Metamorphic Dehydration Research Articles

Cold deep subduction of Indian continental crust and release of ultrahigh‐pressure fluid during initial exhumation: Insights from coesite‐bearing eclogite‐vein systems in Kaghan Valley, Pakistan

AbstractThe ultrahigh‐pressure (UHP) eclogites from the Kaghan Valley in Pakistan, which formed by the deep subduction of the Indian plate beneath the Asian plate in the Eocene, contain complex metamorphic vein systems (including both isolated veins and vein networks), with mineral assemblages of epidote + quartz + kyanite + phengite ± omphacite ± garnet. The investigations on the Kaghan UHP eclogite‐vein systems provide important insights into the mechanism and timing of metamorphic dehydration, fluid flow, and fluid–rock interaction in the deeply subducted Indian continental slab as well as the chemical characteristics of slab‐derived, aqueous fluids. Abundant lawsonite pseudomorphs, characterized by prismatic aggregates of epidote, kyanite, and quartz porphyroblasts, are first recognized in the Kaghan eclogites. This observation, in combination with the occurrence of coesite pseudomorphs in epidote porphyroblasts as well as the coexistence of epidote and coesite in the eclogite zircon, indicates the previous existence of UHP lawsonite in these eclogites. Petrological studies and phase equilibrium modelling reveal clockwise P–T trajectories for the Kaghan eclogites that are featured by prograde vectors in lawsonite‐stability regions with peak conditions of 3.0–3.4 GPa/650–690°C, followed by isothermal decompression and lawsonite breakdown under UHP conditions during the initial exhumation stage. The results of metamorphic evolution, together with in situ epidote and bulk Sr isotopic analyses, indicate that the fluids responsible for vein systems are most likely derived from the breakdown of UHP lawsonite in the eclogites. SIMS U–Pb dating of metamorphic zircons from the eclogites, integrated with the Raman analysis of inclusions in zircons, indicates that the UHP dehydration of eclogites occurred at 46.4 ± 1.2 and 46.8 ± 0.9 Ma. Analyses of hydrothermal zircons from the veins yielded slightly younger ages of 44.7 ± 1.0 and 44.9 ± 1.4 Ma, which represent the timing of fluid flow and/or vein crystallization during exhumation of the UHP rocks. Mass‐balance calculation results, in combination with the vein compositions, show that the fluid flow and fluid‐eclogite interaction led to the transfer of Si, Al, Ca, K, and incompatible trace elements from the eclogites into the fluids, from which the vein systems crystallized. This study indicates cold deep subduction of Indian continental crust along low geothermal gradients (6–7°C/km). The UHP fluid liberation and channelized fluid flow occurred during the initial exhumation of the cold Indian slab and are expected to induce the transfer of H2O and incompatible trace elements from the Indian slab to the Asian lithosphere, which potentially contributes to the formation of post‐collisional magmas. Moreover, we suggest that metamorphic vein systems in UHP lawsonite eclogites offer important constraints on the occurrence and timing of fast slab exhumation in continental subduction‐collision zones.

Post-Collisional Tectonomagmatic Evolution, Crustal Reworking and Ore Genesis along a Section of the Southern Variscan Belt: The Variscan Mineral System of Sardinia (Italy)

Since the early Paleozoic, numerous metallogenic events produced in the Sardinian massif a singular concentration of mineral deposits of various kinds. Among them, the Variscan metallogenic peak represents a late Paleozoic phase of diffuse ore formation linked to the tectonomagmatic evolution of the Variscan chain. Two main classes of ores may primarily be attributed to this peak: (1) mesothermal orogenic-type As-Au ± W ± Sb ores, only found in E Sardinia, and (2) intrusion-related Sn-W-Mo-F and base metals-bearing ores found in the whole Sardinian Batholith, but mainly occurring in central–south Sardinia. Both deposit classes formed diachronously during the Variscan post-compressional extension. The orogenic-type ores are related to regional-scale flows of mineralizing fluids, and the intrusion-related ores occur around fertile intrusions of different granite suites. Metallogenic reconstructions suggest almost entirely crustal processes of mineralization without a significant contribution from the mantle. We summarized these processes with a holistic approach and conceptualized the Sardinian Variscan Mineral System (SVMS), a crustal-scale physical system of ore mineralization in the Sardinian basement. The SVMS required suitable metal sources in the crust and diffuse crustal reworking triggered by heat that allowed (a) the redistribution of the original metal budget of the crust in magmas by partial melting and (b) the production of metal-bearing fluids by metamorphic dehydration. Heat transfer in the Sardinian Variscan crust involved shear heating in lithospheric shear zones and the role of mantle uplift as a thermal engine in an extensional tectonic setting. Lithospheric shear zones acted as effective pathways in focusing fluid flow through a large-scale plumbing system into regional-scale structural traps for ores. Pre-Variscan metal sources of metallogenic relevance may have been (1) the magmatic arc and magmatic arc-derived materials of Ordovician age, extensively documented in E Sardinia crust, and (2) an inferred Precambrian crystalline basement lying under the Phanerozoic crustal section, whose presence has been assumed from geophysical data and from petrological and geochemical characteristics of granite suites. At shallower crustal levels, important contributions of metals may have come from pre-Variscan ore sources, such as the Pb-Zn MVT Cambrian ores of SW Sardinia or the REE-bearing Upper Ordovician paleoplacers of E Sardinia.

