Origin of Mesoarchaean arc-related rocks with boninite/komatiite affinities from southern West Greenland

Abstract

We report whole-rock elemental and Sm–Nd isotope geochemical data from mafic–ultramafic supracrustal rocks from the Nunatak 1390 area in southern West Greenland. Additionally, we report the metamorphic temperature history for these rocks as derived from tourmaline thermometry on a tourmalinite inlier, as well as in situ U–Pb, Hf and O isotopic data from zircons extracted from tonalite–trondhjemite–granodiorite (TTG) gneisses that intruded the mafic–ultramafic sequence.The supracrustal rocks from the Nunatak 1390 area have a minimum age of c. 2900Ma defined by U–Pb zircon ages of cross-cutting aplite sheets of TTG composition. The supracrustal sequence comprises mafic rocks with pillow structures and ultramafic rocks with no evidence of their protolith. They all have amphibolite-facies mineral assemblages and a peak metamorphic temperature of approximately 550°C. The mafic sequence has relatively flat trace element patterns (LaN/SmN of 0.70–2.4) and mostly negative Nb-anomalies (Nb/Nb* of 0.30–1.0) and resembles modern island arc tholeiites. The mafic sequence can be divided into a high- and low-Ti group, where the former group has lower MgO, and significantly higher contents of incompatible elements such as TiO2, P2O5, Zr, Nb and Th. The ultramafic rocks have major and trace element compositions similar to Ti-enriched/Karasjok-type komatiites described in the literature. However, there are no textural indications that the ultramafic rocks from Nunatak 1390 are komatiites sensu stricto.The low-Ti group of the mafic sequence appears to have been derived from a N-MORB source, whereas the high-Ti group and the ultramafic rocks appear to have been derived from a mantle source that is more enriched than the N-MORB source. However, there is no difference in the initial εNd of the mafic and ultramafic rocks. Additionally, assimilation–fractional–crystallisation (AFC) modelling is consistent with this enrichment being caused by introduction of juvenile low-silica adakite (slab–melt) into the mantle source region. Accordingly, we propose that the mafic and ultramafic rocks were derived from a similar type of mantle source, but that the ultramafic rocks were derived from a previously depleted mantle source that was refertilised by slab melts in a subduction zone setting. The high MgO contents of the ultramafic rocks could thus reflect a second stage of partial melting of a refractory mantle in a process similar to that which is suggested for the formation of modern boninites.We propose that the mafic–ultramafic sequence represents an island arc that evolved initially as a juvenile complex (c. 3000Ma). However, inherited zircon grains in aplites and Hf isotope data recorded by the second intrusive TTG phase (c. 2850–2870Ma), show that mixing with older pre-existing crust occurred during this event. Because the regional crust is dominated by TTGs of this younger age, our data suggests that it likely formed by accretion and melting of arcs of different ages and/or contamination of juvenile arcs by pre-existing continental crust rather than entirely by juvenile arc differentiation or melting. Our data thus supports melting of thickened mafic crust in an accretionary setting, rather than direct slab melting, as a mechanism for Archaean crust formation.