Geochronology and Geochemistry of the Nantianwan Mafic–Ultramafic

This article was downloaded by: [China University of Geosciences], [Mr Zhaochong Zhang] On: 15 October 2012, At: 20:59 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

International Geology Review Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tigr20 Geochronology and geochemistry of the Nantianwan mafic–ultramafic complex, Emeishan large igneous province: metallogenesis of magmatic Ni–Cu sulphide deposits and geodynamic setting Meng Wang a , Zhaochong Zhang a , John Encarnacion b , Tong Hou a & Wenjuan Luo a a State Key Laboratory of Geological Processes and Mineral Resources, China University of Geosciences, Beijing, 100083, China b Department of Earth and Atmospheric Sciences, Saint Louis University, 3642 Lindell Boulevard, St. Louis, MO, 63108, USA Version of record first published: 08 Mar 2012.

To cite this article: Meng Wang, Zhaochong Zhang, John Encarnacion, Tong Hou & Wenjuan Luo (2012): Geochronology and geochemistry of the Nantianwan mafic–ultramafic complex, Emeishan large igneous province: metallogenesis of magmatic Ni–Cu sulphide deposits and geodynamic setting, International Geology Review, 54:15, 1746-1764 To link to this article: http://dx.doi.org/10.1080/00206814.2012.668766

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material. International Geology Review Vol. 54, No. 15, November 2012, 1746–1764

Geochronology and geochemistry of the Nantianwan maﬁc–ultramaﬁc complex, Emeishan large igneous province: metallogenesis of magmatic Ni–Cu sulphide deposits and geodynamic setting Meng Wanga , Zhaochong Zhanga*, John Encarnacionb , Tong Houa and Wenjuan Luoa aState Key Laboratory of Geological Processes and Mineral Resources, China University of Geosciences, Beijing 100083, China; bDepartment of Earth and Atmospheric Sciences, Saint Louis University, 3642 Lindell Boulevard, St. Louis, MO 63108, USA (Accepted 17 February 2012)

The Nantianwan mafic–ultramafic complex is situated in the northwest part of the Panxi district, southwest China. It consists predominantly of gabbros, gabbronorites, and lherzolites. LA–ICP–MS U–Pb zircon dating of the gabbronorites yields an age of 259.7 ± 0.6 million years, consistent with the ages of other mafic–ultramafic intrusions in the Emeishan large igneous province (ELIP). Gabbronorites and lherzolites host Cu–Ni sulphide ores. Cumulus texture is pronounced in these rocks, containing magnesium-rich olivine (up to 81.4% forsterite). SiO2 contents of the lherzolites range from 42.93 to 44.18 wt.%, whereas those of the gabbronorites vary between 44.89 and 52.76 wt.%. Analysed samples have low rare earth element (REE) contents (23.22–30.16 ppm for lherzolites and 25.21–61.05 ppm for gabbronorites). Both lherzolites and gabbronorites have similar chondrite-normalized REE patterns, suggesting that they are comagmatic. All samples are slightly enriched in large ion lithophile elements (LILEs, e.g. Rb, Ba, and Sr) relative to high field strength elements (HFSEs, e.g. Nb, Ta, and Ti), very similar to those of ocean island basalts (OIBs). The presence of cumulus textures and geochemical signatures indicates that 87 86 fractional crystallization played an important role in the petrogenesis of these rocks. Initial ( Sr/ Sr)t (t = 260 Ma) ratios and εNd(t) values of the mafic–ultramafic suite vary from 0.70542 to 0.70763, and −0.4 to 1.7, respectively. Compared to the Cu–Ni-bearing Baimazhai and Limahe intrusions in the ELIP, which were considerably contaminated by variable crustal 87 86 materials, the Nantianwan complex exhibits much lower ( Sr/ Sr)t.TheirεNd(t) versus (Th/Nb)PM ratios also indicate that the ore-bearing magmas did not undergo significant crustal contamination. In combination with (Tb/Yb)PM versus (Yb/Sm)PM modelling, we infer that the magmas originated from an incompatible elements-enriched spinel-facies lherzolite that itself formed by interaction between the Emeishan plume and the lithospheric mantle. Most plots of NiO versus Fo contents of olivine suggest that sulphides are separated from the parental magma by liquid immiscibility, which is also supported by bulk-rock Cu/Zr ratios of the lherzolites (7.04–102.67) and gabbronorites (0.88–5.56). We suggest that the gabbronorites and lherzolites experienced undersaturation to oversaturation of sulphur; the latter may be due to fractional crystallization in a high-level magma chamber, accounting for the sulphide segregation. Keywords: Nantianwan complex; geochemistry; metallogenesis; Cu–Ni sulphide; Panxi region

Introduction (platinum group element) sulphide deposits (e.g. Hou et al. The Emeishan igneous complex is a unique large igneous 2011). Particularly, it is well known for the presence of province (LIP) in China, widely recognized by the inter- the world’s largest Fe–Ti–V oxide ore cluster in the Panxi national geoscientiﬁc community. It is one of the three region (e.g. Zhang et al. 2009), where the scale of Cu–Ni– LIPs on Earth that formed near the end of the Permian in (PGE) mineralization is relatively small. In contrast, the widely separated locations, the others being the Siberian Siberian LIP is characterized by the world-class Noril’sk Traps (Sharma 1997; Dobretsov 2005) and the Panjal Cu–Ni–(PGE) sulphide deposits. The difference in miner- Traps of northwestern India (e.g. Bhat et al. 1981). The alization between these two Permian LIPs raises an impor- Emeishan large igneous province (ELIP) is thought to be tant question: what caused the different features of the ∼

Downloaded by [China University of Geosciences], [Mr Zhaochong Zhang] at 20:59 15 October 2012 genetically related to a major 260 Ma plume event (e.g. ore deposits: geologic settings, source composition, depth Thompson et al. 2001; Xu et al. 2004; Ali et al. 2010). of melting, lithospheric/crustal contamination, or magma Although the Permian ELIP is not as large as other LIPs chamber processes? (e.g. Zhang et al. 2006), it is one of the richest in min- Recently, small-scale Cu–Ni–(PGE) sulphide miner- eral resources and the only province in the world that hosts alization has been recognized in the Nantianwan maﬁc– both magmatic Fe–Ti–V oxide ores and Cu–Ni–(PGE) ultramaﬁc complex in the Pingchuan area, Yanyuan

*Corresponding author. Email: [email protected]

county of Sichuan province, in the western part of the of clastic, carbonate, and metavolcanic rocks (SBGMR ELIP. Current exploration drilling is in progress by the 1991). The early Sinian consists of clastic rocks and 604 Geological Team, Geological and Mineral Resources felsic volcanic rocks, while the late Sinian Dengying Bureau, Sichuan, China. Sparsely disseminated Cu–Ni Formations consist of clastic rocks in the lower part and sulphide ores with a thickness of ∼350 m were recog- phosphorous-bearing carbonate rocks in the upper part. nized in one borehole. Unlike other Cu–Ni–(PGE) sulphide The Early Cambrian strata are characterized by clastic and deposits in the ELIP, such as the Limahe, Zhubu, and carbonate rocks, whereas the Middle–Late Cambrian strata Baimazhai, which are hosted by small sills, the volume are characterized by limestones (Zhulinping Formation) of the Nantianwan mafic–ultramafic intrusion is relatively and dolomitic limestones (Gaojiaping Formation). The large (39 km2). The prospect for large-scale Cu–Ni–(PGE) Early Silurian strata mainly consist of argillaceous rocks sulphide mineralization resembling those large Cu–Ni– and sandstones (Zhongcao Formation). The Ordovician (PGE) sulphide deposits such as Noril’sk and Sudbury strata consist of carbonate (Hongshiya Formation) and (Naldrett 1997, 1999) has aroused extensive interest. argillaceous rocks, whereas the Carboniferous strata con- However, except for a limited geological survey conducted sist of silicalites (Daopingzi Formation) and limestones at the scale of 1:50,000 and 1:200,000 in the Pingchuan (Maping Formation). The Permian Emeishan basaltic suc- area, no other detailed information has been documented cession unconformably overlies the limestones of the Early to date. Permian Maokou Formation (YBGMR 1990). In the Panxi In this article, we present the first geochronology, min- region, extensive erosion and thinning of the Middle eral chemical, bulk-rock major + trace element, and Sr–Nd Permian limestone in the central part of the ELIP indi- isotopic compositions, aimed at constraining the nature of cate kilometre-scale regional uplift linked to the rising the sources of the intrusions and the petrogenesis, which plume (He et al. 2003; Xu et al. 2004). Numerous in turn provide some key constraints on the metalloge- mafic–ultramafic intrusions are exposed along several nesis of Cu–Ni sulphide mineralization. These new data N–S-trending faults. These intrusions host major world- not only shed new light on the petrogenesis of the ELIP class Fe–Ti oxide deposits (Zhou et al. 2005, 2008). that we can apply to other provinces, but may also allow refinement of previously proposed exploration models for the same type of deposits in the region and around the Geology of the Nantianwan complex and associated world. Cu–Ni ores The Nantianwan complex, situated at the northwestern margin of the Panxi region, approximately 35 km to the Regional geological setting southwest of Xichang City, is bounded by the Jinhe-Jinghe The ELIP is exposed over a large part of southwest fault. The WNW-trending complex has a large surface China, including Yunnan, Guizhou, Sichuan, and Guangxi exposure of 39 km2 that is 9 km in length and 3–5 km provinces, and northern Vietnam, from the eastern mar- in width (Figure 2). There is a fault to the south of the gin of the Tibetan Plateau to the western margin of complex running in a NW–SE direction. The complex the Yangtze Craton (Xu et al. 2004). It forms a mas- intruded argillaceous rocks and sandstones of the Early sive Permian–Triassic succession of volcanic rocks along Silurian Zhongcao Formation. In the study area, Cambrian the western margin of the Yangtze Craton (Chung and to Triassic strata are exposed, such as the Cambrian Jahn 1995; Xu et al. 2001, 2004; He et al. 2003). The Zhulinping Formation, Gaojiaping Formation, Ordovician Emeishan volcanic succession comprising predominantly Hongshiya Formation, Carboniferous Maping Formation, basaltic flows and pyroclastics, with minor amounts of Daopingzi Formation, and Triassic strata. picrite and trachyte/rhyolite, is a several hundred metres to The Nantianwan mafic–ultramafic complex consists 5 km-thick bimodal suite (SBGMR 1991; Chung and Jahn of gabbros (∼60 vol.%), gabbronorites (∼25 vol.%), 1995), exposed in an area of ∼2.5 × 105 km2, and is asso- and lherzolites (∼15 vol.%). The gabbros are intruded Downloaded by [China University of Geosciences], [Mr Zhaochong Zhang] at 20:59 15 October 2012 ciated with numerous ultramafic–mafic to felsic alkaline by the gabbronorites in the central part of the com- intrusions (Figure 1). plex, and several small lherzolites locally intrude the The basement of the Yangtze Craton locally comprises gabbronorites, with a total outcrop area of ∼0.5 km2. the Archaean-Palaeoproterozoic Kangding Complex, com- Thus, the intrusive sequence for the Nantianwan posed of granulite–amphibolite facies metamorphic rocks, mafic–ultramafic complex is inferred to be as follows: and the Mesoproterozoic Huili Group or its equivalents, the gabbro→gabbronorite→lherzolite. Yanbian Group, which consists of metasedimentary rocks Ni–Cu sulphide ores are only hosted by the lherzo- interbedded with felsic and mafic metavolcanic rocks. lites and gabbronorites (Figures 3A and 3B). Since the The basement is overlain by a thick sequence (>9km) Nantianwan complex associated with Cu–Ni mineraliza- of Sinian (610–850 Ma) to Permian strata composed tion is still under exploration, the ore reserves and grade 1748 M. Wang et al.