Geochemical variations of anatectic melts in response to changes of P–T–H2O conditions: Implication for the relationship between dehydration and hydration melting in the Himalayan orogen

The anatectic mechanism is the most fundamental criterion in determining the petrogenetic relationship between migmatite, granulite and granite at convergent plate margins. Although two main anatectic mechanisms, namely water-absent (dehydration) and water-fluxed (hydration) melting, have been identified in many collisional orogens, their spatiotemporal relationship and genetic connection remain obscure. To address this issue, we conduct an integrated study of forward phase equilibrium and trace element modelling for partial melting of metapelite in the Himalayan orogen. The results are used to decipher changes of phase relation and melt composition in response to changes of temperature, pressure and water content. Pressure and water content mainly affect the solidus and evolving residual mineral assemblages, whereas temperature controls the anatectic reaction. Hydration melting at lower pressure and temperature would produce anatectic melts with relatively high SiO2, FeOT + MgO, CaO, K2O, and Ba contents and SiO2/Al2O3 ratios, but low Al2O3 and Na2O contents and Na2O/K2O and Rb/Sr ratios. In contrast, dehydration melting at higher pressure and temperature would generate anatectic melts with relatively high Al2O3 and Na2O contents and Na2O/K2O and Rb/Sr ratios, but low SiO2, FeOT + MgO, CaO, K2O, and Ba contents and SiO2/Al2O3 ratios. A comparison of geochemical compositions between the modelled melts and the two groups of well-characterized leucogranites in the Himalayan orogen indicates that the coeval dehydration and hydration melting would likely occur at the deeper and shallower crustal levels, respectively. In combination with the secular evolution of metamorphic facies series in the Himalayan orogen, we propose a spatiotemporally coupled dehydration-hydration melting mechanism. The early metamorphism at lower geothermal gradients would result in metamorphic dehydration in the lower crust and hydration in the upper crust during the syn-collisional period in the Early Cenozoic. As a result, the later metamorphism at higher geothermal gradients would lead to dehydration melting in the lower crust and hydration melting in the upper crust during the post-collisional period in the Late Cenozoic. This study highlights the applicability of phase equilibrium and trace element modelling to reveal the geochemical variations of anatectic melts in response to changes of P–T–H2O conditions. This provides new insights into the relationship between dehydration and hydration melting in collisional orogens.

The REY geochemistry of phosphorites during metamorphism of the Haizhou Group, NW Yangtze Block, China

Rare earth elements (REE including Y—REY) in phosphorites have been used widely as a proxy for seawater chemistry at the time of their formation. However, it is unclear whether metamorphism can lead to REE remobilization and alteration of primary seawater signatures recorded in phosphorites. In this study, we examine similarly aged metamorphosed phosphorite ore deposits from the northeastern margin of the Yangtze Block (Jinping and Xinpu ore sections), with contemporaneously deposited unmetamorphosed equivalents in the southwestern margin of the Yangtze Block (Weng'an ore section). New insights have been provided into the migration of rare earth elements and the redistribution of Zn and O isotopes in metamorphosed phosphorite, which suggests that the varying degrees of metamorphic fluid–rock reaction during the metamorphism process significantly reduced the δ18O value of phosphate (13.7‰ ± 2.2‰ of Jinping phosphate and 7.3‰ ± 1.8‰ of Xinpu phosphate) and added light Zn isotope. These metamorphic fluids were likely generated by the metamorphic dehydration of upper slab uplift during the plate subduction in the Triassic. The negative correlation between the O (and Zn) isotope compositions and REY contents in P-rich rocks suggests that REY were likely re-enriched during the recrystallization process of apatite. Preferential migration of LREE led to significant LREE and MREE enrichment in Haizhou metamorphosed P-rich rocks compared to Weng'an unmetamorphosed phosphorites.

Boron isotope behavior during metamorphic dehydration and partial melting of continental crust: Constraints from tourmaline in metapelites and leucocratic dikes, Himalayan orogen

