Abstract
ABSTRACT An orientation study around the NICO Co-Au-Bi deposit in the Great Bear magmatic zone of NW Territories, Canada, was initiated in 2007 to establish a practical guide to geochemical and mineralogical exploration for iron oxide copper-gold deposits in glaciated terrain. Bedrock and till samples were collected up-ice, proximal and down-ice from mineralization and host rocks, to characterize their indicator mineral signatures. Results demonstrate that gold grain abundance, size and shape, as well as magnetite and hematite composition, have the best potential to fingerprint the mineralization at NICO. Pristine-shaped gold grains indicative of a local bedrock source and a short distance of glacial transport are relatively abundant in till samples collected immediately down-ice from several mineral occurrences at NICO and none were recovered up-ice. Iron oxide composition using preliminary discriminant diagrams shows some potential, using Ni/(Mn+Cr) versus Ti+V plots. In particular, magnetite and hematite from till samples collected over, or directly down-ice of, the NICO deposit have lower Ti+V compositions compared to magnetite and hematite from till collected up-ice from mineralization. Potential non-ferromagnetic indicator minerals are either not chemically stable in surface sediments (arsenopyrite, chalcopyrite, pyrite), not sufficiently coarse-grained or resistant to glacial transport (bismuthinite, tourmaline, ferroactinolite), not abundant enough in the mineralized bedrock (scheelite, molybdenite, cobaltite, allanite), or not sufficiently heavy (tourmaline) to be useful at NICO but may be at other deposits in the region or elsewhere in glaciated terrain. The development of indicator mineral methods, together with till geochemistry, will be tested with further sampling over the Great Bear magmatic zone.
The Great Bear magmatic zone (GBMZ) in the NW Territories is now considered the most prospective setting for iron oxide copper-gold (IOCG) deposits in Canada (Corriveau 2007). It includes two economic IOCG deposits: the magnetite-group NICO Co-Au-Bi deposit and the nearby hematite-group Sue-Dianne Cu-Ag-Au deposit (Fig. 1). Moreover, many past-producing vein-type U, Ag and Cu mines and Kiruna-type showings in the GBMZ have recently been re-examined and shown to be parts of large polymetallic IOCG systems within an IOCG- porphyry-epithermal continuum (Mumin et al. 2007, 2010; Montreuil et al. 2010; Corriveau et al. 2010). As part of a joint government-industry-academia research project, the NICO deposit was selected as a test site to characterize the heavy mineral and geochemical signature of IOCG±U deposits in derived glacial sediments and determine if some heavy minerals have potential to be useful indicator minerals for IOCG deposits in glaciated terrain. A very small number of studies about indicator mineral tracing have been conducted near IOCG deposits (i.e. Belousova et al. 2002; Sarala et al. 2007) and to our knowledge, this is the first detailed study that describes the distribution and characteristics of potential indicator minerals in till derived from IOCG deposits. At the NICO deposit, mineralized zones and hydrothermally-altered host rocks are exposed at surface, ore mineralogy is well known, and Fortune Minerals Ltd. infrastructures at Lou Lake provided logistical support in an otherwise remote, poorly-accessible area (Corriveau et al. 2007). The mineralogy and chemistry of heavy minerals from bedrock and till samples collected in 2007 in the vicinity of the NICO deposit are presented here.
The Great Bear magmatic zone bedrock geology, mineral occurrences and past-producing mines (modified from Corriveau et al. 2007). The NICO deposit is located in the southern part of the GBMZ. Inset map locates the Wopmay orogen (Bear Province) and adjacent Slave craton at the western end of the Canadian Shield.
REGIONAL SETTING
The NICO deposit is located at the south end of the GBMZ in the Proterozoic Bear Structural Province of the Canadian Shield, about 160 km NW of Yellowknife, NW Territories, Canada (Fig. 1). It is an economically significant source of Co-Au-Bi-Cu-Fe with pre-production reserves of 31.7 Mt at 0.91 g/t Au, 0.12% Co, 0.16% Bi and 0.04 % Cu (Fortune Minerals Ltd. news release, January 14, 2010). Mineralization at NICO consists of three surface and sub-surface tabular zones, up to 1.5 km in length, that are hosted in hydrothermally-altered and locally brecciated marine siltstone and wacke of the c. 1.88 Ga Treasure Lake Group below their unconformity with overlying 1.86 Ga felsic volcanic rocks of the Faber Group (Fig. 2; Goad et al. 2000a, b; Gandhi & van Breemen 2005). The sedimentary rocks have been metamorphosed at c. 1.88 Ga during the short-lived Calderian orogeny prior to Great Bear magmatism (cf. synthesis of Wopmay Orogen by Hildebrand et al. 2010). Sub-volcanic granitoid intrusions of the Marian River Batholith underlie and intrude the Treasure Lake and Faber groups. Both groups are also intruded by porphyry dyke swarms (Goad et al. 2000a; Mumin et al. 2007, 2010). Brecciation and intense K-feldspar alteration are common at the interface of the porphyry dykes with the amphibole-magnetite-bitotite-altered Treasure Lake Group rocks they intrude. The youngest hydrothermal effects observed in the general area of the NICO deposit are emplacement of giant quartz vein complexes and quartz stockwork, and quartz-epidote veining and alteration (Gandhi & Lentz 1990; Gandhi et al. 2000). The quartz veins comprise specular hematite, Cu sulphides, pitchblende (uraninite), pyrite and sphalerite while their wall rocks are commonly altered to chlorite and hematite and silicified. To the NE and SE of the NICO deposit, Bi, Co, Cu, Au, Ni and U mineral occurrences are common and hosted within hydrothermally-altered Treasure Lake and Faber group rocks (Fig. 2). Most mineral occurrences are hosted within coalescing magnetite-rich hydrothermal systems that include albite, amphibole, biotite, calcite, chlorite, K-feldspar, magnetite and/or hematite replacive alteration assemblages, breccia and/or veins.
Mineral occurrences and regional geology of the NICO deposit area (modified from Goad et al. 2000b) and sample site locations (see Table 1 for bedrock sample numbers collected at till sampling sites). Locations of sample sites close to the NICO deposit are shown on Figure 3.
Summary of bedrock sample descriptions
Mineral occurrences and detailed bedrock geology of the NICO deposit area (modified from Goad et al. 2000b) and sample site locations. Bedrock samples collected at till sampling sites and at mineral occurrences are listed in Table 1.
