Sulﬁde Rb-Sr, Re-Os and In Situ S Isotopic Constraints on Two Mineralization Events at the Large Hongnipo Cu Deposit, SW China

: The Hongnipo deposit, a newly discovered large copper deposit in the Kangdian copper belt, SW China, is hosted in the Paleoproterozoic Hekou Group. This deposit contains ~4200 Mt of Cu ores, with an average grade of 1.42 wt.% Cu. Orebodies occur mainly as stratiform, stratoid and lenticular forms. Mineralization consists predominantly of high grade vein-type ores and low grade laminated ores. Field relationships indicate vein-type mineralization crosscuts laminated mineralization and host rocks, indicating that there were at least two mineralization events during the formation of the deposit. Pyrite separates from the laminated ores yield a Rb-Sr isochron age of 1552 ± 80 Ma, with a highly radiogenic initial 87 Sr / 86 Sr ratio of 0.71214 ± 0.00081, indicating a major contribution from the ore-hosting rocks. Sulﬁdes from the laminated ores have δ 34 S values ranging from − 1.8% (cid:24) to 11.4% (cid:24) , with the vast majority in the range of 5.3% (cid:24) to 11.4% (cid:24) , suggesting the mixed derivation of sulfur from seawater sulfates and magmatic ﬂuids. Chalcopyrite separates from the vein-type ores have a Re-Os isochron age of 794.8 ± 7.9 Ma. The initial 187 Os / 188 Os (2.8 ± 1.2) and γ Os ( + 2202) values are slightly lower than the average values of continental crust, indicating a major metal source of the Hekou Group with minor mantle input. Sulﬁdes from the vein-type ores have δ 34 S values that range from − 10.3% (cid:24) to 4.0% (cid:24) and cluster between 0% (cid:24) to 2.2% (cid:24) , which implies a signiﬁcant contribution of magmatic-sourced sulfur with minor biogenic sulfur. Two major mineralization events have been identiﬁed. The Rb-Sr age of the laminated ores likely records a VMS mineralization event at ~1.6 Ga. The much younger Re-Os age is considered to represent the timing of an important mineralization event, which is likely related to the Neoproterozoic magmatism and / or metamorphism and represents a newly documented mineralization event to be targeted by exploration.


Introduction
The Kangdian copper belt, which extends for~300 km in the southwest of the Yangtze Block [1][2][3][4], is one of the most important iron-copper-gold production areas in China. The Lala copper ore field ( Figure 1) is one of the most important and typical ore fields within this belt. Two large deposits (Luodang and Hongnipo, with Cu resources of~0.56 and~0.64 Mt, respectively) and several medium-small  [70]); (b) Regional geologic map showing Precambrian strata, intrusions and major Fe-Cu and Cu deposits in the Kangdian copper belt (modified after [16,59]).

Deposit Geology
The Hongnipo copper deposit, with central geographic coordinates of 101°57′00″E and 26°11′30″N, is situated in the southwestern part of Huili County (Figure 1b). The exposed strata are the late Paleoproterozoic Luodang and Changchong Formations, Mesoproterozoic Heishan Formation, Triassic Baiguowan Formation, and overlying Quaternary ( Figure 2). The Luodang Formation is sparsely distributed in the eastern margin of the deposit, and is composed predominantly of garnet-biotite schist, carbonaceous slate, quartz albitite, biotite-quartz schist, muscovite-quartz schist, two-mica quartz schist, and metatuff. Lenses of marble occur locally in this formation. The Changchong Formation, in conformable contact with the Luodang Formation, is exposed to the west and northeast of the deposit, and its lithological associations are similar to those  [70]); (b) Regional geologic map showing Precambrian strata, intrusions and major Fe-Cu and Cu deposits in the Kangdian copper belt (modified after [16,59]).

