Partial melting in the Inzie Head gneisses: the role of water and a petrogenetic grid in KFMASH applicable to anatectic pelitic migmatites

A sequence of prograde isograds is recognized within the Dalradian Inzie Head gneisses where pelitic compositions have undergone variable degrees of partial melting via incongruent melting reactions consuming biotite. Three leucosome types are identified. At the lowest grades, granitic leucosomes containing porphyroblasts of cordierite (CRD‐melt) are abundant. At intermediate grades, CRD‐melt mingles with garnetiferous leucosomes (GT‐melt). At the highest grades, CRD‐melt coexists with orthopyroxene‐bearing leucosomes (OPX‐melt), while garnet is conspicuously absent. The prograde metamorphic field gradient is constrained to pressures of 2–3 kbar below the CRD‐melt isograd, and no greater than 4.5 kbar at the highest grade around Inzie Head. A petrogenetic grid, calculated using thermocalc , is presented for the K₂O–FeO–MgO–Al₂O₃–SiO₂–H₂O (KFMASH) system for the phases orthopyroxene, garnet, cordierite, biotite, sillimanite, H₂O and melt with quartz and K‐feldspar in excess. For the implied field gradient, the reaction sequence predicted by the grid is consistent with the successive prograde development of each leucosome type. Compatibility diagrams suggest that, as anatexis proceeded, bulk compositions may have been displaced towards higher MgO content by the removal of (relatively) ferroan granitic leucosome. An isobaric (P = 4 kbar) T–a_H2O diagram shows that premigmatization fluids must have been water‐rich (a_H2O > 0.85) and suggests that, following the formation of small volumes of CRD‐melt, the system became fluid‐absent and melting reactions buffered a_H2O to lower values as temperatures rose. GT‐ and OPX‐melt formed by fluid‐absent melting reactions, but a maximum of 7–11% CRD‐melt fraction can be generated under fluid‐absent conditions, much less than the large volumes observed in the field. There is strong evidence that the CRD‐melt leucosomes could not have been derived by buoyantly aided upwards migration from levels beneath the migmatites. Their formation therefore required a significant influx of H₂O‐rich fluid, but in a quantity insufficient to have exhausted the buffering capacity of the solid assemblage plus melt. Fluid : rock ratios cannot have exceeded 1 : 30. The fluid was channelled through a regionally extensive shear zone network following melt‐induced failure. Such an influx of fluid at such depths has obvious consequences for localized crustal magma production and possibly for cordierite‐bearing granitoids in general.     

Origin,age, and significance of deep‐seated granulite‐facies migmatites in the Barrow zones of Scotland,Cairn Leuchan,Glen Muick area

Petrological modelling of granulite‐facies mafic and semipelitic migmatites from Cairn Leuchan, northeast Scotland, has provided new constraints on the pressure (P) and temperature (T) conditions of high‐grade metamorphism in the type‐locality Barrow zones. Phase diagrams constructed in the Na₂O–CaO–K₂O–FeO–MgO–Al₂O₃–SiO₂–H₂O–TiO₂–O₂ system have constrained the P?T conditions of peak metamorphism in the Glen Muick region of the upper sillimanite zone (Sill+Kfs) to have been at least ～840°C at ～9 kbar (high‐P granulite facies). These conditions are ～120°C and ～3 kbar higher than those recorded by lower sillimanite zone (Sill+Ms) units located only a few kilometres away to the southeast at Glen Girnock, indicating the presence of a significant thermal and barometric high exposed within the Scottish Dalradian, and supporting previous suppositions of a potential tectonic break between the two regions. U–Pb zircon geochronology performed on these mafic migmatites produced ages of c. 540–470 Ma from grains with both igneous and metamorphic morphological characteristics. Their basaltic protoliths likely formed during a period of volcanism dated at c. 570 Ma, associated with passive‐margin extension prior to the onset of Iapetus Ocean closure, and high‐grade metamorphism and partial melting is interpreted to have taken place at c. 470 Ma, synchronous with sillimanite‐grade metamorphism recorded elsewhere in the Dalradian. These high‐grade Cairn Leuchan lithologies are interpreted as representing a fragment of Grampian Terrane lower crust that was exhumed via displacement along a steeply dipping tectonic discontinuity related to the Portsoy–Duchray Hill Lineament, and are not pre‐Caledonian Mesoproterozoic basement, as suggested by some previous studies. Veins within some mafic migmatites in the Cairn Leuchan area, composed almost entirely (>80%) of garnet, with minor quartz, plagioclase, amphibole, and clinopyroxene, are interconnected with leucosomes and are interpreted to represent former garnet‐bearing melt segregations that have been locally drained of almost all melt. Thus, mafic components of the lower crust, currently underlying relatively lower grade metasediments exposed to the southeast, may represent a potential source rock for widely documented, post‐orogenic felsic plutons, sills, and dykes that occur throughout the Grampian Terrane.     