The surface expression of the NICO ore body is a large sub-circular valley referred to as the Bowl Zone (Fig. 3; Goad et al. 2000a). It exposes meta-siltstone and meta-arkose with intense, pervasive, stratabound replacement alteration, veining, and brecciation and their overlying K-feldspar altered rhyolite (Goad et al. 2000a). Mineral assemblages of alteration in the metasedimentary rocks consist predominantly of iron-oxide (magnetite-dominant), hornblende and biotite with and without K-feldspar that record polyphase calcic-iron to potassic-iron alteration (e.g. Discovery Zone). Stratabound tourmaline alteration and retrograde hematite or carbonate alteration occur locally. The pervasively-altered metasedimentary rocks are crosscut by calcite veins, quartz-hematite veins (e.g. #25 Zone) and arsenopyrite veins with or without magnetite (e.g. #25 Zone) as well as by porphyry dykes, some of which have intense potassic (K-feldspar) alteration at their margins (e.g. #2 Zone) and disseminated sulphides. Potassic alteration and potassic-iron alteration in the form of pervasive K-feldspar and magnetite-K-feldspar replacement alteration occur within the volcanic rocks. Tourmaline breccias occur within volcanic rocks and porphyries and are particularly well developed at the southern end of the NICO deposit (e.g. ‘Portal A’ road and beyond). Chlorite and sericite occur at the periphery of the deposit. Minor mineralization extends into the overlying volcanic rocks and porphyry dykes as veinlets and veins (including the # 1 and #3 zones). The various styles of IOCG-type alteration and mineralization record considerable spatial overlap and overprinting in both prograde and retrograde sequences and a continuum with porphyry system (Goad et al. 2000a; Corriveau et al. 2010; Mumin et al. 2010). Ore mineralogy mainly includes Fe-, As-, Co- and Cu-sulphides, and native Au and Bi. Iron oxides, dominated by magnetite with lesser amount of retrograde hematite, comprise approximately 20% of mineralized zones and its host alteration zones, and result in a variety of coincident geophysical anomalies, namely a regional positive Bouguer-gravity anomaly and a positive total-field magnetic-gradient anomaly (Goad et al. 2000a). The NICO deposit is now considered to be a member of the IOCG deposit type because of features such as abundant iron oxides, hydrothermal mineralization, Au as primary economic metal and regional- to deposit-scale intense K, Ca and Fe alteration (Gandhi et al. 1996; Goad et al. 2000a; Corriveau et al. 2010). Potentially resistive minerals documented at NICO and the various mineral occurrences in the area comprise altogether arsenopyrite, Au-Bi-Te alloys, native bismuth, bismuthinite, bismuth telluride, bornite, chalcopyrite, cobaltite, emplicite, gold, ilmenite, loellingite, magnetite, molybdenite, scorodite, pyrite, pyrrhotite, scheelite, sphalerite, native silver, tennantite, uraninite, wittichenite and wolframite (Mathieu 1966; Gandhi & Lentz 1990; Goad et al. 2000a; Sidor 2000). Chemically, the various hosts of sulphide minerals within the IOCG systems and its variety of mineralized zones are known to be enriched in As, Au, Bi, Co, Fe and S.
Metasedimentary rocks of the Treasure Lake Group, their overlying volcanic sequence and the variety of hydrothermal alteration zones they host generally form prominent ridges striking SE–NW. These ridges have relief exceeding 100 m and are dominated by exposed bedrock and thin, discontinuous till (<2 m thick). Great Bear intrusive rocks form more poorly drained lowlands covered by generally thin till. Striations, roches moutonnées and crescentic fractures indicate the area was influenced primarily by the Laurentide Ice Sheet flowing to the west-southwest during the last Wisconsinan glaciation, from an ice mass centered in Keewatin (McMartin et al. in press). Pebble lithology data show that surface till composition predominantly has a local (<1–5 km) provenance from the underlying bedrock (McMartin et al. in press). Late during deglaciation, Glacial Lake McConnell occupied the area from w. 8.5–10.5 ka BP (Dyke 2004) as a result of glacioisostatic depression reversing the regional drainage in the Great Bear, Great Slave and Athabasca lake basins (e.g. Lemmen et al. 1994; Smith 1994). Evidence for reworking of glacial sediments by glaciolacustrine processes, to varying degrees, is present at all elevations as veneers of silt and clay in topographic depressions, specifically around Lou Lake and other lake basins, or as veneers of winnowed till and littoral sands overlying the glacial deposits. Since the retreat of the glacial lake 8 500 years ago, extensive discontinuous permafrost and brunisolic static cryosols have developed on the glacial sediments. Vegetation today is dominated by a mixed deciduous and conifer open forest cover.
METHODS
Field procedures
A total of thirteen 5–12 kg till samples were collected from twelve sites in the vicinity of the NICO deposit in 2007 (Figs 2 and 3). Samples were collected in the upper C horizon soils developed on till from hand dug pits, at an average depth of 50 cm, to obtain relatively unaltered material. Permafrost or sometimes bedrock was encountered at the bottom of most holes. Most minerals are typically weakly oxidized in C horizon soils of northern Canada except for the most labile minerals which may be partly or completely destroyed by postglacial weathering (e.g. McMartin & Campbell 2009). Three till samples were collected immediately down-ice (<150 m) from exposed mineralized zones (Fig. 3; 07MOB005, 90 m west of #25 Zone; 07MOB006, 150 m west of #3 Zone; 07MOB007, 20 m west of Discovery Zone). On the NW side of the Bowl Zone (Fig. 3), one 6 m thick section exposing a glacial diamicton in a borrow pit was sampled above (07MOB009: 180 cm depth) and below (07MOB010: 350 cm depth) the zone of oxidation to evaluate the effects of post-glacial surface weathering on till composition. Four till samples were collected at different sites away from mineralization: two samples along Lou Lake (Fig. 2) over potassic-altered porphyritic volcanic rocks of the Faber Group (07MOB012) and monzonite of the Great Bear batholith (07MOB013); and two samples over magnetite-altered metasedimentary rocks and porphyry dyke (07MOB008 and 07MOB011) (Fig. 3). Four till samples (07MOB001 to 004) were collected approximately 10 km NE from the NICO deposit area over altered metasedimentary and weakly altered porphyritic volcanic rocks of the Treasure Lake and Faber groups (Fig. 2). Although these samples were collected up-ice from NICO, they can not be considered as representing non altered background sites since additional mineral occurrences and alteration zones are located in their vicinity, including one mineral occurrence 120 m up-ice from sample 07MOB001 (Tan U-Fe showing).