Deposit Geology
The Hongnipo copper deposit, with central geographic coordinates of 101 • 57 00"E and 26 • 11 30"N, is situated in the southwestern part of Huili County (Figure 1b). The exposed strata are the late Paleoproterozoic Luodang and Changchong Formations, Mesoproterozoic Heishan Formation, Triassic Baiguowan Formation, and overlying Quaternary ( Figure 2). The Luodang Formation is sparsely distributed in the eastern margin of the deposit, and is composed predominantly of garnet-biotite schist, carbonaceous slate, quartz albitite, biotite-quartz schist, muscovite-quartz schist, two-mica quartz schist, and metatuff. Lenses of marble occur locally in this formation. The Changchong Formation, Minerals 2020, 10, 414 5 of 23 in conformable contact with the Luodang Formation, is exposed to the west and northeast of the deposit, and its lithological associations are similar to those of the Luodang Formation. The Heishan Formation, exposed in the northwest of the deposit, comprises carbonaceous slate, sericite phyllite, and marble. This formation is in fault contact with the Changchong Formation to the SE. The Baiguowan Formation, locally distributed in the southwest corner of the deposit, unconformably overlies the Changchong Formation. It mainly contains basal conglomerate, gritstone, shale, feldspar-quartz sandstone and siltstone. The Quaternary eluvial, alluvial and aqueoglacial sediments cover the entire central part of the deposit. The EW-striking Hongnipo synclinorium and NNE-striking faults (F 13  of the Luodang Formation. The Heishan Formation, exposed in the northwest of the deposit, comprises carbonaceous slate, sericite phyllite, and marble. This formation is in fault contact with the Changchong Formation to the SE. The Baiguowan Formation, locally distributed in the southwest corner of the deposit, unconformably overlies the Changchong Formation. It mainly contains basal conglomerate, gritstone, shale, feldspar-quartz sandstone and siltstone. The Quaternary eluvial, alluvial and aqueoglacial sediments cover the entire central part of the deposit. The EW-striking Hongnipo synclinorium and NNE-striking faults (F13, F27, F30) ( Figure 2) are the major structures at Hongnipo.
Extensive gabbro sills and dikes intrude the strata, but no other intrusive rocks have been identified. Intrusive breccias genetically related to the gabbro occur sporadically along boundaries between the strata and the intrusions ( Figure 2). Additionally, abundant gabbro dikes and intrusive breccias also occur at various depths below the deposit ( Figure 3). It is worth noting that fractures adjacent to the gabbro dikes, especially those in the hanging wall, commonly contain some mineralization. Extensive gabbro sills and dikes intrude the strata, but no other intrusive rocks have been identified. Intrusive breccias genetically related to the gabbro occur sporadically along boundaries between the strata and the intrusions ( Figure 2). Additionally, abundant gabbro dikes and intrusive breccias also occur at various depths below the deposit ( Figure 3). It is worth noting that fractures adjacent to the gabbro dikes, especially those in the hanging wall, commonly contain some mineralization. showing the geometry of the orebodies and their relationships to the host rocks (modified after [26]).
Five orebodies occur as stratiform, stratoid and lenticular, and are hosted in the Changchong and Luodang Formations ( Figure 3). These orebodies have an average copper grade of 1.42%. The majority are strata bound, dipping 10°-20° to the southwest [26]. There are two ore varieties, namely, laminated and vein-type ores (Figure 4a-d), the latter of which is vastly more important in terms of metal reserves of the deposit. In the laminated ores, sulfides always occur parallel to the foliation (Figure 4a), indicating that they may have formed contemporaneously with the host rocks. Typical of the vein-type ores are sulfide-calcite±quartz veins that fill the fractures of the host rocks ( Figure  4b) or overprint early laminated ores along the foliation (Figure 4c). It is common that vein-type mineralization crosscuts the laminated mineralization (Figure 4d), suggesting that economic Cu mineralization postdated the low-grade laminated mineralization. Ore minerals in the laminated ores generally consist of abundant subhedral to anhedral pyrite and minor chalcopyrite (Figure 5a). Arsenopyrite, pyrrhotite and sphalerite are scarce (Figure 5b,c). showing the geometry of the orebodies and their relationships to the host rocks (modified after [26]).
Five orebodies occur as stratiform, stratoid and lenticular, and are hosted in the Changchong and Luodang Formations ( Figure 3). These orebodies have an average copper grade of 1.42%. The majority are strata bound, dipping 10 • -20 • to the southwest [26]. There are two ore varieties, namely, laminated and vein-type ores (Figure 4a-d), the latter of which is vastly more important in terms of metal reserves of the deposit. In the laminated ores, sulfides always occur parallel to the foliation (Figure 4a), indicating that they may have formed contemporaneously with the host rocks. Typical of the vein-type ores are sulfide-calcite±quartz veins that fill the fractures of the host rocks ( Figure 4b) or overprint early laminated ores along the foliation (Figure 4c). It is common that vein-type mineralization crosscuts the laminated mineralization (Figure 4d), suggesting that economic Cu mineralization postdated the low-grade laminated mineralization.  showing the geometry of the orebodies and their relationships to the host rocks (modified after [26]).
Five orebodies occur as stratiform, stratoid and lenticular, and are hosted in the Changchong and Luodang Formations ( Figure 3). These orebodies have an average copper grade of 1.42%. The majority are strata bound, dipping 10°-20° to the southwest [26]. There are two ore varieties, namely, laminated and vein-type ores (Figure 4a-d), the latter of which is vastly more important in terms of metal reserves of the deposit. In the laminated ores, sulfides always occur parallel to the foliation (Figure 4a), indicating that they may have formed contemporaneously with the host rocks. Typical of the vein-type ores are sulfide-calcite±quartz veins that fill the fractures of the host rocks ( Figure  4b) or overprint early laminated ores along the foliation (Figure 4c). It is common that vein-type mineralization crosscuts the laminated mineralization (Figure 4d), suggesting that economic Cu mineralization postdated the low-grade laminated mineralization.

Sampling
For the present study, 21 ore samples were taken from the drill cores provided by Geological Team 403 of Sichuan Geological and Mineral Exploration and Development Bureau. The sampling locations are shown in Figure 2 and Table 1. The ore type of these samples are also given in Table 1. Six relatively low-grade laminated ores (Hnp019, Hnp024, Hnp026, Hnp031, Hnp178 and Hnp181) were selected from the No.1 orebody for pyrite Rb-Sr dating. They are hosted in quartz albitite or muscovite-quartz schist. Among these samples, no obvious post-ore hydrothermal alteration was identified with the naked eye. The sulfides occur parallel to the foliation, showing a crude stratiform

