SCIENTIFIC COMMUNICATIONS

Transcription

1 2009 Society of Economic Geologists, Inc. Economic Geology, v. 104, pp SCIENTIFIC COMMUNICATIONS A NEW GENETIC MODEL FOR THE GIANT Ni-Cu-PGE SULFIDE DEPOSITS ASSOCIATED WITH THE SIBERIAN FLOOD BASALTS CHUSI LI, EDWARD M. RIPLEY, Department of Geological Sciences, Indiana University, Bloomington, Indiana AND ANTHONY J. NALDRETT Department of Geology, University of Toronto, Ontario, Canada M5S 3B1 Abstract The Kharaelakh intrusion is one of several sill-like, multiphase gabbroic intrusions that host world-class Ni- Cu-PGE deposits in the Noril sk-talnakh region. The sulfide ores of the Kharaelakh intrusion are characterized by elevated δ 34 S values (10 12 ) and high PGE concentrations (e.g., up to10 ppm Pt). The δ 34 S values require addition of crustal S with elevated isotope ratios such as evaporite-bearing country rocks (δ 34 S, ~20 ) and R factors (magma/sulfide mass ratios) of <400 during sulfide segregation, whereas the Pt concentration requires a much higher R factor (>2,000) if Pt content in the magma is similar to the values in the coeval, most undepleted lavas (~10 ppb). Such a discrepancy can be explained by sulfide resorption by a new flux of mantle-derived magma into a deep staging chamber to form PGE-enriched magma. Interaction of the PGE-enriched magma with evaporite-bearing country rocks in the plumbing system at higher crustal levels could have produced sulfide liquids with high PGE contents as well as elevated δ 34 S values. Mass-balance calculations indicate that <0.9 wt percent anhydrite assimilation from evaporite-bearing country rocks is required to explain the elevated δ 34 S values of the sulfide ores. This model, although developed purely based on data from the Kharaelakh deposit, can also be applied to other coeval Ni-Cu-PGE deposits. For the coeval Noril sk-i and Talnakh deposits different depths of anhydrite assimilation in the respective plumbing systems are required. This model differs from the previously proposed models in detail but reenforces the idea that chalcophile element depletion in continental flood basalts is a useful exploration tool for Ni-Cu-PGE deposits associated with coeval subvolcanic intrusions. Introduction The focus of current debate on the genesis of Ni-Cu-PGE sulfide ores in several subvolcanic intrusions, including the Kharaelakh intrusion in the Noril sk-talnakh region, Siberia, is on the contributions of chalcophile elements from the coeval flood basalts. A research group led by Naldrett (Naldrett, 1992; Naldrett et al., 1992, 1996; Naldrett and Lightfoot, 1999) proposed a conduit model that emphasizes a direct link between the ore-bearing intrusions and a specific suite of the coeval lavas (i.e., the Nadezhdinsky suite) that is depleted in chalcophile elements including Ni, Cu, Au, and platinumgroup elements (PGE; Brügmann et al., 1993; Lightfoot and Keays, 2005). The various versions of the conduit model have been criticized by other researchers because of differences between the intrusive and extrusive rocks shown by liquidus phase relationships (Latypov, 2002), S isotopes (Ripley et al., 2003), and Sr-Nd isotopes plus trace elements (Wooden et al., 1993; Czamanske et al., 1995; Arndt, 2005; Arndt et al., 2003, 2005). To explain some of these differences Arndt and coworkers (Arndt, 2005; Arndt et al., 2003, 2005) proposed that the ore-bearing intrusions were not the conduits of the chalcophile element-depleted lavas and that the depleted lavas lost their chalcophile elements to sulfide liquid that segregated and remained in a deep staging chamber, not to the Corresponding author: , cli@indiana.edu sulfide ores in the ore-bearing intrusions. The different genetic models render different implications for global mineral exploration. In this communication we propose a new genetic model that can explain the combination of elevated δ 34 S values and high PGE concentrations in the sulfide ores that have not been explained by the previous models, and the occurrence of magmatic anhydrite-sulfide assemblages in the Kharaelakh intrusion (Li et al., 2009). To avoid confusion that may arise from the presence of anhydrite of different origins in the intrusion, a detailed description of anhydrite textures will be given first. Geologic Background U-Pb zircon dating has indicated that the ore-bearing intrusions in the Noril sk-talnakh region, Siberia, are contemporaneous with the voluminous Late Permian continental flood basalts in the region (Kamo et al., 1996, 2003). The most important ore-bearing intrusions are the Noril sk-1, Talnakh, and Kharaelakh intrusions (Fig. 1a). The Kharaelakh intrusion was also called the northwestern Talnakh intrusion by some researchers (Naldrett, 1992; Naldrett et al., 1992, 1996; Li et al., 2003). All mine geologists now refer to it as the Kharaelakh intrusion. In both the Noril sk and Talnakh areas, weakly mineralized intrusions are also present. Although the ore-bearing (e.g., Kharaelakh) and weakly mineralized (e.g., Lower Talnakh) intrusions are spatially close to each other /09/3812/

