Synsedimentary to Diagenetic Cu ± Co Mineralization in Mesoproterozoic Pyritic Shale Driven by Magmatic-Hydrothermal Activity on the Edge of the Great Falls Tectonic Zone – Black Butte, Helena Embayment, Belt-Purcell Basin, USA: Evidence from Sulfide Re-Os Isotope Geochemistry

The ca. 1,500 to 1,325Ma Mesoproterozoic Belt-Purcell Basin is an exceptionally preserved archive of Mesoproterozoic Earth and its paleoenvironmental conditions. The Belt-Purcell Basin is also host to world-class base metal sediment-hosted mineralization produced in a variety of settings from the rift stage of basin evolution through the subsequent in ﬂ uence of East Kootenay and Grenvillian orogenies. The mineral potential of this basin has not been fully realized yet. New rhenium-osmium (Re-Os) data presented here for chalcopyrite, pyrite, and black shale contribute to re ﬁ ne a robust genetic model for the origin of the Black Butte copper ± cobalt ± silver ( Cu ± Co ± Ag ) deposit hosted by the ca. > 1,475Ma Newland Formation in the Helena Embayment of the Belt-Purcell Basin in Montana, USA. Chalcopyrite Re-Os data yield an isochron age ( 1,488 ± 34 Ma , unradiogenic initial 187 Os/ 188 Os composition Os i ‐ chalcopyrite = 0 : 13 ± 0 : 11 ) that overlaps with the geological age of the Newland Formation. Further, the Re-Os data of synsedimentary to diagenetic massive pyrite yield evidence of resetting with an isochron age ( 1,358 ± 42 Ma ) coincident with the timing of the East Kootenay orogeny. The unradiogenic Os i-chalcopyrite at ca. 1,488Ma ( 0 : 13 ± 0 : 11 ) argues for derivation of Os from a magmatic source with a 187 Os/ 188 Os isotopic composition inherited from the upper mantle in the Mesoproterozoic ( Os mantle1 , 475Ma = 0 : 12 ± 0 : 02 ). The unradiogenic Os i-chalcopyrite also suggests limited contamination from a continental crustal source. This source of Os and our new sulfur isotopic signatures of chalcopyrite ( – 4.1 to +2.1 ‰ -VCDT) implies a dominantly magmatic source for metals. We integrate our new results and previously published geological and geochemical evidence to conceptualize a genetic model in which Cu and metals were largely contributed by moderate-temperature, reduced magmatic-hydrothermal ﬂ uids carrying reduced sulfur species with a magmatic origin and ﬂ owing as highly metalliferous ﬂ uids within the shale sequence. A subsidiary derivation of metals during thermally forced shale diagenesis is possible. Chalcopyrite mineralization replaced locally massive synsedimentary to diagenetic pyrite units close to the sediment-water interface, i.e., an ideal locus where magmatic-hydrothermal ﬂ uids could cool and the solubility of chalcopyrite would fall. We suggest that Cu mineralization was coeval with the timing of an enhanced thermal gradient in the Helena Embayment triggered until ca. 1,455Ma by tholeiitic dike swarm that intruded into Archean basement rocks and intersected the NE-SW-trending Great Falls Tectonic Zone.


