[Pw_forum] quadrupole monents & GIPAW & a problem

[Davide Ceresoli]

Dear Zibi,
     could you check with me whether your crystal is the gamma
phase? experiments report Cq=0.042 MHz and eta=0.0 (Vorotilova,
Dmitrieva and Samoson, 1987). B is an nearly cubic environment
and the tetragonal symmetry should yield an axial tensor, hence
vanishing eta.

Anyway, I recognize that B.pbe-tm-gipaw.UPF has some problems.
It did work for NMR of small molecules, but I never tested it
in solid state systems, I'm sorry.

The last thing to check, is the tensor convention. I think I'm
using the Haeberlen convention, which is the same used by the
Simpson code 
(http://anorganik.uni-tuebingen.de/klaus/nmr/index.php?p=conventions/csa/csa).

HTH.

Best,
     Davide



On 09/11/2012 06:00 PM, Zbigniew Lodziana wrote:
> Dear users & developers,
>
> While calculating quadrupole moments with GIPAW I have experienced a strange
> output that confuses me and it is difficult for interpretation, at least to me.
>
> Performing calculations for LiBO2 (with occupancies = fixed!) efg calculations
> provide for Boron:
> (a)	Cq=   0.0208 MHz    eta= 0.00000
> (b)	Cq=   2.0747 MHz    eta= 0.00000
> (c)	Cq=   0.0228 MHz    eta= 0.00000
>
> Where
> (a) is with TM potential
> [https://sites.google.com/site/dceresoli/pseudopotentials/B.pbe-tm-new-gipaw-dc.UPF?attredirects=0];
>
> (b) also TM
> [https://sites.google.com/site/dceresoli/pseudopotentials/B.pbe-tm-gipaw.UPF?attredirects=0];
>
> (c) with ultrasoft.
>
> Experimental value is Cq=2.56, eta=0.6; theoretical report Cq=2.552, eta=0.54
>
> Test calculations with other more accurate potentials unfortunately do not
> solve problem.
>
> And the problem is that the worst description of the valence states provides
> (at least numerically) best Cq; besides it is difficult to understand how
> small changes in the valence state might result in so large – two orders of
> magnitude – changes in the core region. For other systems changes are even 3
> orders.
>
> Structural parameters are reasonable.
>
> For oxygen I did not found such a large variations, for Al Cq is reasonable.
>
> Does anyone experienced similar problems or has any hint where the problem
> could origin form?
>
> Thank you in advance,
> Best regards,
>
> Zibi
>
>
> Pw.x – 5.0
> Gipaw.x 5.0
> (version 4.3a gives similar results)
> -----------------------------------
> Pw.x
>
> &control
>      calculation = 'scf'
>      prefix = 'LiBO2'
>      restart_mode = 'from_scratch'
>      pseudo_dir = './pseudo/'
>      outdir = './tmp/'
>      verbosity = 'high'
>      wf_collect=.true.
> /
> &system
>      ibrav = 0
>      celldm(1) = 1.0
>      nat = 16
>      ntyp = 3
>      ecutwfc = 110
>      ecutrho = 1000
>      spline_ps = .true.
>      occupations  =  'fixed'
> /
> &electrons
>      diagonalization = 'david'
>      diago_thr_init = 1e-4
>      mixing_mode = 'plain'
>      mixing_beta = 0.1
>      conv_thr =  1e-10
> /
>
> ATOMIC_SPECIES
> Li 6.941  Li.pbe-tm-gipaw-dc.UPF
> B  10.811 B.pbe-tm-gipaw.UPF
> O  15.999 O.pbe-tm-new-gipaw-dc.UPF
>
> CELL_PARAMETERS (alat=  1.00000000)
>     8.020588949   0.000000000   0.000000000
>     0.000000000   8.020588949   0.000000000
>     0.000000000   0.000000000  12.497695405
>
> ATOMIC_POSITIONS (crystal)
> Li       0.000000000   0.000000000   0.500000000
> Li       0.500000000   0.000000000   0.249999998
> Li       0.500000000   0.500000000   0.000000000
> Li       0.000000000   0.500000000   0.750000002
> B        0.000000000   0.000000000   0.000000000
> B        0.500000000   0.000000000   0.750000000
> B        0.500000000   0.500000000   0.500000000
> B        0.000000000   0.500000000   0.250000000
> O        0.154649384   0.249987860   0.125014658
> O        0.845350616   0.750012140   0.125014658
> O        0.249987860   0.845350616   0.874985342
> O        0.750012140   0.154649384   0.874985342
> O        0.345350616   0.250012141   0.625014658
> O        0.654649384   0.749987859   0.625014658
> O        0.749987859   0.345350616   0.374985342
> O        0.250012141   0.654649384   0.374985342
>
> K_POINTS automatic
> 4 4 4   1 1 1
> ------------------------------------------
> Gipaw.x
>
> &inputgipaw
>          job = 'efg'
>          prefix = 'LiBO2'
>          tmp_dir = './tmp/'
>          iverbosity = 11
>          spline_ps = .true.
>          Q_efg(1) = 4.06    ! 7Li
>          Q_efg(2) = 4.06    ! 11B
>          Q_efg(3) = 2.55    ! 17O
> !       q_gipaw    = 0.01
> /
> Q_efg for Li might not be perfect but shall not matter.
>
>

-- 
+--------------------------------------------------------------+
   Davide Ceresoli
   CNR Institute of Molecular Science and Technology (CNR-ISTM)
   c/o University of Milan, via Golgi 19, 20133 Milan, Italy
   Email: davide.ceresoli at istm.cnr.it
   Phone: +39-02-50314276, +39-347-1001570 (mobile)
   Skype: dceresoli
+--------------------------------------------------------------+