Sampling
For the present study, 21 ore samples were taken from the drill cores provided by Geological Team 403 of Sichuan Geological and Mineral Exploration and Development Bureau. The sampling locations are shown in Figure 2 and Table 1. The ore type of these samples are also given in Table 1. Six relatively low-grade laminated ores (Hnp019, Hnp024, Hnp026, Hnp031, Hnp178 and Hnp181) were selected from the No.1 orebody for pyrite Rb-Sr dating. They are hosted in quartz albitite or muscovite-quartz schist. Among these samples, no obvious post-ore hydrothermal alteration was identified with the naked eye. The sulfides occur parallel to the foliation, showing a crude stratiform mineralogical partitioning. Petrographic observation revealed that mineral assemblages consist of abundant subhedral to anhedral pyrite, with minor chalcopyrite, quartz and calcite. In addition, it is worth mentioning that two stages of pyrite were observed in samples Hnp024 and Hnp026. In these samples, the early stage pyrite is locally enclosed by the late stage pyrite and/or show signs of metasomatism ( Figure 5d). Six vein-type ores (Hnp484, Hnp487, Hnp488, Hnp490, Hnp493 and Hnp495) with high-grade Cu were selected from the No.1 orebody for chalcopyrite Re-Os geochronology. Irregular sulfide veins crosscut the wall rocks of marble, and are characterized by the mineral assemblage comprising dominant chalcopyrite, with minor pyrite, quartz and calcite. No later hydrothermal overprint was found in hand specimens and corresponding polished thin sections. Pure pyrite and chalcopyrite separates were handpicked under a binocular microscope after crushing, followed by magnetic and heavy liquid separation. These separates were then ground to about 200 mesh in an agate mortar. For sulfur isotope work, three representative laminated ores (Hnp017, Hnp022 and Hnp287) with minor laminated sulfides and six typical veined ores (Hnp053, Hnp223, Hnp483, Hnp499, Hnp503 and Hnp504) were made into polished thin sections. They share similar structural features, mineral associations and host rocks with those of samples used for Rb-Sr and Re-Os analyses.

Analytical Methods
The pyrite Rb-Sr isotope analyses were carried out at the Analytical Laboratory of the Beijing Research Institute of Uranium Geology, Beijing, China, following the procedures and methods described in the literature [71]. Pyrite grains were washed ultrasonically in analysis-grade alcohol and ultrapure water. Subsequently, accurately weighed pyrite powder (0.03−0.1 g) was loaded into Teflon vial (15 mL), and then wetted with a little high purity water. Proper amounts of 87 Rb and 84 Sr spikes were added into the vial, and then they were shaken to be uniform. After adding 3 mL hydrofluoric acid and 4 drops of perchloric acid and shaking the bottle, it was capped. Later, the bottle containing solution was heated on a hot plate with temperature around 120 • C for 24 h, until the sample dissolved completely. Hydrofluoric acid was expelled by heating it to 180 • C, following the operations of removing the cap and evaporating the sample. The internal wall of the tubular bottle was washed twice with 2 mL of 6 mol/L hydrochloric acid and then dried down at about 120 • C. Then, 1 mL of 0.5 mol/L hydrochloric acid was added into the tubular bottle to redissolve the sample, after which, the sample was transferred to plastic centrifuge tube. Finally, the supernatant used for the separation of rubidium and strontium was obtained through centrifugation for 6−10 min. The supernatant was then poured into the cation exchange column. A small amount of Milli-Q water was used to wash the corners of the column. Afterwards, 16 mL of 2 mol/L hydrochloric acid was used to elute and remove impurities such as potassium, sodium and iron. Rubidium was desorbed into a quartz beaker (10 mL) by the use of 5 mL of 2 mol/L hydrochloric acid. Moreover, 12 mL of 2 mol/L hydrochloric acid together with 2 mL of 3 mol/L hydrochloric acid was used to elute and remove ions such as calcium, magnesium and iron. Lastly, strontium was desorbed into a quartz beaker (10 mL) using 6 mL of 3 mol/L hydrochloric acid. Pure rubidium and strontium were loaded on single rhenium filaments with TaF 5 as activator. Rubidium and strontium isotopic compositions were then determined on a Phoenix TIMS instrument, in static multicollector mode. The internal precision of 87 Sr/ 86 Sr ratio is better than 0.002%. Mass fractionation corrections in Sr isotopic analyses were done with 86 Sr/ 88 Sr value of 0.1194. During this study, the 87 Sr/ 86 Sr ratio for NBS 987 Sr standard was 0.710250 ± 7 (2σ). Both of the blanks for Rb and Sr are 0.2 ng within the total procedure. Data regression for Rb-Sr isochron age was done with the Isoplot software [72], using errors of 2% for 87 Rb/ 86 Sr ratios and the within-run measurement precision for 87 Sr/ 86 Sr values (2σ).
The chalcopyrite Re-Os isotope analyses were performed at the Re-Os Laboratory of the National Research Center of Geoanalysis, Chinese Academy of Geological Sciences, Beijing, China, utilizing isotope dilution-negative thermal ionization mass spectrometry (ID-NTIMS). Detailed analytical procedures used in this study are found in Du et al. [73,74]. A precisely weighed sample, together with a predetermined amount of mixed 190 Os and 185 Re spike was loaded into a Carius tube, and then digested by reverse aqua regia. The tube was sealed and then heated in an oven at 200 • C for 24 h. After cooling, scratching, rewarming, and breaking off the tube's neck, the contents were poured into a distillation flask, and the residue was washed out into the flask with 40 mL of Milli-Q water. Osmium was separated and purified using distillation and microdistillation [75]. Rhenium was extracted from the residue by acetone in an 8 mol/L NaOH solution. The acetone phase was then transferred into a 100 mL beaker with 2 mL of Milli-Q water. The beaker was heated on a hot plate at 50 • C to expel acetone. When the acetone was exhausted, the heating temperature was increased to evaporate the solution to dryness. Afterwards, 0.8 mol/L HNO 3 was added to neutralize and dissolve the residue. An appropriate amount of solution obtained from the previous step was added to an anion exchange column, which had been equilibrated with 0.8 mol/L HNO 3 . Then, 3 mL of 0.8 mol/L HNO 3 , 2 mL of 1mol/L HCl were added to elute impurities, and 1 mL of Milli-Q water was used to remove the HCl. Lastly, 3 mL of 4 mol/L HNO 3 was used to elute Re into a 7 mL Teflon vial. Re and Os concentrations and isotope compositions were determined using a Thermo Fisher Scientific Triton plus mass spectrometer, operating in negative ion detection mode [76]. Purified Re and Os were loaded onto platinum filaments with Ba(OH) 2 as an ion emitter and measured as ReO − 4 and OsO − 3 , respectively. The signals of Re and Os were subsequently received by Faraday cups and secondary electron multiplier, respectively. Oxygen isotope interference was removed by means of a gradual spectral stripping method, according to the natural abundance of oxygen isotopes and statistical equal probability model [77]. The instrumental mass fractionation of Os was calibrated by normalizing the measured 192 Os/ 188 Os ratio to 3.08271 [78]. The mass fractionation of Re was corrected using normal Re ( 185 Re/ 187 Re = 0.59738) as an external standard determined at the same temperature as the samples. Moreover, Re isotopic fractionation correction uncertainties of 0.1% were included in the calculations, even though the mass bias for ReO − 4 is small using NTIMS. On the basis of blank runs analyzed together with samples, the total procedural blanks for Re and Os were about 3 pg and 0.5 pg, respectively. Uncertainties for Re-Os isotopic dating were determined by error propagation of uncertainties in Re and Os mass spectrometer measurement and mass fractionation, bank abundances and isotopic compositions, spike calibrations, and weighing uncertainties of samples and spike. In addition, an in-house sulfide Re-Os isotope references JCBY from Jinchuan Cu-Ni sulfide deposit was used for quality control of the whole procedure. Its results are identical with the recommended value within analytical uncertainty for Os isotopic compositions, as well as Re and Os concentrations.
Sulfides' in situ sulfur isotopic composition measurements were conducted on a Nu Plasma II MC-ICP-MS instruments at the State Key Laboratory of Geological Process and Mineral Resources (GPMR), China University of Geosciences (Wuhan). The detailed analytical procedure and conditions are given in Liu et al. [79]. In brief, the sulfides on polished thin sections were ablated by a Resonetics-S155 laser ablation system, with a laser energy fluence of 3 J/cm 2 , a beam diameter of 33 µm, a repetition rate of 10 Hz and an ablation time of 40 s. Ablated sulfide aerosols were carried by high purity helium and mixed with argon and minor nitrogen, before entering the plasma. Using the 34 S/ 32 S ratios of samples and the internal laboratory standard pyrite (WS-1) (δ 34 S: + 0.9% ), the δ 34 S values of sulfides from ore samples were obtained by using the sample-standard bracketing approach.