Twenty-seven representative 0.5–2 kg bedrock samples were collected for indicator mineral recovery at most drift sampling sites, in the vicinity of and at nearby surface occurrences, and from the sub-surface “Portal” area at various bulk sample stock piles (see Table 1 and Figs 2 and 3). To determine what the regional and local rocks may contribute to till in terms of heavy minerals, bedrock samples included mineralized rocks from the NICO deposit (n=15), barren host rocks in the immediate NICO area (n=5), mineralized metasedimentary rocks from the Tan showing (n=1), and unmineralized and least-altered regional rocks (n=6) (Table 1). Detailed till and bedrock sample locations and descriptions are provided in McMartin et al. (in press).
Analytical procedures
Unmineralized or weakly mineralized bedrock samples were milled to <2 mm at a commercial laboratory and the mineralized rocks were milled to <2 mm at Overburden Drilling Management Ltd (ODM). All crushed bedrock samples were then processed at ODM for heavy mineral separation, panning and indicator mineral picking (Fig. 4). Samples were crushed and processed in order of increasing content of visible sulphides to minimize the potential for indicator mineral carryover between samples. Crushed vein quartz (i.e. mineral blank) was processed between bedrock samples in order to reduce and monitor contamination from laboratory equipment. The crushed bedrock fraction was pre-concentrated with respect to density using a shaking table (McClenaghan 2011). Each sample was tabled twice to increase gold grain recovery. Visible gold and sulphide grains recovered from this table concentrate and/or by subsequent panning were counted and described, then returned to the table concentrate after examination. Heavy liquid separation using methylene iodide (SG 3.2) was used to produce a final heavy mineral concentrate (HMC) from the table concentrate. After a ferromagnetic separation, the non-ferromagnetic HMC (NFM-HMC) was sieved to obtain the sand fraction (0.25–2 mm) for picking. The ferromagnetic fraction (FM-HMC) was sieved (0.25–2 mm) for further picking.
Generalized flow sheet showing steps in bedrock and till sample processing to recover indicator minerals.
Till samples were processed at ODM to recover heavy minerals, using a method similar to that used for the bedrock samples (Fig. 4). Samples were disaggregated and sieved to obtain the <2 mm (matrix) fraction prior to processing using a double-run across the shaking table. Blank samples were not inserted. After heavy liquid and ferromagnetic separations, the NFM-HMCs of till samples were sieved at >0.25 mm to obtain the 0.25–2 mm fraction. Only half of this fraction was picked for indicator minerals (the other half was used for geochemistry); therefore, all grain counts for till in the 0.25–2 mm NFM-HMC fraction were normalized by a factor of 2.
Prior to indicator mineral examination and selection, the NFM-HMCs recovered from till and bedrock samples were sieved to 0.25–0.5 mm, 0.5–1 mm and 1–2 mm. For the bedrock samples, only the 0.25–0.5 mm fraction was examined as it consisted of individual minerals grains; the 0.5–1 and 1–2 mm fractions consisted of impure lithic fragments containing the same heavy minerals. The 0.25–0.5 mm fraction was further sorted with an electromagnetic separator into fractions with different paramagnetic characteristics to help reduce the volume of concentrate to be visually examined (Averill & Huneault 2006). All fractions were examined under a stereoscopic microscope at ODM to determine the abundance of potential IOCG indicator minerals, metamorphosed or magmatic massive sulphide indicator minerals (MMSIMs) and kimberlite indicator minerals (KIMs). Checks were performed on selected grains using SEM-energy dispersive x-ray spectrometer (EDS) to confirm mineral identity. Grains considered having possible IOCG affinities (mainly sulphides and silicates) were mounted for further study. Because of their significant abundance in some bedrock samples, no more than 20–40 representative grains of the same mineral species were selected per sample for further study.
A total of 94 grains from till samples and 532 grains from bedrock samples from the NFM-HMCs were mounted on 25-mm epoxy-impregnated stubs and analyzed to confirm their identity and quantify their chemical composition. Analyses were conducted at the Geological Survey of Canada Microbeam Laboratory, Ottawa, using a CAMECA SX50 electron microprobe (EMP) equipped with four wavelength-dispersive spectrometers. Operating conditions were 20 kV accelerating voltage, and 10 nA beam current using a focused spot. Count times on peak were 10 seconds, with 5 seconds off-peak. The raw data were processed with the ZAF matrix correction. Standards comprise a range of natural and synthetic pure metals, simple oxides and simple compounds. The analysed grains were classified (or re-classified when necessary) on the basis of their chemical composition. Theoretical chemical compositions of mineral end-members (Le Maitre 1982) were used to calculate cut-off values (at approximately 50:50 mol %) for members of binary solid solution series. For minerals that contain substantial amounts of more than two end-members, the threshold values were lowered accordingly.
Grains from the 0.25–2 mm ferromagnetic fraction (FM-HMC) were visually picked randomly for each sample. As a first pass to test the usefulness of using iron oxide composition to fingerprint IOCG mineralization, five grains per till sample for a total of 65 grains and ten grains per bedrock sample for a total of 265 grains were mounted on 25-mm epoxy-impregnated stubs at SGS Mineral Services, Lakefield. The grains were analyzed at Université Laval using a CAMECA SX-100 5-WDS electron microprobe under a beam of 15 kV at 100 nA, using a range of natural and synthetic standards. After counting over the peak for 20–30 seconds, background was measured on both sides for 10 seconds at offset from the peak empirically determined to improve sensitivity. Matrix corrections were computed using CAMECA's implementation of the PAP method. These settings yield minimum detection limits as low as 20 ppm for elements such as K, Ca, Al, Si, Ti and Mg, 50 ppm for Mn, Cr and V, 200 ppm for Cu and Ni and 400 ppm for Zn. The optimized analytical routine allows analysis of one spot in c. 20 min. Of the 65 till and 265 rock sample grains, only 46 and 211, respectively, were analyzed for minor and trace elements because some of the ferromagnetic fraction grains consisted of other minerals with ferromagnetic mineral inclusions too small for accurate analysis. The complete tabling, picking and mineral chemistry results for both the 0.25–2 mm NFM-HMCs and FM-HMCs are provided in McMartin et al. (in press).
