Estimating the Comparative Volatility of Compounds with a Common Ion

[Veritas_C&E]

I've come to accept the notion that it isn't straightforward to predict volatility from the sheer atomic composition of a compound. I'm wondering, however, about the case of anticipating the volatility of compounds mainly consisting of a common ion (i.e. whose main structure is atomically identical):

• Can we generally rank the volatility of different compounds with a common ion, based only on the atomic weight of the compound (assuming each would bind in the same numbers with the same number of common ions, like CuCl2 and ZnCl2)? Is this a wrong rule, a good rule of thumb (true most of the time), or an absolute rule (true all of the time)?

• Can we generally anticipate the respective volatility of these compounds as a proportion of a sort (to the weight of the molecule for instance)? Again, would this be a wrong rule, a good rule of thumb (true most of the time), or an absolute rule (true all of the time)?

As a practical example, I'm considering separating a mix of a bunch of metal oxalate complexes by sublimation. I would first like to know if this at all would work well, and try to have a reliable method to anticipate the respective volatility of different oxalate complexes, to known how closely I should control temperature and other experimental conditions.