Contributions to Mineralogy and Petrology最新文献

The dehydration of antigorite is an important reaction in subduction zones with implications on both geochemical and geophysical processes. In this experimental study we focus on the onset of antigorite dehydration and investigate various chemical and physical parameters as possible drivers for the fluid release. We performed hydrostatic and co-axial Griggs experiments on antigorite serpentinites with variable chemical composition and microstructures at high-pressure and high-temperature conditions across the antigorite dehydration (1.5 GPa, 620–670 °C). For these conditions, our thermodynamic models predict the formation of olivine from magnetite decomposition and partial dehydration of antigorite. Detailed analyses of the run products reveal limited magnetite decomposition. Antigorite dehydration is restricted to samples that have been deformed. Nano-sized olivine and orthopyroxene formed locally in oblique dehydration bands and exhibit neither a clear crystallographic preferred orientation nor a topotactic relation with precursor antigorite. We argue that limited local dehydration in our experiments is related to strain and variations in reaction kinetics. Systematic investigation excludes mineralogical and chemical heterogeneities, and temperature gradients as reaction driving potentials. The structural relation of the dehydration bands suggests deformation-related dehydration, which is supported by numerical simulations that couple reaction kinetics with mechanical work rate and self-consistently predict dehydration bands. In this scenario, strain concentration due to applied axial stress locally increases the internal energy of antigorite to reach the activation energy of the dehydration reaction, enabling dehydration. This study highlights the importance of coupled mechanical and chemical processes and provides a mechanistic framework for deformation-induced dehydration of antigorite.

The Acasta Gneiss Complex (AGC) in northwestern Canada comprises Earth’s oldest known evolved crust, with zircon U–Pb ages up to 4.03 Ga. Several pulses of crustal generation and metamorphism are preserved in tonalitic and granitic gneisses spanning over one billion years, along with mafic and ultramafic rocks of unknown age. Major elements, trace elements and radiogenic isotope signatures have been invoked to suggest that these rocks preserve the local onset of horizontal tectonic processes. However, the behavior and influence of volatiles, which have a defining role in modern arc magmatism, remain unconstrained. Here we combine new whole-rock major- and trace-element data with multiple sulfur isotope analyses in 4.0–2.9 Ga Acasta gneisses and spatially associated mafic and ultramafic rocks to investigate the petrogenesis of the AGC. We use a recently-published major element-based melt hygrometer to estimate dissolved water contents for all published plagioclase-saturated Acasta meta-igneous rocks, and find modes at < 0.5 wt.% and 5 wt.% H₂O, similar to modern arc magmas. Tholeiitic and calc-alkaline trends are both present, with the former being more prominent in the oldest (ca. 4.0 Ga) samples and in mafic rocks. Zircon trace element oxybarometry reveals a shift towards more oxidized magmatic conditions by 3.75 Ga. Sulfur isotopes record a limited range in δ³⁴S values, suggesting a common igneous end-member at ~  + 1 ‰, and positively correlate with calculated H₂O contents, with more positive values (up to + 5‰) appearing in the Paleoarchean (< 3.6 Ga). The Eoarchean (4.0–3.6 Ga) δ³⁴S values are consistent with a precursor Hadean crust having an enriched sulfur isotope signature, possibly resulting from hydrous alteration or from isotopic fractionation during its formation. The temporal progression to more positive δ³⁴S values is consistent with a shift towards more hydrous and oxidized magmatic differentiation. Most samples have near-zero Δ³³S that fall along a mass-dependent fractionation (MDF) array, but one 3.5 Ga metasedimentary sample has a negative MIF Δ³³S signature of -0.60 ± 0.01 ‰. Additionally, two granitic gneisses dated at 3.3 and 2.9 Ga preserve small positive MIF Δ³³S values of + 0.08 ± 0.02 ‰, which could reflect recycling of sedimentary material via subduction by 3.3 Ga. Overall, our data indicate that the Acasta Gneiss Complex preserves several modes of crustal generation evolving over time, with an increasing importance of deep hydrous magmatism by 3.75 Ga and of sedimentary inputs by 3.3 Ga.

