Dust from collisions: A way to probe the composition of exo-planets?

The crustal-type material of group A would possibly require the stripping of upper layers by grazing-style hit-and run encounters or high energy events like evaporation/condensation in giant collisions. In tidal disruptions or the involvement of predominantly icy/water bodies the resulting mineral dust would originate mainly in one of the involved planetesimals. This could allow attributing the observed composition to a specific body (such as e.g. Eta Corvi).