2 292 SCIENTIFIC COMMUNICATIONS a Noril'sk RUSSIA 88 o E A Kharaelakh Intrusion projected to surface Basalt Kharaelakh Fault N Gabbroic dyke A' Soil 69.5 o N 5 km b km 0 A A' -0.5 Tungusskaya series sediments Devonian sediments Basalt Talnakh -1 Gabbro Picritic / taxitic gabbro with disseminated sulfide Massive sulfide Kharaelakh Lower Talnakh 1000 m Horizontal FIG. 1. (a). Simplified geologic map of the Talnakh region. (b). Cross section of the Kharaelakh ore-bearing intrusion. (Fig. 1a), their S isotope compositions as well as their radiogenic isotopes and trace elements are different. The mineralized intrusions have δ 34 S values from 6 to 12 per mil, whereas the weakly mineralized intrusions have δ 34 S values that are generally <6 per mil (Grinenko, 1985). In the Talnakh area the ore-bearing intrusions occur in both the Devonian evaporate-bearing sedimentary strata and the overlying coal-bearing sedimentary strata (Fig. 1b). In the Noril sk area all ore-bearing intrusions are above the Devonian evaporate-bearing sedimentary strata. The internal structures of the ore-bearing intrusions, including the Kharaelakh intrusion, have been described by Czamanske et al. (1995). These intrusions all comprise multiple rock units that are suggestive of multiple magma injections (Czamanske et al., 1995; Li et al., 2003). The studies of the overlying basaltic sequence by Brügmann et al. (1993) and Lightfoot and Keays (2005) indicated that the lower and middle parts of the Nadezhdinsky suite (Nd 1-2 ) are significantly depleted in chalcophile elements compared to the overlying Morongovsky suite. The Nd 1-2 lavas and the weakly mineralized intrusions, which are considered to have predated emplacement of the ore-bearing intrusions (Zen ko and Czamanske, 1994; Czamanske et al., 1995), are characterized by high Th/Nb and γ Os, and low ε Nd, suggesting that the assimilation of old granitoid crust or sedimentary rocks derived from old granitoid crust (Lightfoot et al., 1990, 1993; Wooden et al., 1993; Walker et al., 1994; Hawkesworth et al., 1995; Horan et al., 1995; Fedorenko et al., 1996, Arndt et al., 2003). The trace element ratios of the ore-bearing intrusions are significantly different from those of the Nd 1-2 lavas but broadly similar to those of the overlying Morongovsky lavas (Naldrett et al., 1992, 1996). However, the Morongovsky lavas and ore-bearing intrusions are different in estimated parental magma compositions (Latypov, 2002), 87 Sr/ 86 Sr, ε Nd (Arndt, 2005), and S isotopes (Ripley et al., 2003). Samples and Analytical Methods Two (K53, K58) of the samples used in this study are from drill core P5536 which is located in the eastern part of the Kharaelakh intrusion. The mineralogical and S isotope variations in this drill core are given in Li et al. (2003). Another sample (KH-2) is from drill cores M690 that is located in the western part of the intrusion. Sample K53 is a taxitic gabbro (Fig. 2a) and samples K58 and KH-2 (Fig. 2c) are picritic gabbros. Taxitic and picritic gabbros are Russian terms; they correspond to variable-textured and olivine gabbros in western /98/000/ $

3 SCIENTIFIC COMMUNICATIONS 293 a 1 cm c 1 b K53 1 cm Cp KM-1 KH-2 FIG. 2. Photographs of hand specimens showing hydrothermal anhydrite patches in altered gabbro from the Kharaelakh intrusion (a), hydrothermal intergrowth of anhydrite and chalcopyrite in the aureole of the Kharaelakh intrusion (b), and sulfide bleb-bearing picritic gabbro (olivine gabbro) from the Kharaelakh intrusion (c). = anhydrite, Cp = chalcopyrite. literature (e.g., Li et al., 2003). Sample KM-1 is from the aureole of the Kharaelakh intrusion at the Komsomolsky mine. ydrite in the samples was identified using transmitted and reflected light microscopy, followed by backscattered electron imagery and wavelength dispersive analysis using a CAMECA SX50 electron microprobe. Textures of ydrite There are three types of anhydrite in the Kharaelakh intrusion: (1) xenoliths, (2) hydrothermal, and (3) magmatic. The first two types were reported previously by Gorbachev and Grinenko (1973). ydrite-bearing evaporite xenoliths up to 0.5 m in thickness are most common in the western part of the intrusion. The second type is also called metasomatic anhydrite by many Russian geologists (e.g., Gorbachev and Grinenko, 1973). This type of anhydrite occurs within the intrusion as well as in the aureole of the intrusion (Fig. 2b). It is not known whether or not the different occurrences are related to the same hydrothermal activity. Magmatic anhydrite was not reported in the western literature until recently by Li et al. (2009). Hydrothermal anhydrite in the Kharaelakh intrusion occurs as irregular patches up to ~1 cm in diameter, commonly bounded by microfractures, and surrounded by secondary hydrous silicates (Fig. 2a). The hydrous silicates include hydrogrossular, pectolite, prehnite, thomsonite, and xonotlite. Their compositions determined by electron microprobe analysis are listed in Table 1. Minor calcite, pyrite, chalcopyrite, and magnetite are present within some of the secondary hydrous silicate assemblages. Magmatic anhydrite is hard to recognize in hand specimens due to its small size and similar color to plagioclase which is /98/000/ $

4 294 SCIENTIFIC COMMUNICATIONS TABLE 1. Compositions of Secondary Silicates in the Kharaelakh Intrusion Mineral Hydrograssular Pectolite Prehnite Thomsonite Xonotlite SiO TiO Al 2O MgO CaO MnO FeO Na 2O K 2O H 2O 1 not calc Total Cation 24O 24O 24O 80O 24O Si Ti Al Mg Ca Mn Fe Na K Total Calculated by stoichiometry ubiquitous in the rocks. For example, in a picritic gabbro (sample KH-2) from the Kharaelakh intrusion, rare magmatic anhydrite crystals are found in thin sections under microscopic observation but are not visible in split drill core specimens (Fig. 2c). Li et al. (2009) reported the typical textures of magmatic anhydrite in the Kharaelakh intrusion such as planar grain boundaries between anhydrite, olivine, and augite, and inclusions of anhydrite in augite and vice versa. Here we report the textures of magmatic anhydrite in association with magmatic sulfides in the Kharaelakh intrusion. Figure 3a and b illustrates a typical magmatic