Search

Alkalic basalts from continental regions have been widely used to fingerprint chemically and lithologically distinct components in the mantle. However, superimposed on mantle source effects are the effects of intra-crustal magma differentiation, which may be underestimated when exclusively considering fractional crystallization (FC) in melt-rich magma chambers. Here we present extensive mineralogical and geochemical data for mafic cumulates and dolerite dikes from central Tibet to demonstrate that tholeiitic magma differentiation occurred in crystal-rich mushes and melt-filled fractures. The mafic cumulates provide a snapshot of the percolation of water-poor (<0.9 wt% H2O), oxide-undersaturated, tholeiitic melts through slightly hotter mushes (50–60% crystals) with a framework of primitive olivine ± plagioclase crystals. The highly corroded olivines indicate that thermo-chemical disequilibrium between the later percolating melts and pre-existing crystal framework triggered melt–mush reactions in an open system. High-temperature (>1060°C) reactions were accompanied by a decreasing melt mass, causing the initial tholeiitic melts to evolve into alkalic mafic melts during slow cooling. This process involved the dissolution and reprecipitation of olivine, plagioclase, and clinopyroxene as well as the late crystallization of high-TiO2 (up to 3.4 wt%) clinopyroxene and kaersutite. In contrast, the same tholeiitic magmas (i.e. dolerite dikes) that intruded the cold country rocks along fractures produced silica-rich residual melts saturated in Fe–Ti oxides and sodic plagioclase by FC during rapid cooling. The residual melt was always in equilibrium with the instantaneous crystal during closed-system FC, so there was no reaction between the interstitial melt and adjacent crystal even after a high degree of crystallization. Alkalic mafic melts that can be formed by melt–mush reactions exhibit strong enrichment of incompatible element contents (e.g. Ti and alkali and rare earth elements) and ratios (e.g. Ce/Y), but little Si enrichment and Mg depletion, compared with the parental tholeiitic magma, which cannot be reproduced by simple FC. We argue that tholeiitic melt–mush reactions at mid- to lower-crustal pressures (≥0.5 GPa) are a new mechanism for the generation of interstitial alkalic mafic melts in continental mush reservoirs. Continental alkalic basalts could be produced by this mechanism if such interstitial melts were concentrated into melt lenses by compaction and then transported rapidly upward to the surface by focused flow. [ABSTRACT FROM AUTHOR]

Two generations of garnet are recognized in a granite and a pegmatite from the Gangdese orogen in southeastern Tibet on the basis of a combined study of petrography, major and trace element profiles, and garnet O isotopes. Zircon U-Pb dating and Hf-O isotope compositions also help constrain the origin of both granite and pegmatite. The first generation of garnet (Grt-I) occurs as residues in the center of garnet grains, and it represents an early stage of nucleation related to magmatic-hydrothermal fluids. Grt-I is dark in backscattered electron (BSE) images, rich in spessartine, and poor in almandine and grossular. Its chondrite-normalized rare earth element (REE) patterns show obvious negative Eu anomalies and depletion in heavy REE (HREE) relative to middle REE (MREE). The second generation of pegmatite garnet (Grt-II) occurs as rims of euhedral garnets or as patches in Grt-I domains of the pegmatite, and it crystallized after dissolution of the preexisting pegmatite garnet (Grt-I domains) in the presence of the granitic magma. Compared with Grt-I, Grt-II is bright in BSE images, poor in spessartine, and rich in almandine and grossular contents. Its chondrite-normalized REE patterns exhibit obvious negative Eu anomalies but enrichment in HREE relative to MREE. The elevation of grossular and HREE contents for Grt-II relative to Grt-I domains indicate that the granitic magma had higher contents of Ca than the magmatic-hydrothermal fluids. The garnets in the granite, from core to rim, display homogenous profiles in their spessartine, almandine, and pyrope contents but increasing grossular and decreasing REE contents. They are typical of magmatic garnets that crystallized from the granitic magma. Ti-in-zircon temperatures demonstrate that the granite and pegmatite may share the similar temperatures for their crystallization. Grt-II domains in the pegmatite garnet have the same major and trace element compositions as the granite garnet, suggesting that the pegmatite Grt-II domains crystallized from the same granitic magma. Therefore, the pegmatite crystallized at first from early magmatic-hydrothermal fluids, producing small amounts of Grt-I, and the fluids then mixed with the surrounding granitic magma. The U-Pb dating and Hf-O isotope analyses of zircons from the granite and pegmatite yield almost the same U-Pb ages of 77–79 Ma, positive εHf(t) values of 5.6 to 11.9, and δ18O values of 5.2 to 7.1‰. These data indicate that the granite and pegmatite were both derived from reworking of the juvenile crust in the newly accreted continental margin prior to the continental collision in the Cenozoic. [ABSTRACT FROM AUTHOR]

