Graphite interlayer distance: Difference between revisions

From VASP Wiki
No edit summary
 
(20 intermediate revisions by 2 users not shown)
Line 1: Line 1:
Description: determine the interlayer distance of graphite in the stacking direction using the method
{{Template:Bulk_systems - Tutorial}}
of Tchatchenko and Scheffler to account for van der Waals interactions.
----


Semilocal DFT at the GGA level underestimates
== Task ==
long-range dispersion interactions.
This problem causes a bad overestimation of
graphite lattice in the stacking direction
(8.84 A (PBE) vs. 6.71 A (exp)).


In this example, dispersion correction method  
In this example you will determine the interlayer distance of graphite in the stacking direction using the method of Tchatchenko and Scheffler to account for van der Waals interactions.
of Tchatchenko and Scheffler (PRL 102, 073005
(2009)) is used to cope with this problem.


Optimal length of the lattice vector c normal to
Semilocal DFT at the GGA level underestimates long-range dispersion interactions.
the stacking direction is determined in a series
This problem causes a bad overestimation of graphite lattice in the stacking direction: 8.84 Å (PBE) vs. 6.71 Å (exp).
of single point calculations with varied value
of c (all other degrees of freedom are fixed at
their experimental values).  


The computed c vs. energy dependence
In this example, the [[Tkatchenko-Scheffler method|dispersion correction method of Tchatchenko and Scheffler]] is used to cope with this problem.
is written in the file results.dat and can be
visualized e.g. using xmgrace. The optimal
value can be obtained using the attached
utility (python with numpy or Numeric is needed):


<tt>./utilities/fit.py results.dat</tt>
== Input ==
 
=== POSCAR ===
 
graphite
1.0
1.22800000 -2.12695839  0.00000000
1.22800000  2.12695839  0.00000000
0.00000000  0.00000000  7.0
4
direct
    0.00000000  0.00000000  0.25000000
    0.00000000  0.00000000  0.75000000
    0.33333333  0.66666667  0.25000000
    0.66666667  0.33333333  0.75000000
 
=== INCAR ===
 
{{TAGBL|IVDW}} = 20         
{{TAGBL|LVDW_EWALD}} =.TRUE.
{{TAGBL|NSW}} = 1
{{TAGBL|IBRION}} = 2
{{TAGBL|ISIF}} = 4
{{TAGBL|PREC}} = Accurate
{{TAGBL|EDIFFG}} = 1e-5
{{TAGBL|LWAVE}} = .FALSE.
{{TAGBL|LCHARG}} = .FALSE.
{{TAGBL|ISMEAR}} = -5
{{TAGBL|SIGMA}} = 0.01
{{TAGBL|EDIFF}} = 1e-6
{{TAGBL|ALGO}} = Fast
{{TAGBL|NPAR}} = 2
 
=== KPOINTS ===
 
Monkhorst Pack
0
gamma
16 16 8
0 0 0
 
== Running this example ==
 
To run this example, execute the <code>run.sh</code> bash-script:
<pre>
#
# To run VASP this script calls $vasp_std
# (or posibly $vasp_gam and/or $vasp_ncl).
# These variables can be defined by sourcing vaspcmd
. vaspcmd 2> /dev/null
 
#
# When vaspcmd is not available and $vasp_std,
# $vasp_gam, and/or $vasp_ncl are not set as environment
# variables, you can specify them here
[ -z "`echo $vasp_std`" ] && vasp_std="mpirun -np 8 /path-to-your-vasp/vasp_std"
[ -z "`echo $vasp_gam`" ] && vasp_gam="mpirun -np 8 /path-to-your-vasp/vasp_gam"
[ -z "`echo $vasp_ncl`" ] && vasp_ncl="mpirun -np 8 /path-to-your-vasp/vasp_ncl"
 
#
# The real work starts here
#
 
rm results.dat
 
for d in 6.5 6.6 6.65 6.7 6.75 6.8 6.9 7.0
do
  cat>POSCAR<<!
graphite
1.0
1.22800000 -2.12695839  0.00000000
1.22800000  2.12695839  0.00000000
0.00000000  0.00000000  $d
4
direct
  0.00000000  0.00000000  0.25000000
  0.00000000  0.00000000  0.75000000
  0.33333333  0.66666667  0.25000000
  0.66666667  0.33333333  0.75000000
!
  $vasp_std
  cp OUTCAR OUTCAR.$d
  energy=$(grep "free  ene" OUTCAR.$d|awk '{print $5}')
  echo $d $energy >> results.dat
done
</pre>
 
The optimal length of the lattice vector ''c'' normal to the stacking direction is determined in a series of single point calculations with varied value of ''c'' (all other degrees of freedom are fixed at their experimental values).
 