Figure 1. Geological sketch map of Cu–Ni–(PGE)-bearing mafic–ultramafic intrusions in western Sichuan Province (modified from Liu et al. 2008a).

are not yet established. However, based on petrographic recognized in olivine as mantle xenocrysts or in xeno- observation, the mineralized rocks contain ∼2.5 modal% liths, such as strained, kink-banded crystals, have not pentlandite, ∼2 modal% chalcopyrite, and ∼1 modal% been observed. The gabbronorite predominantly consists

Downloaded by [China University of Geosciences], [Mr Zhaochong Zhang] at 20:59 15 October 2012 pyrrhotite. These minerals are sparsely to densely dissemi- of plagioclase (Pl; 35–45 vol.%), Cpx (35 vol.%), Opx nated throughout the gabbronorites. (10 vol.%), and minor olivine and biotite (Bi). Poikilitic The lherzolites are mainly composed of variable sizes texture (Figure 3D) and diabasic texture (Figure 3E) (0.3 up to 2.5 mm in length) of cumulus olivine crys- are ubiquitous in these rocks. Petrographic observa- tals (∼45 vol.%; Figure 3C). Some of the clinopyroxenes tions suggest that the initial crystallization sequence (Cpx, ∼25 vol.%) and orthopyroxenes (Opx, ∼15 vol.%) of the minerals is Ol→sulphide→Cpx→Opx→Pl→Bi are partly replaced by tremolite and chlorite. Minor (Figure 3F). anhedral plagioclases are observed as interstitial phases. The gabbro is dark grey to grey black. Gabbros have a Chromite usually occurs as an accessory mineral. Most diabasic or granular texture and consist mainly of clinopy- cumulus olivines are euhedral and partly altered into roxene (∼50%) and plagioclase (25–30%) with minor serpentine along fractures. Characteristics that are only hornblende and biotite, as well as accessory Fe–Ti oxide. International Geology Review 1749

101.86°E 101.96°E 27.67°N

Triassic sediments

Permian Emeishan basalts

Limestones of Maping Formation

Silicolites of Daopingzi Formation

Mudstones of Zhongcao Formation

Carbonates of Hongshiya Formation

Dolomites of Gaojiaping Formation

Sediments of Zhulinping Formation

Dolomites of Dengying Formation

Gabbros

Gabbronorites Picritic porphyries

Lherzolites Fault 27.47°N

Figure 2. Geological map of Nantianwan complex (modiﬁed from the 1:50,000 geological map of Pingchuan area, 1990).

Clinopyroxenes are usually euhedral to subhedral and Sr–Nd isotope analyses. Details of the preparation and partly replaced by actinolite and tremolite. Plagioclase is analytical techniques follow. subhedral to anhedral and usually partly altered by sericite and saussurite. In other Cu–Ni–(PGE) sulphide deposits, the ores Zircon U–Pb dating always occur in the lower parts of the mafic–ultramafic Zircons were separated from the sample DY (gabbronorite) intrusions or contact zone between intrusions and country using heavy liquid and magnetic techniques and then puri- rocks such as in the Panxi area (e.g. Luo 1981; Liang et al. fied by handpicking under a binocular microscope. Zircon 1998; Zhang et al. 1998; Zhong et al. 2002), where the grains were picked and mounted on adhesive tape then Baimazhai and Limahe intrusions intruded the Ordovician enclosed in epoxy resin and polished to about half of and the Proterozoic Huili Group (Wang et al. 2006; Zhang their diameter. In order to observe textures of the pol- et al. 2009), respectively. The Nantianwan complex, in con- ished zircons, cathodoluminescence (CL) imaging was trast, intruded the Early Silurian sequence, and the ores carried out using a Hitachi S3000-N scanning electron appear to be quite sparsely disseminated through most of microscope (SEM; Hitachi, Tokyo, Japan) with a Mono Downloaded by [China University of Geosciences], [Mr Zhaochong Zhang] at 20:59 15 October 2012 the intrusion. CL3 cathodoluminescence system for high-resolution imaging and spectroscopy at the Institute of Geology, Chinese Academy of Geological Sciences, Beijing, China. Sampling and analytical method Zircon U–Pb isotopic analyses were performed on Ten samples were systematically collected from one drilled a Finnigan Neptune multi-collector ICP–MS with a borehole. The samples were initially checked for weather- Newwave UP213 laser-ablation system at the Institute ing and traces of alteration, which were removed before the of Mineral Resources, Chinese Academy of Geological rocks were reduced to chips. One sample of gabbronorites Sciences. Helium was used as the carrier gas to enhance the from the Nantianwan intrusion was chosen petrograph- transport efficiency of the ablated material. The analyses ically for zircon U–Pb isotopic dating, nine for major were conducted with a beam diameter of 25 µm with element and trace element analyses, and six for whole-rock a 10 Hz repetition rate and a laser power of 2.5 J/cm2 1750 M. Wang et al.

Figure 3. Handspecimen photographs and microphotographs (cross-polarized light) of samples from the Nantianwan complex. (A) Drilling samples, scarce nickel–copper sulphide; (B) disseminated structure in gabbronorite; (C) cumulate texture and mineral parage- nesis in lherzolite; (D) feldspar-bearing poikilitic texture in gabbronorite; (E) diabasic texture in gabbronorite; (F) opaque minerals ﬁlled in the ﬁssures in olivine, and surrounded by orthopyroxene, in gabbronorite. Abbreviations: Opx, orthopyroxene; Cpx, clinopyroxene; Ol, olivine; Pl, plagioclase; Py, pyrite; Po, pyrrhotine; Chl, chalcopyrite.

(Hou et al. 2009). The masses 206Pb, 207Pb, 204(Pb + Hg), were deleted. The zircon Plesovice was analysed as an and 202Hg were measured by multi-ion-counters, while the unknown and yielded a weighted mean 206Pb/238U age masses 208Pb, 232Th, 235U, and 238U were collected using of 337 ± 2 million years (2SD, n = 12), which is in 206 /238 Downloaded by [China University of Geosciences], [Mr Zhaochong Zhang] at 20:59 15 October 2012 a Faraday cup. Zircon GJ1 was used as the standard and good agreement with the recommended Pb U age of zircon Plesovice was used to optimize the machine. U, 337.13 ± 0.37 million years (2SD) (Sláma et al. 2008). The Th, and Pb concentrations were calibrated using 29Si as age calculation and plotting of concordia diagrams were the internal standard and zircon M127 (U: 923 ppm; Th: performed using Isoplot/Ex 3.0 (Ludwig 2003; Figure 4). 439 ppm; Th/U: 0.475; Nasdala et al. 2008) as the external The results are presented in Table 1. standard. The in-house software program, ICPMSDataCal, produced by Liu et al. (2008b), was used for off-line selec- tion and integration of background and analysis signals, Electric microprobe analyses time-drift correction and quantitative calibration for the Electron microprobe analyses were determined for some trace element analyses and U-Pb dating. Correction for olivines in the wehrlites using a JEOL JXA-8230 common Pb was omitted because of the high 206Pb/204Pb Superprobe at the EMPA Laboratory of Analysis Centre of ratios (>1000). Data with abnormally high 204Pb counts Mineral and Rocks of the Institute of Mineral Resources, International Geology Review 1751

Figure 4. 207Pb/235U–206Pb/238U concordia diagram of zircons from the Nantianwan complex.

Chinese Academy of Geological Sciences. Operating con- in distilled HF+HClO4 in 15 ml Savillex Teﬂon screw- ditions were set at 15 kV at 10 nA beam current. Natural cap beakers. Precision for most elements was typically minerals and synthetic pure oxides from SPI Supplies Inc. better than 5% relative standard deviation, and the mea- (West Chester, PA, USA) were used as standards. For sured values for Zr, Hf, Nb, and Ta were within 10% of the pyroxene, the calibration standards used were hornblende certiﬁed values (Dulski, 1994). The detailed sample prepa- (for Si, Ti, Al, Fe, Ca, Mg, Na, and K), fayalite (for Mn), rations, instrument operating conditions, and calibration and Cr2O3 (for Cr). For plagioclase, the standards used procedures follow those established by Qi and Grégoire were hornblende (for Si, Ti, Al, Fe, Ca, and Mg), albite (for (2000). Two standards (granite GSR-1, basalt GSR-3) were Na), orthoclase (for K), and fayalite (for Mn). Precision is used to monitor the analytical quality. better than 1% for element oxides.

