[Pw_forum] cann't get the LO-TO spliting using dynmat.x

[Paolo Giannozzi]

On Saturday 27 August 2005 17:38, sun jason wrote:

> I can't get the LO mode when I use dynmat.x, the result is like this: 
> ============================================================
>     0.0000    0.0000    0.0001 1055.3982 1055.3982 1055.3982
> ============================================================
> my input file for dynmat.x is like that:
> =====================================================
>  phonons of cubic BN at 0.0000000,0.0000000,0.0000000
>  &input
>   amass(1)=10.81100,
>   amass(2)=14.00674,
>   asr=.true.,
>   fildyn='bn.dynG',
>   filout='bn.out',
>   filmol='bn.molden'
>  &end
> 0.0000000,0.0000000,0.0000000
> =====================================================

From the header of pwtools/dynmat.f90:

!  Input data (namelist "input")

!  q(3)      real    calculate LO modes (add nonanalytic terms) along
!                    the direction q (default: q=(0,0,0) )

so q(i) must be specified inside the namelist as 
  q(1)=..., q(2)=..., q(3)=...
If you do not specify anything, or specify 0, the non-analytical part
will not be added. For BN any direction will yield the same TO-LO
splitting.

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy