Selenium isotope evidence for pulsed flow of oxidative slab fluids
Affiliations | Corresponding Author | Cite as | Funding information
Published by the European Association of Geochemistry
under Creative Commons License CC BY-NC-ND 4.0
Keywords: selenium isotopes, sulfur, sulfides, recycling, subduction, redox, slab fluids
- Share this article
Article views:467Cumulative count of HTML views and PDF downloads.
- Download Citation
- Rights & Permissions
top
Abstract
Figures
 Figure 1 Combined S and Se Pourbaix diagrams for 25 °C and 1 bar pressure. Large arrow shows required minimum redox variation for selenite-selenide transition that requires higher redox variation than the sulfide-sulfate transition for slab fluids with pH of up to 9 in subduction zones (Galvez et al., 2016). |  Figure 2 Selenium isotope composition of Raspas samples and AOC from 1256D vs. (a) Pb/Ce, (b) Rb/Ce, and (c) Cs/Ce (Halama et al., 2010; John et al., 2010), and (d) δ98/95Mo variation. Mantle includes representative peridotites (Varas-Reus et al., 2019) and MORB (Yierpan et al., 2020). The large grey arrow (Figs. 2, 3) represents our interpretation of the effects of slab mantle dehydration and increasing interaction of subducted crust with slab fluids. The source of Mo is HP serpentinite (Chen et al., 2019), whereas Se is derived mostly from the AOC. As HP serpentinites dehydrate and flush the overlying subducted AOC, rutiles in the subducted crust preferentially retain low δ98/95Mo as the complementary, high δ98/95Mo fluid is removed into the mantle wedge (Freymuth et al., 2015; Chen et al., 2019). Recrystallising, early sulfides preferentially incorporate light Se isotopes and later stage sulfides become isotopically heavier. Repeated sulfide dissolution-recrystallisation produces a spectrum of low and high δ82/76Se in the flushed AOC. |  Figure 3 BSE images and semi-quantitative element mapping of sulfides in Raspas eclogites: (a) a partially resorbed Co-bearing pyrite, showing high Cu and Fe contents at the recrystallised chalcopyrite rim and (b) an example of a chalcopyrite with a narrow and oxidised rim, indicated by the lower S and Cu contents. Trends of δ82/76Se vs. concentrations of (c) Cl, (d) Sc and (e) Cu in prograde eclogites are interpreted to result from this changing fluid-mineral chemistry. Symbols are the same as those used in Figure 2. |
| Figure 1 | Figure 2 | Figure 3 |
top
Introduction
The development of a habitable Earth with its atmosphere-ocean system is closely linked to plate tectonics and the efficient exchange of elements between interior and exterior reservoirs. Key aspects are the subduction cycles of water, carbon and sulfur and the redox state of arc magmas. It is particularly debated if the oxidised nature of arc magmas is due to oxidised slab components or related to secondary processes during the evolution of arc melts (e.g., Lee et al., 2012
Lee, C.-T., Luffi, A.P., Chin, E.J., Bouchet, R., Dasgupta, R., Morton, D.M., Le Roux, V., Yin, Q.-Z., Jin, D. (2012) Copper systematics in arc magmas and implications for crust-mantle differentiation. Science 336, 64–68.
; Chen et al., 2019Chen, S., Hin, R.C., John, T., Brooker, R., Bryan, B., Niu, Y., Elliott, T. (2019) Molybdenum systematics of subducted crust record reactive fluid flow from underlying slab serpentine dehydration. Nature Communications 10, 4773.
; Tollan and Hermann, 2019Tollan, P., Hermann, J. (2019) Arc magmas oxidized by water dissociation and hydrogen incorporation in orthopyroxene. Nature Geoscience 12, 667–671.
). Likewise, different models for sulfate- and 34S-enriched arc lavas involve either slab-derived sulfur flux into the overlying mantle wedge or crustal assimilation by ascending magma (e.g., Lee et al., 2012Lee, C.-T., Luffi, A.P., Chin, E.J., Bouchet, R., Dasgupta, R., Morton, D.M., Le Roux, V., Yin, Q.-Z., Jin, D. (2012) Copper systematics in arc magmas and implications for crust-mantle differentiation. Science 336, 64–68.
; Pons et al., 2016Pons, M.L., Debret, B., Bouilhol, P., Delacour, A., Williams, H. (2016) Zinc isotope evidence for sulfate-rich fluid transfer across subduction zones. Nature Communications 7, 13794.
). Recent in situ S isotope investigations of sulfides from exhumed high pressure rocks reveal prograde subduction related mobilisation and re-entrapment of S during (partial) sulfide breakdown and recrystallisation, resulting in differences in δ34S of almost 40 ‰ (Evans et al., 2014Evans, K.A., Tomkins, A.G., Cliff, J. Fiorentini, M.L. (2014) Insights into subduction zone sulfur recycling from isotopic analysis of eclogite-hosted sulphides. Chemical Geology 365, 1–19.
; Su et al., 2019Su, W., Schwarzenbach, E., Chen, L., Li, Y., John, T., Gao, J., Chen, F., Hu, X. (2019) Sulfur isotope compositions of pyrite from high-pressure metamorphic rocks and related veins (SW Tianshan, China): Implications for the sulfur cycle in subduction zones. Lithos 348-349, 105212.
; Li et al., 2020Li, J.-L., Schwarzenbach, E.M., John, T., Ague, J.J., Huang, F., Gao, J., Klemd, R., Whitehouse, M.J., Wang, X.-S. (2020) Uncovering and quantifying the subduction zone sulfur cycle from the slab perspective. Nature Communications 11, 514.
; Walters et al., 2020Walters, J.B., Cruz-Uribe, A.M., Marschall, H.R. (2020) Sulfur loss from subducted altered oceanic crust and implications for mantle oxidation. Geochemical Perspectives Letters 13, 36–41.
), with contrasting data for different localities suggesting either oxidised, SO2- and CO2-rich or reduced, H2S-rich fluid involvement (see also Piccoli et al., 2019Piccoli, F., Hermann, J., Pettke, T., Connolly, J.A.D., Kempf, E.D., Vieira Duarte, J.F. (2019) Subducting serpentinites release reduced, not oxidized, aqueous fluids. Scientific Reports 9, 19573.
).To further elucidate the redox role of slab-derived components, we combine the novel tool of stable selenium isotopes with a suite of well characterised high pressure rocks of oceanic origin. In comparison to other redox sensitive elements like Mo, S and Fe, Se combines the characteristics of being chalcophile and a trace element that does not act as a major constituent of sulfide. Selenium, therefore, participates in sulfide dissolution-recrystallisation processes within the slab as a witness without being biased as a major constituent of a mineral. Moreover, the transition to selenite (oxidised Se species; SeO32−; see S-Se species stability fields in Fig. 1) occurs at even higher fO2 compared to the transition to sulfate (oxidised S species; SO42−). Thus, Se isotope fractionation reflects a strong reduction of at least selenite to selenide from originally higher fO2 than the sulfate-sulfide transition (Fig. 1). Selenium isotopes may thus be useful in constraining the conditions during slab fluid-mediated sulfide redistribution, in particular, if more pronounced redox variations occur in fluids that so far remained beyond the scope of S isotope systematics (König et al., 2019
König, S., Eickmann, B., Zack, T., Yierpan, A., Wille, M., Taubald, H., Schoenberg, R. (2019) Redox induced sulfur-selenium isotope decoupling recorded in pyrite. Geochimica et Cosmochimica Acta 244, 24–39.
) and previously employed, lithophile (e.g., Mo-Li-N-Tl) stable isotopes in subduction related studies.Figure 1 Combined S and Se Pourbaix diagrams for 25 °C and 1 bar pressure. Large arrow shows required minimum redox variation for selenite-selenide transition that requires higher redox variation than the sulfide-sulfate transition for slab fluids with pH of up to 9 in subduction zones (Galvez et al., 2016
Galvez, M.E., Connolly, J.A.D., Manning, C.E. (2016) Implications for metal and volatile cycles from the pH of subduction zone fluids. Nature 539, 420–424.
