Results for ⁴⁰Ar/³⁹Ar dating are summarized in age spectra plots in Figure 4 and data are summarized in Table 1. Samples CPL-04-22, 04-25, and 04-27 come from the same drill hole (CPL-95-3) with CPL-04-22 and 04-25 both being black argillite containing a distinct fabric and also containing thin quartz-carbonate-pyrite veinlets. In contrast, sample CPL-04-27 is a dark grey siltstone with some disseminated carbonate spots and represents an alteration zone around a mineralized vein system. Of the three age spectra, two (CPL-04-22, 27) are almost identical and are discussed first. Both spectra show increasing ages that commence near 200 Ma and progress with increasing temperature steps to a pronounced hump around 360 to 380 Ma. This hump profile may indicate the effects of local argon recoil (McDougall and Harrison 1999). Following these humps in the spectra, a small quasi-plateau incorporating about 25–30% of the gas fraction occurs for both samples at around 350 Ma. The highest temperature steps, yielding 10–20% of gas released, give progressively younger ages down to about 320–310 Ma. The third sample, CPL-04-25, from drill hole CPL-95-3 yielded an age spectrum that shows a monotonic increase from an initial age of < 200 Ma to a plateau at 349.3 ± 2.5 Ma which incorporates 51.4 % of the gas. The uniformly low Ca/K ratios (< 0.1) for the whole age spectra in all of the above samples, despite their irregular profiles, indicates that gas was released dominantly from a single K-rich phase, which is interpreted to be mica.

The fourth whole-rock sample (CPL-04-40), from diamond drill hole CPL-95-2, is a fine-grained, laminated, grey-green siltstone with thin (< 5 mm), bedding concordant, quartz-carbonate-specular hematite veins. The sample yielded an age spectrum that shows a monotonic increase in age from an initial step at < 200 Ma to around 380–370 Ma for the last 50% of gas released. Importantly, this sample has a uniformly low Ca/K ratio (≤ 0.1) for the whole spectrum and indicates gas release from a single, K-rich phase, probably mica.

The final two samples dated represent material from near the old workings preserved in dump piles. Sample CPL-99-10C is an altered siltstone with disseminated muscovite that occurs adjacent to a sulphide-bearing carbonate vein with pyrite > chalcopyrite. The sample yielded a flat age spectrum, although the initial 8% of the gas liberated indicates younger ages than the plateau. A plateau age of 322.1 ± 1.6 Ma is defined by 81.8% of the gas released, which is similar to the integrated age of 319.4 ± 1.6 Ma. The uniformly low Ca/K ratios (< 0.1) for the gas fractions defining the plateau age indicates the gas was released from a single, K-rich phase, inferred to be the muscovite seen in the sample.

Sample CPL-2004-2C is a muscovite separate extracted from a pervasively altered siltstone adjacent to a massive carbonate-sulphide (pyrite > chalcopyrite) vein. In thin section, muscovite is seen to form radiating clots of coarse muscovite intergrown with fresh, hydrothermal carbonate. This sample is considered to represent the wall-rock component of the main vein system at the deposit site. The muscovite separate yielded a flat age spectrum with a plateau age of 309.5 ± 1.6 Ma for 80% of the gas liberated, identical to the integrated (309.5 ± 1.6 Ma) and correlation (307.9 ± 3.4 Ma) ages. The uniformly low Ca/K ratio (< 0.1) for this sample is consistent with the gas being liberated from a single, K-rich phase.

Pyrite from a massive carbonate-sulphide (pyrite > chalcopyrite) vein sample was separated into five individual fragments and each was analyzed for its Re and Os isotopic composition (Table 2). The pyrite has a high Re/Os ratio and the Os is highly radiogenic, the so-called "low level highly radiogenic" (LLHR; Stein et al. 2000) sulphide. The data display a well-defined linear relationship in an isochron diagram (Fig. 5) and regression of the data defines an age of 323 ± 8 Ma with MSWD value of 0.4 and initial ¹⁸⁷Os/¹⁸⁸Os value of 1.0 ± 0.4 (2σσ uncertainties).