The nature, location, longevity and pressure-temperature conditions of different crustal melt reactions during orogenesis provide constraints on the structure, mechanical strength and exhumation of orogenic middle crust as well as element mobilisation and crustal differentiation. The Himalayan orogen offers a natural laboratory for studying crustal melting by exposing both migmatites and leucogranites in its structurally highest levels. We combine previous frameworks that link petrography or bulk geochemistry to melt reaction with in-situ trace-element analyses of large-ion lithophile elements in feldspar, mica, and garnet, U-Th-Pb isotopes in monazite and zircon and thermometry calculations in samples from the Badrinath region of the Garhwal Himalaya. Our samples naturally fall into three groups that we interpret as having formed by fluid-present melting of muscovite (Group 1, all migmatites; 650–750 °C), muscovite dehydration melting (Group 2, migmatites, leucosomes and leucogranites; 730–800 °C) and biotite dehydration melting (exemplified by a single leucogranite that contained zoned and inclusion-rich garnet and porpyroblastic K-feldspar). Geochronological data suggest that melting occurred over 20 Ma, with different samples experiencing different reactions and capturing different parts of the record at different times. Despite experiencing the same thermal history, individual outcrops typically only record one melting reaction instead of a progression through fluid-present melting followed by muscovite-dehydration melting. We interpreted this as being due to local compositional variations and availability of fluids. Our results show that petrographic observations and the mineral chemistry record are similar between (source) migmatites and (product) granites, but that fluid-present reactions are only documented in migmatites.

Nearly all geochemical applications of ionic radii appeal to the classic tabulation of Shannon (Acta Crystallographica, A32:751–767, 1976). In that work, smoothing was applied to the crystallographic systematics of lanthanides to ensure consistent decreases in cationic radii with increasing atomic number—the lanthanide contraction. Recent work has now updated preferred radii, based on a much larger database of crystal structures. However, values have not been smoothed, and several average radii violate the lanthanide contraction principle. Here, we propose a set of effective ionic radii for trivalent lanthanides using simple regressions based in part on atomic theory, and verify that these radii satisfy theoretical principles of lattice strain models. We then use prior analysis of partitioning to propose an ionic radius for Y³⁺. The ionic radius of Sc³⁺ cannot be refined to higher precision using partitioning data because its ionic radius is so different from other rare-earth elements (REE) and because it does not necessarily substitute into the same crystallographic sites as other REE. Recommendations are provided for ionic radii for REE-poor silicates and for monazite, xenotime, zircon, and apatite.

In this study 14 experiments were performed in a cold seal pressure vessel at conditions of 850–975 °C and 150–200 MPa to understand Cl and F partitioning between amphibole and dacitic to rhyolitic melts in arc settings. The starting compositions for the experiments were chosen to replicate basaltic andesite samples taken from Shiveluch volcano and a more evolved andesite. We find that F is moderately compatible in amphibole for crustal conditions, with D_F^amph/melt = 4.5–11, and Cl is incompatible, with D_Cl^amph/melt = 0.11–0.25. This adds to the growing experimental database for halogen partitioning between amphibole and melt, but some incongruity remains between mantle condition and crustal condition studies. We performed forward crystallization modeling on the effects of amphibole crystallization on arc volcanic halogen contents and find that the high Cl/F ratios observed in some olivine hosted melt inclusions at arcs may be explainable by the uptake of F into amphibole, providing further evidence for cryptic amphibole crystallization.