Introduction
A comprehensive knowledge of the timing and lithostratigraphic position of hydrothermal fluid flow and mineralization processes throughout the sedimentary infill of intracontinental rifts is paramount in the identification of sedimentary basins that may contain giant sediment-hosted base metal mineralization (e.g., [1]). The preponderance of strata-bound massive pyrite deposits with economic concentrations of base metals (copper (Cu), zinc (Zn), and lead (Pb)) within Mesoproterozoic strata has been tied to increasing marine concentrations of sulfate following the rise of atmospheric oxygen [2][3][4][5][6]. In a graben extending eastward from the center of the Mesoproterozoic Belt-Purcell Basin, Canada-USA (Helena Embayment; Figure 1), the Black Butte (formerly known as Sheep Creek) copper ± cobalt ± silver (Cu ± Co ± Ag) deposit is hosted by massive pyritic shale of the Newland Formation. The Black Butte deposit is thought to have formed during diagenesis in connection with the interaction between biogenically and/or thermally sourced reduced sulfur and metalliferous hydrothermal fluids during basin development [6,7].
The conceptualization of a genetic model for base metal sediment-hosted deposits is based on the identification of six key components: (i) the origin of fluids and their chemistry (redox conditions-reduced versus oxidized fluids, temperatures, and type of ligand to transport metals); (ii) the source(s) of metals (e.g., Cu, Zn, and Pb ± Co ± Ag); (iii) the trigger(s) for a decrease of the solubility of metals; (iv) a reducing agent and/or source of reduced sulfur to fix the metals, in particular in the case of oxidized metalliferous fluids; (v) a geodynamic engine responsible for fluid flow and convergence of the components listed above for the mineralizing processes, and (vi) the absolute timing of base metal mineralization that holds a central role in connecting points (i) to (v). Here, we attempt to constrain the timing of chalcopyrite mineralization and the source of metals as well as assess possible controls on mineralization exerted by (i) continental tholeiitic magmatism coeval with the development of the Belt-Purcell intracontinental rift and (ii) tectonic structures in the long-lived Great Falls Tectonic Zone (GFTZ) in the Archean Wyoming Province, on the southern edge of the Helena Embayment. To this end, we present new petrographic evidence, Re-Os and sulfur isotope data for chalcopyrite associated with massive pyrite beds and intercalated black shale and siltstone, complemented by new Re-Os isotope data for the host pyritic shale sequence at Black Butte.  [8][9][10][11], extend over 25 km along the Volcano Valley-Buttress fault system, and 8 km southward in the Newland Formation within a structural block along a major bend in the fault system ( Figure 2(a)) [7]. This structural block is located on the northern margin of the Helena Embayment, which is an asymmetric, E-NE-trending branch of the main Mesoproterozoic Belt-Purcell Basin in Canada and the USA ( Figure 1) [12]. The Belt-Purcell Basin was initiated at ca. 1,500 to 1,485 Ma as an intracontinental rift during breakup of the Nuna supercontinent [1,13,14]. The Helena Embayment comprises a shallow carbonate shelf in the north, with a southern edge related to a steep and active fault scarp against the uplifted basement of the Archean Wyoming Province [12]. A NW-trending, low-K tholeiitic dike swarm dated at ca. 1,455 Ma (i.e., currently available best age estimate from whole-rock Rb-Sr geochronology) lies within Archean basement rocks to the south of the Helena Embayment and the Belt-Purcell Basin ( Figure 1) [12,15]. The dike swarm has the same strike orientation as ca. 1,468 Ma gabbroic sills found in rift-facies marine turbidites in the main Belt-Purcell Basin, in particular within footwall sediments of the ca. 1,475 Ma synsedimentary Sullivan deposit (U-Pb TIMS cassiterite geochronology, [16]) hosted in the Lower Aldridge Formation, southeastern British Columbia, Canada ( Figure 1) [1,17,18]. The Newland Formation is interpreted as (1) being slightly younger than the Prichard Formation (i.e., lateral equivalent to the Lower Aldridge Formation) and (2) older than the >1,454 Ma Revett Formation (SHRIMP U-Pb geochronology of bentonite) [19] in northwestern Montana and Idaho in the main Belt-Purcell Basin ( [20]; detailed stratigraphic correlations in [7], their Figure 2). In light of these stratigraphic correlations with the northernmost and westernmost part of the main Belt-Purcell basin, maximum and minimum ages of ca. 1,470 and 1,454 Ma have been reported and utilized for the Newland Formation in previous works [7]. However, new detrital zircon U-Pb geochronology data from the northern tip of the main Belt-Purcell basin in British Columbia suggest that the deposition of the Creston Formation, which overlies the Upper Aldridge Formation [7], was rapid between ca. 1,475 and 1,467 Ma [21]. The base of the Creston Formation is directly correlated with the base of the Spokane Formation which overlies the Newland Formation in the northern Helena Embayment, in particular in the Black Butte area. Therefore, the Newland Formation may be older than ca. 1,475 Ma.  [7,12]. In addition, N-trending synsedimentary faults steered debris-flow sedimentation that, in turn, controlled fluid flow and the formation of synsedimentary to early diagenetic massive to bedded fine-grained pyrite units within topographic lows flanked by shale horsts (Figure 2(a)) [7]. The sedimentary to early diagenetic nature of the finegrained pyrite units is supported by the petrography of both matrix-and clast-supported sedimentary breccias that 2 Lithosphere contain poorly sorted, angular, granule-to cobble-sized clasts of the laminated striped shale and pyrite strata, including clasts of broken, pyrite-walled tube structures ( Figure 2(b), e.g., sample 28-SC) [6,7]. The microlaminated, organic-rich black shale of the Newland Formation (1.1 to 1.4 wt.% C at the Black Butte deposit, 0.1 to 3.4 wt.% C outside the Cu-mineralized areas) [9] is interbedded with centimeter-scale homogeneous muddy turbidites [9]. Collectively, these facies record deep-water, organicrich sedimentation [22,23]. Outside of the massive sulfide zones, black shale contains ubiquitous, disseminated pyrite framboids and finely crystalline pyrite blebs within laminae characterized by compactional draping of pyrite framboids [9]. The broad range of δ 34 S values of disseminated pyrite (-8.7‰ to +36.3‰ versus the Vienna Cañon Diablo Troilite (VCDT)) may be compatible with bacterial mechanisms of sulfate reduction [9]. The redox conditions for the mid-Proterozoic ocean suggest that the deposition of the Newland Formation took place in an anoxic but sulfate-bearing (2-6 mM) water column [6,[24][25][26]. The discharge from vents of metalliferous hydrothermal fluids into the overlying, predominantly sulfidic water column resulted in the formation of decimeter-scale tubular structures with pyrite walls [6]. Those vents, which released saline hydrothermal fluids denser than seawater (cf. [27] in [6]), were populated by sulfate-reducing bacteria [9,28,29]. At the sediment-water interface on the seafloor of the Helena Embayment, the interplay between those sulfate-reducing organisms and vents of metalliferous hydrothermal fluids steered the dynamics and redox conditions of bottom waters (i.e., transition from episodic sulfate-rich waters due to episodic marine input to predominantly sulfate-poor sulfidic waters) [ which formed within the surface sediments, characterize the sedimentary and syndiagenetic stage at Black Butte [6]. Early diagenetic barite laths in the laminated black shale (δ 34 S = +14:0 to +18.0‰) [6] precipitated below the sediment-water interface from residual pore-fluid sulfate enriched in 34 S as a result of excess of sulfate reduction with respect to the diffusion of sulfate from coeval seawater (δ 34 S = +10:0 to +12.0‰) into sediments [6,30].

