Search

Natural carbonatite is associated with various types of deposits, including rare earth elements (REEs) and high-field strength elements (e.g., Nb). Although the spatial relationship between fluorine enrichment and the formation of carbonatite-type deposits is well-established, the underlying mechanism remains unclear. In this study, we investigate the influence of fluorine on the partitioning of trace elements in carbonatite melts (CM) from the mantle to the crust. Elemental partitioning experiments were conducted at 3 GPa and 1100–1400 °C, both with and without fluorine. We determined trace element partition coefficients for the mineral-carbonatitic melt pairs involving the following minerals: clinopyroxene, olivine, spinel, perovskite, magnetite, calcite solid solution, magnesio wüstite. The addition of fluorine influences the types and stability of Ti-rich oxides. Fluorine-bearing systems crystallize magnetite, while fluorine-free systems crystallize perovskite. Compared to fluorine-free compositions, systems with 1.55–2.40 wt% fluorine in the melt exhibited significantly lower (clinopyroxene, spinel, olivine)-melt partition coefficients for REEs and HFSEs by half to one order of magnitude. The decreased uptake of these elements by clinopyroxene, spinel, and olivine may be attributed to changes in their speciation within the melt phase, consistent with the formation of REE-F complexes. Using our partition coefficients, we simulated the differences in trace element characteristics between carbonatite melts derived from the mantle under fluorine-bearing and fluorine-free conditions. The fluorine-bearing carbonatitic melt extracted from mantle peridotite displayed significantly higher trace element compositions compared to fluorine-free carbonatitic melt, with trace element concentrations being 5–10 times higher. We simulated the trace element characteristics of carbonatitic melts released from subducted sediments and ocean crusts and found that these ocean crusts or sediments can release fluorine-rich carbonatite melts that are naturally enriched in REE due to their high fluorine and REE contents. Therefore, the subduction of sediments or the recycling of ocean crusts is highly favorable for the formation of carbonatite-hosted REE deposits. This can explain why the majority of carbonatite-hosted REE deposits are typically found in continental margins with subducted materials signals in source region. Furthermore, we modeled the impact of different mineral crystallization and fractionation in the shallow crust on the trace element compositions of carbonatitic melts. Late-stage magma crystallization and differentiation of silicate and carbonate minerals are favorable for the further enrichment of ore-forming materials (e.g., REEs). [ABSTRACT FROM AUTHOR]

Seafloor abyssal peridotites are products after partial melting and melt-rock interaction, providing constraints on upper mantle processes and heterogeneity. Despite extensive studies of ophiolitic peridotites and a burgeoning dataset for dredged abyssal peridotites, limited studies have been undertaken on abyssal peridotite drill core samples. This omission means a lack of comparison between modern ocean mantle lithosphere and that of ophiolites, where field context is possible in the latter, but not the former. Ocean Drilling Program (ODP) Leg 153 near the Kane fracture zone, and ODP Leg 209, near the Fifteen-Twenty fracture zone, both on the mid-Atlantic Ridge but ∼1000 km distant from one another, recovered serpentinized mantle peridotites. We report whole-rock major-, trace- and highly siderophile-element (HSE: Re, Pd, Pt, Ru, Ir, Os) abundances and 187Os/188Os ratios of peridotites from three cores (maximum depth of 200.8 m) from the two Legs to explore mantle heterogeneity within the Atlantic Ocean basin. Peridotites from ODP Leg 153 are relatively fertile, with Al 2 O 3 contents ranging from 1.1 to 2.7 wt%. They are characterized by relatively radiogenic 187Os/188Os (0.1259 ± 0.0017; 2 S.D.) and uniform HSE abundances. The ODP Leg 209 peridotites are highly refractory, with Al 2 O 3 contents ranging from 0.1 to 0.8 wt%, and imply that they record an extensively depleted mantle. They have 187Os/188Os of 0.1207 ± 0.0023 (2 S.D.) and fractionated HSE abundances, indicative of depletion occurring > 1 billion years ago. Marked heterogeneity both in the extent and timing of depletion are therefore evident along a large segment of the mid-Atlantic ridge that produces zero-age basalts. Measurement of Os isotopes in spatially associated mid-ocean ridge basalts (MORB) is challenging due to their sub ng/g Os contents. Instead, Nd isotopes of MORB are used to show that they are unlikely to have formed directly from abyssal peridotite-like sources. Abyssal peridotites therefore represent long-lived refractory residues during mantle upwelling with the sources of MORB being more enriched. Comparison between drilled abyssal peridotites and ophiolitic mantle reveals larger compositional variations in the latter. The overall higher melting extent in ophiolite mantle is likely reflective of a more complex life cycle than the modern oceanic lithosphere. [ABSTRACT FROM AUTHOR]

We apply the stable potassium isotope system (41K/39K) to well-studied Mariana arc lavas, in which inter-island geochemical variability has been interpreted to reflect near constant addition of an aqueous fluid flux, that dominantly samples the subducting mafic oceanic crust and variable amounts of sediment melt addition to the sub-arc mantle wedge. The nature of the sediment component in the Mariana arc lavas remains enigmatic, with a mixture of melts from all subducted sediment lithologies and the altered-mafic oceanic crust seeming likely, but some interpretations point towards a component from melting of the volcaniclastic sediment alone. We present K isotopic data on a set of well-characterised Mariana arc lavas and sediment samples (from Ocean Drilling Program site 801). Our data show that the majority of Mariana arc lavas are isotopically heavy (ẟ 41K), by up to ∼0.2 ‰, relative to mid-ocean ridge basalt (MORB), but most of the subducting sediment samples are slightly isotopically lighter than, or indistinguishable from MORB. The volumetrically important volcaniclastic sedimentary unit however is significantly isotopically lighter than MORB, by ∼0.8 ‰, which reflects marine diagenetic processes. Thus, volcaniclastic material significantly influences the bulk sediment K isotope composition. We show that ẟ 41K compositions of the Mariana arc lavas can be reproduced by the addition of an aqueous fluid with isotopically heavy K (relative to MORB) and an additional fraction of an incompatible-element-enriched isotopically light melt component. Modelling of K isotopes together with K/La and radiogenic Nd indicate that the melt component is best explained by a mix of melts from bulk sediment and altered-mafic oceanic crust. Our results show that distinctive K isotopic variations in subduction zone inputs and K isotopic fractionation during dehydration reactions makes K a useful tracer of subduction zone processes. [ABSTRACT FROM AUTHOR]

The alteration of ocean crust through hydrothermal seawater circulation facilitates chemical exchange between Earth's surface and interior. Hydrothermal alteration leads to uranium (U) removal from seawater and net U uptake by the ocean crust, particularly during low temperature alteration that occurs on the vast ocean ridge flanks away from the spreading axes. Determining the timescales of U uptake and its associated 238U/235U signature has important implications for understanding U exchange processes during subduction and recycling into the mantle. Here we study the U systematics of ∼15 million year-old ocean crust drilled at Site 1256 on the eastern flank of the East Pacific Rise. Analysis of cores from the upper ∼1300 m of intact ocean crust at this site, reveal large variability in U concentrations and 238U/235U ratios. Many of the samples from the upper ∼600 m of extrusive lavas have elevated U concentrations and 238U/235U ratios lower than seawater, consistent with mechanisms of U uptake under relatively oxidised conditions. Samples from the underlying sheeted dikes and gabbros show evidence for hydrothermal U mobilisation, but negligible net U uptake. In contrast, in the transition zone between the extrusive lavas and the sheeted dikes, samples revealed large U enrichments and high 238U/235U ratios above seawater. This is consistent with uptake of the reduced U+4 species under relatively reducing conditions from seawater-derived hydrothermal fluids. In addition, large secular disequilibrium in 234U/238U ratios from samples in the lava-dike transition and upper sheeted dikes give evidence for U mobility within the last ∼1.5 million years, likely driven by deep channelled flow of seawater-derived hydrothermal fluids combined with preferential leaching of 234U from the rock matrix. Both the total estimated U uptake and mean 238U/235U at Site 1256 is lower than similar estimates from significantly older (>100 million years) altered ocean crusts at drill Sites 801 and 417/418. This shows the variable total U uptake and 238U/235U ratio in altered ocean crust over time, which needs to be taken into consideration when estimating global U budgets. [ABSTRACT FROM AUTHOR]

Potassium (K) isotopic compositions for a suite of crustal rocks (plutonic unit, sheeted dike complex, and volcanic sequence) in the Troodos ophiolite, Cyprus, were determined to investigate fluid-transport processes during subduction inception. The gabbros and plagiogranites in the plutonic unit have restricted K isotopic compositions (0.01–0.20 ‰ and −0.17 to 0.02 ‰, respectively), whereas diabases in the sheeted dike complex have variable δ41K values from −0.24 to 1.13 ‰. The δ41K values of the volcanic sequence range from −0.26 to −0.07 ‰ in the lower pillow lavas, and from −0.71 to 0.34 ‰ in the upper pillow lavas. These δ41K values are clearly distinct from those of subducting sediments and altered oceanic crust, ruling out direct crustal input as the cause of the unusually heavy K isotopic compositions. Rather, the high K/Th, K/U, and Sr/Th ratios of the investigated samples, together with binary mixing modelling, reveal that the anomalously heavy K isotopic compositions in Troodos rocks are attributed to the influx of fluids during subduction inception. Notably, the δ41K values in the crustal rocks of the Troodos ophiolite are more variable and extend to higher values than those of arc-front and rear-arc lavas of the Izu arc and arc lavas of Martinique in the Lesser Antilles, indicating continuous K isotope fractionation during slab dehydration. Thus, the heavy K isotopic compositions of the Troodos ophiolite reflect K isotope fractionation during dehydration of a shallow slab within a newly formed subduction system, where heavy K isotopes preferentially enter into the aqueous fluids. Later, progressive breakdown of K-bearing phases at deeper levels results in isotopically lighter hydrous melts, which can be delivered to the arc. [ABSTRACT FROM AUTHOR]

Understanding the subduction-driven recycling of economically important chalcophile metals (e.g., Cu, Au, Ag, and platinum-group elements, PGEs) and volatiles (e.g., sulfur) is critical for assessing models of ore formation. However, whether the sulfide remains stable in recycled altered oceanic crust (AOC) in the mantle is unclear, and the influence of AOC-derived metasomatic agents on the extraction efficiency of chalcophile elements from the deep mantle remains poorly understood. Here we report whole-rock Cu-Ag-Au-Re-PGEs of Cenozoic alkaline basalts at Hannuoba, eastern China, which were derived from the interaction of carbonated eclogite (recycled AOC)-derived melts with lithospheric peridotites. Volatile-rich primitive melts from carbonated eclogite are poor in strongly chalcophile metals (e.g., ∑PGE < 1 ng/g, Au < 1 ng/g, Cu < 30 μg/g). Remarkably, primitive melts show significant fractionation of elements with variable chalcophile affinities, such as lower Cu/Ag ratios than those in primitive mantle and MORBs. These features reflect the sequestration of these metals in residual mono-sulfide solid solution (MSS) during low-degree melting of the AOC in the deep mantle. Given the retention of sulfides in the AOC, partial melts derived from it are rich in sulfur but poor in strongly chalcophile metals. The Hannuoba alkaline melts scavenged some strongly chalcophile metals from peridotites (especially PGEs) during their ascent. However, this did not necessarily lead to the formation of chalcophile metal-enriched melts (e.g., Au < 1 ng/g and Cu < 60 μg/g). These processes appear to be common in eastern China, as evidenced by the low abundances of Cu and Au found in many primitive intraplate alkaline basalts. We propose that strongly chalcophile elements in subducted AOC are efficiently transported into, and stored in, the deep mantle, with a small proportion returning to the exosphere through volcanism. Moreover, the chalcophile metal-poor and volatile-rich melts of subducted AOC could also explain the addition of abundant sulfur (secondary sulfides), but limited chalcophile metals to the metasomatized mantle. [ABSTRACT FROM AUTHOR]

The petrogenesis of basaltic arc lavas with heavy Mo isotopic compositions is controversial. Given that oceanic crust preferentially loses isotopically heavy Mo during subduction at forearc depth and the remaining Mo is isotopically light and locked in rutile, fluids derived from such crust at subarc depth cannot account for the heavy Mo isotopic compositions of basaltic arc lavas. Serpentinites can contain substantial Mo and volatiles and are a potential source of isotopically heavy Mo in basaltic arc lavas. However, the Mo isotopic composition of serpentinites and the Mo isotope behavior during serpentinite dehydration are poorly constrained. Here, we present Mo contents and isotope data for serpentinites from the South Sandwich Trench-Fracture Zone intersection, and deserpentinized peridotites and high-pressure meta-serpentinite veins from the western-central Alps and Betic Cordillera, Spain. Most of these samples have heavy Mo isotopic compositions (δ98/95Mo NIST 3134 , ranging from −0.26 ‰ to 1.84 ‰) with low Ce/Mo (0.10–23.98) and Ce/Pb (0.17–3.59) ratios, which are distinct from the compositions of eclogite-facies metabasalts and metapelites. In addition, no systematic variations in Mo contents and Mo isotopic compositions occur for the studied samples, indicating limited Mo isotope fractionation during serpentinite dehydration. The above results suggest that serpentinites and their derived fluids at subarc depths can serve as the main isotopically heavy Mo sources for basaltic arc lavas. The Mo isotope data with Ce/Mo and Ce/Pb systematics suggest that serpentinite-derived fluids or melting of serpentinite diapirs at subarc depths can explain the heavy Mo isotopic compositions of basaltic arc lavas. Furthermore, recycled oceanic slab (oceanic crust + serpentinized mantle) would has limited influence on the Mo systematics of the mantle, which explains why the mantle has maintained a nearly constant Mo isotopic composition since 3.5 Ga. [ABSTRACT FROM AUTHOR]