).The Raspas Complex in SW Ecuador resembles a deeply subducted (∼60 km) ophiolite complex that formed during the early Cretaceous at the South American margin. Trace element and radiogenic isotope data revealed that the ophiolite consists of depleted abyssal oceanic mantle, oceanic crust (MORB-type eclogites), associated seamounts (blueschists), and sedimentary cover (John et al., 2010
John, T., Scherer, E.E., Schenk, V., Herms, P., Halama, R., Garbe-Schonberg, D. (2010) Subducted seamounts in an eclogite-facies ophiolite sequence: the Andean Raspas Complex, SW Ecuador. Contributions to Mineralogy and Petrology 159, 265–284.
; Halama et al., 2011Halama, R., John, T., Herms, P., Hauff, F., Schenk, V. (2011) A stable (Li, O) and radiogenic (Sr, Nd) isotope perspective on metasomatic processes in a subducting slab. Chemical Geology 281, 151–166.
). The entire complex was metamorphosed at HP/LT conditions (∼600 °C and ca. 2.0 GPa) and most of the rocks experienced subduction related, prograde dehydration related intra-slab fluid flow during burial (John et al., 2010John, T., Scherer, E.E., Schenk, V., Herms, P., Halama, R., Garbe-Schonberg, D. (2010) Subducted seamounts in an eclogite-facies ophiolite sequence: the Andean Raspas Complex, SW Ecuador. Contributions to Mineralogy and Petrology 159, 265–284.
; Halama et al., 2011Halama, R., John, T., Herms, P., Hauff, F., Schenk, V. (2011) A stable (Li, O) and radiogenic (Sr, Nd) isotope perspective on metasomatic processes in a subducting slab. Chemical Geology 281, 151–166.
; Herms et al., 2012Herms, P., John, T., Bakker, R.J., Schenk, V. (2012) Evidence for channelized external fluid flow and element transfer in subducting slabs (Raspas Complex, Ecuador). Chemical Geology 310-311, 79–96.
; Chen et al., 2019Chen, S., Hin, R.C., John, T., Brooker, R., Bryan, B., Niu, Y., Elliott, T. (2019) Molybdenum systematics of subducted crust record reactive fluid flow from underlying slab serpentine dehydration. Nature Communications 10, 4773.
; Urann et al., 2020Urann, B.M., Le Roux, V., John, T., Beaudoin, G.M., Barnes, J.D. (2020) The distribution and abundance of halogens in eclogites: An in situ SIMS perspective of the Raspas Complex (Ecuador). American Mineralogist 105, 307–318.
). The Raspas complex thus represents an exceptional example of a rather complete sequence of an oceanic lithosphere that has experienced long lived oceanic subduction and is not compromised by potential effects of subsequent continent-continent collision.top
Samples
The samples analysed here comprise seven eclogites, one metapelite, and three high pressure serpentinites from the Raspas Complex (Supplementary Information Tables S-1 and S-2). They have been thoroughly investigated for their petrology and all samples record evidence for hydration prior to subduction, but have been mostly selected to avoid any detectable exhumation related overprint (John et al., 2010
John, T., Scherer, E.E., Schenk, V., Herms, P., Halama, R., Garbe-Schonberg, D. (2010) Subducted seamounts in an eclogite-facies ophiolite sequence: the Andean Raspas Complex, SW Ecuador. Contributions to Mineralogy and Petrology 159, 265–284.
; Halama et al., 2011Halama, R., John, T., Herms, P., Hauff, F., Schenk, V. (2011) A stable (Li, O) and radiogenic (Sr, Nd) isotope perspective on metasomatic processes in a subducting slab. Chemical Geology 281, 151–166.
; Herms et al., 2012Herms, P., John, T., Bakker, R.J., Schenk, V. (2012) Evidence for channelized external fluid flow and element transfer in subducting slabs (Raspas Complex, Ecuador). Chemical Geology 310-311, 79–96.
). A comprehensive major, trace element, Rb-Sr, Sm-Nd, Lu-Hf geochronological and stable Mo-Li-N-Tl isotope dataset is also available (e.g., Halama et al., 2010Halama, R., Bebout, G.E., John, T., Schenk, V. (2010) Nitrogen recycling in subducted oceanic lithosphere: The record in high- and ultrahigh-pressure metabasaltic rocks. Geochimica et Cosmochimica Acta 74, 1636–1652.
; John et al., 2010John, T., Scherer, E.E., Schenk, V., Herms, P., Halama, R., Garbe-Schonberg, D. (2010) Subducted seamounts in an eclogite-facies ophiolite sequence: the Andean Raspas Complex, SW Ecuador. Contributions to Mineralogy and Petrology 159, 265–284.
; Chen et al., 2019Chen, S., Hin, R.C., John, T., Brooker, R., Bryan, B., Niu, Y., Elliott, T. (2019) Molybdenum systematics of subducted crust record reactive fluid flow from underlying slab serpentine dehydration. Nature Communications 10, 4773.
; Su et al., 2019Su, W., Schwarzenbach, E., Chen, L., Li, Y., John, T., Gao, J., Chen, F., Hu, X. (2019) Sulfur isotope compositions of pyrite from high-pressure metamorphic rocks and related veins (SW Tianshan, China): Implications for the sulfur cycle in subduction zones. Lithos 348-349, 105212.
). We also analysed four serpentinites and gabbros from the deepest Pacific drillcore available (IODP-1256D; Table S-1). We combine this sample set with an average composition of marine sediment from the literature to provide a reliable estimate of subduction input signatures.top
Selenium Isotope Signatures of Prograde Metamorphic Rocks
Compared to the relatively small range for mantle samples including average MORB with δ82/76Se (δ82/76Se = [(82Se/76Se)sample/(82Se/76Se)NIST3149 – 1] × 1000) of −0.16 ± 0.12 ‰ (2 s.d.; Yierpan et al., 2019
Yierpan, A., König, S., Labidi, J., Schoenberg, R. (2019) Selenium isotope and S-Se-Te elemental systematics along the Pacific-Antarctic ridge: Role of mantle processes. Geochimica et Cosmochimica Acta 249, 199–224.
), peridotites of −0.03 ± 0.07 ‰ (Varas-Reus et al., 2019Varas-Reus, M.I., König, S., Yierpan, A., Lorand, J.-P., Schoenberg, R. (2019) Selenium isotopes as tracers of a late volatile contribution to Earth from the outer Solar System. Nature Geoscience 12, 779–782.
) and the limited variability observed in seawater overprinted altered ocean crust (AOC) with δ82/76Se of −0.35 to −0.07 ‰ (Pacific drillcore samples; Table S-1), eclogites record a large variation in δ82/76Se from −1.89 to +0.48 ‰ (Fig. 2; see Supplementary Information for analytical methods). High pressure serpentinites and metapelites show δ82/76Se of +0.07 to +0.39 ‰, overlapping with the highest δ82/76Se values of eclogites. δ82/76Se values show no systematic variations with Se concentrations (6 to 106 ng·g−1) and samples with the highest Se contents span the entire Se isotope range of the dataset. High pressure serpentinites show highest relative abundances of fluid mobile elements like Pb, Rb, Cs (Fig. 2a–c) or Ba, Li (not shown) when plotted as ratios with fluid immobile Ce as the denominator (see also Fig. S-1). All eclogites show a trend towards lower δ82/76Se with decreasing Pb/Ce, Rb/Ce and Cs/Ce (Fig. 2a). Eclogites with the lowest δ82/76Se sample also display the lowest δ98/95Mo (Fig. 2d; Mo data by Chen et al., 2019Chen, S., Hin, R.C., John, T., Brooker, R., Bryan, B., Niu, Y., Elliott, T. (2019) Molybdenum systematics of subducted crust record reactive fluid flow from underlying slab serpentine dehydration. Nature Communications 10, 4773.
) but two samples with only intermediate δ98/95Mo show the most positive δ82/76Se. Additionally, eclogites with the lowest δ82/76Se show the highest contents of Cu, Cl, Sc (Fig. 3) and N as well as the smallest δ15N range (Fig. S-2; Halama et al., 2010Halama, R., Bebout, G.E., John, T., Schenk, V. (2010) Nitrogen recycling in subducted oceanic lithosphere: The record in high- and ultrahigh-pressure metabasaltic rocks. Geochimica et Cosmochimica Acta 74, 1636–1652.
; Urann et al., 2020Urann, B.M., Le Roux, V., John, T., Beaudoin, G.M., Barnes, J.D. (2020) The distribution and abundance of halogens in eclogites: An in situ SIMS perspective of the Raspas Complex (Ecuador). American Mineralogist 105, 307–318.