anhydrite-sulfide assemblage in a thin section under both reflected and transmitted light. In this assemblage anhydrite crystals occur as randomly orientated crystals and form a framework; sulfides (pyrrhotite, pentlandite, and chalcopyrite) occur in the interstices. Elongated anhydrite crystals at the margins of the assemblage have sharp contacts with granular augite crystals. In some cases sulfides in the assemblage are completely isolated from silicate minerals by anhydrite crystals (Fig. 3c). Figure 3d through f illustrates the textures of hydrothermal anhydrite assemblages in the Kharaelakh intrusion. In relatively large alteration patches anhydrite commonly occurs as needles with other secondary hydrous silicates and calcite (Fig. 3d). In relatively small alteration patches finer grained anhydrite crystals occur intergrown with secondary hydrous silicates (Fig. 3e). In this type of assemblage backscattered electron image (Fig. 3f) and energy dispersive X-ray spectrum are needed for mineral identification. Origin of High δ 34 S Values With the exception of two sulfide-poor gabbroic samples (<500 ppm S) which have δ 34 S values between 1.5 and 3.6 per mil, all sulfide-mineralized samples from the Kharaelakh intrusion analyzed to date yield δ 34 S values from 10 to 12 per mil (Grinenko, 1985; Li et al., 2003, 2009). The positive correlation between S content and δ 34 S value indicates a link between sulfide saturation-segregation and contamination with 34 S-enriched S. However, some Russian geologists (e.g., Godlevsky and Likachev, 1986) have proposed that the high δ 34 S values are of mantle origin rather than being related to crustal contamination. Anomalously high δ 34 S values have not been found in the coeval basalts to date (Ripley et al., 2003). The δ 34 S values in the basalts are similar to those of the two sulfide-poor gabbroic samples from the Kharaelakh intrusion. The significant difference between the basalts and the sulfide-mineralized intrusive rocks is inconsistent with the hypothesis of anomalously high δ 34 S values in the mantle beneath the Siberian platform. Degassing may affect S isotope compositions in volcanic and subvolcanic systems (e.g., Mandeville et al., 2009). Ripley et al. (2003) discussed the possible effects of degassing on the S isotope compositions of volcanic rocks in the Noril sk area. During degassing the S isotope compositions of magma is critically dependent on the temperature, pressure, the mass of the vapor lost, the composition of the vapor, and the f O2 condition which controls speciation in the melt and vapor. At f O2 conditions where sulfate is the predominant species in the magma the loss of a vapor can lead to 34 S enrichment in the magma. However, Ripley et al. (2003) showed that the vast majority of volcanic rocks in the Noril sk region are characterized by δ 34 S values less than 5 per mil. Abundant sulfide ores in the Kharaelakh intrusion indicate that the ore-bearing intrusive rocks have certainly not degassed to the extent of the relatively low δ 34 S coeval lavas. It is therefore extremely unlikely that the elevated δ 34 S values found in the sulfide ores of the Kharaelakh intrusion are a result of degassing. Grinenko (1985) proposed assimilation of sulfur in sour gas as the cause of elevated δ 34 S values in the sulfide ore-bearing /98/000/ $

5 SCIENTIFIC COMMUNICATIONS 295 c Reflected light (-) Sample K58 Transmitted light (+) Sample K58 d Transmitted light (+) Sample K58 Xonotlite Hgs Calcite 1 mm Transmitted light (+) Sample K53 e Hgs Cpx Pl f Back-scattered electron image Prehnite Pectolite Cpx 0.5 mm Hgs Transmitted light (+) Sample K mm FIG. 3. Microphotographs of magmatic and hydrothermal anhydrites in the Kharaelakh intrusion. (a), (b), and (c). Magmatic anhydrite-sulfide assemblages. (d), (e), and (f). Hydrothermal anhydrite. = anhydrite, Cp = chalcopyrite, Cpx = clinopyroxene, Hgs = hydrograssular, Pn = pentlandite, Po = pyrrhotite. intrusions in the Noril sk region. However, sulfur in sour gas commonly occurs as reduced species such as H 2 S. It is extremely unlikely that addition of reduced S species into a mantle-derived magma can drive magma to sulfate saturation, as indicated by the presence of magmatic anhydrite in the Kharaelakh intrusion (Li et al., 2009). Therefore, we support the hypothesis made previously by many other authors (e.g., Naldrett et al., 1992, 1996; Li et al., 2003, 2009) that assimilation of anhydrite-bearing evaporite country rocks is responsible for the elevated δ 34 S values of sulfide ores in the intrusion. Because the melting temperature of anhydrite (>1,400ºC) is significantly higher than the temperature of a basaltic magma (1,200º 1,300ºC), it is unlikely that the magma of the Kharaelakh intrusion could have assimilated anhydrite by melting. Chemical dissolution by magma that was originally unsaturated with anhydrite is a more plausible mechanism /98/000/ $

