Description: While the mantle roots directly beneath Archean cratons have been relatively well studied because of their economic importance, much less is known about the genesis, age, composition and thickness of the mantle lithosphere beneath the regions that surround the cratons. Despite this knowledge gap, it is fundamentally important to establish the nature of relationships between this circum-cratonic mantle and that beneath the cratons, including the diamond potential of circum-cratonic regions. Here we present mineral and bulk elemental and isotopic compositions for kimberlite-borne mantle xenoliths from the Parry Peninsula and Central Victoria Island, Arctic Canada. These xenoliths provide key windows into the lithospheric mantle underpinning regions to the North and Northwest of the Archean Slave craton, where the presence of cratonic material has been proposed. The mantle xenolith data are supplemented by mineral concentrate data obtained during diamond exploration. The mineral and whole rock chemistry of peridotites from both localities is indistinguishable from that of typical cratonic mantle lithosphere. The cool mantle paleogeotherms defined by mineral thermobarometry reveal that the lithospheric mantle beneath the Parry Peninsula and Central Victoria Island terranes extended well into the diamond stability field at the time of kimberlite eruption, and this is consistent with the recovery of diamonds from both kimberlite fields. Bulk xenolith Se and Te contents, and highly siderophile element (including Os, Ir, Pt, Pd and Re) abundance systematics, plus corresponding depletion ages derived from Re-Os isotope data suggest that the mantle beneath these parts of Arctic Canada formed in the Paleoproterozoic Era, at ∼2 Ga, rather than in the Archean. The presence of a diamondiferous Paleoproterozoic mantle root is part of the growing body of global evidence for diamond generation in mantle roots that stabilized well after the Archean. In the context of regional tectonics, we interpret the highly depleted mantle compositions beneath both studied regions as formed by mantle melting associated with hydrous metasomatism in the major Paleoproterozoic Wopmay-Great Bear-Hottah arc systems. These ∼2 Ga arc systems were subsequently accreted along the margin of the Slave craton to form a craton-like thick lithosphere with diamond potential thereby demonstrating the importance of subduction accretion in building up Earth’s long-lived continental terranes.

Description: Platinum-group elements (PGE), rhenium and osmium isotope data are reported for basalts from Deep Sea Drilling Project cores in the Philippine Sea Plate (PSP). Lithophile trace element and isotopic characteristics indicate a range of source components including DMM, EMII and subduction-enriched mantle. MORB-like basalts possess smooth, inclined chondrite-normalised PGE patterns with high palladium-PGE/iridium-PGE ratios, consistent with previously published data for MORB, and with the inferred compatibility of PGE. In contrast, while basalts with EMII-type lithophile element chemistry possess high Pt/Ir ratios, many have much lower Pd/Ir and unusually high Ru/Ir of 〉10. Similarly, back-arc samples from the Shikoku and Parece-Vela basins have very high Ru/Ir ratios (〉30) and Pd/Ir as low as 1.1. Such extreme Pd/Ir and Ru/Ir ratios have not been previously reported in mafic volcanic suites and cannot be easily explained by variable degrees of melting, fractional crystallisation or by a shallow-level process such as alteration or degassing. The data appear most consistent with sampling of at least two mantle components with distinct PGE compositions. Peridotites with the required PGE characteristics (i.e. low Pd, but relatively high Ru and Re) have not been documented in oceanic mantle, but have been found in sub-continental mantle lithosphere and are the result of considerable melt depletion and selective metasomatic enrichment (mainly Re). The long-term presence of subduction zones surrounding the Philippine Sea Plate makes this a prime location for metasomatic enrichment of mantle, either through fluid enrichment or infiltration by small melt fractions. The Re-Os isotope data are difficult to interpret with confidence due to low Os concentrations in most samples and the uncertainty in sample age. Data for Site 444A (Shikoku Basin) give an age of 17.7+/-1.3 Ma (MSWD = 14), consistent with the proposed age of basement at the site and thus provides the first robust radiometric age for these samples. The initial 187Os/188Os of 0.1298+/-0.0069 is consistent with global MORB, and precludes significant metasomatic enrichment of Os by radiogenic slab fluids. Re-Os data for Sites 446A (two suites, Daito Basin) and 450 (Parece-Vela Basin) indicate ages of 73, 68 and 43 Ma, which are respectively, 30, 17 and 〉12 Ma older than previously proposed ages. The alkalic and tholeiitic suites from Site 446A define regression lines with different 187Os/188Osinitial (0.170+/-0.033 and 0.112+/-0.024, respectively) which could perhaps be explained by preferential sampling of interstitial, metasomatic sulphides (with higher time-integrated Re/Os ratios) by smaller percentage alkalic melts. One sample, with lithophile elements indistinguishable from MORB, is Os-rich (146 pg/g) and has an initial 187Os/188Os of 0.1594, which is at the upper limit of the accepted OIB range. Given the Os-rich nature of this sample and the lack of evidence for subduction or recycled crust inputs, this osmium isotope ratio likely reflects heterogeneity in the DMM. The dataset as a whole is a striking indication of the possible PGE and Os isotope variability within a region of mantle that has experienced a complex tectonic history.