Petrogenesis And Metallogenetic Implications Of Aadakites In The Gangdese Porphyry Copper Belt

The Gangdese is a large porphyry Cu(Mo) ore belt in China, with major progresses in exploration in recent years. Most notable is the Miocene porphyry copper belt distributed close to the eastern Himalayan syntaxis in the eastern segment of Gangdese. In this thesis, I studied 4 large-super large deposits in the Tibetan Plateau, including the Qulong, Jiama, Chongjiang and Bangpu deposits. The orebodies of these deposits are hosted in the Miocene intrusive complex including granodiorite porphyry, monzogranite porphyry, and granite porphyry. In addition to the porphyry Cu(Mo) deposits, there are also a lot of other deposits, forming a Oligocene polymetallic W-Mo-Cu-Au belt on the southern margin of Gangdese belt, immediately north of the Yarlung Tsangpo Suture zone, represented by Nuri Skarn Cu-W-Mo deposit and Mingze Porphyry-Skarn Mo-Cu deposit. These deposits are generally characterized by adakitic magmas, including acidic, and intermediate-acidic intermediate-basic igneous rocks. Previous studies found close associations between adakites and these porphyry copper deposits. It is, however, still controversial as regards to which type of adakite was responsible to the porphyry mineralization. Where do the ore-forming adakites occur in the deposits? These questions are of critical importance for revealing the metallogenic mechanism and for prospecting exploration targets in depths. The nature of magma sources for the Miocene adakitic intrusions and the main ore-controlling factors for these porphyry Cu mineralization have been debated. One of the important reasons is that “ore-bearing porphyry” was usually taken as “ore-forming porphyry”. So, it is good for deepening our understanding of porphyry deposit to identify the real “ore-forming porphyries”.In this thesis, I selected the typical porphyry-skarn Cu±Mo deposits, including Qulong, Chongjiang, Bangpu, Jiama and Mingze, as the research objects and carried out comprehensive geochemical studies, e.g., whole rock major and trace elements, Sr-Nd isotopes, zircon U-Pb dating, Hf-O isotopic compositions and apatite major and trace elements compositions of these deposits. Combining with the latest development, this study focuses on the following issues:(1) the nature of magma sources for the Miocene adakitic intrusions closely related to the porphyry-skarn deposits;(2) the discrimination between ore-bearing adakites and ore-forming adakites;(3) the association of these porphyry copper deposits with the subduction of the Indian Ocean Ninetyeast Ridge.Qulong porphyry Cu deposit is the largest Cu deposit in China so far discovered, with total reserves of 10.6 Mt Cu@0.5% and 0.5 Mt Mo@0.03%. The Qulong porphyries have geochemical characteristics of typical adakites, with Sr=259–1195 ppm, Y=1.91–9.12 ppm, Yb=0.2–0.92 ppm, Sr/Y=49–202. All the Qulong adakites studied here have low Mg O(< 2 wt.%), high K2O(2-6 wt.%), with Si O2 contents mostly higher than 64 wt.%. These are dramatically different from ore-forming adakites in the circum-Pacific region and other places in general. The ore-bearing and barren adakites share similar incompatible element and rare earth element patterns, enrichment in incompatible elements, e.g., large ion lithophile elements(LILE), such as Rb, Th, U and Pb, light rare earth elements(LREE), and depleted in high field strength elements(HFSE), e.g., Nb, Ta and Ti and heavy rare earth elements(HREE). A majority of samples show weak or no-Eu anomalies, reflecting the presence of residual garnet and the absence of plagioclase in the source region. Most of the samples belong to arc-adakites(slab melts) in a Sr/Y-La/Yb discrimination, but plotted close to Dabie adakites(continental crust melts). Several ore-bearing samples even plot in the field of Dabie adakites. These characteristics support the model of partial melting of both subducted young oceanic slab and the lower continental crust, with the former as the main contributor. The Sr-Nd isotopes of adakites show small variations, with(87Sr/86Sr)i of about 0.705 and εNd(t) varying from 0 to 0.6. Zircons from these two types of adakites have similar δ18O(ranging from 5.1 to 7.3‰ and 4.0 to 7.4‰, respectively) and εHf(t)(ranging from 1.9 to 10.4 and 5.6 to 9.3 respectively), mostly plotting close to MORB. In-situ zircon Hf-O isotopic measurements for the most samples yield a binary mixing trend between mantle- and crustal-derived melts, with mantle-related sources as the main