To determine the factors that control the oxygen content of sulphide melts in the upper mantle, mono-sulphides were equilibrated with mantle peridotite assemblages, with varying FeO and NiO contents, between 3 and 13 GPa and 1300 to 1819 °C. Carbonates were added as a flux and the use of graphite capsules allowed oxygen fugacities (fo2) to be estimated for most of the experiments. Sulphide melt O concentrations were in the range 0.2–3.7 wt. %, coexisting with silicate assemblages with olivine Fe/(Fe + Mg) ratios between 0.04 and 0.25. Except in Ni-bearing experiments, variations in the sulphur/metal cation ratio did not affect the sulphide melt O contents, which also appeared to be independent of fo2. The silicate FeO contents, temperature and pressure were found to be the main controls on O contents. Nickel lowers the O content, although all Ni-bearing experiments produced sulphides with lower sulphur/metal ratios, making it difficult to categorically separate the effect of Ni alone. A preliminary geothermometer expression was developed based on the O content of the sulphide melts and the iron oxide concentrations of coexisting olivine and orthopyroxene. With this expression, the experimental temperatures are reproduced to within 74 K for Ni-free experiments and within 135 K for Ni-bearing samples. Using measurements of the O contents of sulphide inclusions of peridotite affinity in diamonds from the Lac de Gras kimberlite field, a plausible average entrapment temperature of 1318 ± 48 °C is calculated. Mantle peridotite assemblages would be expected to contain sulphide melts with approximately 0.4–0.6 wt. % O along a typical mantle adiabat down to 200 km depth.

The Pyroxenite Marker, a thin, orthopyroxene-dominated marker horizon, is observed towards the top of the Main Zone of the Bushveld Complex, where the last voluminous influx of magma into the Bushveld Complex is thought to have occurred. In an attempt to constrain the Nd-isotopic composition of the magma added at the level of the Pyroxenite Marker, a total of 21 whole-rock samples from a borehole (BH7771) drilled on the Central Sector of the Eastern Limb of the Bushveld Complex were analysed for their Sr–Nd isotopic ratios. Modelling suggests that the added magma had a unique Sr (87Sr/86Sri = 0.7063–0.7067) and Nd (ƐNdi on the order of − 5.9) isotopic composition, distinct from any of the rocks constituting the layered sequence below the Pyroxenite Marker. Dispersion of data points around the modelled isotopic (melt–melt) mixing curves is interpreted to reflect the incorporation of minerals derived from either the incoming or resident magmas into individual rock layers occurring across the Pyroxenite Marker interval, either in response to the mixing of minerals settling through a stratified magma column, or potentially through the intrusion and mixing of crystal-laden magmas with unique isotopic compositions from a sub-Bushveld staging chamber.

Geochemical and isotopic data are presented for ~ 32 Ma-old high-K andesites and dacites from the Alpine Chain. The samples consist of plagioclase, amphibole, titanomagnetite and rare biotite and quartz. Geochemical and isotope data indicate that slab-derived fluids, sediment melts and presumably AFC processes involving continental crust played a key role in the petrogenesis of the high-K rocks. A contribution of fluids is suggested based on the overall enrichment of large-ion lithophile elements and related high Ba/La, Ba/Zr, Ba/Th, Ba/Nb and Pb/Nd, sometimes distinctively higher than average continental crust. Positively correlated Ba/Nb–Th/Nb relationships, low Ce/Pb, low Nb/U and a negative correlation of Pb isotopes with Ce/Pb and Nb/U and positive ∆ 7/4 and ∆ 8/4 values similar to GLOSS imply the additional involvement of a sediment-derived melt. Negatively correlated Nb/Ta–Zr/Hf ratios at overall low Nb/Ta (13–7.5) are best explained by parental magma differentiation involving amphibole and biotite in a continental arc system. The samples have moderately unradiogenic Nd (εNd: – 2.0 to – 6.7) and radiogenic 87Sr/86Sr isotope compositions (0.7085–0.7113), moderately radiogenic Pb isotope compositions (206Pb/204Pb: 18.50–18.72; 207Pb/204Pb: 15.59–15.65; 208Pb/204Pb: 38.30–38.67), and elevated δ18O values (+ 6.5 to + 9.1 ‰). Epsilon Hf isotope values range from + 2.5 to – 4.0. Negative εHf(t) and εNd(t) values and 206Pb/204Pb ratios are correlated with elevated K2O abundances that indicate enrichment in K2O is related to AFC processes. The offset of εHf at a given εNd points to involvement of aged garnet-bearing crustal lithologies. The latter feature is qualitatively consistent with modification of unexposed primary basaltic andesites by AFC processes involving deep crustal material. In conclusion, in an Alpine context, inferred unexposed primitive high-K basaltic to andesitic melts are generated in the mantle wedge through fluid infiltration from the descending slab where fluids may have caused also partial melting of sedimentary rocks that mixed with evolving andesite–dacite compositions towards shallow-level intrusive and extrusive rocks. High-K and related trace element and isotope features thus result from a combination of already elevated values with participation of fluids and melts and probably AFC processes.

