Fluids liberated from subducting slabs are critical in global geochemical cycles. We investigate the behaviour of Mo during slab dehydration using two suites of exhumed fragments of subducted, ...oceanic lithosphere. Our samples display a positive correlation of δ
Mo
with Mo/Ce, from compositions close to typical mantle (-0.2‰ and 0.03, respectively) to very low values of both δ
Mo
(-1‰) and Mo/Ce (0.002). Together with new, experimental data, we show that molybdenum isotopic fractionation is driven by preference of heavier Mo isotopes for a fluid phase over rutile, the dominant mineral host of Mo in eclogites. Moreover, the strongly perturbed δ
Mo
and Mo/Ce of our samples requires that they experienced a large flux of oxidised fluid. This is consistent with channelised, reactive fluid flow through the subducted crust, following dehydration of the underlying, serpentinised slab mantle. The high δ
Mo
of some arc lavas is the complement to this process.
Sulfur belongs among H
O, CO
, and Cl as one of the key volatiles in Earth's chemical cycles. High oxygen fugacity, sulfur concentration, and δ
S values in volcanic arc rocks have been attributed to ...significant sulfate addition by slab fluids. However, sulfur speciation, flux, and isotope composition in slab-dehydrated fluids remain unclear. Here, we use high-pressure rocks and enclosed veins to provide direct constraints on subduction zone sulfur recycling for a typical oceanic lithosphere. Textural and thermodynamic evidence indicates the predominance of reduced sulfur species in slab fluids; those derived from metasediments, altered oceanic crust, and serpentinite have δ
S values of approximately -8‰, -1‰, and +8‰, respectively. Mass-balance calculations demonstrate that 6.4% (up to 20% maximum) of total subducted sulfur is released between 30-230 km depth, and the predominant sulfur loss takes place at 70-100 km with a net δ
S composition of -2.5 ± 3‰. We conclude that modest slab-to-wedge sulfur transport occurs, but that slab-derived fluids provide negligible sulfate to oxidize the sub-arc mantle and cannot deliver
S-enriched sulfur to produce the positive δ
S signature in arc settings. Most sulfur has negative δ
S and is subducted into the deep mantle, which could cause a long-term increase in the δ
S of Earth surface reservoirs.
The Erro Tobbio olivine-antigorite serpentinites and associated dehydration veins represent hydrated oceanic mantle rocks that escaped complete dehydration and recycling into the mantle after ...subduction to ~ 550–600 °C and 2.0–2.5 GPa. These rocks thus offer valuable insights into the petrological evolution of a slice of hydrated oceanic mantle and the geochemical cycling down to intermediate subduction zone depths. Our study emphasises the role of brucite upon rock-buffered hydration and subduction dehydration employing bulk and in situ chemical data sets combined with petrology.
Bulk rock data reveal a coherent mantle peridotite slice affected by variable melt depletion and refertilisation. Subsequent fluid-rock interaction stages proceeded isochemically with respect to SiO2, i.e., without significant SiO2 enrichment characteristic for hydrothermal ocean floor serpentinisation. Relicts of low-T mesh textures after olivine and preservation of precursor mineral and low-T hydration geochemical features indicate a lack of subsequent fluid and metamorphic overprinting, even on scales of tens of micrometres. Fluid-mobile element enrichments are modest with exceptions for B and W. Enrichment signatures of U/Cs < 1 and Rb/Cs of 4–26 are characteristic of shallow forearc hydration within or atop the slab by fluids derived from breakdown of clays or first dehydration of altered oceanic crust with a subordinate sedimentary pore fluid component. Overall, the geochemical and petrological changes of the Erro Tobbio peridotites during fluid-rock interactions were rock-buffered, in contrast to fluid-buffered hydration accompanied with significant SiO2 metasomatism at, e.g., mid ocean ridges.
Silica-neutral rock-buffered serpentinisation resulted in prominent brucite formation upon olivine hydration. In absence of excess SiO2, subsequent serpentine transformation of chrysotile/lizardite to antigorite likely produced even more brucite. Rock-buffered fluid-rock interactions thus provide a mechanism for stabilising brucite in subduction zone serpentinites, presumably along hydration fronts and within deeper sections of the oceanic lithospheric mantle. Finally, brucite + antigorite dehydration produced up to 40 vol% of metamorphic olivine and prominent olivine + Ti-clinohumite + magnetite vein networks at temperatures <550–600 °C, prior to complete antigorite breakdown. Wall rocks released alkali elements, B, Cr, As, Sb, and Ba into the dehydration fluids, along with substantial Sr, REE and HFSE redistribution into vein minerals.
