Changing the pair potential for Ni-Si using MEAM

Hello all,

I am trying to model the interface between a Nickel layer and a Silicon layer using the MEAM potential. A similar system has been modeled before by Baskes et al (their paper is “Atomistic calculations of composite interfaces” in the journal Modelling Simul. Mater. Sci. Eng. 2 (1994) 505-518).

The only parameters they vary for the inter-layer interactions are rho0 and alpha, which I can update easily enough in the parameter file. However, they give a different formulation of the pair potential (phi) for the Ni-Si interaction. That does not seem like something I can update using the parameter file. Do I have to go into the source code and manually code the pair interaction for Ni-Si or is there a built-in method to account for this in LAMMPS?

Thanks for any suggestions,
Michael

Greg can answer the MEAM part of your Q. I don’t know
how difficult it would be to modify the pair term inside
the MEAM lib.

If you can separate the pair interaction out of MEAM,
then you could possibly use pair hybrid/overlay to
effectively compute MEAM + pair. The pair part
could be a potential already in LAMMPS, or you
could tabulate it and use pair table, if it is truly
a pairwise interaction.

Steve

Hi Michael,

I think the capability you’re looking for is already in LAMMPS/MEAM. The particular form of the pair potential for cross-element interactions depends on the alloy reference structure, which is the L12 structure for Ni3Si as given in the Baskes paper. For many simple alloy structures, where each element only sees neighbors of the opposite type (like the NaCl structure), there’s a straightforward formula. But for more irregular structures, including L12, the particular form does have to be handled separately.

Fortunately, L12 has already been coded. Line 296 in meam_setup_done.F (or thereabouts – it depends if I’m looking at the latest version) has the pair potential for L12 that matches equation (10) in the Baskes paper. Just make sure to specify the “l12” reference structure, and make sure your input files are set up such that Ni is element 1 in the interaction, and Si is element 2.

Regards,

Greg

Greg Wagner

Manager, Thermal/Fluid Science and Engineering Department

Sandia National Laboratories, Livermore, CA

Tel: (925) 294-2180 Fax: (925) 294-3410

Email: gjwagne@…3…