The petrogenesis of silicic (> 63 wt.% SiO₂) magmas found in continental flood basalt provinces involves mechanisms ranging from fractional crystallisation to crustal anatexis. In the western Deccan Traps, India, a swarm of tholeiitic basalt and basaltic andesite dykes exposed in the Vikramgad-Murbad area likely fed the 2 km-thick, exclusively tholeiitic Western Ghats volcanic sequence. Thermobarometric calculations for the dykes indicate co-crystallisation of plagioclase (An_{81 − 55}) (1181–1083 ± 14 °C), clinopyroxene (Mg# 78 − 63) (1176–1115 ± 35 °C) and olivine (Fo_{81 − 55}) (1168–1135 ± 23 °C), at pressures of 3.5–0.1 (± 2.2) kbar (1σ errors). A universal groundmass feature of these dykes is interstitial rhyolitic glass, containing innumerable microlites of plagioclase (An_{42 − 21}), anorthoclase (An₆Ab_{69 − 66}Or₂₈₋₂₅), Fe-rich augite (Mg# 46 − 12), pigeonite (Mg# 32 − 15), Fe-olivine (Fo_15 − 4), Fe-Ti oxides, and apatite. Textures, whole-rock geochemistry, mineral and glass chemistry, thermobarometry, mass balance calculations, and rhyolite-MELTS modelling show that advanced (~ 76–94%) closed-system fractional crystallisation of anhydrous to hydrous (0.5 wt.% H₂O) tholeiitic dyke magmas at low pressures (~ 3–0.5 kbar) could generate the residual rhyolitic melts. Rapid cooling and solidification of the dykes at a shallow crustal depth quenched these melts to glass and thus trapped them within the crystal mush. This explains the notable lack of silicic extrusive or intrusive units in the Western Ghats volcanic sequence, despite the widespread production of silicic melts at depth. Such interstitial silicic glasses, common in tholeiitic flood basalts worldwide, are evidence for silicic magma genesis by fractional crystallisation, though without residual silicic melt segregation, intrusion and eruption.

The role of mantle vs. crustal contributions to A-type granitoids are heavily debated, in line with the wide compositional diversity among these magma types. The Cambrian post-orogenic Venda Nova and Várzea Alegre Plutons of the Araçuaí belt, SE Brazil, turn out to represent an endmember case where the clear “crustal” isotopic signature is already present in the primitive mantle melts and hence mantle source. The mildly alkaline Venda Nova Pluton comprises wehrlites, gabbronorites, diorites, quartz syenites and titanite granites with mineral (clinopyroxene, amphibole, apatite) and bulk rock age-corrected εNd (500 Ma) values of -11.8 to -9.5 and ⁸⁷Sr/⁸⁶Sr (500 Ma) values of 0.7069–0.7080. The more calc-alkaline Várzea Alegre Pluton comprises gabbronorites, monzodiorites, charnockites, and monzo- and syeno-granites with mineral and bulk rock εNd (500 Ma) of -10.1 to -7.6 and ⁸⁷Sr/⁸⁶Sr (500 Ma) of 0.7067–0.7079 in the gabbronorites, but more variable values (0.7075–0.7097) in the charnockites and granites. Mixing models between the Venda Nova gabbronorites and potential crustal assimilants, combined with a mineral fractionation model, show that granite genesis primarily resulted from crystal fractionation from gabbronoritic melts with little crustal assimilation. For Várzea Alegre, the greater scatter in εNd points to a more heterogeneous mantle source, while the larger range of ⁸⁷Sr/⁸⁶Sr points to some localized assimilation of country rock metapelites or S-type granitoids. In both cases, the least differentiated gabbroic rocks already show geochemically enriched isotopic signatures which lack corresponding crustal compositions in the Araçuaí belt. We attribute this crustal signature to a lithospheric mantle source that was metasomatized both in ancient times to generate strongly negative Nd isotopes and more recently by fluids/melts related to subduction along the ~ 600 Ma Rio Doce arc, now incorporated into the Araçuaí belt. This contribution highlights that apparently “crustal” isotopic signatures in A-type granitoids can be derived from metasomatic enrichment of the lithospheric mantle rather than crustal assimilation.

