Is tersoff potential enough to describe properties of Si

Dear LAMMPS users,

Is pair_style tersoff together with the potential file Si.tersoff enough to describe the properties of Si? Do I need to set the bond, angle, and dihedral types? Thanks.

Regards,

Kira Banks

Dear LAMMPS users,

Is pair_style tersoff together with the potential file Si.tersoff enough to
describe the properties of Si?

yes. well, within the limitation of the given parameterization, that is.

Do I need to set the bond, angle, and
dihedral types?

does the documentation tell you to do so anywhere? no.
in fact, if you look through the mailing list archive, you'll
come across discussions, how that can create problems
through the implied excluded interactions on the neighborlists
from the bonds, angles and dihedrals.

you should read some of the papers on the parameterization
of tersoff potential and see how it models interactions.

cheers,
     axel.

Hi Kira,

Is pair_style tersoff together with the potential file Si.tersoff enough to
describe the properties of Si?

Depends on what properties you wish to describe and model. Tersoff
produces Si properties as good as said in its publication.

Do I need to set the bond, angle, and

dihedral types? Thanks.

No, you don't; just use atom_style atomic. Tersoff is a pure atomic
empirical potential and it will determine bonds and angles for you,
but not dihedrals.

Dear Alex,

Thanks a lot.

Bests,

Kira

Dear Ray,

Thanks a lot.

Bests,

Kira