Contributions to Mineralogy and Petrology最新文献

Upper Cretaceous to Miocene continental volcanism in NE Brazil spans 350 km in a N–S direction and 60 km in width, forming the Macau-Queimadas alignment (MQA). This study combines fieldwork, petrography, geochemistry, and Sr–Nd–Pb isotopes to explore its origin and evolution. The MQA consists of volcanic and hypabyssal mafic rocks intruding Cretaceous and Precambrian basement rocks, divided into two groups: (i) alkaline (foidite to trachy-basalt); and (ii) subalkaline (basalt and basaltic andesite). Both are sodic and LREE-enriched, with distinct La/Yb ratios. The alkaline group reflects an asthenospheric source (Nd model age of 1.1–0.4 Ga), while the subalkaline group incorporates an older lithospheric component (Nd model age of 2.1–1.2 Ga). These magmas originated from picritic parental melts, with < 15% melting for the alkaline group and ~ 25–30% melting for the subalkaline group, derived from spinel- to garnet-bearing peridotite. Differentiated series formed by successive small melt volumes, with some samples undergoing crustal fractional crystallization of clinopyroxene + olivine + plagioclase (alkaline group), and clinopyroxene + orthopyroxene + Ca-plagioclase (subalkaline group). The persistence of basaltic magmatism over ~ 90 Myr indicates sustained upper mantle melting. The alignment of volcanics, its association with a positive geoid anomaly, and its parallelism with the Mid-Atlantic Ridge suggest the MQA may represent an aborted ridge that never progressed to an oceanic stage.

The Marion Rise, located in the central portion of the Southwest Indian Ridge (SWIR), marks a relief high but is overall covered with a thin crust, and thus is inferred to be supported by depleted buoyant mantle. However, direct evidence of the regional mantle compositions from abyssal peridotites are still rare for such a hypothesis. This study presents whole rock and mineral compositions of 34 abyssal peridotites dredged from 7 sites between the Discovery and Indomed fracture zones on the Marion Rise. The samples are divided into low-Cr# (Cr# = 0.23–0.33) and high-Cr# (Cr# = 0.40–0.57) groups. The high-Cr# group samples display highly refractory characteristics (whole rock Al₂O₃ contents down to 0.52 wt%), which are reinforced by the depleted pyroxene compositions that indicate partial melting of up to > 18%. Nonetheless, the overall high extents of melting indicated by the peridotites are inconsistent with the regional thin crust, hence require an inherited origin of the melting signatures. Moreover, the Ti and Yb (Y) concentrations of clinopyroxenes (orthopyroxenes) in the high-Cr# group are too depleted to be residues of anhydrous melting at mid-ocean ridges, but instead suggest for a hydrous melting history near subduction zones. Collectively, we fill in a piece of the puzzle of mantle heterogeneity beneath the SWIR, by providing solid evidence for the existence of a highly refractory mantle beneath the Marion Rise. These mantle components carry subduction-modified characteristics, and very likely have a recycled mantle wedge origin.

The recent eruptions at Kīlauea Volcano (Hawai‘i) raised some fundamental questions on the longevity and the preservation of eruptible magma batches left over from previous eruptions. Fingerprinting magma batches at Kīlauea through time with bulk and glass compositions is challenging. Narrow compositional changes (e.g., Nb/Y ratio) in matrix glasses occur over time because of repeated magma mixing, and residence timescales of stored evolved magmas in the lower East Rift Zone are underconstrained. To evaluate the diversity in composition and the minimum residence timescales in Rift Zone magmas, we analyzed major and trace elements in plagioclase and matrix glasses from selected samples that erupted in the first weeks of the 2018 Kīlauea eruption. Plagioclase crystals in these samples represent mixed populations with a range in composition spanning An_30-80, corresponding to rhyodacitic to basaltic compositions and temperatures from 950 to 1200 °C. Diffusion modeling of Mg in these plagioclase crystals indicate minimum crystal residence timescales that range from < 1 to ~ 480 years. The complex zoning patterns in plagioclase (and resorptions) together with the protracted storage timescales from diffusion modeling imply that magmas from the East Rift Zone accumulated various plagioclase populations recording magma mixing events that occurred a few years to a few centuries before the 2018 eruption. The diversity of the magma batches (observed with An-Mg compositions in plagioclase) erupted in a single eruption offers research pathways to potentially estimate the frequency, volume and eruptibility of these evolved magmas, thereby refining the risk in the region.