Display omitted
•Olivine antigorite serpentinites record rock-buffered (de-)hydration histories.•Brucite is central to water and element cycling during forearc dehydration.•Rock-buffered serpentinisation limits SiO2 metasomatism, favouring high brucite modes.
Permeabilities in the subducting slab appear to be too low and dihedral angles between fluid and relevant minerals too high to allow for porous flow, hence fluid channelization is critical for the ...understanding of subduction zone fluid fluxes. In this review we will outline how fluid channelization controls reaction rates and element redistributions during metamorphism of the subducting plate as well as trace element compositions of subduction-related fluids during flow.
Channelized fluid flow predicts that from a rock point of view, most formerly subducted material will show only very limited evidence for fluid flow, consistent with the rarity of observed high fluid fluxes in subduction-related rocks. Aqueous fluid produced by dehydration reactions will not percolate through large rock volumes, but rather will be carried away from the dehydration sites by a veining network. Indeed evidence for significant aqueous-fluid fluxes have been found in high-pressure veins with adjacent selvages. In such selvages, large lithophile elements (LILE's) generally show the highest mobilities, followed by light (L) rare earth elements (REE) and then heavy (H) REE. Compared to high field strength elements (HFSE), even Th shows higher mobilities.
From a fluid point of view, equilibrium between aqueous fluid and surrounding rock will only be approached at sites of fluid production and mineral reaction. However, this fluid can be significantly modified while moving upwards through a veining network where the wallrocks are out of equilibrium with the fluid. In a subducting slab, such reactive fluid flow can preferentially dissolve minerals and release their trace elements (e.g. Ba in phengite, Th and La in monazite). The degree of change in aqueous-fluid composition will depend on the amount of fluid–mineral surface interaction. The chemical exchange reactions will not be possible to model by trace element partition coefficients alone, instead future models need to incorporate kinetic parameters such as surface reaction rates.
Intermediate depth seismicity in subduction zones often occurs in the form of two slab‐parallel bands. We estimated the seismic P to S wave velocity ratio within the shallowest part of the lower ...seismicity zone (LSZ) in the mantle of the subducting slab of the Central Andean subduction system at 50‐km depth, 30 km below the Moho, using local earthquake data. We find an exceptionally high VP/VS value larger than ∼2.0 that cannot be explained by a realistic solid lithology but requires the presence of fluid‐filled porosity. This implies that the incoming Nazca plate must be partially hydrated to this depth below the seafloor. We introduce a state‐of‐the‐art petrophysical model that takes into account the thermodynamic and poroelastic effects of dynamic metamorphic mineral dehydration at 1.8 GPa and consider anisotropic effects. The model shows that a high VP/VS value generally indicates that the medium is near the percolation threshold, that is, that porosity must be interconnected. This result is consistent with observations from outcrops of paleosubduction zones, laboratory experiments, and numerical simulations. It follows that the shallowest part of the LSZ of the Central Andes must reside at a temperature at which mineral dehydration reactions take place, here between 430 and 500 ° C. For the first time, we can confirm that the observations of transient dehydrating fluid‐filled vein structures with a pore volume in the order of only 10−3 are reasonable for the LSZ and enough to allow for effective drainage.
Plain Language Summary
Beneath the Central Andes, small earthquakes occur along a band 30 km within the subducting Nazca tectonic plate that slowly sinks below the South American plate. Such earthquake bands occur worldwide, and it has been proposed that they happen, because there is water bound in the solid rocks there, which would be released when the subducting plate heats up while it sinks into the Earth. We test this hypothesis by determining a rock property that can indicate whether liquid water is present from the arrival times of earthquake waves. We then theoretically calculate how this property would be for different kinds of rock that can be present at this specific location within the Earth. It results that the rock property that we derived in the earthquake band can best be explained by the presence of liquid water that is stored in thin interconnected veins there. Other researchers saw in laboratory experiments, in computer simulations, and in the field that rocks that release water look exactly like this, so that their and our pieces of the puzzle fit together well. With this knowledge, we can understand better how water sinks into the Earth and how it finds its way back to the surface.