Crustal rocks in collisional orogens often underwent the metamorphic transition from dehydration to anatexis in response to an increase of geothermal gradients, but it remains obscure how the two types of metamorphism influenced each other. This issue can potentially be addressed by an integrated study of mineralogy and geochemistry for tourmaline in protoliths and their anatectic products. In this paper, we present new results from comprehensive analyses of major-trace elements and B isotopes in tourmaline together with whole-rock Sm–Nd isotopes in well-characterized metapelites and leucocratic dikes from the Himalayan orogen. The leucocratic dikes show whole-rock εNd(t) values of −15.7 to −15.4, falling within an εNd(t) range of −17.7 to −13.8 for the metapelites. In the metapelites, metamorphic tourmaline (Tur-M) is dravitic and has relatively low δ11B values of −12.0 to −9.0‰. As the metamorphic grade increases, the δ11B values of Tur-M decrease, which is linked to progressive metamorphic dehydration. In the leucocratic dikes, three types of tourmaline are identified by different petrological and geochemical criteria. The tourmaline cores (Tur-LC) are schorlitic and have relatively high δ11B values of −8.7 to −7.7‰, and this type is ascribed to an anatectic origin. The tourmaline rims (Tur-LR) corroding the Tur-LC are dravitic and have relatively high δ11B values of −9.6 to −7.6‰, and this type is attributed to a fluid origin. Another type of anatectic tourmaline (Tur-L) occurs as grains with homogeneous oscillatory zoning that are dravitic and have slightly higher δ11B values of −10.4 to −5.9‰. The anatectic Tur-LC in the leucocratic dikes has higher δ11B values than the metamorphic Tur-M in the metapelites. We attribute this difference to the addition of external fluids with high δ11B values to the metapelite source at the overlying crustal levels. The influxed fluids most likely originated from metamorphic dehydration of lithologically similar metapelites at deeper crustal levels, as evidenced by the Mg-rich characteristics of Tur-L and Tur-LR. The metamorphic fluids would ascend into shallow crustal levels through faults and shear zones, which could not only trigger fluid-fluxed (hydration) melting of the metapelites but also result in fluid activity in later stages. This is further verified by phase equilibrium modelling and model calculation of B geochemical behavior during both metamorphic dehydration and partial melting. On this basis, a spatiotemporally coupled dehydration-hydration melting mechanism is proposed to link the early dehydration metamorphism to the later anatectic metamorphism in the collisional orogen.

Slab dehydration and potassium-lithium recycling in the forearc indicated by potassium and lithium isotope compositions of exhumed metabasites

Slab dehydration at subduction zones has broad implications for arc magmatism, mantle heterogeneity and elemental cycling. We present the petrography, mineral chemistry and whole-rock potassium (K) and lithium (Li) isotopic compositions of subducted metabasites exhumed from serpentinite mud volcanoes from the Mariana forearc that were recovered during International Ocean Discovery Program Expedition 366, to evaluate the capability of K and Li isotopes for tracking slab dehydration and the evolution of the slab-derived fluids at shallow subduction zones. The results show that the δ41K values of HIMU-type alkali metabasites recovered from the shallow-sourced site (i.e., Fantangisña, −0.34 ± 0.06‰, 2SD, n = 3) are slightly higher than those from the deep-sourced site (i.e., Asùt Tesoru, −0.42 ± 0.20‰, 2SD, n = 8). In addition, all the HIMU-type metabasites have homogeneous radiogenic isotope compositions (e.g., Sr, Nd and Hf), and the clinopyroxenes therein show similar trace elements patterns, indicating they have a common mantle source. These observations together with the positive correlation between δ41K and δ11B suggest that K isotope fractionation most likely resulted from metamorphic dehydration. In contrast, no such correlation was observed for Li isotopes (δ7Li ranging from +3.7‰ to +6.6‰). The Li isotope compositions can be modeled by low-temperature fluid/rock interactions with varying water/rock ratios, implying that the Li isotopic system was mainly controlled by fluid metasomatism within the subduction channel. Notably, EM-II-type metabasites have the lightest K isotope compositions (−0.72‰ to −0.67‰) and heaviest Li isotope compositions (+8.1‰ to +8.3‰) in this study. These anomalous signatures were accompanied with the lowest Nb/Th, εNd(t) and εHf(t), indicating that the K and Li isotope data of EM-II-type metabasites were indicative of their mantle source (e.g., EM-II mantle reservoir). Rayleigh fractionation modeling illustrates that the deeply subducted metabasites are characterized by high K and Li concentrations, and low δ41K, but MORB-like δ7Li. Thus, they may transfer large amounts of fluid-mobile elements into the deep mantle that contributes to the K isotopic heterogeneity in the mantle.

Percolation of Low‐Dimensional Water at Crystalline Interfaces Mediates Fluid Migration in Subducting Slabs

AbstractDuring subduction, metamorphic dehydration reactions in the downgoing slab release fluids, generating fluid overpressure. It has been suggested that fluid is driven to flow upward by buoyancy, but a sufficiently high permeability allowing formation of a fluid percolation network is required. Traditionally, fluid percolation has been identified based on the textural equilibrium assumption by measuring the dihedral angle at the triple junction of grains. According to this theory, grain boundaries generally cannot be infiltrated by fluid, and only the grain edge can form a fluid flow channel. We argue that this theory is insufficient because we have found that water from fluid can be adsorbed into the crystalline interface, that is, a layered mineral interlayer, a crack, or a grain boundary. The high pressure in a subducting slab drives water adsorption into the crystalline interface, forming a low‐dimensional fluidic phase, and thus fluid percolation is achieved. Because water adsorbed in the interface is fluidic, water diffusion drives fluid transport in the subducting slab. Due to water adsorption, fluid overpressure at the dehydration front may release, so that dehydration embrittlement may be excluded. Stable water adsorption in the subduction‐slab conditions is determined here by combining molecular dynamics simulations and thermodynamic calculations. Analysis based on simulations shows that water adsorption requires crystalline surfaces which do not form hydrogen bonds well.