Petrogenetic relations among titanium‐rich minerals in an anatectic high‐P mafic granulite

The petrogenetic relations among Ti‐rich minerals in high‐grade metabasites is illuminated here through a detailed petrological investigation of an anatectic garnet–clinopyroxene granulite from the Grenville Province, Ontario, Canada containing rutile, titanite and ilmenite in distinct microtextural settings. Garnet porphyroblasts exhibit zoned Ti concentrations (up to 0.15 wt% TiO₂ in their cores), as well as a variety of rutile inclusion types, including clusters of small, variably elongate grains and thin (≤1 μm) oriented needles. Calcite inclusions in garnet, commonly observed surrounding garnet cores containing quartz and clinozoisite, indicate the presence of evolving C–O–H fluids during garnet growth and suggest that the rutile clusters may have formed from subsequent Ti diffusion and rutile precipitation within existing fluid inclusions. Titanite forms large subhedral crystals and typically occurs where the primary garnet–clinopyroxene assemblage is in contact with leucosome containing megacrystic hornblende, silvialitic scapolite and calcic plagioclase. Many titanite crystals exhibit marginal subgrains that correspond with sharp changes in their major and trace element composition, likely related to a dissolution–precipitation or recrystallization process following primary crystallization. Clinopyroxene–ilmenite symplectite coronas surround titanite in most locations, likely forming from reaction with the hornblende‐plagioclase matrix (±fluids/melt). Integration of multi‐equilibria thermobarometry and Zr thermometry in rutile and titanite with phase equilibrium modelling allows definition of a clockwise P–T path evolving to peak pressures of ~1.5 GPa at ~750°C during garnet and rutile growth, followed by peak temperature conditions of ~1.2 GPa and ~820–880°C associated with melt‐present titanite growth, and finally cooling and decompression to regional amphibolite facies conditions (~1.0 GPa and ~750°C) associated with the formation of clinopyroxene–ilmenite symplectites surrounding titanite. P–T pseudosections calculated for the pristine (leucosome‐ and titanite ‐free) metabasite bulk composition reproduce much of the prograde phase relations, but predict rutile as the stable Ti‐rich mineral at the peak thermal conditions associated with melt‐present titanite growth. The P–M(CaO) and T–M(CaO) models show that bulk CaO concentrations have a significant effect on the stability ranges of titanite and rutile. Increased bulk CaO tends to stabilize titanite to higher pressure and temperature at the expense of rutile, with a ≥15% increase in CaO producing the observed titanite‐bearing assemblage at high‐P granulite facies conditions. Thus, the model results are consistent with the textural observations, which suggest that titanite stability is associated with a chemical exchange between the host metabasite and a Ca‐rich melt.     

Osumilite–melt interactions in ultrahigh temperature granulites: phase equilibria modelling and implications for the P–T–t evolution of the Eastern Ghats Province,India

The exposed residual crust in the Eastern Ghats Province records ultrahigh temperature (UHT) metamorphic conditions involving extensive crustal anatexis and melt loss. However, there is disagreement about the tectonic evolution of this late Mesoproterozoic–early Neoproterozoic orogen due to conflicting petrological, structural and geochronological interpretations. One of the petrological disputes in residual high Mg–Al granulites concerns the origin of fine‐grained mineral intergrowths comprising cordierite + K‐feldspar ± quartz ± biotite ± sillimanite ± plagioclase. These intergrowths wrap around porphyroblast phases and are interpreted to have formed by the breakdown of primary osumilite in the presence of melt trapped in the equilibration volume by the melt percolation threshold. The pressure (P)–temperature (T) evolution of four samples from three localities across the central Eastern Ghats Province is constrained using phase equilibria modelling in the chemical system Na₂O–CaO–K₂O–FeO–MgO–Al₂O₃–SiO₂–H₂O–TiO₂–Fe₂O₃ (NCKFMASHTO). Results of the modelling are integrated with published geochronological results for these samples to show that the central Eastern Ghats Province followed a common P–T–t history. This history is characterized by peak UHT metamorphic conditions of 945–955 °C and 7.8–8.2 kbar followed by a slight increase in pressure and close‐to‐isobaric cooling to the conditions of the elevated solidus at 940–900 °C and 8.5–8.3 kbar. In common with other localities from the Eastern Ghats Province, the early development of cordierite before osumilite and the peak to immediate post‐peak retrograde reaction between osumilite and melt to produce the intergrowth features requires that the prograde evolution was one of contemporaneous increasing pressure with increasing temperature. This counter‐clockwise (CCW) evolution is evaluated for one sample using inverse phase equilibria modelling along a schematic P–T path of 150 °C kbar^?1 starting from the low P–T end of the prograde P–T path as constrained by the phase equilibria modelling. The inverse modelling is executed by step‐wise down temperature reintegration of sufficient melt into the residual bulk chemical composition at the P–T point of the 1 mol.% melt isopleth at each step, representing the melt remaining on grain boundaries after each prograde drainage event, to reach the melt connectivity transition (MCT) of 7 mol.%. The procedure is repeated until a plausible protolith composition is recovered. The result demonstrates that clastic sedimentary rocks that followed a CCW P–T evolution could have produced the observed mineral assemblages and microstructures preserved in the central Eastern Ghats Province. This study also highlights the role of melt during UHT metamorphism, particularly its importance to both chemical and physical processes along the prograde and retrograde segments of the P–T path. These processes include: (i) an increase in diffusive length scales during the late prograde to peak evolution, creating equilibration volumes larger than a standard thin section; (ii) the development of retrograde mineral assemblages, which is facilitated if some melt is retained post‐peak; (iii) the presence of melt as a weakening mechanism and the advection of heat by melt, allowing the crust to thicken; and (iv) the effect of melt loss, which makes the deep crust both denser and stronger, and reduces heat production at depth, limiting crustal thickening and facilitating the transition to close‐to‐isobaric cooling.     

