Evolution of the upper mantle of the Earth's Moon: Neodymium and strontium isotopic constraints from high-Ti mare basalts

Isotopic studies of mare basalts have led workers to conclude that their sources are heterogeneous on both large and small scales. Furthermore, these studies have led workers to postulate that depletion within the lunar mantle occurred early in its evolution and was a result of accumulation of mafic minerals from a LREE-enriched magma ocean. High-Ti basalts from the Apollo 1 I and 17 landing sites and ilmenite basalts from Apollo 12 are secondary evidence of this extreme, early depletion event. KREEPy rocks are the complementary enriched component in the Moon. A total of fourteen high-Ti basalts have now been analyzed from the Apollo 11 landing site for neo- dymium and strontium isotopes. A Sm-Nd internal isochron on basalt 10058 yields an age of 3.70 + 0.06 Ga, similar to 40Ar/39Ar ages of other Group B 1 basal@.. A compilation of all previously determined ages on Apollo 11 high-Ti basalts indicates four distinct phases of volcanism at 3.85 + 0.02 Ga (Group B2), 3.71 f 0.02 Ga (Group B3), 3.67 + 0.02 Ga (Group Bl), and 3.59 f 0.04 Ga (Group A). Whole- rock Sm-Nd isotopic data for all Apollo 11 high-Ti basalts form a linear array, which yields the age of the Moon (4.55 f 0.30 Ga). A similar regression of all uncontaminated high-Ti basalts from the Moon (both Apollo 11 and Apollo 17) yields an age of 4.46 + 0.17 Ga. Both arrays are interpreted as average source ages of the high-Ti basalts and are consistent with the formation of these sources by precipitation of cumulates from a magma ocean early in the history of the Moon. These new strontium and neodymium isotopic data, coupled with previously published data, are con- sistent with a two-component model for the upper mantle of the Moon. These two components include mafic adcumulates precipitated from a magma ocean prior to 4.4 Ga and small amounts (~2%) of trapped, KREEPy, late-stage, magma ocean differentiates. The mafic adcumulate evolves from 4.5 Ga, with '47Sm/'44Nd = 0.318 and *'Rb/%r = 0.005 to extremely radiogenic neodymium isotopic ratios and very unradiogenic strontium isotopic ratios. The KREEPy trapped liquid has a 14'Sm/ '44Nd = 0.168 and "Rb/%r = 0.235 and thus, evolves toward very unradiogenic neodymium and radiogenic strontium isotopic ratios. Because the KREEPy trapped liquid is enriched in both rubidium and the REEs by over an order of magnitude compared to the mafic adcumulate, trapping of even small proportions of this liquid in the adcumulate will control the radiogenic isotopic composition of the source. The apparent heterogeneity in the source regions of mare basalts could be caused by trapping of variable, yet small, proportions of this LILE-enriched liquid in the cumulate pile.