). No correlations between δ82/76Se and Li or Tl isotopes are observed (Fig. S-2).Figure 2 Selenium isotope composition of Raspas samples and AOC from 1256D vs. (a) Pb/Ce, (b) Rb/Ce, and (c) Cs/Ce (Halama et al., 2010
Halama, R., Bebout, G.E., John, T., Schenk, V. (2010) Nitrogen recycling in subducted oceanic lithosphere: The record in high- and ultrahigh-pressure metabasaltic rocks. Geochimica et Cosmochimica Acta 74, 1636–1652.
; John et al., 2010John, T., Scherer, E.E., Schenk, V., Herms, P., Halama, R., Garbe-Schonberg, D. (2010) Subducted seamounts in an eclogite-facies ophiolite sequence: the Andean Raspas Complex, SW Ecuador. Contributions to Mineralogy and Petrology 159, 265–284.
), and (d) δ98/95Mo variation. Mantle includes representative peridotites (Varas-Reus et al., 2019Varas-Reus, M.I., König, S., Yierpan, A., Lorand, J.-P., Schoenberg, R. (2019) Selenium isotopes as tracers of a late volatile contribution to Earth from the outer Solar System. Nature Geoscience 12, 779–782.
) and MORB (Yierpan et al., 2020Yierpan, A., König, S., Labidi, J., Schoenberg, R. (2020) Recycled selenium in hot spot-influenced lavas records ocean-atmosphere oxygenation. Science Advances 6, eabb6179.
). The large grey arrow (Figs. 2, 3) represents our interpretation of the effects of slab mantle dehydration and increasing interaction of subducted crust with slab fluids. The source of Mo is HP serpentinite (Chen et al., 2019Chen, S., Hin, R.C., John, T., Brooker, R., Bryan, B., Niu, Y., Elliott, T. (2019) Molybdenum systematics of subducted crust record reactive fluid flow from underlying slab serpentine dehydration. Nature Communications 10, 4773.
), whereas Se is derived mostly from the AOC. As HP serpentinites dehydrate and flush the overlying subducted AOC, rutiles in the subducted crust preferentially retain low δ98/95Mo as the complementary, high δ98/95Mo fluid is removed into the mantle wedge (Freymuth et al., 2015Freymuth, H., Vils, F., Willbold, M., Taylor, R.N., Elliott, T. (2015) Molybdenum mobility and isotopic fractionation during subduction at the Mariana arc. Earth and Planetary Science Letters 432, 176–186.
; Chen et al., 2019Chen, S., Hin, R.C., John, T., Brooker, R., Bryan, B., Niu, Y., Elliott, T. (2019) Molybdenum systematics of subducted crust record reactive fluid flow from underlying slab serpentine dehydration. Nature Communications 10, 4773.
). Recrystallising, early sulfides preferentially incorporate light Se isotopes and later stage sulfides become isotopically heavier. Repeated sulfide dissolution-recrystallisation produces a spectrum of low and high δ82/76Se in the flushed AOC.Figure 3 BSE images and semi-quantitative element mapping of sulfides in Raspas eclogites: (a) a partially resorbed Co-bearing pyrite, showing high Cu and Fe contents at the recrystallised chalcopyrite rim and (b) an example of a chalcopyrite with a narrow and oxidised rim, indicated by the lower S and Cu contents. Trends of δ82/76Se vs. concentrations of (c) Cl, (d) Sc and (e) Cu in prograde eclogites are interpreted to result from this changing fluid-mineral chemistry. Symbols are the same as those used in Figure 2.
top
Selenium Isotope Vestige of Slab Dehydration
Selenium isotopes do not fractionate significantly during melting, fractional crystallisation and metasomatism (Varas-Reus et al., 2019
Varas-Reus, M.I., König, S., Yierpan, A., Lorand, J.-P., Schoenberg, R. (2019) Selenium isotopes as tracers of a late volatile contribution to Earth from the outer Solar System. Nature Geoscience 12, 779–782.
; Yierpan et al., 2019Yierpan, A., König, S., Labidi, J., Schoenberg, R. (2019) Selenium isotope and S-Se-Te elemental systematics along the Pacific-Antarctic ridge: Role of mantle processes. Geochimica et Cosmochimica Acta 249, 199–224.
). The δ82/76Se range of eclogites and clear positive covariations of δ82/76Se with several fluid proxies (Fig. 2a–c) therefore suggest a non-magmatic origin of these signatures. Moreover, since the eclogite facies mineral assemblages show no evidence of subsequent retrograde greenschist facies overprint (except for one sample selected here; John et al., 2010John, T., Scherer, E.E., Schenk, V., Herms, P., Halama, R., Garbe-Schonberg, D. (2010) Subducted seamounts in an eclogite-facies ophiolite sequence: the Andean Raspas Complex, SW Ecuador. Contributions to Mineralogy and Petrology 159, 265–284.
; Herms et al., 2012Herms, P., John, T., Bakker, R.J., Schenk, V. (2012) Evidence for channelized external fluid flow and element transfer in subducting slabs (Raspas Complex, Ecuador). Chemical Geology 310-311, 79–96.
), the δ82/76Se and their preserved fluid related trends must be attributed to dehydration processes prior to exhumation. Analysed AOC extends to slightly lower δ82/76Se and the high pressure serpentinites to slightly higher δ82/76Se than mantle, while both assemblages maintain a much narrower range than the eclogites. Light S isotope compositions of pyrite in subducted rocks from SW Tianshan, China could be traced back to the alteration of oceanic crust and subsequent, fluid driven sulfide recrystallisation events with prograde metamorphism (Su et al., 2019Su, W., Schwarzenbach, E., Chen, L., Li, Y., John, T., Gao, J., Chen, F., Hu, X. (2019) Sulfur isotope compositions of pyrite from high-pressure metamorphic rocks and related veins (SW Tianshan, China): Implications for the sulfur cycle in subduction zones. Lithos 348-349, 105212.
). Such abyssal seawater-rock interaction therefore also explains the low average δ82/76Se in our AOC, due to scavenging related partial reduction of preferentially light Se isotopes in an entirely oxic water column (e.g., Johnson, 2004Johnson, T.M. (2004) A review of mass-dependent fractionation of selenium isotopes and implications for other heavy stable isotopes. Chemical Geology 204, 201–214.
). In conclusion, following subduction of AOC with moderately low δ82/76Se, higher δ82/76Se of the high pressure serpentinites as well as the large δ82/76Se range of the eclogites are best explained by slab dehydration during prograde metamorphism at the blueschist-eclogite transition.top
Role of Oxidised Slab Fluids
In fluid-solid systems, isotopes of Se are significantly fractionated during partial reduction of oxidised Se6+ to Se4+ or further to Se2− (Johnson, 2004
Johnson, T.M. (2004) A review of mass-dependent fractionation of selenium isotopes and implications for other heavy stable isotopes. Chemical Geology 204, 201–214.
). Sulfides thus preferentially incorporate isotopically light Se (e.g., König et al., 2019König, S., Eickmann, B., Zack, T., Yierpan, A., Wille, M., Taubald, H., Schoenberg, R. (2019) Redox induced sulfur-selenium isotope decoupling recorded in pyrite. Geochimica et Cosmochimica Acta 244, 24–39.
). Even at 25 °C and 1 atm, the transition from Se2- to Se0 and Se4+ starts at higher redox potential (Eh) and thus require higher fO2 than the sulfide-sulfate transition (Fig. 1). This difference remains with slightly elevated slab fluid pH of up to 9 in subduction zones (Galvez et al., 2016Galvez, M.E., Connolly, J.A.D., Manning, C.E. (2016) Implications for metal and volatile cycles from the pH of subduction zone fluids. Nature 539, 420–424.
). At present, Se6+/∑Se equilibrium values for substantially higher temperatures and pressures are unavailable and therefore the quantification of fO2 from the inferred presence of oxidised Se in slab fluids requires further experimental work. However, such experimentally derived constants exist for S6+/∑S and were previously used to infer a range of slab fluid FMQ of +1.0 to +1.4 at 1.5 GPa from observed sulfate as well as anhydrite saturation in sub-arc mantle peridotites and melt inclusions (Bénard et al., 2018Bénard, A., Klimm, K., Woodland, A.B., Arculus, R.J., Wilke, M., Botcharnikov, R.E., Shimizu, N., Nebel, O., Rivard, C., Ionov, D.A. (2018) Oxidising agents in sub-arc mantle melts link slab devolatilisation and arc magmas. Nature Communications 9, 3500.