Phase equilibria modelling of residual migmatites and granulites: An evaluation of the melt‐reintegration approach

Suprasolidus continental crust is prone to loss and redistribution of anatectic melt to shallow crustal levels. These processes ultimately lead to differentiation of the continental crust. The majority of granulite facies rocks worldwide has experienced melt loss and the reintegration of melt is becoming an increasingly popular approach to reconstruct the prograde history of melt‐depleted rocks by means of phase equilibria modelling. It involves the stepwise down‐temperature reintegration of a certain amount of melt into the residual bulk composition along an inferred P–T path, and various ways of calculating and reintegrating melt compositions have been developed and applied. Here different melt‐reintegration approaches are tested using El Hoyazo granulitic enclaves (SE Spain), and Mt. Stafford residual migmatites (central Australia). Various sets of P–T pseudosections were constructed progressing step by step, to lower temperatures along the inferred P–T paths. Melt‐reintegration was done following one‐step and multi‐step procedures proposed in the literature. For El Hoyazo granulites, modelling was also performed reintegrating the measured melt inclusions and matrix glass compositions and considering the melt amounts inferred by mass–balance calculations. The overall topology of phase diagrams is pretty similar, suggesting that, in spite of the different methods adopted, reintegrating a certain amount of melt can be sufficient to reconstruct a plausible prograde history (i.e. melting conditions and reactions, and melt productivity) of residual migmatites and granulites. However, significant underestimations of melt productivity may occur and have to be taken into account when a melt‐reintegration approach is applied to highly residual (SiO₂ <55 wt%) rocks, or to rocks for which H₂O retention from subsolidus conditions is high (such as in the case of rapid crustal melting triggered by mafic magma underplating).     

An experimental study of grain scale melt segregation mechanisms in two common crustal rock types

Creation of pathways for melt to migrate from its source is the necessary first step for transport of magma to the upper crust. To test the role of different dehydration‐melting reactions in the development of permeability during partial melting and deformation in the crust, we experimentally deformed two common crustal rock types. A muscovite‐biotite metapelite and a biotite gneiss were deformed at conditions below, at and above their fluid‐absent solidus. For the metapelite, temperatures ranged between 650 and 800 °C at P_c=700 MPa to investigate the muscovite‐dehydration melting reaction. For the biotite gneiss, temperatures ranged between 850 and 950 °C at P_c=1000 MPa to explore biotite dehydration‐melting under lower crustal conditions. Deformation for both sets of experiments was performed at the same strain rate (ε.) 1.37×10^?5 s^?1. In the presence of deformation, the positive ΔV and associated high dilational strain of the muscovite dehydration‐melting reaction produces an increase in melt pore pressure with partial melting of the metapelite. In contrast, the biotite dehydration‐melting reaction is not associated with a large dilational strain and during deformation and partial melting of the biotite gneiss melt pore pressure builds more gradually. Due to the different rates in pore pressure increase, melt‐enhanced deformation microstructures reflect the different dehydration melting reactions themselves. Permeability development in the two rocks differs because grain boundaries control melt distribution to a greater extent in the gneiss. Muscovite‐dehydration melting may develop melt pathways at low melt fractions due to a larger volume of melt, in comparison with biotite‐dehydration melting, generated at the solidus. This may be a viable physical mechanism in which rapid melt segregation from a metapelitic source rock can occur. Alternatively, the results from the gneiss experiments suggest continual draining of biotite‐derived magma from the lower crust with melt migration paths controlled by structural anisotropies in the protolith.     

Low-P–high-T metamorphism and the role of heat transport by melt migration in the Higo Metamorphic Complex, Kyushu, Japan

This paper characterizes the metamorphic thermal structure of the Higo Metamorphic Complex (HMC) and presents the results of a numerical simulation of a geotherm with melt migration and solidification. Reconstruction of the geological and metamorphic structure shows that the HMC initially had a simple thermal structure where metamorphic temperatures and pressures increased towards apparent lower structural levels. Subsequently, this initial thermal structure has been collapsed by E–W and NNE–SSW trending high‐angle faults. Pressure and temperature conditions using the analysis of mineral assemblages and thermobarometry define a metamorphic field P–T array that may be divided into two segments: the array at apparent higher structural levels has a low‐dP/dT slope, whereas that at apparent lower structural levels has a high‐dP/dT slope. This composite array cannot be explained by heat conduction in subsolidus rocks alone. Migmatite is exposed pervasively at apparent lower structural levels, but large syn‐metamorphic plutons are absent at the levels exposed in the HMC. Transport and solidification of melt within migmatite is a potential mechanism to generate the composite array. Thermal modelling of a geotherm with melt migration and solidification shows that the composite thermal structure may be formed by a change of the dominant heat transfer from an advective regime to a conduction regime with decreasing depth. The model also predicts that strata beneath the crossing point will consist of high‐grade solid metamorphic rocks and solidified melt products, such as migmatite. This prediction is consistent with the observation that migmatite was associated with the very high‐dP/dT slope. The melt migration model is able to generate the very high‐dP/dT segment due to the high rate of heat transfer by advection.     

