Diversity of Net-Textured Sulfides in Magmatic Sulfide Deposits: Insights from the Eagle’s Nest Ni-Cu-(Platinum Group Element) Deposit, McFaulds Lake Greenstone Belt, Superior Province, Canada

Understanding the textures of the sulfides in magmatic Ni-Cu-(platinum group element; PGE) deposits is important to learn how they formed, for determining variations in tonnages and grades for mining, and for designing beneficiation strategies. Although some deposits are relatively homogeneous in terms of sulfide content and/or texture (e.g., Dumont: Eckstrand, 1975; Jinchuan: Tang, 1993; Mt. Keith: Barnes, 2007), other deposits are commonly zoned (e.g., Kambalda: Ewers and Hudson, 1972; Gresham and Loftus-Hills, 1981) or exhibit complex variations indicating multistage ore emplacement (e.g., Raglan: Lesher, 2007; Alexo: Houlé et al., 2011; Eagle’s Nest: Mungall et al., 2010; this study).

The textures of sulfide assemblages vary with abundance of sulfide and the nature and form of the sulfide and gangue components. Disseminated mineralization typically comprises <15 wt % sulfide dispersed between silicate phases and may be fine or coarse (blebby or globular), lightly to heavily distributed, and uniformly or nonuniformly (patchy) dispersed. Net-textured mineralization typically comprises 15–60 wt % sulfide that forms a more or less continuous film between silicate phases (most commonly fine-grained olivine) and may also be uniformly or nonuniformly (patchy) distributed and may contain orthopyroxene or clinopyroxene oikocrysts. Semimassive ores typically contain 50 to 80 wt % sulfide and may contain blebs of former silicate melt (emulsion texture), cumulus silicates, anteliths (fragments derived from earlier crystallization products in the system), local/exotic xenoliths, and/or veins. Massive sulfides typically contain >80 wt % sulfide consisting of various minerals (commonly pyrrhotite, chalcopyrite, and pentlandite) and similar gangue phases.

The origins of many of these textures have been reviewed by Barnes et al. (2017, 2018), and, of these, the formation of net texture is the most complicated and the least understood. Several hypotheses have been proposed, none of which are mutually exclusive: (1) dynamic flow segregation of silicate melt, olivine, and sulfide liquid (e.g., Hudson, 1972), (2) static density segregation of silicate melt, olivine, and sulfide liquid (“billiard-ball model”: Naldrett, 1973; Usselman et al., 1979), (3) physical transport/deposition of olivine and sulfide liquid in more or less existing proportions (Tang, 1993; De Waal et al., 2004), (4) downward (gravitational) percolation of sulfide liquid into a network of cumulus olivine (e.g., Chung and Mungall, 2009; Mungall et al., 2010; Barnes et al., 2018) during or subsequent to mechanisms 1 and 2, and/or (5) surface tension-driven infiltration of sulfide into adjacent host rocks (Huminicki, 2004).

The Eagle’s Nest deposit in the McFaulds Lake greenstone belt of northern Ontario (Fig. 1) contains a wide range of subfacies of net-textured mineralization. Many parts are texturally and mineralogically well preserved, and some contacts between the different net textures are gradational. These textural variations are exposed in thousands of meters of diamond drill core, making Eagle’s Nest an excellent place to study the origin of the various subfacies of net texture and their constraints on ore emplacement and localization. The specific goals of this paper are to (1) characterize the textural and sulfide mineralogical variations, (2) determine if whole-rock geochemical data can be used to distinguish between sulfide textures, and (3) use textural variations and sulfur isotopes to constrain potential S sources and emplacement of the Ni-Cu-(PGE) mineralization.

The 2.83–2.66 Ga McFaulds Lake greenstone belt is located at the edge of the Hudson Bay lowlands within the northwestern Archean Superior province (Fig. 1A). The McFaulds Lake greenstone belt contains metasedimentary rocks, mafic-felsic metavolcanic rocks, and ultramafic, mafic, and felsic plutonic rocks (Mungall et al., 2010; M.G. Tuchscherer et al., unpub. report, 2010; Metsaranta et al., 2015; Metsaranta and Houlé, 2020). A 2736–2732 Ma mafic-ultramafic suite, the Ring of Fire intrusive suite, has been defined by geophysical surveys, diamond drilling, geochemistry, and U-Pb single zircon geochronology (Metsaranta et al., 2015; Houlé et al., 2020).