It is well established that Indian- and Pacific-type mantle domains observed in the present-day upper mantle are both long-lived and can be traced back to at least the Early Paleozoic, with their contrasting isotopic compositions inherited from mantle beneath the Tethys and Panthalassic oceans, respectively. But, the long-term evolution of the Tethyan-Indian oceanic mantle remains loosely constrained due to sparse sampling of the Tethyan oceanic lithosphere, which makes it unclear if there has been temporal variation of its diagnostic DUPAL anomaly throughout the Phanerozoic, and what caused any changes. To address these issues, we report new geochemical data, including bulk-rock major and trace elements and Sr-Nd-Hf-Pb isotopes, for samples from Tethyan ophiolites located in the Tibetan Plateau. These results, combined with existing data for ophiolites across the Tethyan domain and Indian oceanic crust, demonstrate that the DUPAL anomaly in the Tethyan-Indian oceanic mantle is long-lasting, but its strength is weakening through time. The DUPAL anomaly in the Tethyan-Indian upper mantle was most pronounced in the Paleozoic with a homogeneous Indian-type asthenosphere underlying the Proto/Paleo-Tethys oceans, but has become less obvious since the Mesozoic with an isotopically heterogeneous asthenosphere beneath the Neo-Tethys and Indian oceans containing Pacific-type mantle fragments. We propose that the declining DUPAL anomaly in the Tethyan-Indian oceanic mantle since the Mesozoic was most likely associated with inflow of Pacific-type mantle during breakup of Pangea and Gondwana, when interaction between Indian- and Pacific-type mantle was more efficient. On the other hand, Indian-type asthenosphere might have been squeezed out into the Panthalassic-Pacific oceanic mantle during closure of the Tethys Ocean. This transfer may have contributed, at least in part, to the DUPAL isotopic anomaly detected in the Panthalassic-Pacific oceanic mantle. Our findings indicate that tectonically-induced asthenospheric flow permits interactions between Indian- and Pacific-type mantle domains and contributes to evolving mantle compositions. [ABSTRACT FROM AUTHOR]

Metasomatism of the sub-continental lithospheric mantle (SCLM) can result in enrichment in volatile elements introduced via subduction of seawater-altered oceanic lithosphere. Subduction-related metasomatism can introduce chlorine isotope heterogeneity to the SCLM inherited from the subducted source, which can be used to characterize the source of metasomatism. This study presents Cl isotope compositions (δ37Cl) of variably metasomatized continental lithospheric mantle xenoliths from the Colorado Plateau and southern Rio Grande Rift (southwestern United States). Fertile mantle xenoliths that represent asthenosphere that has recently been emplaced at the base of the North American lithosphere have δ37Cl values of 0.0 ‰ to +0.1 ‰, suggesting that the convecting mantle has an average δ37Cl value similar to seawater and chondrites. In contrast, metasomatized xenoliths span a range of δ37Cl values from +0.1 ‰ to +1.9 ‰, extending above the average convecting mantle value. The range in δ37Cl values can be explained by mixing between a subduction-related fluid with δ37Cl values between +1.0 ‰ and +1.9 ‰ and a shallow crustal fluid with a δ37Cl value less than +0.3 ‰. The most likely source of the high δ37Cl subduction-related signature is the lower crust in the subducted Farallon plate. Lower crustal Cl likely entered the SCLM either through direct dehydration of lawsonite ± phengite in the deeply subducted Farallon plate or from a mixed serpentinite-oceanic crust source in the subducted Farallon. Regardless of the mechanism, this study provides evidence for subduction of crustal Cl into the mantle, which can introduce isotopic heterogeneity into the SCLM and has implications for the volatile cycling in the HIMU mantle and SCLM. [ABSTRACT FROM AUTHOR]

Understanding deep nitrogen (N) cycling better requires investigating the delivery of N to subduction systems via various lithologies. Input to subduction zones through mafic rocks is more voluminous and massive as compared to sedimentary rocks which calls for a thorough investigation of the behavior of N in metabasic rocks. Here we estimate the delivery of N to subduction zones at forearc depths by investigating the geochemistry of amphibolites and epidote-blueschists from the Central Qiangtang Metamorphic Belt in Tibet where the metabasic rocks likely represent the transition from oceanic to continental subduction. The rocks contain 21–147 ppm N with δ15N values from +1.8 ‰ to +10.0 ‰, and 147 ppm N is the highest that has been reported in a metabasic rock thus far. Given the N abundances for most of the rocks are much higher than those of altered oceanic crust (i.e. basalts, sheeted dikes and gabbros; 6.0 ± 4.7 ppm), the N is likely neither magmatic nor was introduced in the rocks during hydrothermal alteration prior to subduction. This is confirmed by the K 2 O/Th versus Ba/Th, Th/U versus Th and Ba/Rb versus K 2 O plots where these rocks align with the trend of metamorphic fluid alteration rather than seafloor hydrothermal alteration. A two-step process led to N acquisition in the metabasic rocks. In the first step, the metabasic rocks acquired their N from metasediment-derived fluids during metamorphism in the subduction channel. In the second step, some of the metabasic rocks underwent N loss and concomitant enrichment in 15N due to devolatilization within the subduction channel. We modeled the N fluxes at forearc depths in 55 modern-day subduction zones via metasedimentary and metabasic rocks assuming their minimum, median and maximum N concentrations to assess their relative importance in delivery of N to subduction zones. We find that metabasic rocks supply comparable fluxes of N to metasedimentary rocks at forearc depths, even though metasedimentary rocks have at least an order of magnitude higher N abundance than metabasic rocks. The importance of metabasic rocks in N delivery reinforces the need to investigate the behavior of N in them from more locations globally to improve our understanding of deep N cycling. [ABSTRACT FROM AUTHOR]

The Mascarene Islands in the western Indian Ocean, encompassing La Réunion, Mauritius, and Rodrigues, are the recent (<10 Ma) surface expressions of the Réunion hotspot. Ocean island basalts (OIB) from these islands exhibit a remarkably homogeneous long-lived radiogenic isotopic composition, coinciding with the convergence field of many global OIB trends in the mantle array. Réunion plume-related OIB therefore provide one of Earth's most pristine representations of this "focal zone" component, which may have a primordial heritage. Besides this signature, Mascarene lavas have been suggested to retain contributions from sources with distinct compositions, including: (1) Archaean-aged zircons assimilated from continental crust within the oceanic lithosphere by trachytic magmas from Mauritius; (2) more deeply recycled continental crust components preserved by elevated 87Sr/86Sr and 208Pb/206Pb in lavas from the Piton des Neiges volcano of Réunion; and (3) an isotopically depleted mantle component resulting from interaction with Central Indian Ridge material. In this study we use Sr-Nd-Pb isotope systematics, along with major and trace element compositions of basaltic lavas from all three Mascarene Islands to investigate the relationship of their sources to well-characterized mantle endmembers. Among the Mascarene Islands, Rodrigues lavas are the most enriched in highly incompatible elements, likely reflecting shallower and lower degrees of partial melting than Réunion or Mauritius. Combined Sr-Nd-Pb isotope compositions indicate that lavas from the Older Series of Mauritius resemble those from Réunion, whereas lavas from the Younger and Intermediate Series, together with Rodrigues, are consistent with contributions from an isotopically depleted component. In addition, the Pb isotopic compositions of Rodrigues samples require an additional contribution from a component with a long-term enrichment in its Th/U ratio. Based on isotope mixing models, direct assimilation of continental crust embedded within the oceanic lithosphere is unlikely to account for the Pb isotopic variation of Rodrigues. A metasomatized mantle component, previously envisaged as a "fossil" Réunion plume near the Central Indian Ridge, is partially able to reproduce the trace element signature, but not the observed Sr-Nd-Pb isotopic compositions of Rodrigues. Instead, small proportions (<5%) of an EM1-like component provide a preferred endmember. The composition and origin of this component, which is exclusively reflected within Rodrigues lavas, may constitute a new geochemical feature of the Mascarene Islands and is consistent with geodynamical predictions that small-scale enriched mantle domains may be widespread in the source regions of hotspot volcanoes. [ABSTRACT FROM AUTHOR]

Compositions of arc magmas depend on several factors and are often thought to reflect conditions in the mantle wedge and at the slab-mantle interface. However, in continental arc settings, magmas are also influenced by assimilation of continental crust. Here, we present measurements and modeling of 238U-230Th activity ratios, Sr, Nd, Hf, and Pb isotopic compositions, and major and trace element concentrations in young, historic lavas erupted from Reventador, an active stratovolcano in the Ecuadorian Andes. In arc lavas, 238U-230Th disequilibria are often assumed to reflect processes occurring in the mantle wedge, as U and Th behave differently in this relatively oxidized and fluid-rich environment. Enhanced mobility of hexavalent U in aqueous fluids results in (230Th/238U) < 1 and elevated (238U/232Th), which are common in arc lavas. However, the majority of Reventador lavas have (230Th/238U) = 1.0–1.1 and (238U/232Th) = 0.94–1.12, the latter of which is considerably lower than the depleted mantle ((238U/232Th) ∼ 1.5, Sims and Hart, 2006). While this Th enrichment could be due to melting of subducted oceanic crust, approximately linear trends between (230Th/232Th), wt. % SiO 2 , and radiogenic isotope ratios indicate otherwise. We argue that crustal assimilation lowers long-term, or time-integrated, (238U/232Th) in Reventador magmas. To quantify the effects of assimilation we modeled stepwise assimilation and fractional crystallization. Observed trends between (230Th/232Th), 87Sr/86Sr, ε Nd , ε Hf , and 208Pb/206Pb can be reproduced by up to 10% assimilation, which is consistent with previous regional studies. However, reproducing the full spectrum of isotopic diversity among Reventador lavas requires heterogeneous basalt compositions. In a broader context, this study emphasizes the need to consider crustal processes when examining continental arcs. While 238U-230Th disequilibria develop in the mantle wedge, they can be overwritten by subsequent interaction with continental crust. [ABSTRACT FROM AUTHOR]

Seafloor hydrothermal vents represent potential sources of Mo and other biologically relevant transition metals to the global ocean, complementing continental runoff. Here, we use a combination of experimental, theoretical, and field-sampling approaches to investigate the behavior of Mo in basalt-hosted seafloor hydrothermal systems to provide insight into the processes controlling Mo concentrations in hydrothermal fluids and to derive estimates of vent fluid Mo concentrations and fluxes. Results of this study demonstrate that reaction fluids generated from 350 °C, 500 bar hydrothermal basalt alteration experiments contain 775–801 nmol/kg Mo and are thus comparable to a recently collected time series of natural seafloor vent fluids that contained 200–220 nmol/kg Mo at 302 °C and 29–30 nmol/kg at 281–282 °C (Evans et al., 2023). Synchrotron-based analyses of experimentally altered basalt produced in this study and additional natural samples of altered oceanic crust originally collected from Pito Deep Rift reveal the presence of Mo-rich particles consistent with trace molybdenite. Comparisons of Mo:Cu ratios in natural vent fluids and near-vent sediment trap samples from Main Endeavour Field indicate that vent fluid Mo is readily incorporated into buoyant plume particles and advected out of the near-vent field, analogous to previous mass balance studies of Cu in this region. Thermodynamic calculations of molybdenite solubility in the context of mineral-buffered hydrothermal fluids and comparisons with natural and experimental hydrothermal fluids suggest that high-temperature vent fluids contain 30–1500 nmol/kg Mo. While a minor component of the modern Mo budget, hydrothermal Mo fluxes are estimated to have constituted 0.3–200× the contemporaneous continental weathering fluxes prior to the ∼2.4 Ga ago "Great Oxidation Event" and widespread oxidative continental weathering. Overall, identification of hydrothermal vents as a source of Mo-rich plume particles with potential for dispersal into the wider marine environment has significant implications for hypotheses regarding the co-evolution of Life and Earth's environments, specifically the form and availability of Mo in anoxic Archean-Eon oceans, where Mo-dependent enzymatic pathways are thought to have emerged and subsequently evolved. [ABSTRACT FROM AUTHOR]

Uptake of nitrogen (N) during hydrothermal alteration of oceanic crust makes the altered oceanic crust (AOC) a significant N reservoir in parallel to seafloor sediments. While N enrichment in the upper (basaltic) oceanic crust during low-temperature alteration has been widely observed, the N characteristics of moderate- to high-temperature altered gabbroic oceanic crust, which accounts for ∼70 vol% of the crustal material in subducting slabs, have been rarely studied. The lack of these data resulted in large uncertainties in calculating the N input flux of global subducting slabs and modeling the N quantities released through the arc and subducted into the deep mantle. To fill this gap, we examined the N concentrations and isotope compositions of 38 altered gabbroic rocks recovered by ODP/IODP drillings from three oceans, i.e., Hole 735B at the Atlantis Bank in the Southwest Indian Ridge, Hole 1309D at the Atlantis massif in the Mid-Atlantic Ridge, and Hole 1415P at the Hess Deep in the East Pacific Rise. All the gabbroic samples show significant N enrichment with mostly positive δ15N values (in average: 7.4±3.8 ppm and +0.6±2.0‰ for Hole 735B, 5.3±2.1 ppm and +1.4±1.3‰ for Hole 1309D and 7.9±1.9 ppm and +2.0±1.8‰ for Hole 1415P), which are comparable to those of altered basalts from global oceanic crust. The N concentrations and δ15N values of these gabbroic rocks can be readily explained by mixing between minor inherited mantle N and mainly secondary N derived from seawater. The secondary N-hosting minerals in altered gabbroic rocks are likely plagioclase, amphibole and chlorite formed during moderate- to high-temperature alteration stages. Using these new data, we obtained a global nitrogen input flux of 20.6 - 1.3 + 0.9 × 109 mol·yr−1 to 29.9±2.2 × 109 mol·yr−1 for subducting AOC, in which 50% – 64% is contributed by the altered gabbroic oceanic crust. Integrating with previous estimate of N input flux from seafloor sediments, a total N input flux of 74.9 - 1.3 + 0.9 – 84.2±2.2 mol·yr−1 was estimated for global subducting slabs. Mass balance between this input N flux and various estimates of N output flux in global arcs gave a large range (from 47% – 53% to 10% – 20%) for the fraction of slab N to be subducted beyond the sub-arc depth to the deeper mantle. The upper end (47% – 53%) of this range is more consistent with the results from other constraints, such as thermal structural control and comparison between altered gabbros and meta-gabbros. [ABSTRACT FROM AUTHOR]

