The ABINIT Users Mailing List ( CLOSED )

[David Tompsett <dat36@cam.ac.uk>]

Dear Matthieu, thanks for the advice.

I am revamping my input files to address some of the issues that you 
raised. I have a few questions as to how to do this and can answer some 
of your questions too:

matthieu verstraete wrote:
> Hello David,
>
> interesting problem, and it will be good to see a comparison with CASTEP.
>
> 1) is your system metallic? And do you have enough unoccupied bands.
> It sounds like the rest of your things are well converged, so I'll
> presume so.
>   
The system is metallic. Single band system at Fermi level. I had about 8 
bands above the Fermi level in the initial ground state SCF calculation.

> 5) If the system is really metallic, you probably need a lot more
> k-points. This could be the source of error.
> Are you sure the 8x8x8 + 4 shifts is equivalent to the kptrlatt grid
> you use for the response functions? I don't think so (it should be -8
> 8 8   8 -8 8    8 8 -8), and this could explain why the RF is
> (spectacularly) bad.
>   
I am trying to address this in my new calculation by using a 16x16x16 
grid with shiftk 0.0 0.0 0.5 (since the unit cell is hexagonal) for both 
the ground state and rf calc.
One question I have is: If my q-vector for the rf calc of the density 
matrix is Gamma then should the Gamma pt. (0.0 0.0 0.0) be a part of the 
k-pt grid? I have not been able to figure this out from the tutorials.

A further question is on tractability. Ultimately one of my goals is to 
calculate the electron-phonon coupling of the material. With 13 atoms in 
the unit cell there are many displacements. ABINIT does not have 
ultrasoft pseudopotentials and so calculations can be long. I have 
access to a cluster where I may use up to 64 cores for a max of 12 
hours. Should this sort of problem be tractable in ABINIT?

I will post the results with the comparison to CASTEP once I am able to 
get some sane numbers from my dynamical matrix at Gamma.

Thanks again for you help,
David.