Paleoenvironmental Conditions and Sedimentation in an
A second stage of mineral precipitation coincided with formation of the tube walls with rapid replacement by mottled (δ 34 S mean = +10:3 ± 12:5‰, 1 standard deviation (SD), n = 108, up to +45‰) and colloform (δ 34 S = +10:8 ± 7:1‰, 1 SD, n = 63, up to +34.0‰) pyrite [6]. Rare euhedral marcasite, hypothetically indicating some biogenically induced, transient, locally acidic and oxic conditions [6,[31][32][33], overgrows the pyrite walls. The sulfur isotopic composition of  colloform and mottled pyrite is consistent with quantitative reduction of seawater sulfate that, in turn, favors pyrite precipitation steered by the iron supply from hydrothermal fluids. With subsequent recharge of the water column in sulfate via oxic seawater ingress from the main Belt-Purcell Basin, a rejuvenation of the activity of sulfate-reducing bacteria took place, thereby inducing mottled pyrite precipitation with highly variable sulfur isotopic compositions [6].

Proposed Timing of Cu Mineralization.
In the Belt-Purcell Basin, metalliferous hydrothermal and magmatichydrothermal fluids circulated in the main branch of this basin at least from the time of Zn-Pb-Ag mineralization in the Sullivan deposit at ca. 1,475 Ma (U-Pb TIMS cassiterite age) [16] to the time of magmatic-hydrothermal Cu-Co-Au mineralization at ca. 1,349 Ma in the Idaho cobalt belt (Re-Os isochron age of cobaltite) [39]. At Black Butte, Cu mineralization is restricted to the massive sulfide zones (Upper, Middle, and Lower Sulfide Zones; Figures 2(a) and 2(b)), in which massive pyrite layers are present [7,9]. The black shale interbedded with massive pyrite beds is bereft of sulfides other than disseminated pyrite framboids and blebs of bacterial origin [9]. Not only does the black shale lack base metal sulfides (outside some rare replacement of disseminated pyrite by chalcopyrite) [40] but there is also no petrographic evidence for brittle-style veins and stockworks that could have formed as the result of fluids introduced epigenetically in the Newland Formation (Tintina Montana, Inc. staff, pers. comm., 2015).
Copper mineralization is proposed to have formed in the subsurface within the sedimentary sequence during continuous basin development and burial of the hydrothermal vents while diagenetic barite was being thermally reduced [6,7]. Copper mineralization predates deformation and reactivation of the Volcano Valley fault during the Cretaceous to Paleocene Laramide orogeny [7,19]. Therefore, the absolute age of the Cu ± Co ± Ag mineralization might span the timing of diagenesis of the Newland Formation and the transition to orogenic sedimentation and burial metamorphism during the East Kootenay orogeny and termination of sedimentation in the Belt-Purcell Basin [20,41]. In addition, the Re-Os isotopic composition was determined on one pyrite sample from the massive fine-grained pyrite, between the MSZ and USZ, which formed during the earliest (i.e., synsedimentary) sulfide generation, and one finegrained pyrite sample associated with barite at the top of the USZ. The shale sample, which lacks pyrite and chalcopyrite, was collected below the MSZ (Figure 2(b)); this sample was subdivided into two 5 cm long shale subsamples ("A, B," ca. 20 g each). The two shale subsamples were processed to determine their Re-Os isotopic composition in order to calculate the initial 187 Os/ 188 Os isotopic composition of the shale for the age interval proposed for deposition of the Newland Formation.

Sulfide Petrography and Quality Control of Mineral
Separates. Polished thin sections of chalcopyrite-and/or pyrite-bearing samples were studied by transmitted and reflected light microscopy in order to establish paragenetic relationships. These relationships served as a basis for optimizing sample preparation, including determination of the step-wise methodology using a Frantz Isodynamic Separator at various currents to produce monomineralic mineral separates (see Section 3.3). After sample preparation, an aliquot 5 Lithosphere of each chalcopyrite mineral separate was embedded in epoxy to determine qualitatively the purity of the mineral separates. These mounts were studied by using a reflected light microscope.

