A Combined Re-Os and Pt-Os Isotope and HSE Abundance Study of Ru-Os-Ir Alloys from the Kunar and Unga Placer Deposits, the Taimyr Peninsula, Polar Siberia

: In order to provide further insights into the origin of Ru-Os-Ir alloys, this study presents new highly siderophile element (HSE: Re, Os, Ir, Ru, Pt, and Pd) abundance and 187 Re-187 Os and 190 Pt-186 Os isotope data for detrital grains of native Ru-Os-Ir alloys in placer deposits of the Ku-nar and Unga Rivers, which display a close spatial association with the Kunar dunite–harzburgite complex in the northern part of the Taimyr Peninsula in the Polar Siberia. The study utilized electron microprobe analysis, negative thermal ionization mass-spectrometry (N-TIMS) and laser ablation multiple-collector inductively coupled plasma mass-spectrometry (LA MC-ICP-MS). The primary nature of the Ru-Os-Ir alloys is supported by the occurrence of euhedral inclusions of high-Mg olivine (Fo 92–93 ) that fall within the compositional range of mantle olivine. The LA MC-ICP-MS data show similar average initial 187 Os/ 188 Os and γ 187 Os(740 Ma) values for PGM assemblages from the Kunar and Unga deposits of 0.1218 ± 0.0010, − 0.18 ± 0.85, and 0.1222 ± 0.0025, +0.10 ± 2.1, respectively. These values are identical, within their respective uncertainties, to the initial 187 Os/ 188 Os value of the Ru-Os-Ir alloy grain measured by N-TIMS (0.1218463 ± 0.0000015, γ 187 Os(740 Ma) = − 0.1500 ± 0.0012). The combined 187 Re-187 Os isotopic data for all studied grains ( γ 187 Os(740 Ma) = − 0.02 ± 1.6) indicate evolution of the Kunar and Unga mantle sources with a long-term chondritic 187 Re/ 188 Os ratio of 0.401 ± 0.030. In contrast to the 187 Os/ 188 Os data, the initial 186 Os/ 188 Os value of 0.1198409 ± 0.0000012 ( µ 186 Os(740 Ma) = +34 ± 10) obtained for the same Ru-Os-Ir alloy grain by N-TIMS is suprachondritic and implies evolution of the Kunar and Unga mantle source(s) with a long-term suprachondritic 190 Pt/ 188 Os ratio of 0.00247 ± 0.00021. This value is ~40% higher than the average chondritic 190 Pt/ 188 Os ratio of 0.00180 and indicates long-term enrichment of the Kunar source in Pt over Os. Establishing the source of this enrichment requires further investigation.


Introduction
Abundances of the highly siderophile elements (HSE = Os, Ir, Ru, Rh, Pt, Pd, Re, and Au) and HSE-based isotopic systems ( 187 Re-187 Os and 190 Pt-186 Os) are important tools for geochemical investigations of the mantle. Because Re and Os have strongly contrasting partitioning behavior during mantle melting and magma differentiation, e.g., [1][2][3][4], the 187 Re-187 Os isotopic system is particularly useful in (i) distinguishing between crustal and mantle sources of the HSE, (ii) tracking melt extraction events, and (iii) studying long-term evolution of mantle sources. This tool can be applied at both the whole-rock and individual mineral (i.e., chromite, Ru-Os-Ir alloy, sulfide) scales. To gain further insight into the origin of these detrital Ru-Os-Ir alloys, we obtained HSE abundance and 187 Os/ 188 Os isotopic data for eight alluvial grains of Os-rich alloys from the Kunar River placer deposit, which is closely associated with the ultramafic rocks of the Kunar ultramafic complex, and for fourteen Ru-Os-Ir alloy grains from the Unga River placer deposit. This dataset is supplemented by a high-precision 186 Os/ 188 Os isotopic analysis of an Os-rich alloy sample from the Kunar deposit. These new data are compared with the published data for Ru-Os-Ir alloys of the Unga placer deposit [17] and other globally distributed placer deposits, e.g., [14,15]. We integrate an application of To gain further insight into the origin of these detrital Ru-Os-Ir alloys, we obtained HSE abundance and 187 Os/ 188 Os isotopic data for eight alluvial grains of Os-rich alloys from the Kunar River placer deposit, which is closely associated with the ultramafic rocks of the Kunar ultramafic complex, and for fourteen Ru-Os-Ir alloy grains from the Unga River placer deposit. This dataset is supplemented by a high-precision 186 Os/ 188 Os isotopic analysis