Eclogitic diamond formation can be associated with interactions between subducted slabs and ambient mantle. To gain a better understanding of the extent of chemical exchange between slabs and ambient mantle and the mechanisms of coeval diamond formation, we examined 16 diamonds with eclogitic garnet inclusions from the Koidu kimberlite complex in Sierra Leone (West African Craton). We analyzed the 16 garnet inclusions for major element, trace element and oxygen isotope (δ18O) compositions and their host diamonds for carbon (δ13C) and nitrogen (δ15N) isotope compositions. The garnet inclusions have δ18O values ranging from +5.4 to +12.1‰ (median = +11.3‰) and all but two have δ18O values ≥+9.9‰. Such high δ18O values indicate a link to protoliths that had undergone extensive low-temperature alteration by seawater, which occurs in the uppermost basaltic layer of oceanic crust, prior to subduction. Diamonds hosting the high δ18O garnets have a crustal δ13C signature (−29.6 to −19.4‰) paired with δ15N values (−5.6 to +1.3‰, with one outlier at +9.9‰) implying typically only minor 15N enrichment. This apparent decoupling of δ18O and δ13C from δ15N indicates diamond formation in an eclogitic substrate where garnet oxygen isotope compositions are inherited from altered low-pressure protoliths, diamond carbon is principally derived from biogenic carbonate and organic matter, and diamond nitrogen is variably added by an external, mantle-derived fluid that prompted diamond formation. Of the two remaining garnet inclusions, one has mantle-like δ18O (+5.4‰) and a positive Eu anomaly, suggesting derivation from gabbroic protoliths originally located deep in the oceanic crust where significant alteration by seawater did not occur. Based on mantle-like δ13C (−4.7‰) and δ15N (−6.9‰), the associated diamond formed from mantle-derived fluids/melts. Relative to the other Koidu garnet inclusions, the final garnet has a small majorite component (formation pressure ∼8 GPa), high Mg# (79.0), elevated Cr# (0.90), and low δ18O (+6.3‰), suggesting encapsulation during infiltration of slab-derived melts into surrounding asthenospheric peridotite, associated with a high degree of chemical exchange between slab- and mantle-derived components. [ABSTRACT FROM AUTHOR]

Stable barium (Ba) isotopes are emerging as tracers for the recycling of crustal material into the mantle. Small but significant Ba isotope variations have been found in global MORB (δ 138/134Ba values of −0.04‰ to 0.15‰), but the processes causing these Ba isotope variations remain poorly understood. In addition, uncertainties still exist in the estimate of the Ba isotope composition of the depleted upper mantle. Here, we present a systematic study of Ba isotopes for well-characterized MORB glass samples from the South Mid-Atlantic Ridge (SMAR) between 5 and 11°S, which span a wide range of radiogenic isotope ratios and trace element contents. Our results show that the northernmost segment A0 basalts with highly depleted radiogenic isotope compositions have δ 138/134Ba values ranging from −0.02 to 0.05‰. In comparison, the δ 138/134Ba values of the isotopically enriched basalts from the A1-A4 segments vary from 0.02 to 0.11‰. Thus, while small, the enriched components involved can be observed to modify the Ba isotope composition of the mantle source beneath A1-A4 segments. Combining our new measurements with literature data, this study places new constraints on the Ba isotope composition of the depleted upper mantle, as well as the origin of Ba isotope heterogeneity in MORB. The Ba isotope composition of the depleted upper mantle in the absence of recycled components is estimated to have a δ 138/134Ba value of 0.03 to 0.05‰, which can be adopted as the baseline for using Ba isotope ratios as a tracer of mass transfer processes between the crust and mantle. In comparison, global MORB have δ 138/134Ba values ranging from −0.04‰ to 0.15‰. No global correlation is observed between Ba isotope ratios and geochemical parameters that record mantle enrichment, indicating that the Ba isotope variations in MORB cannot be simply ascribed to the addition of some single component, such as sediment. Models for melting and mixing between recycled materials and depleted mantle suggest that the variable Ba isotope compositions of the enriched MORB require a contribution from recycled altered oceanic crust and crustal sedimentary materials. Therefore, these recycled components both play important roles in the chemical budget of the convecting upper mantle, especially for incompatible elements such as Ba. [ABSTRACT FROM AUTHOR]

High-temperature hydrothermal alteration of oceanic crust is one of the two main sources of potassium (K) to the oceans, with modern flux estimates ranging roughly from ∼8 to 30 % of the K flux from rivers. Despite the role of high-temperature hydrothermal fluids in the global seawater K budget, little is known about its effect on the K isotopic composition of the oceans. Here we present stable K isotope measurements (δ41K SRM3141a) of globally distributed high-temperature hydrothermal fluids from three mid-ocean ridge systems: the East Pacific Rise (n = 21), the Juan de Fuca Ridge (n = 34), and the Mid-Atlantic Ridge (n = 18). We find a strong correlation between δ41K and Mg/K ratios, consistent with conservative mixing between a high-temperature hydrothermal fluid (i.e., Mg = 0) and seawater as fluids ascend to the seafloor and/or due to seawater entrainment during sampling. The δ41K of end-member hydrothermal fluids is found to range between −0.80 ‰ and 0.07 ‰, with an average value of −0.36 ± 0.30 ‰ (2σ, n = 38). Most (∼76 %) of the variability in end-member fluid δ41K compositions observed here can be explained by high-temperature fluid-rock K exchange, with little (∼0.2 ‰) to no K isotope fractionation between hydrothermal fluid and altered crust. Larger deviations from the average end-member hydrothermal fluid value are likely to result from processes other than high-temperature fluid-rock exchange, such as (1) low-temperature hydrothermal reactions during fluid recharge, (2) reaction of fluids with local sedimentary sources, and (3) phase separation. The K contents of end-member fluids vary considerably, from ∼1 to 38 mM, thus a K-weighted average of −0.37 ± 0.24 ‰ (2σ, n = 38) is estimated to represent the δ41K composition of the global hydrothermal K flux. Our results suggest that K sourced from axial hydrothermal alteration does not contribute to the elevated 41K/39K of seawater compared to bulk silicate Earth (BSE). In addition, subduction of oceanic crust altered under high-temperature conditions is unlikely to be a significant source of K isotopic heterogeneity to Earth's mantle. [ABSTRACT FROM AUTHOR]

Subduction is a key element in the carbon cycle between the Earth's surface and deep interior. In addition to pressure (P) and temperature (T), oxygen fugacity (ƒO 2) is an intensive variable that can change the speciation, stability, and mobility of carbon-bearing phases in subduction zones. However, tracking redox evolution and associated carbon speciation in high-pressure metamorphic rocks from subduction zones remains challenging. In this study, we identified CH 4 -rich fluid inclusions in ultra-high-pressure (UHP; i.e., coesite-bearing) eclogites (the UHP-CH 4 eclogite) and CO 2 -rich fluid inclusions in high-pressure (HP; i.e., quartz-bearing) eclogites (the HP-CO 2 eclogite) from the Chinese southwestern (SW) Tianshan palaeosubduction zone, respectively. Combining the garnet-clinopyroxene oxybarometry and comprehensive thermodynamic calculations, we revealed that the ƒO 2 of the UHP-CH 4 eclogite decreases to a minimum of ∼FMQ − 3 at peak P–T conditions (2.8 GPa, 525 °C) and CH 4 is expected to form in graphite-saturated C H O fluids. Whereas during the early-exhumation stage (2.5 GPa, 600 °C), the UHP-CH 4 eclogite is predicted to buffer higher ƒO 2 conditions of ∼FMQ − 0.9. In contrast, the HP-CO 2 eclogite, which has a higher bulk CO 2 content (∼10 wt%) and elevated Fe3+/ΣFe ratio (0.24) than the UHP-CH 4 eclogite, suggesting a stronger oxidative seafloor alteration of the HP-CO 2 eclogite protolith before subduction, stabilizes graphite and CO 2 -rich fluids at peak ƒO 2 conditions between FMQ and FMQ + 1 (2.6 GPa, 500 °C). These data demonstrate that the thermal structure (P–T conditions) of the subduction zone and the composition of the slab are first-order controls on the redox evolution of subducting slabs and the changing speciation of carbon in slab fluids buffered by the metamorphic rock system. Therefore, we propose that the cold subduction of a weakly altered oceanic crust (e.g., UHP-CH 4 eclogites from the SW Tianshan) provides ideal conditions for the formation of ultra-deep abiotic CH 4. [ABSTRACT FROM AUTHOR]

The compositions of ocean island basalts (OIB) can record signatures of ancient subducted crust that has been recycled through the mantle. Based on their radiogenic isotope, major and trace element compositions, several geochemically-extreme mantle compositions are suggested to host components of recycled crust. Barium (Ba) stable isotopes can serve as a tracer of crustal recycling into the mantle sources of OIB because 1) Ba isotopes are fractionated in surface crustal reservoirs and 2) Ba is highly enriched in crust relative to the mantle. Therefore, the Ba isotopic compositions of the subducted crust could be detected in OIB. Here, we characterized Ba isotopic compositions of fresh OIB from hotspots sampling a range of mantle compositions. We first explore δ 138/134Ba (permil deviation of the 138Ba/134Ba ratio from the standard SRM3104a) in a suite of Hekla samples and demonstrate that magmatic differentiation does not impact the δ 138/134Ba values of OIB over MgO concentrations from 2.3 to 8.0 wt%, and higher MgO lavas are unlikely to experience δ 138/134Ba fractionation. After excluding OIB samples that experienced low-temperature alteration—which may shift the δ 138/134Ba values of samples away from their mantle sources—the δ 138/134Ba values of Samoan (including EM2 and high 3He/4He) lavas range from -0.07 to 0.07‰, while for Cook-Austral (HIMU) lavas the δ 138/134Ba values range from 0.00 to 0.11‰, and for St. Helena (HIMU) samples the δ 138/134Ba values range from 0.02 to 0.05‰. High 3He/4He (> 20 Ra, ratio to atmosphere) lavas from Vestfirðir of Iceland (-0.09 to 0.14‰) and Samoa (-0.02 to 0.01‰) also exhibit Ba isotopic variability. An extreme EM2 lava examined here (143Nd/144Nd = 0.512582) has lighter Ba isotopic composition (-0.07 ± 0.02‰, 2SE) than the depleted mantle sampled by MORB (0.05 ± 0.05‰, 2SD). The Samoan lavas show a possible positive trend between δ 138/134Ba values and 143Nd/144Nd ratios, and can be modeled as a binary mixing between the high 3He/4He (FOZO) mantle and recycled heterogeneous continental crust. In contrast, the extreme HIMU composition (206Pb/204Pb > 21.7) shows heavier Ba isotope signatures (0.11 ± 0.02‰, 2SE); δ 138/134Ba values positively correlate strongly with 206Pb/204Pb ratios in HIMU lavas, consistent with a contribution from recycled oceanic crust. [ABSTRACT FROM AUTHOR]

The fate of subducted crustal nitrogen (N) in modern-style subduction zones remains hotly debated from nearly complete return to the surface via arc volcanoes to net ingassing into the deep mantle beyond the sub-arc depth. One of the major obstacles for this controversy is the lack of constraint on the input N flux from altered oceanic crust (AOC). While notable N enrichment by low-temperature hydrothermal alteration has been widely reported in the upper volcanic section of AOC, N behavior during high-temperature alteration of the underlying intrusive oceanic crust (i.e., sheeted dikes and gabbros) remains poorly understood. Here we examined the bulk-rock N concentrations and isotopic compositions of the AOC (with a main focus on sheeted dikes and gabbros, yet including basalts from sections that have not been examined before) from ODP/IODP Hole 1256D and DSDP/ODP Hole 504B, which are two reference sites for subducting material into the warm Central America (CA) subduction zone. The results show that the intrusive sections have comparable N concentrations and isotopic compositions to their overlying upper volcanic rocks at 1256D (sheeted dikes: 16.4±7.2 ppm and −1.3±1.5‰; gabbros: 11.3±5.6 ppm and +1.3±1.0‰; basalts: 11.8±3.8 ppm and +0.9±1.7‰) and 504B (sheeted dikes: 7.6±2.1 ppm and +2.1±1.7‰; basalts: 8.0±3.8 ppm and +2.4±2.9‰). These data can be readily explained by mixing between inherited mantle N and secondary N mostly derived from seawater/sediments with a minor contribution from abiotic N 2 reduction. The secondary N in sheeted dikes and gabbros was mainly incorporated at moderate- to high-temperature (>250 °C) alteration stages and likely resides in chlorite, secondary plagioclase and amphibole. These new data clearly indicate that the intrusive section of AOC is a non-negligible N reservoir. Using these data, together with previously published N concentration data of basalts from Holes 1256D and 504B and of subducting sediments from Hole 1039B offboard the CA trench, we obtained a total N input flux of 3.1±0.6 × 109 mol‧yr−1 to 4.0±0.8 × 109 mol‧yr−1 into the CA trench. Integrating recently published N output flux of 0.58 × 109 to 1.4±0.6 × 109 mol‧yr−1 at the CA arc, an N recycling efficiency of 15±3–45±5% was yielded for the CA margin, indicating that up to 55±5–85±3% of subducted slab N is possibly transported beyond the sub-arc depth in this warm subduction zone. Based on the new finding of comparable N enrichment between basalts and intrusive oceanic crust, our modeling also gave a revised N recycling efficiency of 12±10–17±12% for the cold Izu-Bonin-Mariana (IBM) subduction zone, which is consistent with previous estimate of 4–17%. While these new estimates do not provide a conclusive constraint on the effect of subduction-zone thermal structure on N recycling, they consistently suggest that a large fraction of slab N (>50%) can survive the sub-arc filter and be delivered to the deeper mantle in global subduction zones. [ABSTRACT FROM AUTHOR]