A sample of pervasively bleached and altered siltstone from drill core (Fig. 6a) was observed during petrographic and imaging studies to contain an anomalous amount of subhedral to euhedral, normally zoned zircon grains (Fig. 6b). The zircon was considered to possibly represent hydrothermal growth based on its abundance and habit and it was therefore separated for dating, the results of which are summarized in Table 3. Crushing of the sample yielded a multi-grain fraction (n = 40) of small, colorless zircon grains which have a U content of 242 ppm, Th content of 68 ppm, and Th/U ratio of 3.5 that is typical of igneous zircon. The individual Pb/U ages obtained (i.e., ²⁰⁶Pb/²³⁸U, ²⁰⁷Pb/²³⁵U) are highly discordant, as illustrated by a standard concordia plot shown in Figure 6c, but the calculated ²⁰⁷Pb/²⁰⁶Pb age is 1634 ± 11.2 Ma.

Two samples containing monazite were selected for in situ dating using the Th-Pb chemical dating method (Montel et al. 1996; Williams et al. 1999). In this method, monazite samples with sufficient Th and of appropriate age can be dated using mineral analyses obtained from the electron microprobe and model ages calculated. As part of the procedure, the monazite grains were first characterized chemically to select material for dating. Representative chemical analyses for monazite grains along with their chrondrite-normalized rare-earth element (REE) plots and representative back scattered electron (BSE) images are provided in Table 4 and Fig. 7, respectively. The monazite grains are all Ce-rich versus the Ca- and Th-rich members (i.e., cheralite and brabantite; Förster 1998) of the phosphate group and have uniform chemistry. However, the BSE images indicate complex growth histories with variably sieved cores observed in some cases and euhehdral overgrowths common. The chondritic profiles and REE abundances are uniform for all the grains analyzed in the four different samples and the chemical data are typical of monazite from both igneous and metamorphic settings (e.g., Bea 1996; Förster 1998; Spear and Pyle 2002).

X-ray maps indicate that the monazite grains are variably zoned with irregular-shaped core areas and euhedral overgrowths, as indicated from the initial observations from BSE imaging (Fig. 8). A total of twelve ages were obtained for seven grains from two samples (Table 5); two of the samples analyzed have insufficient Th and Pb concentrations to provide age determinations The data indicate ages from 387 Ma to a low of 225 Ma with three apparent groupings: (1) 397–370 Ma (4 ages), (2) 360–330 Ma (4 ages), and (3) two at 272 and 225 Ma.

The results of ⁴⁰Ar/³⁹Ar dating indicate that several geological events are represented by the samples. The unaltered or barren whole rock samples record young ages of ≤ 200 Ma for initial Ar release, with the higher temperature components recording ages between 350 Ma and 380–370 Ma. The oldest plateau age of 380–370 Ma, recorded in CPL-04-40, is slightly younger than a U-Pb zircon age of 389 Ma for felsic volcanic rocks in the Guysborough Group and similar to ⁴⁰Ar/³⁹Ar ages for whole-rock Horton Group samples from north of the Roman Valley Fault just east of the study area, which Reynolds et al. (2004) attributed to a detrital muscovite component. Based on the available information, this age would be consistent with either of the following scenarios. Firstly, inheritance via a detrital component derived from, for example, mica in granitoid rocks of the Meguma terrane to the south or metamorphic rocks to the north (e.g., Cape Porcupine Complex; White et al. 2001), which contains muscovite of the appropriate age (Reynolds et al. 2004). In regards to the detrital component, Murphy and Hamilton (2000) reported U-Pb ages of appropriate age (i.e., 380–370 Ma) in their dataset for detrital zircons from the Devono-Carboniferous Horton Group in the St. Marys basin located south of the study area. A second explanation is that the ⁴⁰Ar/³⁹Ar age spectrum represents a relict depositional age for the host rocks. Given the absence of detrital muscovite in the dated sample and abundance of fine-grained mica based on petrographic study, the latter interpretation is cautiously considered to be the most likely.

