Subduction metasomatism and collision-related metamorphic dehydration controls on the fertility of porphyry copper ore-forming high Sr/Y magma in Tibet

Abstract

Processes generating fertile magma and associated porphyry Cu deposits in post-collision tectonic settings remain uncertain. Dabu, located in the Gangdese porphyry Cu belt of southern Tibet, is a typical porphyry Cu–Mo deposit that formed in a post-collision setting. The deposit records two stages of magmatism: an Eocene (46±1Ma) barren monzogranite (EG); and a fertile Miocene (15±1Ma) monzogranite porphyry (MMP). These intrusions have similar Sr–Nd (87Sr/86Sr(i)=0.70514 to 0.70573, εNd(t)=−2.78 to +0.77) and zircon Hf (εHf(t)=+3.46 to +8.85) isotopic compositions that fall within the range of the Palaeocene–Eocene Gangdese Batholith and the coeval Linzizong volcanics and plot on a mixing line between Tethyan basalts and Indian Ocean pelagic sediments. These features, coupled with the variable Sr/Nd, Sr/Th, Ba/Th, and Rb/Y ratios of the intrusions, are indicative of a common source involving basaltic lower crustal melts underplated during the Palaeocene and Eocene. The basaltic material was sourced from a mantle wedge that was metasomatized by Neo-Tethyan slab-derived fluids and oceanic sediments. The MMP, however, formed under a higher magmatic oxygen fugacity than did the EG, as indicated by their higher zircon Ce4+/Ce3+ ratios (87-1112) and higher fO2 values (ΔFMQ=5), determined using ilmenite–magnetite mineral pairs. The decompression and discharge of SO2 during the Linzizong volcanic event and the fractionation of magnetite could account for the low oxygen fugacity (Ce4+/Ce3+=58–164; ΔFMQ=−0.3) of the EG as well the limited development of porphyry deposits associated with the Gangdese Batholith.The MMP and Miocene mineralization-related intrusions within other Gangdese porphyry Cu deposits have high Sr/Y and V/Sc ratios and a negative correlation between Dy/Yb and SiO2, all of which are indicative of the crystallization of hornblende from a hydrous mafic melt. The EG and Gangdese Batholith, however, have low Sr/Y and V/Sc ratios and decreasing Sr concentrations over a wide range of SiO2 contents, indicating that these intrusions formed from dry melts that had undergone significant plagioclase fractionation. Late-crystallized hornblende is generally present as small crystals within early formed feldspar phenocrysts in the EG, which is in good agreement with a rapid increase followed by a decrease in Y and Dy concentrations at ~60wt.% SiO2. In addition, apatite in MMP has F (2.66–3.72%) and Cl (0.06–0.53%) volatile contents that are higher than those of apatite from EG (0.85–1.50% F and 0.02–0.03% Cl). These geochemical and mineralogical characteristics suggest that EG formed from a relatively dry magma with <4.0% H2O, whereas MMP formed from a more hydrous magma with >5.5% H2O. It is here proposed that long-lived subduction caused the high oxygen fugacity conditions of the arc magma by adding oxidized components to the sub-arc mantle in addition to F, Cl, S, and Cu. Long-term metamorphic dehydration resulted from the India–Asia collision, and subsequent crust thickening and shortening was responsible for the high water contents by continuously replenishing the basaltic melts at the base of the lower crust. Both subduction-related metasomatism and late-collision-related metamorphic dehydration controlled the genesis of fertile magma that formed the post-collision porphyry Cu–Mo deposits of southern Tibet.