Partial melting due to breakdown of an epidote‐group mineral during exhumation of ultrahigh‐pressure eclogite: An example from the North‐East Greenland Caledonides

In the North‐East Greenland Caledonides, P–T conditions and textures are consistent with partial melting of ultrahigh‐pressure (UHP) eclogite during exhumation. The eclogite contains a peak assemblage of garnet, omphacite, kyanite, coesite, rutile, and clinozoisite; in addition, phengite is inferred to have been present at peak conditions. An isochemical phase equilibrium diagram, along with garnet isopleths, constrains peak P–T conditions to be subsolidus at 3.4 GPa and 940°C. Zr‐in‐rutile thermometry on inclusions in garnet yields values of ~820°C at 3.4 GPa. In the eclogite, plagioclase may exhibit cuspate textures against surrounding omphacite and has low dihedral angles in plagioclase–clinopyroxene–garnet aggregates, features that are consistent with former melt–solid–solid boundaries and crystallized melt pockets. Graphic intergrowths of plagioclase and amphibole are present in the matrix. Small euhedral neoblasts of garnet against plagioclase are interpreted as formed from a peritectic reaction during partial melting. Polymineralic inclusions of albite+K‐feldspar and clinopyroxene+quartz±kyanite±plagioclase in large anhedral garnet display plagioclase cusps pointing into the host, which are interpreted as crystallized melt pockets. These textures, along with the mineral composition, suggest partial melting of the eclogite by reactions involving phengite and, to a large extent, an epidote‐group mineral. Calculated and experimentally determined phase relations from the literature reveal that partial melting occurred on the exhumation path, at pressures below the coesite to quartz transition. A calculated P–T phase diagram for a former melt‐bearing domain shows that the formation of the peritectic garnet rim occurred at 1.4 GPa and 900°C, with an assemblage of clinopyroxene, amphibole, and plagioclase equilibrated at 1.3 GPa and 720°C. Isochemical phase equilibrium modelling of a symplectite of clinopyroxene, plagioclase, and amphibole after omphacite, combined with the mineral composition, yields a P–T range at 1.0–1. 6 GPa, 680–1,000°C. The assemblage of amphibole and plagioclase is estimated to reach equilibrium at 717–732°C, calculated by amphibole–plagioclase thermometry for the former melt‐bearing domain and symplectite respectively. The results of this study demonstrate that partial melt formed in the UHP eclogite through breakdown of an epidote‐group mineral with minor involvement of phengite during exhumation from peak pressure; melt was subsequently crystallized on the cooling path.     

Granulite facies metamorphism and subsolidus fluid-absent reworking, Strangways Range, Arunta Block, central Australia

Orthopyroxene‐rich quartz‐saturated granulites of the Strangways Range, Arunta Block, central Australia, record evidence of two high‐grade metamorphic events. Initial granulite facies metamorphism (M₁, at c. 1.7 Ga) involved partial melting and migmatization culminating in conditions of 8.5 kbar and 850 °C. Preservation of the peak M₁ mineral assemblages from these conditions indicates that most of the generated melt was lost from these rocks at or near peak metamorphic conditions. Subsequent reworking (M₂, at c. 1.65 Ga) is characterized by intense deformation, the absence of partial melting and the development of orthopyroxene–sillimanite ± gedrite‐bearing mineral assemblages. Gedrite is only present in cordierite‐rich lithologies where it preferentially replaces M₁ cordierite porphyroblasts. Pseudosection calculations indicate that M₂ occurred at subsolidus fluid‐absent conditions (a_H2o ～ 0.2) at 6–7.5 kbar and 670–720 °C. The mineral assemblages in the reworked rocks are consistent with closed system behaviour with respect to H₂O subsequent to M₁ melt loss. M₂ reworking was primarily driven by increased temperature from the stable geotherm reached after cooling from M₁ and deformation‐induced recrystallization and re‐equilibration, rather than rehydration from an externally derived fluid. The development of the M₂ assemblages is strongly dependent on the intensity of deformation, not only for promoting equilibration, but also for equalizing the volume changes that result from metamorphic reactions. Calculations suggest that the protoliths of the orthopyroxene‐rich granulites were cordierite–orthoamphibole gneisses, rather than pelites, and that the unusual bulk compositions of these rocks were inherited from the protoliths. Melt loss is insufficient to account for the genesis of these rocks from more typical pelitic compositions. In quartz‐rich gneisses, however, melt loss along the M₁ prograde path was able to modify the bulk rock composition sufficiently to stabilize peak metamorphic assemblages different from those that would have otherwise developed.     

The cordierite‐bearing anatectic rocks of the higher Himalayan crystallines (eastern Nepal): low‐pressure anatexis,melt productivity,melt loss and the preservation of cordierite

Cordierite‐bearing anatectic rocks inform our understanding of low‐pressure anatectic processes in the continental crust. This article focuses on cordierite‐bearing lithologies occurring at the upper structural levels of the Higher Himalayan Crystallines (eastern Nepal Himalaya). Three cordierite‐bearing gneisses from different geological transects (from Mt Everest to Kangchenjunga) have been studied, in which cordierite is spectacularly well preserved. The three samples differ in terms of bulk composition likely reflecting different sedimentary protoliths, although they all consist of quartz, alkali feldspar, plagioclase, biotite, cordierite and sillimanite in different modal percentages. Analysis of the microstructures related to melt production and/or melt consumption allows the distinction to be made between peritectic and cotectic cordierite. The melt productivity of different prograde assemblages (from two‐mica metapelite/metagreywacke to biotite‐metapelite) has been investigated at low‐pressure conditions, evaluating the effects of muscovite v. biotite dehydration melting on both mineral assemblages and microstructures. The results of the thermodynamic modelling suggest that the mode and type of the micaceous minerals in the prograde assemblage is a very important parameter controlling the melt productivity at low‐pressure conditions, the two‐mica protoliths being significantly more fertile at any given temperature than biotite gneisses over the same temperature interval. Furthermore, the cordierite preservation is promoted by melt crystallization at a dry solidus and by exhumation along P‐T paths with a peculiar dP/dT slope of about 15–18 bar °C^?1. Overall, our results provide a key for the interpretation of cordierite petrogenesis in migmatites from any low‐P regional anatectic terrane. The cordierite‐bearing migmatites may well represent the source rocks for the Miocene andalusite‐bearing leucogranites occurring at the upper structural levels of the Himalayan belt, and low‐P isobaric heating rather than decompression melting may be the triggering process of this peculiar peraluminous magmatism.     

