| Abstract | The cold, rigid, and melt-depleted mantle underlying Archean cratons plays an important role in the preservation of the overlying continental crust and is one of the main sources of diamonds. However,
with the possible exception of rare earth elements (REE) and platinum group-elements (PGE), the concentrations and host mineral phases for many other critical trace elements within lithospheric mantle remain very poorly understood. Here we address
that knowledge gap, presenting new electron microprobe and laser-ablation inductively-coupled-plasma mass-spectrometry results for a suite of mantle xenoliths (n = 12) and olivine xenocrysts (n = 376) from the Jericho, Muskox, and Voyageur
kimberlites (northern Slave craton, Canada). Low-temperature (<1000?°C) harzburgite xenoliths and olivine xenocrysts suggest that the shallowest portions of the garnet-bearing mantle (less than or equal to 160 km) underlying the northern Slave craton
is chemically depleted and becomes increasing re-fertilized from 160 to 200 km. High-temperature (>1000 °C) garnet and clinopyroxene crystals with Ti/Eu ratios > > 1000, and olivine xenocrysts suggest that interaction with ultramafic silicate melts
is the most likely mechanism to re-fertilize melt-depleted peridotite with incompatible elements toward the base of the lithosphere (~200 km). In contrast, lower temperature garnet and clinopyroxene with Ti/Eu ratios <1000 are more likely related to
metasomatism by carbonatitic melts and/or fluids. Carbonatitic metasomatism is also interpreted as the preferred explanation for the trend of Nb (4?ppm)- and Ta (185 ppb)-rich concentrations of olivine xenocrysts sampled from mid-lithosphere depths
(~140 km). With the exception of a few elements that substitute into the olivine crystal structure during sub-solidus re-equilibration (e.g., Ca, Cr, Cu, Na, Sc, V, Zn), most other olivine-hosted trace elements do not systematically vary with depth.
Instead, we interpret olivine-hosted trace element concentrations that are significantly above the analytical detection and/or quantification limits to reflect trapped fluid (e.g., As, Mo, Sb, Sn), base-metal sulphide (e.g., Ag, Au, Bi, Pd, Pt, Se,
Te), and other mineral inclusions (e.g., U, Th) rather than enrichments of these elements due to substitution reactions or analytical artefacts. We interpret that these inclusions occur in olivine throughout the garnet stability field, but are
relatively rare. As a result, these trapped carbonatitic, proto-kimberlite, and/or other ultramafic silicate melts do not represent a significant source for the suite of trace elements that become enriched to economic levels in the crust. |