The ABINIT Users Mailing List ( CLOSED )

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

Hello,

Just one or two comments:
 I may mistake but I feel that 30 Ha for norm conserving pseudo of Ti amd O is small. I wonder  if your pseudo are correct for many other properties than lattice parameter.  I would recommend you to check one or two other known properties. In you case, computing elastic constant would certainly be quite helpfull. If you find them again softer than  experiment you will have to blame your pseudo. Otherwise it means that your phonon computations have a flaw...

regards

PMA

On Nov 26, 2007 1:08 PM, lan haiping <lanhaiping@gmail.com> wrote:

Hi ,All !
 I performed a series test calculations on TiO2 rutile, and want to examine its lattice dynamical properties.
Now, i have some questions about the calculation results when compared with published data. Would you please give me
some hints or comments ?
 My calculation setting parameters for TiO2 rutile  are following :
      ecut 30 Ha,   MP grid for BZ sampling is 4x4x6, and experimental lattice parameters as input settings, my pseudopotentials
are generated by opium package with  LDA functional.
   After cell optimization ,   structural parameters  are  good compared with experimental data:   a is 4.562( expt 4.587) Ang , c is 2.933(expt 2.954) Ang ,and u is 0.3041.
 With  the optimized structure, i then performed response function calculation with high SCF criterions (tolvrs = 1.0E-18) for gamma point. The phonon frequencies are following ( in cm-1)    :
- -4.218215E-03 -3.926303E-03  0.000000E+00   (acoustic  modes)
  9.043052E+01  
        1.106313E+02 (TO)     7.534477E+02 (LO)
       1.160615E+02(TO)    3.465211E+02(LO)
      3.758941E+02 (TO)   4.323807E+02(LO)
      4.778716E+02 (TO)  7.860030E+02(LO)
1.323684E+02
3.847799E+02 
4.117380E+02
4.613717E+02
6.068729E+02
8.118167E+02
 These values are well soften than the published results , such as Professor Gonze' s PRB 50 13379.  
I then tried to increase the ecut value , but it seems no effect .    Would you please tell me how to improve these
values ?  i also performed a test on tio2 anatase with the almost same setting parameters, the calculated phonon frequencies seem
having the same soften trends when compared with published results.  Is this problem   related with pseudopotentials ?, or other ?

Bests,
Hai-Ping










--
Hai-Ping Lan
Department of Electronics ,
Peking University , Bejing, 100871
lanhaiping@gmail.com, hplan@pku.edu.cn

-- 
Pierre-Matthieu Anglade