Whether auriferous hydrothermal fluid is generated during crustal metamorphism or by subcrustal processes to form orogenic gold deposits is hotly debated. Noble gas isotope and halogen abundance ratios can provide important constraints to help resolve this controversy. The Danba gold deposit is a rare example of a Jurassic age hypozonal orogenic gold deposit on the western margin of the Yangtze Craton, China. It is characterized by thick barren quartz veins with fractures infilled by thin auriferous quartz veinlets. The syn -gold quartz veinlets are associated with plagioclase and pyrrhotite and are distinguished from the barren quartz vein hosts on the basis of cathodoluminescence, trace-element content and Raman spectroscopy and microthermometry of fluid inclusions. The dominant fluid inclusions in the gold-bearing quartz and plagioclase are H 2 O–CH 4 with vapor fill of ∼ 15–45 vol% and high pressure-corrected trapping temperatures of ∼ 500 °C, low salinities (<4 wt% NaCl equiv.) and high Mg/Fe ratios. Crushing syn -gold quartz and plagioclase released fluid inclusions with 40Ar/36Ar ratios of between 490 and 710, molar Br/Cl ratios of (0.3–0.7) × 10−3 and molar I/Cl ratios of (18–60) × 10−6. Crushing syn -gold pyrrhotite released fluid inclusions with 3He/4He ratios of 0.21–0.38 Ra (Ra = air value, 1.39 × 10−6) that are higher than typical crustal values of < 0.1 Ra, and 40Ar/36Ar of 330–640 that are similar to the gold-related quartz-hosted inclusions. The combined noble gas and halogen data do not favor an auriferous fluid derived from K- and U-rich Proterozoic granitic basement, which is expected to have radiogenic 40Ar/36Ar and 3He/4He ratios. Instead, the data are more consistent with a fluid source from underlying mantle lithosphere containing a high proportion of altered ocean crust (AOC). The auriferous fluid that formed the Danba gold deposit is thus considered to have been derived from a source comprising both metasomatized mantle lithosphere that had been fertilized in a Neoproterozoic subduction event and a portion of relict AOC beneath the western margin of the Yangtze Craton. In contrast with syn -gold fluid inclusions, those in pre-gold barren quartz are H 2 O–CO 2 dominated. Crushing pre-gold quartz released fluid inclusions with 40Ar/36Ar ratios of 2410–5450, Br/Cl ratios of (1.2–1.5) × 10−3 and I/Cl ratios of (60–80) × 10−6 that are distinctly different to the gold-ore stage. The higher 40Ar/36Ar ratios of the pre-gold fluids are consistent with a source for the barren host quartz veins from prograde greenschist- to amphibolite-facies metamorphism of the Proterozoic granitic basement. [ABSTRACT FROM AUTHOR]

Molybdenum isotope is a diagnostic tracer for crustal and mantle components in arc magmatism. However, the mechanism of Mo isotopic variation in arc magmas is still debated, e.g., input of different subduction components into the mantle wedge versus isotopic fractionation during dehydration of subducted slab. Here we present whole-rock Mo-Sr-Nd-Hf-Pb isotopic data for the Continental Arc Basalt (CAB) and Back Arc Basalt (BAB) from Java, Indonesia, to investigate the role of slab inputs in Mo isotopic variation of the Sunda arc magmatism. The CAB samples have variable K 2 O contents (0.44–2.49 wt%) and are mainly classified as calc-alkaline series, while the BAB samples are shoshonitic with markedly high K 2 O contents (2.12–6.90 wt%) relative to the CABs. The Java CABs and BABs have similar Mo isotopic compositions (δ98/95Mo = −0.65 to −0.07‰ and −0.66 to −0.07‰, respectively, relative to NIST SRM3134), suggesting that such a significant Mo isotopic variation should not be caused solely by the isotopic fractionation during the subduction. Instead, δ98/95Mo values of the Java basalts positively correlate with Pb isotopic ratios. This implies that the Mo isotopic variations in the Java arc rocks should result from the metasomatism in the mantle wedge by hybrid agents, including varying proportions of melts from subducted sediments (with heavy Mo isotope) and melts from the subducted altered upper oceanic crust (SAOC) (with light Mo isotope). The light Mo isotope of the Java arc rocks, compared with the Mariana arc basalts, suggests that melts from the SAOC have much lighter Mo isotopic compositions than the components from the lower oceanic crust. Thus, Mo isotope has great potential to distinguish the components from the subducted upper and lower oceanic crust. The Java CABs show along-arc variations in Mo-Sr-Nd-Hf-Pb isotopes, which is related closely with the thermal status of the subducted slab. Upwelling of the asthenosphere due to the slab tearing beneath the Java arc might have enhanced the partial melting of subducted sediments nearby the slab window. The complicated subduction system in the Sunda arc has strongly controlled the geochemical composition of arc magmas, which changes with input of different subduction components into the mantle wedge along arc. [ABSTRACT FROM AUTHOR]

A simple process-based model of the Li, Mg, Ca, Sr and alkalinity cycles in the ocean has been developed with the aim of better understanding the information contained in the evolution of Cenozoic seawater chemistry. In the model, changes in seawater chemistry are forced by changing climatic (temperature) and tectonic (ocean crust accretion rate and continental weatherability) conditions. These drive time-varying element and isotope fluxes from continental and seafloor weathering, mid-ocean ridge hydrothermal systems and sedimentation. Inverse modelling reveals that a range of model parameter values can simultaneously reproduce the key features of the Cenozoic history of seawater chemistry for all species considered. In addition to Cenozoic cooling, a decrease in the rate of ocean crust accretion and/or an increase in continental weatherability are required to fit the data. However, paleoseawater compositions do not allow the relative roles of these tectonic forcings to be determined. While other processes (e.g., changes in surface lithology) may have played a role in generating the observed variations in seawater chemical and isotopic composition, they are not required to explain the general tends in the data. The increase in Mg/Ca over the Cenozoic in acceptable models is largely driven by a decrease in the Ca/alkalinity of the net continental, diagenetic and hydrothermal fluxes, and not by an increase in the Ca/alkalinity of the carbonate mineral sink. Thus, the increase in seawater Mg/Ca over the Cenozoic is readily explained without any decrease in dolomite formation. Notably, all acceptable models have a stronger temperaturedependence of seafloor than continental weathering. This means that climatically driven changes in chemical fluxes in the model come dominantly from seafloor, not continental, weathering. [ABSTRACT FROM AUTHOR]

The Marquesas islands are an age-progressive volcanic chain in French Polynesia formed by partial melting above a mantle plume. Previous studies have demonstrated the presence of an enriched mantle (EM2) endmember mixed with a high-3He/4He (up to 14.4 R A) component in Marquesas post-shield lavas. In contrast, Marquesas shield lavas sample a dilute high-μ (HIMU) type component with lower 3He/4He (∼8 R A). We report new incompatible trace element and highly siderophile element (HSE: Os, Ir, Ru, Pt, Pd, Re) abundances and osmium isotope data for bulk rocks from the islands of Fatu Hiva, Hiva Oa and Nuku Hiva, for which He, Sr, Nd and Pb isotope compositions have been reported previously (Castillo et al., 2007). As the HSE are chiefly hosted in sulfide phases in the mantle and all but Re are compatible during partial melting, they offer information on both sulfide-rich and melt-depleted mantle lithologies within the Marquesas plume. Primitive mantle-normalized HSE abundances show depletions in Ru and Pd, and fractionated (Ir/Os) N in fourteen of the seventeen analyzed samples. Ruthenium depletions and Ir-Os fractionations are likely caused by parental magma fractional crystallization. Palladium depletions are not easily explained by magmatic or secondary alteration processes and are instead likely to be a feature of the source mantle which was partially depleted in the more incompatible HSE by previous melting events within the Marquesas plume. Osmium isotopic systematics are mildly radiogenic relative to chondrites and are identical within uncertainties for shield-building Fatu Hiva (γOs = +2.9 ± 1.6; 2 SD; n = 7) and post-shield Hiva Oa (+2.7 ± 2.2; n = 9), showing no evidence for significant assimilation of radiogenic oceanic crust. The osmium budget of Marquesas lavas is controlled by a volumetrically dominant peridotite endmember that is incompatible element depleted while the Sr, Nd, and Pb budgets are strongly influenced by a volumetrically minor, incompatible element enriched EM2 component. The Nd-Sr-Os isotope systematics of the lavas can be explained by mixing of 0.3–0.7% of an EM2 type melt into the Marquesas source. Marquesas lavas offer further support that the high-3He/4He "focus zone" (FOZO) mantle has 187Os/188Os mildly more radiogenic than the depleted upper mantle, similar to estimates of the primitive mantle. Osmium isotope systematics offer a useful tool with which to characterize more depleted components within the ocean island basalt-source mantle that may be difficult to detect using lithophile isotope signatures. [ABSTRACT FROM AUTHOR]

Despite being volumetrically minor components, carbonate-rich ultramafic magmas like aillikites represent good candidates to investigate the compositional variations in plume and/or lithospheric mantle sources because they represent low-degree melts which preferentially sample highly fusible components including recycled crustal material. To gain new insights into the composition of the plume-related magmas and, more broadly, the petrogenesis of ultramafic lamprophyres, we have undertaken the first comprehensive study of bulk rock and mineral (olivine and Ti-magnetite) highly siderophile element (HSE) abundances and Re-Os isotopes combined with in situ major-, trace-element and Sr-Nd isotope analyses of apatite and perovskite from the Permian Wajilitag aillikites of the Tarim large igneous province, China. The Wajilitag aillikites have high PPGE (Pt and Pd) contents relative to IPGE (Os, Ir and Ru), which can be ascribed to low-degree partial melting and/or fractionation of olivine and laurite. Measured 187Os/188Os ratios are moderately to highly radiogenic (0.186–0.313) with age-corrected γOs values up to +113. In situ Sr and Nd isotope analyses of apatite phenocrysts (87Sr/86Sr (i) = 0.70349–0.70384; εNd (i) = +1.3 to +4.9) and fresh perovskite grains (87Sr/86Sr (i) = 0.70340–0.70390; εNd (i) = +1.3 to +3.8) exhibit limited variability both within and across samples from different aillikite dykes and the only volcanic pipe in the area. These Nd isotopic values resemble those from bulk-rock samples (εNd (i) = +1.9 to +5.2), whereas Sr in apatite and perovskite extends to marginally less radiogenic values than the bulk-rock compositions (87Sr/86Sr (i) = 0.70362–0.70432). The moderately depleted Sr-Nd isotope compositions of magmatic apatite and perovskite, and the previously reported mantle-like C isotope values of these samples suggest that the aillikites and their carbon probably derived from a sub-lithospheric (plume) source with minimal contribution of deeply subducted material. Conversely, the radiogenic Os isotope compositions of the Tarim aillikites and separated minerals require some contribution from recycled crustal material in the plume source. Mass balance calculations suggest that the radiogenic Os isotopes and moderately depleted Sr-Nd isotopes can be reproduced by less than one third of eclogite component addition to a moderately depleted mantle source. We conclude that the combination of complementary isotopic systems can enlighten contributions from different components to mantle-derived magmas and, in this case, clarifies the occurrence of carbon-free subducted oceanic crust in the Tarim plume. [ABSTRACT FROM AUTHOR]

Ocean island basalts (OIBs) with Zn isotopic ratios higher than the normal mantle (δ66Zn = 0.17 ± 0.08‰) or mid-ocean ridge basalts (MORBs; δ66Zn = 0.27 ± 0.06‰) generally also have an enriched Sr-Nd isotopic signature, suggesting carbonate-bearing eclogites, whose protolith is inferred to be subducting altered oceanic crust, in their mantle source. On the contrary, continental intraplate basalts with high δ66Zn usually show depleted Sr-Nd isotopic signatures (i.e., decoupled Zn-Sr-Nd isotopic composition). To elucidate the origin of the decoupled Zn-Sr-Nd isotopic composition in continental intraplate basalts, we report the discovery of both coupled and decoupled Zn-Sr-Nd isotopic data for a suite of Cenozoic continental intraplate basalts from the Zhejiang province, Southeast China. These basalts display clear spatial and temporal geochemical variations, with early-stage inland low-silica samples presenting moderately enriched Sr-Nd isotopic signatures and high δ66Zn (coupled Zn-Sr-Nd isotopic composition, similar to OIBs), and later-stage coastal high-silica samples that display a pronounced δ66Zn decrease with increasing SiO 2 and 87Sr/86Sr and with decreasing alkali contents and 143Nd/144Nd (decoupled Zn-Sr-Nd isotopic composition). The early-stage basalts with coupled high Zn-Sr-Nd isotopic signatures are also more enriched in incompatible elements than any other basalts from eastern China reported so far. We explain the spatial and temporal geochemical variations of these basalts as the result of two main melting events: 1) the low-silica early-stage magmatism mostly occurs inland and results from high-pressure partial melting of a carbonated eclogite-bearing asthenospheric mantle. Because of the presence of a thick lithosphere limits the melting of the depleted mantle component, the signature of the Zn-Sr-Nd isotopically enriched, and more fusible carbonated eclogite is preserved. 2) At the later stage, magmatism mostly occurs on the coast where the subcontinental lithosphere is thinner. Hence, decompression melting progresses to shallower depth, resulting in an increase of the contribution from the depleted peridotite matrix and a dilution of the signal from the isotopically enriched fusible component. Further upwelling and in-situ melting at the base of the subduction-modified sub-continental lithospheric mantle (SCLM) explains both the decoupled Zn-Sr-Nd isotopic signature of the coastal basalts and their major and trace element variability. We further propose that decompression melting is driven by small-scale convection resulting from variations of lithospheric thickness. Our data highlight the importance of dynamic melting of carbonated eclogite-bearing asthenosphere and subsequent lithospheric melting in preservation and destruction of the coupled enriched Zn-Sr-Nd isotopic signature of carbonated eclogite component and generation of the apparent decoupled Zn-Sr-Nd isotopic signal commonly observed in continental intraplate basalts. [ABSTRACT FROM AUTHOR]

