# Can tersoff/table be used for arbitrary g(theta)?

The Tersoff angular function g(theta) has an analytical form that only
allows one angle (strictly two angles: positive and negative values of
the angle). This works well for "tetrahedron-based" structures like
diamond, zinc-blende, and wurtzite, that have coordination 4 (so the
angle is 109.2 degrees).

But it doesn't make any sense for structures like hcp with
coordination 12, where multiple angles are involved: +/- 30, +/- 60,
180 degrees, etc.

Can I construct a "numerical" g(theta) with these multiple angles and
use with tersoff/table in lammps? so I can use tersoff with an hcp
structure?

MEAM is another option but I just found out MEAM in lammps doesn't
support 'wurtzite' structure.

(I'm trying to develop a two-component potential that works with
wurtzite, while its single-component works with hcp.)

F

Correction: hcp angles like +/-60, +/- 120, 180 degrees, etc.

i cannot provide a specific answer to your search for a suitable potential file, but i would like to point out, that there is a tersoff/mod potential with a modified (more flexible) angular term.
if you need something beyond that, it would probably be the cleanest approach to follow that example and create a similar, derived class where you replace the angular potential functional form with a different one you deem suitable to your needs.

developing empirical potentials is a process full of trial and error. so you will have to make experiments and tests and see what would work and what is not helping.

axel.

Thanks so much Axel.

I will go ahead and review the literature, as well as try to follow an
example, so I could derive from it and modify.

Best regards,
F