Rb-Sr Isotopes
The Rb-Sr isotope data of six pyrite samples are listed in Table 2. They have variable contents of Rb (0.022 to 0.446 ppm) and Sr (0.269 to 1.74 ppm), with 87 Rb/ 86 Sr ratios from 0.1447 to 1.1854. The scatter observed probably suggests the heterogeneity of the initial Sr isotopic compositions of the pyrite [80] or inclusions [81][82][83]. Both Rb and Sr are lithophile elements and are incompatible within pyrite [1]. Combined with the typical sieve texture (Figure 5f), the minerals within pyrite may be responsible for the variation.  Figure 6) and 87 Rb decay constant of 1.42 × 10 −11 a −1 .
Minerals 2019, 9, x FOR PEER REVIEW 10 of 23 ratios of samples and the internal laboratory standard pyrite (WS-1) (δ 34 S: + 0.9‰), the δ 34 S values of sulfides from ore samples were obtained by using the sample-standard bracketing approach.

Rb-Sr Isotopes
The Rb-Sr isotope data of six pyrite samples are listed in Table 2. They have variable contents of Rb (0.022 to 0.446 ppm) and Sr (0.269 to 1.74 ppm), with 87 Rb/ 86 Sr ratios from 0.1447 to 1.1854. The scatter observed probably suggests the heterogeneity of the initial Sr isotopic compositions of the pyrite [80] or inclusions [81][82][83]. Both Rb and Sr are lithophile elements and are incompatible within pyrite [1]. Combined with the typical sieve texture (Figure 5f), the minerals within pyrite may be responsible for the variation.
Regression of all six Rb-Sr isotopic data gives a highly imprecise model 3 [72] date of 1530 ± 880 Ma (MSWD = 289; Figure 6). Most of the scatter is attributable to Hnp024 and Hnp026. The isochron dating method requires that the samples have identical initial isotopic composition and the isotopic system remained closed after mineral crystallization [84]. In view of these requirements and our petrographic evidence (Figure 5d), samples Hnp024 and Hnp026 should be precluded for isochron age calculation. A best-fit line through the remaining four analyses defines a much more precise model 1 age of 1552 ± 80 Ma ( Figure 6), with an initial 87 Sr/ 86 Sr ratio of 0.71214 ± 0.00081 (MSWD = 3.2). Using the age of 1552 Ma, the calculated initial 87 Sr/ 86 Sr ratios for the remaining samples are 0.712330, 0.712010, 0.711957, and 0.712236 (Table 1), with an average value of 0.712133.  Figure 6) and 87 Rb decay constant of 1.42 × 10 −11 a −1 .