Shooting at a moving target: phase equilibria modelling of high‐temperature metamorphism

Thermodynamic modelling and calculation of P–T pseudosections are commonly employed for quantifying the P–T evolution of metamorphic rocks. A key assumption involved in interpreting a P–T pseudosection is that the bulk‐rock composition used is representative of the effective bulk composition (EBC) from which apparently equilibrated mineral assemblages grew. Choosing an EBC can be difficult in cases where the rock has evolved significantly throughout the P–T history and has become domainal for whatever reason (e.g. loss of fluid and/or melt), particularly at suprasolidus conditions. During partial melting, melt migration may not only change the bulk composition by melt loss but also may generate local variations due to the variable consumption/loss of melt from domain to domain to create volumes of rock that were once internally equilibrated in the presence of a grain boundary melt, but which departed from equilibrium as inter‐granular mobility was slowed by local reductions in melt volume. As well as careful consideration of an EBC, the results of thermodynamic modelling are highly dependent on the specific thermodynamic data set and solution models used, as updates to these data sets may lead to substantially different calculated phase equilibria. This contribution addresses: (1) how consideration of evolving EBCs at multiple scales of observation can be used to resolve the history of complex high‐grade rocks, and (2) how use of different thermodynamic data sets and a–x models (i.e. thermocalc ds5.5 v. ds6) can result in different interpretations of metamorphic evolution. This study investigates the evolution of a mineralogically heterogeneous and texturally complex hand sample of granulite from the Gruf Complex (Central Alps). At the hand‐specimen scale, an EBC can be identified and used to constrain the P–T conditions at which the ‘whole rock’ was last in mutual equilibrium, in the presence of intergranular melt that has subsequently been lost or consumed. Smaller macrodomains (~cm scale) and microdomains (~mm scale) can be identified that represent subsequent evolution during and after melt channelization and loss, and P–T pseudosections can be calculated for the compositions of these domains. Using this approach reveals that the sample experienced a clockwise P–T path marked by near‐isothermal decompression following attainment of peak UHT conditions (~960 °C, 8.5 kbar). The approach enables construction of a P–T history of a rock for which P–T pseudosections are otherwise difficult to interpret. Thermodynamic modelling using ds6 yields similar results to those stated above, but suggests: (1) near‐isothermal decompression occurred over a wider pressure range (~0.5 kbar v. 1.5 kbar), and (2) that not all microdomains record this part of the P–T evolution.     

Cordierite as a sensor of fluid conditions in high-grade metamorphism and crustal anatexis

Cordierite H₂O and CO₂ volatile saturation surfaces derived from recent experimental studies are presented for P–T conditions relevant to high‐grade metamorphism and used to evaluate fluid conditions attending partial melting and granulite formation. The volatile saturation surfaces and saturation isopleths for both H₂O and CO₂ in cordierite are strongly pressure dependent. In contrast, the uptake of H₂O by cordierite in equilibrium with melts formed through biotite dehydration melting, controlled by the distribution of H₂O between granitic melt and cordierite, D_w[D_w = wt% H₂O (melt)/wt% H₂O(Crd)], is mainly temperature dependent. D_w = 2.5–6.0 for the H₂O contents (0.4–1.6 wt percentage) typical of cordierite formed through biotite dehydration melting at 3–7 kbar and 725–900 °C. This range in D_w causes a 15–30% relative decrease in the total wt% of melt produced from pelites. Cordierite in S‐type granites are H₂O‐rich (1.3–1.9 wt%) and close to or saturated in total volatiles, signifying equilibration with crystallizing melts that achieved saturation in H₂O. In contrast, the lower H₂O contents (0.6–1.2 wt percentage) preserved in cordierite from many granulite and contact migmatite terranes are consistent with fluid‐absent conditions during anatexis. In several cases, including the Cooma migmatites and Broken Hill granulites, the cordierite volatile compositions yield a_H2O values (0.15–0.4) and melt H₂O contents (2.2–4.4 wt%) compatible with model dehydration melting reactions. In contrast, H₂O leakage is indicated for cordierite from Prydz Bay, Antarctica that preserve H₂O contents (0.5–0.3 wt%) which are significantly less than those required (1.0–0.8 wt%) for equilibrium with melt at conditions of 6 kbar and 860 °C. The CO₂ contents of cordierite in migmatites range from negligible (< 0.1 wt%) to high (0.5–1.0 wt%), and bear no simple relationship to preserved cordierite H₂O contents and a_H2O. In most cases the cordierite volatile contents yield total calculated fluid activities (a_H2O + a_CO2) that are significantly less than those required for fluid saturation at the P–T conditions of their formation. Whether this reflects fluid absence, dilution of H₂O and CO₂ by other components, or leakage of H₂O from cordierite is an issue that must be evaluated on a case‐by‐case basis.     