Rb–Sr and Sm–Nd isotope analyses Major and trace element analyses Rb–Sr and Sm–Nd isotopic compositions were obtained After screening under the microscope, relatively fresh using a Finnigan MAT-262 multi-collector mass spec- samples were selected and sawn into slabs and the central trometer at the Institute of Geology, Chinese Academy parts were used for whole-rock analyses. Specimens were of Geological Sciences. The samples (∼100 mg) were crushed in a steel mortar and ground in a steel mill to weighed and spiked before dissolution with mixed isotopic powders of ∼200 mesh. Major elements were acquired tracers. They were dissolved overnight using HF and through analysis of a fused glass disc using a scanning HNO3, evaporated to dryness, and then followed by wavelength dispersive X-ray ﬂuorescence spectrometer oven dissolution in fresh HF and HNO3 for 7 days at at the Key Laboratory of Orogenic Belts and Crustal 160◦C. Separation of Rb and Sr were carried out with a Downloaded by [China University of Geosciences], [Mr Zhaochong Zhang] at 20:59 15 October 2012 Evolution, Ministry of Education, School of Earth and cation-exchange column (packed with AG50Wx8). Sm and Space Sciences, Peking University, Beijing, China. The Nd were further puriﬁed using a second cation-exchange analytical uncertainties are less than 1%, estimated from column (packed with AG50Wx12). Total procedural the repeated analyses of two standards (andesite GSR- blanks were ∼200 pg for Sr and ∼50 pg for Nd. The mass 2 and basalt GSR-3). Loss on ignition was determined fractionation corrections for Sr and Nd isotopic ratios ◦ gravimetrically after heating the samples at 980 Cfor were based on 88Sr/86Sr of 8.37521 and 146Nd/144Nd 30 min. of 0.7219, respectively. During our analyses, measured Trace elements were determined by solution ICP–MS 87Sr/86Sr ratios for standard NBS987 were 0.710242 ± performed at the ICP–MS Laboratory at the National 0.000012, and measured 143Nd/144Nd ratios for the JMC Research Centre for Geoanalysis, Beijing, China. After standard were 0.511124 ± 0.000010 (in 2σ uncertainty for complete dissolution, powders (∼40 mg) were dissolved 18 analyses). 1752 M. Wang et al. σ U1 238 / Pb 206 σ U1 235 / Pb 207 σ 26.85 261.77 1.15 260.40 1.10 − Pb 1 206 / Pb 207 U 238 / Pb 206 U 238 / Pb 206 U 235 / Pb 207 U 235 / Pb 207 Pb 206 / Pb 207 Pb 206 / Pb 207 U Downloaded by [China University of Geosciences], [Mr Zhaochong Zhang] at 20:59 15 October 2012 / ) U–Th–Pb isotopic ratio Age (million years) –6 10 × Contents ( Dy-13Dy-14 1826.43Dy-15 705.42 601.29Dy-16 1326.32 2.59 415.47Dy-17 670.04 912.14Dy-18 1.45 0.05075 1.98 925.25 668.49Dy-19 208.94 0.04818 521.27Dy-20 1.36 1769.10 0.05082 0.00015 206.40 1.77 1019.55 2013.72 0.05049 0.00029 1.01 1.74 0.28823 0.00025 811.62 0.05006 2.48 0.06731 0.05135 0.28044 0.00016 0.00184 0.28853 0.00026 0.05120 0.00223 0.28580 0.01191 0.00017 0.04118 0.00197 0.28393 0.00025 0.04221 0.00164 0.30680 0.00022 0.29132 0.04116 0.00209 0.00022 0.29082 0.04104 0.01163 0.00187 0.00018 227.85 0.04113 0.00206 0.00019 0.04130 109.35 0.04114 231.55 0.00021 0.93 0.04118 0.00077 0.00023 216.74 12.96 11.11 257.16 198.23 0.00021 251.01 855.55 257.47 12.04 257.40 1.45 11.11 1.77 250.07 374.82 255.25 1.55 260.12 4.63 253.77 266.52 271.70 1.29 260.04 1.39 11.11 259.60 1.65 1.39 9.04 259.26 1.09 259.21 1.47 259.83 260.87 1.15 1.62 259.88 1.32 4.78 260.14 1.44 1.32 Table 1. LA–ICP–MS U–Pb isotope compositions of zircon in gabbronorites from the Nantianwan complex. Sample ThDy-1Dy-2 2073.52Dy-3 1083.35 UDy-4 809.74 1.91 1807.85Dy-5 568.22 2341.68 Th 885.27Dy-6 0.05280 1.43 1022.91 938.03Dy-7 2.04 568.05Dy-8 2.50 464.75 0.05199 0.00039 0.05312Dy-9 1.80 815.91 310.65 1563.44 0.05355Dy-10 0.00066 491.72 0.29811 1.50 0.05040 0.00038 899.06Dy-11 557.58 1458.07 1.66 0.00089Dy-12 1.74 0.05002 1171.54 0.00368 0.29438 465.89 929.48 0.29917 0.00032 0.04994 3905.22 1.20 674.14 1.57 0.05175 0.30256 2050.01 0.04099 0.00347 0.00040 1.74 0.00297 0.28532 1.90 0.05019 0.05145 0.00032 0.00569 0.00029 0.00047 0.04112 0.05079 0.28303 0.04088 0.00241 0.05173 0.28322 0.00025 0.00034 0.04100 0.00022 0.29254 320.43 0.00260 0.00032 0.00036 0.04109 0.00012 0.00208 0.28390 0.00030 0.29124 0.00214 0.04108 283.40 0.00024 0.28837 21.30 344.50 0.29408 0.04116 0.00228 0.00242 0.04102 353.76 0.00021 264.92 0.00269 34.26 213.04 0.00147 0.00019 16.67 0.04102 0.04105 0.00018 2.88 34.26 262.00 194.53 0.04115 265.75 0.04122 0.00024 0.00018 14.81 190.82 258.99 268.40 2.72 275.99 0.00020 2.32 18.51 0.00018 254.87 2.89 211.19 4.43 259.76 261.18 14.81 258.30 12.96 253.06 1.90 231.55 259.03 272.29 1.38 253.21 1.97 12.96 10.19 260.56 2.05 259.58 1.86 21.29 1.65 253.75 259.53 1.68 259.53 1.47 257.28 260.04 1.80 1.90 259.13 1.29 2.12 1.18 259.17 259.36 1.12 259.99 1.49 1.13 1.26 International Geology Review 1753

Results ranges from 42.93 to 44.18%, whereas the SiO2 content LA–ICP–MS U–Pb dating of the gabbronorites is relatively higher, from 44.89 to 52.76%, as expected. Except for two analyses (HC-1 and Zircons in our sample DY (gabbronorite) are generally HC-4), most of our samples have relatively low TiO con- transparent, euhedral, and prismatic (Figure 5). On the CL 2 tents (0.19–0.87 wt.%). For lherzolites, gabbronorites, and images, most zircon grains display long prismatic shapes gabbros, MgO contents decrease from 27.82 to 5.45% (more than 100 µm in length) and oscillatory zoning. (Figure 6). Additionally, the SiO ,AlO ,TiO, and Twenty analyses were obtained in total with variable U con- 2 2 3 2 K O+NaO contents increase whereas total FeO contents of 206–2050 ppm, Th contents of 465–3905 ppm, 2 2 tents remain constant with decreasing MgO contents, and high Th/U ratios (1.01–2.59), showing characteristics consistent with the fractionation of olivine and pyroxene of typical igneous zircons (Hanchar and Rundnick 1995; (Figure 6). Hoskin and Black 2000; Corfu et al. 2003; Grant et al. In general, all samples have similar primitive mantle- 2009). Except for one discordant analysis (DY-14), 19 anal- normalized trace element and chondrite-normalized rare yses show a consistent age of 259.7 ± 0.6 million years earth element (REE) patterns with chondrite-normalized (mean square weighted deviation = 0.13). This can be REE patterns similar to ocean island basalts (OIB) and interpreted as the crystallization age for the gabbronorite Lijiang picrites (Figure 7). The total REE contents of the in the Nantianwan complex. three types of rocks are relatively low (23.2–142.2 ppm) and are relatively enriched in light REE relative to heavy Mineral chemistry LREE. The samples lack significant Eu anomalies suggest- Analyses of olivine, clinopyroxene, and plagioclase in ing the lack of significant plagioclase fractionation, except lherzolite samples are listed in Table 2. The Mg-number for three lherzolite samples (δEu = 0.79–0.91; δEu = × / + [100 Mg (Mg Fe), molar] in the olivine varies 2Eun/(Smn+Gdn)) with slightly negative Eu anomalies. from 79.1 to 81.4. Olivine from lherzolite has much Most of the samples are slightly enriched in large ion higher CaO content (>0.1 wt.%) than that of typ- lithophile elements (LILEs, e.g. Rb, Ba, and Sr) relative to ical mantle xenoliths (Thompson and Gibson 2000). high field strength elements (HFSEs) with sizeable Nb–Ta The orthopyroxene analyses show small compositional and Ti troughs. ranges of Wo0.99–4.25En77.62–81.49Fs17.46–18.28 belonging to bronzite. The plagioclases show a compositional range of An58.8–80.5,Ab19.1–40.2,Or0.2–1.3, and are, therefore, Sr and Nd isotope compositions bytownites to andesines. Samples from the Nantianwan complex have age-corrected 87 86 ( Sr/ Sr)t (t = 260 Ma) values varying from 0.70542 to Bulk-rock major and trace element data 0.70763 and age-corrected εNd(260 Ma) values ranging 87 86 According to the major and trace element analyses listed from –0.4 to 1.7 (Table 4). Obviously, the ( Sr/ Sr)t of in Table 3, in general, the SiO2 content of lherzolites the Nantianwan complex is higher than those of typical Downloaded by [China University of Geosciences], [Mr Zhaochong Zhang] at 20:59 15 October 2012

Figure 5. Cathodoluminescence (CL) images of zircons of the Nantianwan complex (circles are laser points, numbers are points order). 1754 M. Wang et al.

Table 2. Electron microprobe composition of representative minerals in Nantianwan lherzolites (wt.%).

Sample no. Mineral SiO2 Na2OMgOAl2O3 CaO FeO MnO TiO2 NiO Total Fo

DPZ-2-2 Ol Ol 39.69 0.02 42.85 0.01 0.01 18.33 0.22 0.02 0.11 101.25 80.42 DPZ-2-4 Ol Ol 39.61 0.00 42.74 0.00 0.00 19.14 0.23 0.02 0.23 101.98 80.08 DPZ-2-4 Ol Ol 38.93 0.00 43.10 0.00 0.00 17.70 0.19 0.00 0.23 100.15 81.42 DPZ-2-5 Ol Ol 39.25 0.00 43.39 0.01 0.00 18.17 0.17 0.00 0.11 101.12 81.13 DPZ-2-7 Ol Ol 39.36 0.00 42.71 0.03 0.03 18.89 0.23 0.05 0.13 101.47 80.28 DPZ-2-12 Ol 2 Ol 38.68 0.00 42.86 0.01 0.01 18.78 0.27 0.04 0.20 100.88 80.43 DPZ-3-1 Ol Ol 39.37 0.07 41.33 0.02 0.00 19.46 0.38 0.04 0.12 100.89 79.27 DPZ-3-2 Ol Ol 39.14 0.01 41.40 0.00 0.00 18.60 0.26 0.01 0.13 99.55 80.03 DPZ-3-3 Ol 2 Ol 39.79 0.06 41.32 0.02 0.00 19.67 0.22 0.05 0.13 101.24 79.08 DPZ-3-3 Ol 3 Ol 38.88 0.09 41.26 0.00 0.04 19.36 0.26 0.05 0.12 100.09 79.32 DPZ-1-6 Ol 39.02 0.04 42.89 0.03 0.04 19.09 0.31 0.00 0.12 101.54 80.17 DPZ-1-9 Ol 39.05 0.01 42.89 0.00 0.00 17.99 0.28 0.00 0.10 100.32 81.10 DPZ-1-18 Ol 39.34 0.00 43.34 0.00 0.02 18.75 0.28 0.00 0.13 101.88 80.62 Wo En Fs DPZ-2-10 Opx 55.79 0.00 30.32 1.03 0.72 11.91 0.32 0.06 0.02 100.36 1.37 80.45 18.18 DPZ-2-11 Opx 56.57 0.02 30.91 0.67 0.52 11.68 0.13 0.01 0.06 100.61 0.99 81.49 17.46 DPZ-2-5 Opx 55.28 0.03 29.73 1.28 1.45 11.46 0.18 0.32 0.01 100.00 2.78 79.65 17.49 DPZ-3-5 Opx 55.29 0.00 29.53 1.50 1.62 12.11 0.15 0.30 0.05 100.93 3.11 78.61 18.28 DPZ-1-19 Opx 55.06 0.11 29.56 1.58 2.25 11.82 0.29 0.11 0.04 101.19 4.25 77.62 17.75 DPZ-1-20 Opx 53.91 0.10 29.08 1.49 1.76 11.90 0.29 0.25 0.04 99.23 3.40 78.01 18.24 DPZ-1-11 Cpx 51.84 0.42 16.38 2.95 21.72 5.59 0.14 0.41 0.01 100.15 43.68 45.84 8.96 An Ab Or DPZ-3-5 Pl 1 Pl 48.68 3.06 0.01 31.48 15.28 0.31 0.01 0.01 0.00 99.04 72.73 26.37 0.90 DPZ-3-5 Pl 4 Pl 51.22 4.32 0.00 29.31 12.80 0.18 0.00 0.03 0.00 98.14 61.29 37.45 1.27 DPZ-3-5 Pl 5 Pl 52.43 4.63 0.01 29.23 12.26 0.23 0.00 0.06 0.04 99.13 58.77 40.16 1.07 DPZ-3-5 Pl 6 Pl 48.45 2.72 0.03 31.89 15.75 0.25 0.02 0.02 0.03 99.32 75.64 23.66 0.69 DPZ-3-5 Pl 7 Pl 49.51 3.18 0.01 31.18 14.73 0.19 0.00 0.01 0.00 98.93 71.48 27.88 0.64 DPZ-1-1 Pl 47.67 2.27 0.00 33.06 16.55 0.41 0.01 0.05 0.00 100.45 79.68 19.79 0.53 DPZ-1-2 Pl 47.82 2.68 0.02 32.68 15.97 0.37 0.00 0.03 0.00 99.99 76.42 23.23 0.35 DPZ-1-3 Pl 45.70 2.27 0.47 31.19 17.35 0.38 0.01 0.04 0.01 97.80 80.51 19.07 0.42 DPZ-1-24 Pl 49.99 3.47 0.06 31.24 14.33 0.29 0.01 0.06 0.01 99.88 69.06 30.24 0.70 DPZ-1-25 Pl 47.88 2.72 0.01 32.59 15.84 0.33 0.00 0.02 0.01 99.88 75.91 23.59 0.50

+ Note: Fo = 100 × Mg/(Mg + Fe2 ).

