Abstract:
The solubility of both β-whitlockite and α-whitlockite has been experimentally determined between 1200 and 1400°C at 1 atm using a wide range of natural rocks and synthetic mixtures as starting materials. The solubility of both phases depends strongly on melt composition, decreasing systematically with increasing silica content and aluminosity. Experiments also show that α-whitlockite contains much more Na (0.5-6.3 wt.% Na2O) than β-whitlockite (<0.5 wt.% Na2O). Lunar low- and high-Ti mare basalts are far below the saturation limit of whitlockite and need 90-99% fractionation of olivine, pyroxene, plagioclase, and ilmenite to precipitate whitlockite, whereas KREEP basalts need less but at least 80-95% fractionation. It is shown here that both lunar mafic and felsic immiscible melts, the former enriched in Fe, REE, P, U, and Th, and the latter in Si and K, are undersaturated in whitlockite, and further fractionation of fayalite, ilmenite, plagioclase, and K-feldspar is required to reach the saturation limit. Thus, lunar whitlockite must have crystallised from highly fractionated residual melts. Lunar whitlockite, which is low in Na (0.09-0.49 wt.% Na2O), crystallised originally as β-whitlockite from low-temperature residual melts. In contrast, meteoritic whitlockite contains more Na (0.5-3.3 wt.% Na2O and therefore, had formed initially as α-whitlockite at higher temperatures and transformed into β-whitlockite upon cooling. It is proposed that the interior of the Martian mantle and crust was enriched in volatiles in its early history (4.6-1.3 Ga), but has become essentially dry and very depleted in water and halogens at least since the last 180 Ma. Calculations show that the Earth's crust and mantle as a whole contains only 5% of the total P of the Earth, and the remaining 95% is stored in the core. In contrast, the crust and mantle of Mars are much more enriched in P and contain as much as 43% of the Martian total P budget, with the remaining 57% being distributed in the relatively smaller Martian core. This difference in the distribution of P among planetary shells must have resulted from a more oxidising environment during the accretion and early evolution of Mars compared to the more reducing conditions under which Earth formed.