Microbially influenced formation of anhydrite at low temperature

Calcium sulfate minerals are abundant in nature – on Earth and on Mars – and important in several fields of material sciences. With respect to gypsum and bassanite, anhydrite represents the anhydrous crystalline phase in the CaSO4–H2O system. Despite years of research, the formation of anhydrite in the laboratory at low temperature remains challenging and, in the geological record, this mineral is mostly interpreted as a secondary phase that form through metamorphic dehydration of gypsum. Here, we present the results of laboratory precipitation experiments showing that anhydrite can form at 35 °C from evaporated seawater through a microbially influenced mineralization process. The experiments were conducted in the presence of extracellular polymeric substances (EPS) produced by bacterial strains isolated from a modern evaporitic environment, the Dohat Faishakh sabkha in Qatar. Without organic molecules, only gypsum formed in parallel control experiments. This finding provides a possible explanation for the origin of several natural occurrences of anhydrite that cannot be satisfactorily explained by existing models and reveals a new precipitation pathway that may have industrial applications.

Progressively elevated geothermal gradients for Miocene metamorphism in the Himalayan orogen

The property of regional metamorphism at convergent plate margins is substantially related to the change of geothermal gradients, resulting in different types of crustal reworking through metamorphic dehydration and partial melting. The relationship between crustal metamorphism and thermal evolution of collisional orogens can thus be deciphered by investigating the change of metamorphic thermobaric ratios in both space and time. This is illustrated by a combined study of field observation, petrographic observation, whole-rock and mineral major and trace elements, laser ablation−inductively coupled plasma−mass spectrometry zircon and monazite U-Th-Pb isotopes and trace elements, and phase equilibrium modeling of granitic gneisses from the Cona area, southern Tibet, in the Himalayan orogen. The results reveal variations in spatial level, protolith age, petrological composition, and metamorphic pressure−temperature−time path along an N-S transect across the Higher Himalayas. On this basis, three types of granitic gneisses were identified for the Miocene metamorphism. Type I gneiss was formed at 16.2−14.7 Ma under upper amphibolite facies conditions of 720−735 °C and 0.77−0.82 GPa, Type II gneisses were formed at 18.3−14.3 Ma under granulite facies conditions of 765−795 °C and 0.99−1.06 GPa, and Type III gneisses were formed at 21.8−13.0 Ma under high-pressure granulite facies conditions of 850−875 °C and 1.40−1.45 GPa. These rocks underwent anatectic metamorphism at lower to upper crustal depths from 22 to 13 Ma, and their metamorphic thermobaric ratios increase from 586−803 °C/GPa at 22−18 Ma to 878−967 °C/GPa and finally to 1071 °C/GPa at 15−13 Ma. This increase corresponds to a progressive increase in metamorphic geothermal gradients from 16.1−22.0 °C/km to 24.1−26.5 °C/km and finally to 29.4 °C/km. Therefore, the crustal rocks in the Himalayan orogen experienced the three types of regional metamorphism at the three different geothermal gradients in the Miocene. The progressively elevated geothermal gradients are attributed to continental rifting due to asthenospheric upwelling consequential to thinning of the lithospheric mantle in the collisional orogen. This has important implications for the tectonic evolution of collisional orogens in the post-collisional stage.

The petrogenetic relationship between migmatite and granite in the Himalayan orogen: Petrological and geochemical constraints

