[Pw_forum] error on phonon calculation (root not converged)

[Liping YU]

Nicola Marzari wrote:

>>I have a feeling (could be wrong) that this might be due to LDA-PP of
>>Strontium.
>>Since i did the same calculation for PbTiO3 using the same PPs of Ti and
>>O as
>>those in SrTiO3 and it worked well. So I appreciate if someone could
>>send me
>>your Sr LDA-PP (if you have one) that i could try.
>>    
>>
>
>
>Hi Liping,
>
>it's indeed baffling. Do check that your coordinates are scaled
>correctly - i.e. that the atoms in cartesian coordinates are
>were they should be, that the axes are what they should be.
>  
>
Yes. see below:
1x1x1:
bravais-lattice index = 1
lattice parameter (a_0) = 7.2766 a.u.
unit-cell volume = 385.2939 (a.u.)^3
number of atoms/cell = 5
number of atomic types = 3
kinetic-energy cutoff = 30.0000 Ry
charge density cutoff = 270.0000 Ry
convergence threshold = 1.0E-08
beta = 0.7000
number of iterations used = 8 plain mixing
Exchange-correlation = SLA PZ NOGX NOGC (1100)
celldm(1)= 7.276637 celldm(2)= 0.000000 celldm(3)= 0.000000
celldm(4)= 0.000000 celldm(5)= 0.000000 celldm(6)= 0.000000

crystal axes: (cart. coord. in units of a_0)
a(1) = ( 1.000000 0.000000 0.000000 )
a(2) = ( 0.000000 1.000000 0.000000 )
a(3) = ( 0.000000 0.000000 1.000000 )
......

Cartesian axes

site n. atom positions (a_0 units)
1 Sr tau( 1) = ( 0.0000000 0.0000000 0.0000000 )
2 Ti tau( 2) = ( 0.5000000 0.5000000 0.5000000 )
3 O tau( 3) = ( 0.0000000 0.5000000 0.5000000 )
4 O tau( 4) = ( 0.5000000 0.0000000 0.5000000 )
5 O tau( 5) = ( 0.5000000 0.5000000 0.0000000 )

2x2x1
bravais-lattice index = 6
lattice parameter (a_0) = 14.5533 a.u.
unit-cell volume = 1541.1754 (a.u.)^3
number of atoms/cell = 20
number of atomic types = 3
kinetic-energy cutoff = 30.0000 Ry
charge density cutoff = 270.0000 Ry
convergence threshold = 1.0E-08
beta = 0.7000
number of iterations used = 8 plain mixing
Exchange-correlation = SLA PZ NOGX NOGC (1100)
celldm(1)= 14.553273 celldm(2)= 0.000000 celldm(3)= 0.500000
celldm(4)= 0.000000 celldm(5)= 0.000000 celldm(6)= 0.000000

crystal axes: (cart. coord. in units of a_0)
a(1) = ( 1.000000 0.000000 0.000000 )
a(2) = ( 0.000000 1.000000 0.000000 )
a(3) = ( 0.000000 0.000000 0.500000 )
...,,,

Cartesian axes

site n. atom positions (a_0 units)
1 Sr tau( 1) = ( 0.0000000 0.0000000 0.0000000 )
2 Sr tau( 2) = ( 0.5000000 0.0000000 0.0000000 )
3 Sr tau( 3) = ( 0.0000000 0.5000000 0.0000000 )
4 Sr tau( 4) = ( 0.5000000 0.5000000 0.0000000 )
5 O tau( 5) = ( 0.2500000 0.2500000 0.0000000 )
6 O tau( 6) = ( 0.2500000 0.7500000 0.0000000 )
7 O tau( 7) = ( 0.7500000 0.2500000 0.0000000 )
8 O tau( 8) = ( 0.7500000 0.7500000 0.0000000 )
9 Ti tau( 9) = ( 0.2500000 0.2500000 0.2500000 )
10 Ti tau( 10) = ( 0.2500000 0.7500000 0.2500000 )
11 Ti tau( 11) = ( 0.7500000 0.2500000 0.2500000 )
12 Ti tau( 12) = ( 0.7500000 0.7500000 0.2500000 )
13 O tau( 13) = ( 0.0000000 0.2500000 0.2500000 )
14 O tau( 14) = ( 0.0000000 0.7500000 0.2500000 )
15 O tau( 15) = ( 0.5000000 0.2500000 0.2500000 )
16 O tau( 16) = ( 0.5000000 0.7500000 0.2500000 )
17 O tau( 17) = ( 0.2500000 0.0000000 0.2500000 )
18 O tau( 18) = ( 0.7500000 0.0000000 0.2500000 )
19 O tau( 19) = ( 0.2500000 0.5000000 0.2500000 )
20 O tau( 20) = ( 0.7500000 0.5000000 0.2500000 )

>Next, if you use the same identical input files, but you put
>Pb instead of Sr, do you get identical dielectric constants
>from 1x1x1 and 2x2x1 ?
>  
>
No, for PbTiO3, it's tetragonal structure, not cubic one as SrTiO3. The
followings are
dielectric constants of PTO in tetragonal 1x1x1 unit cell and 2x2x1
supercell.
Note for no reason I used 4x4x4 k-point grid for both unit cell and
super cells.

1x1x1 unit cell,
( 7.981121520 0.000000000 0.000000000 )
( 0.000000000 7.981121520 0.000000000 )
( 0.000000000 0.000000000 7.542263658 )

2x2x1 supercell,
( 7.775208874 0.000000000 0.000000000 )
( 0.000000000 7.775208874 0.000000000 )
( 0.000000000 0.000000000 7.491522972 )

Maybe I should also try same k-point (4x4x4) sampling for SrTiO3 in two
different cells, although they are not equivalent, and see what will happen.

Also I'll try what you suggested. Thank you very much!

>If you do, I admit that one would be tempted to blame the
>pseudopotential, but it is still unlikely. Have a look at the
>eigenvalues in the two calculations - do they match exactly ?
>
>Let us know what you find,
>
>			nicola
>
>---------------------------------------------------------------------
>Prof Nicola Marzari   Department of Materials Science and Engineering
>13-5066   MIT   77 Massachusetts Avenue   Cambridge MA 02139-4307 USA
>tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu
>_______________________________________________
>Pw_forum mailing list
>Pw_forum at pwscf.org
>http://www.democritos.it/mailman/listinfo/pw_forum
>  
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20060410/27ba4c3d/attachment.htm