). The sulfate-sulfide transition at conditions relevant to the Raspas eclogite suite at 2 GPa and 600 °C requires ca. FMQ +2.5. If significant ΔEh between lower sulfate and selenite stabilities persist at 600 °C and 2 GPa, the occurrence of oxidised Se points to above FMQ +2 for slab fluids involved here. This would agree with previous estimates for slab fluid fO2 above FMQ +2 based on Mo isotope data of Raspas eclogites (Chen et al., 2019Chen, S., Hin, R.C., John, T., Brooker, R., Bryan, B., Niu, Y., Elliott, T. (2019) Molybdenum systematics of subducted crust record reactive fluid flow from underlying slab serpentine dehydration. Nature Communications 10, 4773.
). More accurate constraints require experimental work, yet Se isotope data can already be attributed to more pronounced minimum redox variations than those required for stability of oxidised S and even Mo in the subducting slab, as supported by other studies arguing for variably pronounced oxidation potentials of fluids in subduction zones (e.g., Bénard et al., 2018Bénard, A., Klimm, K., Woodland, A.B., Arculus, R.J., Wilke, M., Botcharnikov, R.E., Shimizu, N., Nebel, O., Rivard, C., Ionov, D.A. (2018) Oxidising agents in sub-arc mantle melts link slab devolatilisation and arc magmas. Nature Communications 9, 3500.
; Chen et al., 2019Chen, S., Hin, R.C., John, T., Brooker, R., Bryan, B., Niu, Y., Elliott, T. (2019) Molybdenum systematics of subducted crust record reactive fluid flow from underlying slab serpentine dehydration. Nature Communications 10, 4773.
; Piccoli et al., 2019Piccoli, F., Hermann, J., Pettke, T., Connolly, J.A.D., Kempf, E.D., Vieira Duarte, J.F. (2019) Subducting serpentinites release reduced, not oxidized, aqueous fluids. Scientific Reports 9, 19573.
; Li et al., 2020Li, J.-L., Schwarzenbach, E.M., John, T., Ague, J.J., Huang, F., Gao, J., Klemd, R., Whitehouse, M.J., Wang, X.-S. (2020) Uncovering and quantifying the subduction zone sulfur cycle from the slab perspective. Nature Communications 11, 514.
; Walters et al., 2020Walters, J.B., Cruz-Uribe, A.M., Marschall, H.R. (2020) Sulfur loss from subducted altered oceanic crust and implications for mantle oxidation. Geochemical Perspectives Letters 13, 36–41.
).top
Sulfide Dissolution-Recrystallisation
High pressure, intra-slab fluid flow occurs as channelised and in pulses, variably affecting slab mineralogy (e.g., Herms et al., 2012
Herms, P., John, T., Bakker, R.J., Schenk, V. (2012) Evidence for channelized external fluid flow and element transfer in subducting slabs (Raspas Complex, Ecuador). Chemical Geology 310-311, 79–96.
; John et al., 2012John, T., Gussone, N., Podladchikov, Y.Y., Bebout, G.E., Dohmen, R., Halama, R., Klemd, R., Magna, T., Seitz, H.-M. (2012) Volcanic arcs fed by rapid pulsed fluid flow through subducting slabs. Nature Geoscience 5, 489–492.
; Chen et al., 2019Chen, S., Hin, R.C., John, T., Brooker, R., Bryan, B., Niu, Y., Elliott, T. (2019) Molybdenum systematics of subducted crust record reactive fluid flow from underlying slab serpentine dehydration. Nature Communications 10, 4773.
). The lack of systematic covariations between δ82/76Se and other stable isotope signatures of Li, N, and Tl in Raspas eclogites points to different controls on each isotope signature. Kinetically controlled Li isotope fractionation is related to diffusion during fluid-rock interaction (Halama et al., 2011Halama, R., John, T., Herms, P., Hauff, F., Schenk, V. (2011) A stable (Li, O) and radiogenic (Sr, Nd) isotope perspective on metasomatic processes in a subducting slab. Chemical Geology 281, 151–166.
). Lighter N and Tl isotope signatures of Raspas eclogites compared to MORB indicate preservation of low temperature AOC signatures during subduction (Halama et al., 2010Halama, R., Bebout, G.E., John, T., Schenk, V. (2010) Nitrogen recycling in subducted oceanic lithosphere: The record in high- and ultrahigh-pressure metabasaltic rocks. Geochimica et Cosmochimica Acta 74, 1636–1652.
; Shu et al., 2019Shu, Y., Nielsen, S.G., Marschall, H.R., John, T., Blusztajn, J., Auro, M. (2019) Closing the loop: subducted eclogites match thallium isotope compositions of ocean island basalts. Geochimica et Cosmochimica Acta 250, 130–148.
). In contrast to these lithophile element based isotope systems, chalcophile Se substitutes for S in sulfides and witnesses repeated sulfide dissolution-recrystallisation events as fluids pulse through the slab (Evans et al., 2014Evans, K.A., Tomkins, A.G., Cliff, J. Fiorentini, M.L. (2014) Insights into subduction zone sulfur recycling from isotopic analysis of eclogite-hosted sulphides. Chemical Geology 365, 1–19.
; Su et al., 2019Su, W., Schwarzenbach, E., Chen, L., Li, Y., John, T., Gao, J., Chen, F., Hu, X. (2019) Sulfur isotope compositions of pyrite from high-pressure metamorphic rocks and related veins (SW Tianshan, China): Implications for the sulfur cycle in subduction zones. Lithos 348-349, 105212.
; Li et al., 2020Li, J.-L., Schwarzenbach, E.M., John, T., Ague, J.J., Huang, F., Gao, J., Klemd, R., Whitehouse, M.J., Wang, X.-S. (2020) Uncovering and quantifying the subduction zone sulfur cycle from the slab perspective. Nature Communications 11, 514.
). This is supported by the mineral assemblages of the eclogites that show sulfides in different stages of dissolution and recrystallisation (Fig. 3a,b) and covariations between δ82/76Se and trace elements (Fig. 3c–e). Eclogite 46-2, for which a partially recrystallised pyrite with highest Co contents and high Cu (chalcopyrite) rims is shown (Fig. 3a), has the lowest bulk δ82/76Se and highest Cl contents (Fig. 3c). This sample also contains abundant apatite, with up to 300 μg·g−1 Cl (Urann et al., 2020Urann, B.M., Le Roux, V., John, T., Beaudoin, G.M., Barnes, J.D. (2020) The distribution and abundance of halogens in eclogites: An in situ SIMS perspective of the Raspas Complex (Ecuador). American Mineralogist 105, 307–318.
), suggesting that breakdown of hydrous Cl complexes, possibly cause co-precipitation of cations like Sc that partition into matrix minerals (e.g., Sc into omphacite and garnet, Fig. 3d). This coincides with recrystallisation of sulfides that increase Cu contents in eclogites (Fig. 3a,e). Whereas Mo isotopes in the bulk eclogites record the integrated and unidirectional signature of interaction between fluid and rutile-bearing rocks (Fig. 2d; Chen et al., 2019Chen, S., Hin, R.C., John, T., Brooker, R., Bryan, B., Niu, Y., Elliott, T. (2019) Molybdenum systematics of subducted crust record reactive fluid flow from underlying slab serpentine dehydration. Nature Communications 10, 4773.
), δ82/76Se signatures in subducted rocks seem to capture snapshots of repeated and localised sulfide dissolution-recrystallisation effects during highly oxidised, channelised intra-slab fluid flow.We compared our data to closed system equilibrium and Rayleigh-type fractionation models, which express Se isotope behaviour during partial reduction of Se incorporated into (re)-crystallising sulfides from migrating slab fluids. Sulfides that partly dissolve, recrystallise and dissolve again are neither entirely removed nor fully remixed into the system and hence neither closed nor open system Rayleigh fractionation alone, but a combination of the two can adequately express the entire natural process. This was also previously inferred for such prograde metamorphic conditions for S isotopes (Evans et al., 2014
Evans, K.A., Tomkins, A.G., Cliff, J. Fiorentini, M.L. (2014) Insights into subduction zone sulfur recycling from isotopic analysis of eclogite-hosted sulphides. Chemical Geology 365, 1–19.
