Journal of Metamorphic Geology最新文献

Subduction infancy corresponds to the first few million years of the start of subduction following heat transfer from the incipient mantle wedge towards the slab-top, as witnessed by metamorphic soles which represent slivers of oceanic crust metamorphosed up to granulite facies conditions welded beneath obducted ophiolites. In this study, integrated petrological, geochemical, mineralogical, geochronological, and thermodynamic studies were carried out on samples from the Yushugou high-temperature metamorphic ophiolitic complex (YHTM) in the South Tianshan Accretionary Complex (STAC), where a massive exposure of coherent granulite accompanied by a thick peridotite body is preserved. Bulk-rock compositions and Sr–Nd–Hf isotopes demonstrate the petrogenesis of meta-basalts with oceanic island basalt (OIB)-like and mid-ocean ridge basalt (MORB)-like affinities, with little infiltration by subduction-derived melts and/or fluids (e.g., no negative Nb–Ta anomalies). Thermodynamic modelling and U–Pb chronology reveal that the YHTM meta-basalts experienced granulite facies metamorphism of ~840–940°C and ~0.92–1.02 GPa at c. 392 Ma and then possibly reheating and zircon alteration in the Carboniferous. In addition, detrital zircons in sedimentary host rocks of the YHTM show limited Precambrian records and offer maximum depositional ages of c. 410–400 Ma together with the oldest Palaeozoic cluster around c. 470–450 Ma. It is suggested that the YHTM granulites could be of Ordovician–Silurian protolith and such an age pattern significantly deviates from those of adjacent terranes (the Central Tianshan, STAC, and North Tarim Craton). Combined with the compilation of pressure–temperature–time estimates of the YHTM and ages of regional ophiolites, arc intermediate-mafic rocks, A-type granites, and deformation, a model of induced, temporarily northward, intra-oceanic subduction initiation is proposed, which probably occurred along the previously existing weak zone close to a seamount or oceanic plateau in the earliest middle Devonian during the northward subduction of the South Tianshan Ocean (STO). Anomalously high geothermal gradients could be triggered by asthenosphere upwellings, further facilitating the formation of OIB-type metamorphic soles. The YHTM, which represents the remnant of metamorphic soles and associated ophiolites, was finally emplaced to the north margin of the STAC during the relatively long-term (c. 160 million years) accretion and continuous subduction of the STO before its closure. This finding also presents a new natural example of OIB-type metamorphic soles as a snapshot of fossil intra-oceanic subduction infancy during the complex evolutionary history of the STO.

The formation of most jadeitites and other jadeite-rich rocks (jadeitoids) during subduction is thought to occur by precipitation (P-type) or metasomatism (R-type) by infiltration of Na-Al-Si-rich aqueous fluids because of the compositional similarity of the rocks to inferred subduction fluids. Whether these rocks can form by isochemical metamorphism (I-type) during subduction is still hotly debated. A characteristic of I-type jadeitoid is that it exhibits a similar prograde metamorphic record as associated eclogite, in contrast to P- and R-type jadeitite and jadeitoids that are typically enclosed in serpentinite derived from the mantle wedge and either lack a prograde metamorphic history (R-type and P-type) or probably experience a prograde history (R-type) that is difficult to discern owing to the high variance of the jadeite-dominated assemblages and alteration by subduction fluids. The recently discovered Baqing (eastern-central Tibet) jadeitoid is enclosed by quartzo-feldspathic schist and has a peak metamorphic assemblage of almandine + jadeite/omphacite + phengite/paragonite + rutile + quartz, similar to eclogite. Abundant mineral inclusions in almandine, especially rutile inclusions with increasing Zr contents from the core to rim of almandine, provide an opportunity to further decode the jadeitoid-forming processes. In this study, pseudosections and Zr-in-rutile thermometry, together with conventional geothermobarometers, were employed to decipher the metamorphic history of Baqing jadeitoids. Two analysed Baqing jadeitoids exhibit a similar clockwise P–T path, starting from early metamorphic conditions of 5–7 kbar, 350–440°C, to different peak conditions (27–29 kbar, 730–760°C, or 20–23 kbar, 670–710°C), followed by relatively consistent retrograde metamorphic conditions of 6–7 kbar, 530–600°C. This result indicates a similar subduction history to the Baqing eclogite. In addition, the Baqing jadeitoids show similar geochemical characteristics to some Na-rich, K-depleted and Ca-depleted sedimentary rocks or plagiogranite. Therefore, we propose an isochemical genesis for the Baqing jadeitoid, rather than a metasomatic origin.

