# What type of potential is suitable for ion bombardment in Molecular Dynamics?

I would like to study ion bombardment on targets with molecular dynamics, especially on oxides. So far I ran conclusive tests with Lennard-Jones potentials, but I would like to obtain better quantitative results. The literature on the potentials for oxides recommends the use of a Buckingham potential ($$U = A e^{-r/\rho} - C/r^6$$) and a massless core-shell model, where the cores and shells have charges $$q_\text{core}$$ and $$q_\text{shell}$$. The cores and the shells are bonded by harmonic bonds ($$E = \frac{1}{2} k r^2$$). The fitting of the potential and bonding on experimental data gives the parameters $$A$$, $$\rho$$, $$C$$, $$q_i$$ and $$k$$. Note that the interactions defined by the Buckingham potential are for shell-shell interactions.

This is all good, but I still have a few questions about the way to proceed, and I hope that some of you can help me clarify those points.

• First I am unsure about the general method to follow for testing a potential. I am using LAMMPS, and the way I am going with this so far is: I input the initial positions of the atoms as well as the bonds between cores and shells. Then I input the type of potential (Buckingham + core-shell) and give the parameters mentioned above. So far so good. Then I perform an equilibration run at room temperature. Here I am not sure what to expect. If all is good, am I supposed to have a crystal where the atoms are vibrating locally, but with a stable overall structure? When I reach this point, should I run other tests for comparison to experimental data?

• Second, I have some doubts about if this type of potential is well-suited for ion bombardment. The literature gives fit parameters for M-O (metal-oxygen) interactions as well as for O-O, but not for M-M. This might be because the metals don't bond together in the crystal, but what if during the bombardment a metal atom gets dislocated and collides with another metal atom? Surely this interaction should also be defined. Any idea?

• Third and last, also related to the relevance of this type of potential in the context of ion bombardment. How do I define the interactions between ions and target atoms M and O? Should I also use a Buckingham potential with a core-shell model? If yes, I guess that I should try to find individual parameters in the literature for the interactions ion-M and ion-O. If no, do you have any suggestion?

Thank you very much in advance for your answers and ideas. I hope that this discussion will also be helpful to other people starting with molecular dynamics.

Julien.