; Supplementary Information and Fig. S-3). As such, the result of repeated sulfide dissolution-recrystallisation cycles during continuous and pulsed fluid flow is the most plausible scenario that generates the δ82/76Se spectrum of eclogites. This encompasses low δ82/76Se for early stage, only partially re-dissolved sulfides and high values for late stage and/or significantly reworked, i.e. re-dissolved sulfides (Figs. 2, 3, S-3).top
Implications for the Se Subduction Cycle
Our Se data can be reconciled with a repetitive cycle of sulfide dissolution and recrystallisation during oxidising fluid flow through the subducted AOC as previously suggested for S (e.g., Evans et al., 2014
Evans, K.A., Tomkins, A.G., Cliff, J. Fiorentini, M.L. (2014) Insights into subduction zone sulfur recycling from isotopic analysis of eclogite-hosted sulphides. Chemical Geology 365, 1–19.
; Su et al., 2019Su, W., Schwarzenbach, E., Chen, L., Li, Y., John, T., Gao, J., Chen, F., Hu, X. (2019) Sulfur isotope compositions of pyrite from high-pressure metamorphic rocks and related veins (SW Tianshan, China): Implications for the sulfur cycle in subduction zones. Lithos 348-349, 105212.
). Small scale Se isotope heterogeneities are however likely re-homogenised at the slab scale with continuous sulfide reworking (see Graphical Abstract). We therefore speculate that any significant volume of Se, and by analogy also S, that ultimately crosses the slab-mantle wedge zone into the sub-arc mantle might not be isotopically distinct from an initial subduction input. In other words, considering the big picture mass balance, slab dehydration may affect the mass fraction of Se without necessarily affecting its isotope signature. Preservation of the overall isotope signature of the slab, which is further subducted into the deeper mantle, is able to explain high δ82/76Se of plume-related lavas that trace recycled sedimentary components subducted from a redox stratified Proterozoic ocean (Yierpan et al., 2020Yierpan, A., König, S., Labidi, J., Schoenberg, R. (2020) Recycled selenium in hot spot-influenced lavas records ocean-atmosphere oxygenation. Science Advances 6, eabb6179.
), as well as low δ82/76Se of Mariana arc lavas that include recycled materials subducted from a fully oxygenated modern ocean (Kurzawa et al., 2019Kurzawa, T., König, S., Alt, J.C., Yierpan, A., Schoenberg, R. (2019) The role of subduction recycling on the selenium isotope signature of the mantle: Constraints from Mariana arc lavas. Chemical Geology 513, 239–249.
).top
Acknowledgements
SK acknowledges ERC Starting Grant O2RIGIN (636808). M. Liesegang is thanked for EPMA support and IODP for the drillcore samples. Helpful comments by two reviewers and particularly editor H. Marschall significantly improved this contribution.
Editor: Horst R. Marschall
top
References
Bénard, A., Klimm, K., Woodland, A.B., Arculus, R.J., Wilke, M., Botcharnikov, R.E., Shimizu, N., Nebel, O., Rivard, C., Ionov, D.A. (2018) Oxidising agents in sub-arc mantle melts link slab devolatilisation and arc magmas. Nature Communications 9, 3500.
Show in context However, such experimentally derived constants exist for S6+/∑S and were previously used to infer a range of slab fluid FMQ of +1.0 to +1.4 at 1.5 GPa from observed sulfate as well as anhydrite saturation in sub-arc mantle peridotites and melt inclusions (Bénard et al., 2018).
View in article
More accurate constraints require experimental work, yet Se isotope data can already be attributed to more pronounced minimum redox variations than those required for stability of oxidised S and even Mo in the subducting slab, as supported by other studies arguing for variably pronounced oxidation potentials of fluids in subduction zones (e.g., Bénard et al., 2018; Chen et al., 2019; Piccoli et al., 2019; Li et al., 2020; Walters et al., 2020).
View in article
Chen, S., Hin, R.C., John, T., Brooker, R., Bryan, B., Niu, Y., Elliott, T. (2019) Molybdenum systematics of subducted crust record reactive fluid flow from underlying slab serpentine dehydration. Nature Communications 10, 4773.
Show in context It is particularly debated if the oxidised nature of arc magmas is due to oxidised slab components or related to secondary processes during the evolution of arc melts (e.g., Lee et al., 2012; Chen et al., 2019; Tollan and Hermann, 2019).
View in article
A comprehensive major, trace element, Rb-Sr, Sm-Nd, Lu-Hf geochronological and stable Mo-Li-N-Tl isotope dataset is also available (e.g., Halama et al., 2010; John et al., 2010; Chen et al., 2019; Su et al., 2019).
View in article
The source of Mo is HP serpentinite (Chen et al., 2019), whereas Se is derived mostly from the AOC.
View in article
As HP serpentinites dehydrate and flush the overlying subducted AOC, rutiles in the subducted crust preferentially retain low δ98/95Mo as the complementary, high δ98/95Mo fluid is removed into the mantle wedge (Freymuth et al., 2015; Chen et al., 2019).
View in article
This would agree with previous estimates for slab fluid fO2 above FMQ +2 based on Mo isotope data of Raspas eclogites (Chen et al., 2019).
View in article
More accurate constraints require experimental work, yet Se isotope data can already be attributed to more pronounced minimum redox variations than those required for stability of oxidised S and even Mo in the subducting slab, as supported by other studies arguing for variably pronounced oxidation potentials of fluids in subduction zones (e.g., Bénard et al., 2018; Chen et al., 2019; Piccoli et al., 2019; Li et al., 2020; Walters et al., 2020).
View in article
High pressure, intra-slab fluid flow occurs as channelised and in pulses, variably affecting slab mineralogy (e.g., Herms et al., 2012; John et al., 2012; Chen et al., 2019).
View in article
This coincides with recrystallisation of sulfides that increase Cu contents in eclogites (Fig. 3a,e). Whereas Mo isotopes in the bulk eclogites record the integrated and unidirectional signature of interaction between fluid and rutile-bearing rocks (Fig. 2d; Chen et al., 2019), δ82/76Se signatures in subducted rocks seem to capture snapshots of repeated and localised sulfide dissolution-recrystallisation effects during highly oxidised, channelised intra-slab fluid flow.
View in article
The entire complex was metamorphosed at HP/LT conditions (∼600 °C and ca. 2.0 GPa) and most of the rocks experienced subduction related, prograde dehydration related intra-slab fluid flow during burial (John et al., 2010; Halama et al., 2011; Herms et al., 2012; Chen et al., 2019; Urann et al., 2020).
View in article
Evans, K.A., Tomkins, A.G., Cliff, J. Fiorentini, M.L. (2014) Insights into subduction zone sulfur recycling from isotopic analysis of eclogite-hosted sulphides. Chemical Geology 365, 1–19.
Show in context Recent in situ S isotope investigations of sulfides from exhumed high pressure rocks reveal prograde subduction related mobilisation and re-entrapment of S during (partial) sulfide breakdown and recrystallisation, resulting in differences in δ34S of almost 40 ‰ (Evans et al., 2014; Su et al., 2019; Li et al., 2020; Walters et al., 2020), with contrasting data for different localities suggesting either oxidised, SO2- and CO2-rich or reduced, H2S-rich fluid involvement (see also Piccoli et al., 2019).
View in article
This was also previously inferred for such prograde metamorphic conditions for S isotopes (Evans et al., 2014; Supplementary Information and Fig. S-3).
View in article
In contrast to these lithophile element based isotope systems, chalcophile Se substitutes for S in sulfides and witnesses repeated sulfide dissolution-recrystallisation events as fluids pulse through the slab (Evans et al., 2014; Su et al., 2019; Li et al., 2020).
View in article
Our Se data can be reconciled with a repetitive cycle of sulfide dissolution and recrystallisation during oxidising fluid flow through the subducted AOC as previously suggested for S (e.g., Evans et al., 2014; Su et al., 2019).
View in article
Freymuth, H., Vils, F., Willbold, M., Taylor, R.N., Elliott, T. (2015) Molybdenum mobility and isotopic fractionation during subduction at the Mariana arc. Earth and Planetary Science Letters 432, 176–186.