Aluminosilicates (kyanite, sillimanite and andalusite) are useful pressure–temperature (P–T) indicators that can form in a range of rock types through different mineral reactions, including those that involve partial melting. However, the presence of xenocrystic or inherited grains may lead to spurious P–T interpretations. The morphologies, microtextural positions, cathodoluminescence responses and trace element compositions of migmatite-hosted kyanite from Eastern Bhutan were investigated to determine whether sub-solidus kyanite could be distinguished from kyanite that crystallised directly from partial melt, or from kyanite that grew peritectically during muscovite dehydration reactions. Morphology and cathodoluminescence response were found to be the most reliable petrogenetic indicators. Trace element abundances generally support petrographic evidence, but protolith bulk composition exerts a strong control over absolute element abundance in kyanite. Sample-normalised concentrations show distinctive differences between petrogenetic types, particularly for Mg, Ti, V, Cr, Mn, Fe and Ge. LA-ICP-MS element maps, particularly combined to show Cr/V, provide additional information about changing geochemical environments during kyanite growth. Most kyanite in the studied migmatitic leucosomes is of sub-solidus origin, with less widespread evidence for peritectic crystallisation. Where present, grain rims commonly crystallised directly from the melt; however, entire grains crystallised exclusively from melt are rare. The presence of kyanite in leucosomes does not, therefore, necessarily constrain the P–T conditions of melting, and the mechanism of growth should be determined before using kyanite as a P–T indicator. This finding has significant implications for the interpretation of kyanite-bearing migmatites as representing early stages of melting during Himalayan evolution.

Stishovite is a key mineral for understanding the deep Earth water cycle because of its potential as a main carrier for water into the transition zone and lower mantle. During subduction-related metamorphism of basaltic oceanic crust, stishovite stabilizes at 8–9 GPa and comprises 10–25 vol% of the bulk mineralogy, with some experimental studies indicating that stishovite can accommodate 3.5 wt% H2O or more in the transition zone and upper lower mantle. This large water solubility has been explained by a hydrogarnet substitution mechanism (1Si4+ ↔ 4H+) and/or the incorporation of interstitial molecular water. To investigate water speciation and hydrogen isotope behavior, we synthesized partially deuterated hydrous stishovite at 9 GPa and 450 °C in a multi-anvil press (MA). The hydrous stishovite contains on average 1.69 ± 0.05 wt% water, which is consistent with earlier MA studies but is significantly lower than the 3.5 wt% reported from in situ diamond anvil cell (DAC) studies made at higher pressures and temperatures. 1H MAS NMR spinning sideband characteristics suggest a high abundance of interstitial molecular water in hydrous stishovite, while the presence of a hydrogarnet defect cannot be ruled out. Unit-cell volumes and deuterium enrichment in the quenched hydrous stishovite indicate that ~ 45% of the water is lost from the stishovite upon quenching and decompression of the experiment, consistent with a higher solubility. This implies that the pristine water contents of a P–T–fO2 equilibrated hydrous stishovite cannot be quenched to 1 atm and room temperature from classical MA experiments. We further present a capillary-based recovery method for fluid from experimental capsules, allowing direct determination of the D/H ratio of the experimental fluid and indirect determination of the hydrous stishovite. Using Rayleigh modeling to account for the quench-related water loss, we find that, at 450 °C and 9 GPa, deuterium is 3.5–4.5 times enriched in hydrous stishovite relative to coexisting aqueous fluid. This is opposite of what is commonly observed for mineral–fluid pairs above 300 °C, rendering hydrous stishovite a potential sink for deuterium and decreasing the D/H ratio of coexisting aqueous fluids. Partial decomposition (30–60%) of hydrous stishovite during mantle upwelling and production of primary basaltic melts could be accompanied by high-temperature D/H fractionation, decreasing the hydrogen isotope composition of such melts towards “mantle-like” δD values between −75 and −220‰.