Houlé et al. (2019, 2020) subdivided the Ring of Fire intrusive suite into (1) an areally more extensive Ekwan River subsuite composed of gabbro to ferrogabbro and pyroxenite containing magmatic Fe-Ti-V-P mineralization (e.g., Butler, Thunderbird) and (2) an areally more restricted Koper Lake subsuite composed of dunite, peridotite, chromitite, pyroxenite, and gabbro containing Cr (e.g., Black Thor, Black Label, Black Creek, Big Daddy, Black Horse, Blackbird) and Ni-Cu- (PGE) (e.g., Eagle’s Nest, Eagle Two, Blue Jay) mineralization (Fig. 1B).

The 50- to 1,500-m-thick complex of ultramafic rocks hosting the Cr and Ni-Cu-(PGE) mineralization in the Ring of Fire intrusive suite appears to be contiguous over at least 16 km and has been defined as the Esker intrusive complex (Houlé et al., 2019, 2020), which includes the Black Thor intrusion, the Double Eagle intrusion, and the Eagle’s Nest dike (Fig. 1B). The apparent northeast-southwest orientation of the Eagle’s Nest deposit projected to surface in Figure 1B reflects variations in dip along the steep plunge of the dike; on surface the dike is oriented north-south. The contacts and layering in the Esker intrusive complex, including the stratiform Cr mineralization in the Black Thor intrusion and Double Eagle intrusion, are subvertical, and the lithologies fractionate toward the south-southeast (e.g., M.G. Tuchscherer, unpub. report, 2010; Carson et al., 2015; Houlé et al., 2020), so the Esker intrusive complex is interpreted to young to the south-southeast and to have been structurally rotated toward that direction from an originally subhorizontal orientation (e.g., Houlé et al., 2020).

The Black Thor intrusion is underlain by a trough-shaped feeder that extends up to ~1 km to the north and hosts the Blue Jay (formerly AT-12) and Blue Jay Extension (formerly AT-12 Extension) Ni-Cu-(PGE) occurrences (Farhangi et al., 2013). It contains a Late Websterite phase that invaded the lower units of the Black Thor intrusion, including the Black Label chromite horizon (Spath et al., 2015; Spath, 2017). The Double Eagle intrusion is underlain by the Eagle’s Nest dike, which hosts the Eagle’s Nest Ni-Cu-(PGE) deposit. It has not been physically linked to the Double Eagle intrusion, and there is a significant structural break (i.e., Frank shear zone) between Eagle’s Nest and the Double Eagle intrusion.

All of the rocks in the Esker intrusive complex, including the Eagle’s Nest area, have been metamorphosed to lower to middle greenschist facies (Metsaranta and Houlé, 2020). Relict igneous chromite is often preserved, and relict igneous olivine and pyroxene are locally preserved, but all sulfides have reequilibrated to low-temperature assemblages (see review by Naldrett, 2004). Because relict igneous textures are well preserved, for simplicity rocks will be referred to by their igneous names.

The rocks in the McFaulds Lake greenstone belt are very poorly exposed, so all of the work in this study focused on diamond drill cores. All 80 mineralized drill core intervals were examined, and the mineralized parts of 53 representative cores through all accessible parts of the system were logged in detail in February 2017 and August 2017 at the Noront Resources Esker exploration camp. One hundred sixty-three mineralized samples composed of 20- to 30-cm segments of half-sawn NQ (4.5 cm diam) core and 35 unmineralized samples composed of ~20-cm segments of whole or half-sawn core were taken for detailed macroscopic, microscopic, and geochemical study. Sample preparation and petrographic, micro-X-ray fluorescence (μXRF), mineral chemical, whole-rock geochemical, and S isotope analytical methods are given in Appendix 1. A list of samples analyzed for whole-rock geochemistry is given in Appendix Table A1.

Sulfide mineralization at the Eagle’s Nest deposit occurs in several broad textural facies (disseminated, net, and semimassive to massive sulfides), which can be divided into several subfacies (see Table 1; Figs. 2, 3; section on mineralized rocks below).