> Matthieu
>
> On Mon, May 4, 2009 at 2:08 PM,  <dat36@cam.ac.uk> wrote:
>   
> > Dear All,
> >            I am having trouble finding correct phonon frequencies at Gamma for
> > the d-metal system Ag5Pb2O6. I have calculated them before in the CASTEP
> > plane-wave code and wanted to repeat the calucation in ABINIT as a sanity 
> > check
> > before using ABINIT to consider more properties.
> >
> > With the following input file I am able to calculate an accurate band
> > structure, and I then use the WFK file from the SCF calculation as the input 
> > to
> > my response function calculation.
> > # Ag5Pb2O6 BS
> > #
> > # Computation of the band structure.
> > # First, a SCF density computation, then a non-SCF band structure calculation.
> >
> > ndtset 2
> >
> > #Dataset 1 : usual self-consistent calculation
> > kptopt1 1          # Option for the automatic generation of k points,
> >                   # taking into account the symmetry
> > nshiftk1 4
> > shiftk1  0.5 0.5 0.5  # These shifts will be the same for all grids
> >         0.5 0.0 0.0
> >         0.0 0.5 0.0
> >         0.0 0.0 0.5
> > ngkpt1  8 8 8
> > prtden1  1         # Print the density, for use by dataset 2
> > toldfe1  1.0d-8
> >
> > #Dataset 2 : the band structure
> > iscf2    -2
> > getden2  -1
> > kptopt2  -4
> > nband2   100
> > ndivk2   28 19 29 45     # 10, 12 and 17 divisions of the 3 segments, 
> > delimited
> >                       # by 4 points.
> > kptbounds2 0.0 0.0 0.0
> > 0.5 0.0 0.0
> > 0.333333 0.333333 0.0
> > 0.0 0.0 0.0
> > 0.0 0.0 0.5
> >
> > tolwfr2  1.0d-16
> > #enunit2  1             # Will output the eigenenergies in eV
> >
> >
> > #Definition of the unit cell
> > acell 3*1.8897        # This is equivalent to   1 Angstrom each
> > rprim
> >        5.137609105410803      -2.966200000000002       0.000000000000000
> >        0.000000000000000       5.932400000000000       0.000000000000000
> >        0.000000000000000       0.000000000000000       6.410500000000000
> >
> >
> > #Definition of the atom types
> > ntypat 3
> > znucl 8 47 82
> >
> >
> > #Definition of the atoms
> > natom 13           # There are two atoms
> > typat 1 1 1 1 1 1 2 2 2 2 2 3 3
> > xred              # This keyword indicate that the location of the atoms
> >                  # will follow, one triplet of number for each atom
> >    0.6222000122070313   0.0000000000000001   0.6888999938964855
> >   -0.0000000000000001   0.6222000122070313   0.6888999938964855
> >   -0.6222000122070311  -0.6222000122070314   0.6888999938964854
> >   -0.0000000000000001  -0.6222000122070314  -0.6888999938964855
> >   -0.6222000122070311   0.0000000000000001  -0.6888999938964855
> >    0.6222000122070313   0.6222000122070313  -0.6888999938964854
> >    0.0000000000000000   0.0000000000000000   0.2413000017404568
> >    0.0000000000000000   0.0000000000000000  -0.2413000017404568
> >    0.5000000000000000  -0.0000000000000000   0.0000000000000000
> >    0.0000000000000000   0.5000000000000001   0.0000000000000000
> >   -0.5000000000000000  -0.5000000000000001  -0.0000000000000000
> >     
>
>   
> >    0.6666666666666667   0.3333333333333334   0.5000000000000001
> >   -0.6666666666666667  -0.3333333333333334  -0.5000000000000001
> > #Definition of the planewave basis set
> > ecut 40.0         # Maximal kinetic energy cut-off, in Hartree
> >
> > #Definition of the SCF procedure
> > nstep 100          # Maximal number of SCF cycles
> >       # 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.
> >
> > occopt 7
> > tsmear 0.007
> >
> >
> >
> > The input for the response function calculation for the density matrix at 
> > Gamma
> > is:
> > # Crystalline Ag5Pb2O6: computation of the dynamical matrix at Gamma
> > #
> >
> > #Response-function calculation, with q=0
> >  rfphon   1            # Will consider phonon-type perturbation
> >  rfatpol   1 13          # All the atoms will be displaced
> >   rfdir   1 1 1        # Along all reduced coordinate axis
> >    nqpt   1            # One wavevector is to be considered
> >     qpt   0 0 0        # This wavevector is q=0 (Gamma)
> >  kptopt   2            # Automatic generation of k points, taking
> >                        # into account the time-reversal symmetry only
> >  tolvrs   1.0d-8       # SCF stopping criterion
> >    iscf   5            # Self-consistent calculation, using algorithm 5
> >  irdwfk   1            # Read the ground-state wavefunctions
> >
> >