Several samples record plateau ages of about 350 Ma, the most convincing being in sample CPL-04-25 where the last 51% of the gas liberated defines an age of 349 Ma. The 350 Ma ages occur in rock samples with well-developed cleavage, therefore, this age is interpreted to indicate the time of muscovite growth during regional deformation in the area. This age is some 10 Ma older than the timing of cleavage formation in Horton Group sedimentary rocks to the north determined by Reynolds et al. (2004). The age discrepancy may indicate that the older, more deeply buried rocks of the Guysborough Group were heated above the blocking temperature of muscovite (i.e., > 350 °C; McDougall and Harrison, 1999) before they were thrust over the Horton Group rocks during dextral compression along the CCFS (Webster et al.1998). This sequence of events would account for the difference in ages for the deformation and heating of the two sequences. On a more regional scale, this deformation is similar in age to syntectonic plutonism and continued movement along the CCFS in the Cobequid Highlands west of the study area (Pe-Piper et al. 2004; Fig. 1a).

The lack of younger plateau ages at 320 Ma for the whole-rock samples from drill core material which is altered and veined (Kontak 2006) is significant. The dated material records alteration in the form of secondary mineral growth (e.g., quartz, albite, carbonate), but evidently the temperature of the fluids attending this alteration was either not hot enough or of sufficient duration to cause diffusive loss of argon in the samples. Given the restricted nature of the veins, their generally narrow width (< 1 m) and high-level setting (Kontak 2006), it is apparent that the temperature away from the main ore zone was below the 250–300 °C required to cause resetting in fine-grained, micaceous sedimentary rocks (McDougall and Harrison 1999). However, the two samples that record new growth of hydrothermal muscovite do reflect a thermal event at 320–310 Ma, which is considered to be coincident with vein-related mineralization given the restriction of the alteration to the immediate wall rock to mineralized veins. The ages for these two samples, both yielding excellent plateaus, do not overlap and due to this we note the following points. Firstly, the two samples were positioned closed to each other during irradiation and the similar values for the flux monitors (i.e., J-values, Table 1) essentially rule out analytical problems. Secondly, the muscovite grains were sufficiently coarse (i.e., 100s microns) that recoil would not be considered a potential problem (McDougall and Harrison 1999). Thirdly, given that the two samples come from the same mineralized zone, differential cooling or heating of samples is considered unlikely. Finally, the whole-rock sample is dominated by secondary hydrothermal muscovite, thus any contribution of radiogenic Ar from fine-grained metamorphic muscovite is considered not to have been significant. In addition, the excellent plateau for this sample contrasts with the irregular nature of the age spectra for whole rock samples which might be expected to be seen if outgassing of grains of different generations had occurred. Therefore, based on the two plateau ages it is concluded that the best estimate for the timing of hydrothermal activity based on ⁴⁰Ar/³⁹Ar dating is 320–310 Ma.

The result of Re-Os dating of vein pyrite provides a time for pyrite formation and, hence, hydrothermal activity at 323 Ma. This age overlaps within error that recorded for growth of hydrothermal muscovite in the wall rocks adjacent ore veins based on ⁴⁰Ar/³⁹Ar dating. The coincidence of these ages provides further support that the Re-Os method is a reliable and robust means of directly dating sulphide mineralization, as documented elsewhere (Arne et al. 2001; Morelli et al. 2004, 2005). Also relevant is that other workers have demonstrated the concordance of Re-Os sulphide ages with ages obtained using other chronometers (i.e, Rb/Sr, ⁴⁰Ar/³⁹Ar) by either dating the mineralization directly (e.g., Schneider et al. 2007) or indirectly (Selby and Creaser 2001, Masterman et al. 2004). Thus, the apparent difference of the Re-Os pyrite and ⁴⁰Ar/³⁹Ar age for sample CPL-2004-2C cannot be attributed to differences of calibration of the two chronometers. The agreement of the Re-Os and ⁴⁰Ar/³⁹Ar chronometers, along with the robust nature of the Re-Os system in pyrite (Creaser 20), provides for the first time a direct age for the mineralization along the CCFS of about 320 Ma and allows interpretation of this mineralization in the context of its metallogenic significance.

