ABINIT Discussion Forums

Hi all,

I'm a new user of abinit. After read some explanations of the rprim, I've some some questions on the rprim's non-uniqueness.

Say, for the Wurtzite (B4) Structure, the rprim can be taken as follows:
(In the following form, I use comma to separate each matrix elements only to increase the readability. )

\frac{{\sqrt 3 }}{2}, \frac{1}{2}, 0
- \frac{{\sqrt 3 }}{2}, \frac{1}{2}, 0
0, 0, 1

The above rprim is used by the tutorial of AlAs, one can see telast_1.in for detail.

On the other hand, from the following webpage,

http://cst-www.nrl.navy.mil/lattice/struk/b4.html

I find the following primitive vectors taken for Wurtzite (B4) Structure:

{A_1} = \frac{1}{2}aX - \frac{{\sqrt 3 }}{2}aY
{A_2} = \frac{1}{2}aX + \frac{{\sqrt 3 }}{2}aY
{A_3} = cZ

In the latter case, the rprim should be written as follows:

\frac{1}{2}, - \frac{{\sqrt 3 }}{2}, 0
\frac{1}{2}, + \frac{{\sqrt 3 }}{2}, 0
0, 0, 1

I want to know, whether both of these two forms of rprims are right for Wurtzite (B4) Structure?

Thanks in advance.

Hongsheng.

They are equivalent, just rotating x to y. There is always a large degree of freedom in choosing the primitive vectors (see the first chapters of Ashcroft+Mermin)

m.

mverstra wrote:They are equivalent, just rotating x to y. There is always a large degree of freedom in choosing the primitive vectors (see the first chapters of Ashcroft+Mermin)

m.

If so, the different primitive vectors will make the corrresponding xred dataset diferrent also, am I right?

Another issue is: in my mind, the values of all the elements of xred should be range from 0 to 1. But in the Materials Studio, I take Si as the example to generate the .cell file which is the input file for castep, then I use abinit-6.6.1\util\users\Cell2Abinit.py to obtain the corresponding structure information for abinit. The following is the partial content of the original .cell file generated by Materials Studio:

----------------
%BLOCK LATTICE_CART
-0.000000000000001 2.715007197043869 2.715007197043870
2.715007197043869 -0.000000000000001 2.715007197043871
2.715007197043870 2.715007197043870 0.000000000000001
%ENDBLOCK LATTICE_CART

%BLOCK POSITIONS_FRAC
Si 0.0000000000000000 0.0000000000000000 0.0000000000000000
Si 0.2499999999999999 1.2500000000000000 0.2500000000000001
%ENDBLOCK POSITIONS_FRAC
-----------------

When I run the Cell2Abinit.py on this .cell file, I obtain the following outputs:

----------------------
python Cell2Abinit.py si.cell
####################################################
# Atoms Stucture created by CellToAbinit Converter !
####################################################

# Definition of the unit cell
#The length of the primitive vectors
# 1 Bohr=0.5291772108 Angstroms
# acell 2.715007 2.715007 2.715007 Angstroms
acell 5.130620 5.130620 5.130620
rprim
-0.000000 1.000000 1.000000
1.000000 -0.000000 1.000000
1.000000 1.000000 0.000000

#Definition of the atom types
# 2 kind of atoms
# ['Si']
ntypat 1
znucl Needed

#Definition of tha atoms
natom 2
typat 1 1
xred
0.0 0.0 0.0
0.25 1.25 0.25

##### End of CellToAbinit Conveter ! #####
----------------------

We can see that the value of the element of the xred is more than 1, is this reasonable or not?

Thanks in advance.

Regards.

Hi

Your cell is periodic, so it shouldn't be a problem, unless abinit stops if it detects a value greater than 1. But I don't think so since you can put negative values for xred.

This should work

Boris

Hi Hongsheng,

If so, the different primitive vectors will make the corrresponding xred dataset diferrent also, am I right?

no, not necessarly, the reduced coordinates are defined with respect to the primitive translations. So, if you rigidely rotate your primitive translations you can keep your xred unchanged, it's equivalent to rotate the whole structure (lattice+atoms) which, physically, changes nothing.

Another issue is: in my mind, the values of all the elements of xred should be range from 0 to 1. But in the Materials Studio, I take Si as the example to generate the .cell file which is the input file for castep, then I use abinit-6.6.1\util\users\Cell2Abinit.py to obtain the corresponding structure information for abinit

I suggest you to use the geometry builder of abinit ( see for example in the doc the input variable brvltt) which allows you to build your structure from its crystallographic description which you can find, for example, in the Pearson handbook.