The initial Hf isotope composition of zircon in high-grade metamorphic rocks is an important tracer for protolith origin as well as for the growth and evolution of continental crust. However, less attention has been paid to the behavior of Lu-Hf isotopes in peritectic reactions during partial melting of the continental crust. In this paper, we present a combined study of zircon U-Pb ages, Lu-Hf isotopes and trace elements as well as petrology and whole-rock geochemistry for migmatites close to the eastern Himalayan syntaxis in southeastern Tibet. The results show that peritectic zircons in the migmatites have significantly higher 176Hf/177Hf ratios than their protolith zircons. More interestingly, the 176Lu/177Hf ratios and the contents of high field strength elements (such as Nb, Ta, and Hf) and heavy rare earth elements are significantly elevated in the peritectic zircons compared to the protolith zircons. This suggests that garnets with high Lu/Hf ratios and other minerals (such as titanite, ilmenite, amphibole, and biotite) with moderate Lu/Hf ratios were decomposed during crustal anatexis, contributing to the Hf isotope composition of peritectic zircons. This is confirmed by petrographic observations that garnet and biotite occur as residues or remnants in mesosomes (biotite gneiss) and some leucosomes (felsic veins), but are absent in the melanosomes of the target migmatites. Therefore, the peritectic zircons were produced together with anatectic melts through peritectic reactions and acquired the elevated 176Lu/177Hf ratios compared to the inherited zircons from the protoliths if no peritectic garnet was coevally produced during the crustal anatexis. In this regard, the greatly elevated Hf isotope compositions of peritectic zircons in anatectic granites cannot faithfully reflect the Hf isotope composition of parental rocks. As such, great care must be taken when using Hf isotope ratios to trace the nature of parental rocks with respect to magmatic processes and crustal evolution. • Peritectic zircon have higher Lu/Hf and Hf isotopic ratios than protolith zircons. • Peritectic zircon also have higher HFSE contents than protolith zircons. • Minerals with high Lu/Hf ratios than zircon were decomposed during crustal anatexis. [ABSTRACT FROM AUTHOR]

• Triassic rocks in northern Qiangtang Block include basalts-rhyolites (237–235 Ma). • Basalts were originated from metasomatized mantle sources. • Rhyolites were generated from newly formed basaltic lower crust. • The formation of studied suite was related to Paleo-Tethys oceanic slab subduction. Triassic magmatic rocks are widely exposed in the northern Qiangtang Block in Tibet. However, their petrogenesis and relationship to the evolution of Tethys Ocean are matters of debate. Here we report zircon U-Pb ages, mineral compositions, whole-rock major and trace elements, and Nd isotope for newly reported Triassic volcanics in order to provide constraints on the formation of these rocks in Tethys ocean's evolution. The volcanic suite consists of grey-porphyritic basalts (B1), black-aphyric basalts (B2), and rhyolites. SIMS, LA-ICPMS zircon U-Pb age dating shows that they were erupted at 237–235 Ma. All rocks show no continuous trends in major-element compositions, and are enriched in Light Rare Earth Elements (LREE) relative to heavy REE (HREE) with various Eu-anomalies. Basalts have narrow εNd(t) range (−0.90 to −0.01), slightly higher than those of rhyolites (−4.36 to −0.66). Given similar Nd-isotope and REE patterns, B1 and B2 basalts were probably produced from a common parental magma by mineral accumulation and fractional crystallization during late- and early-stages of magma evolution, respectively. Based on indicative element-ratios of slab-derived materials, we suggest that parental magmas of these basalts were derived by partial melting of metasomatized mantle source with subducted oceanic slab component. Rhyolites were generated by partial melting of basaltic lower-crust as a result of underplating by magma that corresponds to the studied basalts. Given spatial–temporal distribution of Triassic magmatic-metamorphic rocks, and geological background, we suggest that magmatism of studied suite was triggered by northward subduction of the Paleo-Tethys ocean, and recorded late-stage of Paleo-Tethys Wilson cycle. [ABSTRACT FROM AUTHOR]