of an Os-rich alloy sample from the Kunar deposit. These new data are compared with the published data for Ru-Os-Ir alloys of the Unga placer deposit [17] and other globally distributed placer deposits, e.g., [14,15]. We integrate an application of electron probe microanalysis (EPMA) of HSE, Fe, Ni, and Cu abundances in Ru-Os-Ir alloys and solid inclusions within these PGM, and 187 Os/ 188 Os LA MC-ICP-MS isotopic analysis of Ru-Os-Ir alloys supplemented by high-precision 186,187 Os/ 188 Os N-TIMS isotope analysis of native osmium. The objectives of the research were to (i) obtain HSE abundances in the PGM from both localities, and (ii) constrain the source(s) of the HSE in these PGM and HSE composition of that source. This study is part of a larger effort to elucidate the mantle versus crustal origin of HSE/PGM in oceanic and subcontinental mantle settings, e.g., [7,8,11,12,[14][15][16][17]21,[37][38][39][40][41][42][43][44][45][46][47][48][49][50][51].

Geological Background and Sample Location
Dunite-harzburgite and dunite-peridotite massifs of the Taimyr Peninsula are represented by multiple elongated bodies confined to five ultramafic belts, i.e., Chelyuskin, Malinovsk, Zhdaniinsk, Moskvichevsk, and Stanovoi [35,36,[52][53][54][55][56]. The first four represent relicts of the supra-subduction zone ophiolite of the Late Riphean (ca. 740 Ma) suture located on the border of the North Kara microcontinent and the Siberian Craton. The dunite-peridotite massifs of the Stanovoi belt record an earlier event (ca. 960 Ma), manifested by the transformation of the regime of the passive continental margin into an active one that occurred in the Early Neoproterozoic.
Ultramafic rocks of the Kunar complex ( Figure 2), together with the other NE-trending bodies that form the Chelyuskin ultramafic belt, are controlled by the zone of the Main Taimyr Fault [52]. The ultramafic bodies of the Kunar complex are composed of serpentinized harzburgites, dunites, and associated chromitites; these trend northeast for about 70 km (Figure 2). These bodies usually have tectonic contacts with metavolcanics, zones of serpentinite melange, greenschists, and listvenites [36]. Other rocks in the Chelyuskin ophiolite belt are represented by tholeiitic basalts, dolerites of a dyke-sill complex, and sheeted plagiogranites [35]. The geology of the Chelyuskin ophiolite belt has been described in previous studies [36,52,57].
Despite the lack of direct geochronological information on the ultramafic rocks, the protrusive nature of the ultramafic bodies, confinement to the Late Riphean formations, and findings of chromite in conglomerates of the Lower sub-formation of Laptevsk Suite (R 3 lp 1 ) allowed the emplacement of ultramafic bodies to be attributed to the pre-Laptev time.
U-Pb and Sm-Nd isotopic studies indicate that the plagiogranites of the Chelyuskin ophiolite formed between 720 and 740 Ma [35,56,58]. Taking into account the above structural, petrological, and geochronological constraints, we assume the minimum formation age of 740 Ma for the ultramafic rocks of the Kunar complex.
In this study, we analyzed (i) 12 grains of Ru-Os-Ir alloys ranging in size from 0.25 to 2 mm, obtained by panning the Quaternary deposits in the middle part of the Kunar River stream (Figure 2), and (ii) 33 grains of Ru-Os-Ir alloys that range in size between 0.25 and 0.5 mm, which were obtained from a gold production concentrate in the area of the Unga River located between Cape Chelyuskin and Cape Pronchishcheva ( Figure 2). The Unga placer deposit area is characterized by polygenic and polychronous placers of gold, which are represented by Jurassic conglomerates of riverbed valleys, Paleogene weathering crusts, Paleogene-Neogene coastal-marine sediments, and Quaternary eluvial-deluvial formations [27,28]. The significant content of chromite (4.7-17.5 g/m 3 [28]) in the heavy fraction of the production concentrates and the spatial proximity of the Kunar complex imply that chromitites of the Kunar complex are the most probable source of the PGM studied.  [59] with locations of the PGM grains studied.