6 (δ 34 S = 20 ) 296 SCIENTIFIC COMMUNICATIONS Mass-Balance Constraints The δ 34 S value of sulfide ores in the Kharaelakh intrusion vary mostly between 10 and 12 per mil (Li et al., 2003). The δ 34 S values of anhydrite country rocks in the region are ~20 per mil (Gorbachev and Grinenko, 1973). The content of sulfur in mantle-derived magmas in continental rift settings such as the Deccan magmas is estimated to be ~0.14 wt percent based on the analysis of melt inclusions in olivine phenocrysts (Self et al., 2008). To produce a contaminated magma with a δ 34 S value of 12 per mil, the magma needs to assimilate 0.89 wt percent anhydrite if the content of S in the uncontaminated magma is 0.14 wt percent and the δ 34 S value of the uncontaminated magma is 0 per mil. The amount of anhydrite assimilated increases to 0.96 wt percent if the content of S in the uncontaminated magma is 0.15 wt percent. The relationship between δ 34 S values in contaminated magmas and the amounts of anhydrite assimilated by the magmas are illustrated in Figure 4. ydrite assimilation by magma will increase f O2 in the magma through the following equilibria: and CaSO 4 solid CaSO 4 melt, (1) CaSO 4 melt + 9FeO melt FeS melt + 4Fe 2 O 3 melt + CaO melt. (2) Equation (2) can be used to calculate the increase of Fe 2 O 3 /FeO in the magma due to anhydrite assimilation. The increase of f O2 in the contaminated magma can be calculated using the MELTS program of Ghiorso and Sack (1995) in which the f O2 -(Fe 2 O 3 /FeO)-T relationship is implemented. Figure 5 illustrates the change of f O2 due to anhydrite assimilation by a magma with composition similar to the average composition of the Noril sk-type sills (Zen ko and Czamanske, 1994) and initial f O2 at 2 log units below the buffer of fayalitemagnetite-quartz (i.e., FMQ-2). The assumed original oxidation state is within the range of midocean ridge basalts wt% anhydrite assimilated δ 34 S = 12 Contaminated magma δ 34 S = 10 δ 34 S = wt% S in initial magma (δ 34 S = 0 ) FIG. 4. Two-component mixing calculation showing the produced δ 34 S value, the amount of anhydrite assimilation, and the initial content of S in magma. Oxidation state of contaminated magma ( FMQ) Oxidation state of initial magma: FMQ wt% anhydrite assimilated FIG. 5. Relationship between change in oxidation state of a basaltic magma initially at FMQ-2 and the amount of anhydrite assimilated by the magma based on reaction (2) and the MELTS program of Ghiorso and Sack (1995) in which the f O2 -(FeO/Fe 2O 3)-T relationship is implemented. (Christie et al., 1986) and the Hawaiian ocean island basalts (Rhodes and Vollinger, 2005). Our calculations indicate that f O2 in the magma will increase from FMQ-2 to FMQ+1.5 after 0.9 wt percent anhydrite assimilation (Fig. 5). The redox changes due to anhydrite assimilation are likely maximum values because in addition to FeO, potential organic matter from country rocks (Arndt et al., 2005; Jugo and Lesher, 2005) may also serve as reduction agents. In addition to total pressure, temperature, and composition, the content of S dissolved in silicate melts at S saturation depends on the oxidation state because that controls sulfur speciation. Results from measured wavelength shifts of sulfur Kα X-rays (Carroll and Rutherford, 1988; Wallace and Carmichael, 1994) indicate that under reducing conditions (< FMQ+1.5) S is dissolved predominantly as S 2 (>90% of total S), under oxidizing condition (> FMQ+2) S is dissolved predominantly as S 6+ (>90% of total S), and under transitional oxidation states (between ~ FMQ+1.5 and ~ FMQ+2) both S 2 and S 6+ species are important in silicate melts. At S saturation, the average concentration of sulfur dissolved in a basaltic magma at 1,300ºC and 10 kbars total pressure is 0.14 wt percent under reducing conditions (line a-b, Fig. 6) and 1.45 wt percent under oxidizing conditions (line c-d, Fig. 6; Jugo et al., 2005). These values do not vary with f O2 if other parameters remain unchanged (Jugo et al., 2005). Incidentally, the former is the same as the value estimated for the parental magma of the Deccan Traps (Self et al., 2008) as well as a magma with composition similar to the average composition of the Noril sk-type sills (Zen ko and Czamanske, 1994) under the conditions of 1 kb and FMQ+1.2, calculated using the empirical equation of Li and Ripley (2005) in conjunction with the MELTS program of Ghiorso and Sack (1995) for the estimation of liquid temperature. Due to lack of experiments, little is known about the contents of S in magma at S saturation under the transitional oxidation states. The dashed line b- c in Figure 6 illustrates a situation where the content of S in magma at S saturation under the transitional oxidation states is assumed to increase with increasing f O2. Line e-h in the /98/000/ $

7 SCIENTIFIC COMMUNICATIONS 297 wt% S a S content in hydrous felsic magma at S-saturation o C, 2 kb (Carroll and Rutherford, 1987) S content in anhydrous mafic magma at S-saturation o 1300 C, 10 kb (Jugo et al., 2005) e Reducing Oxidation state, FMQ Oxidizing figure represents the increase of total S with increasing amounts of anhydrite assimilation by a magma with initial S content of 0.14 wt percent and oxidation state of FMQ-2. During the early stages of anhydrite assimilation (from point e-g), the contaminated magma becomes sulfide saturated, and immiscible sulfides develop within it. However, further assimilation of anhydrite (after point g) drives the contaminated magma below S saturation. The S saturation history of the contaminated magma can be further divided into two parts: the first part from point e to f during which the R factor decreases with increasing anhydrite assimilation, and the second part from point f to g during which the R factor increases with increasing anhydrite assimilation due to decreasing amount of immiscible sulfide liquid. The results of massbalance calculations (see Figs. 4, 5) indicate that after point f the δ 34 S values of the contaminated magmas are higher than 12 per mil. Since the majority of the sulfide ores in the Kharaelakh intrusion and other ore-bearing intrusions in the region have δ 34 S values below this value (Grinenko, 1985; Li et al., 2003), we envision that the process from point f to g was not important for the formation of the deposits. The R factors from