The initial ¹⁸⁷Os/¹⁸⁸Os ratio, derived from regression of the Re-Os isotopic data (Fig. 5), of 1.0 ± 0.4 can be used to infer the nature of the reservoir for Os in the mineralizing system. This value compares, for example, to initial ¹⁸⁷Os/¹⁸⁸Os ratios derived from regression of Re-Os isotopic data for sulphides from a variety of other sediment-hosted mineralized settings: (1) 0.2 ± 0.2 for Zn-Pb ore at the Red Dog SEDEX deposit, Alaska (Morelli et al. 2004); (2) 0.37 ± 0.27 for the auriferous quartz vein mineralization at Murantau, Uzbekistan (Morelli et al. 2007); (3) 1.04 ± 0.16 for orogenic Au mineralization, Victoria, Australia (Arne et al. 2001); and 0.83 ± 0.15 and 0.38 ± 0.16 for Meguma gold veins (Morelli et al. 2005). These initial ¹⁸⁷Os/¹⁸⁸Os ratios indicate that a mixture of source reservoirs, from mantle to crustal, are involved in these mineralized settings. For the Copper Lake setting, the high initial ¹⁸⁷Os/¹⁸⁸Os ratio suggests a dominantly crustal reservoir for the Os, which is similar to that for some of the Au settings noted above and, importantly, precludes a significant mantle contribution to the Os component of this system. This conclusion is consistent with stable isotopic analyses (¹⁸O, ³⁴S) for the Copper Lake vein system (Kontak 2006) which indicate that the mineralizing fluids are of crustal origin.

The single U-Pb zircon age reported indicates that the dated sample is dominantly of detrital origin, as there may be a small component of overgrowth, and thus does not preclude the presence of a hydrothermal component to the zircon in the sample. Although the data obtained are highly discordant, the ²⁰⁷Pb/²⁰⁶Pb age of 1634 ± 11.2 Ma is noteworthy as it provides some indication of the provenance for the sedimentary rocks of the Guysborough Group sedimentary rocks, which is currently lacking. Existing databases for detrital zircon exist for a large part of the Meguma (Meguma Group, White Rock and Torbrook formations) and Avalon terranes, as well as the overlapping Horton Group (summaries in Murphy and Hamilton 2000; Murphy et al. 2004). Based on these data, a likely source for detrital zircon would be the Avalon terrane, albeit locally derived in the Antigonish Highlands where an appropriate detrital zircon age population has been reported.

The origin of monazite in IOCG deposits is important, as enrichment in rare-earth elements is, in some cases, a chemical signature of this mineralization style (Williams et al. 2005). Thus, purpose of monazite analysis was twofold, first to investigate its origin, that is detrital versus hydrothermal, and secondly to see if it could constrain the age of hydrothermal activity and rare element mobility. Twelve monazite Th-Pb ages fromr two samples indicate a large spread, from 397 Ma to 225 Ma, which indicates either different periods of monazite growth or variable resetting of the monazite grains.

The older ages are consistent with timing of deformation in the Meguma Group to the south (e.g., Kontak et al. 1998) and suggest, therefore, a detrital origin since monazite is a common accessory phase in metasedimentary rocks. The middle age population is similar within error to whole-rock ⁴⁰Ar/³⁹Ar ages that record metamorphism in the region based on our data and earlier work (Reynolds et al. 2004), with some overlap with the age of hydrothermal activity related to mineralization. The youngest ages (≤ 300 Ma) record yet another period of mona-zite growth. In summary, although not providing precise ages, the chemical dating of monazite in two samples indicates a multi-stage history for the mineral with growth both pre-dating and post-dating the time of mineralization at Copper Lake.