Analytical Techniques
The analytical work was performed at the Institute of Geology and Geochemistry, Ural Branch of the Russian Academy of Sciences (IGG UB RAS), Ekaterinburg, Russia, Australian Research Council Centre of Excellence for Core to Crust Fluid Systems (ARC CoE CCFS)/ARC National Key Centre for Geochemical Evolution and Metallogeny of Continents (GEMOC ARC National Key Centre) in the Department of Earth and Planetary Sciences at Macquarie University, Sydney, Australia, and the Isotope Geochemistry Laboratory (IGL) at the Department of Geology, University of Maryland College Park, USA.
Initially, the morphology of the PGM grains, represented by individual crystals and polymineral aggregates, was documented by the scanning electron microscopy (SEM).

Analytical Techniques
The analytical work was performed at the Institute of Geology and Geochemistry, Ural Branch of the Russian Academy of Sciences (IGG UB RAS), Ekaterinburg, Russia, Australian Research Council Centre of Excellence for Core to Crust Fluid Systems (ARC CoE CCFS)/ARC National Key Centre for Geochemical Evolution and Metallogeny of Continents (GEMOC ARC National Key Centre) in the Department of Earth and Planetary Sciences at Macquarie University, Sydney, Australia, and the Isotope Geochemistry Laboratory (IGL) at the Department of Geology, University of Maryland College Park, USA.
Initially, the morphology of the PGM grains, represented by individual crystals and polymineral aggregates, was documented by the scanning electron microscopy (SEM). The grains were then mounted and polished, described, and analyzed by electron probe microanalysis (EPMA) using a CAMECA SX-100 equipped with five WDS spectrometers and a Bruker energy dispersive spectrometer system at Common Use Center "Geoanalyst" of the Institute of Geology and Geochemistry, Ural Branch of the Russian Academy of Sciences, Ekaterinburg, Russia. Quantitative WDS analyses were performed at 25 kV accelerating voltage and 20 nA beam current, with a beam diameter of about 1 µm. The following X-ray lines and standards have been used: RuLα, RhLα, PdLβ, OsMα, IrLα, PtLα, NiKα (all native element standards), FeKα, CuKα, SKα (chalcopyrite), and AsLα (sperrylite). Corrections were performed for the interferences involving Ru-Rh, Ru-Pd, and Ir-Cu. Only Os, Ir, Ru, Rh, Pt, Fe, and Ni were found to be above the detection limits of the EPMA analysis. A total of 128 EPMA analyses of Ru-Os-Ir alloys were performed. Additional details of the analytical protocols used are given in Badanina et al. [60].
Thirty one in situ Re-Os isotope analyses of Ru-Os-Ir alloys were carried out at the Geochemical Analysis Unit, CCFS/GEMOC laboratories, Macquarie University, Sydney, Australia, using analytical methods described in detail by Marchesi et al. [42], González-Jimenéz et al. [61] and Malitch et al. [48]. These analyses employed a New Wave/Merchantek UP 213 laser ablation system attached to a Nu Plasma multi-collector ICP-MS instrument. Laser ablation was carried out with a frequency of 4 Hz, energies of 1-2 mJ/pulse and a spot size of 40 µm. A standard NiS bead (PGE-A) with 199 ppm Os [62] and 187 Os/ 188 Os = 0.1064 [63], along with a native osmium (i.e., Os 1.0 ) from the Guli massif [64], were analyzed between the samples to monitor and correct for mass-fractionation. The isobaric interference of 187 Re on 187 Os was corrected by measuring the 185 Re peak and using 187 Re/ 185 Re = 1.6742. All analyzed grains have very low Re contents (typically, 185 Re/ 188 Os < 0.0003), allowing for the isobaric interference of 187 Re on 187 Os to be accurately and precisely corrected for. The Os isotopic data were reduced using the Nu Plasma time-resolved software, which allows the selection of the most stable intervals of the signal for integration. The external reproducibility of 187 Os/ 188 Os for the PGE-A standard during the period of data collection was 0.10652 ± 0.00013 (0.12% relative; 2SD; n = 15). Repeated analyses of a crystal of native osmium, which has been used to estimate the accuracy of the LA MC-ICP-MS measurements, yielded 187 Os/ 188 Os = 0.12452 ± 0.00004 (2SD, n = 27).
