[apexaviour]

Mg(II)F₂ has a rutile structure like TiO₂. It forms the same structure as MnF₂, NiF₂, CoF₂ etc.. which are all transition metal cations and have d-orbitals forming your octahedral crystal field (ie surrounded by 6 anions). The d-orbitals split into e_g and t_2g in this field and the bonding is fairly straight forward.

Anyway... magnesium is element 12, no d orbitals. Okay thinking about it now the p-orbital lobes could line up pointing in 6 directions... hmmm is that the case or am I way off? Is there some hybrisation with the Mg 3s? Or am I barking up the wrong tree here...

Jebus I really need to find out more about the bonding in this compound. Any recommended reading or papers would be appreciated... I've exhausted my searching ability on web of science.

Cheers 

-experimental physics grad student desperately trying to write up thesis.

[cheese (MSW)]

If you had taken your BSc in chemistry () you would have learned early on that an important factor in the determination of the structure of solids that are mainly ionic is the radius ratio [google] r+/r- where r+ is the radius of the cation and r- is the radius of the anion.   Electrostatic attraction is nondirectional (unlike covalent bonding): the more anions that can be packed around a cation the more energetically favorable it is; opposing this is the electrostatic repulsion between the anions.  It is predicted that if r+/r- is between 0.4 - 0.7, 6:6 (NaCl lattice) 6:3 (TiO2) will be preferred, whereas if r+/r- is >0.7,  8:8 (CsCl) 8:4 (CaF2) is predicted to be more stable.   There is only a rough correlation between r+/r- predictions and what is observed; the exceptions are rationalized that covalent contributions tip the structure in favor of 6:6 (AX) or 6:3 (AX2) coordination.  Note that Na^+ in NaCl is six-coordinate and doesn’t have any d AOs to talk about.
So calculate r+/r- for your cmpds [the data at http://www.webelements.com/  appears reliable] and confirm that radius ratio goes a long way in determining the structures of MF2 cmpds (note that CaF2 with the larger Ca^2+ ion has 8:8 coordination).  The r+ for Mg^2+ and say Mn^2+ are similar.
Lattice Energies (+ve values, kJ mol^-1) MgF2 2926 (calc) 2978(exp);  MnF2 2644(calc) –(exp)
(CRC Handbook).
For NiF2, CoF2 there is added CFSE that you are familiar with (F^- is however weak field ligand).  For  MnF2 (high spin d^5) and ZnF2 (d^10) there is no CFSE, but the cmpds are still stable.
The following is somewhat  outside my area of expertise.)  There is some interaction of the empty 3p AOs on Mg^2+ with the filled 2p AOs of F^- to give the valence band, but the interaction is small because there is a large energy difference between interacting AOs.  MgF2 is an insulator with a large band gap.
Further reading at your level: J. E. Huheey, E. A. Keiter, R. L. Keiter Inorganic Chemistry 4th ed. Chpt 4.

[apexaviour]

I'm not sure I could have hoped for a better answer. Everything you wrote either directly or as a nucleation point, filled in pretty much all the gaps and questions I had. It also helped me remember some of the two years of undergrad chemistry I thought I'd forgotten.

You helped me plough through a section of my thesis I was really stumped and frustrated with. My sincerest thanks!