Alkaline basalts occur widely in intraplate settings and carbonate-bearing mantle sources such as carbonated peridotites are increasingly regarded to play a key role in their formation. Carbonated eclogites, most likely the products of subducted carbonate-bearing altered oceanic crust, are probable alternative ingredients in the mantle sources of many intraplate alkaline basalts, highlighting the importance of the subduction-driven deep carbon cycle. However, this widely proposed hypothesis remains enigmatic because the recognition of low-MgO primitive alkaline basalts predicted by experiments is scarce. Here we show that Cenozoic continental intraplate alkaline basalts occurring above the stagnant oceanic slab in the mantle transition zone beneath the Hannuoba region, eastern China, display geochemical features consistent with their origin as low-degree partial melts of carbonate-bearing eclogites. Their MgO contents correlate positively with CaO, Ba/Th and Ti/Eu, but negatively with Dy/Yb and ε Nd. Remarkably, the most primitive alkaline basalts are characterized by low MgO (<5.25 wt.%), low heavy rare earth elements and Sc contents, low CaO/Al 2 O 3 (<0.41), low Ti/Eu (<3380), but Dy/Yb (>7.1) higher than those of ocean island basalts (OIBs). These features cannot be ascribed to differentiation from high-MgO alkaline basalts because significant amounts of crystallization of clinopyroxene and garnet did not occur during ascent. Differentiation also cannot account for the correlations of time-integrated Sr-Nd isotopes with MgO, Dy/Yb and Ba/Th. Instead, the linear correlations mainly reflect strong interaction between ascending primitive alkaline melts and the lithospheric mantle. The compositions of primitive alkaline basalts reflect the key control of garnet and clinopyroxene in the mantle residue (eclogites), and the Ti, Zr and Hf anomalies further indicate the critical effect of carbonates in the eclogite source. Partial melting of the carbonate-bearing eclogites likely occurred in the uppermost asthenosphere. The production of alkaline basalts with low MgO contents by partial melting of carbonate-bearing eclogite below the peridotite solidus in an intraplate setting has been overlooked and the magmas were instead often considered to be highly evolved. Recycled altered oceanic crust thus may not only cause metasomatism of the deep mantle but may also serve as a direct source of mafic melts. These results on natural rocks support the experiment-based model for subducted altered oceanic crustal material and also indicate its diverse fate in the mantle. [ABSTRACT FROM AUTHOR]

Inclusion-bearing diamonds from the Koidu kimberlite complex, Sierra Leone (West African Craton) were analyzed in situ for carbon and nitrogen isotope compositions, nitrogen concentrations and nitrogen aggregation states. In a suite of 105 diamonds, 78% contain eclogitic mineral inclusions, 17% contain peridotitic mineral inclusions, and 5% − an unusually high proportion − contain co-occurring eclogitic and peridotitic mineral inclusions indicating a mixed paragenesis. Major and trace element compositions of mineral inclusions from two mixed paragenesis diamonds (one with omphacite + Mg-chromite, the other with eclogitic garnet + forsteritic olivine) were determined. The presence of positive Eu anomalies in centrally located omphacite and eclogitic garnet inclusions indicates derivation from subducted protoliths, formed as igneous cumulates in lower oceanic crust. Mg-chromite (Cr# 85.5; Mg# 65.2) and olivine (Mg# 94.5) inclusions, located in outer portions of the mixed paragenesis diamonds, have compositions indicative of derivation from strongly depleted cratonic peridotites. Given that the olivine Mg# of 94.5 is the highest reported to date for the West African Craton, the eclogitic and peridotitic inclusions in these mixed paragenesis diamonds cannot have precipitated during infiltration of peridotitic substrates by eclogite-derived fluids, as the consequent fluid-rock interaction should lead to Mg# lower than that for the original peridotitic diamond substrate. The different origins of eclogitic and peridotitic inclusions could be explained by physical transport of their host diamonds from eclogitic into peridotitic substrates, possibly along high-strain shear zones, before renewed diamond growth. Based on the δ13C–δ15N systematics of the entire inclusion-bearing diamond suite from Koidu, three major compositional clusters are identified. Cluster 1 (eclogitic diamond cores; δ13C = –33.2 to −14.4‰ and δ15N = −5.3 to +10.1‰) bears the isotopic signature of recycled crustal material (± a mantle component). Cluster 2 (peridotitic diamonds and including the core of a diamond containing omphacite + Mg-chromite; δ13C = −6.0 to −1.1‰ and δ15N = −4.2 to +9.7‰) likely involves mixing of carbon and nitrogen from subducted and mantle sources. Cluster 3 (rims of eclogitic diamonds and including the eclogitic garnet + olivine included diamond and the rim of the omphacite + Mg-chromite included diamond; δ13C = −7.8 to −3.6‰ and δ15N = −7.9 to −2.1‰) matches convecting mantle-derived fluids/melts. The distinct isotopic signatures of the three diamond clusters, together with differences in nitrogen aggregation and cathodoluminescence response between diamond cores and rims, suggest episodic diamond growth during multiple fluid/melt pulses. [ABSTRACT FROM AUTHOR]

Understanding the Earth's geological nitrogen (N) cycle requires an understanding of how N behaves during dehydration of subducted crust. We present the first in situ measurements of N in silicate minerals by secondary ion mass spectrometry, focusing on high pressure rocks representing subducted oceanic crust. We investigate the distribution of N between mineral phases, and combine analyses of N with other trace and major elements in order to constrain the behaviour of N during fluid-rock interaction. The data confirm that white mica (phengite, paragonite) is the primary host for N, containing up to 320 μg/g, whereas minerals including clinopyroxene, amphibole and epidote contain < 5 μg/g N. Chlorite can also contain N (up to 83 μg/g) and may play a previously unrecognised role in subduction zone N cycling. Bulk rock N concentrations estimated from mineral N concentrations and mineral modes are consistent with N concentrations measured by bulk combustion, which confirms that most N is hosted within silicate minerals and not along grain boundaries or in fluid inclusions. Bulk rock N contents correlate with K 2 O (N/K 2 O = 19.3 ± 2.0). Using N/K 2 O ratios and the average K 2 O of altered oceanic crust, the flux of N subducted in upper oceanic crust is estimated to be 0.6 - 2.4 × 10 11 g / yr , which is consistent with but at the lower end of previous estimates. The data were also used to investigate the behaviour of N during fluid-rock interaction. Open system fluid-rock interaction modelling was used to model the evolution of N, B and Li contents during fluid-rock interaction in phengite from a garnet-phengite quartzite. By comparison to data for B and Li, the phengite-fluid partition coefficient for N was estimated to be 0.1–1.5. Separately, the growth of paragonite during fluid-rock interaction in a blueschist was shown to sequester N from phengite and limit bulk N loss to the fluid. The stability of white mica during fluid-rock interaction is therefore critical in controlling the behaviour of N. Nitrogen addition from sediment-derived fluids appears to be an important process in subduction zone rocks. Mafic crust can act as a sink for this N if white mica is stable. This work provides the first natural constraints on the fluid-mineral partitioning behaviour of N at subduction zone conditions and emphasises the complexity of N mobility within subduction zones, with redistribution between different phases and lithologies being important. [ABSTRACT FROM AUTHOR]

Arc lavas and the continental crust exhibit a selective enrichment of W relative to similarly incompatible elements, indicating that subduction zone environments are a tectonic setting where W is mobilized from the subducting slab. Here we present evidence that ultramafic portions of altered oceanic lithosphere are a major sink for W and evaluate how its W budget needs to be included in constraining the global geochemical cycle of W. We report high precision W, HFSE and U-Th data obtained by isotope dilution for altered basement formed at super-slow spreading rates along the Mid-Atlantic Ridge drilled during ODP Leg 209 (between 14°N and 16°N). The oceanic basement consists of abyssal peridotites and associated gabbroic rocks exhumed in a magma starved setting. Hydrothermal alteration styles covered in the sample set are serpentinization, talc alteration, and low-T seawater alteration. The drilled rock portions are among the most depleted in HFSE that have ever been studied. Tungsten concentrations range from 4 ppb in the least altered harzburgites to 500 ppb in strongly serpentinized dunites. Locally, W is significantly enriched relative to U, Th and Ta by factors of up to 100 compared to canonical mantle values, much higher than in mafic portions of altered oceanic crust such as hole 1256D (up to 10). The samples from ODP Leg 209 show strong W enrichment during progressive serpentinization (holes 1268A, 1270 and 1271) and during late-stage oxidative seawater alteration (hole 1270D and 1272A). In contrast, silica metasomatism associated with talc alteration (hole 1268A) is associated with selective W depletion relative to Th and Ta. Trace element modelling indicates that hydrothermal enrichment of W by redistribution is a more likely source than seawater-derived W. Collectively, our data show that altered ultramafic rocks likely constitute an important geochemical reservoir in the global geochemical cycle of W contributing to the enrichment of W found in arc lavas and to the recycling of W into the Earth's mantle. [ABSTRACT FROM AUTHOR]

Reactive porous or focused melt flows are common in crystal mushes of mid-ocean ridge magma reservoirs. Although they exert significant control on mid-ocean ridge magmatic differentiation, their role in metal transport between the mantle and the ocean floor remains poorly constrained. Here we aim to improve such knowledge for oceanic crust formed at slow-spreading centers (approximately half of present-day oceanic crust), by focusing on specific igneous features where sulfides are concentrated. International Ocean Discovery Program (IODP) Expedition 360 drilled Hole U1473A 789 m into the lower crust of the Atlantis Bank oceanic core complex, located at the Southwest Indian Ridge. Coarse-grained (5–30 mm) olivine gabbro prevailed throughout the hole, ranging locally from fine- (<1 mm), to very coarse-grained (>30 mm). We studied three distinct intervals of igneous grain size layering at 109.5–110.8, 158.0–158.3, and 593.0–594.4 meters below seafloor to understand the distribution of sulfides. We found that the layer boundaries between the fine- and coarse-grained gabbro were enriched in sulfides and chalcophile elements. On average, sulfide grains throughout the layering were composed of pyrrhotite (81 vol.%; Fe 1-x S), chalcopyrite (16 vol.%; CuFeS 2), and pentlandite (3 vol.%; [Ni,Fe,Co] 9 S 8), which reflect paragenesis of magmatic origin. The sulfides were most commonly associated with Fe-Ti oxides (titanomagnetites and ilmenites), amphiboles, and apatites located at the interstitial positions between clinopyroxene, plagioclase, and olivine. Pentlandite exsolution textures in pyrrhotite indicate that the sulfides formed from high-temperature sulfide liquid separated from mafic magma that exsolved upon cooling. The relatively homogenous phase proportion within sulfides along with their chemical and isotopic compositions throughout the studied intervals further support the magmatic origin of sulfide enrichment at the layer boundaries. The studied magmatic layers were likely formed as a result of intrusion of more primitive magma (fine-grained gabbro) into the former crystal mush (coarse-grained gabbro). Sulfides from the coarse-grained gabbros are Ir-Platinum Group Element-rich (PGE; i.e., Ir, Os, Ru) but those from the fine-grained gabbros are Pd-PGE-rich (i.e., Pd, Pt, Rh). Notably, the sulfides from the layer boundaries are also enriched in Pd-PGEs, and therefore elevated sulfide contents at the boundaries were likely related to the new intruding melt. Because S concentration at sulfide saturation level is dependent on the Fe content of the melt, sulfide crystallization may have been caused by FeO loss, both via crystallization of late-precipitating oxides at the boundaries, and by exchange of Fe and Mg between melt and Fe-bearing silicates (olivine and clinopyroxene). The increased precipitation of sulfide grains at the layer boundaries might be widespread in the lower oceanic crust, as also observed in the Semail ophiolite and along the Mid-Atlantic Ridge. Therefore, this process might affect the metal budget of the global lower oceanic crust. We estimate that up to ∼20% of the Cu, ∼8% of the S, and ∼84% of the Pb of the oceanic crust inventory is accumulated at the layer boundaries only from the interaction between crystal mush and new magma. [ABSTRACT FROM AUTHOR]

Pervasive alteration of basaltic oceanic crust by heated seawater at greenschist facies conditions produces two contrasting hydrothermal rocks. "Spilites", consisting of chlorite + albite + quartz ± actinolite ± epidote, occur typically with regional extents. Locally spilites are metasomatically transformed to "epidosites" consisting of epidote + quartz + titanite + hematite or magnetite. Both alteration types have been proposed as markers of deep hydrothermal upflow in sub-seafloor convection cells, and as sources of the ore metals in basalt-hosted seafloor massive sulfide deposits. Little direct evidence is available for the chemical compositions of these fluids in their states deep in the upflow zones prior to their discharge at the seafloor. To better characterize them we have conducted a field, petrographic and fluid inclusion study of the lavas, sheeted dikes and plagiogranites in the Semail ophiolite, with supporting samples from the Troodos ophiolite. Our results show that both the spilite- and epidosite-forming fluids were single-phase aqueous liquids during the hydrothermal alteration. At some sites their salinity is 3.1–3.2 wt.% NaCl eq , which we take to represent the chlorinity of Cenomanian seawater in the Semail realm. At other sites salinities are as low as 2.4 wt.% NaCl eq or as high as 5.7 wt.% NaCl eq , attributable to liquid–vapor separation and partial remixing deep in the crust along the dew curve of seawater, prior to ascent of the fluids to the sites of fluid inclusion trapping. Hypersaline brines, often accompanied by vapor, are restricted to plagiogranites in both the Semail and Troodos ophiolites and they represent magmatic–hydrothermal fluids that pre-date and are genetically unrelated to the spilite and epidosite alteration. The volcanostratigraphic locations of the samples constrain their maximum depths to 1470–3600 m below seafloor during alteration. The range of possible fluid trapping pressures for all samples is 31–68 MPa. Trapping temperatures vary between sites from 145 to 440 °C for spilite fluids and 255 to 435 °C for epidosite fluids. Quantitative analyses of 12 elements in individual fluid inclusions by LA-ICP-MS define the chemical characters of the two alteration fluids. The Br/Cl ratio in the spilite fluid is the same as in modern seawater and the other elements fit expectations from seawater–basalt experiments at elevated temperature. Accordingly, concentrations of Li, B, Na, Cl, K, Br and Sr in the spilite fluid match those in modern black-smoker vent fluids in basaltic crust. Exceptions are Ca and Fe, which are enriched in the spilite fluid. As these elements may precipitate below or at the seafloor prior to vent sampling, we conclude that the spilite fluids are plausible feeders of basalt-hosted black-smoker vents. The epidosite fluid has broadly similar elemental concentrations to the spilite fluid, but vastly lower Fe, reflecting the highly oxidized state of epidosites. This suggests that epidosite fluids are incapabable of forming basalt-hosted seafloor massive sulfide deposits. [ABSTRACT FROM AUTHOR]