E-MORB and OIB (Stille et al. 1983; Saunders et al. Discussion 1988). Compared with those of the Noril’sk intrusion in the The nature of the mantle source 251 Ma Siberian LIP, the Sr and Nd isotopic ratios of the Olivine with higher forsterite (Fo) contents (up to 0.81) in Nantianwan complex are relatively restricted. These val- the Nantianwan complex may have crystallized from a ues are also slightly higher than those of EMI, but much Mg-rich melt. The most Mg-rich olivine core (Fo in lower than those of EMII (Stille et al. 1983; Saunders 81 sample DPZ-2) corresponds to olivine that crystallized et al. 1988). The ε (260 Ma) values and (87Sr/86Sr) Nd t from relatively primary mantle melts. Assuming a K (Fe– ratios of the Nantianwan complex overlap the values for the D Mg)ol-liq of 0.30 ± 0.03 (Roeder and Emsile 1970), the Emeishan ﬂood basalts (Figure 8). The complex has higher 87 86 parental magma that would have been in equilibrium with Downloaded by [China University of Geosciences], [Mr Zhaochong Zhang] at 20:59 15 October 2012 ( Sr/ Sr) values than those of Siberian trap basalts. t olivine with Fo can be estimated by the formula W However, compared with the Cu–Ni-bearing Baimazhai 81 MgO = 0.56095 × K × Fo/(1–Fo) × W (Zhang and and Limahe intrusion in the ELIP (Xu et al. 2001; Hanski D FeO Wang 2003). The FeO values correspond to the FeO con- et al. 2004; Xiao et al. 2004; Wang et al. 2006), the tent of the parental magma on the assumption that 15% Nantianwan complex exhibits much lower (87Sr/86Sr) and t of the total iron oxide content is ferric. According to higher ε (t). In contrast, the Sr and Nd isotopic data Nd such estimations and the MgO content (27.82 wt.%), we of the Fe–Ti–V oxide ore-bearing intrusions in the ELIP conclude that the primary magma is probably basaltic (represented by Taihe, Baima, and Panzhihua intrusion) and contains 6.68 wt.% MgO, which is lower than the overlap with or are close to the high ε (260 Ma) end of Nd average of the ELIP ﬂood basalts (8 wt.%; Zhang and ore-bearing intrusions in the ELIP. Wang 2002). International Geology Review 1755

Table 3. Major (wt.%) and trace (ppm) elements analyses of the Nantianwan complex.

Sample DPZ-1 DPZ-2 DPZ-3 DPZ-4 DPZ-7 DPZ-9 NTG-1 NTG-3 NTG-4

Lithology Lherzolite Gabbronorite

SiO2 42.93 43.40 44.18 46.89 47.11 47.19 44.89 48.36 48.37 Al2O3 8.67 7.25 8.47 23.21 17.12 20.57 20.81 21.04 19.64 TFe2O3 10.87 10.09 10.08 4.64 5.96 7.00 6.10 4.00 6.17 CaO 7.29 7.78 8.11 11.30 12.27 9.21 10.59 11.27 9.95 MgO 27.36 27.82 26.58 9.08 13.29 9.04 11.97 10.99 9.24 K2O 0.11 0.11 0.22 0.19 0.19 0.21 0.16 0.22 0.24 Na2O 1.10 0.86 1.30 3.37 2.19 4.65 3.19 2.10 4.07 MnO 0.17 0.16 0.16 0.07 0.13 0.12 0.10 0.09 0.12 TiO2 0.26 0.23 0.28 0.28 0.35 0.69 0.60 0.19 0.59 P2O5 0.05 0.04 0.04 0.05 0.05 0.14 0.10 0.04 0.10 LOI 0.12 0.93 <0.01 0.64 0.95 0.90 1.08 1.51 1.34 Total 98.92 98.66 99.42 99.72 99.61 99.71 99.59 99.80 99.83 Mg# 0.82 0.83 0.83 0.78 0.80 0.70 0.78 0.832 0.73 La 2.11 2.09 2.81 3.53 3.31 5.45 4.07 2.45 3.44 Ce 4.89 4.68 6.28 7.55 7.43 12.60 9.29 5.26 8.12 Pr 0.68 0.64 0.85 1.00 1.03 1.74 1.38 0.70 1.19 Nd 3.25 2.81 3.69 4.47 4.73 7.90 6.54 3.09 5.57 Sm 0.98 0.83 1.12 1.22 1.43 2.32 2.03 0.89 1.80 Eu 0.29 0.29 0.35 0.48 0.53 1.00 0.74 0.41 0.77 Gd 1.28 1.14 1.46 1.59 1.94 3.27 2.80 1.27 2.60 Tb 0.22 0.19 0.25 0.26 0.30 0.50 0.42 0.19 0.40 Dy 1.43 1.27 1.58 1.62 1.96 3.12 2.70 1.29 2.57 Ho 0.30 0.27 0.33 0.32 0.41 0.64 0.56 0.26 0.56 Er 0.88 0.82 1.04 1.02 1.21 2.01 1.63 0.85 1.66 Tm 0.12 0.11 0.14 0.14 0.16 0.27 0.22 0.12 0.21 Yb 0.84 0.74 0.90 0.93 1.08 1.79 1.41 0.76 1.47 Lu 0.12 0.11 0.14 0.14 0.17 0.24 0.22 0.12 0.22 Y 7.82 7.23 9.22 9.34 10.80 18.20 15.60 7.55 14.70 Rb 2.98 3.12 5.36 4.22 4.57 7.56 5.60 5.62 9.24 Ba 48.9 58.1 70 163 162 987 519 705 763 Th 0.44 0.52 0.7 0.66 0.57 0.76 0.79 0.47 0.41 U 0.13 0.12 0.2 0.17 0.15 0.22 0.17 0.13 0.12 Nb 1.17 1.11 1.59 2.03 1.80 4.09 2.67 1.17 2.31 Ta 0.14 0.11 0.13 0.15 0.13 0.25 0.24 0.09 0.17 Sr 47.20 46.80 44.70 190.00 163.00 213.00 173.00 190 180 Cs 0.11 0.50 0.20 0.24 0.30 5.25 0.61 0.85 0.63 Pb 4.45 4.67 0.90 2.10 2.13 1.29 1.30 2.61 2.20 Zr 18.80 17.20 26.00 27.40 25.90 53.60 41.50 18.60 39.20 Hf 0.69 0.63 0.84 0.86 0.84 1.62 1.27 0.65 1.29 Cr 3257 2733 2720 298 1316 480 485 495 324 Co 175.00 165.00 150.00 56.30 65.80 63.60 54.40 47.30 46.20 Ni 1094 1364 662 134 191 168 219 140 81.4 Sc 24.8 36.3 41.7 26.6 49.8 39.3 27.9 38.4 51.2 V 156 147 166. 133 196 282 200 132 261 Cu 1190 1766 183 52.6 144 46.9 44.1 92.2 36.7 Ga 7.57 6.72 7.50 15.90 13.30 19.4 16.6 14.60 17.20 REE 25.21 23.22 30.16 33.61 36.49 61.05 49.61 25.21 45.28

Downloaded by [China University of Geosciences], [Mr Zhaochong Zhang] at 20:59 15 October 2012 Cu/Zr 63.30 102.67 7.04 1.92 5.56 0.88 1.06 4.96 0.94 (La/Nb)PM 1.87 1.95 1.83 1.80 1.91 1.38 1.58 2.17 1.55 (Th/Ta)PM 0.84 1.07 0.86 0.96 0.94 0.85 1.15 0.89 0.84 δEu 0.79 0.91 0.83 1.05 0.97 1.11 0.95 1.18 1.08

# Note: TFe2O3 is total iron as FeO; Mg = Mg/(Mg+Fe); δEu = 2Eun/(Smn+Gdn).

In a diagram of (Tb/Yb)PM versus (Yb/Sm)PM, melts peridotite. But the parental magma of the Nantianwan of spinel peridotite are markedly different from melts of complex was likely generated by partial melting of incom- garnet peridotite (Figure 9). Our data plot closer to a patible element-enriched spinel-facies lithospheric mantle melt range for spinel peridotite than to one for garnet and with moderate partial melting of ∼10%. This is similar 1756 M. Wang et al.