Preparation of Sulfide Mineral
Separates. All samples were cut into slabs that were thoroughly cleaned using silicon carbide grit, ultrapure (18.2 MΩ) water, and ethanol to remove any metal traces introduced by hammering or sawing. The shale sample was split into two aliquots that were powdered in an automated agate mill. All sulfide samples were crushed using a zirconia ceramic dish and puck and sieved through disposable, home-made nylon sieves to produce 70-200 and +70 mesh-size fractions. A Frantz Isodynamic Separator was used to produce magnetic (M) and nonmagnetic (NM) subfractions from the 70-200 mesh fractions by applying successive amp currents with 15°side slope and 10°forward slope (protocol adapted from [39]), as follows: (1) 0.3 and 0.6 amp currents were used to collect pyrite fractions from the pyrite-only samples and (2) a 1.3 amp current was used for the chalcopyrite-bearing samples from which chalcopyrite was collected in the M1.3 fraction. The sulfide species were then isolated from remaining gangue minerals into final sulfide mineral separates by heavy liquid separation using Sodium Polytungstate (SPT, specific gravity 2.86) with a minimum of ten rinses using ultrapure (18.2 MΩ) water, followed by a rinse with ethanol prior to dry down in an oven at 60°C overnight.
3.4. Re-Os Isotope Geochemistry. For each analysis, between 200 and 550 mg of sulfide (pyrite or chalcopyrite) mineral separate or shale powder was weighed and transferred into a thick-walled borosilicate Carius tube [42]. Each sulfide aliquot was dissolved in inverse Aqua Regia (~3 mL of 11 N HCl and~6 mL of 16 N HNO 3 ), together with a known amount of 185 Re + 190 Os spike solution at 220°C for 24 hours. Each shale aliquot was dissolved in~8 mL of a mixture of CrO 3 -H 2 SO 4 with a known amount of 185 Re + 190 Os spike solution at 220°C for 48 h (Canadian Centre for Isotopic Microanalysis, University of Alberta, Edmonton, Canada, and Laboratory for Sulfide and Source Rock Geochemistry and Geochronology, Durham University, Durham, UK). The Re-Os laboratory protocol used in the present work is described in full in Selby et al. [43] and Cumming et al. [44].
Rhenium and Os isotopic compositions were determined by negative thermal ionization mass spectrometry (N-TIMS) using a Thermo Scientific Triton mass spectrometer at the Canadian Centre for Isotopic Microanalysis, University of Alberta, Edmonton, Canada (pyrite and shale), and at the Arthur Holmes Laboratory, Durham University, Durham, UK (chalcopyrite). Rhenium was measured as ReO 4 in static mode on Faraday collectors; Os was measured as OsO 3 in peak-hopping mode on SEM with a constant flow of oxygen [45,46]. Quality of the measurements of the Re and Os isotopic compositions of pyrite and shale was monitored by repeated analyses of in-house Re ( 185 Re/ 187 Re = 0:59774 ± 0:00065, n = 23) and Os ("AB-2," 187 Os/ 188 Os = 0:10682 ± 0:00009,  Table 2. The presented uncertainties listed in Table 2 include propagated uncertainties in the standard, spike calibrations, mass spectrometry measurements, and blanks.

Sulfur Isotopic Composition of the Chalcopyrite Mineral
Separates. About 5 to 10 mg of chalcopyrite mineral separate was utilized for each sulfur isotopic analysis. Sulfides were analyzed by standard techniques [47]

Finite Mixture Modeling of Re-Os Data and Isochron
Regression. To statistically assess whether the data may belong to multiple regression lines, we iteratively fit a series of finite mixture models to the data, where each finite mixture represents a linear regression with an intercept and slope (i.e., a potential isochron) [48]. We test whether the maximum number of mixture components can lie between 1 and 5 (i.e., 1 to 5 potential and discrete regression lines) and compare the results using a model comparison score to choose the best-fitting model (Integrated Completed Likelihood Criterion, ICL parameter). We utilize the 187 Re/ 188 Os and 187 Os/ 188 Os ratios and the error correlation factor rho for each data point. Those parameters are conventionally used to plot Re-Os data in the 187 Os/ 188 Os vs. 187 Re/ 188 Os space as error ellipses. In the present modeling exercise, we represent each data point of a sulfide aliquot using 100 points to account for the elliptical uncertainty in the 187 Re/ 188 Os and 187 Os/ 188 Os ratios (Figures 3(a) and 4(a)), but assume that all 100 points belong to a single aliquot and its uncertainty ellipse must belong to the same isochron.
In detail, we run the R script provided by Davies et al. [48], which utilizes the Flexmix package [49]. All text in quotation marks that follows represents inputs into the R script [48]. We run the model with 100 points for each sulfide aliquot, to capture the analytical uncertainty in each measurement, but assume that all 100 points belong to a single aliquot and therefore the same isochron. Hence, we set the parameters "iterate = 50" and "random.effects = TRUE". We test for the possibility that the data can be explained by up to five possible isochrons, using the expectationmaximization (EM) algorithm in the Flexmix package. We allow up to 1,000 possible iterations of the EM algorithm to allow it to converge ("n:iter = 1,000"). In our case, all runs for of the EM algorithm converged (Tables 1 and 2). To account for variability in the initiation of the EM algorithm, we repeated 100 times the search for 1 to 5 components ("n:rep = 100").
We used Isoplot v. 4.15 [50] to report all Re-Os dates as Model 1 or Model 3 isochrons through statistical regression of the Re-Os data, which are reported at the 2σ level (95% level of confidence), in the 187 Os/ 188 Os vs. 187 Re/ 188 Os space using the decay constant of 187 Re (λ 187 Re = 1:666e -11 ± 5:165e -14 a -1 ) [51]. Error propagation was also investigated using the Monte Carlo approach [52]. In this technique, a prescribed number of isochrons are constructed from the input data and their corresponding probability density function (analytical uncertainty of the 187 Re/ 188 Os and 187 Os/ 188 Os ratios and their error correlation rho). The Re-Os date and estimate of the initial 187 Os/ 188 Os ratio are cross-plotted yielding a probabilistic distribution that includes an analytical uncertainty. Given the use of error propagation rho, the analytical uncertainty presented above inherently includes both random and systematic uncertainties. Further, model uncertainties include the assumptions that all samples have nearly identical ages and the same initial 187 Os/ 188 Os composition, and for those aliquots, the Re-Os system remained closed.

