| Abstract: | The formation of earliest stable planetary crust poses the following dilemma. All surface crusts formed by direct quenching of planetary-scale magma are mafic and dense, and will founder. All felsic crystals light enough to float will form only after protracted differentiation and will grow too slowly to be separated from their parent magma on the short timescale of near-surface residence. Early crust must be intrinsically stable to accretionary bombardment in order to survive, hence it must float. One solution is to make pop-up felsic crust on a rapid time scale by collecting the lighter evolved liquid, not the crystals. Such a liquid is produced by solute rejection at depth from growing mafic crystals, and it rises by compositional convection. It stirs and cools the magma and reduces the accretionary thermal maximum, perhaps to zero. It has a high viscosity and high barrier to nucleation, hence may coalesce metastably. Near the surface, rapid crystallization will follow spontaneous nucleation. Rafts and rockbergs will coalesce to form the earliest crust. On the moon, this is anorthositic. On the earth, the presence of water enhances the silica content of the network-rich liquid and gives rise to tonalite. Existing ancient tonalites may reflect an original tonalitic crust. Modern oceanic rhyolites and trondhjemites may owe their existence to a similar process of liquid collection, and could therefore serve as a test of the crust-forming hypothesis. |