Partial melting of crustal rocks contributes to the formation of granites in collisional orogens. It is generally recognized that residual granulites are compositionally complementary to crustal-derived granites, yet the petrogenetic relationship between migmatites and granites remains obscure. To shed light on this issue, a comprehensive study of field geology, petrographic observations, whole-rock major and trace elements, mineral major elements, zircon and monazite U–Th–Pb isotopes and trace elements, and monazite SmNd isotopes was carried out for the Higher Himalayan leucosomes, leucocratic dikes, and leucogranites from the Cona area in eastern Himalaya. These rocks were formed by partial melting of the Higher Himalayan metasedimentary rocks and granitic gneisses in the Miocene, and are characterized by distinct petrological and geochemical features. Leucosomes and leucocratic dikes, which were derived from different source rocks (granitic gneisses and metasedimentary rocks, respectively), share many similarities. Both of them are rich in K-feldspar and have high K2O, Ba, and Sr contents and CaO/(CaO + Na2O) ratios, but low Rb/Sr and Sm/Nd ratios; LREE contents correlate positively with Th contents, but negatively with P2O5 contents; zircon and monazite have low MREE/HREE ratios. In contrast, leucogranites, which originated from mixed sources of metasedimentary rocks and granitic gneisses, are rich in plagioclase and have low K2O, Ba, and Sr contents and CaO/(CaO + Na2O) ratios, but high Rb/Sr and Sm/Nd ratios; LREE contents correlate positively with Th and P2O5 contents; zircon and monazite have high MREE/HREE ratios. The K2O-rich leucosomes and leucocratic dikes and the Na2O-rich leucogranites do not show geochemical signatures indicative of a link owing to melt evolution processes, suggesting that the two main groups are the products of distinct and independent anatectic processes in the source. The leucosomes and leucocratic dikes are associated with water-fluxed (hydration) melting at lower temperatures and pressures, whereas the leucogranites are related to water-absent (dehydration) melting at higher temperatures and pressures. In this regard, the leucosomes and leucocratic dikes in migmatites and the leucogranites likely represent two different types of anatectic products, which were respectively derived from partial meltig of the hydrated crustal rocks at the shallower level and the dehydrated crustal rocks at the deeper level. The hydrated-dehydrated crustal architecture is ascribed to the influx of water into the shallower crustal rocks caused by metamorphic dehydration at the deeper crustal level. The spatiotemporally coupled dehydration-hydration anatectic processes may be a common mechanism for the formation of coherent granite, migmatite, and granulite in collisional orogens.

Tourmaline chemistry, boron, and strontium isotope systematics trace multiple melt–fluid–rock interaction stages in deeply subducted continental crust

The generation, transport, and recrystallization of slab-derived melts/fluids play a critical role in the deep recycling of elements in subduction zones. While boron (B) isotope systematics have been invoked as an important tracer of these processes, its behavior during metamorphic dehydration and partial melting of deeply subducted continental slabs, and the partitioning of B isotopes between minerals and melts/fluids is not fully understood. Here, we investigate these processes through an in situ study of the major, trace-element, and B-Sr isotope variations in different occurrences of tourmaline in migmatite from the Yuka terrane, North Qaidam orogen (China), which resulted from partial melting of a continental slab at different stages during subduction and exhumation. Based on textural and detailed high-resolution X-ray mapping studies, tourmaline was classified into four paragenetic generations (Tur-I, Tur-II, Tur-III, and Tur-IV). Dravitic Tur-I occurs in melanosomes and shows increasing Fe, Ca, and Ti contents from the core to the outer rim. In addition, it has relatively homogeneous Sr isotope values (0.7407–0.7416) and decreasing δ11B values (-3.8 to −8.6 ‰) and XMg (Mg/(Mg + Fe)) ratios, indicating formation in a rock buffered by an aqueous fluid during the prograde to peak metamorphism. Schorlitic Tur-II occurs within selvage zones between melanosomes and leucosomes, and yields high-Fe values and low δ11B (-13.5 to −10.9 ‰) and more variable 87Sr/86Sr (0.7343–0.7418) values, indicating crystallization in the presence of a hydrous melt external derived from breakdown of Fe-rich mineral(s) during partial melting of the subducted slab. Dravitic Tur-III formed in the matrix and also enveloped Tur-II. It shows homogeneous 87Sr/86Sr values (0.7411–0.7420) and decreasing δ11B values (-6.8 to −9.9 ‰) and XMg ratios as well as increasing Fe and Ti contents from core to outer rim. Formation of Tur-III reflects a transitional stage from hydrous melt to aqueous fluid during exhumation. Tur-IV in the leucosomes is essentially a schorl-dravite solid solution with small amounts of Ca. Its 87Sr/86Sr values (0.7402–0.7416) and δ11B values (-11.4 to −8.5 ‰) are intermediate between the respective values of Tur-II and Tur-III. The formation of Tur-IV likely results from interaction between melt and fluid and, based on its chronological sequence, and is interpreted to have formed during the exhumation stage of the Yuka terrane.Overall, the variable XMg ratios and δ11B values in the different generations of tourmalines are a consequence of the evolution of the melt/fluid at different depths within the deeply subducted slab. Decreasing δ11B values from Tur-I to Tur-II and Tur-III are controlled by the breakdown of different minerals during partial melting or metamorphic dehydration of the subducted slab, while the co-variations of the elemental geochemistry and B isotopic compositions of tourmaline reflect different depths of formation during subduction and exhumation of the lithosphere. These observations suggest tourmaline may serve as a useful tracer of multiple melt/fluid–rock interactions and of boron cycling in continental subduction zones. The heterogeneity of δ11B in melts/fluids at different depth levels of the continental subducted slab may also result in locally variable B isotope values in syn- and post-collisional magmas.