The computed ''c'' vs. energy dependence is written in the file <code>results.dat</code> and can be visualized e.g. using ''xmgrace''.
The optimal value can be obtained using the attached utility (python with numpy or Numeric is needed):
 
./utilities/fit.py results.dat
 
This should yield:


  200 iterations performed
  200 iterations performed
Line 34: Line 113:
         d0(A):  6.65603352689
         d0(A):  6.65603352689


The computed value of 6.66 A agrees well with  
The computed value of 6.66 &Aring; agrees well with experiment (6.71 &Aring;).
experiment (6.71 A).


Details of implementation of TS in VASP + a number
== Download ==
of tests:
[[Media:GraphiteDistance ts.tgz| graphiteDistance_ts.tgz]]
Bucko et al., PRB 87, 064110 (2013).
 
== Used INCAR Tags ==
{{TAG|ALGO}}, {{TAG|EDIFF}}, {{TAG|EDIFFG}}, {{TAG|IBRION}}, {{TAG|ISIF}}, {{TAG|ISMEAR}}, {{TAG|IVDW}}, {{TAG|LCHARG}}, {{TAG|LVDW_EWALD}}, {{TAG|LWAVE}}, {{TAG|NPAR}}, {{TAG|NSW}}, {{TAG|PREC}}, {{TAG|SIGMA}}


== Download ==
{{Template:Bulk_systems}}
[http://www.vasp.at/vasp-workshop/examples/graphiteDistance_ts.tgz graphiteDistance_ts.tgz]
----
[[VASP_example_calculations|To the list of examples]] or to the [[The_VASP_Manual|main page]]


[[Category:Examples]]
[[Category:Examples]]

Latest revision as of 13:25, 14 November 2019

Task

In this example you will determine the interlayer distance of graphite in the stacking direction using the method of Tchatchenko and Scheffler to account for van der Waals interactions.

Semilocal DFT at the GGA level underestimates long-range dispersion interactions. This problem causes a bad overestimation of graphite lattice in the stacking direction: 8.84 Å (PBE) vs. 6.71 Å (exp).

In this example, the dispersion correction method of Tchatchenko and Scheffler is used to cope with this problem.

Input

POSCAR

graphite
1.0
1.22800000 -2.12695839  0.00000000
1.22800000  2.12695839  0.00000000
0.00000000  0.00000000  7.0
4
direct
   0.00000000  0.00000000  0.25000000
   0.00000000  0.00000000  0.75000000
   0.33333333  0.66666667  0.25000000
   0.66666667  0.33333333  0.75000000

INCAR

IVDW = 20           
LVDW_EWALD =.TRUE. 
NSW = 1 
IBRION = 2
ISIF = 4
PREC = Accurate
EDIFFG = 1e-5
LWAVE = .FALSE.
LCHARG = .FALSE.
ISMEAR = -5
SIGMA = 0.01
EDIFF = 1e-6
ALGO = Fast
NPAR = 2

KPOINTS

Monkhorst Pack
0
gamma
16 16 8
0 0 0

Running this example

To run this example, execute the run.sh bash-script:

#
# To run VASP this script calls $vasp_std
# (or posibly $vasp_gam and/or $vasp_ncl).
# These variables can be defined by sourcing vaspcmd
. vaspcmd 2> /dev/null

#
# When vaspcmd is not available and $vasp_std,
# $vasp_gam, and/or $vasp_ncl are not set as environment
# variables, you can specify them here
[ -z "`echo $vasp_std`" ] && vasp_std="mpirun -np 8 /path-to-your-vasp/vasp_std"
[ -z "`echo $vasp_gam`" ] && vasp_gam="mpirun -np 8 /path-to-your-vasp/vasp_gam"
[ -z "`echo $vasp_ncl`" ] && vasp_ncl="mpirun -np 8 /path-to-your-vasp/vasp_ncl"

#
# The real work starts here
#

rm results.dat

for d in 6.5 6.6 6.65 6.7 6.75 6.8 6.9 7.0
do
  cat>POSCAR<<!
graphite
1.0
1.22800000 -2.12695839  0.00000000
1.22800000  2.12695839  0.00000000
0.00000000  0.00000000  $d
4
direct
   0.00000000  0.00000000  0.25000000
   0.00000000  0.00000000  0.75000000
   0.33333333  0.66666667  0.25000000
   0.66666667  0.33333333  0.75000000
!
  $vasp_std
  cp OUTCAR OUTCAR.$d
  energy=$(grep "free  ene" OUTCAR.$d|awk '{print $5}')
  echo $d $energy >> results.dat 
done

The optimal length of the lattice vector c normal to the stacking direction is determined in a series of single point calculations with varied value of c (all other degrees of freedom are fixed at their experimental values).

The computed c vs. energy dependence is written in the file results.dat and can be visualized e.g. using xmgrace. The optimal value can be obtained using the attached utility (python with numpy or Numeric is needed):

./utilities/fit.py results.dat

This should yield:

200 iterations performed
Ch-square: 4.30305519481e-09
---------

       E0(eV):         -37.433456779
       d0(A):  6.65603352689

The computed value of 6.66 Å agrees well with experiment (6.71 Å).

Download

graphiteDistance_ts.tgz