4.2.
Re-Os Isotope Geochemistry of Black Shale, Pyrite, and Chalcopyrite. The two black shale aliquots have Re and total Os contents of 11.8 to 13.3 ppb and 285 to 324 ppt, respectively (Table 2). Common Os, for which the best approximation is given by contents in 192 Os, represents ca. 14% of total Os contents. The Os budget in these shale aliquots is thus dominated by radiogenic 187 Os. These shale aliquots have relatively similar 187 Re/ 188 Os values (567-588) and highly radiogenic (14.5-15.0) 187 Os/ 188 Os ratios. For a depositional age in excess of 1,475 Ma for the Newland Formation, the calculated range of initial 187 Os/ 188 Os ratios (Os i ) for those aliquots is 0:342 ± 0:004 and 0:354 ± 0:003. These Os i values must be positive and greater than the value of the 187 Os/ 188 Os ratio for the mantle in that age range (Os mantle 1,475 Ma = 0:12 ± 0:02, calculated for the hypothetical primitive Mesoproterozoic upper mantle at 1,475 Ma by using present-day 7 Lithosphere

Evaluation of Re-Os Dates and Timing of Pyrite and
Chalcopyrite Mineralization. The Model 3 apparent "isochrons" for pyrite and chalcopyrite have very high MSWD (MSWD pyrite = 59 and MSWD chalcopyrite = 68) with dates and initial 187 Os/ 188 Os ratios of unknown geological integrity. These apparent "isochrons" are interpreted as resulting from an excess scatter around a regression line because several aliquots may not have had closed Re-Os systematics since the time of sulfide precipitation. Such excess scatter could be related to partial resetting of several aliquots or to multiple time components in a dataset that record discrete geological processes in time and/or initial 187 Os/ 188 Os isotopic composition [48,55].
In the following, we remediate this situation by deciphering and weighing against independent geological evidence, the three mixture components identified for both pyrite and chalcopyrite using the Integrated Completed Likelihood Criterion (ICL criterion) and the cluster analysis (pyrite: Figures 3(a) and 3(b); chalcopyrite: Figures 4(a) and 4(b)). We treat here data for pyrite and chalcopyrite separately. To choose the optimum number of potential mixture com-ponents, we compared the results for different numbers using the ICL criterion. The larger the negative value of the ICL parameter, the better the model statistically describes the data. We plot the ICL parameter for both the pyrite and chalcopyrite data separately (Figures 3(b) and 4(b)). In both cases, it can be seen that the ICL and the other model fitting parameters (i.e., logLik, AIC, BIC) ( [48]; Tables 3 and 4) collectively suggest three mixture components to describe the pyrite and chalcopyrite data, respectively. For the pyrite data, the lowest ICL value is found so that a model of 3 mixture components would sufficiently describe the data (Figure 3(b) and Table 3). In the case of the chalcopyrite data, we choose 3 mixture components, even though 4 provide a slightly lower ICL value. This is because the model for 4 components actually converged to 3 components (Figure 4(b) and Table 4).

Pyrite.
In light of the ICL criterion and cluster analysis, finite mixture modeling identified three discrete components for the Re-Os data of pyrite (Figures 3(a) and 3(b)). For each data cluster, a gross Re-Os date estimate is based on linear regression of the number of data points (n) belonging to each cluster (Figure 3(a)). Those estimates range from 1,354 Ma (cluster 3, n = 4, 36.4% of the data, samples py-05 and py-17, modeled Os i = 0:82), 1,310 Ma (cluster 1, n = 4, 36.4% of the data, sample py-05, modeled Os i = 1:41), and 1,174 Ma (cluster 2, n = 3, 27.2% of the data, sample py-05, modeled Os i = 1:55). None of those gross date estimates coincide with the age of deposition of the Newland Formation at ca. >1,475 Ma comprising fine-grained sediments and associated synsedimentary to diagenetic pyrite. To interpret these results, we computed Re-Os isochron dates for clusters with at least four defining data points using Isoplot v.4.15 [50] and the Monte Carlo simulation-based application [52].
The gross Re-Os estimate defined by three data points in "cluster 2" at ca. 1,174 Ma is inconclusive. This date might be a pseudoisochron age. Therefore, more work is warranted to investigate if this date could be related to a partial resetting of Re-Os systematics in pyrite (e.g., [55]) caused by thermal events during the Grenvillian orogeny recognized worldwide between ca. 1,190 and 1,006 Ma [17,39,41,56,58,61,62]. This Grenvillian event, which has been recently confirmed to have impacted geological units located in northwestern present-day North America, was responsible for widespread hydrothermal activity causing mineral precipitation (or recrystallization) in the Belt-Purcell Basin between ca. 1,160 and 1,040 Ma [21,39,59,63].
In conclusion, we infer that thermal events associated with the East Kootenay orogeny may have impacted and caused partial resetting of Re-Os systematics in synsedimentary to diagenetic pyrite with an assumed age equivalent to the depositional age of the Newland Formation. Due to tectonic and thermal events that partially reset the Re-Os clock in discrete pyrite samples, the initial 187 Os/ 188 Os ratio in pyrite increased from an unknown value at the time of synsedimentary to diagenetic pyrite mineralization at ca. >1,475 Ma to 0:81 ± 0:28 at ca. 1,358 Ma. This evolution of initial 187 Os/ 188 Os ratios identifies a crustal contribution in radiogenic 187 Os during the East Kootenay orogenic episode and associated (magmatic-)hydrothermal events affecting the Re-Os systematics of an original synsedimentary to diagenetic pyrite.