The origin of enriched mantle 1 (EM1) component in the mantle has long been debated. EM1 source of basalts has been variously attributed to recycled sub-continental lithosphere, recycled lower continental crust, or recycled oceanic crust plus pelagic sediments. Molybdenum stable isotope systematics has the potential to resolve this debate, because of the significant isotopic differences between mantle and crustal reservoirs. Here we present molybdenum isotope data for EM1-type continental intraplate basalts from Nuominhe (NMH), Northeast China. The NMH basalts display significant variations in both δ98/95Mo (98Mo/95Mo ratio relative to NIST SRM 3134, −0.49 to −0.15‰) and Mo/Ce (0.009–0.037), and strong correlations between δ98/95Mo, Mo/Ce and Hf isotope and trace elements. EM1 end-member of the NMH basalts shows the lightest δ98/95Mo and lowest Mo/Ce values. The involvement of continental crust and sub-continental lithospheric mantle cannot account for the light Mo isotopic compositions of the basalts. By contrast, the recycled oceanic crust with pelagic sediments, characterized by low δ98/95Mo and Mo/Ce values, is the satisfactory candidate for the EM1 component in the NMH basalts. Our results also suggest a residual slab reservoir with light δ98/95Mo values (<−0.5‰) in the mantle beneath Northeast China, representing the complementary reservoir to heavy Mo isotope continental crust. Identification of this reservoir makes the Mo isotope system a powerful tracer of crust-mantle interactions. [ABSTRACT FROM AUTHOR]

At mid-ocean ridges (MORs), seawater carrying dissolved sulfate (SO 4) infiltrates the oceanic crust. Hydrothermal fluid emissions from such systems have much lower δ34S and sulfur is mostly present as reduced sulfide, albeit in lower total sulfur concentrations than in seawater. This has been explained by anhydrite formation and sulfate reduction based on petrographic evidence and mass balance considerations. Here, we utilize the chemical and stable isotope (δ34S, δ18O) systematics in natural anhydrite and pyrite from various locations along the submarine and on-land section of the Mid-Atlantic ridge near Iceland to quantify the key variables that control anhydrite formation and sulfate recycling in the oceanic crust. Hydrothermal anhydrite exhibited δ34S values of +20.6 ± 1.0‰ and δ18O values between +2.4 to +25.3‰. Volcanogenic anhydrite in encrustations showed δ34S values of −1.7 to +21.4‰ and δ18O values between +1.4 and +38.0‰. Hydrothermal pyrite exhibited δ34S values ranging from +3.4 and +19.7‰. Comparison of the natural dataset with results from geochemical isotope modelling revealed that δ34S and δ18O values of anhydrite and pyrite were dependent on the isotope composition of the source fluid, extent of water–rock interaction, temperature, and redox conditions. Departures of δ34S and δ18O values in anhydrite from the source fluid were caused by progressive fluid-basalt interaction where lower δ34S and δ18O values reflected a change in sources of S and O from solely fluid to basaltic origin. The δ18O values of anhydrite were additionally affected by temperature. Quantitative formation of anhydrite mainly occurred at temperatures < 150 °C, whereas at elevated temperatures (>200 °C) reduction of seawater-sulfate to H 2 S and subsequent pyrite precipitation were found to limit anhydrite formation. Extending our calculations to the oceanic crust revealed that the majority of seawater-sulfate is sequestered into anhydrite (3–38 Tg S yr−1) in vicinity of MORs at < 200 °C at shallow depth (<1500 m), with only a small portion of seawater-derived SO 4 discharged by high-temperature hydrothermal vents (0.1–3.4 Tg S yr−1). However, sequestration of sulfur by anhydrite is not long-lasting due to retrograde dissolution of anhydrite. The removal of anhydrite upon cooling and aging of the crust may result in a return back to the oceans of 10–60% of the sulfur originally sequestered in anhydrite upon hydrothermal alteration in vicinity of MORs. [ABSTRACT FROM AUTHOR]

Serpentinite is one of the most important sources of volatiles and fluid-mobile elements in subduction zones. Attempts have been made to discriminate serpentinite-derived fluid from other sources (e.g., sediments, altered oceanic crusts) using a variety of geochemical tools, but such differentiation has proven to be changeling because fluids from different lithologies are dynamically mixed in the subduction channel. Serpentinites are essentially distinguished by high Cr contents and variable degrees of 53Cr excess. Given high Cr mobility in Cl-rich metamorphic fluids based on experimental determination, Cr elemental and isotope compositions could be potential markers for fluid-mediated mass transfer from serpentinites. Here, we test this hypothesis by analyzing Cr concentrations and isotope compositions of meta-basalts, as well as country mica schists, marbles and serpentinites from the southwestern Tianshan Orogen. The meta-basalts have highly variable Cr concentrations (30.0–625 ppm) and resolvable δ53Cr variations (−0.25‰ to −0.05‰), while mica schists and marbles have similar isotope compositions and low Cr contents (δ53Cr of −0.15‰ to −0.14‰ with Cr contents of 89.5–110 ppm for mica schists, δ53Cr of −0.21‰ to −0.17‰ with Cr contents of 4.1–8.73 ppm for marbles, respectively). Most of the investigated meta-basalts were carbonated. The positive correlations of loss on ignition (LOI), CO 2 , Sb and Sb/Ce with Cr contents, as well as with δ53Cr for these carbonated meta-basalts indicate that the metamorphic fluids are at least partially sourced from serpentinites and sedimentary carbonates. High pressure serpentinites display similar δ53Cr variations from −0.19‰ to −0.02‰, but high Cr contents ranging from 934 to 4920 ppm. Chromium isotope variations of these serpentinites could not be generated by serpentinization, due to opposite trends of Cr isotopes and serpentinization indexes compared with previous observations. Instead, dehydration of serpentinites during subduction/exhumation accounts for Cr isotope variations, generating fluids with high Cr concentrations and relatively lighter isotope compositions. According to a Rayleigh dehydration model, the estimated Cr isotope fractionation factor (α fluid-residue) between fluids and residual serpentinites is approximately 0.99995 –0.99975, implying that Cr species in the subducted serpentine derived fluid is probably as complexes of Cr3+-Cl− H 2 O(OH–). The mixing model reveals that investigated carbonated meta-basalts were sourced from initial evolved oceanic basalts metamorphosed by fluids from serpentinite dehydrations, with contributions from sedimentary carbonates. This work emphasizes that coupled Cr element and isotope systems present the potential ability to discriminate the complex fluid sources in subduction zones, particularly serpentinites. [ABSTRACT FROM AUTHOR]

To constrain the behavior of K isotopes during the low-temperature alteration of oceanic crust and reveal its impact on the global K cycle, we measured the K isotopic compositions of 49 fresh and altered basalts recovered from the Integrated Ocean Drilling Program (IODP) Sites U1365 (~100 Ma) and U1368 (~13.5 Ma) in the South Pacific Ocean. The measured basalts representing the uppermost oceanic crust have been subjected to low-temperature (<150 °C) alteration of different types and intensities as reflected by different secondary mineral modal abundances and element mobility. Both Sites have remarkably low sedimentation rates (0.71 and 1.1 m/Myr, respectively). The altered basalts from both Sites show higher K 2 O contents and K/Nb ratios than fresh basalts, suggesting the addition of K during low-temperature alteration. The measured δ41K values of the altered basalts vary over a large range (−0.76‰ to −0.17‰) compared to unaltered basalts, indicating significant K isotopic fractionation during low-temperature alteration of the oceanic crust. We propose that the K isotopic characteristics of the altered basalts were caused by the combined effects of equilibrium and kinetic isotopic fractionations between seawater and alteration minerals during the incorporation and adsorption of K into alteration minerals. A weighted average δ41K value of −0.40 ± 0.33‰ (2sd) is given for the low-temperature altered oceanic crust (AOC) at Sites U1365 and U1368, which is indistinguishable from the fresh N-MORB value (−0.44 ± 0.17‰), but significantly lower than its K source, i.e., seawater (δ41K = +0.12 ± 0.07‰, 2sd). Therefore, low-temperature AOC serves as an important sink for K and is a cause for the heavy K isotopic composition of seawater. In addition, the heterogeneous K isotopic compositions of the low-temperature AOC, together with large scale fractionation during various crustal processes, further highlights the utility of the K isotope system to trace subducted crustal materials. [ABSTRACT FROM AUTHOR]

• Sulfides in high-pressure rocks host significant Cu, As, Ag, Cd and Te. • The Pb budget of high-pressure rocks is governed by silicates, not sulfides. • Co, Ni, Zn, Ga and Pb may be lost during lawsonite, epidote and amphibole breakdown. • Ge, Tl, Sn and Mo are largely retained in the eclogitic slab residue. • Co, Ni and As zoning in pyrite record variable degrees of fluid–rock interaction. Subduction zones are the site of long-term chemical cycling between Earth's surface and interior. During subduction, dehydration and melting may partition mobile elements into the overlying mantle wedge and arc system. Volcanic arcs also host much of the world's base (e.g. , Cu, Sn, Sb, and Mo) and precious (e.g. , Au, Ag) metal deposits. However, slab contributions to the chalcophile and siderophile element (CSE) budget of arc magmas and their associated ore deposits remain controversial. Although previous studies have identified the mobility of some CSE in slab fluids, few studies have examined the contribution from sulfides to the CSE budget of the slab. Here we report in situ laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) analyses of Cr, Co, Ni, Cu, Zn, Ga, Ge, As, Mo, Ag, Cd, In, Sn, Sb, Te, Tl, Pb, and Bi in sulfides and silicates in thirteen rocks from six exhumed high-pressure (HP) terranes worldwide. Combined with data from the literature, we demonstrate that sulfides host nearly the entire Cu, As, Ag, Cd, and Te budget of subducted rocks, whereas Co, Ni, Zn, Ga, Ge, Mo, Sn, and Tl are dominantly hosted in silicate and oxide phases. Prograde sulfide-breakdown during subduction is expected to liberate Cu, As, Ag, Cd, and Te into the overlying arc-mantle wedge system, whereas Co, Ni, Zn, Ga, and Pb may be partially lost during lawsonite, epidote, and amphibole dehydration reactions. In contrast, Ge, Tl, Sn, and Mo are largely retained in the eclogitic slab residue. We also demonstrate that CSE zoning in pyrite may be used to track high pressure (HP) metasomatism. Oscillatory Co and Ni zoning in metasomatic pyrite is linked to transitions between fluid and rock buffered regimes, consistent with seismic- and dehydration-induced cyclic fluctuations in fluid pore pressure and flux. Thus, minor and trace CSEs in HP rocks may also be a useful tool to constrain slab fluid migration. [ABSTRACT FROM AUTHOR]

The role of water on melting in the mantle wedge is still debated due to large uncertainty on the estimates of H 2 O flux beneath arcs. B has been proven as an effective proxy for water flux because B and H 2 O show similar chemical behaviors during subduction. The Habahe mafic dikes from the Chinese Altai were emplaced within a narrow area (<20 km from south to north) during the northward subduction of the Junggar Ocean in the middle Paleozoic. These dikes have been classified into four types with distinct geochemical and Sr-Nd-Hf-Pb isotopic compositions, which originated from mantle sources metasomatized by different subduction components, including melts from subducted sediments (Type-I, Type-IV), fluids from subducted sediments (Type-II), and melts from subducted oceanic crust (Type-III). We present B content and isotope data for the Habahe mafic dikes to investigate the influence of subduction components on melting in the mantle wedge. Type-I and -III mafic dikes all have negative δ11B values (−7.7‰ to −5.0‰) with variable B contents (3.65–13.4 ppm) and B/Nb ratios (2.10–7.39), indicating B isotopically light features for the subducted sediments and oceanic crust. Type-II mafic dikes have lower B contents (3.97–9.90 ppm) and higher B/Nb ratios (7.07–14.4) than Type-I mafic dikes, with a wide range of δ11B values from −7.8‰ to −2.7‰. This suggests that their mantle source may have been metasomatized by fluids from subducted serpentinite besides fluids from subducted sediments. Type-IV mafic dikes have higher B contents (17.0–27.5 ppm) and B/Nb ratios (25.0–40.8), and heavier B isotopic compositions (δ11B = −2.9‰ to +3.5‰) than Type-I mafic dikes. This indicates involvement of fluids from the slab serpentinite in metasomatism of their mantle source in addition to melts from the subducted sediments. The Habahe mafic dikes show wide range of B/Nb ratios, suggesting that different amounts of water were added into their mantle sources. These dikes exhibit variable Zr/Yb and Nb/Yb ratios, and constantly low TiO 2 /Yb, indicating their formation through different degrees melting of depleted mantle sources. Their Zr/Yb and Nb/Yb ratios are negatively correlated with B/Nb, which reflects elevation of the melting degree of their mantle sources as increasing water input. Similar trends are also observed in basalts from global arcs and their major and trace elements correlate well with B/Nb ratios. Thus, water flux should play an important role on melting in the mantle wedge and control magma compositions of the arcs. [ABSTRACT FROM AUTHOR]