Granitic rocks are crucial in revealing continental crust formation and evolution. The classic S- and I-type granitoid classification (based upon metasedimentary vs. metaigneous protolith) is widely adopted, but is not always straightforward when diagnostic minerals (e.g., cordierite for S-type and amphibole for I-type) are absent. In this study, zircon oxygen isotope and water content were measured simultaneously with Secondary Ion Mass Spectrometry (SIMS) on I- and S-type granitoids from various tectonic settings, including the New England Orogens (NEO), the Diancangshan–Ailaoshan and the Himalayan orogens, and the South China Block. Our results indicate the S-type samples are characterized by lower water-in-zircon content (median 14–170 ppm) than the I-type samples (median 449–885 ppm). Also, inherited I-type zircon cores in S-type samples from the NEO have higher water-in-zircon content than the zircon rims, confirming the different water content between the two granitoid types. We propose that water-in-zircon content is a potential tool to distinguish I- and S-type granitoids. Nonetheless, the low water-in-zircon content of S-type granitoids has likely no correlation with high magma water content.

In this paper, we present mineralogical, whole-rock geochemical, Sr and Pb isotopic, and zircon trace elemental and U–Pb geochronological data for the newly-discovered Pengco listvenite in central Tibet, in an attempt to probe the petrogenesis of listvenite and the tectonic environment. The listvenite is composed of magnesite, quartz and minor dolomite and Cr-spinel. Chemistry of abundant protogenetic Cr-spinel in the Pengco listvenite as well as the whole-rock Cr and Ni concentrations and low rare-earth element contents indicates the harzburgite protolith. Addition of Ca and K and redistribution of Si at sample scale suggest that the formation of Pengco listvenite was not an isochemical process. Serpentinization could partly accounts for the enrichments in some fluid mobile elements before listvenitization. The εHf(t) values of xenocrystic zircons in the Pengco listvenite are dominantly negative, suggesting a continental crustal source. The zircon age spectra and high whole-rock initial 87Sr/86Sr values (0.7094–0.7107, and one up to 0.7193) and high 207Pb/204Pb (15.750–15.759), 208Pb/204Pb (39.145–39.215) and 206Pb/204Pb (18.542–18.567) of the Pengco listvenite are similar to those of the Qiangtang continental sediments/upper crust, indicating the terrigenous source of reactant fluids and fluid-assisted solid-state transfer during the formation of listvenite. The Pengco listvenite is likely formed in the forearc extensional environment of the Bangong Meso-Tethyan ocean during the Late Jurassic.

Fluocerite has been shown experimentally and by thermodynamic modeling to be an important phase for the evolution of magmatic-hydrothermal REE mineralizing systems, but one that is rarely observed in nature. The reason for the latter is poorly understood. The newly discovered Taipingzhen REE deposit in Henan province, Central China, is unique for containing abundant fluocerite and thus provides a great opportunity to explore the role of fluocerite in REE mineralization. Fluocerite was the earliest REE mineral to crystallize in this deposit and was replaced extensively by bastnäsite. It has a higher La/Nd ratio than the later-depositing REE minerals, e.g., bastnäsite and monazite. This decrease in the La/Nd ratio with time is interpreted to have resulted from the fluocerite-mediated fractionation of REE (fluocerite prefers the lightest REE during its deposition from a high temperature, sulfate-rich hydrothermal fluid, in which the REE were transported as sulfate complexes). This study provides compelling evidence that fluocerite is, indeed, an important, early crystallizing mineral in magmatic-hydrothermal REE systems, which elsewhere was replaced entirely by bastnäsite during subsequent fluid evolution. Moreover, the study shows that the bastnäsite, which replaced fluocerite, inherited its high La/Nd ratio, suggesting that this ratio could be used as a tool with which to identify the former presence of fluocerite. The preservation of fluocerite in the Taipingzhen deposit is interpreted to have been due to the lack of significant cooling until a late stage of mineralization, at which time most of the fluid was lost from the system.