Re-Os Isotopes
Re and Os data of six chalcopyrite samples in which the blank corrections have been made are shown in Table 3. All samples contain low concentrations of Re and 187 Os, with large concentration Regression of all six Rb-Sr isotopic data gives a highly imprecise model 3 [72] date of 1530 ± 880 Ma (MSWD = 289; Figure 6). Most of the scatter is attributable to Hnp024 and Hnp026. The isochron dating method requires that the samples have identical initial isotopic composition and the isotopic system remained closed after mineral crystallization [84]. In view of these requirements and our petrographic evidence (Figure 5d), samples Hnp024 and Hnp026 should be precluded for isochron age calculation. A best-fit line through the remaining four analyses defines a much more precise model 1 age of 1552 ± 80 Ma ( Figure 6), with an initial 87 Sr/ 86 Sr ratio of 0.71214 ± 0.00081 (MSWD = 3.2). Using the age of 1552 Ma, the calculated initial 87 Sr/ 86 Sr ratios for the remaining samples are 0.712330, 0.712010, 0.711957, and 0.712236 (Table 1), with an average value of 0.712133.

Re-Os Isotopes
Re and Os data of six chalcopyrite samples in which the blank corrections have been made are shown in Table 3. All samples contain low concentrations of Re and 187 Os, with large concentration ranges of 0.7317-17.23 ppb and 0.0083-0.1471 ppb, respectively. The abundances of common Os are low, but relatively uniform (0.00329-0.00742 ppb). The 187 Re/ 188 Os ratios vary between 873.8 and 11225, and the 187 Os/ 188 Os ratios range from 15.82 to 152.5. The regression of six 187 Re/ 188 Os-187 Os/ 188 Os data points with the corresponding uncertainty correlation value, yields a Re-Os model 3 [72] age of 874 ± 310 Ma. The scatter of these data around the regression line is extremely large (MSWD = 954; Figure 7a). A high MSWD value is indicative of the existence of geological scatter about the isochron in excess of the analytical uncertainty [85]. Most of the scatter is accounted for by samples Hnp487 and Hnp493, which deviate significantly from the regressive trend, compared with the remaining data ( Figure 7a). Given the absence of petrographic evidence for post-ore hydrothermal overprint, the possible cause of deviation of the two samples is heterogeneity in their initial 187 Os/ 188 Os ratios, in comparison to the other four samples left. Considering the fundamental requirements for isochron dating [84], samples Hnp487 and Hnp493 are omitted. Regressing all the remaining data results in a model 1 (assumes that the scatter regarding to the linear regression only results from assigned uncertainty; [72]) Re-Os date of 794 ± 12 Ma (2σ, initial 187 Os/ 188 Os = 2.8 ± 1.2, MSWD = 2.9; Figure 7a). Using the calculation method given in Zhu and Sun [17], Selby et al. [86] and Li et al. [87], the abundance and uncertainty of radiogenic 187 Os ( 187 Os r ) of samples (Hnp484, Hnp488, Hnp490 and Hnp495) are derived from the corresponding total 187 Os and error ( Table 3). The 187 Os budget is mainly composed of 187 Os r (81.95%-98.12%; Table 3), which satisfies the characteristics of low-level highly radiogenic sulfide [87]. As a result, the 187 Re versus 188 Os r data can be applied to determine a Re-Os date [88]. The 187 Re versus 188 Os r data yield a model 1 four-point date of 794.8 ± 7.9 Ma (2σ, initial 187 Os = −0.0002 ± 0.00075 ppb, MSWD = 2.4; Figure 7b), which is within the uncertainty of the 187 Re/ 188 Os versus 187 Os/ 188 Os isochron age.

Sulfur Isotopes
Sulfide isotopic data are listed in Table 4 and illustrated in Figure 8. Sulfides from the laminated ores have δ 34 S values ranging from −1.8‰ to 11.4‰. Except for spot Hnp017-1 with a negative δ 34 S value of −1.8‰, both pyrite and chalcopyrite have high positive δ 34 S values. The δ 34 S values of sulfides from the vein-type ores range from −10.3‰ to 4.0‰, but mostly between 0‰ and 2.2‰. Additionally, there are two samples, especially sample Hnp504, which have δ 34 S values characterized by strong negative values (down to −10.3‰). Overall, most of the δ 34 S values of the sulfides from the laminated ores are much higher than those from vein-type ores (Figure 9). Using the 187 Re (sample with 187 Re/ 188 Os ratio exceeding 2800) and 187 Os r data, calculated model ages are 774 ± 17 Ma, 795 ± 22Ma and 775 ± 19 Ma, respectively (Table 3), and the weighted mean age is 779 ± 26 Ma (MSWD = 1.3). The relatively large uncertainties for model ages are attributed to the introduction of large errors caused by the blank correction for the extremely low level of Re and 187 Os in chalcopyrite [88]. Nevertheless, they can still provide additional information for mineralization ages. The 187 Re versus 187 Os r isochron age (Figure 7b) is within uncertainty and in accordance with both the four-point 187 Re/ 188 Os versus 187 Os/ 188 Os age (Figure 7a) and the weighted mean age.
Furthermore, based on the values of initial 187 Os/ 188 Os of 2.8 and 187 Re versus 187 Os isochron age of 794.8 Ma, an approximate γ Os value (+2202) is calculated using the formulas shown in Jiang et al. [89] and Walker et al. [90].