contributor. Magmatic zircons from these two series of Qulong intrusions have U-Pb ages of 16.6±0.5 Ma–17.0±0.6 Ma and 16.7±0.3 Ma–17.4±0.4 Ma, respectively, which are identical to each other within analytical errors but are systematically older than although marginally overlap with the Re-Os isochron ages of 15.36±0.21–16.41±0.48 Ma. The similarities of CeⅣ/Ⅲ(871–1457 and 910–1220 respectively) and δEu in zircons from ore-bearing and barren adakites indicate high oxygen fugacity in their magma sources, having geochemical characteristics of typical subduction genesis. Importantly, no sharp contact relationship between the ore-bearing and barren adakites was found. However, barren adakites have very low Cu contents and are not responsible to the material source. Based on the above evidences, neither the ore-bearing nor the barren adakites so far sampled from Qulong porphyry deposit are responsible to the Cu mineralization. Combining with the recent researches, the monzogranite porphyry have age of 15.3±0.6 Ma, identical to the ore-forming age, in the central phase may be the most likely ‘ore-forming adakites’. The similarities of CeⅣ/Ⅲ(871–1457 and 910–1220 respectively) and δEu in zircons from ore-bearing and barren adakites indicate high oxygen fugacity in their magma sources, having geochemical characteristics of typical subduction genesis. Ore-bearing adakites are poor in sulfur and rich in gold, with abundant magnetite-hematite and anhydrite. Such phenomena are seen mostly in porphyry copper deposits related to plate subduction in active arc environments, but not in post collisional settings. Therefore, Qulong porphyry cannot be explained reasonably through part melting of either thickened lower continental crust / newly underplated crust or relict of previously subducted Jurassic oceanic crust dehydrated long time ago. Apatites from Qulong adakites have similar F and Cl concentrations, indicating a settings different from Yulong and Dexing porphyry copper deposits. Despite the high similarities of geochemical characteristics, apatites from the two types of adakites are different from each other. The REE fractionation of apatites from ore-bearing adakites is weaker than that of barren adakites. In the Th/U-Ce/Pb、Sr/Ce-Sr/Eu、Sr-Y、Sr-Yb and Sr-(La/Yb)n diagrams, apatites from barren adakites overlap with that from mafic microgranular enclaves(MMEs), which are obviously different from that of from ore-bearing adakites, indicating that apatites from ore-bearing adakites suffered from metallogenic hydrothermal, resulting in elements exchange. Apatite may be a potential indicator mineral for distinguishing ore-bearing(ore-forming) from barren intrusions.Chongjiang deposit is a large porphyry Cu deposit in the Gangdese belt, with total reserves of 5 Mt Cu@0.4%. Chongjiang adakites also have low Mg O(< 2 wt.%), high K2O(ranging from 3.4 wt.% to 6.6 wt.%). Similarly to Qulong deposit, most of the samples belong to subduction-related adakites, but plot near the field of thickened LCC-derived adakites. A few ore-bearing samples even plot in the field of thickened LCC-derived adakites, indicating partial melting of both subducted slab and the lower continental crust. Chongjiang ore-bearing and barren adakites have U-Pb ages of 14.9±0.3 Ma-14.8±0.3 Ma and 12.9 Ma and the later represents the emplacement age of the intrusion after mineralization. Ore-bearing and barren porphyries have consistent Sr-Nd isotopes compositions, with(87Sr/86Sr)i ranging from 0.706 to 0.707 and εNd(t) ranging from-3.8 to-2.6 respectively, which are different from Qulong deposit, indicating a slightly different magma source, i.e., Chongjiang having more crust material than that of Qulong deposit. Zircons from the Chongjiang ore-bearing adakites have δ18O ranging from 5.0 to 7.2‰(average 6.2‰)and εHf(t) from(-1.0)to 7.6, which also plot close to the value of MORB. In-situ zircon Hf-O isotopic measurements for the most samples yield a binary mixing trend between MORB- and crustal-derived melts, with MORB sources as the main contributor. F and Cl concentrations of apatites from the ore-bearing porphyries indicate a tectonic environment different from Dexing and Yulong porphyry copper deposits.Bangpu deposit is a large Miocene Mo(Cu) deposit in Gangdese belt, with total reserves of 0.45 Mt Mo and 0.92 Mt Cu. Magmatic zircons from these ore-bearing intrusions have U-Pb ages of 13.9±0.3-14.0±0.2 Ma, which are within the range of other Miocene porphyry deposits in Gangdese belt. Zircons from these ore-bearing intrusions have δ18O