Chalcopyrite.
In light of the ICL criterion and cluster analysis, finite mixture modeling for Re-Os data of chalcopy-rite identified three discrete components (Figures 4(a) and  4(b)). The gross Re-Os date estimates are based on linear regressions of an even number of aliquots (n = 4, 33.3% of the data in each cluster): ca. 1,483 Ma for "cluster 3" (samples cpy-03, cpy-25, cpy-29, and cpy-35, modeled Os i = 0:13); ca. 1,449 Ma for "cluster 1" (samples cpy-03 and cpy-25, modeled Os i = 0:00); and ca. 1,048 Ma for "cluster 2" (samples cpy-24 and cpy-28, modeled Os i = 1:29). As done for pyrite, we computed Re-Os isochron dates for clusters with at least four defining data points using Isoplot v. 4.15 [50] and the Monte Carlo simulation-based [52]. Data in "cluster 3" yield a Model 3 isochron with a Re-Os date of 1,488 ± 62 [62] Ma, (2σ, MSWD = 12) and an initial 187 Os/ 188 Os ratio of 0:13 ± 0:04 (Figure 4(c)). The Monte Carlo simulation of this same data cluster shows that the analytical and model uncertainties contribute to 37% and 63% of the total uncertainty, respectively. This Monte Carlo simulation yields a Re-Os date of 1,488 ± 34 [34] Ma and an initial 187 Os/ 188 Os ratio of 0:13 ± 0:11. This Re-Os date of 1,488 ± 34 [34] Ma overlaps with the geological age of the Newland Formation at ca. >1,475 Ma. Hence, we interpret that chalcopyrite mineralization was potentially contemporaneous (within uncertainty) with sedimentation and diagenesis of the Newland Formation. This interpretation is compatible with independent conclusions proposing a diagenetic origin for chalcopyrite mineralization [7] and its connection with magmatic-hydrothermal activity at and below the sediment-water interface [6].
The regression of data belonging to "clusters 1 and 2" indicates possible disturbance of these samples following the initial closure of the Re-Os isotopic system. Data in "cluster 1" yield a Model 1 isochron with a Re-Os date of 1,460 ± 79 [80] Ma (2σ, MSWD = 0:02) with a geologically unrealistic initial 187 Os/ 188 Os ratio of -0:01 ± 0:10 and a Monte Carlo simulation with Re-Os date of 1,454 ± 113 [113] Ma and an initial 187 Os/ 188 Os ratio of 0:00 ± 0:29 (Supplementary Figure 2). We interpret those unrealistic date and Os i as resulting from a limited disturbance of the Re-Os systematics in the samples belonging to "cluster 1." Data in "cluster 2" define a Model 3 isochron with a highly imprecise Re-Os date of 1,055 ± 264

Sources of Osmium and Reduced Sulfur in Copper
Mineralization. The initial 187 Os/ 188 Os isotopic composition of sulfides may be utilized to trace the source of Os and by inference other metals (e.g., Cu), from potential source(s) [64,65]. This evaluation is possible for a Re-Os dataset with limited scatter and uncertainty of the initial 187 Os/ 188 Os isotopic composition, e.g., the Os i value of 0:13 ± 0:11 for the ca. (ii) the NE-SW-trending GFTZ, which is a tectonic lineament that separates the 3.6 to 2.5 Ga Wyoming Craton to the southeast from the mixed Archean (3.6-2.5 Ga) and Paleoproterozoic (ca. 1.8 Ga) Medicine Hat Craton to the northwest [7,70]. In situ laser ablation inductively coupled mass spectrometry U-Pb ages of zircon in gneiss and amphibolite have identified the presence of 3.2 to 2.4 Ga Archean protoliths that were involved in tectono-metamorphic processes at ca. 1.86 Ga and ca. 1.78 Ga within the GFTZ [70]. At ca. 1.8 Ga, in connection with creation of the GFTZ, this Proterozoic event involved fluid/melt that raised the Re/Os ratio of the subcontinental lithospheric mantle [76]. As such, the long-lived Paleoproterozoic and Archean basement of the GFTZ likely comprises lithologies having high Re/Os ratios. This interpretation is compatible with a lead component that was uniquely identified in barren and mineralized sedimentary belt rocks in the tectonically controlled Helena Embayment. Lead was there derived from the early Archean continental crust in the adjacent Wyoming Craton [77].
Considering as a lower limit the 187 Os/ 188 Os isotopic composition of the primitive Mesoproterozoic upper mantle at ca. >1,475 Ma (Os mantle 1,475= 0:12 ± 0:02), the Os i value of 0:13 ± 0:11 for the ca. 1,488 Ma Re-Os isochron age for synsedimentary to diagenetic chalcopyrite mineralization defines a range of possible initial 187 Os/ 188 Os ratios for chalcopyrite from 0.10 to 0.24 ( Figure 5). This range of initial 187 Os/ 188 Os ratios for chalcopyrite does not overlap with either the radiogenic initial 187 Figure 5). In conclusion, the range of potential initial 187 Os/ 188 Os ratios for chalcopyrite from 0.11 to 0.24 is compatible with (1) a derivation of Os from a markedly unradiogenic magmatic source with a 187 Os/ 188 Os isotopic composition inherited from the upper mantle in the Mesoproterozoic (Os mantle 1,475= 0:12 ± 0:02) and (2) limited contamination in radiogenic 187 Os from a continental crustal source with an estimated 187 Os/ 188 Os isotopic composition of 0:30 ± 0:23 that could cause an increase of initial 187 Os/ 188 Os ratios for chalcopyrite up to 0.24.
In light of (1) our petrographic observations, (2) the unradiogenic, dominantly mantle-derived to low radiogenic source of Os in synsedimentary to early diagenetic Cu mineralization, and (3) considering that Cu mineralization occurred in the subsurface during continuous basin development and burial of the hydrothermal vents ( [6]; this study), a magmatic source of sulfur could be permitted for chalcopyrite mineralization. Indeed, Lyons et al. [9] suggested that a narrow range of δ 34 S signatures clustering around 0‰ in chalcopyrite may be indicative of a magmatic source of sulfur. Pyrite originally precipitated at the site of hydrothermal vents when metals titrated reduced sulfur produced by the metabolism of microorganisms in the water column with large sulfur isotopic fractionation (in excess of 50‰) in the immediate vicinity of metalliferous hydrothermal seepages [6]. Then, in the subsurface, thermal reduction of sulfate in barite contributed H 2 S for base metal sulfide precipitation that was 10 to 20‰ lighter than seawater-derived sulfate in barite [6]. The bulk sulfur isotopic signature of chalcopyrite (-2.3‰, Table 2, this study), which replaced barite in sedimentary pyrite-walled tube structures (sample 35-SC), is compatible with this interpretation by Present et al. [6]. Yet, in all other samples analyzed in the present study, i.e., stringer-type, vein-type, varve-like chalcopyrite (Figure 2), in which barite is absent, it is possible that the sulfur was there predominantly of magmatic origin (-4.1 to +2.1‰), possibly through (1) the leaching of H 2 S from magmatic lithologies and (2) the transport of metals and reduced sulfur rendered possible in moderate temperature magmatichydrothermal fluids (cf. [78]) (see next section). Thus, in the complex hydrodynamics of the restricted basin of the Helena Embayment, we contend that sulfur isotopic signatures of chalcopyrite clustering around 0‰ could identify a magmatic source of reduced sulfur in areas of Cu mineralization where evidence lack that barite was reductively dissolved to yield reduced sulfur for Cu mineralization.