In this work, we studied the phase relationships in the NaAlSi2O6–CaMgSi2O6–CO2 system in the range of 3–6.5 GPa and 900–1500 ℃ using a multianvil press. The duration of most experiments varied from 5 to 9 days. Equilibrium was examined from both sides using oxide-carbonate mixtures and diopside-jadeite glasses with Ag2C2O4 as starting materials. We found that the subsolidus assemblage is represented by clinopyroxene, coesite, dolomite, and CO2 fluid. The solidus of the system passes through 1015 ℃/3 GPa and 1095 ℃/6 GPa, and has a slope of 38 MPa/℃. The near-solidus carbonate melt contains 1–4 mol% SiO2 and 1–4 mol% Na2O. We found that clinopyroxene can be in equilibrium with CO2 fluid under P–T conditions corresponding to the diamond stability field to at least 6–6.5 GPa and 1050–1100 ℃. Yet, the composition of such clinopyroxene corresponds to almost pure jadeite, Jd# 95–97. Thus, the infiltration of CO2 fluid into eclogite under conditions of the continental lithospheric mantle should be accompanied by partial carbonation of clinopyroxene with the formation of a low-Mg carbonate melt with moderate concentrations of sodium. As a result of the reaction, the compositions of clinopyroxenes of eclogites of Groups A and B should shift towards a high-jadeite composition corresponding to eclogites of Group C.

A post-rift Aptian magmatism is recorded in a 500-m-thick sequence of basalts interbedded with marls in the Bacalhau oil and gas field in Santos Basin, SE Brazil. This magmatic section is within the so-called pre-salt sequence of Santos Basin that comprises the major oil and gas reserves of Brazil. This is the first publication of systematic petrological and geochronological data for the Aptian magmatism in Santos. Whole-rock Ar–Ar integrated ages obtained for these basalts are 116.93 ± 0.22 Ma, 116.16 ± 0.10 Ma, 115.21 ± 0.13 Ma and 109.95 ± 0.20 Ma and. As such, they are younger than the rift-related Camboriú basalts in Santos as well as the Paraná-Etendeka basalts and related dike swarms. The Santos basalts comprise a low-Ti tholeiitic suite with La/Nbn (2.7–4.2) and La/Ybn (4.2–5.9) ratios typical of continental flood basalt provinces. The basalts vary in MgO content but show no evidence for cogeneticity by differentiation processes. Lithogeochemical data showed that the Aptian basalts in Santos cannot be related with either the low-Ti, Esmeralda and Gramado suites in Paraná-Etendeka or the low-Ti Lumiar, Serrana, and Costa Azul suites in the Serra do Mar Dike Swarm on the basis of lithogeochemical data. No geochemical and isotopic correlation can be done with the Aptian, Alagoas basalts in Campos basin as well. Initial (115 Ma) isotope ratios (Sr: 0.705747–0.706804; εNd: −5.9 to −2.8; 206Pb/204Pb: 17.61–18.67; 207Pb/204Pb: 15.47–15.58; 208Pb/204Pb: 38.17–38.39; εHf: + 0.3 to −8.2) indicate derivation from a EM1 mantle component in the SCLM. Modal batch partial melting modelling showed that melting occurred within the garnet stability zone. We propose a geodynamic model for the Aptian in Santos in which the melting of the SCLM is related with either the presence of the Tristan da Cunha mantle plume in Aptian time below Santos or stretching of different portions of the blob-rich SCLM itself. This stretching is due to the remaining heat advected from Tristan during the voluminous Early Cretaceous magmatism that gave rise to the Paraná-Etendeka CFB.