The pressure and temperature conditions of the transition from spinel to garnet as the stable aluminous phase in peridotite lithologies of the upper mantle is integral to elucidating the tectonic significance of the ‘garnet signature’ in basalts. It provides an essential constraint on models of mantle partial melting and oceanic crust formation. Existing experimental results on the univariant phase transition in the simple systems MgO-Al₂O₃-SiO₂ (MAS) and CaO-MgO-Al₂O₃-SiO₂ (CMAS) are mutually inconsistent. To resolve this, we have re-determined the P-T coordinates of the univariant transition in both synthetic systems by running experiments containing both systems simultaneously in the piston-cylinder apparatus, along with the MgO-ZnO pressure sensor. These experiments show a ~ 0.4 GPa difference in the pressure of the spinel/garnet phase transition between the two chemical systems at 1400 ºC, double that inferred from a compilation of existing experimental data. Absolute pressure in these experiments can be verified using the MgO-ZnO sensor. The results imply that the thermodynamic data used in recent mineral equations of state based on the Holland-Powell thermodynamic dataset are substantially correct.

Numerical models for the growth of garnet are presented to evaluate the relative significance of reaction-limited growth and diffusion-limited growth following garnet nucleation after significant overstepping of the equilibrium garnet-in reaction. Reactions are only permitted among phases that are adjacent across grain boundaries and the extent of reaction at a given reaction site is scaled to the local amount of chemical affinity available to the two or three reactant phases relative to the grain boundary composition. This local affinity is dissipated as the local reaction proceeds, which changes the composition of the adjacent grain boundary “phase” and sets up chemical gradients that drive diffusion along the grain boundaries. Reactions proceed until all affinity is exhausted at which point the rock is essentially at equilibrium. Two extremes are modeled. Reaction-limited growth is modeled as infinitely rapid grain boundary diffusion whereas diffusion-limited growth is modeled by assuming that reactions proceed infinitely fast such that the supply of nutrients and removal of waste products from a reaction site is restricted by the rate of diffusion. Models are presented with model assemblages chlorite + quartz + garnet and chlorite + quartz + muscovite + biotite + plagioclase + garnet. Reaction-limited models result in garnets displaying well-formed “bell-shaped” Mn zoning profiles with all garnet crystals showing similar amounts of growth and zoning profiles. Diffusion-limited models result in mineral growth or consumption that is texture-sensitive such that the amount of consumption or production of a phase depends on the location of the crystal in the sample and the proximity of other phases. For example, the total amount of garnet continues to increase for the duration of diffusion-limited models although locally an individual garnet crystal may first grow and then be consumed. Mn zoning in models with short diffusion times display distinct “peaks” in the central garnet cores, in contrast to the bell-shaped profiles in reaction-limited models. With increasing diffusion times, these Mn zoning profiles evolve towards bell-shapes. These models demonstrate that diffusion-limited growth of garnet porphyroblasts may result in textural and compositional complexities that are not encapsulated by bulk-rock thermodynamic modeling.

Abyssal peridotites can provide complementary information on the compositional features of the asthenosphere, as the refractory mantle within the asthenosphere contributes little to the genesis of mid-ocean ridge basalts (MORB). Here we present major and trace elements of ~ 70 abyssal peridotites from the Sparsely Magmatic Zone (SMZ) and Eastern Volcanic Zone (EVZ) of the Gakkel Ridge, which are residues of the asthenosphere that have undergone < 15% partial melting. Their clinopyroxenes display LREE-depleted and LREE-flat patterns, the latter of which resulted from refertilization by quasi-instantaneous melts in the melting zone beneath the mid-ocean ridge. Compositions of the Gakkel peridotites are highly variable along the ridge axis, which cannot be attributed to the spatial variation of temperature of the asthenosphere. The estimated degrees of melting of the Gakkel abyssal peridotites are higher than the values inferred by seismic thickness of ocean crust along the SMZ and EVZ. This implies the Gakkel abyssal peridotites inherit ancient melting prior to their entering the Gakkel Ridge, which also causes the crust-mantle decoupling in compositions. Moreover, compositions of the Gakkel peridotites differs significantly from subduction-related peridotites. We suggest the asthenosphere beneath the Gakkel Ridge is highly heterogeneous in compositions, which is the culprit of crust-mantle geochemical decoupling. Enriched MORB erupted in the SMZ region were derived from small amounts of enriched components within the asthenosphere, which cannot be represented by the abyssal peridotites exposed in this region.

The Himalayan orogeny caused partial melting of rocks within the Greater Himalayan Sequence (GHS), forming migmatites. The extensive occurrence of such migmatites in the lower structural level of the GHS (GHS_L) is a distinctive feature of the Western Arunachal Himalaya (WAH), situated in eastern part of the orogen; meanwhile leucogranite is predominantly found in the highest reaches of the GHS_L. A comprehensive multi-method study incorporating field observations, petrography, phase equilibrium modelling, geochemical analysis, and zircon U–Pb and monazite U–Th–Pb geochronology was conducted on migmatitic paragneiss and leucogranites from the GHS_L along the Bomdila-Tawang section of the WAH. P–T pseudosection modelling reveals a clockwise P–T path characterized by prograde burial and heating, significant melt production, and nearly isothermal decompression during melt solidification. Structural observations, including concordant and discordant relationships between leucosomes and gneissic bands, suggest that deformation established pathways for melt migration. Zircon U–Pb dates reveal bimodal protolith ages of ~ 1350 Ma (Ectasian) and ~ 900 Ma (Tonian). Insufficient zircon overgrowth (< 20 μm), likely due to extensive melt extraction during suprasolidus metamorphism, precludes younger age determination. Monazite U-Th-Pb age indicates peak metamorphism of the GHS_L at ca. 25–26 Ma, synchronous with MCT initiation in the WAH. Melt generation at peak metamorphic conditions in the GHS_L reached ~ 16 vol% in stromatic metatexites and ~ 26 vol% in layered diatexites and of these generated melts, > 50% escaped at depths of ~ 30–34 km. This extensive migration formed complex leucosome networks, contributing to regional leucogranite distribution and rheological weakening, enabling ductile flow within the GHS.