In situ age and trace element determinations of monazite, rutile and zircon grains from an ultrahigh temperature (UHT) metapelite-hosted leucosome from the Napier Complex using laser split-stream analysis reveal highly variable behaviour in both the U–Pb and trace element systematics that can be directly linked to the microstructural setting of individual grains. Monazite grains armoured by garnet and quartz retain two concordant ages 2.48 and 2.43 Ga that are consistent with the previously determined ages for peak UHT metamorphism in the Napier Complex. Yttrium in the armoured grains is unzoned with contents of ~700 ppm for the garnet-hosted monazite and in the range 400–1,600 ppm for the monazite enclosed within quartz. A monazite grain hosted within mesoperthite records a spread of ages from 2.43 to 2.20 Ga and Y contents ranging between 400 and 1,700 ppm. This grain exhibits core to rim zoning in both Y and age, with the cores enriched in Y relative to the rim and younger ages in the core relative to the rim. A monazite grain that sits on a grain boundary between mesoperthite and garnet records the largest spread in ages—from 2.42 to 2.05 Ga. The youngest ages in this grain are within a linear feature that reaches the core and is connected to the grain boundary between the garnet and mesoperthite; the oldest ages are observed where monazite is in contact with garnet. Yttrium in the grain is enriched in the core and depleted at the rim with the strongest depletions where monazite is adjacent to grain boundaries between the silicate minerals or in contact with garnet. The unarmoured monazite grains have lower intercept ages of 1.85 Ga, which overlaps with the bulk of ages determined from the rutile and is coincident with a previously reported zircon age obtained through depth profiling from the Napier Complex. The age and chemical relationships outlined above illustrate decoupling between the geochemical and geochronological systems in monazite. Individual grains are suggestive of a range of processes that modify these systems, including volume diffusion, flux-limited diffusion and fluid-enhanced recrystallization, all operating at the scale of a single thin section and primarily controlled by host mineral microstructural setting. These findings illustrate how the development of simple partitioning coefficients (cf. garnet/zircon) and geospeedometry based on experimentally determined diffusion coefficients on grain separates may not be achievable. However, it highlights the utility of combining age and trace element concentrations from multiple accessory minerals with microstructural information when trying to build a complete history of tectonothermal events experienced by an ancient rock system that has undergone a prolonged history of thermal, deformational and fluid flow events.

Recent studies of the Cretaceous lower arc crust exposed in Fiordland, New Zealand, conclude that shear zones are sites of melt migration and mass transfer through the deep crust. Here, we investigate the 4–10 km-wide George Sound Shear Zone, which cuts the Western Fiordland Orthogneiss, comprising two main rock types: two-pyroxene gneiss and hornblende gneiss. Previous studies infer a predominantly igneous origin for the two types of gneiss. However, this study finds that melt-rock interaction within the George Sound Shear Zone formed the hornblende gneiss from the precursor two-pyroxene gneiss. Petrographic analyses of samples collected in transects across the shear zone show hydration reaction textures ranging from rims of hornblende + quartz around pyroxene grains to complete replacement of pyroxene grains. Plagioclase is recrystallized and partially replaced by clinozoisite. Additionally, biotite mode increases towards the higher strain rocks in the shear zone. Backscatter images and polarized light microscopy show microstructures indicative of former melt-present deformation, including (a) interconnected mineral films of quartz and K-feldspar along grain boundaries, (b) grains that terminate with low apparent dihedral angles, (c) interstitial grains, with some (d) undulose extinction in plagioclase and (e) serrated grain boundaries. In addition, zircon microstructures are consistent with Zr mobility, further supporting the former presence of melt; geochemical data show enrichment of Zr in the hornblende gneiss as compared to the two-pyroxene gneiss. From the above observations, it is inferred that a felsic to intermediate hydrous melt migrated through the George Sound Shear Zone reacting with the host two-pyroxene gneiss of the Western Fiordland Orthogneiss. Melt migration along grain boundaries was deformation assisted, (i) causing hydration of pyroxene to hornblende + quartz, and plagioclase to clinozoisite, (ii) increasing proportions of biotite within the shear zone and (iii) causing depletion of Cr + Ni and Zr enrichment in the hydrated rock. Our interpretation is supported by published observations of pegmatite dyke swarms that intruded into the George Sound Shear Zone, the P-T conditions of deformation and characterization of microstructures that contrast sharply with those typically found in mylonitic rocks formed under solid-state metamorphic conditions. Our results confirm that hydrous shear zones within otherwise anhydrous country rock are retrogressive and may represent evidence of melt migration through zones of deformation.

