Supercell calculation

[Inaoton]

Hello sir
    How can i  find the equilibrium lattice constants of a unit cell from supercell calculation? Do i have to divide  by a scaling factor.? And what about the equilibrium volume of the unit cell?

[lknife]

In my opinion, after geometry optimizing the supercell, you can get the equilibrium lattice constant by noticing how many times you repeated your unit cell in each direction. Also, it's a good idea to optimize the geometry with the smallest unit cell. In addition, the lattice constants are different if you choose different unit cell with different symmetry.

[Inaoton]

Thanks a lot .
1)But if i am supposed to model a structure with vacancy/vacancies, can i get a perfect model of the structure with the primitive cell where i delete atom/atoms?
                  Or
do i have to repeat the primitive cell and use substitutional alloy method or VCA (for a good model)?
2)Besides this, if the concerned structure is having a structure with fcc configuration(a=b=c), and if i repeat the cell by 2X2X1, the lattice constant so obtained do i need to scale the resultant lattice constant (after optimizing the geometry) by 1 or  2 to get the equilibrium lattice constant of the unit cell?

[Petr Khomyakov]

You may first calculate the lattice parameters of the perfect crystal for the primitive cell.

For defect calculations, you can then build a supercell by repeating the primitive or conventional unit cell of the perfect crystal with the lattice parameters obtained in the first step. 

Afterwards, you can create a defect(s) in the supercell structure, and do ion relaxation of this created structure with a defect(s).

So, this is how you may model a crystal structure with a defect(s).