60 2.4 Gabbros Gabbronorites 55 Iherzolites 1.6 (wt.%) 2 (wt.%)

50 2 SiO

TiO 0.8 45

40 0 0 5 10 15 20 25 30 0 5 10 15 20 25 30 10 30

20 O (wt.%) 2 5 Na

CaO (wt.%) + 10 O 2 K

0 0 0 5 10 15 20 25 30 0 5 10 15 20 25 30

30 30

20 20 (wt.%) 3 O 2 TFeO (wt.%) TFeO

10 Al 10

0 0 0 5 10 15 20 25 30 0 5 10 15 20 25 30 MgO (wt%) MgO (wt%)

Figure 6. Variation of MgO versus major elements for maﬁc–ultramaﬁc rocks from the Nantianwan complex.

to the Baimazhai intrusion and Emeishan low-Ti basalts, have been contaminated by crustal materials. However, the 87 86 appearing to reflect shallower depth and lower temperature limited range of εNd(t) and initial Sr/ Sr indicates that and pressure in the spinel stability field (Xu and Chung there was little variation in the degree of crustal contami- 2001; Zhang and Wang 2002). nation (Figure 8). This is in contrast to other Ni–Cu–(PGE) deposits, such as Limahe and Baimazhai (Wang et al. 2006; Tao et al. 2008), and Noril’sk in Siberia (Lightfoot et al. Crustal contamination 1994), where the magmas were considerably contaminated Due to the highly incompatible behaviour of Nb, U, Ce, by variable crustal materials (Figure 8). Such an infer- ε / and Pb during mantle partial melting and crystalliza- ence is supported by plots of Nd(t) versus (Th Nb)PM tion processes, the Nb/U and Ce/Pb ratios are sensitive (Figure 10). In this respect, the Nantianwan complex is dis- indexes of crustal contamination (Hofmann 1988). The tinct from other intrusions hosting Ni–Cu–(PGE) ores in presence of orthopyroxene in our samples further sug- the ELIP. gests the involvement of crustal contamination. The Ce/Pb values (1–9.8) and Nb/U values (8–19) of our samples are much lower than those of the typical mantle (25 Constraints on the metallogenesis ± 5, 30–35, respectively) (Hofmann 1988; Rudnick and It is generally accepted that Cu–Ni–(PGE) sulphide Downloaded by [China University of Geosciences], [Mr Zhaochong Zhang] at 20:59 15 October 2012 Fountain 1995). Hence, it seems that the rocks from the deposits are generated by immiscible sulphide melt Nantianwan complex experienced variable crustal con- in response to formation of S-saturated magma from tamination. This conclusion is also supported by slightly S-undersaturated magma. However, there is debate around negative Nb, Ta, and Ti anomalies coupled with ele- the mechanism for S-saturation. Four mechanisms have 87 86 vated ( Sr/ Sr)t. However, these signatures can also be been proposed: (1) rapid decrease of temperature along explained as the result of addition of subduction-related the margins of intrusions (Maier et al. 1998); (2) contam- melts/fluids into the mantle source (Wilson 1989). If that ination by crustal materials (Lightfoot and Hawkesworth is the case, strong negative Nb, Ta, Zr, and Hf anomalies 1997; Ripley et al. 2003; Zhang and Wang 2003); (3) mix- would be expected (Stolz et al. 1996). Since all our sam- ing of different magmas (Lambertt et al. 1998; Xiong ples show slightly negative Nb, Ta, Zr, and Hf anomalies, 1994); and (4) rapid differentiation of magma (Haughton we tend to agree that the primary magma was likely to et al. 1974). International Geology Review 1757

(A) Lijiang picrite 100 Average OIB

10 Rock/chondrite

Iherzolites Gabbronorites Gabbros 1 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

(B) Lijiang picrite 100

Average OIB

Rock/primitive mantle Rock/primitive 1

0.1 Th U Nb Ta K La Ce Pb Pr Sr P Nd Zr Hf Sm Eu Ti Gd Tb Dy Y Ho Er Tm Yb Lu

Figure 7. (A) Chondrite-normalized REE patterns of the Nantianwan complex, normalized values after Taylor and Mclennan (1985). The diagram for those of Lijiang picrite (Zhang et al. 2006) is shown for comparison. Primitive mantle-normalizing values and average OIB pattern are from Sun and McDonough (1989). (B) Primitive mantle-normalized trace elements spider diagram of Nantianwan complex. Normalized values after Sun and McDonough (1989); data of gabbros in (A) from Zeng, L.G., personal communication.

Table 4. Sr and Nd isotopic geochemistry for the Nantianwan complex.

87 86 87 86 87 86 147 144 143 144 143 144 Sample no. Rb Sr Rb/ Sr ( Sr/ Sr)0 ( Sr/ Sr)t Sm Nd Sm/ Nd ( Nd/ Nd)0 ( Nd/ Nd)t εNd(t)

DPZ-1 2.933 29.91 0.2837 0.706806 0.7057800 0.9298 3.112 0.1807 0.512626 0.512318831 0.30 DPZ-2 3.058 42.80 0.2068 0.708228 0.7074801 0.8604 2.88 0.1807 0.512603 0.512295831 −0.15 DPZ-7 1.266 107.10 0.0342 0.706140 0.7060162 1.289 4.409 0.1769 0.512606 0.512325291 0.42 NTG-1 2.350 120.30 0.0565 0.705628 0.7054237 1.918 6.469 0.1793 0.512695 0.512390211 1.69 NTG-3 1.327 131.30 0.0292 0.707740 0.7076342 0.8011 2.791 0.1737 0.512577 0.512281731 −0.43 NTG-4 4.846 134.30 0.1044 0.707066 0.7066884 1.7958 5.804 0.1872 0.512695 0.512376782 1.43 Downloaded by [China University of Geosciences], [Mr Zhaochong Zhang] at 20:59 15 October 2012 143 144 0 143 144 0 Note: Chondrite uniform reservoir (CHUR) values [( Sm/ Nd) CHUR = 0.512638, ( Nd/ Nd) CHUR = 0.1967] are used for the calculation. 87 86 143 144 ( Sr/ Sr)t,( Nd/ Nd)t,andεNd(t) were calculated at 260 Ma.

Many researchers have suggested that crustal contami- well as speeding crystallization, all factors which could nation is the most important factor triggering S-saturation make magmas S-oversaturated. However, the degree of of basalt magmas (e.g. Brügmann et al. 1993; Arndt crustal contamination in the Nantianwan complex is not as et al. 2003; Naldrett 1973, 2004; Lightfoot and Keays high as expected. The presence of cumulus texture, com- 2005), because it can also lead to the addition of sul- bined with the negative correlation of MgO with SiO2, phur and silicon from the wallrocks to the magma systems, Na2O, TiO2, and A12O3 and the roughly positive corre- decreasing iron contents, increasing oxygen fugacity, as lation with FeO (Figure 6), indicates the fractionation of 1758 M. Wang et al.

15 10

Nantianwan Uncontaminated 10 Siberian Traps 5

Basalts in ELIP 5 (t)

5 Nd 0 Fe–Ti–V oxide ε ore-bearing intusions 0 Nantianwan (t) 20 –5 Baimazhai Nd Crustal contamination Limahe ε Cu–Ni–(PGE) Xinjie –5 ore-bearing intusions Panzhihua OIB 50 –10 EMI 0123456 –10 10 EMII 70 UCC (Th/Nb)PM –15 Figure 10. Plots of εNd(t) versus (Th/Nb)PM (data for Xinjie 20 LC intrusion after Zhao et al. 2006; data for Limahe, Panzhihua intru- –20 sions after Zhang et al. 2009; data for Baimazhai after Wang 0.702 0.704 0.706 0.708 0.710 0.712 0.714 2008; normalized values after Sun and McDonough 1989). 87 86 Sr/ Srt melt. Since no evidence for magma mixing (e.g. disequilib- ε 87 /86 Figure 8. Nd(t)versus( Sr Sr)t for the Nantianwan complex. rium mineral assemblage in the rocks) has been observed, Data sources: Cu–Ni–(PGE) sulphide-bearing intrusions (Wang the possibility of a simple magma-mixing model can be 2008; Zhang et al. 2009); the Fe–Ti–V oxide ore-bearing intrusions (Zhang et al. 2009; Pang et al. 2010); Emeishan basalts ruled out. Finally, gypsiferous strata have not been rec- data (Xu et al. 2001; Hanski et al. 2004; Xiao et al. 2004; Wang ognized in the Panxi district unlike the Noril’sk region et al. 2007 ); Siberian Traps (Sharma et al. 1992; Lightfoot et al. (Lightfoot and Hawkesworth 1997). Therefore, contami- 1993; Wooden et al. 1993); dotted lines with numbers showing nationbyS-enriched country rock can be excluded. The the mixing ratio of the mantle and crust material, as the original 87 86 only viable mechanism left is the rapid differentiation of magma and the Yangtze upper crust ( Sr/ Sr = 0.72, εNd(t) = 87 86 magma to facilitate the Cu–Ni mineralization, which is fur- −10) and Yangtze lower crust ( Sr/ Sr = 0.71, εNd(t) =−30) as contaminants (Jahn et al. 1999). ther explored in the discussion below. As stated above, the Nantianwan complex experienced fractional crystallization of olivine and clinopyroxene and slight crustal contamination. However, it is not clear at what stage in the crystallization sequence the ore-bearing magmas became saturated in sulphide. The chemical composition of olivine occurring in the Cu–Ni sulphide ore-bearing intrusions can be used to infer the presence of an immiscible S-rich liquid in equilibrium with the silicate melt (Keays 1995; Naldrett 1997, 1999). Ni and Mg are compatible in early crystallizing olivine and thus decrease in abundance in later phases as crystallization proceeds (Li et al. 2003). The DNi value between solid phase and magma, particularly between olivine and magma, is summarized in a recent review by Bédard (2005); Li et al. (2003) suggested an olivine/melt Figure 9. Variation of (Tb/Yb) versus (Yb–Sm) (after PM PM Ni Shaw 1970; Janney et al. 2000; Salters and Stracke 2004; Ito and D value of ∼7 for S-bearing basaltic magmas based Mahoney 2005; Workman and Hart 2005; Zhang et al. 2006; data on the analyses of MORB samples. In komatiitic magma,

Downloaded by [China University of Geosciences], [Mr Zhaochong Zhang] at 20:59 15 October 2012 for Xinjie intrusion after Zhao et al. 2006; data for Limahe intru- the olivine/melt DNi is between 3 and 5 (Arndt 1977). sion after Zhang et al. 2009; data for Baimazhai after Wang 2008; Based on the available experimental data, a DNi value of data for Emeishan low-Ti basalts after Xu et al. 2001b). 1 and ∼0 for clinopyroxene and plagioclase, respectively, is reasonable for basaltic systems. When olivine is in equi- olivine, clinopyroxene, and orthopyroxene. The complex librium with a sulphide liquid after the segregation of liquid was emplaced into the Silurian Zhongcao Formation along sulphide during silicate crystallization, the olivines will a NW–SE-striking fault, suggesting a relatively deeper contain lower Ni than those that crystallize with no seg- emplacement depth (∼4.5–5.5 km in the light of regional regation of liquid sulphide at given MgO. This relation stratigraphic correlation). Thus, shallow emplacement and can be used to evaluate sulphide liquid composition using rapidly decreasing temperature of the magma are unlikely olivine from sulphide-bearing rock samples (Simkin and factors leading to the formation of an immiscible sulphide Smith 1970). Our analyses show that the Fo contents in the International Geology Review 1759