Key Points
A high seismic VP/VS ratio is observed in a lower band of seismicity in a subducting slab
Active metamorphic mineral dehydration is required to explain this value
Poroelastic modeling indicates the presence of a percolating vein network
Raman spectroscopy has been widely used in mineralogy and petrology for identifying mineral phases. Some recent applications of Raman spectroscopy involve measuring the residual pressure of mineral ...inclusions, such as quartz inclusions in garnet host, to recover the entrapment pressure condition during metamorphism. The crystallographic orientations of entrapped inclusions and host are important to know for the modelling of their elastic interaction. However, the analysis of tiny entrapped mineral inclusions using EBSD technique requires time consuming polishing. The crystallographic orientations can be measured using polarized Raman spectroscopy, as the intensities of Raman bands depend on the mutual orientation between the polarization direction of the laser and the crystallographic orientation of the crystal. In this study, the Raman polarizability tensor of quartz is first obtained and is used to fit arbitrary orientations of quartz grains. We have implemented two rotation methods: (1) sample rotation method, where the sample is rotated on a rotation stage, and (2) polarizer rotation method, where the polarization directions of the incident laser and the scattered Raman signal are parallel and can be rotated using a circular polarizer. The precision of the measured crystallographic orientation is systematically studied and is shown to be ca. 0.25 degrees using quartz wafers and quartz plates that are cut along known orientations. It is shown that the orientation of tiny mineral inclusions (ca. 2–5 μm) can be precisely determined and yield consistent results with EBSD.
Volatile-rich, CI- and CM-like clasts occur in different brecciated achondrite and chondrite groups. The CI-like clasts in HEDs, polymict ureilites, as well as ordinary, CR, and CB chondrites have a ...similar mineralogy, indicating a similar alteration history. However, when viewed in detail, their mineral chemistry shows some minor differences between the clasts from different meteorite groups. For CM-like clasts found in HED meteorites, the clasts are, based on their mineralogy, clearly fragments of CM chondrites. To be able to decipher whether CI- (or CM-)like clasts from different meteorite groups are related to certain meteorite classes known to contain volatiles, we obtained D/H ratios of several clasts from the meteorite groups mentioned above and compared them with those of CI and CM chondrites as well as to unique carbonaceous chondrites such as Bells, Essebi, and Tagish Lake. Considering the δD-values, CM-like clasts in HEDs span a similar range compared to bulk values of CM chondrites, further indicating that CM-like clasts are fragments of CM chondrites. For CI-like clasts a clear distinction can be made: While CI-like clasts in HEDs and ordinary chondrites show a very similar range in their δD-signatures compared to “common” CI chondrites, meaning that these clasts are likely related to CI chondrites, the CI-like clasts in polymict ureilites are enriched in D up to 3000‰; a similarly high enrichment is found for the CI-like clasts in CR chondrites. Thus, although the CI-like clasts in ureilites and CR chondrites likely experienced similar alteration histories as the CI-like clasts found in the other meteorite types, these clasts probably formed in a different region than the CI chondrites and, thus, are more accurately referred to as C1 clasts. Overall, the existence and isotopic signatures of the C1 clasts in several meteorite groups proves the existence of additional primitive, volatile-rich material in the (early) Solar System besides the matter we study as the CI, CM, and CR chondrites. This material was distributed throughout the Solar System very early and might have played an important role for the volatile inventory of the terrestrial planets.
Apatite (Ca5(PO4)3(OH, F, Cl)) is one of the main host of halogens in magmatic and metamorphic rocks and plays a unique role during fluid–rock interaction as it incorporates halogens (i.e. F, Cl, Br, ...I) and OH from hydrothermal fluids to form a ternary solid solution of the endmembers F-apatite, Cl-apatite and OH-apatite. Here, we present an experimental study to investigate the processes during interaction of Cl-apatite with different aqueous solutions (KOH, NaCl, NaF of different concentration also doped with NaBr, NaI) at crustal conditions (400–700°C and 0.2GPa) leading to the formation of new apatite. We use the experimental results to calculate partition coefficients of halogens between apatite and fluid. Due to a coupled dissolution–reprecipitation mechanism new apatite is always formed as a pseudomorphic replacement of Cl-apatite. Additionally, some experiments produce new apatite also as an epitaxial overgrowth. The composition of new apatite is mainly governed by complex characteristics of the fluid phase from which it is precipitating and depends on composition of the fluid, temperature and fluid to mineral ratio. Furthermore, replaced apatite shows a compositional zonation, which is attributed to a compositional evolution of the coexisting fluid in local equilibrium with the newly formed apatite. Apatite/fluid partition coefficients for F depend on the concentration of F in the fluid and increase from 75 at high concentrations (460μg/g F) to 300 at low concentrations (46μg/g F) indicating a high compatibility of F in apatite. A correlation of Cl-concentration in apatite with Cl− concentration of fluid is not observed for experiments with highly saline solutions, composition of new apatite is rather governed by OH− concentration of the hydrothermal fluid. Low partition coefficients were measured for the larger halogens Br and I and vary between 0.7*10−3–152*10−3 for Br and 0.3*10−3–17*10−3 for I, respectively. Br seems to have D values of about one order of magnitude higher than I. These data allow an estimation of the D values for the other halogens based on a lattice strain model which displays a sequence with DF of ∼120, DOH of ∼100, DCl of ∼2.3 DBr ∼0.045, and DI ∼0.0025. Results from this experimental study help to better understand fluid–rock interaction of an evolving fluid, as it enables the composition of hydrothermally derived apatite to be used as a fluid probe for halogens at crustal conditions. It further shows the importance of mineral replacement as one of the key reactions to generate apatite of different composition.
