Decoding the oxygen fugacity of ore-forming fluids from garnet chemistry, the Longgen skarn Pb-Zn deposit, Tibet

The relationship between the compositional zonation of garnet in skarn deposits and the evolution of physi-cochemical conditions of ore-forming fluids is not clear and needs further investigation. This study presents the characteristics of trace elements of garnet from the Longgen skarn Pb-Zn deposit, Tibet with an aim to reveal the evolution trend of the physical and chemical conditions (fO2 and pH) of ore-forming fluids and to explore the fingerprint of garnet trace elements in different types of skarn deposits. Based on robust alteration halo, mineral assemblages and petrographic observation, the garnets of the Longgen skarn Pb-Zn deposit are divided into two types: (1) distal exoskarn garnet near limestone with ore bearing and fine oscillatory zoning (Adr(39.9-75.0); Grt1, Adr represents andradite) and (2) proximal exoskarn garnet (core: Adr(32.9-70.7); rim: Adr(83.3-99.5)) with Pb-Zn-rich mineral and coarse oscillatory zoning (Grt2). According to the total rare earth element (Sigma REE) content in the garnet compositional zone, Grt1 is further divided into three component zones: Grt1-1 (heavy rare earth element (HREE)-enriched chondrite-normalized REE distribution patterns and positive Eu-anomaly (delta Eu > 1, delta Eu = 2*(Eu-sample/Eu-chondrite)/(Sm-sample/Sm-chondrite+ Gd-sample/Gd-chondrite)), Grt1-2 (HREE-enriched REE patterns with negative Eu-anomaly (delta Eu < 1)), and Grt1-3 (light rare earth element (LREE)-enriched REE patterns with delta Eu > 1). Similarly, five component zones were identified in Grt2: Grt2-1 (REE-rich REE patterns with delta Eu < 1), Grt2-2 (LREE-depleted REE patterns with delta Eu < 1), Grt2-3 (steady REE patterns with dEu < 1), Grt2-4 (LREE-enriched REE patterns with delta Eu > 1), and Grt2-5 (LREE-enriched REE patterns with delta Eu < 1) from the core to rim, respectively. The fractionation of LREE with HREE and the behavior of Eu indicate that the physical-chemical conditions during the formation of Grt1-1-Grt1-2 and Grt2-1-Grt2-3 were near neutral (pH = 6-7), and changed to mildly acidic in Grt1-3 and Grt2-4-Grt2-5 (pH < 6-7). In addition, the characteristics of low U (< 1 ppm) and the linear relationship between Sn and Eu indicate that the fO2 of the ore-forming fluid was relatively oxidized in comparison to the garnets found in other skarn deposits, and Grt1-1-Grt1-3 and Grt2-1-Grt2-5 have a gradual increasing oxidative ability. The linear relationship between TREE and Y/Ca indicates that Grt1 and Grt2 grew in a near-closed system. The incorporation of REE3+ into the garnet was mainly controlled by a "yttrogarnet" (YAG)-type substitution, Ca site vacancy, fluid chemistry, and physicochemical conditions of precipitation. However, the compositional zone at the transition zone between Al-rich and Fe-rich segments in Grt2-4 may be formed in an open system, as suggested by the exchange of LREE and Cl-. In conclusion, garnet crystal growth was affected by the crystal chemistry and physicochemical conditions of the skarn deposit. This study demonstrates that factor analysis and discrimination diagrams (log Sn vs. log U, log Ce vs. log U, and log (Ce + Hf) vs. log U) are effective methods for distinguishing different types of skarn deposits by their contents of redoxsensitive trace elements in garnet. [GRAPHICS] .