Fast and Slow Earthquakes in Alaska: Implications from a Three-Dimensional Thermal Regime and Slab Metamorphism

Fast and slow earthquakes are predominantly generated along faults constituting active plate boundaries. Characterized by repeated devastating earthquakes and frequent slow slip events and tremors, the Alaska megathrust presents a chance to understand the complicated dynamics of a subduction system changing from steep to shallow dips associated with enigmatically abundant fast and slow seismic events. Based on three-dimensional thermal modeling, we find that the downgoing metamorphosed oceanic crust containing bound water releases a large amount of fluid and causes the recurrence of fast and slow earthquakes by elevated pore fluid pressure and hydrofracturing. The seismogenic interface and the slow slip events (SSEs) identified beneath the Upper Cook Inlet coincide well with the slab metamorphic dehydration regions. The observed slow earthquakes with quasi-stable fault slips preferentially occur, accompanied by high dehydration and temperature downdip along the transition zone.

Can Post-Subduction Porphyry Cu Magmas Form by Partial Melting of Typical Lower Crustal Amphibole-Rich Cumulates? Petrographic and Experimental Constraints from Samples of the Kohistan and Gangdese Arc Roots

Abstract Sulfide-rich hydrous arc cumulates in the lower crust are considered a likely source for porphyry Cu-forming magmas in post-subduction settings. However, it remains controversial whether or not post-subduction porphyry Cu magmas can be produced by partial melting of lower crustal arc cumulates. Here we performed detailed petrographic and geochemical studies on exhumed lower to middle crustal cumulates from the Gangdese and Kohistan arc sections, and then conducted partial melting experiments of the Gangdese sulfide-rich lithologies at 1.5 GPa and 800°C–1150°C using a piston cylinder press. Both the Gangdese and the Kohistan arc-root samples formed as cumulates in the lower to middle crust, but the Kohistan samples were subsequently brought into the garnet stability field during the India–Eurasia continental collision. Most of the Gangdese samples contain relatively high abundances of magmatic sulfides preserved in the form of Cu-poor monosulfide solid solution. In contrast, most of the Kohistan samples contain very low abundances of Cu-rich sulfides preserved as inclusions in garnet, probably due to intense metamorphic dehydration and up to 2 wt% partial melting in the garnet stability field. The original magmatic sulfides at Kohistan appear to have been removed mainly by metamorphic fluids rather than by partial melts, as the arc-root samples could have experienced only very low degree of partial melting and the sulfur solubility in experimentally produced partial melts is low (≤0.1 wt%) even at oxygen fugacities two log-units above the fayalite-magnetite-quartz buffer. Furthermore, the experimental partial melts contain much less K2O, MgO, Cl and ±S than typical post-subduction porphyry Cu-forming magmas. Hence, we conclude that the role of sulfide-rich arc cumulates in the formation of post-subduction porphyry Cu magmas has been severely overestimated.

Hydration, melt production and rheological weakening within an intracontinental gneiss dome

The Entia Gneiss Complex (EGC) in central Australia represents a deeply exhumed Paleoproterozoic basement terrane that underwent fluid-catalysed transformation culminating in the formation of a migmatitic gneiss dome during the 450–300 Ma intracontinental Alice Springs Orogeny (ASO). Foremost amongst the changes the EGC experienced during this event was the regional-scale conversion of nominally anhydrous granulite and igneous crust to hydrous amphibolite facies migmatite, together with the emplacement of locally voluminous mica-bearing pegmatite swarms. LA–ICP–MS U–Pb dating of zircon, monazite and xenotime demonstrates that pegmatite emplacement dominantly occurred late in the Alice Springs Orogeny (c. 360–300 Ma). The oldest generation of pegmatites are transposed into the regional Palaeozoic gneissic fabric and formed at c. 1770 Ma, indicating they belong to protolith rocks of the EGC. Three subsequent generations of pegmatite have ages that range between 360 and 300 Ma. The oldest of these is parallel to the migmatitic gneissic fabric and dated at c. 360 Ma, whereas the younger two generations are largely undeformed, discordant and intruded throughout the period c. 350–300 Ma. The capability for a comparatively dry intracratonic orogenic core to repeatedly produce melt over an interval of c. 60 Myr requires explanation. We suggest crustal-scale duplex development during the latter stages of the ASO resulted in the juxtaposition of dry and mechanically strong Paleoproterozoic basement over deeply buried Neoproterozoic–Cambrian sedimentary sequences. Fluid derived from the metamorphic dehydration of these sequences led to hydration, partial melting and rheological weakening of the basement slab. Prior to its hydration at c. 360 Ma, the EGC was metamorphically unresponsive, accounting for its comparatively late timing of deformation and partial melting compared to the overall history of the Alice Springs Orogeny. The development of the crustal-scale duplex system, combined with hydrous melting of the low-density basement slab beneath high-density supracrustal rocks, resulted in a thermomechanically unstable architecture that facilitated the formation of an orogen-scale gneiss dome during the waning stages of intracontinental orogenesis.