Show in context As HP serpentinites dehydrate and flush the overlying subducted AOC, rutiles in the subducted crust preferentially retain low δ98/95Mo as the complementary, high δ98/95Mo fluid is removed into the mantle wedge (Freymuth et al., 2015; Chen et al., 2019).
View in article
Galvez, M.E., Connolly, J.A.D., Manning, C.E. (2016) Implications for metal and volatile cycles from the pH of subduction zone fluids. Nature 539, 420–424.
Show in context Large arrow shows required minimum redox variation for selenite-selenide transition that requires higher redox variation than the sulfide-sulfate transition for slab fluids with pH of up to 9 in subduction zones (Galvez et al., 2016).
View in article
This difference remains with slightly elevated slab fluid pH of up to 9 in subduction zones (Galvez et al., 2016).
View in article
Halama, R., Bebout, G.E., John, T., Schenk, V. (2010) Nitrogen recycling in subducted oceanic lithosphere: The record in high- and ultrahigh-pressure metabasaltic rocks. Geochimica et Cosmochimica Acta 74, 1636–1652.
Show in context A comprehensive major, trace element, Rb-Sr, Sm-Nd, Lu-Hf geochronological and stable Mo-Li-N-Tl isotope dataset is also available (e.g., Halama et al., 2010; John et al., 2010; Chen et al., 2019; Su et al., 2019).
View in article
Additionally, eclogites with the lowest δ82/76Se show the highest contents of Cu, Cl, Sc (Fig. 3) and N as well as the smallest δ15N range (Fig. S-2; Halama et al., 2010; Urann et al., 2020).
View in article
Selenium isotope composition of Raspas samples and AOC from 1256D vs. (a) Pb/Ce, (b) Rb/Ce, and (c) Cs/Ce (Halama et al., 2010; John et al., 2010), and (d) δ98/95Mo variation.
View in article
Lighter N and Tl isotope signatures of Raspas eclogites compared to MORB indicate preservation of low temperature AOC signatures during subduction (Halama et al., 2010; Shu et al., 2019).
View in article
Halama, R., John, T., Herms, P., Hauff, F., Schenk, V. (2011) A stable (Li, O) and radiogenic (Sr, Nd) isotope perspective on metasomatic processes in a subducting slab. Chemical Geology 281, 151–166.
Show in context Trace element and radiogenic isotope data revealed that the ophiolite consists of depleted abyssal oceanic mantle, oceanic crust (MORB-type eclogites), associated seamounts (blueschists), and sedimentary cover (John et al., 2010; Halama et al., 2011).
View in article
The entire complex was metamorphosed at HP/LT conditions (∼600 °C and ca. 2.0 GPa) and most of the rocks experienced subduction related, prograde dehydration related intra-slab fluid flow during burial (John et al., 2010; Halama et al., 2011; Herms et al., 2012; Chen et al., 2019; Urann et al., 2020).
View in article
Kinetically controlled Li isotope fractionation is related to diffusion during fluid-rock interaction (Halama et al., 2011).
View in article
They have been thoroughly investigated for their petrology and all samples record evidence for hydration prior to subduction, but have been mostly selected to avoid any detectable exhumation related overprint (John et al., 2010; Halama et al., 2011; Herms et al., 2012).
View in article
Herms, P., John, T., Bakker, R.J., Schenk, V. (2012) Evidence for channelized external fluid flow and element transfer in subducting slabs (Raspas Complex, Ecuador). Chemical Geology 310-311, 79–96.
Show in context The entire complex was metamorphosed at HP/LT conditions (∼600 °C and ca. 2.0 GPa) and most of the rocks experienced subduction related, prograde dehydration related intra-slab fluid flow during burial (John et al., 2010; Halama et al., 2011; Herms et al., 2012; Chen et al., 2019; Urann et al., 2020).
View in article
They have been thoroughly investigated for their petrology and all samples record evidence for hydration prior to subduction, but have been mostly selected to avoid any detectable exhumation related overprint (John et al., 2010; Halama et al., 2011; Herms et al., 2012).
View in article
Moreover, since the eclogite facies mineral assemblages show no evidence of subsequent retrograde greenschist facies overprint (except for one sample selected here; John et al., 2010; Herms et al., 2012), the δ82/76Se and their preserved fluid related trends must be attributed to dehydration processes prior to exhumation.
View in article
High pressure, intra-slab fluid flow occurs as channelised and in pulses, variably affecting slab mineralogy (e.g., Herms et al., 2012; John et al., 2012; Chen et al., 2019).
View in article
John, T., Scherer, E.E., Schenk, V., Herms, P., Halama, R., Garbe-Schonberg, D. (2010) Subducted seamounts in an eclogite-facies ophiolite sequence: the Andean Raspas Complex, SW Ecuador. Contributions to Mineralogy and Petrology 159, 265–284.
Show in context Trace element and radiogenic isotope data revealed that the ophiolite consists of depleted abyssal oceanic mantle, oceanic crust (MORB-type eclogites), associated seamounts (blueschists), and sedimentary cover (John et al., 2010; Halama et al., 2011).
View in article
The entire complex was metamorphosed at HP/LT conditions (∼600 °C and ca. 2.0 GPa) and most of the rocks experienced subduction related, prograde dehydration related intra-slab fluid flow during burial (John et al., 2010; Halama et al., 2011; Herms et al., 2012; Chen et al., 2019; Urann et al., 2020).
View in article
They have been thoroughly investigated for their petrology and all samples record evidence for hydration prior to subduction, but have been mostly selected to avoid any detectable exhumation related overprint (John et al., 2010; Halama et al., 2011; Herms et al., 2012).
View in article
A comprehensive major, trace element, Rb-Sr, Sm-Nd, Lu-Hf geochronological and stable Mo-Li-N-Tl isotope dataset is also available (e.g., Halama et al., 2010; John et al., 2010; Chen et al., 2019; Su et al., 2019).
View in article
Moreover, since the eclogite facies mineral assemblages show no evidence of subsequent retrograde greenschist facies overprint (except for one sample selected here; John et al., 2010; Herms et al., 2012), the δ82/76Se and their preserved fluid related trends must be attributed to dehydration processes prior to exhumation.
View in article
Selenium isotope composition of Raspas samples and AOC from 1256D vs. (a) Pb/Ce, (b) Rb/Ce, and (c) Cs/Ce (Halama et al., 2010; John et al., 2010), and (d) δ98/95Mo variation.
View in article
John, T., Gussone, N., Podladchikov, Y.Y., Bebout, G.E., Dohmen, R., Halama, R., Klemd, R., Magna, T., Seitz, H.-M. (2012) Volcanic arcs fed by rapid pulsed fluid flow through subducting slabs. Nature Geoscience 5, 489–492.
Show in context High pressure, intra-slab fluid flow occurs as channelised and in pulses, variably affecting slab mineralogy (e.g., Herms et al., 2012; John et al., 2012; Chen et al., 2019).
View in article
Johnson, T.M. (2004) A review of mass-dependent fractionation of selenium isotopes and implications for other heavy stable isotopes. Chemical Geology 204, 201–214.
Show in context Such abyssal seawater-rock interaction therefore also explains the low average δ82/76Se in our AOC, due to scavenging related partial reduction of preferentially light Se isotopes in an entirely oxic water column (e.g., Johnson, 2004).
View in article
In fluid-solid systems, isotopes of Se are significantly fractionated during partial reduction of oxidised Se6+ to Se4+ or further to Se2− (Johnson, 2004).
View in article
König, S., Eickmann, B., Zack, T., Yierpan, A., Wille, M., Taubald, H., Schoenberg, R. (2019) Redox induced sulfur-selenium isotope decoupling recorded in pyrite. Geochimica et Cosmochimica Acta 244, 24–39.
Show in context Selenium isotopes may thus be useful in constraining the conditions during slab fluid-mediated sulfide redistribution, in particular, if more pronounced redox variations occur in fluids that so far remained beyond the scope of S isotope systematics (König et al., 2019) and previously employed, lithophile (e.g., Mo-Li-N-Tl) stable isotopes in subduction related studies.
View in article
Sulfides thus preferentially incorporate isotopically light Se (e.g., König et al., 2019).
View in article
Kurzawa, T., König, S., Alt, J.C., Yierpan, A., Schoenberg, R. (2019) The role of subduction recycling on the selenium isotope signature of the mantle: Constraints from Mariana arc lavas. Chemical Geology 513, 239–249.