Sulfur Isotopes
Sulfide isotopic data are listed in Table 4 and illustrated in Figure 8. Sulfides from the laminated ores have δ 34 S values ranging from −1.8% to 11.4% . Except for spot Hnp017-1 with a negative δ 34 S value of −1.8% , both pyrite and chalcopyrite have high positive δ 34 S values. The δ 34 S values of sulfides from the vein-type ores range from −10.3% to 4.0% , but mostly between 0% and 2.2% . Additionally, there are two samples, especially sample Hnp504, which have δ 34 S values characterized by strong negative values (down to −10.3% ). Overall, most of the δ 34 S values of the sulfides from the laminated ores are much higher than those from vein-type ores (Figure 9).

Ages of Two Mineralization Events
Field relationships show vein-type mineralization crosscut laminated mineralization (Figure 4d) at the Hongnipo deposit. Petrographic observation reveals that the percentage of chalcopyrite is significantly different in these two types of ores (Figures 5a-f and 8). This evidence suggests these two ore type might have formed from different metallogenic events.
For all isochron dating methods, including Rb-Sr, the prerequisites are that all samples have the can be ascribed to the interaction between ore-forming fluids and organic matter. Thus, the sulfur of sources of the sulfides in the vein-type ores are mixtures of magmatic sulfur and biogenic sulfur.

Implications for Regional Metallogeny
Our new geochronological data show that the Hongnipo deposit has formed by two mineralization events at ~1.6 and 0.8 Ga. By contrast, previous data of the Luodang deposit from the same ore field recorded four hydrothermal events, including three main mineralization events at ~1.67, ~1.3 and ~1.1 Ga and a local hydrothermal overprint events at ~0.82 Ga (Figure 10). Although the early mineralization event age (~1.6 Ga) in the Hongnipo deposit is slightly younger than that (~1.67 Ga) in the Luodang deposit, they are indistinguishable from the zircon U-Pb ages of their respective host rocks (~1.6-1.7 Ga [5][6][7][8][9][10]). This geochronological evidence, together with the rifting setting [6,7,51,59], mixed sulfur sources of seawater sulfates and magmas and the typical ore structures, indicates that they were likely to have formed through a volcanic-exhalative-sedimentary process. The early stage mineralization is a VMS-type mineralization in the Lala ore field. The late main mineralization event age (~0.8 Ga) in the Hongnipo deposit is much younger than the subsequent main mineralizaion ages (~1.3 and ~1.1-1.0 Ga) in the Luodang deposit. Neoproterozoic magmatism (ages: 850-796 Ma [53,64,65,67,111,112]) and/or metamorphism (ages: ~850 Ma [15,113,114]) is likely responsible for the youngest main mineralization event at Hongnipo.
Many other deposits (e.g., Yinachang, Etouchang and Dahongshan) in the Kangdian copper belt are hosted in comparable volcano-sedimentary sequences and share similar mineralization styles, indicating that they might have had a similar metallogenic history.
A number of geochronological data of other representative deposits in the Kangdian copper belt have been published so far. Ye et al. [115] and Hou et al. [116] reported chalcopyrite Re-Os isochron ages of 1690 ± 99 Ma and 1648 ± 14 Ma, respectively, for the formation timing of the Yinachang deposit. Zhao et al. [1] obtained two groups of Re-Os molybdenite ages, with average ages of 1654 ± 7 Ma and 1451 ± 6 Ma. They suggested that the Yinachang deposit was formed at 1654 ± 7 Ma, and overprinted by a less important and local hydrothermal event. Li and Zhou [117] reported a monazite U-Pb age of 842 ± 9 Ma and interpreted it as the timing of a post-ore hydrothermal event. Zhao et al. [1] also reported a chalcopyrite Re-Os isochron age and a pyrite Rb-Sr isochron age of 1487 ± 110 Ma and 1453 ± 28 Ma, respectively. They suggested that these ages likely recorded the timing of a secondary hydrothermal event. A zircon U-Pb age of 1653 ± 18 Ma for hydrothermal zircons from stockwork chalcopyrite-magnetite ore was proposed to represent the timing of major mineralization that formed the Dahongshan deposit [58]. Zhao et al. [58] also reported pyrite and molybdenite Re-Os ages of 1026 ± 22 Ma and 780 ± 36 Ma, and argued that they recorded the timing of post-ore hydrothermal overprint or remobilization. Song [18] and Jin et al. [57] reported two Re-Os chalcopyrite ages of 1083 ± 45 Ma and 1115 ± 28 Ma, respectively, to constrain the timing of the main mineralization event. Compiling all the geochronological data together (Figure 10), the Kangdian copper belt has

