[Pw_forum] Strange Lowdin Charges

[Wang Beizhou]

Dear Dr. Arles V. Gil Rebaza，

Thanks for your reply. The prefix tags should not be the reason because I have remove some part of prefix in scf input file just for privacy and forgot to change it in lowdin charge input file. Anyway, all the prefix and outdir tags should be same in real calculations.

Best,

Beizhou





-------- 原始信息 --------
由： "Arles V. Gil Rebaza" <arvifis at gmail.com>
日期: 2017/6/18 05:31 (GMT+08:00)
收件人： PWSCF Forum <pw_forum at pwscf.org>
主题： Re: [Pw_forum] Strange Lowdin Charges

Dear Wang, the "prefix" tag in your scf and projwfc input file, must be the same.

Best

Dr. Arles V. Gil Rebaza
IFLP - Argentina.

El sábado, 17 de junio de 2017, Wang Beizhou <magicmonk at hotmail.com<mailto:magicmonk at hotmail.com>> escribió:
Dear all,

I encountered a problem when I calculated lowdin charges for MoS2 edge using projwfc.x of QE-5.4.0 . Part of the results was shown below:
-------------------------------------------------------------------
     Atom #  69: total charge =   5.7346, s =  1.5888, p =  4.1458, d =  0.0000,
     Atom #  70: total charge =   5.7839, s =  1.6118, p =  4.1721, d =  0.0000,
     Atom #  71: total charge =  14.2310, s =  2.3482, p =  6.8669, d =  5.0159,
     Atom #  72: total charge =  14.4296, s =  2.2972, p =  6.9909, d =  5.1415,
     Atom #  73: total charge =   5.8638, s =  1.6250, p =  4.2388, d =  0.0000,
     Atom #  74: total charge =   5.6876, s =  1.5856, p =  4.1019, d =  0.0000,
     Atom #  75: total charge =   5.6885, s =  1.5857, p =  4.1028, d =  0.0000,
     Atom #  76: total charge =  14.4255, s =  2.3021, p =  6.9983, d =  5.1251,
     Atom #  77: total charge =  14.3017, s =  2.3619, p =  6.9347, d =  5.0051,
------------------------------------------------------------------------
As you could see from the results above, the charges for Mo atom(#71) is 14-14.2310<0 while charges of S atom(#70) is 6- 5.7839>0. This is weird and abnormal. I have also tested MoS2 cluster and the results seemed to be reasonable. The input files for SCF and lowdin charges calculations were shown below:
------------------------------------------------
&CONTROL
  calculation='scf',
  restart_mode='from_scratch',
  nstep=55,
  iprint=10,
  etot_conv_thr=1.94d-04,
  forc_conv_thr=1.94d-03,
  pseudo_dir="/disk/rdisk09/upf_files",
  disk_io='low',
  wf_collect=.true.
  outdir='/disk/rdisk09/pw-wfc',
  prefix='pure',
  max_seconds=1700.0D5,
/

&SYSTEM
  ibrav=0,
  nat=96,
  ntyp=2,
!  nbnd=230,
  ecutwfc=30,
  ecutrho=240,
  nspin=1,
  input_dft='pbe',
  occupations='smearing',
  smearing='gaussian'
  degauss=2d-03,
/

&ELECTRONS
  conv_thr=1d-06,
  diago_david_ndim=6,
  mixing_mode='local-TF',
  mixing_ndim=20,
  mixing_beta=0.2d0,
  electron_maxstep=150,
/

&IONS
  ion_dynamics='bfgs',
  upscale=10.0,
!  trust_radius_ini=0.756d0,
!  trust_radius_min=0.378d0,
!  trust_radius_max=0.945d0,
/

ATOMIC_SPECIES
Mo   1.0  Mo.pbe-spn-rrkjus_psl.0.2.UPF
S    1.0  S.pbe-n-rrkjus_psl.0.1.UPF

CELL_PARAMETERS {bohr}
             37.03839294358365    0.00000000000000    0.00000000000000
              0.00000000000000   47.78739424883790    0.00000000000000
              0.00000000000000    0.00000000000000   22.03604620987204

ATOMIC_POSITIONS {bohr}
Mo      16.178904896   0.021011000  16.923677623
Mo      21.147881968   2.974967793  16.894101816
……
……
S       35.139272622  41.823716049  16.906787202
S       35.111475195  47.636308872  17.054520301

K_POINTS gamma
------------------------------------------------------------------------------
&PROJWFC
    outdir='/nfs/disk2/beizhwang3/software/pw-wfc'
    prefix='pure'
    Emin=-20.0, Emax=20.0, DeltaE=0.01
    ngauss=0, degauss=0.002
/
-------------------------------------------------------------------------------
Any advice will be appreciated. Thanks for your kind attention!

Best,

Beizhou Wang
City University of Hong Kong


--
###--------->   Arles V.   <---------###
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20170618/473d1dd1/attachment.html>