[Pw_forum] problem with projwfc.x

[Lorenzo Paulatto]

In data 14 settembre 2009 alle ore 14:11:34, Neel Singh  
<neelphysics at yahoo.in> ha scritto:

Dear Neel,


> in which the  valance states are
> 3S  3  0  2.00
> 3P  3  1  6.00
> 3D  3  2  2.00
> 4S  4  0  2.00
> 4P  4  1  0.00

>  i have a question that how we are getting 1s and 2p states as we have  
> used the pseudo potentials.

You cannot, but you don't need to either. The pseudopotential  
approximation uses the froze-core approximation: the occupation of core  
states is exactly the same as in the isolated atoms.

> And the second one that Ti has 22 electrons,

It has only 12 electrons in valence, and only if you consider 3s and 3p to  
be valence. Pseudopotential calculations only use valence electrons.

> how we can get Ti4d as the
> configuration is 4s2 3d2.

You cannot, because there is no 4d wavefunction in the pseudopotential  
file the code has no way to project the crystal wavefunctions on it. You  
can add a projector for the 4d wavefunction while generating the pseudo,  
but you would get a much slower pseudopotential. You could also modify the  
code to read additional projectors from somewhere and use them in projwfc,  
but you would have to do it yourself. The GIPAW may do something like this  
internally, but I don't know the details.

But are you sure you do want it? The 4d electron in Titanium, in the DFT  
framework, has an energy higher than 5s 5p 6s and 6p; actually it is not  
even a bound state!

best regards

-- 
Lorenzo Paulatto
SISSA  &  DEMOCRITOS (Trieste)
phone: +39 040 3787 511
skype: paulatz
www:   http://people.sissa.it/~paulatto/

     *** save italian brains ***
  http://saveitalianbrains.wordpress.com/