5000 The dispersed sulphide droplets would form the disseminated Cu–Ni sulphide ores. Globally, it is generally No sulphide removal accepted that the generation of magmatic Cu–Ni sulphide 4000 Sulphide removal ore deposits requires crustal contamination of mantle- Olivine from Iherzolites derived magmas (Lesher and Keays 2002; Lightfoot and 3000 Keays 2005; Wilson and Chunnett 2006). More contro- versial is the role of crustal versus. mantle S in both 2000 triggering and maintaining the S-saturated state of the magma. At issue is whether very large ore deposits require Ni in olivine (ppm) Ni in olivine a significant crustal contribution of S. As stated above, 1000 the Nantianwan intrusion is distinct from other large-scale Cu–Ni sulphide deposits in which crustal contamination 0 played an important role in generating an immiscible 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 sulphide melt, such as Sudbury in Canada (Keays and Fo in olivine Lightfoot 2004), Noril’sk in Siberia (Lightfoot and Keays Figure 11. Fo–NiO diagram of olivine (after Simkin and Smith 2005), and Jinchuan in China (Tang 1990). The large 1970). quantity of Cu–Ni–PGE sulphide ores that occur in the Noril’sk intrusion has high initial 87Sr/86Sr ratios, which are attributed in part to the assimilation of anhydrite- olivine from the lherzolites vary from 79.1 to 81.4 ppm, and rich, evaporitic sediments. These sediments are abundant the corresponding Ni contents vary from 0.11 to 0.23 ppm in the sedimentary sequence that underlies the volcanic (Figure 11). Most olivines plot within and below the field sequence (Lightfoot and Hawkesworth 1997). The parental of normal olivine compositions, strongly suggesting that magmas that formed the Noril’sk intrusion ascended to some liquid sulphides have separated from the ore-bearing a higher level chamber within the sedimentary sequence, magma (Naldrett 1989; Naldrett et al. 1992). where it encountered evaporites, assimilated sulphur, and The Cu/Zr ratios compare the concentrations of two segregated immiscible sulphide melts (Arndt et al. 2003). highly incompatible elements, one being a highly chal- Previous studies suggested that the capacity of a cophile element (Cu) and the other (Zr) a non-chalcophile, magma to form an economic Ni–Cu–(PGE) deposit is con- lithophile element (Li and Naldrett 1999). These two ele- trolled mainly by (1) the abundance of ore metals in the ments are similarly incompatible during the early stages magma; (2) the sulphide saturation state of the magma; of fractional crystallization of sulphide-undersaturated and (3) the capacity of the magma to interact with its mafic–ultramafic magmas. Typically, chalcophile metal- surroundings (see discussion by Lesher and Campbell undepleted continental flood basalts have Cu/Zr ratios 1993). In other words, to form an economic Ni–Cu–(PGE) around 1, whereas lavas depleted in chalcophile metals due deposit, the primary mantle-derived magma must contain to sulphide segregation have Cu/Zr ratios less than <1 sufficient ore metals and must be capable of being driven (Lightfoot and Keays 2005). The bulk-rock Cu/Zr ratios in to sulphide saturation. If the magma is too undersaturated lherzolites (7.04–102.67) and gabbronorites (0.88–5.56) of in sulphide and/or does not interact with wallrocks, it may the Nantianwan complex indicate that the parental magmas not reach sulphide saturation until crystallization is well of the gabbronorites and lherzolites went from undersatura- advanced, resulting in only small amounts of fine-grained tion to oversaturation of sulphur, which may have been due disseminated sulphides and very fine grained ‘cloudy’ or to fractional crystallization in a high-level magma cham- ‘dusty’ sulphides. Hence, although an immiscible sulphide ber. This can account for the sulphide segregation, although phase is a normal segregation product in most mafic– the role of crustal contamination cannot be ignored for the ultramafic systems, it usually segregates in only very small formation of the Cu–Ni ores in the Nantianwan complex. amounts because of limitations on the abundance of S in Interestingly, a similar scenario has also been recognized in Downloaded by [China University of Geosciences], [Mr Zhaochong Zhang] at 20:59 15 October 2012 the magma and usually at a late stage during the crystal- the Uitkomst intrusion, Mpumalanga, South Africa (Maier lization of the magma. Thus, Arndt et al. (2005) concluded et al. 1998). that the magma must be driven to sulphide saturation by contamination, with or without the addition of external S, to not only saturate the magma in sulphide but also to form Assessment of ore potential large enough amounts of a molten immiscible sulphide As stated above, the primary basaltic magma underwent liquid. As a consequence, considering the insignificant fractional crystallization of Opx+Cpx+Ol and segregation crustal contamination in the Nantianwan mafic–ultramafic of sulphides. Because of the emplacement depth, the complex and relatively homogeneous fine-grained dissem- fractional crystallization of olivine and clinopyroxene will inated sulphides in the complex, we infer that it does not lead to the formation of sulphide droplets that crystallize. have much potential for large Cu–Ni sulphide deposits. 1760 M. Wang et al.

Constraints on geodynamics Based on the above discussion, it is reasonable to As stated above, the Nantianwan complex is temporally and infer that the primitive magmas of the Nantianwan com- ∼ spatially associated with the Emeishan continental flood plex were formed when large-scale thinning ( 60 km) of basalts, but their genetic relationship has not been dis- the lithosphere resulted from the interaction between the cussed with regard to their evolution, although many other Emeishan mantle plume and the lithosphere. Alternatively, layered mafic–ultramafic intrusions in the area have been because the plume impinged near the margin of the craton, considered to be a part of the ELIP (e.g. Zhou et al. 2002; different parts of the plume head may have interacted with Zhong and Zhu 2006; Zhang et al. 2009) on the basis of the lithosphere of variable initial thickness. Regardless, their close spatial and temporal relationship. since the age of the Nantianwan complex is very close The U–Pb age of the Nantianwan gabbronorites is to the time of other mafic–ultramafic intrusions in Panxi 259.7 ± 0.6 million years, which is very close to the age district and major eruption of the Emeishan flood basalts, of other mafic–ultramafic intrusions in the Panxi area it is fairly clear that impingement of Emeishan mantle (Table 1, Zhong et al. 2002; Wang et al. 2006) and the plume played a crucial role in the petrogenesis (Griffiths time of major eruptions of the Emeishan flood basalts and Campbell 1991). (Zhou et al. 2002; Fan et al. 2004; He et al. 2003). Many continental flood basalts or LIPs, including those hosting world-class Ni–Cu–PGE ore deposits have been Conclusions thought to be genetically related to a mantle plume (e.g. The Nantianwan complex consists predominantly of Noril’sk-Talnakh and the Siberian Traps; Lightfoot et al. gabbros, gabbronorites, and lherzolites. The geochemical 1990; Naldrett et al. 1992; Hawkesworth et al. 1995; characteristics suggest a comagmatic relationship among Lightfoot and Hawkesworth 1997; Lightfoot and Keays the main rock types. The parental magma did not undergo 2005). The ELIP covers an area of more than 300,000 km2 significant crustal contamination, but was subjected to with flood basalts and many associated mafic–ultramafic significant fractional crystallization of olivine, clinopy- intrusions in the western part of the Yangtze Craton in roxene, and minor orthopyroxene, which was the likely southwest China and northern Vietnam (Zhou et al. 2005). crucial factor leading to sulphide oversaturation. This It has been documented that the Emeishan flood basalts contrasts with other large mafic–ultramafic ore-bearing formed from a Permian mantle plume that reached the intrusions in which crustal contamination triggered sep- base of the lithosphere beneath south China at ∼260 Ma aration of a sulphide-rich liquid phase. Thus, we infer (Chung and Jahn 1995; Xu and Chung 2001; Zhou et al. that the ore potential of this intrusion is probably 2002; Zhang et al. 2006). Thus, the Nantianwan complex limited. is generally considered to be a part of the ELIP in terms The primitive magma which formed the Nantianwan of the Emeishan magma-plumbing system at ∼260 Ma. complex was likely to have been derived from an incom- This linkage is evidenced by the Sr–Nd isotope data of patible element-enriched spinel-facies lherzolite of the some samples, which overlap the field of Emeishan basalts lithospheric mantle. The LA–ICP–MS zircon data yield a in Figure 6. In addition, they have overall similar shapes U–Pb age of 258.7 ± 0.6 million years, which is consistent in REE and normalized incompatible element patterns to with the ages of other Emeishan mafic–ultramafic intru- those of OIB and Lijiang picrites (Figure 7). As stated sions, which are interpreted in terms of a mantle plume above, the primitive magma which formed the Nantianwan interacting to a variable extent with the lithosphere. complex was derived from incompatible element-enriched spinel-facies lherzolite. Thus, the magmas were formed at < relatively shallow depth, 80 km (McKenzie and Bickle Acknowledgements 1988). However, the Lijiang picritic lava represents the Aspects of this work were supported by the 973 Project (Grant No. product of the Emeishan plume during the early stage, cor- 2012CB416800), National Natural Science Foundation of China responding to a lithosphere thickness of ∼140 km (Zhang (Grant No. 40925006), Special Fund for Scientific Research in Downloaded by [China University of Geosciences], [Mr Zhaochong Zhang] at 20:59 15 October 2012 et al. 2006). In general, many mineralized intrusions are the Public Interest (200911007-25), ‘the Fundamental Research linked with deep lithospheric faults, where the lithosphere Funds for the Central Universities’, 111 Project (B07011), and PCSIRT. may have been disrupted as a result of tectonic events or located at cratonic margins such as the Noril’sk (Naldrett 1999) and Emeishan (Zhou et al. 2005). Deep faults or other weak cratonic margins may provide the channel for References eruption and emplacement of continental flood basalts that Ali, J.R., Fitton, J.G., and Herzberg, C., 2010, Emeishan large igneous province (SW China) and the mantle plume up- bear geochemical signatures inherited from interaction doming hypothesis: Journal of the Geological Society, v. 167, between plume heads and cratonic subcontinental litho- p. 953–959. spheric mantle (SCLM) over a wide area (Zhang et al. Arndt, N.T., 1977, The partitioning of nickel between olivine and 2008). ultrabasic and basic komatiite liquids, Carnegie Institution International Geology Review 1761

