Chalcophile and Siderophile Element Abundances in Kilbourne Hole Lherzolites: Distinguishing the Signature of Melt Depleted Primitive Mantle from Metasomatic Overprints

Selenium, tellurium and the highly siderophile elements in peridotites have the potential to illustrate planetary scale processes that are opaque to lithophile elements. However, the interpretation of chalcophile and siderophile element abundances relies heavily on the selection of representative mantle material and the determination of what processes have affected these elements since melt depletion. Whole rock and in-situ sulfide data demonstrate that chalcophile and HSE systematics of the upper mantle could be significantly modified through sulfide-metasomatism, particularly by C-O-H-S × Cl fluids[1] or sulfide melts[2] i.e., chalcophile and siderophile element abundances result from a complex interplay between sulfide addition and alteration of pre-existing sulfide. Here we present new bulk-rock S-Se-Te-PGE abundances on a suite (n = 17) of lherzolite and harzburgite xenoliths from Kilbourne Hole, USA[3, 4]. Mineral modal abundances, major element contents and LREE/HREE ratios for 10 of these xenoliths are consistent with varying degrees of melt depletion (≤ 20 %) whereas the remainder appear to have been affected by cryptic metasomatism, refertilization, or melt-rock interaction which affected lithophile element abundances [4]. While sulfur, Se and PGE budgets are primarily controlled by sulfides, 50 × 30% of Te in peridotite may be accounted for by Pt-Pd tellurides[5]. Although most Kilbourne Hole peridotite xenoliths have PGE characteristics consistent with varying degrees of melt depletion and somewhat scattered Se/Te ratios, KH96-24 has Pt-Pd-Te abundances consistent with Pt-Pd-telluride precipitation, in addition to petrographic evidence for alteration by secondary processes[4]. S/Se are well correlated within the suite. However, lherzolites that retain a strong melt-depletion signature have distinctly lower abundances of both S and Se (<65 ppm and <31 ppm respectively) compared to peridotites that have had their lithophile element budgets perturbed. Moreover, Se/Te increases with decreasing Te in progressively melt depleted xenoliths, suggesting that Te is more incompatible during partial melting (cf. [6][7]). In addition, the peridotites with the highest S and Se abundances may have had their chalcogen budget augmented, either by sulfide melts or by precipitation of sulfide from a S-saturated silicate melt. Critically, the simultaneous addition of S and Se can drive chalcophile element compositions to abundances in excess of those that can be attributed to melt depletion alone. In the absence of additional evidence for secondary processes it can be easy to overestimate the range of chalcophile element abundances purely attributable to melt depletion. Consequently, estimates of primitive mantle composition that include peridotites with augmented chalcophile and / or siderophile element budgets may be misleading. Refs: [1] Alard et al. (2011) J. Petrol. 52 (10), 2009-2045. [2] Alard et al. (2002) Earth Planet. Sci. Lett. 203, 651-633. [3] Harvey et al. (2011) Geochim. Cosmochim. Acta 75, 5574-5596. [4] Harvey et al. (2012) J. Petrol. 53 (8), 1709-1742. [5] Lorand & Alard (2010) Chem. Geol. 278, 120-130. [6] Konig et al. (2012) Geochim. Cosmochim. Acta 86, 354-366 [7] Wang & Becker (2013) Nature 499, 328-331