The rare-earth element chemistry determined for the monazite appears uniform (Fig. 7), although imaging indicates apparent zoning and chemical dating indicates a multi-stage growth history for several of the grains. Thus, in this study the chemistry cannot be used to either discriminate source regions for the detrital components (i.e., igneous vs. metamorphic) or distinguish older core regions from subsequent overgrowths.

One of the basic or fundamental aspects of any mineral deposit model begins with the timing of mineralization with respect to the host rock, for example syngenetic versus epigenetic modes of origin. In the case of IOCG mineralization, the relative timing of mineralization is inferred based on geological relationships supported by more detailed petrographic observations. However, determining the absolute age of mineralization, critical to formulating a model, is contingent on the presence of phases amenable to dating. In the present study, four chronometers have been applied to a well-characterized suite of samples from the Copper Lake district and two of these, Re-Os on ore-stage pyrite and ⁴⁰Ar/³⁹Ar on hydrothermal muscovite (i.e., whole rock and mineral separate), provided a reliable age for the hydrothermal activity responsible for mineralization. Thus, for the first time an absolute age has been determined for this important metallogenic domain in Nova Scotia and the age of about 320 Ma can now be assessed in the context of the regional geological evolution of this area.

The timing of mineralization at Copper Lake overlaps the stratigraphic age, i.e., Namurian (327–315 Ma) using time scale of Okulitch (1999) for mineralized and altered host rocks at some of the IOCG-type mineralized localities in Nova Scotia, most notably Londonderry (Wright 1975; Ervine 1994) and Mt. Thom (O’Reilly 2005; Kontak 2006). It is significant, therefore, to note that similar ages, based on ⁴⁰Ar/³⁹Ar dating, as that obtained for mineralization at Copper Lake have also been reported elsewhere in southern Nova Scotia, which indicates that this thermal event is a widespread phenomenon. Firstly, Pe-Piper et al. (2004) reported ages of 330–320 Ma for hydrothermally altered lamprophyric rocks cutting older granites (i.e., 350 Ma) in the Cobequid Highlands (Fig. 1a). The authors suggested that this event may relate to fluids driven by a mid-crustal heat source. In addition, it was also noted that this age is coincident with Alleghanian deformation and uplift of the Cobequid Highlands and corresponds to the mid-Carboniferous break in Nova Scotia, a time of disconformities and unconformities in the stratigraphic record (Calder 1998). Secondly, Kontak et al. (20; also our unpublished data) reported ⁴⁰Ar/³⁹Ar ages for several small mafic intrusions adjacent to the CCFS, some of which intrude Mabou Group (i.e., Namurian) sedimentary rocks (O’Reilly 2005). The ages span the range of 340–300 Ma, hence, in part overlapping the age of mineralization at Copper Lake. Thirdly, several ⁴⁰Ar/³⁹Ar ages have been reported throughout the Meguma terrane that document Alleghanian tectono-thermal activity and fluid movement (e.g., Reynolds et al. 1987; Kontak et al. 1995; Culshaw and Reynolds 1997; Murphy and Collins 20). Thus, the data above indicate that coincident with the age of mineralization at Copper Lake there was widespread mafic magmatism, regional deformation, and movement of hydrothermal fluids. A model involving mobilization of fluids related to emplacement of mid-crustal mafic magma, as suggested by Pe-Piper et al. (2004) for the Londonderry area, is supported by the present study. Based on geochemical and isotopic data, both radiogenic and stable (Kontak 2006 and unpublished data), a crustal source for the mineralizing fluids at both Copper Lake and Mt. Thom is indicated.