Greetings QE developers and users,
I know this question is vague, but does anyone expect reasonable U values
for iron? following the linear response method of Cococcioni and Gironcoli,
I got U values for Fe in Na3Fe(II)PO4CO3 and Na2Fe(III)PO4CO3 (single Fe
site in the asymmetric unit) to be 6.39 and 7.82 eV, which seems much
higher than other Fe containing compounds in phosphates (vaguely speaking).
This leads to positive deviations from experimental observations (e.g. XRD
cell volume and redox potential). can anyone please comment on possible
causes? the input file for unperturbed SCF calculations is included
(diago_thr_init = last value in zero alpha along with charge-density.dat
used for subsequent non zero perturbations).
Q.E. version: 5.2.1
&CONTROL
title = 'Na2FePO4CO3 U det' ,
calculation = 'scf' ,
outdir = '/etc./common/' ,
wfcdir = '/etc./common/' ,
pseudo_dir = '/etc./pseudo/' ,
prefix = 'Na2FePO4CO3U' ,
verbosity = 'high' ,
/
&SYSTEM
ibrav = 12,
A = 8.997 ,
B = 5.163 ,
C = 6.741 ,
cosAB = -0.0027925231 ,
nat = 10,
ntyp = 5,
ecutwfc = 30 ,
ecutrho = 270 ,
lda_plus_u = .true. ,
lda_plus_u_kind = 0 ,
Hubbard_U(1) = 1d-20,
Hubbard_alpha(1) = 1d-20,
space_group = 11 ,
uniqueb = .false. ,
/
&ELECTRONS
conv_thr = 1d-9 ,
startingpot = 'atomic' ,
startingwfc = 'atomic' ,
diagonalization = 'david'
/
ATOMIC_SPECIES
Fe 55.93300 Fe.pbe-sp-van_mit.UPF
P 30.97400 P.pbe-van_ak.UPF
C 12.01100 C.pbe-van_ak.UPF
O 15.99900 O.pbe-van_ak.UPF
Na 22.99000 Na.pbe-sp-van_ak.UPF
ATOMIC_POSITIONS crystal_sg
Fe 0.362400000 0.279900000 0.250000000
P 0.587500000 0.798900000 0.250000000
O 0.569400000 0.096500000 0.250000000
Na 0.740900000 0.251300000 0.002200000
C 0.060500000 0.234500000 0.250000000
O 0.123300000 0.461600000 0.250000000
O 0.143000000 0.031700000 0.250000000
O 0.919500000 0.215700000 0.250000000
O 0.321000000 0.281700000 0.565500000
O 0.431600000 0.674000000 0.250000000
K_POINTS automatic
3 6 5 1 1 1
--
*Hooman Yaghoobnejad*
*Department of Chemistry*
*Missouri University of Science and Technology*
*Rolla, MO 65401*
*USA*
_______________________________________________
Pw_forum mailing list
[email protected]
http://pwscf.org/mailman/listinfo/pw_forum