Furthermore, to evaluate the potential for finer-grained and/or lighter-weight indicator minerals in till, a 1 kg split from each bulk till sample was wet sieved to recover the 0.063–0.25 mm fraction (Fig. 4). The resulting finer fraction, weighing between 57–470 g, was separated at ODM using methylene iodide (SG 3.2). After a ferromagnetic separation, a portion of the liquid light fraction (SG<3.2), the heavy mineral fraction (SG>3.2) and the ferromagnetic fraction were mounted in epoxy on glass slides at the Geological Survey of Canada. The slides were examined using stereoscopic and petrographic microscopes at Consorminex Inc., and the grains counted (NFM grains only). Grain count results are also included in McMartin et al. (in press).
Quality Assurance / Quality Control
One arsenopyrite and two pyrite grains were found in the vein quartz blank processed prior to the mineralized bedrock concentrates at ODM. Seven of the thirteen quartz blank samples crushed and processed with the batch at ODM yielded no heavy mineral grains >0.25 mm. Traces (1–11 grains) of 0.25–0.5 mm arsenopyrite and/or pyrite were found in four of the other six blank concentrates indicating some carryover from the bedrock crusher or during the processing at ODM. Thirty hematite grains were found in the initial vein quartz blank crushed at the commercial laboratory and are presumed to represent crusher carryover from a previous client's samples. One to four pyrite or arsenopyrite grains were found in five of the other twelve vein quartz blanks processed at this commercial laboratory.
HEAVY MINERAL SIGNATURES
Sulphides
Six sulphide species were observed in the 0.25–0.5 mm NFM-HMCs of the bedrock samples in decreasing order of abundance: arsenopyrite >> pyrite > chalcopyrite > bismuthinite = molybdenite = cobaltite ( Table 2). Arsenopyrite is the most abundant As-bearing mineral observed in this fraction. Although arsenide minerals such as loellingite and cobaltite have been observed in mineralized bedrock at NICO (e.g. Gandhi & Lentz 1990; Goad et al. 2000a), only one grain of loellingite (FeAs2) and three grains of cobaltite (CoAsS) were recovered from mineralized bedrock samples in this study. Arsenopyrite was found in the 0.25–1 mm size range of the NFM-HMCs from 15 of the 27 bedrock samples, in trace amounts in one barren host rock sample (CQA-07 445C) and up to about 129,674 grains/kg in mineralized and hydrothermally altered metasedimentary rocks collected at the #25 Zone (CQA-07 443A) and from the sub-surface Portal area (CQA-07 480Q). The grains are fresh, prismatic to massive, and have a silver colour and metallic luster. A few grains are euhedral and some grains contain trace amounts of Co, as confirmed by the SEM. Arsenopyrite sometimes contains inclusions of amphibole and bismuthinite (Fig. 5A). Arsenopyrite is also abundant (up to 1000 grains/sample) in the 0.025–0.150 mm size range of the pan concentrate of several mineralized bedrock samples. Arsenopyrite is entirely absent in the till concentrates, even in those samples located directly down-ice from mineralized zones enriched in arsenopyrite (e.g. #2 and #25 zones).
Summary of heavy minerals in bedrock (SG>3.2; <2mm fraction)
Secondary electron microscope images of non-ferromagnetic mineral grains from bedrock and till samples collected in the NICO deposit area: (A) arsenopyrite grain with ferroactinolite and bismuthinite inclusions, from Discovery Zone metasedimentary rocks; (B) chalcopyrite grain partly altered to goethite, from till collected in the Bowl Zone; (C) tourmaline grain intergrown with quartz, from porphyry breccia; (D) tourmaline grain with Ti-magnetite inclusions from metasedimentary rocks at the Tan showing; (E) silt-sized prismatic crystals of ferroactinolite grain, from Discovery Zone metasedimentary rocks; (F) sub-rounded ferroactinolite grain from till sample collected near the Discovery outcrop.
Bismuthinite was identified by its metallic steel-grey color. It was found in the NFM-HMCs of two mineralized bedrock samples, both collected from the sub-surface Portal area (Table 2). The grains mainly occur as small inclusions in 0.25–0.5 mm sized arsenopyrite grains (Fig. 5A). They are so fine-grained (mostly silt-size) that no individual grains are present in this fraction. Smaller bismuthinite grains were also recovered in abundance (up to 1% of sample) in the same two mineralized bedrock samples from the pan concentrates (<0.1 mm in size). Bismuthinite was recovered only in one till sample collected at the south end of Lou Lake (07MOB013), where a single bismuthinite grain was found in the 0.25–0.5 mm fraction ( Table 3).
Summary of heavy minerals in till (SG>3.2; <2 mm fraction)
Chalcopyrite was found in the NFM-HMCs of eight bedrock samples collected from various mineralized zones of the NICO deposit (Table 2). It occurs generally in trace amounts (a few grains) but up to 310 grains/kg at #3 Zone (CQA-07 487A2) and 3,093 grains/kg in one of the sub-surface Portal samples (CQA-07 465A). The grains are fresh, massive, and have a metallic brass-yellow colour in bedrock. About 20 grains of chalcopyrite were also observed in the pan concentrate (0.1 mm in size) of the bedrock sample collected at the #3 Zone. In till, chalcopyrite occurs in only one sample (07MOB010) but in minor amounts (18 grains/10 kg), mainly in the 0.25–0.5 mm fraction (Table 3). Some grains are fresh but most are partly altered to goethite (Fig. 5B).
Molybdenite was identified by its metallic lead-grey color. It was recovered in the 0.25–0.5 mm NFM-HMCs of only one bedrock sample collected in the sub-surface Portal area (CQA-07 482F: 61 grains/kg). Molybdenite is completely absent from the till sample concentrates.
