Should the oriatation of the cell be like hexagonal lattices

[saray1360]

Dear all,

I have got abit mixed up by the way different articles make supercells. As an example the below article:

http://arxiv.org/PS_cache/cond-mat/pdf/0511/0511274v3.pdf

for making the super cell for the single layer ZnO:

1. How should we make the unit cell?
2. How amny atoms should we consider there in a unit cell? (should it be more than two atoms: 1-zn and 1-O necessarilly, why?)
3. Should the oriatation of the cell be like hexagonal lattices (90 90 120) or we should change it to a cubic one? (90 90 90)


Regards,
Sara

 

[Modey3]

Hello,

It all depends on what you want to calculate, but a supercell should be chosen with calculation time in mind. For simple lattice parameter and band structure structure calculations, try to use the smallest supercell as possible; i.e. minimize the number of atoms in your supercell. If you are planning on doing ab-initio molecular-dynamics you will want to keep the number of atoms as small as possible because calculation time scales with the number of atoms.