Eim potential in short range

LAMMPS LAMMPS General Discussion
1

With regard to all community
In using eim potential, what is the best way to improve short range interactions (high energy collisions)? Is hybrid/overlay eim zbl 0.1 3 a proper style? In this case, i am worry about the distance between 0.1 to about 1 angstrom, where eim potential has NAN value, therefore there will be a great uncontinuously between below an beyond the 1 angstrom. Any suggestion will be appreciated.
Best regards

2

We know that hybrid/overlay adds zbl potential to the main potential in the determined range. Eim potential has a large repulsive value from 1 angstrom. I think, Addition zbl potential to eim below 1 angstrom will result a large discrepancy.

3

There a basically two ways to create an EIM-like potential that uses ZBL as its core.

  • create a custom pair style. you should be able to use the same potential file, since it has all relevant information, including what would be needed for ZBL
  • use hybrid/overlay with a tabulated potential that adds ZBL and subtracts the pairwise part of EIM where it overlaps with ZBL

Of course, the latter would require that the EIM pairwise term is finite where it would be replaced by ZBL. There is an example for creating such a table for EAM in the tools/tabulate folder.

I only know examples of ZBL being used as a replacement for core interactions for non-polar or slightly polar compounds, where the core interactions are not so different. So I am not so sure it is a good choice to do the same, when you have (strong) coulomb interactions as for ions. Have you made a survey of the published literature, if somebody has done something like that before? and whether that produced accurate results?