point e to f for both open and closed systems have been calculated and are illustrated in Figure 7. In the open system case the contaminated magma is withdrawn from the system after segregation of immiscible sulfide liquid due to anhydrite assimilation. In the closed system case no magma leaves the system after anhydrite assimilation. The R factors for these b c f g h FIG. 6. Consequences of increasing anhydrite assimilation by a basaltic magma initially at FMQ-2 and with 0.14 wt percent S. The increase of total S (in both dissolved and immiscible forms) in the magma due to anhydrite assimilation is indicated by curve e-h. Line a-b represents the average content of S in anhydrous mafic magma at sulfide saturation determined by Jugo et al. (2005); line c-d represents the average content of S in anhydrous mafic magma at anhydrite saturation (Jugo et al., 2005); and curve b-c represents the estimated content of S in magma saturated with both sulfide and anhydrite. Compared to anhydrous mafic magma, the contents of S in hydrous felsic magma at S saturation determined by Carroll and Rutherford (1987) are significantly lower mainly due to lower FeO content and temperature. Both systems indicate that the content of S in magma is drastically higher at sulfate saturation than at sulfide saturation. Transitional d R-factor Open system Closed system δ 34 S ( ) in contaminated magma FIG. 7. Relationships between R factor (magma/sulfide mass ratio) and δ 34 S value of contaminated magma after anhydrite assimilation. The solubility of S at sulfide saturation in the magma is assumed to be 0.14 wt percent. In the open system case the magma was completely withdrawn from the system after sulfide segregation. In the closed system case no magma leaves the system after anhydrite assimilation. two different systems vary from 100 to 400. These values are likely the maximum values because the content of S in the magma before anhydrite assimilation may be higher than the assumed value of 0.14 wt percent due to higher FeO content. For example, the sulfur content at sulfide saturation in a magma with composition similar to the average composition of the Noril sk-type sills (Zen ko and Czamanske, 1994) under the conditions of 4 kbars and FMQ-2, calculated using the empirical equation of Li and Ripley (2005) in conjunction with the MELTS program of Ghiorso and Sack (1995) for the estimation of liquid temperature, is 0.16 wt percent. The R factors of 100 to 400 (Fig. 7) are one order of magnitude lower than R factors estimated based on the Pt tenors of the sulfide ores in the Kharaelakh intrusion and initial magma Pt content similar to the value in the coeval, most undepleted lavas (~10 ppb Pt; Naldrett et al., 1996). The Pt tenors in the sulfide ores in the Kharaelakh intrusion vary between 2 and 8 ppm (Naldrett et al., 1996). To form the sulfide liquids with such high Pt concentrations and R factors of <400 the parental magmas must contain Pt >60 ppb (Fig. 8), which is six times the value of the coeval, most undepleted lavas that have been analyzed to date (Brügmann et al., 1993; Lightfoot and Keays, 2005). Sulfide Resorption The most effective way to form a PGE-enriched magma is by resorption of PGE-rich sulfide liquid (Kerr and Leitch, 2005). Among the lavas that are coeval with the Kharaelakh intrusion, the Nd 1-2 lavas are most depleted in PGE, an indication of sulfide segregation at depth (Brügmann et al., 1993; Lightfoot and Keays, 2005). The Nd 1-2 lavas are also characterized by high La/Sm, high SiO 2, and low ε Nd, which indicate contamination with granitoid crust (Lightfoot et al., 1993; Lightfoot and Hawkesworth, 1997). Sulfide saturation in these lavas at depth was thought to be due to contamination with granitic crustal materials in a deep staging chamber /98/000/ $

8 298 SCIENTIFIC COMMUNICATIONS ppb Pt in initial magma Sulfide liquid 8 ppm Pt 6 ppm Pt 4 ppm Pt 2 ppm Pt Parameters Batch #1 Batch #4 Batch #8 Pt D of sulf-liq/magma 20,000 Pt in initial sulf-liq. 200 ppm Mass of initial sulf-liq. 100 g Pt in new magma 10 ppb Each batch of magma 10,000 g Pt = 11 ppb Pt = 16 ppb Pt = 42 ppb R-factor FIG. 8. Relationships between Pt content in sulfide liquid, R factor, and Pt content in initial magma. perhaps in the lower parts of the upper crust (Lightfoot et al., 1993; Lightfoot and Hawkesworth, 1997; Arndt et al., 2005). The results of numerical modeling by Lightfoot and Keays (2005) suggest that the overlying Morongovsky lavas may have equilibrated with sulfide liquid with Pt content in excess of 200 ppm. When a new, S-unsaturated magma from the mantle entered the staging chamber, it would have dissolved FeS from the sulfide liquids. Nickel, Cu, and especially PGE would have stayed with the remaining sulfide liquid because of their much higher sulfide liquid/magma partition coefficients relative to Fe. As magma replenishment continued, the sulfide liquid would have become progressively more PGE (and Ni and Cu) enriched. Since the magma flowing through the chamber remained in equilibrium with the sulfide liquid present there, it would also have become PGE enriched, eventually acquiring a much higher PGE content than the original magma. As an example, Figure 9 illustrates the results of such a process based on a two-component mass balance on Pt and the parameters given in the figure. In our mass-balance calculations we used the maximum sulfide/ magma D Pt given by Stone et al. (1990) and Fleet et al. (1999). The initial Pt content in the sulfide liquid is within the range of values for the sulfide liquids in equilibrium with the Morongovsky magma given by Lightfoot and Keays (2005). The Pt content in the mantle-derived, S-unsaturated magma was fixed at 10 ppb and is similar to the values in the relatively PGE undepleted lavas above the Nadezhdinsky suite (Lightfoot and Keays, 2005). Our calculations indicate that Pt content in the magma will exceed 60 ppb after nearly complete resorption of sulfide liquid in the staging chamber. A New Model In light of the mass-balance constraints given above, we propose a new model for the formation of Ni-Cu-PGE sulfide ores in the Kharaelakh intrusion. The key aspects of the model are illustrated in Figure 10 and summarized here. Early sulfide segregation took place in a deep staging chamber due to contamination with granitic crustal materials in the lower parts of the upper crust, as suggested previously by S in new magma is 350 ppm below S content at sulfide saturation in the magma 90.4% sulf-liq. remaining Pt = 221 ppm 61.5% sulf-liq. remaining Pt = 323 ppm 23% sulf-liq. remaining Pt = 831 ppm FIG. 9. S-unsaturated magma interacting with sulfide liquids in a conduit or sill would dissolve FeS from the sulfide liquids. Nickel, Cu, and especially PGE would stay with the remaining sulfide liquids because of their much higher sulfide liquid/magma partition coefficients relative to Fe. Repetitive batches of magma passing through a system, dissolving out FeS, would steadily upgrade the Ni, Cu, and PGE contents of the residual sulfide liquid. As the sulfide liquids became more PGE enriched, the flowing magma itself would also become PGE enriched, eventually having a much higher PGE content than the original magma. other researchers (Naldrett et al., 1992, 1996; Naldrett and Lightfoot, 1999; Arndt et al., 2003, 2005). The magma then rose to form the weakly mineralized intrusions and erupted to the surface to form the Nd 1-2 lavas, leaving a sulfide liquid with relatively low tenors of Ni, Cu, and PGE in the staging chamber. The PGE-poor sulfide liquid in the chamber was then upgraded in chalcophile elements (PGE, Ni, and Cu) by the Morongovsky magma from the mantle to form a PGErich sulfide liquid. The PGE-rich sulfide liquid remained in the staging chamber while the Morongovsky magma erupted to form the Morongovsky lavas (Fig. 10a). New, S-unsaturated magma from the mantle continued to enter the chamber, progressively dissolved the PGE-rich sulfide liquid in the chamber to form a PGE-enriched magma. The PGE-enriched magma then rose to the upper parts of the upper crust where it reacted with anhydrite-bearing evaporite country rocks and became sulfide saturated, thereby producing immiscible sulfide liquid with high PGE concentrations as well as high δ 34 S values (Fig. 10b). The sulfide liquid became lodged in the hydraulic traps of the plumbing system at Kharaelakh (Fig. 10c). According to Zen ko and Czamanske (1994), the weighted average sulfide content (calculated as FeS) in the Noril sk ore-bearing intrusion and its peripheral sills intercepted by drilling to date is ~0.17 wt percent, which is within the values for the closed system case described above. The weighted average sulfide content in the Kharaelakh-Talnakh ore-bearing intrusions and their peripheral sills intercepted by drilling to date is ~1.6 wt percent, which is 60 percent higher than the maximum value for the closed system case (~1 wt %) described above. In other words, ~40 percent of the magma /98/000/ $

9 SCIENTIFIC COMMUNICATIONS 299 a Mr lavas (slightly depleted in PGE) b Nd1-2 lavas (highly depleted in PGE) ydrite-bearing sedimentary rocks Ore-barren intrusion (e.g., Lower Talnakh) Ore-bearing intrusion (e.g., Kharaelakh) Magma Magma PGE-depleted magma PGE-enriched magma Magma Immiscible sulfide liquid segregation due to assimilation of granitoid crust Magma Sulfide resorption by S unsaturated magma c Sulfide saturation due to assimilation of anhydrite-bearing evaporites PGE-enriched magma Exit to peripheral sills Kharaelakh subvolcanic chamber Sulfide liquid injection into footwall FIG. 10. A new model for the formation of Ni-Cu-PGE sulfide ores in the Kharaelakh intrusion. The sulfide liquid, segregated in a deep staging chamber from the Nd 1-2 magmas, was upgraded by the Morongovsky magma (a), and then dissolved by a new, S-unsaturated magma from the mantle to form a PGE-enriched magma (b). Reaction of the PGE-enriched magma with anhydrite-bearing evaporite country rocks at a higher level produced immiscible sulfide liquids with high PGE tenors as well as elevated δ 34 S values. The sulfide liquids became lodged in the hydraulic traps of the plumbing system at Kharaelakh to form the deposit (c). involved in the formation of the Kharaelakh-Talnakh deposits has not been accounted for by the host intrusions and their peripheral sills intercepted by drilling to date. It is possible that more peripheral sills in the Kharaelakh-Talnakh oreforming system may have not been found due to a lack of drilling beyond the orebodies. Alternatively, the missing magma with elevated δ 34 S values may have erupted to the surface but have not been identified due to insufficient sampling. However, the difference in the estimated parental magma compositions between the intrusive and extrusive rocks (Latypov, 2002) makes the second scenario less likely. Although the above mismatch may be explained by shallowlevel processes such as contamination and fractional crystallization that took place after the magma left the intrusions /98/000/ $

