Implementation of Improved Pedone's interatomic Potential

Dear,

I would like to implement the improved Pedone potential in LAMMPS (https://doi.org/10.1103/PhysRevMaterials.5.045602), referred to in the literature as BMP. I am currently using the original potential, referred to in the literature as PMMCS (https://doi.org/10.1021/jp0611018), which I have implemented without problems since the LAMMPS release of April 17, 2024. The enhanced potential contains the parameters for the Pedone potential for binary oxides, the Buckingham potential for interactions between identical atoms, and also includes terms related to the Harmonic three-body potential parameters (BMP-harm) and Screened harmonic three-body potential parameters (BMP-shrm).

For PMMCS, I’m using something like

pair_style hybrid/overlay pedone 15 coul/long 15
pair_coeff * * coul/long
# Dᵢⱼ(kcal/mol) aᵢⱼ (Å⁻²) rā‚€ (ƅ) Cᵢⱼ (kcal/mol ʲ)
pair_coeff 1 2 pedone 7.853359 2.006700 2.100000 23.060540 # Si–O
pair_coeff 2 2 pedone 0.977652 1.379316 3.618701 507.331880 # O–O

The specific question is how do I complement the interatomic potential file to include the rest of the interactions? In the attached photo is the table with the parameters of the potential I would like to implement. Thanks.

Given the complexity of the improved potential, I don’t see an alternative to writing a new pair style in C++.

Have you check which simulation software the authors of the paper in question are using?

The 3-body part looks like pair_style nb3b/harmonic command — LAMMPS documentation

1 Like

I am affraid they did not specify which software have used. Is it possible to mix Pedone’s and Buckingham’s potentials with the @mkanski 's answer?

You have the paper in front of you not me. So you are in a better position to judge. I suspect that not.

You could ask the (corresponding) author(s).

1 Like

By searching the title of the paper I found the website of the group (CompMaterChem - Downloads), and it seems that they have placed the files needed for lammps simulations there. I did not investigate how to use them, though.

1 Like

Thank you very much @initialize. Your suggestion did work and now I am trying this interatomic potential.

According to @initialize’s reply, yes. It is fully developed in LAMMPS but for me, as a beginner, it is a little bit tricky the way that authors have implemented it. Anyway, thank you for your support.