Dear Ehsan,

   Yes, that is allowed: Using the same exchange-correlation functional 
but different kinds of pseudo potentials. In this case it would even make 
quite well sense, as Cl and Br are "soft" elements and thus would probably 
not increase the requirement for 'ecutwfc'. Please notice that because the 
pp for carbon is of uspp/Vanderbilt type, you also have to insert a value 
for the variable 'ecutrho'.

   Another option would be to use the goodies of the PSlibrary, 
http://qe-forge.org/gf/project/pslibrary/, and generate the required pp's 
yourself.

     Greetings from Zurich,

        apsi

-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-
   Ari P Seitsonen / Ari.P.Seitsonen at iki.fi / http://www.iki.fi/~apsi/
   Institut f?r Chemie der Universit?t Z?rich, CH-8057 Z?rich
   Tel: +41 44 63 54 497  /  Mobile: +41 79 71 90 935


On Tue, 4 Feb 2014, ehsan targholi wrote:

> hi
> Dear all?
> i have one calculation of pw.x that it has tree type of ?atom (C , Cl , Br ) 
> . i want to use of pw91 type of?pseudopotentials. but there is no one type of 
> pw91?pseudopotentials
> for all of the them. if i use?Cl.pw91-mt.UPF for Cl and?Br.pw91-mt.UPF?for Br 
> and?C.pw91-van_ak.UPF?for C , is wrong? or no any problem of this?
> 
> 
> 
> best regard
> ehsan
> graduate student of iust
> 
>

Reply via email to