Pyrite was found in trace to minor amounts in the 0.25–0.5 mm NFM-HMCs of several bedrock samples collected as part of this study. However it does occur in abundance (up to 27 670 grains/kg) in three bedrock samples collected at the#3 and #25 zones and from the Portal area (Table 2). The grains are sub-angular with a brassy colour and a metallic luster, but they are commonly mantled by goethite and some contain very fine, <0.1 mm magnetite inclusions. Seven representative pyrite grains with minor Co from two mineralized bedrock samples were selected for further study. Pyrite is also present generally in trace amounts in the 0.05–0.1 mm size range of eight bedrock samples as identified by panning. In till, pyrite occurs in trace amounts in a few samples, mainly in the 0.25–0.5 mm size range but also in the pan concentrate (0.025–0.2 mm) and in the 0.063–0.25 mm heavy mineral fraction (Table 3). Goethite, an alteration product of iron-bearing minerals, was found in trace amounts in a few samples and in minor amounts (32 grains/10 kg) in till sample 07MOB013. Goethite grains were also identified in both the heavy and the light 0.063–0.25 mm mineral fractions. Being non-distinctive and non-diagnostic of IOCG deposits, goethite grains and the remaining pyrite grains were not submitted for EMP analysis.
Silicates
Tourmaline was recovered from the 0.25–0.5 mm NFM-HMC fraction of three bedrock samples in considerable amounts, varying from 179 to 1088 grains/kg (Table 2). In two of these samples, both from barren brecciated porphyry dyke (CQA-07–445B&C), tourmaline grains are comprised of fractured mineral grains that are <0.05 mm in size and intercalated with small quartz grains (Fig. 5C). The third bedrock sample, from the Tan U-Fe showing, contains 179 grains/kg of dark brown, sand-sized, euhedral to massive tourmaline grains with Ti-magnetite inclusions (Fig. 5D). In till, tourmaline occurs in trace amounts (1–3 grains) in several samples, and the grains observed are black and not visually distinct, except for one till sample collected near the Tan showing (07MOB002) that contains a single distinctive, sand-sized, blue-grey tourmaline. Because the few tourmaline grains in till lack distinct physical characteristics between those collected in the vicinity of the NICO deposit and those from the up-ice samples, the utility of tourmaline as an indicator mineral at NICO is limited. Low quantities of tourmaline (1–6 grains/200 grains counted) were also found in the heavy fine sand fraction (SG>3.2: 0.063–0.25 mm) of six till samples but these have no distinct physical features either.
Ferroactinolite grains were observed, but not systematically counted, in the 0.25–0.5 mm NFM-HMC fraction of seven mineralized bedrock samples, mainly from the sub-surface Portal area, forming the major paramagnetic assemblage mineral in these samples. Ferroactinolite typically forms fragile aggregates of silt-sized prismatic crystals (Fig. 5E) or occurs as small inclusions in arsenopyrite (cf. Fig. 5A). In till, undifferentiated ferroactinolite-actinolite was found in only one sample collected near the Discovery Zone (07MOB007). This sample contains a relatively large amount of rounded grains of prismatic crystals (52 grains/10 kg) in the 0.25–0.5 mm fraction (Fig. 5F). In both the bedrock and till samples, ferroactinolite grains containing inclusions of heavier minerals such as magnetite were recovered with the ferromagnetic fraction.
Other non-ferromagnetic minerals
Other distinctive heavy minerals were observed in minor amounts or only in a few samples (Tables 2 and 3). Allanite, occurring as greenish black columnar crystals, forms the major mineral assemblage with arsenopyrite in one mineralized bedrock sample from the #2 Zone (CQA-07 441D). Scheelite, a glassy colourless mineral, was found in trace to major amounts (3,093 grains/kg) in the 0.25–0.5 mm NFM-HMCs of the two bismuthinite-bearing bedrock samples (CQA-07 458F&465A) but is absent in the till samples. The unoxidized till sample collected at depth near the Bowl Zone (07MOB010) yielded major gedrite (25% of table concentrate: 1,744 grains/10 kg), an alteration mineral that is essentially an aluminous anthophyllite. A significant number of sand-sized grains of bismutite (116 grains/10 kg), an oxidation product of bismuthinite or native bismuth, were also found in the same till sample.
Gold
Only five gold grains, 0.025–0.125 mm in size, were recovered in the pan concentrates of the entire suite of bedrock samples, mainly from the NICO deposit area (Table 2). Gold at the NICO deposit is known to occur in abundance as microscopic grains ranging from <0.001 to >0.1 mm in size, typically as inclusions within sulphides, particularly cobaltian arsenopyrite, or at sulphide and telluride grain boundaries (Goad et al. 2000a, b). Perhaps gold in the processed bedrock samples was too small to be visually seen and/or was encapsulated in sulphide minerals and therefore, was not recovered during the processing. In contrast to bedrock, gold grains (0.025–0.2 mm in size) are relatively abundant in till, averaging 11 grains/10 kg, and up to 39 grains/10 kg in one till sample collected near #25 Zone (07MOB005). Pristine gold grains are abundant in till collected down-ice of the #25 and #3 zones. None of the gold grains from till collected up-ice of NICO are pristine in shape.
Magnetite and hematite
Magnetite and hematite occur in the 0.25–2.0 mm ferromagnetic fraction of all bedrock samples in varied concentrations. The weight of the ferromagnetic fraction varies between 0.2 and 328.6 g per kg, on average 38.6 g (Table 2). It is abundant in the few bedrock samples collected from the Discovery (241 g/kg) and East (160 g/kg) zones and from two sub-surface samples from the Portal area (96 and 329 g/kg). It consists of individual magnetite (dominant) or hematite grains, bedrock fragments with disseminated magnetite (Fig. 6A), non-ferromagnetic minerals with small magnetite inclusions, and magnetite-hematite samples mixtures. In till, the 0.25–2 mm ferromagnetic fraction is generally smaller than in bedrock and varies from 0.5–87.5 g per kg, on average 8.5 g. It is significant in the till sample collected down-ice from the Discovery Zone (07MOB007: 88 g/10kg), forming close to 9% by weight of the <2 mm table concentrate. Similar to bedrock, the ferromagnetic fraction in till consists of individual magnetite (magnetite) or hematite grains (Fig. 6B), polymetallic grains including magnetite, and some hematite-magnetite mixtures. Examination of the 0.063–0.25 mm ferromagnetic fraction of till shows that many magnetite grains are irregular and have a weathered appearance while some grains have a high luster and a conchoidal fracture but few are euhedral. Several magnetite grains in this fraction appear to be intermixed with hematite. Hematite is also present in the 0.063–0.25 mm non-ferromagnetic fraction.
