[Wien] Supercell structure co-ordinates are different in Wien2k and PHON

[Fecher, Gerhard]

Yes, if you use the cell with 12 (NONEQUIV.ATOMS !) and split the position of one of the Fe atoms it's ok
thus you can move one of the V, AL and Fe atoms each independent  of their equilibrium positions
In the same way you may split the position of one of the Fe atoms in the larger supercells.
(It's not a preferential method to do that by hand in the struct files)

The cell has indeed not 12 but 16 atoms (the 4 iron atoms appear twice), that is, it has 12 inequivalent positions occupied by overall 16 atoms.

If you start withthe structure in space group  F m-3m (225)
a)     and make a 1x1x1 supercell with F lattice (result is a cell with 4 positions) and afterwards from this a 1x1x1 cell with P lattice you will have the cell with 16 inequivalent atoms.
b)     and make a 2x2x2 supercell with P lattice you will have the cell with 96 inequivalent atoms (again all Fe appear twice).
c)     and make a 1x1x1 supercell with F lattice and afterwards a 2x2x2 supercell with P lattice you will have the cell with 128 inequivalent atoms.
(use sgroup to check that it finds back your initial structure,  in all cases)

I guess the use of the P, F, and B lattices is not realy intuitive. Example:
The primitive cell of the Fe2VAl compound is fcc with a basis of 4 atoms (in the detailed F m-3m space group (225) you have only "3 independent atoms" occupying the 4a, 4b, 8c positions, 4+4+8=16)
in the fcc lattice you need, however, to occupy 4 positions with the 4 atoms (2+1+1), those positions are: 0,0,0, 1/4,1/4,1/4, 1/2,1/2,1/2, and 3/4,3/4,3/4,3/4  ! (note: this corresponds in 216 for example to 4a, 4b, 4c, 4d)
the primitive cell has not a cubic shape but it is rhombohedral with
a1=a2=a3 = ac/sqrt(2)  and  alpha=beta=gamma=60°, where ac is the lattice constant of the cube.
(Try it, use a P lattice with those parameters and the 4 positions occupied and you will see that sgroup finds F m-3m, with the 3 iequivalent positions occupied)

We learn: NONEQUIV.ATOMS means NONEQUIValent POSITIONS but not number of atoms in the cell.
                                         
My remark on P1 was that you cannot use the 12 atoms cell directly but the one with 16 atoms produced by (a) (this is a setup in P1 symmetry).
There is no need to have P1 before running supercell, see above.

If you keep the atoms in the high symmetry positions then the forces will be Zero (did you ever check youre output and its meaning ?) 
To find the forces you need to move the inequivalent atoms one by one from their equilibrium positions to calculate the forces.
(I would not do that by hand but its a good thing to play with and to see whats going on with symmetry and forces.)
These forces can then be used to calculate the phonons. 
(by the way, don't use mini positions or MSR1a, then the forces will again be zero, hopefully, otherwise youre structure is not stable))

Note: I did not check if besides the monoclinic space group P m (6) something orthorhombic may also describe the output of 1x1x1 supercell for P lattice.

It is interesting to note that the structure can also be described by the primitive cubic space group P m-3m
and occupying the 8g (Fe), 1b, 3a (V) and 1a, 3c (Al) positions (this are 5 independent positions).
(also interesting , this does not work directly with P, use 221 and save it, then change to P, 
structgen in W2web tells "Error - incorrect space group symbol", just continue and sgroup finds again F m-3m)

We learn: P is not equal to P, unfortunately.


Ciao
Gerhard

DEEP THOUGHT in D. Adams; Hitchhikers Guide to the Galaxy:
"I think the problem, to be quite honest with you,
is that you have never actually known what the question is."

====================================
Dr. Gerhard H. Fecher
Institut of Inorganic and Analytical Chemistry
Johannes Gutenberg - University
55099 Mainz
and
Max Planck Institute for Chemical Physics of Solids
01187 Dresden
________________________________________
Von: wien-bounces at zeus.theochem.tuwien.ac.at [wien-bounces at zeus.theochem.tuwien.ac.at] im Auftrag von venkatesh chandragiri [venkyphysicsiitm at gmail.com]
Gesendet: Freitag, 5. Dezember 2014 05:22
An: wien at zeus.theochem.tuwien.ac.at
Betreff: Re: [Wien] Supercell structure co-ordinates are different in Wien2k    and PHON

Dear Dr.Gerhard

Thanks for your quick answer. According to you, I should start with P1 cell,

which could also obtained by splitting the atomic positions of Fe-atoms

of this pm(6) space group (12 atoms). Further, In your mail, I have seen the statement

"For Phonon calculations one would need the P 1 cell,
otherwise one moves (in certain cases)   2 or more atoms at once out of there
equilibrium positions (here Fe atoms)"


Can please, elaborate it (why one needs P1 cell as starting structure (before doing supercell))

for phonon calculations...why the atoms go out of their equilibrium positions, if i start with other than P1 symmetry.



Thanking you and looking forward to your answers

Sincerely,
Ch. Venkatesh