No abstract

No abstract

No abstract

No abstract

No abstract

Himalayan leucogranites of Cenozoic age are generally attributed to partial melting of metasedimentary rocks at low temperatures of <770 °C. It is unknown what the spatial distribution and characteristics of high-temperature (>800 °C) leucogranites are in the Himalayan orogen. The present study reports the occurrence of such leucogranites in the collisional orogen. We use the Ti-in-zircon thermometry in combination with the thermodynamically calibrated relationships of T-aSiO2-aTiO2 to retrieve crystallization temperatures of Miocene (ca. 17 Ma) two-mica granites from Yalaxiangbo, in the eastern Himalaya, SE Tibet. The results give the maximum temperature as high as ∼850 °C for granite crystallization, providing a significant constraint on the nature of thermal sources. Phase equilibrium modeling using metasedimentary rocks as the source rocks indicates that felsic melts produced at ∼850 °C and 6−10 kbar can best match the target leucogranites in lithochemistry. In this regard, the anatectic temperatures previously obtained for the production of Himalayan leucogranites would probably be underestimated to some extent. Such high temperatures are difficult to explain purely by the internal heating of the thickened orogenic crust. Instead, they require an extra heat source, which would probably be provided by upwelling of asthenospheric mantle subsequent to thinning of the orogenic lithospheric mantle by foundering along the convergent plate boundary. Therefore, the Himalayan leucogranites of Miocene age would be derived from partial melting of the metasedimentary rocks in the post-collisional stage.

No abstract

Amphibolite- and granulite-facies metamorphic rocks are common in the eastern Himalayan syntaxis of southeastern Tibet. These rocks are composed mainly of gneiss, amphibolite and schist that underwent various degrees of migmatization to produce leucogranites, pegmatites and felsic veins. Zircon U–Pb dating of biotite gneiss, leucocratic vein and vein granite from the syntaxis yields consistent ages of ∼49 Ma, indicating crustal anatexis during continental collision between India and Asia. Garnets in these rocks are categorized into peritecitc and anatectic varieties based on their mode of occurrence, mineral inclusions and major- and trace-element zoning. The peritectic garnets mainly occur in the biotite gneiss (mesosome layer) and leucocratic veins. They are anhedral and contain abundant mineral inclusions such as high-Ti biotites and quartz, and show almost homogeneous major-element compositions (except Ca) and decreasing HREE contents from core to rim, indicating growth during the P- and T-increasing anatexis. Peak anatectic conditions at 760–800°C and 9–10·5 kbar are well constrained by phase equilibrium calculations, mineral assemblages, and garnet isopleths. In contrast, anatectic garnets only occur in the vein granite. They are round or subhedral, contain quartz inclusions, and exhibit increasing spessartine and trace-element contents from core to rim. The garnet–biotite geothermometry and the garnet–biotite–plagioclase–quartz geobarometry suggest that the anatectic garnets crystallized at ∼620–650°C and 4–5 kbar. Some garnet grains show two-stage zoning in major and trace elements, with the core similar to the peritectic garnet but the rim similar to the anatectic garnet. Mineralogy, whole-rock major- and trace-element compositions and zircon O isotopes indicate that the two types of leucosomes were produced by hydration (water-present) melting and dehydration (water-absent) melting, respectively. The leucocratic veins contain peritectic garnet but no K-feldspar, have lower whole-rock K2O contents and Rb/Sr ratios, higher whole-rock CaO contents and Sr/Ba ratios, and show homogeneous δ18O values that are lower than those of relict zircons, indicating that such veins were produced by the hydration melting. In contrast, the vein granite contains peritectic garnet and K-feldspar, has higher whole-rock K2O contents and Rb/Sr ratios, lower whole-rock CaO contents and Sr/Ba ratios, and shows comparable δ18O values with those of relict zircons, suggesting that this granite were generated by the dehydration melting. Accordingly, both hydration and dehydration melting mechanisms have occurred in the eastern Himalayan syntaxis.