[Pw_forum] ghost atom

MISSAOUI Jamil missaouijamil at yahoo.co.uk
Wed Feb 26 12:01:30 CET 2014


Dear Guido;
Firstly, ihave to say thank you very much for your helps.
But it still somewhat mysterious for me.
I want to know, in defining the pseudopotential of the ghost atom, is it better to generate a new one or just change something in any UPF file to get the desired pseudopotential. In the second case how we can edit it to cancel the value of charge and potential? (i.e. which files of this ones we will edit)


	* PP_INFO
	* PP_HEADER
	* PP_MESH
	* PP_NLCC

	* PP_LOCAL
	* PP_NONLOCAL
	* PP_PSWFC (optional)
	* PP_RHOATOM (optional).
Regards.
Jamil
__________________________________

Jamil MISSAOUI,
PhD. Student,


Faculty of Sciences of Tunis,
University of Manar,

Tunis, Tunisie.


________________________________
 From: Guido Fratesi <fratesi at mater.unimib.it>
To: MISSAOUI Jamil <missaouijamil at yahoo.co.uk>; PWSCF Forum <pw_forum at pwscf.org> 
Sent: Tuesday, 25 February 2014, 12:34
Subject: Re: [Pw_forum] LDOS
 

On 02/24/2014 03:21 PM, MISSAOUI Jamil wrote:

> I want to calculate the local density of states (LDOS) projected on a
> vacancy which is choosen to be located at the center of my supercell. My
> question

Maybe adding to the system at the vacancy site
 a ghost atom (I'm pretty 
sure you can edit the UPF file and set to zero the charge and the 
potential) could work better: then you just do the usual PDOS calculation.

> is how we can define the volume surrounding the vacancy (example a
> sphere of raduis = xA°) containing the points of the FFT grid included
> with irmin and irmax?

A sphere? The LDOS is only implemented for "boxes" with edges parallel 
to crystal axes. You can easily modify the function thetabox(ir)=1 
inside the volume of integration and 0 outside, see PP/src/projwfc.f90

Quick and dirt: approximate your sphere by a cube.


Guido

-- 
Guido Fratesi
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20140226/cb5b097c/attachment.html>


More information about the users mailing list