ABINIT Discussion Forums

Hi, I am a new user of Abinit. I use variable spgroup, acell, angdeg, xred from .cif file to produce the structure of Y2O3, I use input file:
acell 3*20.048
spgroup 206
angdeg 3*90
ntypat 2
znucl 39 8
natom 80
typat 16*1 16*1 48*2
natrd 3
brvltt -1
xred 0.25 0.25 0.25
0.4675 0.0 0.25
0.1087 0.3478 0.1194
ecut 8.0
kptopt 1
ngkpt 4 4 4
nshiftk 2
shiftk 0.25 0.25 0.25
-0.25 -0.25 -0.25
nstep 10
toldfe 1.0d-6
diemac 18.0
and get error message:
abinit : ERROR -
fillcell : The number of atoms obtained from symmetries, 40
is lower than the input number of atoms, natom= 80
This is not allowed.
Action : modify natom or the symmetry data in the input file.

The .cif of Y2O3 comes from ICDD,
......
_cell_angle_alpha 90
_cell_angle_beta 90
_cell_angle_gamma 90
_cell_formula_units_Z 16
_cell_length_a 10.6039
_cell_length_b 10.6039
_cell_length_c 10.6039
......
_chemical_formula_sum 'Y2 O3'
_symmetry_space_group_name_H-M 'I a -3'
......
_atom_site_occupancy
Y1 Y3+ 0.25 0.25 0.25 0.00759909 Uiso 1
Y2 Y3+ 0.4675 0 0.25 0.00696583 Uiso 1
O1 O2- 0.1087 0.34777 0.11949 0.008359 Uiso 1

Help me to solve this problem, thanks.

typat and xred have to agree. When expanded, typat equals 1 1 1 1 ... (i.e. Y1 Y1 Y1 ...) and xred are the Y1 Y2 O coordinates.
Accessorily are you sure about 16 Y1 + 16 Y2 in the unit cell - and not 8 Y1 + 24 Y2; with so many symmetry operations, an atom at 0.25 0.25 0.25 will be "folded back" onto itself many times ?

Kind regards,

Alain

Dear Alain, you are right, the variable typat should be "8*1 24*1 48*2", but the error message in my earlier post are still existing after the variable typat has been modified to "8*1 24*1 48*2".

typat not only labels the atoms in the cell but has to agree with the content of xred. As you give the positions of Y1 Y2 O in xred, typat should be 1 1 2
Look at the input reproduced below - pipe it to abinit - it only takes half a minute to run - and display it with your favorite CIF viewer.



Kind regards,

Alain

Dear Alain, thanks a lot for your post!
You have taught me a method to produce .cif file and xred value of all atoms in unit cell from a convertiional .cif file.
Now I know that the error message are caused by the variable brvltt. In my input file, the brvltt=-1, then the variable natom should be number of atoms in a primitive unit cell So the variable natom should be 40, not 80. In your input file, the brvltt=-1, then the variable natom should be number of atoms in a convertional unit cell. So the natom should be 80. right?
jingjing_lzu

right, brvltt -1 will reduce the unit cell to a primitive cell; as the latter is half the size and contains only 40 atoms. The conventional one contains 80.

Alain

Dear alain, thanks a lot for your help.