Show in context Preservation of the overall isotope signature of the slab, which is further subducted into the deeper mantle, is able to explain high δ82/76Se of plume-related lavas that trace recycled sedimentary components subducted from a redox stratified Proterozoic ocean (Yierpan et al., 2020), as well as low δ82/76Se of Mariana arc lavas that include recycled materials subducted from a fully oxygenated modern ocean (Kurzawa et al., 2019).
View in article
Lee, C.-T., Luffi, A.P., Chin, E.J., Bouchet, R., Dasgupta, R., Morton, D.M., Le Roux, V., Yin, Q.-Z., Jin, D. (2012) Copper systematics in arc magmas and implications for crust-mantle differentiation. Science 336, 64–68.
Show in context It is particularly debated if the oxidised nature of arc magmas is due to oxidised slab components or related to secondary processes during the evolution of arc melts (e.g., Lee et al., 2012; Chen et al., 2019; Tollan and Hermann, 2019).
View in article
Likewise, different models for sulfate- and 34S-enriched arc lavas involve either slab-derived sulfur flux into the overlying mantle wedge or crustal assimilation by ascending magma (e.g., Lee et al., 2012; Pons et al., 2016).
View in article
Li, J.-L., Schwarzenbach, E.M., John, T., Ague, J.J., Huang, F., Gao, J., Klemd, R., Whitehouse, M.J., Wang, X.-S. (2020) Uncovering and quantifying the subduction zone sulfur cycle from the slab perspective. Nature Communications 11, 514.
Show in context Recent in situ S isotope investigations of sulfides from exhumed high pressure rocks reveal prograde subduction related mobilisation and re-entrapment of S during (partial) sulfide breakdown and recrystallisation, resulting in differences in δ34S of almost 40 ‰ (Evans et al., 2014; Su et al., 2019; Li et al., 2020; Walters et al., 2020), with contrasting data for different localities suggesting either oxidised, SO2- and CO2-rich or reduced, H2S-rich fluid involvement (see also Piccoli et al., 2019).
View in article
In contrast to these lithophile element based isotope systems, chalcophile Se substitutes for S in sulfides and witnesses repeated sulfide dissolution-recrystallisation events as fluids pulse through the slab (Evans et al., 2014; Su et al., 2019; Li et al., 2020).
View in article
More accurate constraints require experimental work, yet Se isotope data can already be attributed to more pronounced minimum redox variations than those required for stability of oxidised S and even Mo in the subducting slab, as supported by other studies arguing for variably pronounced oxidation potentials of fluids in subduction zones (e.g., Bénard et al., 2018; Chen et al., 2019; Piccoli et al., 2019; Li et al., 2020; Walters et al., 2020).
View in article
Piccoli, F., Hermann, J., Pettke, T., Connolly, J.A.D., Kempf, E.D., Vieira Duarte, J.F. (2019) Subducting serpentinites release reduced, not oxidized, aqueous fluids. Scientific Reports 9, 19573.
Show in context Recent in situ S isotope investigations of sulfides from exhumed high pressure rocks reveal prograde subduction related mobilisation and re-entrapment of S during (partial) sulfide breakdown and recrystallisation, resulting in differences in δ34S of almost 40 ‰ (Evans et al., 2014; Su et al., 2019; Li et al., 2020; Walters et al., 2020), with contrasting data for different localities suggesting either oxidised, SO2- and CO2-rich or reduced, H2S-rich fluid involvement (see also Piccoli et al., 2019).
View in article
More accurate constraints require experimental work, yet Se isotope data can already be attributed to more pronounced minimum redox variations than those required for stability of oxidised S and even Mo in the subducting slab, as supported by other studies arguing for variably pronounced oxidation potentials of fluids in subduction zones (e.g., Bénard et al., 2018; Chen et al., 2019; Piccoli et al., 2019; Li et al., 2020; Walters et al., 2020).
View in article
Pons, M.L., Debret, B., Bouilhol, P., Delacour, A., Williams, H. (2016) Zinc isotope evidence for sulfate-rich fluid transfer across subduction zones. Nature Communications 7, 13794.
Show in context Likewise, different models for sulfate- and 34S-enriched arc lavas involve either slab-derived sulfur flux into the overlying mantle wedge or crustal assimilation by ascending magma (e.g., Lee et al., 2012; Pons et al., 2016).
View in article
Shu, Y., Nielsen, S.G., Marschall, H.R., John, T., Blusztajn, J., Auro, M. (2019) Closing the loop: subducted eclogites match thallium isotope compositions of ocean island basalts. Geochimica et Cosmochimica Acta 250, 130–148.
Show in context Lighter N and Tl isotope signatures of Raspas eclogites compared to MORB indicate preservation of low temperature AOC signatures during subduction (Halama et al., 2010; Shu et al., 2019).
View in article
Su, W., Schwarzenbach, E., Chen, L., Li, Y., John, T., Gao, J., Chen, F., Hu, X. (2019) Sulfur isotope compositions of pyrite from high-pressure metamorphic rocks and related veins (SW Tianshan, China): Implications for the sulfur cycle in subduction zones. Lithos 348-349, 105212.
Show in context Recent in situ S isotope investigations of sulfides from exhumed high pressure rocks reveal prograde subduction related mobilisation and re-entrapment of S during (partial) sulfide breakdown and recrystallisation, resulting in differences in δ34S of almost 40 ‰ (Evans et al., 2014; Su et al., 2019; Li et al., 2020; Walters et al., 2020), with contrasting data for different localities suggesting either oxidised, SO2- and CO2-rich or reduced, H2S-rich fluid involvement (see also Piccoli et al., 2019).
View in article
Light S isotope compositions of pyrite in subducted rocks from SW Tianshan, China could be traced back to the alteration of oceanic crust and subsequent, fluid driven sulfide recrystallisation events with prograde metamorphism (Su et al., 2019).
View in article
In contrast to these lithophile element based isotope systems, chalcophile Se substitutes for S in sulfides and witnesses repeated sulfide dissolution-recrystallisation events as fluids pulse through the slab (Evans et al., 2014; Su et al., 2019; Li et al., 2020).
View in article
Our Se data can be reconciled with a repetitive cycle of sulfide dissolution and recrystallisation during oxidising fluid flow through the subducted AOC as previously suggested for S (e.g., Evans et al., 2014; Su et al., 2019).
View in article
A comprehensive major, trace element, Rb-Sr, Sm-Nd, Lu-Hf geochronological and stable Mo-Li-N-Tl isotope dataset is also available (e.g., Halama et al., 2010; John et al., 2010; Chen et al., 2019; Su et al., 2019).
View in article
Tollan, P., Hermann, J. (2019) Arc magmas oxidized by water dissociation and hydrogen incorporation in orthopyroxene. Nature Geoscience 12, 667–671.
Show in context It is particularly debated if the oxidised nature of arc magmas is due to oxidised slab components or related to secondary processes during the evolution of arc melts (e.g., Lee et al., 2012; Chen et al., 2019; Tollan and Hermann, 2019).
View in article
Urann, B.M., Le Roux, V., John, T., Beaudoin, G.M., Barnes, J.D. (2020) The distribution and abundance of halogens in eclogites: An in situ SIMS perspective of the Raspas Complex (Ecuador). American Mineralogist 105, 307–318.
Show in context Additionally, eclogites with the lowest δ82/76Se show the highest contents of Cu, Cl, Sc (Fig. 3) and N as well as the smallest δ15N range (Fig. S-2; Halama et al., 2010; Urann et al., 2020).
View in article
This sample also contains abundant apatite, with up to 300 μg·g−1 Cl (Urann et al., 2020), suggesting that breakdown of hydrous Cl complexes, possibly cause co-precipitation of cations like Sc that partition into matrix minerals (e.g., Sc into omphacite and garnet, Fig. 3d).
View in article
The entire complex was metamorphosed at HP/LT conditions (∼600 °C and ca. 2.0 GPa) and most of the rocks experienced subduction related, prograde dehydration related intra-slab fluid flow during burial (John et al., 2010; Halama et al., 2011; Herms et al., 2012; Chen et al., 2019; Urann et al., 2020).
View in article
Varas-Reus, M.I., König, S., Yierpan, A., Lorand, J.-P., Schoenberg, R. (2019) Selenium isotopes as tracers of a late volatile contribution to Earth from the outer Solar System. Nature Geoscience 12, 779–782.