Estimates of broad textural facies distributions are based on S assays from 340 drill hole intersections in the Noront Resources database (over 10,000 samples). The vast majority of samples fall into the disseminated (<13 wt % sulfide) and net-textured (13–33 wt % sulfide) categories (collectively 85% of mineralization), whereas semimassive to massive (50– 100 wt % sulfide) make up the rest of mineralization (Table 1; App. Fig. A1).

The greatest proportion of Ni-Cu-(PGE) mineralization and Ni grade is restricted to the narrow northwest contact of the intrusion, extending almost continuously from surface to ~1,300 m (Fig. 2). In the upper parts of the deposit, the drill hole spacings are 10 to 30 m, making them more definitive; however, below ~450-m depth, the spacings are 50 to 100 m, making them less definitive. Nevertheless, there is generally good agreement between adjacent drill holes, so the results appear to be representative.

The orebody can be subdivided into three major domains, each of which is zoned from north to south (Fig. 2).

Upper embayment: The upper embayment between surface and 270-m depth contains 50 to 70 m of massive sulfide along the basal contact overlain by 50 to 100 m of net, disrupted net, and disseminated mineralization that grades through 30 to 60 m of disseminated sulfide in peridotite into 80 m of barren pyroxenite.

Topographic high: A paleotopographic high between 270- and 350-m depth contains 50 m of barren peridotite and very minor disseminated sulfide.

Lower embayment: A lower embayment between 350- and 1,300-m depth can be subdivided into five subdomains:

1. An upper part between 350 and 450 m that is a 20-m-wide × 30-m-long × 20- to 30-m-thick zone of massive sulfide grading upward to a 25- to 50-m-thick zone of disseminated sulfide;

2. An upper-middle part between 450- and 600-m depth that is underlain by up to 30 cm of 0.05- to 10-m-thick sulfide veins with sharp margins in tonalitic country rocks that is overlain by 50 to 150 m of leopard net texture and minor disrupted net texture and 100 m of barren peridotite and pyroxenite with only very sparse disseminated sulfides;

3. A middle part between 600- and 800-m depth that comprises 30 to 70 m of leopard net, disrupted net, and lesser pinto net and inclusion net textures with little to no massive sulfide and very little disseminated sulfide (the contacts between disrupted net and leopard net-textured zones are very irregular and gradational);

4. A lower-middle part between 800 and 1,050 m that comprises 20 to 60 m of leopard net and 5 to 10 m of disrupted net texture, underlain by 5- to 10-m-thick sulfide veins and overlain by disseminated sulfide and 100 to 150 m of barren peridotite;

5. A lower part between 1,050 and 1,300 m that comprises 20–30 m of leopard net and highly altered leopard net with minor massive sulfide veins and 30 m of barren peridotite—one hole at 1,240-m depth intersected massive sulfide, but drilling has not progressed past this point to define its extent.

The Eagle’s Nest dike is ~500 m wide (N-S) × ≤85 m thick (E-W) × >1,500 m deep (Figs. 1, 2) with a narrow, flattened tube-like morphology. It composed mainly of peridotite and lesser pyroxenite, grading to the southeast from mineralized peridotite through unmineralized peridotite to minor zones of unmineralized pyroxenite, with fine-grained pyroxenitic margins. In some areas the tonalitic country rocks are contact metamorphosed. Allowing for the same structural rotation that has affected the Esker intrusive complex and consistent with the asymmetric lithological zoning, it is interpreted to have originally been a subhorizontal blade-shaped dike (Mungall et al., 2010; Zuccarelli et al., 2018).

The shape of the intrusion varies with depth. From surface to 250-m depth, the intrusion is 500 m wide (thick) and the northwest (original basal) contact defines ~150-m-wide (deep) embayment. Between 250- and 360-m depth, the basal contact protrudes to the southwest, representing an ~50-m-wide (high) topographic feature. Between 360- and 1,500-m depth the intrusion narrows to 50 m wide (thick) and the basal contact defines an ~150-m-wide (deep) embayment.