The island of Holsnøy in the Bergen Arcs, which belong to the Caledonides of western Norway, represents an excellent example of how fluid‐induced eclogitization modifies material deeply buried by ...subduction and continental collision. We produced a new detailed map of the northwestern part of Holsnøy, differentiating not only the magnitude of eclogitization but also the strain intensity at different spatial scales: from the outcrop to the entire massif. Using structural data from eclogite‐facies shear zones and eclogitized low‐strain domains, we show that fluid‐mediated eclogitization not only progresses via the development of shear zones (dynamic eclogitization) but occurs over large areas of the island without associated deformation, creating a characteristic static eclogite‐facies overprint. Static eclogitization preserves the structural features of the granulitic protolith while the rock body is transformed from a granulite‐ to an eclogite‐facies mineral assemblage. The extent of static eclogitization was underestimated strongly in the past, a finding also relevant for the interpretation of seismological images in currently active orogens, where presumably similar processes are currently occurring. In addition, we find that the general structure of the eclogite‐facies shear zones is scale‐independent over several orders of magnitude. Although crustal‐scale eclogite‐facies complexes are rarely preserved without significant modification during exhumation, this implies that similar geometrical configurations are likely produced at the scale of the whole lower crust during subduction or continental collision and therefore shape the crustal geophysical signature.
Key Points
The fluid‐induced eclogitization on Holsnøy is controlled both dynamically by shear zone development and static equilibration
Eclogite‐facies shear zone geometry on Holsnøy is scale‐independent from centimeter to kilometer scale
We investigated the halogen (Cl, F, Br, and I) chemistry of serpentinites that record progressive dehydration during subduction from shallow oceanic environments via increased pressure and ...temperature conditions to complete breakdown of antigorite. The aim is to evaluate the relevance of serpentinites for halogen recycling in subduction zones and for deep mantle recharge of these elements. The halogen compositions of the analyzed samples indicate input from seawater and sedimentary sources during initial serpentinization of either subducting lithospheric mantle during slab bending or forearc mantle by uprising slab fluids. During the first dehydration stage (antigorite
+
brucite
→
olivine
+
H
2O), fluids with high Br/Cl and I/Cl ratios are released resulting in residual serpentinites with lower Br/Cl and I/Cl ratios. Veins associated with this event and with the final antigorite breakdown (antigorite
→
olivine
+
orthopyroxene
+
H
2O) show higher halogen ratios compared to their adjacent wall rocks, and they are similar to those found in arc volcanoes (F/Cl and I/Cl between ca. 0.083–1.5, and ca. 0.00038–0.0013, respectively). All measured deserpentinization samples show a narrow range in δ
37Cl values (between −
0.42‰ and +
0.92‰) overlapping the δ
37Cl values of seafloor serpentinites and confirming that no significant Cl isotope fractionation occurs during subduction dehydration of serpentinites. Our findings document the conservative behavior of halogens during subduction. Mass balance constraints reveal that serpentinites strongly control the halogen chemistry of deep subduction zone fluids and that descent of rock residues after deserpentinization strongly affects the halogen budget of the mantle.
► Halogen concentrations and ratios and Cl isotope data indicate halogen input from sedimentary reservoirs during shallow serpentinization, either at the outer rise during slab bending or in suprasubduction shallow environments. ► During deserpentinization, fluids with high Br/Cl and I/Cl ratios are released which drives the residual serpentinites to lower Br/Cl and Cl/I and during ongoing dehydration. F/Cl ratios increase strongest with increasing degree of dehydration throughout the whole deserpentinization process. The high-pressure slab fluids liberated during the antigorite to olivine reaction have F/Cl and I/Cl ratios that are similar to those of arc volcanoes. ► The overall δ
37Cl values of the deserpentinization fluids are close to 0‰, which are within the range of known arc volcanoes. ► The concentrations and ratios found in the serpentinization–deserpentinization sequence indicate the conservative behavior of halogens during subduction and fluid release. ► Mass balance calculations show that serpentinites seem to strongly control the halogen budget, and likely also the Cl isotope signature, of the deeper subduction zone fluids.
PDF