Secondary electron microscope images of ferromagnetic grains from bedrock and till samples collected in the NICO deposit area: (A) aggregate of magnetite and other minerals from a sub-surface bedrock sample; (B) magnetite grain from till collected at the south end of Lou Lake.
CHEMISTRY OF INDICATOR MINERALS IN THE 0.25–2.0 MM FRACTION
Sulphides
One of the most abundant heavy minerals recovered in mineralized bedrock is arsenopyrite. Compositions of 318 grains range from pure FeAsS to up to 17 wt.% Co trending towards cobaltite (CoAsS) endmember composition. The highest Co values are in grains from bedrock collected at the NICO #3 Zone porphyry, the NICO stockpile and the Discovery Zone magnetite-amphibole-biotite altered metasedimentary rocks (Fig. 7). Arsenopyrite compositions reported by Sidor (2000) for the NICO deposit are similar to those reported here. Sidor (2000) reported a high degree of substitution of Co for Fe particularly in gold-rich ironstone (Co range from 1.2–17.5 wt.%), which also contained endmember cobaltite and slightly arsenian pyrite, whereas unmineralized (Au-poor) ironstone contained arsenopyrite with only up to 4.9 wt.% Co and As-poor pyrite but no cobaltite.
Fe versus Co concentrations in arsenopyrite grains from bedrock (n=318). Co ± Ni substitution for Fe trends towards the cobaltite endmember composition (at c. 35% Co and 0% Fe, not shown).
Seventy-one chalcopyrite grains were analyzed from bedrock; mainly from hydrothermally altered metasedimentary rocks and 22 grains from porphyry dyke samples. Chalcopyrite is near stoichiometric in composition in most ore assemblages and this is also the case here (see McMartin et al. in press). There is no discernible difference in the composition of chalcopyrite from hydrothermally altered metasedimentary rocks versus those from porphyry samples. Eight grains of chalcopyrite from till sample 07MOB010 were analyzed and, similar to those recovered from bedrock, they are near stoichiometric in composition.
Twenty-one molybdenite grains from mineralized and hydrothermally altered metasedimentary rocks in a NICO bulk sample stockpile were analyzed. Molybdenite, like chalcopyrite, is near stoichiometric in composition with only trace amounts of Se (up to 0.6 wt.%) substituting for S (McMartin et al. in press). Seven pyrite grains were analyzed: five from mineralized and hydrothermally altered metasedimentary rocks and two from a mineralized porphyry dyke near the # 25 zone. The pyrite grains are stoichiometric in composition, and the grains from porphyry dyke contain only slightly higher As and Bi than those from the metasedimentary rocks (McMartin et al. in press).
Silicates
Fifty-nine tourmaline grains were analyzed: 20 grains from tourmaline-rich breccia (CQA-07 445C), 19 grains from porphyry (CQA-07 445B) and 20 from hydrothermally-altered metasedimentary rocks at the Tan showing (CQA-07 437A-1). Tourmaline in altered metasedimentary rocks at the Tan showing have considerably higher contents of CaO and Mg-# (closer to the dravite endmember) compared to tourmaline from porphyry and tourmaline breccia, which overlap in their compositional range and correspond to schörl (Fig. 8).
Mg-# versus wt.% CaO for tourmaline grains from bedrock (n=59). Tourmaline from metasedimentary rocks hosting the Tan showing has considerably higher contents of CaO and Mg-# compared to tourmaline from barren porphyry and breccia.
Twenty-seven ferroactinolite/actinolite grains were analyzed from one till sample collected near the Discovery Zone (07MOB007) and from two mineralized bedrock samples collected at the East Zone and Discovery Zone (CQA-07 447A and CQA-07 228A). The amphibole grains from till sample 07MOB007 are Fe-poor actinolite grading to actinolitic hornblende and ferro-actinolitic hornblende (Fig. 9). Mineralized bedrock contains much more Fe-rich amphibole than the till sample, spanning the range from ferro-actinolite to ferro tschermakitic hornblende (Fig. 9). The amphibole compositions for mineralized bedrock are similar to those reported by Sidor (2000) for mineralized and unmineralized ironstones and hornfels at NICO, which have variable but generally low mg-#(<50) and variable Si/Al ratios with Si(IV) ranging from c. 6.0–8.0.
Si(IV) versus Mg/(Mg+Fe2+) for amphibole grains from bedrock and till. The mineralized bedrock samples contain much more Fe-rich amphibole than the till sample, from which most amphibole grains were classified as Fe-poor actinolite.
Ferromagnetic fraction
Of the 46 ferromagnetic grains from till and 211 from bedrock analyzed by EMP, 41 and 186 consisted of magnetite whereas 5 and 25 grains were hematite, in till and bedrock respectively. The chemical composition of the NICO magnetite and hematite can be plotted in a preliminary discriminant diagram defined by Beaudoin & Dupuis (2010) to fingerprint a range of mineral deposit-types based on iron oxide composition. In a Ti+V v. Ni/(Mn+Cr) plot (Fig. 10), most magnetite and hematite grains from bedrock collected in altered metasedimentary rocks at the Tan Fe-U showing have higher Ti+V compositions compared to grains from bedrock collected at or near the NICO deposit. There is a reasonable differentiation between magnetite and hematite from NICO altered metasedimentary rocks and magnetite and hematite from regional metasedimentary rocks. In till, magnetite and hematite from samples collected up-ice of NICO have higher Ti+V compositions compared to grains from till collected over or directly down-ice of the NICO deposit and have a similar composition to many grains from bedrock collected in altered metasedimentary rocks at the Tan Fe-U showing. There is a good correspondence between the composition of grains from bedrock collected in altered NICO metasedimentary rocks and grains from till collected over or directly down-ice of NICO. These preliminary results indicate magnetite and hematite compositions have the potential to fingerprint the mineralization at NICO; however the low number of analyses prevents any further interpretation and more grains per sample must be analyzed to properly evaluate the usefulness of the method.