ranging from 4.72-7.22‰ and 4.0 to 7.4‰ and εHf(t) ranging from-2.3 to 5.6, respectively, mostly between depleted mantle- and continental crustal-derived melts, which imply significant involvement of continental crust components. Zircon εHf(t) from Bangpu ore-bearing intrusions is lower than that of from Qulong deposit, indicating more continental crust components than that of Qulong deposit, which may be the reason for losing its own adakitic characteristics.Jiama skarn-type polymetallic deposit is a super-large ore deposit with proven reserves of 4.5 Mt Cu, 0.45 Mt Mo and 100 t Au. The samples from Talongwei, Xiangbeishan and hidden intrusions studied have Si O2 ranging from 60.20 to 70.77 wt.% and relatively high K2 O contents(4.47–5.31 wt.%). Based on their chemical compositions, they mainly plot in the high-K calc-alkaline field in a K2 O versus Si O2 diagram. Talongwei and Xiangbeishan intrusions have U-Pb ages of 16.0±0.2–14.9±0.3 Ma and 15.9-16.0 Ma respectively; However, the hidden intrusion have a slightly younger age of 15.0±0.2 Ma. Jima intrusions have(87Sr/86Sr)i ranging from 0.705-0.707 and εNd(t) ranging from(-4.5) –(-0.4) respectively, which are different from Qulong deposit, indicating more crust material involved than that of Qulong deposit. Zircons from hidden intrusion have δ18O ranging from 4.88–7.15‰ and εHf(t) ranging from 0.8 to 7.2 respectively, mostly between depleted mantle- and continental crustal-derived melts, with the depleted mantle-related sources as the main contributor.Mingze deposit is a newly explored porphyry Mo-Cu deposit belong to the Oligocene polymetallic W-Mo-Cu-Au belt on the southern margin of Gangdese belt. Mingze Oligocene porphyry is typical monzonite granite, belonging to high-K calc-alkaline-shoshonitic series. According to the tectonic settings and dating results, Mingze deposit is not in the Gangdese Miocene porphyry copper deposit belt. U-Pb age of zircon from the intrusion is from 28.1±0.4 Ma to 28.2±0.5 Ma. The whole rock results of trace element analysis show that they are rich in LILE and LREE e.g. Rb, Th, U and Pb, depleted in HFSE, e.g. Nb, Ta and Ti, and show weak or no-Eu anomalies, reflecting a typical characteristic of subduction-related magmas. All the samples have high Sr and low Y concentration and displaying typical geochemical affinities of adakites. The Sr/Y-(La/Yb)n discrimination indicates that the formation of these adakites is mainly related to the partial melting of thickened lower continental crust, while only a few samples belong to subduction-related adakites. This feature is in well agreement with its features of Mo-rich and Cu-poor. These characteristics make them quite different from adakites from Gangdese porphyry copper deposit with ages less than 23 Ma, suggesting the change of geochemical characteristics and tectonic mechanism. This change is probably caused by the subduction of the Indian Ocean Ninetyeast Ridge. The age of granodiorite from Chengba ore section is from 91.1±1.0 Ma to 90.1±0.7 Ma. The whole rock results of trace element analysis show that they are rich in LILE and depleted in HFSE, but the contents of Rb, Th, U, Pb, Nb and Ta are lower than that of Oligocene monzonite granites. They have smaller REE fractionation than that of Oligocene monzonite granites, have high Sr and low Y concentration and show weak Eu anomalies, reflecting a typical characteristic of subduction-related adakites, different from thickened LCC-derived adakites. The above characteristics show that the adakites in the eastern Gangdese could be formed through partial of slab subduction during Cretaceous, thickened lower continental crust during Oligocene and slab(ridge) subduction.In the whole Miocene Gangdese porphyry ore deposit belt, most of the large porphyry Cu(Mo) deposits formed mainly in the Miocene(12–20 Ma). These ages are varied in different places, showing an eastward older trend in the belt between 87°E and 93°E. The Gangdese Miocene porphyry copper deposits have close association with adakites. The Sr/Y-(La/Yb)n, Sr-Nd isotopes, zircon Hf-O isotopes and high CeⅣ/Ⅲ value demonstrate that the Miocene adakites closely related to the porphyry-skarn deposits are the melting of subducted slab, with significant involvement of continental crust components. The combined characteristics of the distributions of porphyry Cu deposits, chemical characteristics, high oxygen fugacity of these adakites, and change of tectonic mechanism around 23 Ma can best be explained by the subduction of the Indian Ocean Ninetyeast Ridge.