Partial melting experiments on a mildly depleted peridotite (GKR-001) conducted in a piston-cylinder apparatus provide new insights into the solidus of anhydrous peridotite, melting relations and reactions, and residue compositions at 5 GPa. These have important implications for melting conditions relevant to cratonic lithosphere, the petrogenesis of komatiites and orthopyroxene-enriched peridotites. The solidus of GKR-001 was identified at 1575ºC. As melting progressed, clinopyroxene melted out at ~ 1650ºC, closely followed by garnet at ~ 1670ºC and orthopyroxene at ~ 1725ºC. Orthopyroxene remained stable throughout most of the melting interval. Clinopyroxene's incongruent melting initiated peritectic reactions (Ol + Cpx + Grt = Opx + Melt) at melting degrees ≥ 8%, forming additional orthopyroxene in the residue. Clinopyroxene exhausted at around 24% melting, while garnet-out was encountered at slightly higher melting degree (~ 34%). Orthopyroxene reached its maximum content in the residue after ~ 24% melting and was dissolved back at melting degrees > 45%. The melts formed at ≥ 8% melting are Ti-depleted komatiites (30.7–43.6 Al₂O₃/TiO₂), consistent with melting of a slightly depleted peridotitic source. Melts in equilibrium with a garnet-bearing residue have CaO/Al₂O₃ ratios > 1, while those formed above the garnet-out curve have ratios < 1. Residues from GRK-001 exhibit higher SiO₂ contents than those produced after the partial melting of pyrolytic compositions (KR4003). However, high-pressure partial melting of GKR-001, albeit producing quite orthopyroxene-rich residues at moderate degrees of melting, cannot produce the silica-rich peridotites observed in many cratonic lithospheres worldwide. Comparison between thermodynamic modelling and experimental results reveals inconsistencies in the garnet and pyroxenes stability fields and the absolute temperature of the solidus.

Crystal mush systems, often referenced in the context of large silicic magma bodies, involve the reactivation of a near-solidus crystal mush by heat input from mafic injections. This model suggests that interstitial melt is extracted from the mush, leading to the generation of high-silica rhyolites and granites. Such processes have been well-documented in various tectonic settings and contribute to both large-scale eruptions and the formation of granitic plutons. However, in the Mineral Mountains, Utah, the zircon and whole rock geochemical record indicate a different scenario. The presence of sector-zoned zircons and the absence of highly evolved central domains indicative of extraction from a mush suggest rapid magma generation from partial melting of solid granitoids rather than from a long-lived crystal mush. Fractional crystallization and equilibrium partial melting models support derivation from the granitoid bodies, rather than from a common shared parental rhyolitic magma or from coeval basalts. The proposed model, presented here, for rhyolite formation in the Mineral Mountains involves episodic injections of mafic magma into the crust, leading to localized partial melting of different granitoid lithologies. Partial melting up to 30% can produce isolated, ephemeral pools of high-silica melt, which crystallize zircons rapidly and ascend to form rhyolitic domes. This process is distinct from the long-lived crystal mush model, explains the lack of intermediate compositions, and the confinement of mafic eruptions to lower elevations. By integrating geochemical data, zircon morphology, and fractionation modeling, this study provides a comprehensive framework for understanding the magmatic processes at play in the Mineral Mountains.

The minerals zircon, monazite, xenotime and rutile commonly occur in metasedimentary rocks as detrital but also authigenic grains, which can result from different fluid-driven processes. In this study such processes are investigated on the basis of detailed petrographic observations, mineral-geochemical data, and results of in situ U-Pb dating and Nd isotope analyzes in a kaolinitized micaschist. The data provide evidence for the preservation of detrital grains of zircon, rutile and monazite crystallized between ~ 3100 and 1150 Ma, and for authigenic zircon, xenotime, rutile and monazite formed at ca. 530 Ma. The authigenic character is indicated by zircon outgrowths closely intergrown with xenotime and rutile crystals. The outgrowths occur where detrital zircon faces are intensely dissolved in contact with Fe-Mg-rich phengitic muscovite, suggesting the involvement of an aqueous fluid enriched in K-Mg-Fe-Al-Ti-P near the thermal peak at 510 °C and > 0.8 GPa. In contrast, authigenic monazite rims overgrowing rounded cores were formed during the retrograde evolution, as indicated by their occurrence in assemblage with kaolinite, and results of geothermobarometry (T = 280 °C, P < 0.3 GPa). The monazite rims show very low U contents (4–14 µg/g), extremely high Th/U (up to 1670), and nearly identical ¹⁴³Nd/¹⁴⁴Nd_t (0.51184 ± 0.00006) values that markedly differ from those of the detrital cores (¹⁴³Nd/¹⁴⁴Nd_t = 0.51000 ± 0.00050; U = 162–16418 µg/g; Th/U = 2.6–153). The high Th/U points to the involvement of an oxidizing fluid, in line with late goethite formation. Monazite microstructures, and Nd isotope characteristics require a chain of processes, from partial dissolution of detrital monazite, through REE transport and Nd isotope homogenization in an aqueous fluid, to new monazite growth.