Iron oxide composition from till and bedrock samples plotted on a Ti+V v. Ni/(Mn+Cr) diagram to help discriminate magnetite and hematite from the NICO deposit, barren host rocks near NICO and regional rocks (see Table 1 for classification of bedrock samples). Magnetite and hematite from till samples are divided between those samples collected over or <150 m down-ice from NICO, away from NICO (<3 km) and 10 km up-ice of NICO. Till and bedrock samples collected near the TAN showing are also differentiated. For both till and bedrock samples, only Fe-oxide grains having >0.01 Ni/(Mn+Cr) are plotted.
DISCUSSION
Shallow till sampling (<1 m) is an effective exploration method in discontinuous and continuous permafrost terrain of the Canadian Shield and has been widely used for geochemical and indicator mineral sampling by government agencies and exploration companies (e.g. McMartin & McClenaghan 2001 and references therein). However, in weakly oxidized C horizon soil developed on till, easily weathered ore minerals such as sulphides and tellurides are oxidized. Chalcopyrite, which usually survives near-surface weathering longer than the other sulphides (e.g. Peuraniemi 1984; Averill 2011), has been used as an indicator mineral for metamorphosed or magmatic massive sulphide exploration (Averill 2001), and Ni-Cu-PGE and porphyry Cu deposits (Averill 2011). At NICO, many of the sulphides that are highly to moderately abundant in the bedrock (i.e. arsenopyrite, chalcopyrite, pyrite) are absent or rare in till. This scarcity suggests that sulphides are unstable in soils and thus poorly preserved in the C-horizon developed on till at shallow depth. Although arsenopyrite is abundant in mineralized bedrock at NICO, arsenopyrite is absent in C horizons developed on tills indicating it did not survive post-glacial weathering. Chalcopyrite only occurs in one till sample collected below the zone of oxidation at 3.5 m depth in the Bowl Zone. This pattern suggests that chalcopyrite is not a useful indicator of mineralization in till samples at NICO. However perhaps in the northern part of the GBMZ, where soils developed on till are generally thin and immature because of relatively less chemical weathering, or in areas where chalcopyrite is the predominant sulphide in mineralized bedrock, chalcopyrite has better potential as an indicator mineral for IOCG mineralization.
Some of the indicator minerals are present in bedrock and till concentrates as small inclusions intergrown in larger minerals; it is difficult to recover them in the sand-sized HMCs. For example bismuthinite, present as small grains (<0.1 mm) in the pan concentrates of two mineralized bedrock samples, occurs as brittle, acicular silt-sized inclusions in 0.25–0.5 mm sized arsenopyrite grains from the same two samples. When the arsenopyrite is oxidized in the soils, the bismuthinite grains are released and likely destroyed as well. No bismuthinite grains were recovered from the pan concentrates of any till sample indicating that, even if present in sufficiently large concentration and in coarse-grained fractions, bismuthinite does not survive glacial transport and/or post-glacial weathering.
Tourmaline, an accessory alteration mineral at NICO, occurs in the 0.25–0.5 mm NFM-HMCs fraction of two heavily brecciated porphyry dyke samples closed to NICO. In these two samples, tourmaline sand sized grains, classified as schörl, are heavily fractured and intercalated with quartz. The grains are therefore not sufficiently heavy to be readily recovered in the SG>3.2 fraction, or are too small to be recovered in the sand-sized fraction of till samples as they probably break apart during glacial transport. Although tourmaline has shown potential as a resistate indicator mineral because it is able to survive weathering and mechanical dispersal in both deeply weathered and glaciated terrains (e.g. Slack 1982; Ramsden et al. 1993; Averill 2001, 2007), the fragmented nature of the tourmaline grains in crackle breccias or their small size in the altered metasedimentary rocks hamper the use of tourmaline as an indicator mineral for the NICO deposit. In contrast, euhedral to massive sand-sized tourmaline grains with high contents of CaO and Ti-magnetite inclusions occur in abundance in bedrock from the nearby Tan Fe-U showing, up-ice of NICO. Tourmaline has been observed in greater abundance in other IOCG settings within the GBMZ, namely in the Echo Bay District (Mumin et al. 2010) and the De Vries area (Ootes et al. 2010). If sufficiently abundant and coarse-grained in the altered bedrock, and if having distinct physical characteristics in mineralized rocks versus other sources, tourmaline could have some potential as an indicator mineral in the GBMZ. However, because of its mid-density specific gravity (2.96–3.31), a significantly diluted heavy liquid (SG 3.0) should be used to separate and recover tourmaline.
Ferroactinolite, although forming a pervasive alteration mineral in mineralized bedrock at NICO (cf. Corriveau et al. 2010), is scarce or absent in till concentrates. It typically occurs as loose, aggregates of silt-sized prismatic crystals, as small inclusions in other mineral grains (i.e. arsenopyrite) or as part of polymetallic grains in several of the bedrock samples observed. Ferroactinolite was probably crushed during glacial transport and comminution; hence it has a poor preservation potential in till. One till sample collected directly down-ice from the Discovery outcrop does contain a relatively large amount of sand-sized actinolite grains (although Fe-poor) suggesting that the mere presence of these grains in till indicates a close proximity to intensively altered (mineralized?) bedrock.
Some of the indicator minerals at NICO are not sufficiently abundant in bedrock to be detected in till, or are too soft to survive glacial transport. For example molybdenite, a very soft sulphide, is generally scarce in the bedrock concentrates and absent in the tills. Scheelite, a useful and distinctive resistate indicator mineral for skarn, VMS, lode gold and tungsten deposits in both glaciated and deeply weathered terrains (e.g. Lindmark 1977; Averill 2001), occurs as an accessory mineral in mineralized metasedimentary rocks at NICO (Goad et al. 2000a). It was recovered in the two bismuthinite-bearing sub-surface bedrock samples but not in till samples. Significant abundances of gedrite and bismutite occur in the unoxidized till sample in the Bowl Zone. In effect, this till sample is the only one collected at NICO that yielded a distinct suite of heavy minerals (with chalcopyrite) that could be used as an indicator of mineralization.
Relatively large numbers of gold grains were found in till overlying and near mineral occurrences of the NICO deposit area (up to 39 grains/10 kg) in comparison with those collected over barren host rocks (up to 10 grains/10 kg) and background terrain (up to 4 grains/10 kg). The pristine nature of the gold grains in till collected near mineralization likely indicates a local bedrock source and a short distance (10s to 100s metres) of glacial transport. Documenting gold grain abundance, size and shape in till (e.g. Grant et al. 1991; McClenaghan & Cabri 2011) represents a valuable surface exploration method for gold-bearing IOCG deposits in the GBMZ.