of Washington Year Book 76: Washington, D.C., Carnegie He, B., Xu, Y.G., and Chung, S.L., 2003, Sedimentary evidence Institution of Washington, p. 553–557. for rapid, kilometer-scale crustal doming prior to the eruption Arndt, N.T., Czamanske, G., Walker, R.J., ChauveL, C., and of the Emeishan flood basalts: Earth and Planetary Science Fedorenko, V.,2003, Geochemistry and origin of the intrusive Letters, v. 213, p. 319–405. hosts of the Noril’sk-Talnakh Cu-Ni-PGE sulfide deposits: Hofmann, A.W., 1988, Chemical differentiation of the earth: Economic Geology, v. 98, p. 495–515. The relationship between mantle, continental crust and Arndt, N.T., Lesher, C.M., and Czamanske, G.K., 2005, Mantle- oceanic crust: Earth and Planetary Science Letters, v. 90, derived magmas and magmatic Ni-Cu-(PGE) deposits: 100th p. 297–314. Anniversary Volume, Society of Economic Geologists, Hoskin, P.W.O., and Black, L.P., 2000, Metamorphic zircon p. 5–23. formation by solid-state recrystallization of protolith Bédard, J.H., 2005, Partitioning coefficients between olivine and igneous zircon: Journal of Metamorphic Geology, v. 18, silicate melts: Lithos, v. 83, p. 394–419. p. 423–439. Bhat, I., Zainuddin, S.M., and Rais, A., 1981, The Panjal Trap Hou, K.J., Li, Y.H., and Tian, Y.Y., 2009, In situ U–Pb zircon dat- chemistry and the birth of Tethys: Geological Magazine, ing using laser ablation-multi ion counting-ICP-MS: Mineral v. 118, p. 367–375. Deposits, v. 28, no. 4, p. 481–492 (in Chinese with English Brügmann, G.E., Naldrett, A.J., Asif, M., Lightfoot, P.C., and abstract). Gorbachev, N.S., 1993, Siderophile and chalcophile metals Hou, T., Zhang, Z.C., Kusky, T., Du, Y.S., Liu, J.L., and Zhao, as tracers of the evolution of the Siberian Trap in the Noril’sk Z.D., 2011, A reappraisal of the high-Ti and low-Ti classi- region, Russia: Geochimica et Cosmochimica Acta, v. 57, fication of basalts and petrogenetic linkage between basalts p. 2001–2018. and mafic–ultramafic intrusions in the Emeishan Large Chung, S.L., and Jahn, B.M., 1995, Plume-lithosphere interaction Igneous Province, SW China: Ore Geology Reviews, v. 41, in generation of the Emeishan flood basalts at the Permian- p. 133–143. Triassic boundary: Geology, v. 23, p. 889–892. Ito, G., and Mahoney, J.J., 2005, Flow and melting of a het- Corfu, F., Hanchar, J.M., Hoskin, P.W., and Kinny, P., 2003, Atlas erogeneous mantle: 1. Importance to the geochemistry of of zircon textures: Reviews in Mineral and Geochemistry, ocean island and mid-ocean ridge basalts: Earth and Planetary v. 53, p. 468–500. Science Letters, v. 230, p. 29–46. Dobretsov, N.L., 2005, 250 Ma large igneous province of Jahn, B.M., Wu, F.Y., Lo, C.H., and Tsai, C.H., 1999, Crust– Asia: Siberian and Emeishan traps (plateau basalts) and mantle interaction induced by deep subduction of the conti- associated granitoids: Russian Geology and Geophysics, nental crust: Geochemical and Sr–Nd isotopic evidence from v. 46, p. 879–890. post-collisional mafic–ultramafic intrusions of the northern Dulski, P., 1994, Interferences of oxide, hydroxide and chlo- Dabie complex, central China: Chemical Geology, v. 157, ride species in the determination of rare earth elements p. 119–146. in geological samples by inductively coupled plasma–mass Janney, P.E., Macdougall, J.D., Natland, J.H., and Lynch, M.A., spectrometry: Fresenius Journal of Analytical Chemistry, 2000, Geochemical evidence from the Pukapuka volcanic v. 350, p. 194–203. ridge system for a shallow enriched mantle domain beneath Fan, W.M., Wang, Y.J., Peng, T.P., Miao, L.C., and Guo, F., 2004, the South Pacific superswell: Earth and Planetary Science Ar-Ar and U-Pb geochronology of Late Paleozoic basalts Letters, v. 181, p. 47–60. in western Guangxi and its constraints on the eruption age Keays, R.R., 1995, The role of komatiitic and picritic magma- of Emeishan basalts magmatism: Chinese Science Bulletin, tism and S-saturation in the formation of ore deposits: Lithos, v. 49, no. 18, p. 1892–1900 (in Chinese). v. 34, p. 1–18. Grant, M.L., Wilde, S.A., Wu, F.Y., and Yang, J.H., 2009, The Keays, R.R., and Lightfoot, P.C., 2004, Formation of Ni– application of zircon cathodoluminescence imaging, Th-U- Cu–platinum group element sulphide mineralisation in the Pb chemistry and U-Pb ages in interpreting discrete magmatic Sudbury impact melt sheet: Mineral Petrology, v. 82, and high-grade metamorphic events in the North China p. 217–258. Craton at the Archean/Proterozoic boundary: Chemical Lambertt, D.D., Walker, R.J., and Morgan, J.W., 1998, Re-Os Geology, v. 261, p. 155–171. and Sm-Nd isotope geochemistry of the Stillwater complex, Griffiths, R.W., and Campbell, I.H., 1991, Interaction of man- Montana: Implications for the petrogenesis of the J-M Reef: tle plume heads with the Earth’s surface and onset of Journal of Petrology, v. 35, p. 1717–1753. small-scale convection: Journal of Geophysical Research, Lesher, C.M., and Campbell, I.H., 1993, Geochemical v. 96, no. 18, p. 295–310. and fluid dynamic modeling of compositional vari- Hanchar, J.M., and Rundnick, R.L., 1995, Revealing hidden struc- ations in Archean komatiite-hosted nickel sulfide tures: The application of cathodoluminescence and back- ores in Western Australia: Economic Geology, v. 88, scattered electron imaging to dating zircons from lower p. 804–816. Downloaded by [China University of Geosciences], [Mr Zhaochong Zhang] at 20:59 15 October 2012 crustal xenoliths: Lithos, v. 36, p. 289–303. Lesher, C.M., and Keays, R.R., 2002, Komatiite-associated Hanski, I., Erälahti, C., Kankare, M., Ovaskainen, O., and Ni–Cu-(PGE) deposits: geology, mineralogy, geochemistry Sirén, H., 2004, Variation in migration rate among indi- and genesis, in Cabri, L., ed., The geology, geochemistry, viduals maintained by landscape structure: Ecology Letters, mineralogy and beneficiation of the platinum-group elements, v. 7, p. 958–966. Volume 54: Canadian Institute of Mining Metallurgy and Haughton, D.R., Roeder, P.L., and Skinner, B.J., 1974, Solubility Petroleum, p. 579–618. of sulfur in mafic magmas: Economic Geology, v. 69, Li, C., Ripley, E.M., and Mathez, E.A., 2003, The effect of S p. 451–467. on the partitioning of Ni between olivine and silicate melt in Hawkesworth, C.J., Lightfoot, P.C., Fedorenko, V.A., Blake, S., MORB: Chemical Geology, v. 201, p. 295–306. Naldrett, A.J., Doherty, W., and Gorbachev, N.S., 1995, Li, C.S., and Naldrett, A.J., 1999, Geology and petrology of the Magma differentiation and mineralisation in the Siberian Voisey’s Bay intrusion: Reaction of olivine with sulfide and flood basalts: Lithos, v. 34, p. 61–88. silicate liquids: Lithos, v. 47, p. 1–31. 1762 M. Wang et al.

Liang, Y.B., Liu, T.Y., Song, G.R., and Jin, Z.M., 1998, Platinum Naldrett, A.J., 1999, World-class Ni-Cu-PGE deposits: Key fac- group element deposits in China: Beijing, Metallurgical tors in their genesis: Mineralium Deposita, v. 34, p. 227–240. Industry Press, 185 p (in Chinese). Naldrett, A.J., 2004, Magmatic sulfide deposits: Geology, geo- Lightfoot, P.C., and Hawkesworth, C.J., 1997, Flood basalts chemistry and exploration: Heidelberg, Springer Verlag, and magmatic Ni, Cu, and PGE sulfide mineralization: 728 p. Comparative geochemistry of the Noril’sk (Siberian traps) Naldrett, A.J., Lightfoot, P.C., Fedorenko, V.A., Doherty, W., and and West Greenland sequences: Geophysical Monograph, Gorbachev, N.S., 1992, Geology and geochemistry of intru- v. 100, p. 357–380. sions and flood basalts of the Noril’sk Region, USSR, with Lightfoot, P.C., Hawkesworth, C.J., Hergt, J., Naldrett, A.J., implications for the origin of the Ni-Cu ores: Economic Gorbachev, N.S., Fedorenko, V.A., and Doherty, W., 1993, Geology, v. 87, p. 975–1004. Remobilisation of the continental lithosphere by a mantle Nasdala, L., Norberg, N., Schaltegger, U., Schoene, B., Tubrett, plume: Major-, trace element and Sr-, Nd-, and Pb-isotope M.N., and Whitehouse, M.J., 2008, Plesovice zircon-a new evidence from picritic and tholeiitic lavas of the Noril’sk natural reference material for U–Pb and Hf isotopic micro- District, Siberian Trap, Russia: Contributions to Mineralogy analysis: Chemical Geology, v. 249, p. 1–35. and Petrology, v. 114, p. 171–188. Pang, K.N., Zhou, M.F., Qi, L., Shellnutt, G., Wang, C.Y., and Lightfoot, P.C., and Keays, R.R., 2005, Siderophile and chal- Zhao, D.G., 2010, Flood basalt-related Fe-Ti oxide deposits cophile metal variations in flood basalts from the Siberian in the Emeishan large igneous province, SW China: Lithos, v. Trap, Noril’sk Region: Implications for the origin of 119, p. 123–136. the Ni–Cu–PGE sulfide ores: Economic Geology, v. 100, Qi, L., and Grégoire, D.C., 2000, Determination of trace elements p. 439–462. in twenty-six Chinese geochemistry reference materials by Lightfoot, P.C., Naldrett, A.J., Gorbachev, N.S., Doherty, W., inductively coupled plasmamass spectrometry: Geostandard and Fedorenko, V.A., 1990, Geochemistry of the Siberian Newsletter, v. 24, p. 51–63. Trap of the Noril’sk Area, USSR, with implications for the Ripley, E.M., Lightfoot, P.C., Li, C.S., and Elswick, E.R., 2003, relative contributions of crust and mantle to flood basalt Sulfur isotopic studies of continental flood in the Noril’sk magmatism: Contributions to Mineralogy and Petrology, region: Implications for the association between lavas and v. 104, p. 631–644. ore-bearing intrusions: Geochimica et Cosmochimica Acta, Lightfoot, P.C., Naldrett, A.L., Gorbachev, N.S., Fedorenko, v. 67, no. 15, p. 2805–2817. A., and Doherty, W., 1994, Chemostratigraphy of the Roeder, P.L., and Emsile, R.F., 1970, Olivine-liquid equilibrium: Siberian Trap lava, Noril’sk district, Russia: Implications and Contributions to Mineralogy and Petrology, v. 29, p. 275–289. source of flood basalt magma of Sudbury-Noril’sk sympo- Rudnick, R.L., and Fountain, D.M., 1995, Nature and composition sium: Ontario Geological Survey (Special Publication), v. 5, of the continental crust: A lower crustal perspective: Review p. 283–312. of Geophysics, v. 33, p. 267–309. Liu, B.G., Luo, Z.N., Yao, Y., and Zhong, H., 2008a, PGE min- Salters, V.J.M., and Stracke, A., 2004, Composition of the eralization of layered intrusions in Panxi rift region, China: depleted mantle: Geochemistry Geophysics Geosystems, v. 5, Earth Science Frontiers, v. 15, no. 4, p. 269–279 (in Chinese Q05B07 27. with English abstract). Saunders, A.D., Norry, M.J., and Tarney, J., 1988, Origin of Liu, Y.S., Hu, Z.C., Gao, S., Gunther, D., Xu, J., Gao, C., and MORB and chemically depleted mantle reservoirs: Trace Chen, H., 2008b, In situ analysis of major and trace elements element constraints: Journal of Petrology, special volume, of anhydrous minerals by LA-ICP-MS without applying an p. 415–445. internal standard: Chemical Geology, v. 257, p. 34–43. SBGMR (Sichuan Bureau of Geology and Mineral Resources), Ludwig, K.R., 2003, User’s manual for Isoplot 3.00. 1991, Regional geology of Sichuan Province: Beijing, China, A geochronological toolkit for Microsoft Excel Berkeley Geological Publishing House (in Chinese). Geochronology Center Special Publication, v. 4: Berkeley, Sharma, M.A., 1997, Siberian traps, in Mahoney, J.J., and CA, Berkeley Geochronology Center. Coffin, M.F., eds., Large igneous provinces: Continental, Luo, Y.N., 1981, The Emei taphrogenesis of the upper Yangtze oceanic, and planetary flood volcanism: Geophysical Platform in south China: Geological Magazine, v. 127, Monographs, Volume 100: American Geophysical Union, p. 393–405. p. 273–295. Maier, W.D., Barnes, S.J., and De Waal., S.A., 1998, Exploration Sharma, M.A., Basu, R., and Nesterenko, G.V., 1992, Temporal for magmatic Cu-Ni-PGE sulphide deposits: A review of Sr-, Nd-, and Pb-isotopic variations in the Siberian recent advances in the use of geochemical tools, and their flood basalts: implications for the plume-source charac- application to some South African ores: South African teristics: Earth and Planetary Science Letters, v. 113, Journal of Geology, v. 101, no. 3, p. 237–253. p. 365–381. McKenzie, D.P., and Bickle, M.J., 1988, The volume and com- Shaw, D.M., 1970, Trace element fractionation during anatexis: Downloaded by [China University of Geosciences], [Mr Zhaochong Zhang] at 20:59 15 October 2012 position of melt generated by extension of the lithosphere: Geochimica et Cosmochimica Acta, v. 34, p. 237–243. Journal of Petrology, v. 29, p. 625–679. Simkin, T., and Smith, J.V., 1970, Minor-element distribution in Naldrett, A.J., 1973, Nickel sulphide deposits: Their classification olivine: The Journal of Geology, v. 78, no. 3, p. 304–325. and genesis with special emphasis on deposits of volcanic Sláma, J., Kosler, J., Condon, D.J., Crowley, J.L., Gerdes, A., association: Canadian Mining and Metallurgical Bulletin, Hanchar, J.M., Horstwood, M.S.A., Morris, G.A., Nasdala, v. 66, p. 45–63. L., Norberg, N., Schaltegger, U., Schoene, B., Tubrett, M.N., Naldrett, A.J., 1989, Magmatic sulfide deposits: Oxford, Oxford and Whitehouse, M.J., 2008, Plesovice zircon-a new natural University Press, 196 p. reference material for U-Pb and Hf isotopic microanalysis: Naldrett, A.J., 1997, Key factors in the genesis of Noril’sk, Chemical Geology, v. 249, p. 1–35. Sudbury, Jinchuan, Voisey’s Bay and other world-class Song, X.Y., Zhang, C.J., Hu, R.Z., Zhong, H., Zhou, M.F., Ma, Cu-Ni-PGE deposits: Implication for exploration: Earth R.Z., and Li, Y.Z., 2005, Genetic links of magmatic deposits Science, v. 44, p. 281–315. in the Emeishan igneous province with dynamics of mantle International Geology Review 1763

