[Wien] Error while treating Sm 4f states as core

[Xavier Rocquefelte]

Dear Anup,

The method you want to use is an "old" one which is now not anymore 
used, except if you really have no other choice.

It is preferred to use an Hubbard (or an on-site hybrid) correction 
which will allow to correct the DFT error concerning the treatment of 
the 4f states of Sm.

I recommend you to redo the calculation using LDA+U or GGA+U with a 
large U value. Look at the litterature of previous DFT+U calculations 
for Sm. I imagine that a value of about 8 eV should do the job if you 
don't need an exact position of the 4f states but just want to push them 
in the semi-core region. In such a case, you will have to do 
spin-polarized calculations.

Best regards

Xavier

Le 19/10/2019 à 17:45, Anup Shakya a écrit :
> Dear All,
>
> I am trying to perform non-magnetic calculations for SmBi treating Sm 
> 4f electrons in the core region. I have followed the example given for 
> Yb by Prof. Blaha. The steps that I have performed are as follows.
>
> I performed a normal non-magnetic scf calculation using run_lapw and 
> from there found out where the Sm 4f states are from the case. scf 
> file. In case.inc file, I added two lines 4, 3, 5 and 4, -4, 0, (to 
> keep Sm 4f 5/2 and Sm 4f 7/2 states in the core and since Sm in +3 
> configuration has 5 electrons, so I have kept the occupations no as 5 
> for Sm 4f 5/2 and 0 for Sm 4f 7/2) increased the number of orbitals 
> from 14 to 16 and added a downward shift of 0.6 Ry. In case.in1, I 
> changed the value for Sm 4f states from
>
> 3    0.30     0.0010 CONT 1
> to
> 3   -1.00   0.0010 CONT 1,
> so that the Sm 4f states will not be found by lapw1.
> After that in case.in2 I removed 5 electrons and reduced the number NE
> TOT             (TOT,FOR,QTL,EFG,FERMI)
>    -12.0    31.0   0.50 0.05  1   EMIN, NE, ESEPERMIN, ESEPER0, iqtlsave
>
> from 31 to 26 as follows:
>
> OT             (TOT,FOR,QTL,EFG,FERMI)
>    -12.0    26.0   0.50 0.05  1   EMIN, NE, ESEPERMIN, ESEPER0, iqtlsave
> After this I ran dstart first and then did a normal scf run. For the 
> first 39 cycles there was no problem but then an error came in the 
> 40th scf cycle as follows:
>
> LAPW2 - FERMI; weights written
> L2main - QTL-B Error
> L2main - QTL-B Error
>
> cat *.error
>  'l2main' - QTL-B.GT.15., Ghostbands, check scf files
>  'l2main' - QTL-B.GT.15., Ghostbands, check scf files
>
> lse
> 4 -rw-r--r-- 1 root root 54 Oct 19 00:58 uplapw2_29.error
> 4 -rw-r--r-- 1 root root 54 Oct 19 00:58 uplapw2_2.error
> 4 -rw-r--r-- 1 root root 54 Oct 19 00:58 uplapw2_30.error
> 4 -rw-r--r-- 1 root root 54 Oct 19 00:58 uplapw2_31.error
> 4 -rw-r--r-- 1 root root 54 Oct 19 00:58 uplapw2_32.error
> 4 -rw-r--r-- 1 root root 54 Oct 19 00:58 uplapw2_33.error
> 4 -rw-r--r-- 1 root root 54 Oct 19 00:58 uplapw2_34.error
> 4 -rw-r--r-- 1 root root 54 Oct 19 00:58 uplapw2_35.error
> 4 -rw-r--r-- 1 root root 54 Oct 19 00:58 uplapw2_36.error
>
> :FER  : F E R M I - ENERGY(TETRAH.M.)=   0.0628473771
> I followed previous threads and then tried everything as suggested by 
> Prof. Blaha in the link as follows:
> http://susi.theochem.tuwien.ac.at/reg_user/faq/qtlb.html. But I have 
> not been able to overcome this problem.
>
> I have two questions.
>
> 1) Is there any mistake in the steps which I have performed for 
> treating Sm 4f as core states. Should I do spin polarized calculations?
>
> 2) If there is no problems with my steps, then how should I overcome 
> this error. I have tried everything possible from my side. I would be 
> grateful if anyone in the forum could help me to overcome this 
> problem. If you need any more information, please let me know. Looking 
> forward to hearing from you.
>
> Sincerely,
> Anup Pradhan Sakhya (Ph.D.)
> Visiting Post-Doctoral Fellow
> DCMP&MS, TIFR, Mumbai
> Alternative Email:anup.sakhya at tifr.res.in 
> <mailto:rajib.mondal at tifr.res.in>
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> SEARCH the MAILING-LIST at:  http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html

-- 
------------------------
Institut des Sciences Chimiques de Rennes (ISCR)
Univ Rennes - CNRS - UMR6226, France
https://iscr.univ-rennes1.fr/cti/people/permanent-staff/rocquefelte-xavier
------------------------
ICAMM2019 : VASP Workshop and International Materials Modelling Conf
June 26-July 3 2019, Rennes France
icamm2019.sciencesconf.org  : Registration opening soon, see you there!

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20191020/b1cd90ea/attachment.html>