Hi,
Sorry first for a trivial question! I wonder if there is a general rule to how to run a calculation using hybrid functionals with cutting off the long range coulomb interactions. I noticed there is a tag “fock_icutcoul” and as it is mentioned the “OPTION 5” (ERFC for short range interaction) is hard-coded for hybrids. In that case do I need to set the “rcut” value (e.g. HSE06)?
The main idea is calculating the formation energies for the defects using hybrid functional and I also need to optimize the cell and geometry. Here is a part of the input file:
ndtset 1
jdtset 3
paral_kgb 1
enunit 1
#iomode 3
#####################################################################################
A SCF calculation
prtden1 1 # Print the density, for use by dataset 2
tolvrs1 1.0d-18 # This value is way too large for most realistic studies of materials
##prtdos1 3
#########################
Non- SCF for wavefunctions
iscf2 -2
getwfk2 1
getden2 1
tolwfr2 1.0d-21
#########################################################
Definition of the Hartree-Fock calculation
##########################################################
paral_kgb3 0
ixc3 -428 # Calculation with HSE06 functional
getwfk3 2 # Start from previous wavefunctions to ease convergence
fock_icutcoul3 5
rcut3 9.090909
#Additional (and facultative) variables for Hartree-Fock
nnsclohf3 1 # always 1 for exact exchange
nkpthf3 128 # number of k-point in the full-BZ
nband3 64
nbandhf3 46 # number of occupied states
MG These variables are needed to run the test in parallel (bug or feature?)
np_spkpt3 32 # Number of processors for k-point parallelization
nphf3 32 # Number of processors for occupied states parallelization
# The calculation thus requires np_spkpt*nphf processors.
toldfe3 1.0d-6 # Will stop when, twice in a row, the difference
# between two consecutive evaluations of total energy
# differ by less than toldfe (in Hartree)
optcell3 2
ionmov3 2
#getcell3 0
#getxred3 2
ntime3 150
dilatmx3 1.10
ecutsm3 0.5
tolmxf3 1.00d-6
strprecon3 0.5
prtposcar3 1 # Print out VASP-style POSCAR and FORCES for PHON and frophon
Thanks,
Moe