Ages of Two Mineralization Events
Field relationships show vein-type mineralization crosscut laminated mineralization (Figure 4d) at the Hongnipo deposit. Petrographic observation reveals that the percentage of chalcopyrite is significantly different in these two types of ores (Figures 5a-f and 8). This evidence suggests these two ore type might have formed from different metallogenic events.
For all isochron dating methods, including Rb-Sr, the prerequisites are that all samples have the same initial isotopic composition and the isotopic system remains closed after mineral crystallization [84]. Owing to the Rb and Sr budgets of pyrite are predominantly from silicate inclusions, the Rb-Sr age of pyrite essentially represents the timing of the K-bearing minerals that cooled below the closure temperatures of the Rb-Sr isotope system therein [1]. Although previous studies have shown that the Lala area has suffered from upper greenschist facies metamorphism [5], the pyrite Rb-Sr isotope age of 1552 ± 80 Ma, with its large uncertainty, is consistent with the same stage pyrite Pb-Pb age (1669 ± 110 Ma [24]). This indicates that the Rb-Sr isotope systematics of mineral inclusions within pyrite still remained closed. Additionally, the pyrite samples, which defined the well-constrained four-point regression line, were separated from the same ore type collected from the same orebody. Consequently, we suggest that the Rb-Sr isochron age of 1552 ± 80 Ma likely represents the best estimate of the formation timing of the laminated ores.
In this study, we obtained a well-constrained four-point Re-Os chalcopyrite isochron age of 794.8 ± 7.9 Ma (Figure 7b) in this study. No deformation or hydrothermal overprint in chalcopyrite has been detected through field and microscopic observations. This suggests that the chalcopyrite Re-Os system has not been affected by post-ore tectonothermal events, and this age is considered to precisely record the timing of vein-type Cu mineralization. Apparently, it is significantly younger than the laminated mineralization. The field and textural relationships, coupled with our new geochemical data, indicate that at least two mineralizing events were involved in the formation of the Hongnipo deposit.

Laminated Ores
The 87 Sr/ 86 Sr ratio is generally used to discriminate the sources of ore-forming fluids and materials. For example, Chen et al. [92] and Zhao et al. [93] managed to trace distinct fluids sources by comparing the Sr isotopic compositions of apatites with those of other reservoirs. The initial 87 Sr/ 86 Sr value (~0.712) at~1.6 Ga is highly radiogenic, indicating a major contribution from crustal materials. The Paleoproterozoic host rocks with documented 87 Sr/ 86 Sr ratios from 0.711614 to 0.718937 at the time of the laminated mineralization [5], are likely candidates accounting for the radiogenic Sr of the laminated ores. This is in agreement with a previous Pb isotope study of pyrite from this ore type, which demonstrated that ore-forming materials were sourced from the Hekou Group [11].
The heavy sulfur isotope characteristics (up to 11.4% ) for the laminated sulfides are also observed in other deposits of the Kangdian copper belt; for instance, the Dongchuan, Etouchang and Tangdan deposits [59,94]. In addition, such sulfur isotope compositions are comparable to those of sulfides in Emmie Bluff (2.6% -12.5% , most values larger than 5.2% [95]) and Mount Gunson (7.1% -11.7% [96]), both of which are located in the Olympic Dam District of South Australia. Bastrakov et al. [95] proposed evaporates, calcareous host rocks or seawater to be the source of sulfur in the Olympic Dam district. Zhao et al. [94] and Huang et al. [97] suggested that the marine evaporates contributed sulfur during precipitation of sulfides in the Tangdan deposit. Considering the volcanism, marine paleogeographic environment [98] and the δ 34 S values of oceanic sulfates in the late Paleoproterozoic (18% -25% [99]), we infer that sulfur sources of sulfides in the laminated ores are mixtures of seawater sulfates and magmatic fluids.

Vein-Type Ores
As a Cu-affinity element, Os shares similar geochemical behavior with Cu during magmatic evolution and ore-forming fluid migration. Consequently, Os can act as a tracer for the metal provenance, using the distinctive 187 Os/ 188 Os ratios and γ Os values of the crust and mantle as effective discriminants of different sources.
The vein-type ores have an initial 187 Os/ 188 Os value of 2.8 ± 1.2 ( Figure 7a). The large uncertainty is due to the measurement error caused by low levels of common Os [100]. Both it and the γ Os value (+2202) are only slightly smaller than the mean values of continental crust (3.63 and +2758, respectively, [101]), which implies that the ore-forming materials mainly came from the continental crust [102][103][104]. With the Cu concentration of the Hekou Group being several times higher than its Clark Value [105] and the existence of older laminated ores (Figure 4d), the indications are that the copper of the veined ores was likely derived from the Hekou Group. In addition, the gabbro swarms, spatially and temporally associated with the deposit, originated from enriched mantle (EM I) [106], which has a 187 Os/ 188 Os value of 0.152 and a γ Os value of +19.7 [107], may have contributed a portion of the metals to the vein-type mineralization.
Most δ 34 S values of sulfides from the vein-type ores are in the range of 0% to 2.2% (Figure 9), suggesting magmatic-sourced sulfur [108][109][110]. The marked negative δ 34 S values (from −10.3% to −3.6% ) of some veined sulfides largely correspond to the data from the Manxman prospect (−3.6% to −6.5% ) and from the Tangdan deposit (−5.2% to −14.7% ), given in Bastrakov et al. [95] and Huang et al. [97], respectively. Bastrakov et al. [95] suggested that such negative values resulted from involvement of sulfur bearing fluids with negative isotopic values. Huang et al. [97] also suggested that a reduced source, such as carbonaceous slate, provided the sulfur source. In view of the fact that carbonaceous slates often act as the floors of orebodies, the enrichment of light sulfur for these ores can be ascribed to the interaction between ore-forming fluids and organic matter. Thus, the sulfur of sources of the sulfides in the vein-type ores are mixtures of magmatic sulfur and biogenic sulfur.

