The ABINIT Users Mailing List ( CLOSED )

[edwin@gps.caltech.edu]

I'm not sure I understand the question correctly, but I think Abinit usually 
can process coordinates for a non-primitive cell and spit out parameters for 
the primitive cell. This is done by setting brvltt = -1, and optcell = 2. The 
optcell setting is optional I guess, but convenient because the full set of 
unit cell parameters (with rprim normalized) is automatically put in the log 
file at the beginning of Broyden step 0.

Below I've pasted an example input file that starts with the conventional 
hexagonal cell for corundum (Al2O3) and converts to the rhombohedral 
primitive cell. Some unit cells are easier than others -- space groups with 
lots of possible lattice vector arrangements and alternate origins can take a 
certain amount of trial and error.

Apologies if I'm not answering the question you meant to ask.

Edwin Schauble
schauble@ucla.edu

------------------------------------
# Crystalline Al2O3 corundum 

  kptopt   1            # Automatic generation of k points, taking
                        # into account the symmetry
  tolvrs   1.0d-9       # SCF stopping criterion
    iscf   5            # Self-consistent calculation, using algorithm 5
     ixc   11

#######################################################################
brvltt   -1
spgroup 167         
natrd     2
spgorig   1

#Definition of the unit cell
acell  8.989431 8.989431 24.54321 #Conventional unit cell dimensions
angdeg 90.0 90.0 120.0            #Conventional unit cell angles

#Definition of the atom types
ntypat 2               
znucl 13 08           

#Definition of the atoms
  natom 10             # There are ten atoms per primitive cell
  typat 1 2           # The first is of type 1 (Al), 2nd is (O).
 ionmov 2
optcell 2
  ntime 14
   xred       0.0 0.0 0.35218    #Al coords. in conv. unit cell
              0.30625 0.0 0.25   # O coords. in conv. unit cell

fband 0.0              

#Definition of the planewave basis set
ecut    41.0           # Maximal kinetic energy cut-off, in Hartree
ecutsm   1.0
dilatmx  1.02

#Definition of the k-point grid
kptrlatt  2  0  0      # In cartesian coordinates, this grid is simple cubic, 
          0  2  0      # actually corresponds to the so-called 8x8x8 
Monkhorst- 
          0  0  2      # It might as well be obtained through the use of
                       # ngkpt, nshiftk and shiftk . 
shiftk  0.5 0.5 0.5

#Definition of the SCF procedure
nstep 50               # Maximal number of SCF cycles
diemac 4.0 



--------------------------------
Abinit does not have that functionality. It will tell you that you do not 
have a primitive cell but you then have to work it out yourself.

Michel


Le 20/10/06 05:17, « Chol-Jun Yu » <yucj@ghi.rwth-aachen.de> a écrit :

    Dear ABINIT users,

    Does anybody know how to find the primitive unit cell of the single 
crystal?
    This means I need to find the primitive unit cell in order to reduce the
    demand of the calculation.

    Of course, we can perform the calculation using the conventional unit cell
    through the parameter, chkprim = 0 and judge the primitive unit cell of 
the
    simple crystal. For example, we can find one for Si - single crystal 
(diamond
    structure) as Tutorial t3x.in. However, it is a little difficult to find 
one
    of the complex crystal. Can I do it in ABINIT? And then, how can I do it 
in
    ABINIT?

    Best regards,

    Yu.



-- 
***************************************************************
  Michel Cote                          tel: +1 (514) 343-5628     
  Professeur agrégé                   fax: +1 (514) 343-2071               
  Département de physique
  Université de Montréal
  C.P. 6128, succ. Centre-ville
  Montréal (Québec) H3C 3J7          Michel.Cote@umontreal.ca
  Canada                   www.phys.umontreal.ca/~michel_cote
***************************************************************