Origin of Copper Mineralization.
We propose that copper mineralization at Black Butte resulted from (1) a dominantly magmatic-hydrothermal metal source with subsidiary derivation of metals during thermally forced shale diagenesis and (2) flow of highly metalliferous reduced fluids within the shale sequence. Considering a timing for chalcopyrite mineralization at 1,488 ± 34 Ma overlapping with the revised age of the Newland Formation possibly older than 1,475 Ma [7,21], we suggest that Cu mineralization was coeval with the timing of an enhanced thermal gradient in the Helena Embayment, possibly triggered by the NW-trending dike swarm that intruded into the Archean basement rocks to the south of the Helena Embayment until ca. 1,455 Ma [12,15].
The unradiogenic to low radiogenic initial 187 Os/ 188 Os ratio for chalcopyrite (0.11 to 0.24) reflects that metals were brought by magmatic-hydrothermal seepages, possibly carrying H 2 S credits with a magmatic sulfur signature in moderate-temperature magmatic-hydrothermal fluids. Fluid inclusion microthermometry studies in gangue minerals (quartz, barite, and dolomite) at Black Butte identified moderately saline mineralizing fluids (~15 wt.% NaCl equiv.; [34]). Moderate-temperature, reduced fluids with moderate salinities are able to transport barium, metals (i.e., Cu), and reduced sulfur at moderate temperature (>250°-350°C; [78,79]). Such reduced fluids are likely to have been involved in chalcopyrite mineralization at Black Butte based on the buffering capacity of the host black shale.
At the Black Butte deposit, outside of the sulfide zones, no petrographic evidence exists for brittle-style veins or stockwork complexes that could have formed as the result of fluids introduced epigenetically in the Newland Formation (Tintina Montana, Inc., geologists, pers. comm., 2015). In a context of enhanced thermal gradient as described above, the converging effects of subsidence, forced diagenetic maturation, and compaction of shale sequences are known to contribute building up fluid overpressure needed for the flow of metal-bearing connate fluids in shale sequence [80,81]. The continuous basin development and burial of hydrothermal vents proposed here for Black Butte [6,7] could have triggered the necessary heat flow that has been proven elsewhere to facilitate fluid-rock equilibration, enhance diagenetic maturation and metal solubility (especially Cu) from shale source rock, and act as a driver of fluid advection in shale basins [81]. In this context, a subsidiary derivation of metals such as Cu and Os from the crustal shale sequence itself could rationally explain the possibly low radiogenic 187 Os/ 188 Os initial ratio of chalcopyrite up to 0.24. The present interpretation of reduced, metalliferous magmatic-hydrothermal fluids underpinned by our new Re-Os data for chalcopyrite expands on the original suggestion by Graham et al. [7] that metalliferous hydrothermal fluids were sourced and circulated either through the basement or from deeper parts of the basin.

