Linking beryllium enrichment to crystal-melt separation in granitic magmatic systems: Insights from high-silica granites in the southern Great Xing’an Range, NE China

Abstract

High-silica granites are significant carriers of highly incompatible elements and are closely associated with mineralization of the rare metal beryllium. Thus, understanding their origin and evolution is of paramount importance for comprehending the evolution of the continental crust and enrichment processes of beryllium. This study presents zircon U-Pb ages, whole-rock major and trace element compositions, and Nd-Hf-O isotopic data for Early Cretaceous high-silica granite porphyries, monazite U-Pb ages and Nd isotopic data for beryllium-rich quartz veins, and whole-rock and apatite Nd isotopic data for Permian tuffs in the Dongshanwan W-Mo-Be deposit in the southern Great Xing’an Range, northeastern China. Our aim is to elucidate the link between mineralization and magmatism and to further our insight into the processes of crystal-melt separation and the mechanisms driving beryllium enrichment in granitic magmatic systems. Zircon U-Pb dating of the high-silica granite porphyry yields a weighted mean 206Pb/238U age of 141 ± 1 Ma, which is consistent with the age of hydrothermal monazite (ca. 140 Ma) from the beryllium-rich quartz veins. Hydrothermal monazite from the intra-granite porphyry quartz vein has positive εNd(t) values (+0.76 to +1.63) that overlap broadly with whole-rock Nd isotopic compositions of the host granite porphyry [εNd(t) = +1.24 to +1.61]. These characteristics indicate that the Dongshanwan beryllium mineralization was temporally and genetically associated with magmatic-hydrothermal activity of the Dongshanwan high-silica granite porphyries. Our systematic studies suggest that both the Dongshanwan high-silica granite porphyries and coeval Be-barren high-silica granites in the study area are high-K calc-alkaline I-type granites. They originated from a shared magmatic system and formed by the partial melting of a predominantly juvenile medium- to high-K basaltic crustal source, with a minor addition of old crustal components. Rayleigh fractionation modeling indicates that the granitic magma underwent a two-stage crystal-melt separation process. During the first stage, the fractional crystallization of a large proportion of plagioclase, in which beryllium is compatible, not only effectively inhibited beryllium enrichment in the differentiated melt, but also removed a large amount of Ca. During the second stage, minerals in which beryllium is incompatible dominated the fractionating assemblage owing to the low Ca contents in the magma, resulting in a surge in beryllium concentration in the differentiated melt. Our findings reveal that beryllium cannot be appreciably enriched in calc-alkaline granitic magmatic systems until plagioclase fractionation has substantially removed Ca from the system.