For one grain of Ru-Os-Ir alloy (sample T-2), the osmium isotopic composition was determined via negative thermal ionization mass-spectrometry (N-TIMS: [65]) at the IGL, following the analytical procedures detailed in Puchtel et al. [66,67]. To obtain the Os isotopic data, a~1 mg grain of the Ru-Os-Ir alloy, 6 mL of Os-purged, triple-distilled concentrated HNO 3 , and 3 mL triple-distilled concentrated HCl were sealed in a chilled 25 mL Pyrex™ borosilicate Carius Tube and heated to 270 • C for 96 h. Osmium was extracted from the aqua regia acid solution by CCl 4 solvent extraction [68], back-extracted into HBr, and purified via microdistillation [69].
The high-precision measurements of the 186 Os/ 188 Os and 187 Os/ 188 Os ratios were performed in static mode on the Faraday cups of a ThermoFisher Triton ® mass spectrometer. The sample load was run twice; during each run, 1800 ratios were collected, and the results from the two runs were averaged; the resultant in-run uncertainty on the measured 186 Os/ 188 Os and 187 Os/ 188 Os was 4.0 ppm for both ratios (2SE). The possible isobaric interference of 186 W 16  ; this long-term reproducibility was used to assess the true uncertainty on the measured 186 Os/ 188 Os and 187 Os/ 188 Os ratios for the Ru-Os-Ir alloy sample analyzed, also accounting for the fact that the sample was run twice. Since the high-precision 186 Os/ 188 Os ratios obtained at IGL were previously instrumental bias-corrected to a common Johnson-Matthey Os standard 186 Os/ 188 Os value of 0.1198475 [66,67,70,71], the 186 Os/ 188 Os ratio measured in this study was also bias-corrected to the same value using correction coefficient of 1.0000354.
The initial γ 187 Os values for the alloy grains measured by LA ICP-MS were calculated as the per cent (%) deviation of the isotopic composition at the time of the formation of the Ru-Os-Ir alloy grains at 740 Ma relative to the chondritic reference of Shirey and Walker [72] at that time and using the 187 Re/ 188 Os ratios obtained from the LA ICP-MS analysis of the respective alloy grains. To calculate the initial γ 187 Os value for the alloy grain analyzed by N-TIMS, we used the average 187 Re/ 188 Os ratio obtained for all alloy grains from the Kunar placer deposit in this study. The average chondritic Os isotopic composition at the time of the alloy formation was calculated using the 187 Re decay constant λ = 1.666 × 10 −11 year −1 , an early Solar System initial 187 Os/ 188 Os = 0.09531 at T = 4558 Ma, and 187 Re/ 188 Os = 0.40186 [72,73]. To calculate the initial 186 Os/ 188 Os ratio for the Ru-Os-Ir alloy grain analyzed by N-TIMS, the average 190 Pt/ 188 Os ratio obtained from the LA ICP-MS analysis of all alloy grains from the Kunar placer deposit and the 190 Pt decay constant λ = 1.477 × 10 −12 year −1 [74] were used. The initial µ 186 Os value was calculated as the part per million (ppm) deviation of the 186 Os/ 188 Os ratio of the alloy sample at 740 Ma relative to the chondritic reference of Brandon et al. [15] at that time. The latter was calculated using an early Solar System initial 186 Os/ 188 Os = 0.1198269 at T = 4567 Ma and 190 Pt/ 188 Os = 0.00174.
It should be noted that, due to the very low 187 Re/ 188 Os and 190 Pt/ 188 Os ratios in all alloy grains analyzed, age corrections of measured 186 Os/ 188 Os and 187 Os/ 188 Os for radioactive decay of 190 Pt and 187 Re were minimal and usually below the in-run uncertainty of individual analyses.