The tectonic re-equilibration after the Variscan orogeny coincided with widespread early Permian post-collisional magmatism in southern Europe. A full understanding of the origin of this magmatism in the South Variscan realm and its relationship to major tectonic events such as subduction, continental collision, rifting or lithospheric foundering hinges on high-precision geochronological data of the magmatic products. Here, we present new high-precision zircon U–Pb geochronological data obtained by chemical abrasion isotope dilution thermal ionization mass spectrometry (CA-ID-TIMS) for the early Permian Athesian Magmatic District (AMD) in NE Italy. Our analysed zircons from felsic intrusive and volcanic rocks give ages spanning from ca. 281.8 to 277.2 Ma, suggesting that the lifetime of the AMD was significantly shorter than previously reported. Our data, when combined with recent high-precision ages from other South Variscan magmatic systems suggest that the Cisuralian (early Permian) post-collisional magmatism in the Southalpine domain occurred over more than 8 m.y. with the magmatic centres migrating from the western to the eastern Southern Alps. Geochemical and radiogenic isotope modelling of published data for magmatic rocks in the Southern Alps and the Corsica-Sardinia batholith suggest a subduction-enriched mantle source for the South Variscan post-collisional magmatism, with melting occurring under a relatively thin lithosphere at depths of ca. 60 km. Our results point to a significant post-orogenic delamination of the thick lithospheric mantle formed during the Variscan orogeny. In this scenario, the migration of the post-collisional magmatism within the Cisuralian district may be due to the lateral migration of the lithospheric foundering.

Water as structural hydroxyl in olivine plays an important role in determining the water budget of the upper mantle and its numerous physicochemical properties. However, the solubility of water in olivine in the deep upper mantle (i.e., 300–410 km depth), which defines the maximum water content under given conditions, still needs to be known with high precision. We examined the water solubility by annealing experiments under conditions controlled by Fe-FeO buffer and peridotite assemblages at 10–13 GPa and 1100–1450 ºC, using a starting olivine of representative chemistry and different fluid materials. The experimental conditions were broadly consistent with those prevailing in the deep upper mantle. The attainment of equilibrium water incorporation in the H-annealed olivine samples was ensured by H diffusion kinetics, water profile analyses and time-series studies. The annealed samples demonstrate infrared hydroxyl bands at 3650–3000 cm⁻¹, but the relative band patterns are different from those observed in the available H-annealing experiments at 1–7 GPa under otherwise comparable conditions (including starting materials). The obtained solubility of water increases with increasing both temperature and pressure over the run conditions, and differs apparently between the runs equilibrated by different fluids that are relevant to the deep upper mantle and water solubility studies. In general, the water solubility of olivine increases nonlinearly with increasing depth in the upper mantle, and can be described as: C_w = (290 ± 78) × exp ((0.0043 ± 0.0006) × depth (km))– (268 ± 89) (H₂O as coexisting fluid) and C_w = (149 ± 72) × exp ((0.0046 ± 0.0011) × depth (km))–(132 ± 85) (CH₄-H₂O as coexisting fluid), where C_w is water solubility (ppm wt. H₂O). The water solubility of olivine in the realistic upper mantle should be defined from the runs coexisting with CH₄-H₂O, and the highest value is only ~ 800 ± 80 ppm wt. H₂O, implying that the actual water contents of olivine in the upper mantle must be mostly (if not exclusively) lower. The inferred storage capacity of water in peridotite in the upper mantle reaches its maximum of 600 ± 100 ppm wt. H₂O (95% confidence level) at the bottom boundary of ~ 410 km depth, and a minimum of 350 ± 50 ppm wt. H₂O (95% confidence level) is expected at mid-depths of 190–230 km. During the upwelling of relatively water-rich materials from the source regions of enriched mid-ocean ridge basalts or ocean island basalts, hydrous melting would be much easier to trigger at the mid-depths of the upper mantle. The data further suggest that, to produce a pervasive hydrous melting at the ~ 410 km depth, the prevailing water content of the mantle transition zone should be greater than ~ 600 ppm wt. H₂O.