Embrittlement Within Viscous Shear Zones Across the Base of the Subduction Thrust Seismogenic Zone

AbstractGeophysical observations indicate that patches of localized fracturing occur within otherwise viscous regions of subduction plate boundaries. These observations place uncertainty on the possible down‐dip extent of the seismogenic zone, and as a result the maximum magnitude of subduction thrust earthquakes. However, the processes controlling where and how localized fracturing occurs within otherwise viscous shear zones are unclear. We examined three exposures of exhumed plate boundary on Kyushu, Japan, which contain subducted sediments and hydrated oceanic crust deformed at ∼300 to ∼500°C. These exposures preserve subduction‐related viscous deformation, which in two of the studied exposures has a mutually overprinting relationship with quartz veins, indicating localized cyclical embrittlement. Where observed, fractures are commonly near lithological contacts that form viscosity contrasts. Mineral equilibrium calculations for a metabasalt composition indicate that exposures showing cyclical embrittlement deformed at pressure‐temperature conditions near dehydration reactions that consume prehnite and chlorite. In contrast, dominantly viscous deformation occurred at intervening pressure‐temperature conditions. We infer that at conditions close to metamorphic dehydration reactions, only small stress perturbations are required for transient embrittlement, driven by localized dehydration reactions reducing effective stress, and/or locally increased shear stresses along rheological contrasts. Our results show that the protolith composition of the subducting oceanic lithosphere controls the locations and magnitudes of dehydration reactions, and the viscosity of metamorphosed oceanic crust. Therefore, compositional variations might drive substantial variations in slip style.

Fluid-rock interactions at shallow depths in subduction zone: Insights from trace elements and B isotopic composition of metabasites from the Mariana forearc

To reveal the spatial variations of slab-derived fluids and to trace the in-situ dehydration in the shallow subduction zone, we investigated the petrography, mineral chemistry, and whole-rock B isotopes of metabasites that were recovered from the Fantangisña and Asùt Tesoru serpentinite mud volcanoes and originated from the shallow subduction channel at the forearc of the Mariana subduction zone. The alteration mineral assemblages in the investigated metabasites suggest zeolite- to prehnite-pumpellyite-facies metamorphism and lawsonite-blueschist facies metamorphism beneath Fantangisña and Asùt Tesoru seamounts, respectively. The fluid mobile elements (e.g., B, As, Sb, Pb) are preserved in the low grade metamorphic phyllosilicate minerals (e.g., glauconite, pumpellyite, celadonite), thus fixing the B concentrations of the subducted oceanic crust during shallow subduction (<18 km). Both B concentrations (16.7 to 43.9 μg/g) and δ 11 B values (−5.0 to +3.2‰) of the investigated metabasites are significantly higher than fresh normal mid-ocean ridge basalts (N-MORB) and ocean island basalts (OIB), and are generally comparable to the uppermost altered oceanic slab. Notably, the recovered metabasites from the Mariana forearc exhibit a decreasing trend in δ 11 B values with increasing distance to the trench from Fantangisña (62 km) through Asùt Tesoru (72 km) to South Chamorro (78 km) Seamounts. This trend together with Rayleigh dehydration modeling indicate that the 11 B-enriched aqueous fluids were released from the subducting slab during prograde metamorphic dehydration. The estimated B isotopic compositions of the slab-derived fluids released at arc magma genesis depths are generally comparable to that of the Mariana arc lavas. However, the fluids released by dehydration of subducted sediments at shallow depths should be characterized by lower δ 11 B values than fluids released from the slab at depths of magma genesis beneath the island arc. Then, the recycling of the hydrated forearc mantle is necessary to explain the high δ 11 B values of the Mariana arc lavas. While, the variable B isotope compositions of Mariana arc lavas should be controlled by the different ratios of sediment/AOC ratios. • The investigated metabasites experienced low-grade metamorphism. • Both B concentrations and δ 11 B values are significantly higher than their protolith. • δ 11 B values exhibit a decreasing trend with increasing distance to the trench axis. • Subducted serpentine is necessary to explain the high δ 11 B values of the arc lavas.

Sericite 40Ar/39Ar Dating and Indosinian Mineralization in the Liushuping Au–Zn Deposit, West Qinling Orogen, China

The Liushuping deposit, located on the northeast margin of the Bikou Block, is the middle-sized gold-zinc deposit (with ore reserves of 15.67 × 104 t Zn and 2.2 t Au) in the Mianxian–Lueyang–Yangpingguan area. The orebodies occur in the meta-dolomite of the Duantouya and Jiudaoguai formations controlled by the Jiudaoguai syncline. The ore-forming process has experienced hydrothermal period and epigenetic oxidation period, and the hydrothermal period can be divided into two stages. The hydrothermal sericite sample collected from stage 2 yielded a well-defined 40Ar/39Ar isotopic plateau age of 215.70 ± 0.37 Ma, and an 39Ar/36Ar-40Ar/36Ar normal isochron age of 215.35 ± 0.38 Ma, indicating that the metallogenic age of the Liushuping is the Late Triassic (ca. 215 Ma). The ISr(t) of sphalerite is higher than that of the Bikou Group but similar to the Duantouya Formation, indicating that the ore-forming fluids may mainly originate from the metamorphic dehydration of the Duantouya Formation. The Liushuping Au–Zn deposit is consistent with that of the Qinling Indosinian orogeny and mineralization, which are related to oceanic subduction during the Late Triassic.

