A strontium and neodymium isotopic study of Apollo 17 high-Ti mare basalts: Resolution of ages, evolution of magmas, and origins of source heterogeneities

Abstract

A combined Sr and Nd isotopic study of 15 Apollo 17 high-Ti mare basalts was undertaken to investigate geochronological and compositional differences between previously identified magma types (A, B1, B2, and C). Whole-rock and mineral separates for one of the least-evolved Type B1 basalts, 70139, yield Sm-Nd and Rb-Sr isochron ages of 3.71 +/- 0.12 Ga and 3.65 +/- 0.13 Ga, respectively. A more-evolved, Type A basalt, 71539, exhibits a slightly older Sm-Nd isochron age of 3.75 +/- 0.07 Ga and a Rb-Sr isochron age of 3.67 +/-0.10 Ga. Although these two ages are non-resolvable by themselves, compilation of all available geochronological data allows resolution of Type A and B1/B2 ages at high levels of confidence (>99%). The most reliably dated samples, classified according to their geochemical type, yield weighted average ages of 3.75 +/- 0.02 Ga for Type A (N = 4) and 3.69 +/- 0.02 Ga for Type B1/B2 (N = 3) basalts. Insufficient geochronological data are available to place the rare, Type C basalts within this stratigraphy. We propose that age differences correlate with geochemical magma type, and that early magmatism was dominated by eruption of Type A basalts while later activity was dominated by effusion of Type B1 and B2 basalts.Whole-rock isotopic data yield distinct differences in initial Sr and Nd isotopic compositions between Types A, B1, B2, and C basalts. Types A, B1, and C exhibit restricted intra-group compositional variations and lie along well-defined whole-rock isochrons. These data are consistent with petrogenetic models involving closed-system fractionation of observed microphenocrysts from chemically and isotopically distinct parental magmas. In contrast, a wide range of Type B2 initial isotopic compositions indicates mixing of several distinct components during magma evolution.The Sm-Nd whole-rock isochron age for Type A, Bl, and C basalts of 3.79 +/- 0.15 Ga is within error of Apollo 17 eruptive activity. However, the very well-defined Sr whole-rock isochron age of 4.02 +/- 0.05 Ga is 270 to 330 Ma older than eruptive ages. Isotopic and petrological arguments indicate that extensive Rb/Sr fractionation did not occur at the time of melt generation. Therefore, the 4.0 Ga Sr whole-rock isochron age records a significant event at which time geochemical heterogeneities were established within the originally homogeneous basalt source regions. Types A and C sources were enriched in Rb/Sr, with little or no concurrent modification of 87Sr/86Sr, Sm/Nd, or 143Nd/144Nd. Infiltration of similar-aged KREEP magmas into mantle cumulate sources cannot explain both Sr and Nd isotopic data. Instead, we suggest a metasomatic origin in which Rb, transported as a chloride complex in halogen-rich fluids, was preferentially mobilized relative to Sr and the REEs.