Buckingham potentials for MgO

Hypatio

I'm trying to investigate molecular dynamics simulation for MgO, although this research is somewhat new to me. For those of you with university access, one of the papers I am following is Shukla et al. (2008, J. Nuclear Materials) here: http://www.sciencedirect.com/science...22311508003632

I am confused about a few things:

They use the Buckingham potential:

[tex]\psi_{ij}=A_{ij}e^{-r_{ij}/f}-\frac{C_{ij}}{r{ij}^6}[/tex]

And they have used the following parameters:

O^1.7--O^1.7-: A=35686.18, f=0.201, C=32
Mg^1.7+-O^1.7-: A=929.69, f=0.29909, C=0

[1] Is there a reason why C=0 for Mg-O interaction? This would seem to imply there is no van-der-waals attraction between these two atoms?

[2] Why is there no Mg-Mg interaction but there is O-O interaction?

[3] I have looked at the potentials and compared them to the equilibrium positions of the cubic MgO lattice and I've noticed that the energy minimum for O-O interaction is about 2.85 Angstroms, which is close to the distances between oxygens in the MgO lattice (~2.98 Angstroms), but the repulsion of Mg-Mg atoms is high here, so how does the lattice stay together? Is it the coulombic interaction which for some reason isn't discussed in any detail in any of the papers I'm reading?

Useful nucleus

[1] C=0 is just to make things simpler. These potentials are produced by fitting to either experiment or electronic structure calculations. To simply the fitting process it is a common practice to remove the van der Waal terms of some interactions (but not the O-O!).

[2] For the same reason as [1]. It is very common in the field of modelling metal oxides with pair potentials to ignore cation-cation interactions. In the early days (consult the papers by Lewis and Catlow in 70's) people included cation-cation interaction but with time,experience showed that it is not necessary.

[3] Notice that Mg-Mg distance and O-O are both greater than Mg-O distance. The columbic attraction between Mg-O is enough to overcome the repulsion. However, I must say that in a many-body system it is not wise to think in terms of pairwise interactions.

Hypatio

Thanks! A few more:

[1] So, is the O-O interaction just much more energetic than that between cations?

[2] Are there circumstances under which cation-cation interaction becomes important? In particular I'm thinking of dielectric insulators which may have more complicated lattice structures.

[3] I am assuming that the partial charges identified (O: 1.7- and Mg: 1.7+) are meant to be included directly into the coulombic interaction as, for instance,

[tex]F_{Mg-O}=k_e\frac{q_{Mg}q_{O}}{r^2}=k_e\frac{-1.7\cdot 1.7}{r^2}[/tex]

But I want to understand how to compute (or find a database for) partial charges. From here: http://www.chemaxon.com/marvin/help/...ns/charge.html it is said that electronegativity is related to the partial charge by a quadratic function. How would I compute partial charges for Mg-O? Other interactions I want to understand are Si-O, Fe-O, and the cation interactions Si-Fe, Mg-Fe, Mg-Si.