The Earth's mantle exhibits marked chemical heterogeneity. We provide an examination of thallium systematics in ocean island basalts (OIB): new high-precision trace element analyses, including Tl, and Tl isotopic compositions for 48 OIB spanning the entire range of observed Sr-Nd-Hf-Pb isotope ratios. All investigated OIB are characterised by ubiquitous Tl depletion requiring OIB mantle sources to have Tl concentrations as low as 0.2 ng/g, which is an order of magnitude lower than estimates for the primitive mantle and similar to Tl concentrations inferred for the depleted mantle. The low Tl concentrations inferred for OIB mantle sources are interpreted to reflect near quantitative removal of Tl during subduction and inefficient Tl recycling into the deeper mantle. If true, the Tl isotopic composition of surface materials may not be readily translated to the mantle sources of OIB. The new OIB dataset shows a >10 ε-unit range in primary isotopic variation, from ε205Tl = −6.4 to +6.6. However, the majority of samples (32 of 48) are within uncertainty of mantle values (ε205Tl = −2 ± 1), and show no co-variation with radiogenic isotopic composition. Notably, OIB with only minor Tl depletion (11 samples) have Tl isotopic compositions outside the mantle range. The Tl concentration contrast between the mantle and inputs such as sediments and altered basalt is so great that minor additions (<1% by mass) of high-Tl material will dominate the isotopic budget of a lava, with decoupling of Tl and radiogenic isotopic compositions as an expected result. Thallium isotopic compositions of OIB are therefore difficult to link directly to radiogenic isotope variations and the mantle components they may reflect. Indeed, if isotopically distinct Tl from altered oceanic crust and/or sediments were efficiently recycled into the mantle and sampled via OIB, more variation in the Tl isotopic composition of OIB would be expected than is observed. The markedly unsystematic primary Tl isotopic variations in OIB therefore likely reflect the residual Tl isotopic composition of subducted material, and/or Tl acquired en route to the surface via shallow-level crustal assimilation. [ABSTRACT FROM AUTHOR]

In order to examine the seawater-seafloor sediment interactions that influence the chemical composition of seawater through time, we examined hundreds of pore fluid geochemical analyses from 13 clay-rich sedimentary successions drilled by the ODP-IODP. Chemical trends such as monotonous increases in Ca2+, and decreases in Mg2+ and δ18O with depth are traditionally interpreted to result from water-rock interaction. In this view, the release of Ca2+ into fluids and the uptake of Mg2+ and 18O mainly results from the formation of low-temperature clays in the sediment and within underlying basalts. Chloride concentration profiles and isotopic compositions, however, suggest that different processes may influence pore water geochemistry. The data examined here show relatively constant chloride contents but with a systematic decrease in δ37Cl of chlorides with depth from 0 permil (the seawater value) down to −8.5 permil. The δ37Cl data are highly correlated with δ18O (with δ18O down to −5.7 permil). The δ37Cl-depletions of pore fluid chlorides are found in all studied sedimentary piles regardless of tectonic or sedimentary history. These trends cannot be explained by water–rock exchange reactions because minerals formed at low temperature have Cl contents that are too low to compensate for δ37Cl depletions observed in pore fluids. Accordingly, we hypothesize that fluid-specific processes are responsible for the δ37Cl-depletions of the fluids and that δ37Cl-enriched chlorides were expelled out of the sediments into the ocean. After reviewing the fluid-specific processes that are known to change the chlorine isotope ratios in chlorides, we rule out diffusion and gravitational isotope fractionations of chlorides could generate this isotope pattern. The flow of a δ37Cl -depleted fluid from the underlying basaltic basement into the sediments could explain the δ37Cl data. But the mechanism that produces depletion in δ37Cl of the fluid remains unknown. It cannot be chloride exchanges between fluids and rocks. Here we show that compaction-induced ion filtration of chlorides through clay-rich membranes can produce the observed pore fluid δ37Cl-depletions, with isotope fractionation factors ranging from 1.000 to 1.008 between the chlorides of the expelled fluid (the permeate) and those of the residual fluid (the retentate). We find that smectite-rich sediments are associated with higher isotopic fractionation factors, while illite/chlorite-rich sediments are associated with intermediate values and with clay-poor sediments associated with lower values. This suggests that chlorine isotope fractionation might be controlled by surface charge associated with specific clay minerals. Our calculations show that compaction-induced filtration has the capacity to produce 18O-depletion for oxygen isotope fractionation factors between the expelled fluid and the retentate ranging from 1.000 to 1.005. 18O-enrichment in the expelled fluid is in agreement with the experimental data of Haydon and Graf (1986). Overall, although further experimental work on both chlorine and oxygen isotopes is certainly needed, the results of this study indicate that ion-filtration should be considered as a potential mechanism for fractionating isotopic species in sediment pore waters, particularly for oxygen isotope ratios whose variations are often commonly attributed to water–rock exchange. [ABSTRACT FROM AUTHOR]

• The Itasy volcanic field has experienced abundant Pleistocene monogenetic volcanism. • Itasy magmas have geochemical signatures similar to ocean island basalts. • Itasy magmas are produced by melting of upwelling polluted asthenospheric mantle. • The Itasy magmas may be sourced in part by mantle related to the African superplume. We present new major and trace element and Sr, Nd, Pb, Hf, and Os isotope data for Quaternary basaltic lavas and tephra from the Itasy volcanic field, in order to assess the origin and sources of recent volcanism in the central highlands of Madagascar. Our age determinations provide further evidence for Pleistocene volcanism on the island, with ages between ∼31 and 104 ka. Mafic magmas from Itasy exhibit significant geochemical heterogeneity, and they are chemically distinct from those of the neighboring Ankaratra volcanic complex. Trace element signatures of the Itasy volcanic field are very similar to those of ocean island basalts (OIB), with significant enrichment in high field strength elements (e.g. Nb and Ta) and depletion in large ion lithophile elements (e.g. Rb, Cs, and K), and relatively high Nb/U (>45) and Ce/Pb (>25) ratios. Significant intra-volcanic field isotopic heterogeneity is observed, with well-defined negative correlations of 87Sr/86Sr with 206Pb/204Pb and 143Nd/144Nd. These correlations, along with highly radiogenic Os isotopic signatures in many Itasy samples, could be attributed to assimilation of the Precambrian basement. However, with the exception of Os, the elemental and isotopic characteristics of the Itasy magmas are inconsistent with significant crustal assimilation, and instead are interpreted to represent features of their mantle sources. The Itasy Δε Hf and Δ7/4 values fall within the range of MORB and OIB, and indicate derivation primarily from an asthenospheric mantle source. The absence of significantly negative Δε Hf in the Itasy magmas argues against the involvement of ancient recycled oceanic crust or stranded lower crust in their sources, but the presence of a low µ (LOMU) component in some Itasy magmas can be attributed to pollution of the asthenosphere by delaminated subcontinental lithospheric mantle. A simple two-stage, three-component mixing model indicates that a mixture of mantle containing 90% depleted asthenosphere (DMM) and 10% foundered SCLM, which is further mixed with 30 to 50% of a plume-type mantle component, can reproduce the isotopic compositions of the Itasy magmas. Based on our model, it is possible that easily fusible superplume materials might be widely distributed in the upper mantle underneath the EARS, from the Indian Ocean to the Red Sea-Gulf of Aden, and mixed with shallow asthenosphere that has itself incorporated eroded SCLM. These components are heterogeneously distributed at a small spatial scale, similar to that observed recently in other mafic monogenetic volcanic fields in both intraplate and subduction settings. [ABSTRACT FROM AUTHOR]

[Display omitted] Molybdenum (Mo) isotopes can be fractionated during redox-related hydrothermal and Earth's surface processes. Distinct Mo isotope features of different reservoirs, thus, make Mo isotopes as a potential tracer for the recycling of crustal materials, particularly subduction sediments. Here, we report the Mo isotope compositions of 61 early Cretaceous high-K calc-alkaline rocks (defined as adakitic rocks) with distinct source affinities from central-eastern China, including low-Mg adakitic rocks from the Dabie orogen, ore-barren high-Mg adakitic rocks from the South Tan-Lu Fault and ore-bearing high-Mg adakitic rocks from the Lower Yangtze River Belt. Low-Mg adakitic rocks have the low δ98Mo NIST3134 of −0.48‰ to −0.03‰, and both ore-barren and ore-bearing high-Mg adakitic rocks display heavier isotope compositions and larger variations: for example, −0.38‰ to 0.41‰ and −0.58‰ to 1.39‰, respectively. Magma evolutions (e.g., hornblende crystallizations) appear to dominate the Mo isotope variation in low-Mg adakitic rocks but do not account for those in high-Mg adakitic rocks. Some ore-bearing high-Mg adakitic rocks characterized by high S contents and hydrothermal alteration minerals, display positive correlations between S and Mo, Cu contents, δ98Mo values, reflecting the effects of isotopically heavy hydrothermal fluid. Having eliminated the effects of magmatic evolution and hydrothermal alteration, both ore-barren and ore-bearing high-Mg adakitic rocks display multiple-end-member mixing trends according to the correlations of Ce/Mo, Th/Yb, and Ba/Th ratios to δ98Mo. First, a common end-member with extremely low Ce/Mo, low Th/Yb, high Ba/Th and mantle-like δ98Mo values was identified, representing an enriched mantle origin metasomatized by fluids from subduction oxidized sediments. The other end-member of ore-barren high-Mg adakitic rocks displays high Ce/Mo, Ba/Th ratios and low δ98Mo values similar to low-Mg adakitic rocks, thus, it represents the lower continental crust derived melt component. By contrast, the other end-member of ore-bearing high-Mg adakitic rocks displays extremely high δ98Mo (>1.5‰), implying that initial adakitic melts involve reduced sediments in addition to the oceanic mafic crust. Notably, the common enriched mantle end-member for both ore-barren and ore-bearing adakites is characterized by high Mo and Cu concentrations, implying that the incorporation of oxidized sediments into the mantle likely plays a key role in the initial enrichment of the chalcophile metals in the mantle. This work demonstrates that Mo isotopes can be useful for understanding adakitic source affinities and associated Cu-Au-Mo ore-forming processes. [ABSTRACT FROM AUTHOR]

• Li-isotopic composition of altered lavas from DSDP Sites 417–418 reported. • Very light (−2‰) upper lavas and heavier (5–10‰) deeper lavas. • Different sedimentation histories of cores affect [Li] but not δ7Li. • Changing hydrological conditions during alteration required to explain data. The record of the Li-isotopic composition of ancient-seawater has the potential to provide important insight into the Earth system. However, we still have an incomplete understanding of the controls on the magnitude and isotopic compositions of the fluxes of Li into and out of the ocean. An important sink for Li from seawater is through low-temperature hydrothermal alteration of the upper oceanic crust. Here we present a detailed study of three adjacent drill cores in 120 Myr old Atlantic crust (Holes 417A, 417D and 418A) that had very different sedimentation histories. Samples from Hole 417A, drilled on a topographic high and exposed to seawater for >20 Myr before sediment started to accumulate, have much higher Li contents (up to ∼70 ppm) than samples from the other sites where sediment began accumulating immediately after crustal accretion (up to ∼20 ppm). Despite this difference in Li content, the δ7Li of the altered basalts in all cores increases with depth over the upper ∼40 m from around −1‰ at the lava-sediment interface to 6 ± 2‰ at ∼40 m depth, remaining roughly constant with depth after this. If the uppermost lavas equilibrated with seawater, with a bulk fractionation factor of ∼0.980 ± 0.003, then the isotopic composition of seawater 120 Myr ago was ∼19 ± 3‰, within the range of other estimates for Cretaceous seawater. Fluid-rock reaction models, with the hydrological regime varying as a function of depth and sediment accumulation history, can explain the observed rock compositions. However, modelling shows that complexities related to the time-varying hydrological regime, along with the possibility of time-varying δ7Li of seawater, mean that the Li-isotopic composition of altered oceanic lavas cannot be uniquely interpreted. The bulk Li content and isotopic composition of the well-sampled 120 Myr crust in Holes 417A, 417D and 418A suggest that Cretaceous upper oceanic crust was a sink for ∼4×109 mol yr−1 of Li. Surprisingly, the bulk Li-isotopic composition of the lavas sampled at these sites (∼6‰) is not much higher than that of fresh MORB, and within uncertainty of that of high-temperature hydrothermal vent fluids; i.e., high- and low-temperature alteration processes approximately balanced one another for Li-isotopes in the late Cretaceous. This could be fortuitous; alternatively, if the ocean was at steady state the main source of Li to the ocean at this time may have had a δ7Li similar to that of high-temperature hydrothermal fluids. [ABSTRACT FROM AUTHOR]

Constraining the behaviour of Re and Os during eclogite melting is required to understand the Re and Os budget and 187Os/188Os of recycled slabs produced at warm subduction zones. It is particularly relevant to early Earth history, a period during which slab melting could have prevailed over dehydration due to higher mantle temperatures. There are however currently few constraints on Re and Os mobility during slab melting. Accordingly, we measured Os, Re and 187Os/188Os in primitive submarine lavas (Mg# ˃ 0.6) from the western Aleutian Arc. These include strongly adakitic rocks shown to be derived from eclogite melting (high-Mg# andesite, dacites and rhyodacites), as well as non-adakitic rocks (high-Mg# andesites, basaltic andesites and basalts) with variable sediment and fluid-derived slab contributions for comparison. The 187Os/188Os of the adakitic and non-adakitic volcanic rocks vary significantly but largely overlap. In both groups, the most radiogenic values occur in samples with the lowest Os concentrations, thus implicating crustal assimilation as the main cause of Os isotope variations. Adakitic and non-adakitic rocks least affected by crustal assimilation have overlapping 187Os/188Os of 0.141–0.149. We show that the source of the adakites is very unlikely to comprise significant eclogite-derived Os, which suggests no or minimal mobilization of Os during eclogite melting. Eclogitic Os is inferred to be retained in sulphides or replacement phases formed upon sulphide breakdown for which Os has high affinity, such as a platinum-group minerals (PGMs). The small Os budget of the adakites is most likely derived from limited reaction with the mantle wedge during ascent. Degassing has reduced Re contents in most samples, but not for end-member adakites (SiO 2 > 67% and Sr/Y > 200; n = 4) that were erupted at seafloor depths > 2500 m. These undegassed samples have elevated Re concentrations (0.8–1.5 ppb) that are positively correlated with Sr/Y and so are interpreted to be primary magmatic concentrations resulting from the mobilization of Re from the slab. Re could either be derived from the eclogites or from the serpentinite-derived fluids fluxing eclogites during melting. The former scenario would produce recycled residual crusts with lower Re/Os than in unmelted eclogites while the latter would result in Re/Os ranging from similar to higher than prior to melting. In both cases, the Re/Os and therefore the time-integrated 187Os/188Os of residual crust produced at warm subduction zones involving slab melting are likely to be different from that processed at cooler typical modern subduction zones. Therefore, if slab melting was an important process during the early Earth, the use of Re and Os partitioning in modern subduction zones to model the source of magmas comprising old recycled oceanic crust, such as the HIMU (high μ = 238U/204Pb) ocean island basalts (OIBs), might lead to erroneous interpretations. [ABSTRACT FROM AUTHOR]