Phase equilibria constraints on melting of stromatic migmatites from Ronda (S. Spain): insights on the formation of peritectic garnet

Stromatic metatexites occurring structurally below the contact with the Ronda peridotite (Ojén nappe, Betic Cordillera, S Spain) are characterized by the mineral assemblage Qtz+Pl+Kfs+Bt+Sil+Grt+Ap+Gr+Ilm. Garnet occurs in low modal amount (2–5 vol.%). Very rare muscovite is present as armoured inclusions, indicating prograde exhaustion. Microstructural evidence of melting in the migmatites includes pseudomorphs after melt films and nanogranite and glassy inclusions hosted in garnet cores. The latter microstructure demonstrates that garnet crystallized in the presence of melt. Re‐melted nanogranites and preserved glassy inclusions show leucogranitic compositions. Phase equilibria modelling of the stromatic migmatite in the MnO–Na₂O–CaO–K₂O–FeO–MgO–Al₂O₃–SiO₂–H₂–O₂–C (MnNCaKFMASHOC) system with graphite‐saturated fluid shows P–T conditions of equilibration of 4.5–5 kbar, 660–700 °C. These results are consistent with the complete experimental re‐melting of nanogranites at 700 °C and indicate that nanogranites represent the anatectic melt generated immediately after entering supersolidus conditions. The P–T estimate for garnet and melt development does not, however, overlap with the low‐temperature tip of the pure melt field in the phase diagram calculated for the composition of preserved glassy inclusions in garnet in the Na₂O–CaO–K₂O–FeO–MgO–Al₂O₃–SiO₂–H₂O (NCKFMASH) system. A comparison of measured melt compositions formed immediately beyond the solidus with results of phase equilibria modelling points to the systematic underestimation of FeO, MgO and CaO in the calculated melt. These discrepancies are present also when calculated melts are compared with low‐T natural and experimental melts from the literature. Under such conditions, the available melt model does not perform well. Given the presence of melt inclusions in garnet cores and the P–T estimates for their formation, we argue that small amounts (<5 vol.%) of peritectic garnet may grow at low temperatures (≤700 °C), as a result of continuous melting reactions consuming biotite.     

Melt infiltration into quartzite during partial melting in the Little Cottonwood Contact Aureole (UT,USA): implication for xenocryst formation

Melt infiltration into quartzite took place due to generation and migration of partial melts within the high‐grade metamorphic rocks of the Big Cottonwood (BC) formation in the Little Cottonwood contact aureole (UT, USA). Melt was produced by muscovite and biotite dehydration melting reactions in the BC formation, which contains pelite and quartzite interlayered on a centimetre to decimetre scale. In the migmatite zone, melt extraction from the pelites resulted in restitic schollen surrounded by K‐feldspar‐enriched quartzite. Melt accumulation occurred in extensional or transpressional domains such as boudin necks, veins and ductile shear zones, during intrusion‐related deformation in the contact aureole. The transition between the quartzofeldspathic segregations and quartzite shows a gradual change in texture. Here, thin K‐feldspar rims surround single, round quartz grains. The textures are interpreted as melt infiltration texture. Pervasive melt infiltration into the quartzite induced widening of the quartz–quartz grain boundaries, and led to progressive isolation of quartz grains. First as clusters of grains, and with increasing infiltration as single quartz grains in the K‐feldspar‐rich matrix of the melt segregation. A 3D–μCT reconstruction showed that melt formed an interconnected network in the quartzites. Despite abundant macroscopic evidence for deformation in the migmatite zone, individual quartz grains found in quartzofeldspathic segregations have a rounded crystal shape and lack quartz crystallographic orientation, as documented with electron backscatter diffraction (EBSD). Water‐rich melts, similar to pegmatitic melts documented in this field study, were able to infiltrate the quartz network and disaggregate grain coherency of the quartzites. The proposed mechanism can serve as a model to explain abundant xenocrysts found in magmatic systems.     

Melt-producing and melt-consuming reactions in the Achankovil cordierite gneisses, South India

Migmatitic cordierite gneisses within the Achankovil Zone (AZ) of southern Pan‐African India record melt‐producing and subsequent melt‐consuming mineral reactions. Early mineral assemblages Bt‐Sil‐Qtz and Bt‐Sil‐Spl, deduced from inclusion textures in garnet prophyroblasts, break down via successive dehydration melting reactions to high‐T phase assemblages (e.g. Grt‐Crd‐Liq, Opx‐Liq, Spl‐Crd‐Liq). Later back reactions between the restite and the in situ crystallizing melt resulted in thin cordierite coronas separating garnet from the leucosome, and partial resorption of garnet to Opx‐Crd or Crd‐Bt‐Qtz symplectites. Leucosomes generally display a moderate (low‐strain gneisses) to strong (high‐strain gneisses) depletion of alkali feldspar attributed to mineral‐melt back reactions partly controlled by the degree of melt segregation. Using a KFMASH partial petrogenetic grid that includes a melt phase, and qualitative pseudosections for microdomains of high and low Al/Si ratios, the successive phase assemblages and reaction textures are interpreted in terms of a clockwise P–T path culminating at about 6–7 kbar and 900–950 °C. This P–T path is consistent with, but more detailed than published results, which suggests that taking a melt phase into account is not only a valid, but also a useful approach. Comparing P–T data and lithological and isotopic data for the AZ with adjacent East Gondwana fragments, suggests the presence of a coherent metasedimentary unit exposed from southern Madagascar via South India (AZ) and Sri Lanka (Wanni Complex) to the Lützow–Holm Bay in Eastern Antarctica.     

