Constraints on chalcophile element recycling in subduction zones from Selenium isotope systematics

In this cumulative dissertation isotope systematics of the redox-sensitive, chalcophile and moderately volatile element selenium (Se) are used to place further constraints on the recycling of chalcogens in subduction zones. Investigation of Se isotope compositions of low nanogram (ng)-level geological samples is challenging. This is both due to difficulties regarding chemical Se purification from sample matrices and instrumental challenges during mass spectrometric analysis. As a result, only few Se isotope data exist so far for relatively Se-poor samples, such as subduction-related rocks. In comparison, crustal materials such as sediments have several magnitudes higher Se concentrations. Particularly sediments were thus already subject to several Se isotope investigations regarding the mechanisms and the timing of Earth's atmospheric oxygenation. In contrast, little if any work has been dedicated to employ the versatile Se isotopes to study the connection between Earth's surface and interior evolution. By assessing the chalcophile element cycle in subduction zones using Se isotopes, this dissertation takes first steps to close this gap. Given the analytical challenges of Se isotope analyses and the fact that former instruments used to determine Se isotope ratios of igneous rocks (i.e. Micromass IsoProbe – Rouxel et al., 2002, 2004) have become commercially unavailable, a new method is thus required. The first chapter of this dissertation comprises detailed descriptions and advantages of the method developed for the purpose of this study. The presented method allows to use a ThermoFisher Scientific${}^{\circledR }$ NeptunePlus${}^{TM}$ for accurate and precise determination of Se isotope ratios. This involves a double spike to account for isotope fractionation and a hydride generator system for efficient sample introduction. Injection of methane to the plasma source results in two-to three-fold Se signal increase and background signal suppression by a factor of two. This method now allows a systematic investigation of low ng-level geological materials. In the second chapter, the first Se isotope data on island arc lavas from the Mariana arc system are provided. The Se isotope signature of submarine Mariana lavas from the arc and back-arc regions is not affected by degassing and, unlike the elemental Se budget of these lavas, is not affected by magmatic differentiation processes. Hence, the δ${}^{82/76}$Se of submarine arc lavas may retain its source signature. This signature is likely enriched by melt-like and fluid-like subduction components derived from the subducting Pacific crust and overlying sedimentary cover. The Mariana samples show a tendency to become isotopically lighter from the back-arc to arc(-like) lavas, possibly reflecting a decreasing chromatographic effect of the overlying mantle wedge and increasing influence of a fluid-induced signature. Addition of a sediment melt-like subduction component seems to buffer this fluid signature as observed in Mariana arc samples. The large Se isotope compositional range of Mariana lavas is due to the complex contributions of subduction components, but are on average isotopically lighter than the Mariana pre-subduction mantle. Potential slab-derived contributions may be traced back to subduction recycling of isotopically light Se input such as altered sulfide-bearing oceanic crust and modern sediments. Subduction recycling of Se may also have had an impact on the secular Se isotope composition of the Earth's upper mantle given that a considerable shift to lower δ${}^{82/76}$Se average values in sediments has been identified due to the Neoproterozoic Oxygenation Event. Future studies might help to better understand the interplay between Earth's atmospheric oxygenation, Se recycling and isotopic evolution of the mantle through geological time. In the third chapter, new Se isotope data for prograde metamorphic rocks are presented and used to investigate the processes affecting the behavior of chalcophile elements within subducting oceanic crust. Therefore, eclogites, serpentinized peridotites and a metapelite from the Raspas complex, SW Ecuador, were studied. These rocks were subducted to depths of ~60 km and subsequently exhumed as a coherent package. Eclogites have previously been identified to be derived from typical N-type mid-ocean-ridge basalts (N-MORB) that were subject to hydrothermal seafloor alteration. This is also inferred from isotopically light Se isotope compositions that overlaps with those of typical hydrothermally altered basalts. Eclogites influenced by extensive fluid-mobile element enrichment, as identified by low Ce/Pb, Ce/Rb and Ce/Cs, are characterized by heavy Se isotope compositions that approach those of serpentinized peridotites and metapelites. It is thus proposed that, dehydration of subducted oceanic crust and serpentinites leads to fluid release and scavenging of chalcophile elements. Potential sources are hydrothermal sulfides within hydrothermally altered basalt that destabilize during prograde subduction metamorphism. Extensive interaction of slab lithologies with fluids bears the potential to relocate and re-distribute dissolved chalcophile elements within the slab. Further, this provides a realistic mechanism for element transfer to the overlying sub-arc mantle. This could, in principle, explain the Se isotope signature of Mariana lavas (Chapter 2). The results of the three chapters of this dissertation provide evidence that Se isotope systematics are a powerful tool to unravel the link between the behavior of sulfides within subducting oceanic crust and the subduction-related chalcophile element cycle. It is further anticipated that new constraints can be placed on the evolution of Earth's mantle–crust–atmosphere system throughout geological history from the perspective of this novel redox-sensitive and chalcophile isotope system.