The northwest (original lower) and lower-lateral (northeast-southwest) contacts between the ultramafic intrusion and the tonalite country rocks are highly irregular and are commonly characterized by 10- to 50-cm-thick intervals of peridotite containing angular 30- to 60-cm tonalite fragments and by 10- to 30-cm-thick veins of peridotite and pyroxenite intruding tonalite. Lower contacts between massive sulfide and country-rock tonalite are relatively sharp and locally sheared with 2- to 5-mm-thick zones of chalcopyrite along the contact and 1- to 5-cm-thick veins of Cu-rich sulfides that extend ~30 cm into the country rocks. Lower contacts between net-textured sulfides are rich in chalcopyrite and are also sometimes sheared.

The southeast (original upper) and upper-lateral (northeast and southwest) contacts are characterized by a 5- to 10-m-thick fine-grained pyroxenitic margins that have been altered to tremolite-actinolite (occasionally containing 1- to 2-m angular fragments of tonalitic country rock) and 10- to 30-cm veins of pyroxenite, which intrude the adjacent tonalite. The veins are altered to fine-grained tremolite-actinolite, marked by 2- to 5-cm-thick fine-grained ultramafic margins and by 5- to 10-cm-thick margins of baked/melted tonalite. The baked/melted tonalite is marked by local anhedral/truncated plagioclase crystals and pockets of granophyre. There are no significant sulfides along the southeast (original upper) contact of the dike.

The rocks surrounding the Eagle’s Nest dike include tonalite, iron formation, and gabbro. The predominant country rock adjacent to and near the Eagle’s Nest intrusion is tonalite containing ~30% quartz, ~50% plagioclase, and ~20% biotite and hornblende.

Rafts of silicate-oxide-(sulfide) facies iron formation up to 2 to 20 m thick locally occur within the tonalite (mediumgrained grayish-white plagioclase, quartz, and lesser biotite) northeast of Eagle’s Nest (see Fig. 1; Houlé et al., 2020; Metsaranta and Houlé, 2020). The iron formation is composed of interbedded layers dominated by silicates (fine-grained quartz ± stilpnomelane ± grunerite ± allanite), oxide (finegrained magnetite), and lesser chlorite (probable mafic volcaniclastic horizons; Carson et al., 2015) with fine-grained acicular ferristilpnomelane along the contacts between the silicate and oxide layers. Rafts of 25- to 60-m-thick leuco-mesomelagabbroic rocks also occur in the tonalite. The plagioclase is predominantly fine to medium grained but locally porphyritic and is normally altered to chlorite and amphibole (Carson et al., 2015).

Ultramafic cumulate rocks host all of the sulfide mineralization in the Eagle’s Nest deposit. They range in composition from olivine websterite to lherzolite in terms of cumulus mineralogy (App. Fig. A2), and most are variably altered and metamorphosed to serpentine-tremolite-chlorite-magnetite and lesser talc-carbonate-chlorite assemblages. In order of abundance, they include the following:

1. Lherzolites are normally light gray to dark gray and strongly magnetitic. They contain 40 to 75 modal % olivine (2–10 mm; normally altered to serpentine-magnetite along rims and fractures), 10 to 40% oikocrystic (olivine inclusions) orthopyroxene (3–8 mm, often altered to anthophyllite), 10 to 40% cumulus clinopyroxene (0.2–0.7 mm, altered to tremolite), and 3 to 5% intercumulus/cumulus chromite (0.2–0.7 mm, variably altered to and/or overgrown by magnetite). Relict igneous olivine is normally surrounded by serpentinized rims and/or contains serpentine-magnetite filled fractures.

2. Olivine websterites are light bluish-gray and essentially nonmagnetic. They contain 10 to 30% cumulus olivine (2–10 mm, commonly but not always altered to serpentine-magnetite), 20 to 60% oikocrystic orthopyroxene (5–10 mm, altered to talc), and 10 to 50% cumulus clinopyroxene (1–3 mm, altered to actinolite). Altered pyroxenes also contain minor muscovite and biotite.

3. Wehrlites are dark gray-blue and weakly magnetic. They contain 40 to 60% olivine (2–10 mm, altered to serpentine-magnetite along rims and fractures), <5% oikocrystic orthopyroxene (3–8 mm, altered to anthophyllite), 40 to 60% intercumulus clinopyroxene (0.5–4 mm, altered to tremolite), and <2% cumulus 0.2- to 0.7-mm chromite.