Granulite facies metapelitic gneisses collected over a $�200�\times �120$ km exposed area of the Kakamas Domain of the Namaqua–Natal Metamorphic Province in southern Namibia all contain similar garnet–sillimanite–cordierite–biotite–quartz–K-feldspar–ilmenite $�\pm$ plagioclase $�\pm$ magnetite mineral assemblages. These assemblages are interpreted to have equilibrated at suprasolidus retrograde conditions, and most samples contain distinct biotite- or sillimanite-free peak assemblages. Pseudosection modelling constrains extremely uniform residuum solidus conditions of $�5�.�5�\pm �1$ kbar and $�790�\pm �30$°C for the entire Kakamas Domain. Estimated peak metamorphic conditions overlap with these but are more smeared out at between 4 and 7 kbar at 760°C to potentially more than 900°C. The uniformity of residuum solidus conditions is not coincidental, but is a consequence of retrograde re-equilibration due to minor melt retention after peak metamorphism. Re-equilibration could only stop once all retained melt had crystallized, which required the concomitant growth of a hydrous mineral to account for its H₂O component. Biotite is the most stable hydrous mineral in these rocks, such that the residuum $�P$–$�T$ conditions in the Kakamas Domain reflect the upper-$�T$ stability of biotite, and also corresponds to the intersection of the well-known biotite–sillimanite melting reaction that consumed all biotite during prograde metamorphism. The calculated melt fertility of the sample suite indicates that the variable amounts of heat consumed to overcome the latent heat of fusion could have caused a 25°C spread in the peak temperature achieved by the most and least fertile samples. Peak temperature in the Kakamas Domain may have been as much as 100°C higher than residuum solidus conditions for specific samples but cannot be confidently constrained as it is obscured by the effects of both thermal buffering during prograde metamorphism and melt retention during retrograde metamorphism. Both processes are an inescapable part of the evolution of all granulite facies rocks, but their effects are most pronounced in fertile rocks like metapelites that are traditionally the preferred lithology for quantifying the $�P$–$�T$ history of exhumed terranes.

High-temperature–low-pressure metamorphism is commonly associated with intermediate to felsic magmatism in continental orogenic belts. The heat budgets and transfer mechanisms responsible for such elevated temperatures and partial melting of the upper crust are uncertain. The Trois Seigneurs massif, French Pyrenees, preserves a structurally continuous record of Variscan high-temperature–low-pressure metamorphism through a sequence of upper-to-mid-crustal Paleozoic metasedimentary rocks. Conventional thermobarometry and phase equilibria calculations show that metamorphic conditions span ~2.5 kbar, 575°C to suprasolidus conditions of ~6 kbar, 700°C. Peak temperatures depend strongly on depth: temperature gradients of 50–60°C/km are present through the uppermost 12 km of the section; deeper portions (12–20 km) define restricted temperature conditions of ~650–700°C. The lowest-grade metamorphic rocks preserve the largest spread in monazite ²⁰⁶Pb*/²³⁸U dates, from c. 325–285 Ma, while the spread in dates is restricted to c. 305–290 Ma in the highest-grade rocks. Within this spread, each sample yields a well-defined population of monazite ²⁰⁶Pb*/²³⁸U dates with peaks at c. 305 Ma in the andalusite schists, 295 Ma in the sillimanite schists, and 300 Ma in the migmatite sample. Monazite trace-element compositions capture a systematic change with decreasing date and increasing metamorphic grade, including a more negative Eu-anomaly and decreasing Sr concentrations, consistent with co-crystallizing feldspar; increasing HREE and Y contents, consistent with xenotime breakdown; and decreasing Th/U, reflecting increasing U content during breakdown of inherited zircon. Zircon rims from a granite unit that formed via partial melting of the Paleozoic sedimentary package yields a ²⁰⁶Pb/²³⁸U-²⁰⁷Pb/²³⁵U concordia age of 304.1 ± 3.73 Ma. These rims have trace-element compositions reflecting cogenetic apatite and zircon growth during granite formation. Zircon from a calc-alkaline granodiorite intrusion preserves a 40 Ma record of melt-related activity in the lower crust that preceded the regional thermal climax. We interpret these petrochronological data to show that the Trois Seigneurs field gradient including andalusite schist and biotite granite samples represents a genuine geotherm through Variscan orogenic crust during the regional thermal climax at 305 Ma. When combined with constraints from other Pyrenean massifs, the form of the geotherm is consistent with a thermal scenario in which heat is advected to the upper crust by intermediate-composition magmas generated in the lower crust. A simple thermal model for this process indicates that anatexis in the upper crust may plausibly occur within 10 Ma of the initiation of the lower-crustal melting. Such a thermal scenario, however, requires focusing of melt through a fertile lower crust and an elevated Moho heat flux. We sugg