Show in context Mantle includes representative peridotites (Varas-Reus et al., 2019) and MORB (Yierpan et al., 2020).
View in article
Selenium isotopes do not fractionate significantly during melting, fractional crystallisation and metasomatism (Varas-Reus et al., 2019; Yierpan et al., 2019).
View in article
Compared to the relatively small range for mantle samples including average MORB with δ82/76Se (δ82/76Se = [(82Se/76Se)sample/(82Se/76Se)NIST3149 – 1] × 1000) of −0.16 ± 0.12 ‰ (2 s.d.; Yierpan et al., 2019), peridotites of −0.03 ± 0.07 ‰ (Varas-Reus et al., 2019) and the limited variability observed in seawater overprinted altered ocean crust (AOC) with δ82/76Se of −0.35 to −0.07 ‰ (Pacific drillcore samples; Table S-1), eclogites record a large variation in δ82/76Se from −1.89 to +0.48 ‰ (Fig. 2; see Supplementary Information for analytical methods).
View in article
Walters, J.B., Cruz-Uribe, A.M., Marschall, H.R. (2020) Sulfur loss from subducted altered oceanic crust and implications for mantle oxidation. Geochemical Perspectives Letters 13, 36–41.
Show in context Recent in situ S isotope investigations of sulfides from exhumed high pressure rocks reveal prograde subduction related mobilisation and re-entrapment of S during (partial) sulfide breakdown and recrystallisation, resulting in differences in δ34S of almost 40 ‰ (Evans et al., 2014; Su et al., 2019; Li et al., 2020; Walters et al., 2020), with contrasting data for different localities suggesting either oxidised, SO2- and CO2-rich or reduced, H2S-rich fluid involvement (see also Piccoli et al., 2019).
View in article
More accurate constraints require experimental work, yet Se isotope data can already be attributed to more pronounced minimum redox variations than those required for stability of oxidised S and even Mo in the subducting slab, as supported by other studies arguing for variably pronounced oxidation potentials of fluids in subduction zones (e.g., Bénard et al., 2018; Chen et al., 2019; Piccoli et al., 2019; Li et al., 2020; Walters et al., 2020).
View in article
Yierpan, A., König, S., Labidi, J., Schoenberg, R. (2019) Selenium isotope and S-Se-Te elemental systematics along the Pacific-Antarctic ridge: Role of mantle processes. Geochimica et Cosmochimica Acta 249, 199–224.
Show in context Compared to the relatively small range for mantle samples including average MORB with δ82/76Se (δ82/76Se = [(82Se/76Se)sample/(82Se/76Se)NIST3149 – 1] × 1000) of −0.16 ± 0.12 ‰ (2 s.d.; Yierpan et al., 2019), peridotites of −0.03 ± 0.07 ‰ (Varas-Reus et al., 2019) and the limited variability observed in seawater overprinted altered ocean crust (AOC) with δ82/76Se of −0.35 to −0.07 ‰ (Pacific drillcore samples; Table S-1), eclogites record a large variation in δ82/76Se from −1.89 to +0.48 ‰ (Fig. 2; see Supplementary Information for analytical methods).
View in article
Selenium isotopes do not fractionate significantly during melting, fractional crystallisation and metasomatism (Varas-Reus et al., 2019; Yierpan et al., 2019).
View in article
Yierpan, A., König, S., Labidi, J., Schoenberg, R. (2020) Recycled selenium in hot spot-influenced lavas records ocean-atmosphere oxygenation. Science Advances 6, eabb6179.
Show in context Preservation of the overall isotope signature of the slab, which is further subducted into the deeper mantle, is able to explain high δ82/76Se of plume-related lavas that trace recycled sedimentary components subducted from a redox stratified Proterozoic ocean (Yierpan et al., 2020), as well as low δ82/76Se of Mariana arc lavas that include recycled materials subducted from a fully oxygenated modern ocean (Kurzawa et al., 2019).
View in article
Mantle includes representative peridotites (Varas-Reus et al., 2019) and MORB (Yierpan et al., 2020).
View in article
top
Supplementary Information
The Supplementary Information includes:
- Analytical Methods
- Figures S-1 and S-2
- Sulfide Dissolution-recrystallisation: Modelling Parameters
- Sulfide Dissolution-recrystallisation: Evaluation of Model Applied to our Data
- Figure S-3
- Tables S-1 to S-5
- Supplementary Information References
Download the Supplementary Information (PDF).
Figures
Figure 1 Combined S and Se Pourbaix diagrams for 25 °C and 1 bar pressure. Large arrow shows required minimum redox variation for selenite-selenide transition that requires higher redox variation than the sulfide-sulfate transition for slab fluids with pH of up to 9 in subduction zones (Galvez et al., 2016
Galvez, M.E., Connolly, J.A.D., Manning, C.E. (2016) Implications for metal and volatile cycles from the pH of subduction zone fluids. Nature 539, 420–424.
).
Figure 2 Selenium isotope composition of Raspas samples and AOC from 1256D vs. (a) Pb/Ce, (b) Rb/Ce, and (c) Cs/Ce (Halama et al., 2010
Halama, R., Bebout, G.E., John, T., Schenk, V. (2010) Nitrogen recycling in subducted oceanic lithosphere: The record in high- and ultrahigh-pressure metabasaltic rocks. Geochimica et Cosmochimica Acta 74, 1636–1652.
; John et al., 2010John, T., Scherer, E.E., Schenk, V., Herms, P., Halama, R., Garbe-Schonberg, D. (2010) Subducted seamounts in an eclogite-facies ophiolite sequence: the Andean Raspas Complex, SW Ecuador. Contributions to Mineralogy and Petrology 159, 265–284.
), and (d) δ98/95Mo variation. Mantle includes representative peridotites (Varas-Reus et al., 2019Varas-Reus, M.I., König, S., Yierpan, A., Lorand, J.-P., Schoenberg, R. (2019) Selenium isotopes as tracers of a late volatile contribution to Earth from the outer Solar System. Nature Geoscience 12, 779–782.
) and MORB (Yierpan et al., 2020Yierpan, A., König, S., Labidi, J., Schoenberg, R. (2020) Recycled selenium in hot spot-influenced lavas records ocean-atmosphere oxygenation. Science Advances 6, eabb6179.
). The large grey arrow (Figs. 2, 3) represents our interpretation of the effects of slab mantle dehydration and increasing interaction of subducted crust with slab fluids. The source of Mo is HP serpentinite (Chen et al., 2019Chen, S., Hin, R.C., John, T., Brooker, R., Bryan, B., Niu, Y., Elliott, T. (2019) Molybdenum systematics of subducted crust record reactive fluid flow from underlying slab serpentine dehydration. Nature Communications 10, 4773.
), whereas Se is derived mostly from the AOC. As HP serpentinites dehydrate and flush the overlying subducted AOC, rutiles in the subducted crust preferentially retain low δ98/95Mo as the complementary, high δ98/95Mo fluid is removed into the mantle wedge (Freymuth et al., 2015Freymuth, H., Vils, F., Willbold, M., Taylor, R.N., Elliott, T. (2015) Molybdenum mobility and isotopic fractionation during subduction at the Mariana arc. Earth and Planetary Science Letters 432, 176–186.
; Chen et al., 2019Chen, S., Hin, R.C., John, T., Brooker, R., Bryan, B., Niu, Y., Elliott, T. (2019) Molybdenum systematics of subducted crust record reactive fluid flow from underlying slab serpentine dehydration. Nature Communications 10, 4773.
). Recrystallising, early sulfides preferentially incorporate light Se isotopes and later stage sulfides become isotopically heavier. Repeated sulfide dissolution-recrystallisation produces a spectrum of low and high δ82/76Se in the flushed AOC.
Figure 3 BSE images and semi-quantitative element mapping of sulfides in Raspas eclogites: (a) a partially resorbed Co-bearing pyrite, showing high Cu and Fe contents at the recrystallised chalcopyrite rim and (b) an example of a chalcopyrite with a narrow and oxidised rim, indicated by the lower S and Cu contents. Trends of δ82/76Se vs. concentrations of (c) Cl, (d) Sc and (e) Cu in prograde eclogites are interpreted to result from this changing fluid-mineral chemistry. Symbols are the same as those used in Figure 2.