[Pw_forum] projwfc.x cannot produce the correct Lowdin charges or fail completely
Li, Run
run.li at my.und.edu
Fri Oct 3 20:18:46 CEST 2014
Dear all,
I tried to use projwfc.x to produce PDOS and lowdin charges. Two of my systems run into different problems while the others seemed okay. Could you please help me figure out what is going on?
1. A 73 atom molecule Zn(II) Phthalocyanine tetrasulfonic acid in 3D cell. PBE-D calculation with PAW and USPP potential. SCF in m-p smearing using gamma point. Nscf calculation is in the same parameter (no increase of k point, no tetrahedral occupation). Then projwfc.x output file gives really small numbers for psi^2 (<0.005, mostly 0.000), and Lowdin charge less than 0.005. The corresponding .wfc files can correctly generate charge distribution for pp.x, and Bader analysis gives reasonable charges on the atoms.
2. The molecule above put on top of a graphene sheet. PBE-D calculation with PAW and USPP potential. SCF in m-p smearing using gamma point. Case 1: No nscf calculation, directly calculate PDOS, projwfc.x works. Case 2: tetrahedral occupation with automatic k point (2 2 1 1 1 1), projwfc dies after checking Problem Size, gives: APPLICATION TERMINATED WITH THE EXIT STRING: Killed (signal 9)
My input/output files are listed below:
System1, scf//nscf:
&CONTROL
calculation = 'scf' , // calculation = 'nscf' ,
prefix = 'Zn' ,
nstep = 100 ,
forc_conv_thr= 5.0D-3,
pseudo_dir = '/share/apps/espresso/pseudo/',
/
&SYSTEM
ibrav = 8,
celldm(1) = 65.1924,
celldm(2) = 0.989743318610787,
celldm(3) = 0.76996,
nat = 73,
ntyp = 6,
ecutwfc = 47.0D0 ,
input_dft = pbe ,
occupations = 'smearing' ,
degauss = 0.02 ,
smearing = 'methfessel-paxton' ,
london= .true.,
london_s6= 0.75,
london_rcut= 200,
/
&ELECTRONS
electron_maxstep = 200,
scf_must_converge = .true. ,
conv_thr = 1.D-6 ,
adaptive_thr = .false. ,
/
ATOMIC_SPECIES
C 12.00000 C.pbe-n-kjpaw_psl.0.1.UPF
H 1.00000 H.pbe-kjpaw.UPF
O 16.00000 O.pbe-n-kjpaw_psl.0.1.UPF
N 14.00000 N.pbe-n-kjpaw_psl.0.1.UPF
S 32.00000 S.pbe-n-kjpaw_psl.0.1.UPF
Zn 65.39000 Zn.pbe-van.UPF
/
...
System1, PDOS:
&projwfc
outdir='./'
prefix= 'Zn'
io_choice='both'
Emin=-40.0, Emax=0.0, DeltaE=0.2
/
OUTPUT of PDOS:
...
state # 245: atom 73 (O ), wfc 2 (l=1 m= 2)
state # 246: atom 73 (O ), wfc 2 (l=1 m= 3)
k = 0.0000000000 0.0000000000 0.0000000000
e = -29.07327 eV
psi =
|psi|^2 = 0.009
e = -29.07218 eV
psi =
|psi|^2 = 0.005
e = -29.06742 eV
psi =
|psi|^2 = 0.004
e = -29.06691 eV
psi =
|psi|^2 = 0.005
...
|psi|^2 = 0.000
e = -0.21485 eV
psi =
|psi|^2 = 0.000
e = -0.17288 eV
psi =
|psi|^2 = 0.000
Lowdin Charges:
Atom # 1: total charge = 0.0050, s = 0.0004, p = 0.0000, d = 0.0046,
Atom # 2: total charge = 0.0011, s = 0.0003, p = 0.0008, d = 0.0000,
Atom # 3: total charge = 0.0011, s = 0.0003, p = 0.0008, d = 0.0000,
...
System 2, scf:
&CONTROL
calculation = 'scf' ,
prefix = 'Abs' ,
nstep = 100 ,
forc_conv_thr= 5.0D-3,
pseudo_dir = '/share/apps/espresso/pseudo/',
/
&SYSTEM
ibrav = 8,
celldm(1) = 65.1924,
celldm(2) = 0.989743318610787,
celldm(3) = 0.76996,
nat = 521,
ntyp = 6,
ecutwfc = 47.0D0 ,
input_dft = pbe ,
occupations = 'smearing' ,
degauss = 0.02 ,
smearing = 'methfessel-paxton' ,
london= .true.,
london_s6= 0.75,
london_rcut= 200,
/
&ELECTRONS
electron_maxstep = 200,
scf_must_converge = .true. ,
conv_thr = 1.D-6 ,
adaptive_thr = .false. ,
/
...
System2, nscf:
Change to : occupations = 'tetrahedra' ,
K_POINTS automatic
2 2 1 1 1 1
System2, pdos
&projwfc
outdir='./'
prefix='Abs'
io_choice='both'
Emin=-4.3784 Emax=-1.3784, DeltaE=0.02
ngauss=1, degauss=0.02
/
System2, pdos output file
Problem Sizes
natomwfc = 2038
nbnd = 1250
nkstot = 1
npwx = 143602
nkb = 4082
APPLICATION TERMINATED WITH THE EXIT STRING: Killed (signal 9)
Thank you very much,
-----------------------------------------------
Run Li
Ph.D. candicate for Chemistry
Dr. Mark R. Hoffmann research group
University of North Dakota
Cell: (612)-865-9844
Email: run.li at my.und.edu
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20141003/aad1aa34/attachment.html>
More information about the users
mailing list