Conclusions
The Hongnipo copper deposit, a newly discovered large deposit, is roughly strata bound, with orebodies consisting of laminated, vein-type and superimposed ores. Two mineralization events have been identified in this study. Pyrite separates from laminated ores yield a Rb-Sr isochron age of 1552 ± 80 Ma, representing an early mineralization event. In contrast, chalcopyrite separates from veintype ores yield a much younger Re-Os isochron age of 794.8 ± 7.9 Ma, reflecting the main mineralization event.
The highly radiogenic initial 87 Sr/ 86 Sr and 187 Os/ 188 Os values of pyrites from the laminated ores and chalcopyrites from veined ores, respectively, suggest significant contributions of crustal materials. Sulfur isotopic compositions of the laminated and vein-type ores are quite different. The Many other deposits (e.g., Yinachang, Etouchang and Dahongshan) in the Kangdian copper belt are hosted in comparable volcano-sedimentary sequences and share similar mineralization styles, indicating that they might have had a similar metallogenic history.
A number of geochronological data of other representative deposits in the Kangdian copper belt have been published so far. Ye et al. [115] and Hou et al. [116] reported chalcopyrite Re-Os isochron ages of 1690 ± 99 Ma and 1648 ± 14 Ma, respectively, for the formation timing of the Yinachang deposit. Zhao et al. [1] obtained two groups of Re-Os molybdenite ages, with average ages of 1654 ± 7 Ma and 1451 ± 6 Ma. They suggested that the Yinachang deposit was formed at 1654 ± 7 Ma, and overprinted by a less important and local hydrothermal event. Li and Zhou [117] reported a monazite U-Pb age of 842 ± 9 Ma and interpreted it as the timing of a post-ore hydrothermal event. Zhao et al. [1] also reported a chalcopyrite Re-Os isochron age and a pyrite Rb-Sr isochron age of 1487 ± 110 Ma and 1453 ± 28 Ma, respectively. They suggested that these ages likely recorded the timing of a secondary hydrothermal event. A zircon U-Pb age of 1653 ± 18 Ma for hydrothermal zircons from stockwork chalcopyrite-magnetite ore was proposed to represent the timing of major mineralization that formed the Dahongshan deposit [58]. Zhao et al. [58] also reported pyrite and molybdenite Re-Os ages of 1026 ± 22 Ma and 780 ± 36 Ma, and argued that they recorded the timing of post-ore hydrothermal overprint or remobilization. Song [18] and Jin et al. [57] reported two Re-Os chalcopyrite ages of 1083 ± 45 Ma and 1115 ± 28 Ma, respectively, to constrain the timing of the main mineralization event. Compiling all the geochronological data together (Figure 10), the Kangdian copper belt has experienced several mineralization and hydrothermal events as follows: (1) the earliest main mineralization at 1.6-1.67 Ga, which is nearly synchronous with host rock (ages:~1.6-1.7 Ga [6][7][8]10,39,51,59,118]); (2) a local hydrothermal overprint at~1.45 Ga, which is related to early Mesoproterozoic magmatism (age:~1.5 Ga [54,62,119]); (3) a second major mineralization at~1.3 Ga, which may have been caused by volcanism (age: 1270 ± 95 Ma [20]); (4) the third major mineralization at~1.1-1.0 Ga, which is closely associated with late Mesoproterozoic magmatism (ages: 1142-1208 Ma [19][20][21]52,120]); (5) the fourth major mineralization at~0.8 Ga, which is a newly identified main mineralization event in the Kangdian copper belt, and likely related to the Neoproterozoic magmatism (ages:~850 Ma [15,113,114]) and/or metamorphism (ages:~850 Ma [15,113,114]).

Conclusions
The Hongnipo copper deposit, a newly discovered large deposit, is roughly strata bound, with orebodies consisting of laminated, vein-type and superimposed ores. Two mineralization events have been identified in this study. Pyrite separates from laminated ores yield a Rb-Sr isochron age of 1552 ± 80 Ma, representing an early mineralization event. In contrast, chalcopyrite separates from vein-type ores yield a much younger Re-Os isochron age of 794.8 ± 7.9 Ma, reflecting the main mineralization event.
The highly radiogenic initial 87 Sr/ 86 Sr and 187 Os/ 188 Os values of pyrites from the laminated ores and chalcopyrites from veined ores, respectively, suggest significant contributions of crustal materials. Sulfur isotopic compositions of the laminated and vein-type ores are quite different. The majority of the sulfides from the laminated ores have much higher δ 34 S values (5% to 12% ) than those of the vein-type ores (−11% to 4% ). Sulfur of the laminated ores was likely sourced from seawater sulfates and magmatic fluids, while sulfur of the vein-type ores probably came predominantly from magmatic sulfur with minor biogenic sulfur.
The early laminated mineralization identified in the Hongnipo deposit is likely a VMS-type mineralization, and is widespread in the Kangdian copper belt. The late hydrothermal mineralization is likely related to extensive Neoproterozoic magmatism and/or metamorphic events and is responsible for the formation of economic vein-type mineralization. At present, there is no evidence of this later mineralization event at any of the other deposits of the Kangdian copper belt.