The ABINIT Users Mailing List ( CLOSED )

[andika asyuda <venomdi2000@yahoo.com>]

Good morning all abinit user

Thank you for the suggestion. I am studying lesson 3 and 4 now. But  I still have few questions about the primitive cell of GaAs which only has 2 atoms. Does it make my previous code wrong or just ineffective? This is my code

----------------------------------------------------------------------
ndtset 18  udtset 6 3

acell 10 10 10
ecut:? 10    ecut+? 5

#First dataset
   natom?1  8            
   typat?1 1 1 1 1 2 2 2 2        
  ionmov?1  6          
   ntime?1  30          
  tolmxf?1  5.0d-4      
   xcart?1  -5.5297287043E+00 -5.5297287043E+00 -5.5297287043E+00
              4.1484620949E-01 -5.1046616355E+00  4.1484620949E-01
             -5.1046616355E+00  4.1484620949E-01  4.1484620949E-01
              4.1484620949E-01  4.1484620949E-01 -5.1046616355E+00
             -2.9178542661E+00 -2.9178542661E+00 -2.9178542661E+00
              5.5393733310E+00  5.5393733310E+00  3.3068032333E-01
              5.5393733310E+00  3.3068032333E-01  5.5393733310E+00
              3.3068032333E-01  5.5393733310E+00  5.5393733310E+00
  toldff?1  5.0d-5      
   nband?1   18         

#Second dataset :
   natom?2  1           
  typat?2 1       
  nsppol?2  2           
  occopt?2  2          
   nband?2  2 2         
     occ?2  1.0  1.0 1.0 0.0     
  toldfe?2  1.0d-6    
   xcart?2  0.0 0.0 0.0  
  spinat?2  0.0 0.0 1.0  

#third dataset :
   natom?3  1          
  typat?3 2        
  nsppol?3  2           
  occopt?3  2           
   nband?3  3  3          
     occ?3  1.0 1.0 1.0 1.0 1.0 0.0    
  toldfe?3  1.0d-6       
   xcart?3  0.0 0.0 0.0  
  spinat?3  0.0 0.0 1.0  


#Definition of the atom types
ntypat 2         
znucl 31 33          
 
#Definition of the k-point grid
nkpt 1           

#Definition of the SCF procedure
nstep 1000      
diemac 2.0     

---------------------------------------------------------

I have seen the picture of GaAs, and I will try computation with atom position
Zn: 0  0  0
As: 1  1  1

with initial acell 2.5 5 2.5

Thank you for the help

with regard

Andika Asyuda

----- Original Message ----
From: J. J. Ramsey <jjr19@uakron.edu>
To: forum@abinit.org
Sent: Friday, August 1, 2008 9:07:26 PM
Subject: Re: [abinit-forum] convergence study in abinit

----- Original Message ----
> From: "venomdi2000@yahoo.com" <venomdi2000@yahoo.com>
> To: forum@abinit.org
> Sent: Thursday, July 31, 2008 11:24:58 AM
> Subject: [abinit-forum] convergence study in abinit
>
> Good night abinit user
>
> I want to ask about study of convegence in paramete acell and ecut in abinit.
>
> From, lecture 2 abinit tutorial, it shows that the value of parameter ecut and
> acell can be found by convergence study of atomization energy and the length of
> bond. Tutorial uses Hydrogen as case study. I want to know how to implement
> this in another case. hydrogen only has 2 atoms, so it is easy when we try to
> see the convergence of length of atomic bond. My case is GaAs which has 8
> atoms, so observing length of atomic bond is difficult. Is there another way
> for this?

Look at tutorial 3.4, where the optimization of the lattice parameter for silicon is discussed. Also, you don't need 8 atoms. The *primitive* cell for GaAs has only 2 atoms. Obviously, such a unit cell is not shaped as a cube, but the unit cell in an ABINIT simulation doesn't have to be a cube.