The ABINIT Users Mailing List ( CLOSED )

["Anglade Pierre-Matthieu" <anglade@gmail.com>]

Hi,

Did you try to make a converged calculation ? What is the "absolute"
precision of your simulation ?
For instance usualy light d metals require large cutoff energy.

regards

PMA

On Thu, Mar 20, 2008 at 8:14 PM,  <nathaniel.deskins@pnl.gov> wrote:
> Hello all,
>
>   I'm looking at the following gas-phase reaction. TiCl --> Ti + Cl.
>  The ground state for TiCl is a quintet (3 unpaired e-), Ti is a pentet(?) 
> (4 unpaired e-), while Cl is a doublet (1 unpaired e-). I've modeled this 
> reaction with other planewave and gaussian codes. The energy for the 
> reaction varies between 4.1-5.1 eV (depending on basis set and potential) 
> while the Ti-Cl distance is near 2.22-2.23 angstroms. All this was with PBE 
> XC functional.
>
>   When I've tried Abinit with the FHI GGA-PBE Troullier-Martins 
> pseudopotentials I get strange results. My calculated energy is 6.58 eV 
> while the Ti-Cl distance is 2.40 angstroms. This was for a cutoff of 15 
> hartree. For cutoff of 30 hartree the energy is 3.8 eV(?!) and the distance 
> is 2.39 angstroms.
>
>   Either there is an issue with the pseudopotentials, or I'm doing 
> something wrong.
>
>  Below are my input files. Any help is appreciated.
>
>  Thank you,
>  Aaron
>
>
>  ti.in
>  acell 25 25 25    # The keyword "acell" refers to the
>  ntypat 1          # There is only one type of atom
>  znucl 22           # The keyword "znucl" refers to the atomic number of the
>  natom 1           # There are two atoms
>  typat 1          # They both are of type 1, that is, Hydrogen
>  xcart             # This keyword indicates that the location of the atoms
>    0.0 0.0 0.0    # Triplet giving the cartesian coordinates of atom 1, in 
> Bohr
>  ecut 15.0         # Maximal plane-wave kinetic energy cut-off, in Hartree
>  nkpt 1            # Only one k point is needed for isolated system,
>  nsppol 2
>  occopt 2
>  nband 8 8
>  occ 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0
>  spinat 0.0 0.0 4.0  # Initialisation of spin
>  nstep 50          # Maximal number of SCF cycles
>  toldfe 1.0d-6     # Will stop when, twice in a row, the difference
>  diemac 2.0        # Although this is not mandatory, it is worth to
>
>
>  cl.in
>  acell 25 25 25    # The keyword "acell" refers to the
>  ntypat 1          # There is only one type of atom
>  znucl 17           # The keyword "znucl" refers to the atomic number of the
>  natom 1           # There are two atoms
>  typat 1          # They both are of type 1, that is, Hydrogen
>  xcart             # This keyword indicates that the location of the atoms
>    0.0 0.0 0.0    # Triplet giving the cartesian coordinates of atom 1, in 
> Bohr
>  ecut 15.0         # Maximal plane-wave kinetic energy cut-off, in Hartree
>  nkpt 1            # Only one k point is needed for isolated system,
>  nsppol 2
>  occopt 2
>  nband 8 8
>  occ 1 1 1 0 0 0 0 0 1 1 1 1 0 0 0 0
>  spinat 0.0 0.0 1.0  # Initialisation of spin
>  nstep 50          # Maximal number of SCF cycles
>  toldfe 1.0d-6     # Will stop when, twice in a row, the difference
>  diemac 2.0        # Although this is not mandatory, it is worth to
>
>
>  ticl.in
>  acell 25 25 25    # The keyword "acell" refers to the
>  ntypat 2          # There is only one type of atom
>  znucl 22 17          # The keyword "znucl" refers to the atomic number of 
> the
>  natom 2           # There are two atoms
>  typat 1 2         # They both are of type 1, that is, Hydrogen
>  xcart             # This keyword indicates that the location of the atoms
>    0.0 0.0 0.0    # Triplet giving the cartesian coordinates of atom 1, in 
> Bohr
>    4.3 0 0
>  ecut 30.0         # Maximal plane-wave kinetic energy cut-off, in Hartree
>  nkpt 1            # Only one k point is needed for isolated system,
>  nsppol 2
>  occopt 2
>  nband 12 12
>  occ 1 1 1 1 1 1 1 0 0 0 0 0 1 1 1 1 0 0 0 0 0 0 0 0
>  spinat 0.0 0.0 3.0
>        0.0 0.0 1.0
>  ionmov 2
>  tolmxf 5.0d-4
>  ntime 20
>  nstep 50          # Maximal number of SCF cycles
>  toldfe 1.0d-6     # Will stop when, twice in a row, the difference
>  diemac 2.0        # Although this is not mandatory, it is worth to
>



-- 
Pierre-Matthieu Anglade