> > #######################################################################
> > #Common input variables
> >
> > #Definition of the unit cell
> > acell 3*1.8897        # This is equivalent to   1 Angstrom each
> > rprim
> >        5.137609105410803      -2.966200000000002       0.000000000000000
> >        0.000000000000000       5.932400000000000       0.000000000000000
> >        0.000000000000000       0.000000000000000       6.410500000000000
> >
> >
> > #Definition of the atom types
> > ntypat 3          # There is only one type of atom
> > znucl 8 47 82          # The keyword "znucl" refers to the atomic number of 
> > the
> >
> >                  # possible type(s) of atom. The pseudopotential(s)
> >                  # mentioned in the "files" file must correspond
> >                  # to the type(s) of atom. Here, the only type is Silicon.
> >
> >
> > #Definition of the atoms
> > natom 13           # There are two atoms
> > typat 1 1 1 1 1 1 2 2 2 2 2 3 3
> > xred              # This keyword indicate that the location of the atoms
> >                  # will follow, one triplet of number for each atom
> >    0.6222000122070313   0.0000000000000001   0.6888999938964855
> >   -0.0000000000000001   0.6222000122070313   0.6888999938964855
> >   -0.6222000122070311  -0.6222000122070314   0.6888999938964854
> >   -0.0000000000000001  -0.6222000122070314  -0.6888999938964855
> >   -0.6222000122070311   0.0000000000000001  -0.6888999938964855
> >    0.6222000122070313   0.6222000122070313  -0.6888999938964854
> >    0.0000000000000000   0.0000000000000000   0.2413000017404568
> >    0.0000000000000000   0.0000000000000000  -0.2413000017404568
> >    0.5000000000000000  -0.0000000000000000   0.0000000000000000
> >    0.0000000000000000   0.5000000000000001   0.0000000000000000
> >   -0.5000000000000000  -0.5000000000000001  -0.0000000000000000
> >    0.6666666666666667   0.3333333333333334   0.5000000000000001
> >   -0.6666666666666667  -0.3333333333333334  -0.5000000000000001
> > #Definition of the planewave basis set
> > ecut 40.0         # Maximal kinetic energy cut-off, in Hartree
> > diemac 1000000
> >
> > #Definition of the SCF procedure
> > nstep 100          # Maximal number of SCF cycles
> >       # 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.
> > occopt 7
> > tsmear 0.007
> >
> > #Definition of the k-point grid
> > kptrlatt -4  4  4      # In cartesian coordinates, this grid is simple cubic,
> > and
> >          4 -4  4      # actually corresponds to the so-called 8x8x8
> > Monkhorst-Pack grid.
> >          4  4 -4      # It might as well be obtained through the use of
> >                       # ngkpt, nshiftk and shiftk .
> >
> > #Gives the number of band, explicitely (do not take the default)
> > nband  70               # For an insulator (if described correctly as an
> > insulator
> >                       # by DFT), there is no need to include conduction bands
> >                       # in response-function calculations
> >
> > #Exchange-correlation functional
> > ixc 1             # LDA Teter Pade parametrization
> >
> > I obtain the following phonon frequencies:
> >  Phonon frequencies in cm-1    :
> > -  8.173847E+01  8.385767E+01  9.066616E+01  9.157959E+01  9.161629E+01
> > -  1.007493E+02  1.056726E+02  1.056770E+02  1.167382E+02  1.169064E+02
> > -  1.215102E+02  1.215114E+02  1.273660E+02  1.273693E+02  1.359495E+02
> > -  1.436825E+02  1.559343E+02  1.559382E+02  2.079336E+02  2.079848E+02
> > -  2.190082E+02  2.002110E+03  2.002542E+03  2.003094E+03  2.003199E+03
> > -  2.008633E+03  2.010303E+03  2.010572E+03  2.012517E+03  2.022057E+03
> > -  2.022089E+03  2.022721E+03  2.025986E+03  2.037439E+03  2.038421E+03
> > -  2.039042E+03  2.039143E+03  2.039784E+03  2.040022E+03
> >
> > However, I expected to see three acoustic phonons and none above 500cm^-1.
> >
> > There were two types of warning in the output file, for example:
> > -P-0000  hdr_check: WARNING -
> > -P-0000   input nkpt=     128 not equal disk file nkpt=      52
> >
> > -P-0000  cgwf3: WARNING -
> > -P-0000   New trial energy at line       4 =  -4.421412E+02
> > -P-0000      is higher than former: -4.421412E+02
> >
> > Any help is much appreciated.
> >
> > Thanks,
> > David Tompsett.
> >
> >     
>
>
>
>   

--
David A. Tompsett
Quantum Matter Group
Cavendish Laboratory
J. J. Thomson Avenue
Cambridge CB3 0HE
U.K.
Tel: +44 7907 566351 (mobile)
Fax: +44 1223 768140
http://www-qm.phy.cam.ac.uk/