Tourmaline growth in pelitic schist and quartzite: A textural, chemical and B-isotopic study from the Gangpur Schist Belt, eastern India

Tourmaline is the principal repository of boron in crustal rocks and therefore useful for tracing B-cycling during prograde dehydration and retrogression of supracrustal rocks. Here, we use the major-trace element, and B isotope composition of tourmaline from schists, quartzites, and tourmaline-quartz veins of the Gangpur Schist Belt in eastern India to constrain the source of boron and the physicochemical evolution of B-rich fluids during prograde dehydration metamorphism. Tourmaline growth and re-equilibration in rocks of the Gangpur Schist Belt was a multi-stage process involving several fluid sources. The δ11B varies between −6‰ and −18‰, indicating a dominantly continental source for boron. Tourmaline in schists, quartzites, and tourmaline-quartz veins grew over a wide range of P-T conditions and record multiple episodes of metamorphic dehydration between ca. 1.6 Ga and ca. 0.95Ga. The tourmaline in tourmaline-quartz veins and quartzites has lighter B-isotope composition, typical of continental detritus, while those in the schists and quartzites record pelite-dehydration signature with values decreasing gradually from ca. −12‰ in the cores to ca. −17‰ in the rims. Heavier isotopic compositions (δ11B of ca. −6‰) measured in some grains in the pelites and quartzites indicate boron contribution from meta‑carbonate sources. The mixing of a heavier B-rich metacarbonate-derived fluid with pelite-derived metamorphic fluids could explain the lower B-isotope values in such tourmaline. The study also attempts to constrain the controls on the intake of trace elements in tourmaline. The results suggest that the partitioning of Mn, Y, V, Co and Ti in tourmaline is affected by the growth of porphyroblast phases such as garnet, staurolite, and biotite, while Li, Sr, Zn and Sn reflect the signature of the metamorphic fluid.

A likely geological record of deep tremor and slow slip events from a subducted continental broken formation

Fluids in subduction zones play a key role in controlling seismic activity, drastically affecting the rheology of rocks, triggering mineral reactions, and lowering the effective stress. Fluctuating pore pressure is one important parameter for the switch between brittle and ductile deformation, thus impacting seismogenesis. Episodic tremor and slow slip events (ETS) have been proposed as a common feature of the geophysical signature of subduction zones. Their geological record, however, remains scanty. Only the detailed and further characterization of exhumed fossil geological settings can help fill this knowledge gap. Here we propose that fluctuating pore pressure linked to metamorphic dehydration reactions steered cyclic and ETS-related brittle and ductile deformation of continental crustal rocks in the subduction channel of the Apennines. Dilational shear veins and ductile mylonitic shear zones formed broadly coevally at minimum 1 GPa and 350 °C, corresponding to ~ 30–40 km depth in the subduction zone. We identify carpholite in Ca-poor metasediments as an important carrier of H2O to depths > 40 km in cold subduction zones. Our results suggest that the described (micro)structures and mineralogical changes can be ascribed to deep ETS and provide a useful reference for the interpretation of similar tectonic settings worldwide.

Reduction of iron with no systematic isotope fractionation during continental subduction observed in metamorphic rocks from the Dabie-Sulu orogen, China

• No resolvable Fe isotope fractionation in metabasites during prograde dehydration. • Eclogites have Fe 3+ /∑Fe systematically lower than the other metabasites. • Fe reduction likely occurred during eclogitization with simultaneous S oxidization. In order to better understand the mobility and charge balance of Fe in subducted slabs, here we report precise Fe isotopic data, supplemented by Fe 3+ /∑Fe and C-N-S elemental contents, in a well-characterized suite of metamorphic rocks from the Dabie-Sulu orogen, China, including mafic to felsic greenschists, amphibolites and eclogites. There is no systematic Fe isotopic difference among the mafic schists (δ 56 Fe = 0.08 ± 0.08‰, 2SD, N = 13), amphibolites (δ 56 Fe = 0.09 ± 0.04‰, 2SD, N = 19) and eclogites (δ 56 Fe = 0.07 ± 0.06‰, 2SD, N = 27), implying limited Fe mobility during prograde metamorphic dehydration. The overall δ 56 Fe variation of the metabasites and felsic schists ranging from 0.02 ± 0.03‰ to 0.20 ± 0.03‰ most likely mirrors heterogeneity in their Neoproterozoic magmatic protoliths, and can be interpreted by derivation from diverse sources and isotope fractionation during magma differentiation. In a contrary to limited Fe mobility, a decrease in Fe 3+ /∑Fe ratios has been identified for metabasites during eclogitization (on average 0.43 for greenschists, 0.42 for amphibolites and 0.24 for eclogites respectively). Reduction of Fe could have been compensated by oxidation of S where a concurrent expulsion of sulfate seems to be necessary. This study thus supports the hypothesis that Fe-S redox coupling in subducted slabs promotes transfer of oxidizing agents into the mantle wedge in subduction zones.