Results
Investigation of the PGM grains at the Kunar and Unga placer deposits revealed their significant diversity ( Figure 3). Both localities are characterized by dominance of solitary grains of osmium (9 and 18 grains at Kunar and Unga, respectively) and iridium (6 and 5 grains, respectively), which prevail over those of ruthenium (5 grains) and rutheniridosmine (3 grains We suggest the presence of ferroan platinum (disordered face centered cubic structure Fm3m) instead of isoferroplatinum (ordered primitive cubic structure Pm3m), although we did not confirm this by X-ray crystallography data [76].  Table 1. Circles denote locations of laser ablation MC-ICP-MS analyses listed in Table 3; numbers in numerator and denominator correspond to the 187 Os/ 188 Os composition and the measurement error, respectively. (Ir, Os, Ru)-native iridium, RIO-rutheniridosmine, (Ru, Os, Ir)-native ruthenium, (Os, Ir, Ru)-native osmium, Ol-high-Mg olivine. Blue square shows an area detailed in Figure 4.  Table 1. Circles denote locations of laser ablation MC-ICP-MS analyses listed in Table 3; numbers in numerator and denominator correspond to the 187 Os/ 188 Os composition and the measurement error, respectively. (Ir, Os, Ru)-native iridium, RIO-rutheniridosmine, (Ru, Os, Ir)-native ruthenium, (Os, Ir, Ru)-native osmium, Ol-high-Mg olivine. Blue square shows an area detailed in Figure 4.

Osmium Isotope Data
The Os-isotope data for the Ru-Os-Ir alloys from the Kunar and Unga placer deposits are listed in Table 2. The in situ Os-isotope data for 13 Os-rich alloy grains from Kunar show a narrow range of measured 187 Os/ 188 Os values between 0.12033 ± 0.00004 and 0.12244 ± 0.00003, 187 Re/ 188 Os mainly lower than 0.0003, with a mean initial 187 Os/ 188 Os value of 0.1218 ± 0.0010 (2SD, n = 13) and the calculated average γ 187 Os(740 Ma) = −0.18 ± 0.85 (2SD, Table 3). These LA ICP-MS data are indistinguishable, within their respective uncertainties, from the N-TIMS data for the Ru-Os-Ir alloy sample T-2 at Kunar (initial 187 Os/ 188 Os = 0.1218463 ± 0.0000015 with γ 187 Os(740 Ma) = −0.1500 ± 0.0012, Table 4). The PGM assemblage at Unga is characterized by a larger degree of Os-isotope variations ( 187 Os/ 188 Os values range from 0.11848 to 0.12395), with a mean value of 0.1222± 0.0025, γ 187 Os(740 Ma) = +0.10 ± 2.1 (2SD, n = 18, Table 3). The uncertainties on the isotopic ratios and the initial Os isotopic composition are quoted at 2SD based on the long-term reproducibility of the Johnson-Matthey Os standard at IGL [67]. The initial µ 186 Os and γ 187 Os values were calculated for the time of formation of the Ru-Os-Ir alloys at 740 Ma using the parameters specified in the text. The 187 Re/ 188 Os and 190 Pt/ 188 Os ratios used to calculate the initial 186,187 Os/ 188 Os isotopic compositions are the average values for all the Kunar Ru-Os-Ir grains obtained by the LA ICP-MS analysis.

Composition of Ru-Os-Ir and Pt-Fe Alloys at Kunar and Unga
The Ru-Os-Ir alloys were classified according to the nomenclature of Harris and Cabri [75]. In descending order, they are represented by native osmium, iridium, ruthenium, and rutheniridosmine, with considerable variation in Os, Ir, and Ru abundances from one grain to another ( Figure 5, Table 1, an. [1][2][3][4][5][6][7][8]. The sum of Os + Ir + Ru + Pt in an alloy is usually >99 wt.% (Table 1) We suggest the presence of ferroan platinum (disordered face centered cubic structure Fm3m) instead of isoferroplatinum (ordered primitive cubic structure Pm3m), although we did not confirm this by X-ray crystallography data [76].