10 300 SCIENTIFIC COMMUNICATIONS and flowed onward toward the surface, the shallow-level processes cannot explain lower Rb/Sr, 87 Sr/ 86 Sr, and ε Nd in the erupted lavas than in the ore-bearing intrusions (Arndt, 2005). With slight modification our model can also be applied to other Ni-Cu-PGE deposits in the region, including the Noril sk-i deposit which occurs several hundred meters above the evaporite strata. In this case anhydrite assimilation needed to take place before the magma reached the hydraulic trap of the plumbing system at Noril sk-i. The sulfide ores of the Noril sk-i deposit generally have higher Pt tenor (up to 12 ppm: Naldrett et al., 1996) and lower δ 34 S values (8 10 : Li et al., 2003) than the Kharaelakh deposit, which is consistent with less amounts of anhydrite assimilation and thereby higher R factors for the resultant sulfide liquids. Concluding Remarks Our new model adequately explains why the Nd 1-2 lavas, which provided much of the PGE to the sulfide ores in the ore-bearing intrusions, have S isotopes, Sr-Nd isotopes, and trace element ratios different from the ore-bearing intrusions (Lightfoot et al., 1990, 1993; Wooden et al., 1993; Arndt et al., 2003, 2005; Ripley et al., 2003). This is because the Nd 1-2 lavas and the ore-bearing intrusions were not directly linked (Arndt, 2005; Arndt et al., 2003, 2005). Chalcophile element depletion in flood basalts as an exploration tool for important Ni-Cu-PGE deposits in coeval subvolcanic intrusions, which was proposed previously by Naldrett and coworkers (Naldrett, 1992; Naldrett and Lightfoot, 1999) based on different approaches, remains valid according to our new model. Acknowledgments We thank Viktor Radko of the Noril sk Nickel Company and the chief geologist of the Komsomolsky mine for supplying some of the samples used in this study, and Nick Arndt for thoughtful review of the manuscript. This research was supported by the Natural Science Foundation of China ( ), Project-111 from the Ministry of Education of China (B07011), and the Natural Science Foundation of the United States (EAR ). December 3, 2008; February 13, 2009 REFERENCES Arndt, N.T., 2005, The conduits of magmatic ore deposits: Mineralogical Association of Canada Short Course 35, p Arndt, N.T., Czamanske, G.K., Waker, R.J., Chavel, C., and Fedorenko, V.A.,2003, Geochemistry and origin of the intrusive hosts of the Noril sk-talnakh Cu-Ni-PGE sulfide deposits: ECONOMIC GEOLOGY, v. 98, p Arndt, N.T., Lesher, C.M., and Czamanske, G.K., 2005, Mantle-derived magmas and magmatic Ni-Cu-(PGE) deposits: ECONOMIC GEOLOGY 100TH AN- NIVERSARY VOLUME, p Brügmann, G.E., Naldrett, A.J., Asif, M., Lightfoot, P.C., Gorbachev, N.S., and Fedorenko, V.A., 1993, Siderophile and chalcophile metals as tracers of the evolution of the Siberian trap in the Noril sk region, Russia: Geochimica et Cosmochimica Acta, v. 57, p Carroll, M.R., and Rutherford, M.J., 1988, Sulfur speciation in hydrous experimental glasses of varying oxidation state: Results from measured wavelength shifts of sulfur X-rays: American Mineralogist, v. 73, p Christie, D.M., Carmichael, I.S.E., and Langmuir, C.H., 1986, Oxidation states of mid-ocean ridge basalt glasses: Earth and Planetary Science Letters, v. 79, p Czamanske, G.K., Zen ko, T.E., Fedorenko, V.A., Calk, L.C., Budahn, J.R., Bullock, J.H., Jr., Fries, TL., King, B.S.W., and Siems, D.F., 1995, Petrographic and geochemical characterization of ore-bearing intrusions of the Noril sk type, Siberia: With discussion of their origin: Resource Geology Special Issue 18, p Fedorenko, V.A., Lightfoot, P.C., Naldrett, A.J., Czamanske, G.K., Hawkesworth, C.J., Wooden, J.L., and Ebel, D., 1996, Petrogenesis of the flood-basalt sequence at Noril sk, north central Siberia: International Geology Review, v. 38, p Fleet, M.E., Crocket, J.H., Liu, M., and Stone, W.E., 1999, Laboratory partitioning of platinum-group elements (PGE) and gold with application to magmatic sulfide-pge deposits: Lithos, v. 47, p Ghiorso, M.S., and Sack, R.O., 1995, Chemical mass transfer in magmatic processes. IV. A revised and internally consistent thermodynamic model for the interpolation and extrapolation of liquid-solid equilibria in magmatic systems at elevated temperatures and pressures: Contributions to Mineralogy and Petrology, v. 119, p Godlevsky, M.N., and Likhachev, A.P., 1986, Types and distinctive features of ore-bearing formations of copper-nickel deposits, in Friedrich, G.H., Gerkin, A.D., Naldrett, A.J., Ridge, J.D., Sillitoe, R.H., and Vokes, F.M., eds., Geology and metallogeny of copper deposits: Berlin, Springer-Verlag, p Gorbachev, N.S., and Grinenko, L.N., 1973, The sulfur-isotope ratios of the sulfides and sulfates of the Oktyabr sky sulfide deposit, Noril sk region, and the problem of its origin: Geokhimiya, v. 8, p (in Russian) Grinenko, L.N., 1985, Sources of sulfur of the nickeliferous and barren gabbro-dolerite intrusions of the northwest Siberian platform: International Geology Review, v. 28, p Hawkesworth, C.J., Lightfoot, P.C., Blake, S., Naldrett, A.J., Doherty, W., Fedorenko, V.A., and Gorbachev, N.S., 1995, Magma differentiation and mineralisation in the Siberian Continental flood basalts: Lithos, v. 34, p Horan, M.F., Walker, R.J., Fedorenko, V.A., and Czamanske, G.K., 1995, Os and Nd isotopic constraints on the temporal and spatial evolution of flood basalt sources, Siberia: Geochimica et Cosmochimica Acta, v. 59, p Jugo, P.J., and Lesher, C.M., 2005, Redox changes caused by evaporite and carbon assimilation at Noril sk and their role in sulfide precipitation [abs.]: Geological Society of America Abstracts with Programs, v. 39, no. 6, p Jugo, P.J., Luth, R.W., and Richards, J.P., 2005, An experimental study of the sulfur content in basaltic melts saturated with immiscible sulfide or sulfate liquids at 1300ºC and 1.0 GPa: Journal of