A new half-life for 146Sm: Implications for lunar chronology

The timescales and mechanisms of early mantle-crust differentiation on the terrestrial planets and the Moon are often investigated using the 146 Sm- 142 Nd chronometer. Ages derived for the formation of lunar basalt source regions from this system based on the traditional 103 Ma 146 Sm half-life are consistent yielding a weighted average age of 4336 ± 20 Ma that is 40–50 Ma younger than estimates based on crystallization ages of other primordial crystallization products of the lunar magma ocean. To address this issue, we have determined a new 146 Sm half-life of 85.8 ± 0.6 Ma (k = 1) by combining previously reported activity measurements with 146 Sm abundance determinations reported here. The new half-life shifts 146 Sm- 142 Nd ages to older values by several tens of millions of years to 4374 +7.4 / -7.7 Ma. This older age of lunar basalt source formation is now consistent with other independent age estimates of lunar differentiation including ages derived from ferroan anorthosites and model ages for the formation of the late-stage lunar magma ocean differentiation products known as urKREEP. A notable exception, however, is provided by the most ancient ∼4.45 Ma ages determined on a limited number of detrital zircons extracted from lithic and impact melt breccias. If these zircon ages record geologic events, they challenge the validity of the traditional lunar magma ocean model of differentiation which posits that the mare basalt source regions, ferroan anorthosites, and urKREEP are produced during primordial solidification. Under this scenario a mechanism to melt the Moon and produce a global magma ocean must be sought.