The PGM assemblage at Unga is characterized by a larger degree of Os-isotope variations ( 187 Os/ 188 Os values range from 0.11848 to 0.12395), with a mean value of 0.1222 ± 0.0025, γ 187 Os(740 Ma) = +0.10 ± 2.1 (2SD, n = 18, Table 3). The uncertainties on the isotopic ratios and the initial Os isotopic composition are quoted at 2SD based on the long-term reproducibility of the Johnson-Matthey Os standard at IGL [67]. The initial µ 186 Os and γ 187 Os values were calculated for the time of formation of the Ru-Os-Ir alloys at 740 Ma using the parameters specified in the text. The 187 Re/ 188 Os and 190 Pt/ 188 Os ratios used to calculate the initial 186,187 Os/ 188 Os isotopic compositions are the average values for all the Kunar Ru-Os-Ir grains obtained by the LA ICP-MS analysis.
Overall, the 187 Os isotope results identify a restricted range of broadly similar 187 Os/ 188 Os values for PGM assemblages at both Kunar and Unga (Figures 6 and 7, Table 3). Similarly, the average γ 187 Os(T = 740 Ma) values of PGM assemblages at Kunar and Unga are indistinguishable within uncertainty from each other (−0.18 ± 0.85 and +0.10 ± 2.1, respectively, Table 3). The combined average initial 187 Os/ 188 Os ratio for all grains analyzed in this study is 0.1220 ± 0.0020 (γ 187 Os(740 Ma) = −0.02 ± 1.6 (2SD)), which is indistinguishable from the chondritic reference value at that time ( Figure 6).

Provenance of the Ru-Os-Ir Alloys
The provenance of the investigated Ru-Os-Ir alloy grains can be constrained by the close spatial association of the PGM placer deposits with the Kunar dunite-harzburgite complex (Figure 2). The Ru-Os-Ir alloys at Kunar and Unga show similar compositional signatures dominated by osmium and iridium over ruthenium and rutheniridosmine, which is a feature typical of ophiolite-type complexes [77,78]. The primary nature of Ru-Os-Ir alloys is supported by the occurrence of euhedral inclusions of high-Mg olivine Figure 7. Histogram of Os isotopic compositions of the Ru-Os-Ir alloys from (a) Kunar and Unga (this study, n = 32), (b) Unga (LA MC-ICP-MS data from this study and N-TIMS data from [17], n = 46), and (c) Kunar and Unga (data from this and [17] studies, n = 60).

Provenance of the Ru-Os-Ir Alloys
The provenance of the investigated Ru-Os-Ir alloy grains can be constrained by the close spatial association of the PGM placer deposits with the Kunar dunite-harzburgite complex ( Figure 2). The Ru-Os-Ir alloys at Kunar and Unga show similar compositional signatures dominated by osmium and iridium over ruthenium and rutheniridosmine, which is a feature typical of ophiolite-type complexes [77,78]. The primary nature of Ru-Os-Ir alloys is supported by the occurrence of euhedral inclusions of high-Mg olivine (Fo [92][93] ) that fall within the compositional range of mantle olivine (Fo [88][89][90][91][92][93], as represented by the composition of olivine (with a pronounced peak between Fo 93 and Fo 94 ) in mantle peridotite xenoliths on Archean cratons [79][80][81]. The common occurrence of euhedral inclusions of Pt-Fe alloys in Os-Ir-(Ru) alloys is interpreted in light of the large miscibility gaps in the binary systems of Os-Ir, Ir-Os, and Ir-Pt [82][83][84] indicative of their high-temperature origin. Finally, the presence of a ruthenium trend in the mineral compositions of Ru-Os-Ir alloys ( Figure 5) is indicative of high temperature and pressure values that can only be reached under mantle conditions [85]. These data present solid evidence that the Ru-Os-Ir alloys formed under the high P-T conditions and that the observed chemical variations represent primary features of the grains. Thus, the Ru-Os-Ir alloys at Kunar and Unga are considered to be representative of the mantle material derived from the mantle section of the Kunar ophiolitic complex.