• Zircon Hf-O isotopes for three Fe-Ti-V oxide deposits in the Emeishan LIP. • Hongge parental magmas experienced <20% contamination with Precambrian metamorphic rocks. • Carbonate assimilation is detected in the Panzhihua and Taihe deposits. • High-temperature altered oceanic gabbroic crust involved in the plume source of Taihe. Magma-carbonate interaction and recycled oceanic crust in the mantle source are thought to have played a critical role in the genesis of magmatic Fe-Ti-V oxide ore deposits in the Emeishan mantle plume-related large igneous province in southwestern China. To test these hypotheses, we have carried out a combined study of zircon Hf-O isotopes and whole-rock Sr-Nd isotopes on three representative magmatic Fe-Ti-V oxide ore deposits associated with different types of wall-rock in the province: the Hongge and Taihe deposits with granitoid wall-rocks in most places, and the Panzhihua deposit with marble wall-rocks. Diopside-garnet marble xenoliths are common in the host gabbros of the Taihe deposit. The zircon separates used in this study are all from the mineralized units of the host layered mafic–ultramafic intrusions. The selected zircon grains yield U-Pb isotopic ages of 258.2 ± 2.8 Ma for Hongge intrusion, 263.3 ± 2.2 Ma for Taihe intrusion, and 257.6 ± 2.1 Ma for Panzhihua intrusion. The Hf-O isotopes of the zircon separates are ε Hf (t) from 3.1 to 4.7 and δ18O from 5.77 to 6.45‰ for Hongge, ε Hf (t) from 8.3 to 9.8 and δ18O from 4.12 to 4.56‰ for Taihe, and ε Hf (t) from 4.6 to 6.3 and δ18O from 5.49 to 6.10‰ for Panzhihua. The zircon δ18O values of the Taihe deposit are significantly lower than the mantle zircon values (5.3 ± 0.3‰), whereas those of the Hongge and Panzhihua deposits are higher than the mantle values. The initial 87Sr/86Sr and ε Nd (t) of the whole rock samples from the deposits are similar, from 0.7047 to 0.7052 and from 2.3 to 3.1, respectively. The observed decoupling of zircon Hf-O isotopes (low δ18O and high ε Hf (t) values) in the Taihe deposit is most consistent with a high-temperature altered, low δ18O oceanic gabbroic component in the plume source for this deposit. All of the isotope data together show that the parental magmas for these deposits experienced different types of crustal contamination. Contamination with the Precambrian metamorphic country rocks is much more significant in the parental magma for the Hongge deposit than those for the Panzhihua and Taihe deposits. On the other hand, contamination with marble country rocks is more pronounced in the parental magmas for the Panzhihua and Taihe deposits than in the parental magma for the Hongge deposit. An important new finding from this study is that these deposits are not linked to the same component of the mantle plume nor a single type of crustal contamination. [ABSTRACT FROM AUTHOR]

Dehydration of the subducting slab is a crucial process in the generation of hydrous convergent margin magmas, yet the exact processes of how and where the slab dehydrates and how these fluids are transported to the mantle wedge remain obscure. Strontium is a "fluid-mobile" element and as such well suited to investigate the source of slab-derived fluids. We employ mass-dependent Sr isotope systematics (δ88/86Sr; the deviation in 88Sr/86Sr of a sample relative to NIST SRM 987) of primitive arc lavas, in tandem with conventional radiogenic 87Sr/86Sr measurements, as a novel tracer of slab dehydration. To characterise the δ88/86Sr composition of subduction zone inputs, we present new δ88/86Sr data for subducting sediments, altered oceanic crust and MORB. Calcareous sediments are isotopically lighter and carbonate-free sediments are isotopically heavier than mid-ocean ridge basalts (MORB). Samples of the altered oceanic crust display elevated 87Sr/86Sr but only the most intensely altered sample has significantly higher δ88/86Sr than pristine MORB. Mafic arc lavas from the Aegean and Mariana arc invariably have a mass-dependent Sr isotope composition that is indistinguishable from MORB and lower 87Sr/86Sr than upper altered oceanic crust. This δ88/86Sr-87Sr/86Sr signature of the arc lavas, in combination with their high but variable Sr/Nd, can only be explained if it is provided by a fluid that acquired its Sr isotope signature in the deeper, less altered part of the subducted oceanic crust. We propose a model where the breakdown of serpentinite in the slab mantle releases a pulse of fluid at sub-arc depths. These fluids travel through and equilibrate with the overlying oceanic crust and induce wet partial melting of the upper altered crust and sediments. This hydrous melt is then delivered to the mantle source of arc magmas as a single metasomatic component. From mass balance it follows that the slab-derived fluid contributes >70% of the Sr budget of both Mariana and Aegean arc lavas. Whereas this fluid-dominated character is unsurprising for the sediment-poor Mariana arc, the Aegean arc sees the subduction of 3–6 km of calcareous sediments that were found to exert very little control on the Sr budget of the arc magmas and are overwhelmed by the fluid contribution. [ABSTRACT FROM AUTHOR]

• Fluid signatures in Izu arc lavas associated with isotopically heavy Mo. • Slab-derived fluids have heavy Mo isotopic composition of δ98/95Mo = 0.1–0.25‰. • Co-variations with radiogenic isotopes link fluid input and mantle heterogeneity. • Stronger control of slab fluids over more depleted mantle sources. • Heavy Mo lost through slab fluids leaves light δ98/95Mo signature in residual slab. Molybdenum isotope variations in mafic arc lavas have mainly been attributed to the influence of slab-derived components, such as subducted sediment melts and aqueous fluids. The latter have been hypothesised to fractionate Mo isotopes through interaction with the oceanic crust and carry an isotopically heavy signal that is transferred to the source of arc magmas. Thus, understanding Mo isotope systematics in subduction zones requires characterising the Mo isotope composition of slab-derived fluids and their influence on the Mo isotope budget of arc magmas. However, Mo isotope data reported to date show a considerable influence from subducted sediments that complicate accurate constraints being placed on the fluid contribution. We present Mo isotope data for mafic lavas from the Izu arc, a highly depleted oceanic island arc whose magma compositions show a dominant control from slab-derived fluids. The lavas from the Izu volcanic front are isotopically heavier than MORB and the depleted mantle. Their δ98/95Mo (the relative difference in measured 98Mo/95Mo to NIST 3134) systematically varies with indicators for fluid-mobile element enrichment, suggesting that slab-derived fluids in the Izu arc have heavy Mo isotope compositions. Additionally, co-variations with radiogenic 143Nd/144Nd and 176Hf/177Hf point to a relationship between the addition of aqueous fluids and compositional heterogeneity of the sub-arc mantle. We present mass balance models that show that the influence of subduction zone fluids on the trace element pattern of arc magmas is more dominant when these are added to a more depleted and refractory sub-arc mantle, which preferentially melts due to a relatively higher fluid flux. The mass balance of Mo in the Izu arc predicts a light Mo isotope composition for the residual oceanic crust as a result of the preferential removal of isotopically heavy Mo during slab dehydration, consistent with previous suggestions for the Mariana arc and isotopically light Mo previously reported for eclogites. [ABSTRACT FROM AUTHOR]

The westward subduction of the Pacific plate interacted with the northward subduction of the eastern Neo-Tethys plate in the Early Cretaceous. However, the subduction polarities between these two plates in the Early Cretaceous is not well constrained. The proto-Philippine arc is suggested to have originated at the Pacific – Neo-Tethys intersection in the SW Pacific in the latest Jurassic – Early Cretaceous, and can provide information about the ancient interaction between these two plates. In this study, the oldest known Early Cretaceous mantle-derived arc volcanic rocks and slab-melting generated adakites occurring in the Philippine arc have been studied for Pb, Sr, Nd and Hf isotopes to discriminate the provenance of their mantle sources and/or subduction components. The 126–118 Ma Kansi island arc volcanics have Pacific-type Pb isotopes, suggesting that the subduction components in the source are mainly from the subducted Pacific plate. Their Hf-Nd isotopes give support to a mantle source that is isotopically transitional Pacific-Indian type. The formation of the Kansi volcanics dates the subduction of the Pacific plate under the transitional Pacific-Indian mantle as Early Cretaceous. The subsequent generation of the 110 Ma Kansi adakitic diorites with the similar Pacific-type Pb isotopes indicates further partial melting of the subducted Pacific plate. In contrast, the post-110 Ma Lutopan adakitic diorites have Indian-type Pb isotopes, indicating the subsequent subduction of an Indian-type oceanic slab in the eastern Neo-Tethys tectonic regime. Therefore, the proto-Philippine arc is constrained to have originated at the intersection between Indian-type and Pacific-type oceanic crusts and mantle domains. The successive generation of these Early Cretaceous arc rocks in the Philippines indicates that a transformation of the tectonic regime from southwestward Pacific subduction to northeastward subduction of the eastern Neo-Tethys plate in the SW Pacific Ocean occurred at ca. 110 Ma, due to the abrupt change in drifting direction of the Pacific plate. [ABSTRACT FROM AUTHOR]

Subduction zones are critical sites for carbon cycling between Earth's surface and interior. However, how subducted carbon is released and transferred to the surface is not well understood, especially regarding the role of slab-derived fluids in the deep carbon cycle. Here we report Mo and B isotopic data for the Silurian normal arc andesites and adakitic andesites from the Chinese North Tianshan, which represent partial melts of fluid-modified mantle wedge and dehydrated oceanic crust, respectively. The normal arc andesites yielded δ98Mo values (0.33–1.08‰) significantly higher than that (about –0.20‰) of the depleted mantle. Because their limited range of SiO 2 (53.8–55.3 wt.%) precludes differentiation as a cause for their variable δ98Mo values and Mo isotopic fractionation solely by fluid mobilization is limited (≤0.3‰), the elevated δ98Mo values could be ascribed to the incorporation of crustal material with heavy Mo isotopes in the mantle source. Since marine carbonate is featured by both heavy Mo and B isotopes and our normal arc andesites also give heavy δ11B (−1.63 to +4.00‰) values, we consider that marine carbonate was possibly involved as a component of the subducted slab, which modified Mo–B isotopic compositions of the mantle source. The positive correlations between δ98Mo and δ11B and between δ98Mo and Ba/Rb suggest transport of subducted carbonates to the overlying mantle wedge via slab fluids, thus providing robust evidence for transfer of subducted carbon to the overriding plate by carbonate dissolution. In contrast, the younger adakitic andesites have light δ98Mo (−0.48 to −0.27‰) and δ11B (−9.43 to −2.05‰) values, implying an isotopically Mo- and B-light source. Given the preferential transport of heavy 98Mo and 11B to the fluid phase during slab dehydration, their remarkably light δ98Mo and δ11B values support a dehydrated oceanic crust as their magma source. The contrasting Mo–B isotopes for such two kinds of andesites highlight that most carbonates can be removed from the subducted slab to the overriding plate during oceanic subduction. [ABSTRACT FROM AUTHOR]

Dehydration of subducting oceanic crust and underlying serpentinized mantle is potentially the most important source of fluids for element cycling between surface-processed materials and the interior of the Earth. Magnesium (Mg) isotopes -have been proposed as a promising tracer of dehydration of serpentinites because of their high Mg contents and distinctive Mg isotopic signatures relative to the mantle, which is difficult to identify using other isotopes. To constrain the primary controls on Mg isotope cycling in subduction zones, here we analyze a suite of well-characterized sub-arc mantle xenoliths from the Avacha volcano in southern Kamchatka arc, Russia. They are comparable in modal and/or chemical composition to serpentinized fore-arc harzburgites and sub-arc harzburgite xenoliths from the western Pacific. Despite evidence for ubiquitous slab-related fluid metasomatism, the 23 spinel harzburgite xenoliths and four pyroxenite metasomatic veins display mantle-like δ26Mg values from −0.30 to −0.21‰. Mineral separates have similar δ26Mg values, with an average of −0.26 ± 0.04‰ (2SD, n = 17) for olivine, −0.23 ± 0.04‰ (2SD, n = 17) for orthopyroxene, and −0.24 ± 0.10‰ (2SD, n = 5) for late-stage interstitial clinopyroxene. The lack of a clear slab-derived δ26Mg signature in the Avacha sub-arc peridotites is probably due to their high MgO contents and low fluid/rock mass ratios during flux melting and metasomatism. We also compare published δ26Mg data for four volcanic arcs distributed worldwide. This inter-arc comparison reveals that the thermal structure of a subduction zone appears to exert a strong control on the efficiency of Mg isotope cycling from the downgoing slab to the mantle wedge and subsequently to the arc magmas; those arcs that permit large-scale dehydration of Mg-rich serpentinized slab mantle, coupled with available channelized fluid pathways, are more likely to produce isotopically diverse arc magmas. [ABSTRACT FROM AUTHOR]