Molecules (ions) in the unitcell

[Tomás]

I have a molecular crystal with 4 molecules (72 ions) per unit cell. When I am doing the POSCAR file, should the 72 ions included in the unit cell?. When I check the experimental structure with a graphic interface (PLATON) and I add molecules to the original one and draw the unit cell some parts of the molecules are out of this cell and that is my starting point. Of course, when I apply space symmetry operations I build all the crystal irrespectively of the molecules I consider for the beggining, but this is with such graphic interface. Have VASP any problem to simulate the crystal if any parts of the starting molecules (any ions) are out of the unit cell box?

Thanks

[Danny]

VASP uses Periodic Boundary Conditions so if you setup a structure there are no problems to give coordiantes ouside the unitcell, just make sure that PBC doesn't generate overlap of defined atoms(ie you give 2 atoms and they end up in the same place due to PBC).
After the calculation you'll find the coordiantes in the CONTCAR file all to be in the unitcell(though there might be exceptions, when an atom is just across one of the borders.

Danny