I am trying to add 1 atom (more in future..) to the surface of quartz (SiO2)
I use 36 atoms of SiO2 and 1 atom of C.
Thickness of quartz layer is 5.457 A and then layer of vacuum of the thickness 10.914 A
When there is no C atom SCF cycles converges very good.
But when I put this C atom, there is no convergence! Energy goes back and forth all the time in the range [1e-1, 10]
Any thoughts will be appreciated.
Code: Select all
acell 9.993 9.993 16.371 angstrom
angdeg 90 90 120 # corresponds to vectors in rprim (I calculated this)
#Definition of the atom types
ntypat 3
znucl 14 8 6
#Definition of the atoms
natom 37 # number of atoms
typat
1 2 1 2 2 2 2 1 2
1 2 1 2 2 2 2 1 2
1 2 1 2 2 2 2 1 2
1 2 1 2 2 2 2 1 2
3
xred
0.25 0.3962 0.333333333
0.1038 0.2924 0.27777778
0.25 0.1462 0.111111113
0.2076 0 0.055555557
0.1038 0.1038 0.166666667
0.3962 0.1886 0.166666667
0.2924 0.2924 0.055555557
0.5 0.1462 0.222222223
0.3962 0.5 0.27777778
0.75 0.3962 0.333333333
0.6038 0.2924 0.27777778
0.75 0.1462 0.111111113
0.7076 0 0.055555557
0.6038 0.1038 0.166666667
0.8962 0.1886 0.166666667
0.7924 0.2924 0.055555557
1 0.1462 0.222222223
0.8962 0.5 0.27777778
0.25 0.8962 0.333333333
0.1038 0.7924 0.27777778
0.25 0.6462 0.111111113
0.2076 0.5 0.055555557
0.1038 0.6038 0.166666667
0.3962 0.6886 0.166666667
0.2924 0.7924 0.055555557
0.5 0.6462 0.222222223
0.3962 1 0.27777778
0.75 0.8962 0.333333333
0.6038 0.7924 0.27777778
0.75 0.6462 0.111111113
0.7076 0.5 0.055555557
0.6038 0.6038 0.166666667
0.8962 0.6886 0.166666667
0.7924 0.7924 0.055555557
1 0.6462 0.222222223
0.8962 1 0.27777778
0.5 0.5 0.4
chksymbreak=0
chkprim 0
ecut 10.0
#Definition of the k-point grid
kptopt 1
ngkpt 2 2 2
nshiftk 1
shiftk 0 0 0
#Definition of the SCF procedure
nstep 30 # Maximal number of SCF cycles
toldfe 1.0d-3 # Will stop when, twice in a row, the difference
# between two consecutive evaluations of total energy
# differ by less than toldfe (in Hartree)
# This value is way too large for most realistic studies of material
diemac 2.0 # Although this is not mandatory, it is worth to
# precondition the SCF cycle. The model dielectric
# function used as the standard preconditioner
# is described in the "dielng" input variable section.
# Here, we follow the prescriptions for molecules
# in a big box