plume: Journal of Mineralogical and Petrological Sciences, Xiao, L., Xu, Y.G., Mei, H.J., Zheng, Y.F., He, B., and Piranjno, F., v. 25, p. 35–44 (in Chinese with English abstract). 2004, Distinct mantle sources of low-Ti and high-Ti basalts Song, X.Y., Zhou, M.F., Keays, R.R., Cao, Z.M., Sun, M., and from the western Emeishan large igneous province, SW Qi, L., 2006, Geochemistry of the Emeishan flood basalts at China: Implications for plume-lithosphere interaction: Earth Yangliuping, Sichuan, SW China: implication for sulphide and Planetary Science Letters, v. 228, p. 525–546. segregation: Contributions to Mineralogy and Petrology, v. Xiong, Y.L., 1994, The ore-forming materials for magmatic ore 152, p. 53–74. deposits in the light of Re-Os systematics: Earth Science Stille, P., Unruh, D.M., and Tatsumoto, M., 1983, Pb, Sr, Nd, and Frontiers, v. 1, p. 199–203 (in Chinese with English abstract). Hf isotopic evidence of multiple sources for Oahu, Hawaii Xu, Y.G., and Chung, S.L., 2001, The Emeishan large igneous basalts: Nature, v. 304, p. 25–29. province: Evidence for mantle plume activity and melting Stolz, A.J., Jochum, K.P., and Hofmann, A.W., 1996, Fluid- condition: Geochimica, v. 30, p. 1–9 (in Chinese). and melt-related enrichment in the subarc mantle: Evidence Xu, Y.G., Chung, S.L., Jahn, B.M., and Wu, G.Y., 2001, from Nb/Ta variations in island-arc basalts: Geology, v. 24, Petrological and geochemical constraints on the petrogenesis p. 587–590. of the Emeishan Pemo-Triassic Emeishan flood basalts in Sun, S.S., and McDonough, W.F., 1989, Chemical and isotopic southwestern China: Lithos, v. 58, p. 145–168. systematic of oceanic basalts: Implication for mantle com- Xu, Y.G., He, B., Chung, S.L., Menzies, M.A., and Frey, F.A., position and processes: Geological Society, London, Special 2004, Geologic, geochemical, and geophysical consequences Publications, v. 42, p. 313–345. of plume involvement in the Emeishan flood-basalt province: Tang, Z., 1990, Minerogenitic model of the Jinchuan copper and Geology, v. 32, p. 917–920. nickel sulfide deposit: Geoscience, v. 4, no. 4, p. 55–61 (in YBGMR (Yunnan Bureau of Geology and Mineral Resources), Chinese). 1990, Regional geology of Yunnan Province: Beijing, Tao, Y., Li, C., Song, X.Y., and Repley, E.M., 2008, Geological Publishing House, 728 p (in Chinese with English Mineralogical, petrological, and geochemical studies of the abstract). Limahe mafic-ultramafic intrusion and associated Ni-Cu sul- Zhang, C.J., Wang, Y.L., Li, X.L., Xiu, S.Z., and Huang, Y.J., fide ores, SW China: Mineralium Deposita, v. 43, no. 8, 1998, Geochemistry of platinum group elements in mafic- p. 849–872. ultramafic rocks from the Xinjie intrusion: Geochimica, v. 27, Taylor, S.R., and Mclennan, S.M., 1985, The continental crust: its p. 458–466 (in Chinese with English abstract). composition and evolution: Oxford, UK, Blackwell Zhang, M., O’Reilly, S.Y., Wang, K.L., Hronsky, J., and Griffin, Thompson, G.M., Ali, J.R., Song, X., and Jolley, D.W., 2001, W.L., 2008, Flood basalts and metallogeny: The litho- Emeishan basalts, SW China: Reappraisal of the formation’s spheric mantle connection: Earth-Science Reviews, v. 86, type area stratigraphy and a discussion of its significance as a p. 145–174. large igneous province: Journal of Geological Society, v. 158, Zhang, Z.C., Mahoney, J.J., Mao, J.W., and Wang, F.S., 2006, p. 593–599. Geochemistry of picritic and associated basalt flows of the Thompson, R.N., and Gibson, S.A., 2000, Transient high tem- western Emeishan flood basalt Province, China: Journal of peratures in mantle plume heads inferred from magnesian Petrology, v. 47, p. 1997–2019. olivines in Phanerozoic picrites: Nature, v. 407, p. 502–506. Zhang, Z.C., Mao, J.W., Saunders, A.D., Ai, Y., Li, Y., and Zhao, Wang, C.Y., 2008, Origin of the Permian Baimazhai magmatic L., 2009, Petrogenetic modeling of three mafic-ultramafic Ni-Cu-(PGE) sulfide deposits, Yunnan: Implications for the layered intrusions in the Emeishan large igneous province, relationship of crustal contamination and mineralization: SW China, based on isotopic and bulk chemical constraints: Bulletin of Mineralogy, Petrology and Geochemistry, v. 27, Lithos, v. 113, p. 369–392. p. 332–343 (in Chinese with English abstract). Zhang, Z.C., and Wang, F.S., 2002, Geochemistry of the two types Wang, C.Y., Zhou, M.F., and Keays, R.R., 2006, Geochemical of basalts of the Emeishan basalts Province: Evidences for constraints on the origin of the Permian Baimazhai mafic- mantle plume-lithosphere interaction: Acta Geologica Sinica, ultramafic intrusion, SW China: Contributions to Mineralogy v. 76, no. 2, p. 138–147. and Petrology, v. 152, no. 3, p. 309–321. Zhang, Z.C., and Wang, F.S., 2003, A method for identifying Wang, C.Y., Zhou, M.F., and Qi, L., 2007, Permian flood basalts primary magma-Examples from picrite and alkali Basalts: and mafic intrusions in the Jinping (SW China)–Song Da Journal of Jilin University (Earth Science Edition), v. 33, (northern Vietnam) district: mantle sources, crustal contam- no. 2, p. 132–134 (in Chinese). ination and sulphide segregation: Chemical Geology, v. 243, Zhao, L., Zhang, Z.C., Wang, F.S., Hao, Y.L., Ai, Y., and Yang, p. 317–343. T.Z., 2006, Open-system magma chamber: An example from Wilson, A., and Chunnett, G., 2006, Trace element and platinum the Xinjie mafic-ultramafic layered intrusion in Panxi region, group element distributions and the genesis of the Merensky SW China: Acta Petrologica Sinica, v. 22, no. 6, p. 1565–1578 Reef, Western Bushveld Complex, South Africa: Journal of (in Chinese with English abstract). Downloaded by [China University of Geosciences], [Mr Zhaochong Zhang] at 20:59 15 October 2012 Petrology, v. 47, no. 12, p. 2369–2403. Zhong, H., Zhou, X.H., Zhou, M.F., Sun, M., and Liu, B.G., 2002, Wilson, M., 1989, Igneous petrogenesis: A global tectonic Platinum-group element geochemistry of the Hongge Fe-V-Ti approach: London, Unwin Hyman, 466 p. deposit in the Panxi area, Southwestern China: Mineralium Wooden, J.L., Czamanske, G.K., Fedorenko, V.A., Arndt, N.T., Deposita, v. 37, p. 226–239. Chauvel, C., Bouse, R.M., King, B.S.W., Knight, R.J., and Zhong, H., and Zhu, W.G., 2006, Geochronology of layered Siems, D.F., 1993, Isotopic and trace-element constraints on mafic intrusions from the Pan-Xi area in the Emeishan mantle and crustal contributions to Siberian continental flood large igneous province, SW China: Mineral Deposita, v. 41, basalts, Noril’sk area, Siberia: Geochimica et Cosmochimica p. 599–606. Acta, v. 57, p. 3677–3704. Zhou, M.F., Arndt, N.T., Malpas, J., Wang, C.Y., and Kennedy, Workman, R.K., and Hart, S.R., 2005, Major and trace element A.K., 2008, Two magma series and associated ore deposit composition of the depleted MORB mantle (DMM): Earth types in the Permian Emeishan large igneous province, SW and Planetary Science Letters, v. 231, p. 53–72. China: Lithos, v. 103, p. 352–368. 1764 M. Wang et al.

Zhou, M.F., Malpas, J., Song, X.Y., Robinson, P.T., Sun, M., Zhou, M.F., Robinson, P.T., Lesher, C.M., Keays, R.R., Zhang, Kennedy, A.K., Lescher, C.M., and Keays, R.R., 2002, A C.Z., and Malphas, J., 2005, Geochemistry, petrogenesis and temporal link between the Emeishan large igneous province metallogenesis of the Panzhihua gabbroic layered intrusion (SW China) and the end-Guadalupian mass extinction: Earth and associated Fe-Ti-V oxide deposits, Sichuan Province, SW and Planetary Science Letters, v. 196, p. 113–122. China: Journal of Petrology, v. 46, p. 2253–2280. Downloaded by [China University of Geosciences], [Mr Zhaochong Zhang] at 20:59 15 October 2012