4. Harzburgites are uncommon but are normally light gray-blue and strongly magnetic. They contain 60 to 75% cumulus olivine (2–10 mm, altered to serpentine-magnetite along rims and fractures), 20 to 40% oikocrystic orthopyroxene (3–8 mm, altered to anthophyllite), <5% intercumulus clinopyroxene (0.5–4 mm, altered to tremolite), and <2% cumulus/intercumulus 0.2- to 0.7-mm chromite.

Contacts between ultramafic rocks and the country-rock tonalite are often diffuse and highly irregular, with localized brecciation of tonalite clasts into the dike margins. The chilled margins of the dike examined in this study were highly altered to chlorite-biotite, retaining no original mineralogy or textures.

Olivine: Almost all of the olivine in the ores and host rocks at Eagle’s Nest is altered to serpentine-magnetite or talc-magnesite pseudomorphs, but relict igneous olivine is sporadically preserved within zones of late pyroxenite crosscutting net-textured mineralization and peridotite.

Olivine is the most common primary silicate mineral within the Eagle’s Nest dike, occurring as small (2–10 mm) ellipsoidal to prismatic black primocrysts in all of the mineralized and unmineralized rocks forming vermiform networks in some net-textured sulfides and in coarse (2–5 mm) ellipsoidal dunite-peridotite anteliths in inclusion net-textured sulfides. The vermiform networks resemble some of the habits of crescumulate olivine at Kambalda (Lesher, 1983), but the alteration precludes verification that it formed the optically continuous networks that characterize crescumulate olivine. The forsterite (Fo) content of the olivine at Eagle’s Nest is quite restricted and ranges from Fo₈₂ to Fo₈₆.

Pyroxene: The vast majority of pyroxene in the ores and host rocks at Eagle’s Nest is altered. Even where uncommon fresh orthopyroxene is preserved, it is almost entirely crosscut by fractures containing alteration minerals and could not be analyzed. This hampers discrimination between orthopyroxene and clinopyroxene, but orthopyroxene normally forms coarse (0.5–1 cm) oval to circular oikocrysts containing chadacrysts of fresh olivine and is normally altered to gray talc ± serpentine ± anthophyllite psuedomorphs, whereas clinopyroxene normally forms isolated intercumulus crystals and is normally altered to gray-blue tremolite pseudomorphs. Orthopyroxene appears to be most abundant in pinto net-textured sulfides among black serpentinized olivine and interstitial sulfide.

The mineralized lithologies at Eagle’s Nest are identical to the barren lithologies described above with a variable amount of sulfides making a wide range of sulfide textural facies across the deposit. Many of the sulfide facies at Eagle’s Nest (Table 1) are similar to those present in many other komatiite-associated Ni-Cu-(PGE) deposits (e.g., Alexo: Naldrett, 1966; Houlé et al., 2011; Kambalda: Ewers and Hudson, 1972; Raglan: Lesher, 2007; Bazilevskaya, 2009; see review by Barnes et al., 2017), but there are several types that have been not been described previously (e.g., pinto net) or are much less common (inclusion net). Each of the textures in the deposit described below, based on core logging, petrography, and sulfur contents, is shown schematically in Figure 3, and representative samples are illustrated in Figures 4 to 9 (Zuccarelli et al., 2020).

Disseminated sulfides: Disseminated sulfide (Fig. 4) containing ~<5% S and ~<13 wt % sulfide (up to 20 m thick and 125 m strike length) constitutes approximately 5% of the known mineralization in the deposit. It is distributed heterogeneously across the deposit, most commonly between underlying net-textured sulfides and overlying barren host rocks. In this facies pyrrhotite-pentlandite-chalcopyrite ± magnetite ± chromite occur as very small to small (<0.5–40 mm), isolated, irregularly to uniformly dispersed disseminations in the interstitial spaces between olivine and pyroxene in peridotite and pyroxenite (App. Fig. A3A, B).