Petrology, v. 46, p Kamo, S.L., Czamanske, G.K., and Krogh, T.E., 1996, A minimum U-Pb age for Siberian flood-basalt volcanism: Geochimica et Cosmochimica Acta, v. 60, p Kamo, S.L., Czamanske, G.K., Amelin, Y., Fedorenko, V.A., Davis, D.W., and Trofimov, V.R., 2003, Rapid eruption of Siberian flood-volcanic rocks and evidence for coincidence with the Permian-Triassic boundary and mass extinction at 251 Ma: Earth and Planetary Science Letters, v. 214, p Kerr, A., and Leitch, A.M., 2005, Self-destructive sulfide segregation systems and the formation of high-grade magmatic ore deposits: ECONOMIC GEOL- OGY, v. 100, p Latypov, R.M., 2002, Phase equilibria constraints on relations of ore-bearing intrusions with flood basalts in the Noril sk region, Russia: Contributions to Mineralogy and Petrology, v. 143, p Li, C., and Ripley, E.M., 2005, Empirical equations to predict the sulfur content of mafic magma at sulfide saturation and applications to magmatic sulfide deposits: Mineralium Deposita, v. 40, p Li, C., Ripley, E.M., and Naldrett, A.J., 2003, Compositional variations of olivine and sulfur isotopes in the Noril sk and Talnakh intrusions, Siberia: Implications for ore forming processes in dynamic magma conduits: ECO- NOMIC GEOLOGY, v. 98, p Li, C., Ripley, E.M., Naldrett, A.J., Schmitt, A.K., and Moore, C.H., 2009, Magmatic anhydrite-sulfide assemblages in the plumbing system of the Siberian traps: Geology DOI: /G25355A. Lightfoot, P.C., and Hawkesworth, C.J., 1997, Flood basalts and magmatic Ni, Cu, and PGE sulfide mineralization: Comparative geochemistry of the Noril sk (Siberian traps) and West Greenland sequences: American Geophysical Union Monograph 100, p Lightfoot, P.C., and Keays, R.R., 2005, Siderophile and chalcophile metal variations in flood basalts from the Siberian trap, Noril sk region: Implications for the origin of the Ni-Cu-PGE sulfide ores. ECONOMIC GEOLOGY, v. 100, p Lightfoot, P.C., Naldrett, A.J., Gorbachev, N.S., Doherty, W, and Fedorenko, V.A., 1990, Geochemistry of the Siberian trap of the Noril sk area, USSR, with implications for the relative contributions of crust and mantle to flood /98/000/ $

11 SCIENTIFIC COMMUNICATIONS 301 basalt magmatism: Contributions to Mineralogy and Petrology, v. 104, p Lightfoot, P.C., Hawkesworth, C.J., Hergt, J., Naldrett, A.J., Gorbachev, N.S., Fedorenko, V.A., and Doherty, W., 1993, Remobilisation of continental lithosphere by mantle plumes: Major, trace element, and Sr-, Nd-, and Pb-isotope evidence for picritic and tholeiitic lavas of the Noril sk district, Siberian trap, Russia: Contributions to Mineralogy and Petrology, v. 114, p Mandeville, C.W., Webster, J.D., Tappen, C., Taylor, B.E., Timbal, A., Sasak, A., Hauri, E., and Bacon, C.R., 2009, Stable isotope and petrologic evidence for open-system degassing during the climatic and pre-climatic eruptions of Mt. Mazama, Crater Lake, Oregon: Geochimica et Cosmochimica Acta, DOI: /j.gca Naldrett, A.J., 1992, A model for the Ni-Cu-PGE ores of the Noril sk region and its application to other areas of flood basalts: ECONOMIC GEOLOGY, v. 87, p Naldrett, A.J., and Lightfoot, P.C., 1999, Ni-Cu-PGE deposits of the Noril sk region, Siberia: Their formation in conduits for flood basalt volcanism: Geological Association of Canada Short Course Notes 13, p Naldrett, A.J., Lightfoot, P.C., Fedorenko, V.A., Gorbachev, N.S., and Doherty, W., 1992, Geology and geochemistry of intrusions and flood basalts of the Noril sk region, USSR, with implications for the origin of the Ni-Cu ores: ECONOMIC GEOLOGY, v. 87, p Naldrett, A.J., Fedorenko, V.A., Asif, M., Lin, S., Kunilov, V.A., Stekhin, A.I., Lightfoot, P.C., and Gorbachev, N.S., 1996, Controls on the compositions of Ni-Cu sulfide deposits as illustrated by those at Noril sk, Siberia: ECO- NOMIC GEOLOGY, v. 91, p Ripley, E.M., Lightfoot, P.C., Li, C., and Elswick, E.R., 2003, Sulfur isotopic studies of continental flood basalts in the Noril sk region: Implications for the association between lavas and ore-bearing intrusions: Geochimica et Cosmochimica Acta, v. 67, p Rhodes, J.M., and Vollinger, M.J., 2005, Ferric/ferrous ratios in 1984 Mauna Loa lavas: A contribution to understanding the oxidation state of Hawaiian magmas: Contributions to Mineralogy and Petrology, v. 149, p Self, S., Blake, S., Sharma, K., Widdowson, M., and Sephton, S., 2008, Sulfur and chlorine in Late Cretaceous Deccan magmas and eruptive gas release: Science, v. 319, p Stone, W.E., Crocket, J.H., and Fleet, M.E., 1990, Partitioning of palladium, iridium, platinum and gold between sulfide liquid and basaltic melt at 1200ºC: Geochimica et Cosmochimica Acta, v. 54, p Walker, R.J., Morgan, J.W., Horan, M.F., Czamanske, G.K., Krogstad, E.J., Likhachev, A.P., and Kunilov, V.E., 1994, Re-Os isotopic evidence for an enriched-mantle plume source for the Noril sk-type ore-bearing intrusions, Siberia: Geochimica et Cosmochimica Acta, v. 58, p Wallace, P., and Carmichael, I.S.E., 1994, S speciation in submarine basaltic glasses as determined by measurement of SKα X-ray wavelength shifts: American Mineralogist, v. 79, p Wooden, J.L., Czamanske, G.K., Fedorenko, V.A., Arndt, N.T., Chauvel, C., Bouse, R.M., King, B.S.W., Knight, R.J., and Siems, D.F., 1993, Isotopic and trace element constraints on mantle and crustal contributions to characterization of Siberian continental flood basalts, Noril sk area, Siberia: Geochimica et Cosmochimica Acta, v. 57, p Zen ko, T.E., and Czamanske, G.K., 1994, Tectonic controls on ore-bearing intrusions of the Talnakh ore junction: position, morphology, and distribution: International Geology Review, v. 36, p /98/000/ $

12 /98/000/ $