Trigger for Precipitation of
Cu ± Co ± Ag Mineralization. In our model, Cu and metals were largely contributed by moderate-temperature, reduced magmatic-hydrothermal fluids carrying reduced sulfur species with a magmatic origin. A recognized trigger for sulfide precipitation from such fluids is cooling, possibly accompanied by saturation in reduced sulfur, which would depreciate the solubility of chalcopyrite [78]. Chalcopyrite mineralization replaced locally massive synsedimentary to early diagenetic pyrite units that promoted precipitation of chalcopyrite (i.e., a feature previously recognized at the Tharsis deposit in the Iberian Pyrite Belt; [82,83]). This replacement likely occurred close to the sediment-water interface, i.e., an ideal locus where fluids could cool. The thermochemical reductive dissolution of diagenetic barite also took place below the sediment-water 15 Lithosphere interface. Therefore, a pool of reduced sulfur comprising massive pyrite and reductive dissolution of barite was readily available in this setting for magmatic-hydrothermal fluids (themselves carrying H 2 S with a magmatic origin) to cool and contribute precipitation of chalcopyrite ( [6]; this study). The paragenetic association of chalcopyrite, chlorite, quartz, and pyrite lacking significant credits of sphalerite (ZnS) and galena (PbS) indicates that either reduced sulfur contents in the mineralizing fluids did not reach saturation for galena and sphalerite [81], in particular within the Lower Sulfide Zone [7], or the system remained hot enough so that the solubility threshold for limited PbS and ZnS deposition was not reached.
Chalcopyrite precipitation took place after a peculiar mineralization stage comprising Co-bearing sulfarsenides and sulfosalts (i.e., cobaltite, glaucodot, alloclasite, and tennantite) that was bereft of sulfide phases [33]. We suggest that this early Co-Ni stage may have a different origin to the one presented above for chalcopyrite. Metalliferous black shales may be enriched in metals such as Ni, Cu, As, and Co that are tied in organometallic complexes [84,85]. Sedimentequilibrated, saline, and sulfide-deficient fluids could promote the transport of significant concentrations of metals (Cu, Co, As, and Ni), given the intermediate oxygen fugacity and mildly acidic pH generally imposed on by shale source rocks for these metals (see [81]). Thus, we conceptualize that the early Co-Ni-sulfarsenide-sulfosalt stage was related to the forced diagenetic maturation of shale in the Newland Formation caused by the enhanced thermal gradient on the edge of the GFTZ. This enhanced thermal gradient would in turn facilitate the release of acidic, saline, and sulfide-deficient fluids that would leach metals from organometallic complexes indigenous to shale in the Newland Formation.

Summary and Perspectives
Based on textural relationships and isotopic data presented in this study, several key features explain the origin of the Black Butte Cu sulfide mineralization hosted by black shale interbedded with sedimentary to early diagenetic massive pyrite units of the ca. >1,475 Ma Newland Formation in the Helena Embayment of the Mesoproterozoic Belt-Purcell Basin. Chalcopyrite formed at 1,488 ± 34 Ma (Re-Os isochron age) and most likely qualifies as a synsedimentary to diagenetic mineralization overlapping in time and space with formation of massive pyrite units. Copper mineralization at Black Butte is related to magmatic-hydrothermal activity on the edge of the Great Falls Tectonic Zone and an enhanced thermal gradient in possible connection with intrusion of dyke swarms until ca. 1,455 Ma. The initial 187 Os/ 188 Os isotopic composition of chalcopyrite (0:13 ± 0:11) is compatible with (1) derivation of Os from a markedly magmatic source with a 187 Os/ 188 Os isotopic composition inherited from the primitive upper mantle in the Mesoproterozoic (Os mantle 1,475= 0:12 ± 0:02) and (2) limited contamination from a continental crustal source with an estimated 187 Os/ 188 Os isotopic composition of 0:30 ± 0:23 that could cause an increase of initial 187 Os/ 188 Os ratios for chalcopyrite up to 0.24. This evidence favors a magmatic-hydrothermal metal source with subsidi-ary derivation of metals during thermally forced shale diagenesis and flow of highly metalliferous reduced fluids within the shale sequence.
Reduced, magmatic fluids with moderately high salinities probably transported metals (i.e., Cu) and reduced sulfur at moderate temperature (>250°-350°C) while buffered by the capacity of the host black shale. As magmatic-hydrothermal fluids cooled close to and below the sediment-water interface, chalcopyrite mineralization either (1) replaced locally massive synsedimentary to early diagenetic pyrite units, (2) used reduced sulfur species produced by thermochemical reductive dissolution of diagenetic barite, or (3) utilized the magmatic source of reduced sulfur in the moderate-temperature fluids in areas where Cu mineralization has sulfur isotope signatures clustering around 0‰ and did not proceed from reductive dissolution of barite.
In light of the work by Graham et al. [7], Present et al. [6] and the present contribution, it appears that critical controls on the genesis and metal endowment of stratiform Cu ± Co sulfide deposits in Mesoproterozoic shale sequences are (1) the presence of long-lived tectonic features like the Great Fall Tectonic Zone able to focus magmatic-hydrothermal fluid flow and permit intrusion of dyke swarms, (2) the impact of an enhanced thermal gradient on the thermal maturity of basinal rocks, and (3) the dynamics of bottom water with transient ingress of sulfate-bearing seawater in unconsolidated sediments governing the activity and health of microorganism communities at the loci of metalliferous hydrothermal venting. Finally, it has been shown that the continental crust had become particularly enriched in copper through Paleoproterozoic and Mesoproterozoic magmatic activity [86][87][88]. Future research building on the findings for the Black Butte deposit might demonstrate that the preponderance of strata-bound massive pyrite deposits with economic concentrations of base metals (Cu-Zn-Pb) within Mesoproterozoic strata is related to enhanced magmatic activity globally during this era. mineral separation, and sulfide and shale Re-Os isotope geochemistry procedures and mass spectrometry analyses, with contribution by RAC for several pyrite Re-Os analyses. TES and NJS ran the finite mixture modeling, with contribution by CCM. AB ran sulfur isotope geochemistry on samples provided by NJS. NJS wrote the manuscript with comments by all other authors. All authors reviewed and approved the manuscript.