Disseminated sulfides occur as lightly disseminated, medium disseminated, heavily disseminated, patchy disseminated, and blebby disseminated subfacies. Lightly disseminated sulfide occurs as 0.5 to 5 wt % sulfide distributed in very fine (0.2–0.5 mm) grains. Medium disseminated sulfide occurs as 5 to 10 wt % sulfide in small to medium (0.5–10 mm) irregular wisps. Heavily disseminated sulfide occurs as 10 to 13 wt % sulfide in 10- to 30-mm irregular wisps. Patchy disseminated sulfide (8–10 wt % sulfide) occurs as very small to small (<0.5– 30 mm), isolated, unevenly to uniformly dispersed irregular patches. Blebby disseminated sulfide (2–5 wt % sulfide) occurs as small to medium (20–40 mm) rounded to subrounded blebs in the interstitial spaces between olivine and pyroxene in peridotite and pyroxenite. The sulfides within coarser blebs are occasionally fractionated with chalcopyrite-rich parts and pyrrhotite-pentlandite–rich parts, whereas finer blebs are rarely systematically fractionated. Some disseminated ores contain characteristic chromite clasts.

A typical example of 8% patchy disseminated sulfide is shown in Figure 4A. In the scanned optical image, sulfides are present in wisps and small blebs. In the S-Fe-Ca μXRF map (Fig. 4B), sulfide phases (yellows and oranges) are more abundant in olivine-rich domains (dark green), than in pyroxene domains. Clinopyroxene-rich domains contain ellipsoidal peridotite (dark and medium green) and chromitite (bright green) clasts. Some of the latter are embayed. In the Cr-Fe-Ca μXRF map (Fig. 4C), it is possible to identify two types of chromitite clasts: (1) net-textured chromitite where chromite occurs as clusters of single crystals (red to yellow) within olivine-rich domains (dark to dark green) and (2) rounded ellipsoidal massive chromitite clasts exhibiting varying degrees of alteration to ferrichromite (i.e., red = Cr-rich core, yellow = Cr-poor rim). These textures are more complex than the more uniformly disseminated sulfides in most Ni-Cu-(PGE) deposits (see Barnes et al., 2017).

Net-textured sulfides: Net-textured sulfides (Figs. 5–8) containing approximately 5 to 13% S and 13 to 33 wt % sulfide (up to 40 m thick and 130 m strike length) constitute ~80% of known mineralization along the entire length of the deposit. They consist of 70 to 85% tightly packed cumulate olivine and pyroxene with 13 to 33% interstitial sulfides forming ~0.2-mm-thick films (App. Figs. A3E, F, A4E, F) between olivine and ~0.5- to 1-mm-wide patches (App. Figs. A3C-H, A4A-F) between olivine ± pyroxene. All net textures occasionally exhibit 0.5- to 5-mm-thick, parallel foliations of sulfide that crosscut the sulfides and silicates at 50° to 80° to the core axis and ~60° to a weak igneous lamination defined by the olivine aggregates. An example of this is shown in Figure 5, which was drilled subparallel to the basal contact of the intrusion. Because the sawn surface is not necessarily perpendicular to the foliation (nonoriented core), the sulfide foliation may be perpendicular to oblique to the original basal contact of the intrusion, but the igneous lamination cannot be any less oblique (corresponding to ~60° to 90° to original basal contact).

Leopard net-textured sulfides: Leopard net-textured sulfide (Fig. 5) is the most common subfacies of net texture. It comprises ~50% of all net-textured sulfides and contains ~10 to 13% S and ~25 to 33 wt % sulfide. It consists of ~65–85% cumulus olivine and lesser intercumulus altered clino/orthopyroxene and ~25 to 33% interstitial sulfide forming thin films between olivine (App. Fig. A3C-F) and triangular-shaped patches between olivine ± altered clino/orthopyroxene. Olivine and clino/orthopyroxene are bimodal, ranging in size from 1–3 to 5–15 mm. The coarser crystals and lesser crystal aggregates are what define this texture as leopard net, as the large black pseudomorphs of serpentine-magnetite after olivine among the network of yellowish sulfides gives the appearance of a leopard coat. The sulfide assemblage is composed of pyrrhotite > pentlandite > chalcopyrite.

