[Pw_forum] optimizing cell parameters and choosing k-points of graphite

Masoud Nahali masoudnahali at live.com
Fri Jul 2 13:17:02 CEST 2010




One of the origins of this bad convergence is the selection of ecurho. For ultra-soft pseudo-potentials,the ecurho should be 10X of ecutwfc
and I had not considered it well. 

 
very best wishes
Masoud Nahali
SUT
 


Dear Quantum Espresso users
 
I have done a 'vc-relax' calculation to find the cell parameters of graphite.  The selected k-points are (4*4*4) and automatic.
 
obtained cell parameters:
  a         = 2.469,
  b         = 2.469,
  c         = 7.069,
 
then I used these values to 'relax' the atoms in cell (k-points were 4*4*4 too) . Finally I used the obtained cell parameters to find the proper k-points for graphite.
I performed some 'scf' calculations with different k-points and in whole of them the cell parameters and positions were what is obtained from the relax part with   4*4*4 k-points.
 
'scf' energy with different k-points:
 
k-points     energy	
 
(4*4*4): -45.5732791969 Ry 
 
(5*5*5): -45.57637408 Ry
 
(6*6*6): -45.57592043 Ry
 
(7*7*7): -45.57420434 Ry
 
(8*8*8): -45.57474553 Ry
 
(9*9*9): -45.57500776 Ry
 
(10*10*10): -45.57454272 Ry
 
(11*11*11): -45.57464463 Ry
 
(12*12*12): -45.57480118 Ry
 
(13*13*13): -45.57462800 Ry
 
(14*14*14): -45.57463705 Ry
 
(15*15*15): -45.57473135 Ry
 
(16*16*16): -45.57465384 Ry
I appreciate if one help me to find the answers to these questions:
1. Is the whole procedure plausible ?2. why a energy convergence is not occurred after selecting these high values of k-points?3. How can I find the proper k-points safely?
the input file for 'scf' calculation:
&CONTROL
  calculation  = "scf",
  pseudo_dir   = "/home/koa/soft/qe4.2/espresso-4.2/pseudo",
  outdir       = "/home/koa/tmp",
  etot_conv_thr= 1.0D-4,
  forc_conv_thr= 1.0D-3,
  nstep=150,
  /
&SYSTEM
  ibrav     = 4,
  a         = 2.469,
  b         = 2.469,
  c         = 7.069,
  cosab     = -0.5,
  cosac     = 1.0,
  cosbc     = 1.0,
  nat       = 4,
  ntyp      = 1,
  ecutwfc   = 30.D0,
  ecutrho   = 200.D0
  occupations = 'smearing'
  smearing ='mp', 
  degauss = 0.01,
  nspin = 2,
  starting_magnetization(1)= 1.0,
  /
&ELECTRONS
  conv_thr    = 1.D-6,
  mixing_beta = 0.7D0,
  diagonalization = "david",
/
&IONS
 ion_dynamics="bfgs"
/
ATOMIC_SPECIES
C  12.0107  C.pbe-van_ak.UPF
ATOMIC_POSITIONS {angstrom}
C        0.000000000   0.000000000   0.000000000
C        0.000000000   1.425534302   0.000000000
C        0.000000000   0.000000000   3.534553118
C        1.234548920   0.712767151   3.534553118
 
K_POINTS {automatic}
16 16 16 1 1 1
 
the vc-relax input: 
&CONTROL
  calculation  = "vc-relax",
  pseudo_dir   = "/home/koa/soft/qe4.2/espresso-4.2/pseudo",
  outdir       = "/home/koa/tmp",
  etot_conv_thr= 1.0D-4,
  forc_conv_thr= 1.0D-3,
  dt=70,
  nstep=150,
  /
&SYSTEM
  ibrav     = 4,
  a         = 2.456,
  b         = 2.456,
  c         = 6.696,
  cosab     = -0.5,
  cosac     = 1.0,
  cosbc     = 1.0,
  nat       = 4,
  ntyp      = 1,
  ecutwfc   = 30.D0,
  ecutrho   = 200.D0
  occupations = 'smearing'
  smearing ='mp', 
  degauss = 0.01,
  nspin = 2,
  starting_magnetization(1)= 1.0,
  /
&ELECTRONS
  conv_thr    = 1.D-6,
  mixing_beta = 0.7D0,
  diagonalization = "david",
/
&IONS
 ion_dynamics="bfgs"
/
&CELL
cell_dynamics = 'damp-w',
press = 0.0 ,
/
ATOMIC_SPECIES
C  12.0107  C.pbe-van_ak.UPF
ATOMIC_POSITIONS {angstrom}
 C                  0.00000000    0.00000000    0.00000000
 
 C                  0.00000000    1.41797226    0.00000000
 
 C                  0.00000000    0.00000000    3.34800000
 
 C                  1.22800000    0.70898613    3.34800000
 
K_POINTS {automatic}
4 4 4 1 1 1
 
very best wishes
 

  		 	   		  
_________________________________________________________________
Hotmail: Free, trusted and rich email service.
https://signup.live.com/signup.aspx?id=60969
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20100702/f8f826bb/attachment.html>


More information about the users mailing list