Subduction initiation is recorded by upper plate magmatism and lower plate metamorphism, that is, supra-subduction zone (SSZ) ophiolite–metamorphic sole pair. Here, we report geochemical and geochronological data as well as P–T calculations of amphibolites (metamorphic sole) and hornblende gabbros (SSZ ophiolite) from the Saga ophiolitic mélange in Tibetan Plateau. Amphibolites show trace element contents compatible with normal-mid-ocean ridge basalt (N-MORB), indicating that the protolith of amphibolite formed in a MOR setting. Instead, hornblende gabbros show significant high field strength elements (HFSEs) negative anomalies, enriched large ion lithophile elements (LILEs) and high zircon ε_Hf(t) values, suggesting they formed by fluid-induced partial melting of a depleted mantle. Thermobarometry and phase equilibrium modelling suggest two stages of metamorphism for garnet–clinopyroxene amphibolites: (I) a peak metamorphic stage (~1.9 GPa and 1000°C) and (II) a retrograde metamorphic stage (1.1–1.6 GPa and 800–1000°C). Zircon U–Pb ages of amphibolite and hornblende gabbro are 128.8 ± 5.1 Ma and 128.1 ± 1.5 Ma, respectively, suggesting subduction initiation within the eastern Neo-Tethys occurred no later than 128 Ma and SSZ ophiolite formed at ~128 Ma. Apatite U–Pb ages of amphibolite and hornblende gabbro are 121.8 ± 2.1 Ma and 117.5 ± 4.5 Ma, respectively. Titanite U–Pb age of amphibolite is 122.2 ± 1.5 Ma. Overall, our data suggest that the metamorphic sole and SSZ ophiolite were exhumed since 128–118 Ma, and finally exhumed into the ophiolitic mélange.

Sedimentary-derived (S-type) granites are an important product of orogenic metamorphism, and a range of subtypes can be recognized by differences in field occurrence, mineralogy and geochemistry. These subtypes can reflect variations of initial protolith composition, partial melting reactions, pressure and temperature of anatexis, or magmatic processes that occur during ascent through the crust (e.g. mineral fractional crystallization or crustal assimilation). Together, these diverse factors complicate geological interpretation of the history of peraluminous felsic melt fractions in orogenic settings. To assess the influence of these factors, we performed integrated field investigation, petrology, geochemistry, geochronology and phase equilibrium modelling on a series of leucosomes within migmatite associated with different S-type granites within the Khondalite belt, North China craton (NCC), which is an archetypal collisional orogen. Three types of leucosome are recognized in the east Khondalite belt: leucogranitic leucosome, K-feldspar (Kfs)-rich granitic leucosome and garnet (Grt)-rich granitic leucosome. Phase equilibrium modelling of partial melting and fractional crystallization processes indicate that the leucogranitic leucosomes were mostly produced through fluid-present melting, Kfs-rich granitic leucosomes are produced through muscovite dehydration melting with 3 vol.% garnet fractional crystallization, and Grt-rich granitic leucosomes are produced through biotite dehydration melting with 20–40 vol.% K-feldspar fractional crystallization and up to 20 vol.% peritectic garnet entrainment. Mineral fractional crystallization and peritectic mineral entrainment occur in the source during melting, and play equally important roles in partial melting mechanisms in terms of affecting the geochemical compositions of granitic melts. Thus, we suggest that peraluminous felsic magmas preserved in collisional orogens are dominantly produced by fluid-absent melting in the middle to deep continental crust, although extraction of low-volume melt fractions from an anatectic source region at shallower depths during fluid-present melting can also generate small amounts of S-type granites that subsequently crystallize at high structural levels in the crust.