Near-isothermal decompression within a clockwise P–T evolution recorded in migmatitic mafic granulites from Clavering Ø, NE Greenland: implications for the evolution of the Caledonides

The high grade rocks (metapelites and metabasites) of Clavering Ø represent the easternmost exposures of granulites in the Palaeozoic Caledonian Orogen of East Greenland. Mafic granulites which occur as sheet‐like bodies and lenses within metapelitic migmatites and orthogneiss complexes have experienced migmatisation and mineral equilibria which define a clockwise P–T path incorporating a near‐isothermal decompression segment. Textures demonstrate the existence of early garnet‐clinopyroxene‐melt assemblages which equilibrated at >8–11 kbar and 850–915 °C. Subsequently, decompression melting led to formation of orthopyroxene‐plagioclase‐melt assemblages at conditions below >8–11 kbar. Continued syn‐deformational decompression is indicated by a combination of both static and syn‐deformational recrystallization textures which generated finer grained orthopyroxene‐plagioclase assemblages. P–T constraints indicate these assemblages equilibrated at c. 5.0–6.5 kbar at 850–915 °C. These data are consistent with the rocks undergoing a stage of rapid tectonic‐induced exhumation involving some 3.0–4.5 kbar (c.10–12 km) uplift as part of a clockwise P–T path in a collisional setting.     

Deep subduction of mantle-derived garnet peridotites from the Su-Lu UHP metamorphic terrane in China

Garnet peridotites from the southern Su‐Lu ultra‐high‐pressure metamorphic (UHPM) terrane, eastern China, contain porphyroblastic garnet with aligned inclusions comprising a low‐P–T mineral assemblage (chlorite, hornblende, Na‐gedrite, Na‐phlogopite, talc, spinel and pyrite). Orthopyroxene porphyroblasts show fine exsolution lamellae of clinopyroxene and minor chromite. A clinopyroxene inclusion in garnet shows some orthopyroxene exsolution lamellae. Both the rims of porphyroblastic pyroxene and garnet and the matrix pyroxene and garnet crystallized at the expense of olivine. This is interpreted as a result of metasomatism of the peridotites by an SiO₂‐rich melt at UHP conditions. A chromian garnet further overgrew on the rims of the garnet. The X_Mg values (Mg/(Mg+Fe)) of porphyroblastic garnet decrease from core to rim and vary in different peridotite samples, while the compositions of both the porphyroblastic and the matrix pyroxene are similar in terms of Ca–Mg–Fe. The Mg‐rich cores of porphyroblastic garnet and orthopyroxene record high temperatures and pressures (c. 1000 °C, ≥5.1 GPa), whereas the matrix minerals, including the rims of porphyroblasts, record much lower P–T (c. 4.2 GPa, c. 760 °C). Sm–Nd data give apparent isochron ages of c. 380 Ma and negative ε_Nd(0) values (c.?9). These dates are considered meaningless due to isotopic disequilibrium between garnet cores and the rest of the rocks. The isotopic disequilibrium was probably caused by metasomatism of the peridotites by melt/fluids derived from the coevally subducted crustal materials. On the other hand, the Rb–Sr isotopic systems of phlogopite and clinopyroxene appear to have reached equilibrium and record a cooling age of c. 205 Ma. It is suggested that the garnet peridotites were originally emplaced into a low‐P–T environment prior to the c. 220 Ma continental collision, during which they were subducted together with crustal rocks to mantle depth and subjected to UHP metamorphism. An important corollary is that at least some of the coevally subducted crustal rocks in the Su‐Lu terrane have been subjected to peak metamorphism at P–T conditions much higher than presently estimated (≥2.7 GPa, ≤800 °C).     

Retrograde processes in migmatites and granulites revisited

Many migmatites and granulites preserve evidence of a clockwise P–T evolution involving decompression (decrease in P) while close to the thermal peak. The extent of post‐thermal peak reaction is influenced by several factors, including: (1) the P–T path in relation to invariants in the system and the Clapeyron slopes of the equilibria; (2) the rate of cooling; and (3) the availability of fluid (H₂O‐rich volatile phase or melt) for fluid‐consuming reactions. Reaction may occur between products of a prograde (increasing T) fluid‐generating reaction as the same equilibrium is re‐crossed in the retrograde (decreasing T) sense. In general, reaction reversal or ‘back reaction’ requires the P–T path to approximate isobaric heating and cooling, without significant decompression, and evolved fluid to remain within the equilibration volume. The larger the decompression segment in the P–T evolution, the more chance there is of crossing different reactions along the retrograde segment from those crossed along the prograde segment. For common pelite compositions, we may generalize by considering three pressure regimes separated by the [Spl, Ms, H₂O] invariant in KFMASH (approximately 9 kbar) and the intersection of muscovite breakdown with the H₂O‐rich volatile phase‐saturated solidus (approximately 4 kbar). Reaction reversal cannot occur along P–T paths that traverse around one of these points, but may occur along P–T paths confined to one of the three regimes in between. Additionally, above the solidus, melt segregation and loss potentially change the composition of the equilibration volume; and, the size of the equilibration volume shrinks with decreasing T. Since the proportion of melt to residue in the equilibration volume may change with decreasing size, the composition of the equilibration volume may change throughout the supra‐solidus part of the retrograde segment of the P–T evolution. If melt has been lost from the equilibration volume, reaction reversal may not be possible or may be only partial; indeed, the common preservation of close‐to‐peak mineral assemblages in migmatite and granulite demonstrates that extensive reaction with melt is uncommon, which implies melt isolation or loss prior to crossing potential melt‐consuming reactions. Water dissolved in melt is transported through the crust to be exsolved on crystallization at the solidus appropriate to the intrinsic a(H₂O). This recycled water causes retrogression at subsolidus conditions. Consideration of the evidence for supra‐solidus decompression‐dehydration reactions, and review of microstructures that have proven controversial, such as corona and related microstructures, selvage microstructures and ‘late’ muscovite, leads to the conclusion that there is more than one way for these microstructures to form and reminds us that we should always consider multiple working hypotheses!     

