146Sm-142Nd formation interval for the lunar mantle

Abstract

Now-extinct 146Sm (t1/2= 103Ma) was present in the early solar system, and lunar basalts might be expected to have inherited small anomalies in 142Nd abundances from mantle sources with variable Sm/Nd ratios established during the early global differentiation of the Moon. 142Nd/144Nd ratios were measured for several lunar basalts, and values of ϵNd142, (deviations from terrestrial 142Nd/144Nd in parts in 104) calculated for them. The ϵNd142 values show very small variations attributable to 146Sm decay. Neutron capture near the lunar surface also modified both the samarium and neodymium isotopic compositions of the basalt samples. The calculated neutron fluences range from ∼0 to ∼7.8 × 1016 n/cm2. Both thermal and epithermal neutron fluences were calculated from the Sm isotopic compositions of the basalts and used to correct the ϵNd142 values for neutron capture by neodymium. Well-resolved radiogenic enrichments ϵNd142 and +0.17 ± 0.08 (2σ), respectively, were measured for lunar meteorite Asuka 881757 and Apollo 17 basalt 74255, respectively, for which the neutron fluences were nil. Basalts 70135 and 75075, which were exposed to small, measureable neutron fluences of ∼2.5 × 1015 and ∼7.7 × 1015 n/cm2, respectively, have neutron-corrected ϵNd142 values of +0.25± 0.15 and +0.29 ± 0.11, respectively. The average value of radiogenic ϵNd142 for the three high-Ti basalts from Apollo 17 is +0.22 ± 0.06. Neutron capture effects were greater for the other basalts, but can be corrected using the measured neutron fluences derived from the samarium isotopic data. The neutron-corrected ϵNd142 for Apollo 12 ilmenite basalt 12056, consistent with its derivation from a highly depleted mantle source also. Three low-Ti basalts from Apollo 12 and Apollo 15 (12038, 15076, 15555), have neutron corrected ϵNd142 which are not resolved from zero, and average +0.04 ± 0.06. A fourth low-Ti basalt, 12039, has neutron-corrected ϵNd142, but is subject to the largest neutron correction of ∼0.32 ε-units. KREEP basalt 14078 was exposed to only a small neutron fluence of ∼6.4 × 1015 n/cm2, and has a small deficit of 142Nd, corresponding to ϵNd142. The average neutron-corrected ϵNd142 of three KREEP basalts is −0.05 ± 0.04. The ϵNd142 values of the basalts correlate with 147Sm/144Nd values for their source regions as calculated from their ages and initial ϵNd142 values. 147Sm- 143Nd and 146Sm- 142Nd systematics were combined in a three-stage model yielding an isochron equation for the lunar mantle formation interval, which was calculated to be 238−40+56 Ma (2σ). Sm/Nd ratios for the source regions also are derived from the model and compare favorably with values from conventional geochemical models of mare basalt petrogenesis. If the Moon were formed by impact of a large, Mars-sized planetesimal with the Earth, the impact was early enough that the lunar mantle cooled to neodymium isotopic closure by ∼4.32 Ga ago. The bulk lunar ϵmoon142 evaluated from the mantle isochron at (147Sm/144Nd)chur = 0.1967 is −0.01 ± 0.03 (2σ), consistent with derivation of lunar and terrestrial neodymium from a common reservoir.