Iron oxides are abundant in IOCG deposits, resistant to weathering and mechanical transport, and separate easily from heavy mineral concentrates using magnetic methods. For these reasons, they have the potential to be used as indicator minerals for mineral exploration (Beaudoin & Dupuis 2010). Furthermore, the differing chemical composition of iron oxides from a number of representative deposits world-wide allows definition of discriminant diagrams useful to fingerprint the signatures of various mineral deposit-types (Beaudoin & Dupuis 2010). At NICO, ferromagnetic minerals occur in abundance in mineralized bedrock and the large quantity of 0.25–2 mm sized FM-HMCs in till collected down-ice from the Discovery Zone is significant. Magnetite and hematite composition using the Ti+V versus Ni/(Mn+Cr) plot shows significant promise to differentiate mineralized metasedimentary rocks from regional metasedimentary rocks, and till collected over or directly down-ice of NICO from regional till. The EMP results on magnetite and hematite should be used with caution however since the number of analyses is insufficient to properly define the trends observed.
RECOMMENDATIONS FOR FURTHER WORK
The relatively low number of bedrock and till samples collected as part of this first orientation study adds some limitations to the development of IOCG indicator mineral methods in glaciated terrain. Additional bedrock and till samples are required to properly define the heavy mineral signature of the IOCG mineralization and derived glacial sediments at NICO and to establish the mineralogical signature of non altered background rocks. As part of the current project, a suite of additional till and bedrock samples were collected around the NICO deposit and in background terrain within the Slave Province in 2009 and 2010. Furthermore, due to the extraordinary range of polymetallic IOCG systems and intrinsic complexity, detailed bedrock and till sampling was completed in the vicinity of the Sue-Dianne deposit and near additional mineral occurrences across the GBMZ. Analysis of these samples is ongoing and research is continuing on analytical method development to better fingerprint the IOCG mineralization.
For example, further tests will be conducted to determine the optimum grain size, number of grains and other sample preparation methods for using iron oxide mineral chemistry in mineral exploration over the GBMZ (cf. Beaudoin et al. 2009). Better differentiating the bedrock samples with respect to alteration zoning and mineralization vectoring, and applying these differentiations to drift prospecting, using a combination of mineralogy and geochemistry, will also help in developing vectors to IOCG mineralization in the GBMZ.
More specifically at NICO, some of the indicator minerals observed are too small for conventional recovery by visual picking of the 0.25–2 mm fraction (i.e. bismuthinite, tourmaline). Although heavy mineral slides of the 0.063–0.25 mm fraction from till samples were prepared and examined under the microscope, newer techniques such as quantitative evaluation of materials by rapid scanning electron microscope (e.g. MLA, QEMSCAN) are now available that could help in identifying and quantifying the mineralogy of the smaller (<0.25 mm) HMC fractions (i.e. Gottlieb et al. 2000; Oberthür et al. 2008; McClenaghan 2011).
The application of trace element determinations in indicator minerals is a growing field of studies (Griffin et al. 1989; Belousova et al. 2002). Although the EMP plays a critical role in providing major and minor element data for mineral grains, the use of laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS) is recommended to measure trace element concentrations in indicator minerals for properly evaluating geological processes and environments (Jackson 2009). In-situ trace element analysis of the most resistive minerals (i.e. tourmaline, magnetite) by LA-ICP-MS has been initiated on the NICO bedrock and till samples and further work using this more sensitive technique will be completed on samples collected in 2009 and 2010.
CONCLUSION
Studies of the heavy mineral signature of the NICO deposit area and overlying glacial sediments represent the first detailed study of IOCG indicator minerals in glaciated terrain. The results from this study demonstrate that gold grain counts and iron oxide composition have the best potential to fingerprint the NICO IOCG mineralization. Gold grains are more abundant in till samples collected immediately down-ice from several surface occurrences at NICO and no pristine gold grains were found in till up-ice of the NICO deposit. Using trace elements signatures of Fe-oxides to fingerprint IOCG deposits and identify metal-rich debris down-ice of the deposit is also a potential indicator mineral method for IOCG exploration. Particularly magnetite and hematite from till samples collected up-ice of NICO have higher Ti+V compositions compared to magnetite and hematite from till collected over or directly down-ice of the NICO deposit. The non-ferromagnetic heavy minerals at NICO are not particularly effective as indicator minerals of IOCG mineralization. They are either not chemically stable in surface glacial sediments, not sufficiently coarse-grained to be recovered using tabling or visually distinctive to be readily identified, not sufficiently abundant in the mineralized bedrock, or not sufficiently heavy, or not robust enough to survive glacial transport. However abundant ferroactinolite and ferromagnetic grains in a till sample collected directly down-ice of the Discovery Zone provides a good vector to mineralized bedrock in this area. Furthermore, the presence of chalcopyrite, gedrite and bismutite grains in a till sample collected at 3.5 m depth over the Bowl Zone suggests that sampling the less-weathered material in the deeper portions of the C-horizon soils developed on till, if possible, may be more effective than shallow surface till sampling. The potential utility of Fe-oxides and other unusual, distinctive and propitious heavy minerals as indicators of IOCG deposits, together with till geochemistry, will be further tested with additional bedrock and drift samples collected around the NICO and Sue-Dianne deposits and elsewhere across the GBMZ.
Acknowledgments
This research was conducted within the Targeted Geoscience Initiative 3 and the Geo-mapping for Energy and Minerals Program of the Geological Survey of Canada, in collaboration with the NW Territories Geoscience Office through funding from the Strategic Investments in Northern Economic Development Program at Indian and Northern Affairs Canada. We are grateful to Fortune Minerals for logistical support at NICO, K. Venance and P. Hunt (GSC) for electron microprobe and SEM analysis, Overburden Drilling Management Ltd. (S. Averill) for the customized heavy mineral processing, I. Kjarsgaard for mineral identifications, C. Dupuis for magnetite and hematite EMP analysis, and the DIVEX research network for supporting the iron oxide mineral chemistry studies. K. Kelley, R. Paulen, and B. McClenaghan are thanked for providing careful reviews.
- © 2011 AAG/Geological Society of London