The ABINIT Users Mailing List ( CLOSED )

[<annie_beya@yahoo.fr>]

Dear all,

I have been trying to compute the piezoelectric tensor of GaAs (following the
tutorial and the test case tK).
1. I am using abinip 5.6.5 and the fhi pseudopotentials 31-Ga.LDA.fhi and
33-As.LDA.fhi
I got the ddb files and analysed them with anaddb. but the value that i get 
for
the piezo constant is -0.23C/m2 that is 82 times higher than the expected
value.

I had some convergence problems for the ddk calculation then i used nline and
iprcel, following some suggestions from this forum. could somebody have a look
on my input file (see below) and tell me what could be wrong? the values that 
i
use for acell and ecut are the converged values.

2. Using the semi-core electrons as valence electrons (again the fhi psp with
3d electrons for Ga) gave me optimized values for the lattice parameters that
are very close to theoretical ones (I mean a precision of 0.1%). But I could
not perform any response function calculation to get the piezo constant, again
due to convergence problems and band filling. I tried the solutions provided 
to
other users, but they did not help. I thought maybe there are better psp to be
used or better ways. Do you have any suggestions?

3.I will use the berry phase formulation of the polarisation to get the piezo
constants. I am doing response function to compare results from both
calculations (berry phase approach and response function) and I would like to
be sure i am using the good parameters and understand how abinit works.

4. I noticed this warning that comes quite often and i followed the 
suggestions
on the forum to set the same kptopt to both ground state calculations and
response function calculations. but it did not help, and induced other
troubles. Is this warning critical? What can I do to solve the problem?

hdr_check: WARNING -
-P-0000   input nkpt=    6912 not equal disk file nkpt=     182
hdr_check: WARNING -
-P-0000   Restart of self-consistent calculation need translated 
wavefunctions.
-P-0000   Indeed, critical differences between current calculation and
-P-0000   restart file have been detected in:
-P-0000         * the number, position, or weight of k-points
-P-0000         * the format of wavefunctions (istwfk)

thanks for your reply,

Annie

#GaAs in zincblende structure

ndtset 3 
jdtset 1 2 3 
# Set 1 : initial self-consistency

kptopt1   1
tolvrs1 1.0d-18
iscf1 5

#set 2 the ddk 
rfelfd2 2  # only derivative of the gs wf
rfdir2 1 1 1
nqpt2 1
qpt2 0.0 0.0 0.0 #only the gamma point
getwfk2 1
getddk2 2
kptopt2 3
iscf2 -3
tolwfr2 1.0d-20

# Set 3 : response-function strain calculation

rfphon3 1
rfatpol3 1 2 
rfelfd3 3
rfdir3 1 1 1
nqpt3 1
qpt3 0 0 0
rfstrs3  3
getwfk3 1
getddk3 2
kptopt3 3
tolvrs3 1.0d-16
iscf3 5

#common input data
paral_rf 1 # activate parallelization over perturbations
ngroup_rf 2 # number of groups for  parallelization over perturbations

   ecut   50.0
   iscf   5
   diemac 12.0
   iprcel 49
   nline 15

   ixc    7
  nband   8
  occopt 1

  ngkpt   12 12 12
  nshiftk   1
  shiftk   0.0 0.0 0.0
  nstep   200

  acell    3*10.5 # this is the optimized value
  rprim  0.0  0.5  0.5
       0.5  0.0  0.5
       0.5  0.5  0.0
  natom   2
 ntypat   2 
  znucl  31 33
  typat   1 2
   xred   0.0 0.0 0.0
          1/4 1/4 1/4