Large-scale melt-depletion in granulite terranes: an example from the Archean Ashuanipi Subprovince of Quebec

This study uses field, petrographic and geochemical methods to estimate how much granitic melt was formed and extracted from a granulite facies terrane, and to determine what the grain‐ and outcrop‐scale melt‐flow paths were during the melt segregation process. The Ashuanipi subprovince, located in the north‐eastern Superior Province of Quebec, is a large (90 000 km²) metasedimentary terrane, in which > 85% of the metasediments are of metagreywacke composition, that was metamorphosed at mid‐crustal conditions (820–900 °C and 6–7 kbar) in a late Archean dextral, transpressive orogen. Decrease in modal biotite and quartz as orthopyroxene and plagioclase contents increase, together with preserved former melt textures indicate that anatexis was by the biotite dehydration reaction: biotite + quartz + plagioclase = melt + orthopyroxene + oxides. Using melt/orthopyroxene ratios for this reaction derived from experimental studies, the modal orthopyroxene contents indicate that the metagreywacke rocks underwent an average of 31 vol% partial melting. The metagreywackes are enriched in MgO, CaO and FeO_t and depleted in SiO₂, K₂O, Rb, Cs, and U, have lower Rb/Sr, higher Rb/Cs and Th/U ratios and positive Eu anomalies compared to their likely protolith. These compositions are modelled by the extraction of between 20 and 40 wt %, granitic melt from typical Archean low‐grade metagreywackes. A simple mass balance indicates that about 640 000 km³ of granitic melt was extracted from the depleted granulites. The distribution of relict melt at thin section‐ and outcrop‐scales indicates that in layers without leucosomes melt extraction occurred by a pervasive grain boundary (porous) flow from the site of melting, across the layers and into bedding planes between adjacent layers. In other rocks pervasive grain boundary flow of melt occurred along the layers for a few, to tens of centimetres followed by channelled flow of melt in a network of short interconnected and structurally controlled conduits, visible as the net‐like array of leucosomes in some outcrops. The leucosomes contain very little residual material (< 5% biotite + orthopyroxene) indicating that the melt fraction was well separated from the residuum left in situ as melt‐depleted granulite. Only 1–3 vol percentage melt remained in the melt‐depleted granulites, hence, the extraction of melt generated by biotite dehydration melting in these granulites, was virtually complete under conditions of natural melting and strain rates in a contractional orogen.     

Archaean TTGs as sources of younger granitic magmas: melting of sodic metatonalites at 0.6–1.2 GPa

Two natural, low K₂O/Na₂O, TTG tonalitic gneisses (one hornblende-bearing and the other biotite-bearing) were partially melted at 0.8–1.2 GPa (fluid-absent). The chief melting reactions involve the breakdown of the biotite and hornblende. The hornblende tonalite is slightly less fertile than the biotite tonalite, but melt volumes reach around 30% at 1,000°C. This contrasts with results of most previous work on more potassic TTGs, which generally showed much lower fertility, though commonly producing more potassic melts. Garnet is formed in biotite-bearing tonalitic protoliths at P > 0.8 GPa and at > 1.0 GPa in hornblende-bearing tonalitic protoliths. All fluid-absent experiments produced peraluminous granitic to granodioritic melts, typically with SiO₂ > 70 wt.%. For the biotite tonalite, increasing T formed progressively more melt with progressively lower K₂O/Na₂O. However, the compositions of melts from the hornblende tonalite do not vary significantly with T. With increasing P, melts from the biotite tonalite become less potassic, due to the increasing thermal stability of biotite. For the hornblende tonalite, again there is no consistent trend. Fluid-absent melting of sodic TTGs produces melts with insufficient K₂O to model the magmas that formed the voluminous, late, potassic granites that are common in Archaean terranes. Reconnaissance fluid-present experiments at 0.6 GPa imply that H₂O-saturated partial melting of TTGs is also not a viable process for producing magmas that formed these granites. The protoliths for these must have been more potassic and less silicic. Nevertheless, at granulite-facies conditions, sodic TTGs will produce significant quantities of broadly leucogranodioritic melt that will be more potassic than the protoliths. Upward abstraction of this melt would result in some LILE depletion of the terrane. Younger K-rich magmatism is unlikely to represent recycling of TTG crust on its own, and it seems most likely that evolved crustal rocks and/or highly enriched mantle must be involved. Electronic supplementary material The online version of this article (doi:) contains supplementary material, which is available to authorized users.