Very high-K KREEP-rich clasts in the impact melt breccia of the lunar meteorite SaU 169: New constraints on the last residue of the Lunar Magma Ocean

Abstract

In the impact melt breccia (IMB) of Sayh al Uhaymir (SaU) 169, the most KREEP-rich lunar meteorite to date (Gnos et al., 2004), clasts of a new type of lithologies were discovered, consisting of Ca-poor and Ca-rich pyroxenes (60.8vol.%), Ba-rich K-feldspar (27.9vol.%), phosphates (5.6vol.%), Nb-rich ilmenite (4.0vol.%), zircon (1.2vol.%) and minor sulfide (0.6vol.%). These mafic lithic clasts are more enriched in KREEP component (∼1500×CI) than the host meteorite and are highly enriched in potassium. They are referred to as very high-K (VHK) KREEP lithology, and probably most close to the last residual liquid of the Lunar Magma Ocean without significant dilution by other Mg-rich magmas. The fine-grained matrix of the SaU 169 IMB has very similar mineral chemistry to the VHK KREEP lithology, but contains abundant plagioclase with trace K-feldspar. The matrix shows decoupling of K from the REEP-like component; however, it cannot be simply interpreted by mixing the VHK KREEP lithology with anorthosites, which should have diluted the REEP-like component with the same proportion.SIMS Pb–Pb dating was conducted on zircons in various petrographic settings and with different crystal habits. All analyses show a main age peak at 3921±3Ma and a smaller one at 4016±6Ma. The main age peak is identical to the previous Pb–Pb age by Gnos et al. (2006) and U–Pb age by Liu et al. (2009), dating the catastrophic shock event contributed to the formation of SaU 169 IMB. The older ages are consistent with the previous report of an older bulk U–Pb age by Kramers et al. (2007), suggestive of presence of relict crystals in a few large zircon grains. The VHK KREEP clasts predated the fine-grained matrix, but have the same zircon Pb–Pb ages as the latter within the analytical uncertainties. Plagioclase was converted to maskelynite whereas zircon was shocked to diaplectic glass, probably by a second event at ∼2.8Ga. However, the identical zircon Pb–Pb ages of the amorphous parts and the remained crystalline areas indicate no resetting of Pb–Pb isotopes by the later shock metamorphism, or there was another severe impact event postdated solidification of the fine-grained matrix within a few million years.