Os Isotopic Composition of the Mantle as Evidenced by the Ru-Os-Ir Alloys
The primary nature of the Ru-Os-Ir alloys and their formation at high temperatures implies that the Os-isotopic composition of these PGMs reflects that of their source region at the time of their formation. The LA MC-ICP-MS data from this study show similar average initial 187 Os/ 188 Os values for both PGM assemblages at Kunar and Unga (0.1218 ± 0.0010, γOs(740 Ma) = −0.18 ± 0.84, and 0.1222 ± 0.0025, γOs(740 Ma) = +0.10 ± 2.1, respectively, Table 3, Figure 7a). These values are identical, within uncertainty, to the initial 187 Os/ 188 Os value for the Ru-Os-Ir alloy obtained by N-TIMS (0.1218463 ± 0.0000015, γOs(740 Ma) = −0.1500 ± 0.0012, Table 4). The combined average initial 187 Os/ 188 Os ratio for all grains analyzed in this study is 0.1220 ± 0.0020 (γ 187 Os(740 Ma) = −0.02 ± 1.6, 2SD), which is identical to that for Ru-Os-Ir alloys from the Unga deposit studied previously [0.1220 ± 0.0054, n = 28, Figure 7b]. No correlation between the chemical and Os isotopic composition of the PGM was found. This is also evidenced by the restricted range of 187 Os/ 188 Os values for intimately intergrown native iridium and rutheniridosmine pair that forms part of the primary PGM assemblage at Kunar (0.12205-0.12213, Figure 3a).
For the Ru-Os-Ir alloy sample T-2 analyzed to the highest precision, and which is also representative of the entire population of alloy grains from the Kunar and Unga placer deposits, a Re-Os model age can be calculated. Since the studied alloy grain has a very low estimated Re content ( 187 Re/ 188 Os = 0.00024 ± 10), we estimated the T RD model age, which is calculated assuming that a melt-depletion event removed all Re from the sample at the time of its formation, and, hence, growth of 187 Os was completely terminated at that time [72]. This calculated T RD model age is 764 ± 2 Ma, which is consistent with the estimate for the age of the Kunar dunite-harzburgite complex at 740 Ma [56,58].
Using the accepted emplacement age of 740 Ma, the long-term 187 Re/ 188 Os ratio, with which the mantle source of the Kunar dunite-harzburgite complex evolved, can be estimated. In order to model the time-integrated evolution of Re/Os in the Kunar mantle source, the average initial 187 Os/ 188 Os ratio of the studied Ru-Os-Ir alloy grain samples of 0.1220 ± 20 has been used to calculate the minimum 187 Re/ 188 Os ratio required to evolve to this Os isotopic composition by 740 Ma, and assuming formation of this mantle domain shortly after the start of the Solar System. Evolution of the Kunar mantle source from an early Solar System 187 Os/ 188 Os = 0.09531 at 4558 Ma [72,73] to the initial 187 Os/ 188 Os ratio of 0.1220 ± 20 at 740 Ma requires 187 Re/ 188 Os ratio of 0.401 ± 0.030. This time-integrated 187 Re/ 188 Os ratio for the Kunar mantle source is within the range of that for chondritic meteorites (a bulk chondrite average 187 Re/ 188 Os = 0.410 ± 0.051 (±2SD), as compiled from the data of Walker et al. [91] and Fischer-Gödde et al. [92]).
We note that Os-isotope data at Kunar and Unga are consistent with a Neoproterozoic age for the formation of the Chelyuskin ophiolite, which is correlated with coeval ophiolites of other Arctic regions that mark the opening of the Paleo-Pacific Ocean and the breakup of the Neoproterozoic supercontinent Rodinia between 900 and 700 Ma [93].
The initial 186 Os/ 188 Os value of 0.1198409 ± 0.0000012 obtained for the Ru-Os-Ir alloy sample T-2 at Kunar is 34 ± 10 ppm higher than this value in the chondritic reference of Brandon et al. [15] at that time. This indicates evolution of the mantle source with timeintegrated suprachondritic Pt/Os ratio. Using the early Solar System 186 Os/ 188 Os = 0.1198269 at 4567 Ma [15] requires a source with a 190 Pt/ 188 Os = 0.00247 ± 21 to have evolved to its 186 Os/ 188 Os = 0.1198409 ± 12 at 740 Ma. Thus, in contrast to the calculated Re/Os ratio, the required minimum 190 Pt/ 188 Os ratio is~40% higher than the average 190 Pt/ 188 Os = 0.00180 ± 17 (2SD) in bulk chondritic meteorites, as compiled from the data of Horan et al. [94], Brandon et al. [15,95], Fischer-Gödde et al. [92], and van Acken et al. [96].