is just a diagonal matrix = w*I. You have to
tune w so that desired Yb valency is obtained. Simple programming + small
UNIX script will do it.

Regards
Pavel

_________________________________________________
Dr. Pavel Novak
Department of Magnetism and Superconductivity
Institute of Physics AS CR
Cukrovarnicka 10, 162 53 Praha 6, Czech Republic
tel: +420 2 20 318 532
e-mail: novakp at fzu.cz

On Wed, 26 Jul 2006, Stefaan Cottenier wrote:

>
> All of these three methods give the charge within the muffin tin
> spere. Which sphere sizes are you using? I guess the 4f states are not
> completely confined within the spheres, and your 6.78 corresponds to 7
> electrons (a little bit of them being in the interstitial region). If
> that is true, then you have divalent Yb with 14 electrons.
>
> I never did it, but you can try to force trivalency by putting 13
> electrons in the density matrices, converging with -orbc (density
> matrices fixed), then release it and converge again with -orb -- and
> now hoping that this solution will be kept. I'm curious whether that
> will work. Has anybody different suggestions?
>
> Stefaan
>
>
> > Thank you for the prompt reply. I followed your instructions to find out the
> > 4f occupancy. Here are my results for YbAl3:
> >
> > :PCS001: PARTIAL CHARGES SPHERE =  1 S,P,D,F
> > :QTL001: 1.0705 2.9374 0.3427 6.7838  -up
> > :QTL001: 1.0705 2.9374 0.3427 6.7838  -dn
> >
> > :TRA001:  TRACE of UPUP MATRIX=   6.78385   0.00000
> > :TRA001:  TRACE of UPDN MATRIX=   0.00000   0.00000
> > :TRA001:  TRACE of DNDN MATRIX=   6.78385   0.00000
> >
> > FERMI ENERGY=   0.54730 -from .outputtup
> >  0.54800    ...     0.94   6.7828 -up
> >  0.54800    ...     0.94   6.7828 -dn
> >
> >  It seems that there are about 13.5 -13.6 electrons in  the 4f level,
> > thirteen of which should be below Femi energy and one partly occupied level
> > should be "pinned" at the Fermi level (correct me if I'm wrong!)- but all I
> > am geting are the filled states. I also want to mention that LSDA+U
> > calculation on a Ce compound gave one occupied state below fermi energy
> > (large,visible peak) and the rest of  the unoccupies states above EF.
> > It is also puzzeling to me that there is no difference between -up and -dn
> > states.
> > How can I "force" a divalent or trivalent solution?
> >
> > Thanks again,
> > Zsolt
> >
> >
> >
> >
> > ----- Original Message -----
> > From: "Stefaan Cottenier" <Stefaan.Cottenier at fys.kuleuven.be>
> > To: <wien at zeus.theochem.tuwien.ac.at>
> > Sent: Wednesday, July 26, 2006 11:02 AM
> > Subject: Re: [Wien] LSDA+U on Yb compounds
> >
> >
> >>
> >> Can you verify explicitely whether you have 13 or 14 4f-electrons? Do
> >> this by looking at the partial charges for f (:QTL), or by summing the
> >> diagonal elements of the up and dn density matrices, or by looking at
> >> the value of the integrated partial f-dos at the fermi level in
> >> case.outputt (and preferably do it by all three methods, they should
> >> give the same answer).
> >>
> >> If you really have 14 electrons, you might have ended up with divalent
> >> Yb. If you have 13 electrons, perhaps your unoccupied state is diffuse
> >> and not at first sight recognizable as a peak.
> >>
> >> Stefaan
> >>
> >>> I am performing LSDA+U calculations on a series of Yb compounds in which
> >>> the
> >>> valence of Yb is close to 3+. I was expecting to get a splitting between
> >>> the
> >>> occupied and unoccupied 4f states, but my DOS only shows the occupied
> >>> 4f-states located well below the Fermi level, without any empty
> >>> 4f-states.
> >>> I'm doing the following:
> >>>
> >>> runsp_lapw  -cc 0.001;
> >>> runsp_lapw -so -cc 0.001;
> >>> runsp_lapw -so -orb -cc 0.001.
> >>>
> >>> for DOS I do:
> >>> x lapw1 -orb -up/dn
> >>> x lapwso -up -orb
> >>> x lapw2 -qtl -so -up/dn.
> >>>
> >>>
> >>> It also seems that with LSDA+U there is no difference between -up and -dn
> >>> states. (and I'm not using runsp_c_lapw).
> >>> Is there any wien2k user who was able to observe the empty 4f states of
> >>> Yb,
> >>> located at or above Fermi level?
> >>>
> >>> Thanks,
> >>> Zsolt Rak
> >>>
> >>> _______________________________________________
> >>> Wien mailing list
> >>> Wien at zeus.theochem.tuwien.ac.at
> >>> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> >>>
> >>>
> >>
> >>
> >>
> >> --
> >> Stefaan Cottenier
> >> Instituut voor Kern- en Stralingsfysica
> >> K.U.Leuven
> >> Celestijnenlaan 200 D
> >> B-3001 Leuven (Belgium)
> >>
> >> tel: + 32 16 32 71 45
> >> fax: + 32 16 32 79 85
> >> e-mail: stefaan.cottenier at fys.kuleuven.be
> >>
> >>
> >> Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm
> >>
> >> _______________________________________________
> >> Wien mailing list
> >> Wien at zeus.theochem.tuwien.ac.at
> >> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> >>
> >>
> >
> > _______________________________________________
> > Wien mailing list
> > Wien at zeus.theochem.tuwien.ac.at
> > http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> >
> >
>
>
>
> --
> Stefaan Cottenier
> Instituut voor Kern- en Stralingsfysica
> K.U.Leuven
> Celestijnenlaan 200 D
> B-3001 Leuven (Belgium)
>
> tel: + 32 16 32 71 45
> fax: + 32 16 32 79 85
> e-mail: stefaan.cottenier at fys.kuleuven.be
>
>
> Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>