On the origin of cratonic `high-mu' isotopic signatures

Some Archean cratons (i.e. Slave, Wyoming) contain Neoarchean granitoids with initial Pb isotopic compositions indicative of derivation from sources characterized by high time-integrated U/Pb ratios (high-mu [1]). Single-stage high-m precursor source reservoir separation from the depleted mantle occurred no later than 3.9 Ga [2]. However, multi-stage separation could have occurred in the Hadean, suggesting that recycling or reworking of Eoarchean/Hadean crust played a significant role in the generation of Neoarchean granitic crust in many cratons. The Sm-Nd system is similar to the U-Pb system in that it has a short-lived parent-daughter pair (146Sm-142Nd) that is sensitive to very early differentiation events, as well as a long-lived parent-daughter pair (147Sm-143Nd) that is sensitive to differentiation throughout all of Earth history. The 103 Ma half-life of 146Sm makes it sensitive only to Sm/Nd fractionation that occurred in the Hadean, providing a useful tracker for very early differentiation events. Indeed, evidence for Neoarchean remelting of ancient crust in another craton has come from analyses of the paired Sm-Nd isotope systems from the Hudson Bay terrane of the northeastern Superior Province. These results indicate that the source of 2.7 Ga Hudson Bay terrane granitoids was Hadean mafic crust, and not Eoarchean felsic crust [3]. Here, we present new data from Neoarchean granites located in the Slave and Wyoming cratons, along with modeling of the dual paired-isotope systems of U-Pb and Sm-Nd to achieve a tighter constraint on the composition of the precursors and the timing of their melting. Combining our newly collected 142Nd data with the high-m signature of these Neoarchean rocks, we evaluate precursor source separation ages along with the source Sm/Nd and U/Pb compositions. In the simplest end-member scenarios, use of the 142Nd system allows us to test whether the cratonic high-mu signature was created by melting of Hadean mafic crust or Eoarchean felsic crust. Differences between these models have major implications for the longevity of mafic crust on the ancient Earth as well as the growth rate and recycling history of the continents. [1] Oversby, 1978, EPSL; [2] Kamber et al., 2003, CMP; [3] O'Neil and Carlson, 2017; Science