Aluminium's ore bauxite, again due to the high reactivity of aluminium and it's places of occurrence, has a bad habit of picking up various other kinds of oxides as impurities including SiO2, TiO2, iron oxides etc.
Due to magnetic nature of impurities themselves, magnetic separation is out
gravitational separation is risky, because due to such a large sample space of impurities
it's quite likely that there is a type of oxide impurity with similar density as that of bauxite
So that again goes..and froth floatation is only for sulphide ores
3
So, all popular physical separations have failed