The ABINIT Users Mailing List ( CLOSED )

[Xavier Gonze <gonze@pcpm.ucl.ac.be>]

Dear Marcel,

As indicated by Resul, Aldo and Mathieu, the limiting behaviour of the

phonon branches for q->0 DOES depend on the direction ! 

For polar materials (for which there is a LO-TO splitting), 

only in simple cubic structures like zinc-blende (e.g. GaAs),

you will find that the limiting value is the same irrespective of the direction.

For wurtzite or other tetragonal or trigonal (or less symmetrical structure),

the behaviour depends on the direction. The TO frequency will not be affected

by the direction, but the LO frequency will depend on the direction. 

As an example, see the quartz phonon band structure in

Interatomic force constants from first-principles : the case of a-quartz.

X. Gonze, J.-C. Charlier, D.C. Allan and M.P. Teter. Phys. Rev. B (Rapid Communication) 50, 13035-13038 (1994).

For the underlying theory, see the section VIII.  B , especially equation 62, of

Dynamical matrices, Born effective charges, dielectric permittivity tensors, and interatomic force constants from density-functional perturbation theory.

X. Gonze and C. Lee. Phys. Rev. B 55, 10355-10368  (1997). 

Best wishes,

Xavier

On 07 May 2007, at 18:55, Marcel Mohr wrote:

Dear Resul, Aldo and Mathieu

thank you for your replies. I checked the output. For dataset 3 (i.e. the el field ) calculation I obtain different energies at the different directions (in plane/perp plane).

  Phonon wavevector (reduced coordinates) :  0.00000  0.00000  0.00000

 Phonon frequencies in cm-1    :

-  4.111561E+00  4.113461E+00  1.151039E+01  3.257683E+01  3.257684E+01

-  1.241432E+02  1.744005E+02  1.780705E+02  1.780705E+02  1.810052E+02

-  1.810053E+02  1.930965E+02

  Phonon at Gamma, with non-analyticity in the

  direction (cartesian coordinates)  1.00000  0.00000  0.00000

 Phonon frequencies in cm-1    :

-  4.112028E+00  4.194273E+00  1.151039E+01  3.257683E+01  3.257684E+01

-  1.241432E+02  1.744005E+02  1.780705E+02  1.810052E+02  1.810053E+02

-  1.930965E+02  2.047706E+02

  Phonon at Gamma, with non-analyticity in the

  direction (cartesian coordinates)  0.00000  1.00000  0.00000

 Phonon frequencies in cm-1    :

-  4.112977E+00  4.193344E+00  1.151039E+01  3.257683E+01  3.257684E+01

-  1.241432E+02  1.744005E+02  1.780705E+02  1.810052E+02  1.810053E+02

-  1.930965E+02  2.047706E+02

  Phonon at Gamma, with non-analyticity in the

  direction (cartesian coordinates)  0.00000  0.00000  1.00000

 Phonon frequencies in cm-1    :

-  4.111561E+00  4.113461E+00  1.163689E+01  3.257683E+01  3.257684E+01

-  1.241432E+02  1.780705E+02  1.780705E+02  1.810052E+02  1.810053E+02

-  1.930965E+02  2.089019E+02

I guess they should all be equal, independent of direction?

Cheers Marcel

On Mon, 7 May 2007, Aldo Humberto Romero wrote:

I agree with Mathieau.. did you check if your frequencies

from the anaddb analysis in Gamma match the values obtained

from the electric field perturbation?.. I am sure they will not and

you will have to use the ones obtained from the electric field

perturbation calculation.

-aldo.

Dear Aldo, dear Mathieau

was the data for the Gamma point obtained from anaddb ? or the

frequencies were obtained in a different calculation with polarization.

Yes, from anadbb, and corrected (as in tutorial)

All frequencies were obtained in 1 calculation run.

Maybe I should use `rfelfd3 = 1` ?

I posted anadbb input and (some) abinit input below.

Did you use the anaddb keyword to symmetrize the ifc's?..

If you mean the parameter 'symdynmat' , I didn't specify it and used v.

5.2.3, so I guess it was 0. (at least that's what the online manual says).

Will it help to switch it on?

Cheers Marcel

Attachment:

==========

!!!Anaddb input file

!Input file for the ifc code. Analysis of the AlAs DDB

!Flags

  ifcflag   1     ! Interatomic force constant flag

!Wavevector grid number 1 (coarse grid, from DDB)

   brav    4      ! Bravais Lattice : 1-S.C., 2-F.C., 3-B.C., 4-Hex.)

   ngqpt   6  6  4   ! Monkhorst-Pack indices

   nqshft  1         ! number of q-points in repeated basic q-cell

   q1shft  0.0 0.0 0.0

!Effective charges

      asr   1     ! Acoustic Sum Rule. 1 => imposed asymetrically

   chneut   1     ! Charge neutrality requirement for effective charges.

!Interatomic force constant info

   dipdip  1      ! Dipole-dipole interaction treatment

   ifcana  1      ! Analysis of the IFCs

   ifcout 50      ! Number of IFC's written in the output, per atom

   natifc  4      ! Number of atoms in the cell for which ifc's are

analysed

    atifc  1 2 3 4     ! List of atoms

###abinit input file

    ndtset   24

#Set 1 : ground state self-consistency

   getwfk1   0            # Cancel default

   kptopt1   1            # Automatic generation of k points, taking

                          # into account the symmetry

     nqpt1   0            # Cancel default

   tolvrs1   1.0d-18      # SCF stopping criterion (modify default)

   rfphon1   0            # Cancel default

#Q vectors for all datasets

#Complete set of symmetry-inequivalent qpt chosen to be commensurate

# with kpt mesh so that only one set of GS wave functions is needed.

#Generated automatically by running GS calculation with kptopt=1,

# nshift=0, shiftk=0 0 0 (to include gamma) and taking output kpt set

# file as qpt set. Set nstep=1 so only one iteration runs.

      nqpt   1            # One qpt for each dataset (only 0 or 1 allowed)

                          # This is the default for all datasets and must

                          #  be explicitly turned off for dataset 1.

snipped 24 q-points<

#Set 2 : Response function calculation of d/dk wave function

     iscf2   -3         # Need this non-self-consistent option for d/dk

   kptopt2   2          # Modify default to use time-reversal symmetry

   rfphon2   0          # Cancel default

   rfelfd2   2          # Calculate d/dk wave function only

   tolvrs2   0.0        # Cancel default for d/dk

   tolwfr2   1.0d-22    # Use wave function residual criterion instead

#Set 3 : Response function calculation of Q=0 phonons and electric field

pert.

   getddk3   2          # d/dk wave functions from last dataset

   kptopt3   2          # Modify default to use time-reversal symmetry

   rfelfd3   3          # Electric-field perturbation response only

#Sets 4-10 : Finite-wave-vector phonon calculations (defaults for all

datasets)

    getwfk   1          # Use GS wave functions from dataset1

    kptopt   3          # Need full k-point set for finite-Q response

    rfphon   1          # Do phonon response

   rfatpol   1 4        # Treat displacements of all atoms

     rfdir   1 1 1      # Do all directions (symmetry will be used)

    tolvrs   1.0d-8     # This default is active for sets 3-10

snipped the rest<

-- 

Dr. Aldo Humberto Romero

FULL PROFESSOR CINVESTAV - Unidad Queretaro

Libramiento Norponiente 2000

Real de Juriquilla, CP 76230, Qro, Queretaro

Mexico

Phone: + 52 442 441 4909

Fax: + 52 442 441 4938

Email: aromero@qro.cinvestav.mx

WWW:  qro.cinvestav.mx/~aromero