Abundant magnetite occurs as irregularly distributed grains (>0.5 mm) throughout both sulfide and silicates. Sulfide blebs rarely occur within magnetite crystals but exhibit internal fractionation (chalcopyrite at the inferred top, pyrrhotite-pentlandite at the inferred base).

A typical example of leopard net-textured sulfide is shown in Figure 5. In the scanned optical image, sulfides are evenly distributed except within serpentinized olivine mesocrysts (dark black) and fine-grained aggregates (dark grayish-black phases) (Fig. 5A). In the Ni-Cu-S map (Fig. 5B), abundant chalcopyrite (green) forms an anastomosing network around very fine and very coarse olivine (black), most of the pentlandite (magenta) occurs more sporadically, and most of the pyrrhotite (blue) defines a foliation oriented ~60° to the core axis and ~60° to a weak igneous lamination defined by the olivine aggregates. In the Cr-Fe-Ca map (Fig. 5C), coarse serpentinized olivine (black) is oikocrystic, containing fine chromite (red) and ferrichromite (yellow) chadacrysts or xenoliths, comprising clasts of chromite-bearing dunite. The matrix is fine disseminated clinopyroxene (blue) and Fe-bearing sulfides (green).

Inclusion net-textured sulfides: Inclusion net-textured sulfide (Fig. 6) is an uncommon variety of net-textured sulfide (<5% of total net-textured sulfide) and contains ~4 to 11% S and ~10 to 28 wt % sulfide. It has all the same features of leopard net texture with the exception of a wider variety of clinopyroxene, ranging from 10 to 15% of silicates, and the presence of up to 300-mm-long silicate inclusions. The inclusions are typically dunite-peridotite in composition (likely anteliths) and vary from subrounded to rounded, differing from the lesser olivine aggregates (i.e., leopard spots) in leopard net texture due to their highly irregular distribution. The contacts between inclusions and net-textured sulfide are irregular, due to the infiltration of sulfides between olivine-pyroxene grains in the clasts themselves. One particularly large (30 cm) inclusion in inclusion net texture is exceptionally angular (App. Fig. A5A) but exhibits the same irregular and diffuse edge as other clasts within inclusion net texture (App. Fig. A3G, H).

In the optical image (Fig. 6A), sulfide is not evenly distributed and is controlled mainly by the distribution of inclusions and to lesser extent olivine. In the Ni-Cu-S map (Fig. 6B), pentlandite (magenta), pyrrhotite (blue), and lesser chalcopyrite (cyan or green) occur both in large domains with silicates (reddish black in this image) and in fine pyrrhotite-pentlandite–rich filaments oriented ~90° to the core axis (subparallel to the dike contact). Chalcopyrite, silicates, and lesser pyrrhotite-pentlandite dominate the other domains. In the Cr-Fe-Ca map (Fig. 6C), it is clear that the silicate phases include coarse-grained olivine (black) containing small inclusions of chromite (red), orthopyroxene (opalescent), medium-grained clinopyroxene (blue), and fine-grained olivine (black) within sulfide (green). The 1-cm inclusion in this sample is a serpentinized peridotite composed of serpentine (black) with magnetite (green) veinlets and fine chromite inclusions (red).

Pinto net-textured sulfides: Pinto net-textured sulfide (Fig. 7) is an uncommon variety of net-textured sulfide. It makes up <1% total net-textured sulfide and contains ~4 to 12% S and ~10 to 30 wt % sulfide. Pinto net-textured sulfide most commonly occurs as 1- to 2-m zones among areas of disrupted net texture or where leopard net texture meets pyroxene regions. It is similar to leopard net in that it consists of small serpentinized olivine surrounded by interstitial sulfides but differs with large (0.3–10 mm) subhedral-euhedral oikocrysts of talc-carbonate–altered pyroxene (appearing whitish gray in core) with distinctive optical properties indicating original orthopyroxene composition. The oikocrysts of orthopyroxene are what define this texture as pinto net, as the large gray-white oikocrysts among the network of black serpentinized olivine (App. Fig. A4A, B) and yellowish sulfides are similar to the appearance of white spots on the otherwise dark coat of a pinto-textured horse. The oikocrysts contain both fresh olivine and serpentinized olivine. The interstitial sulfides consist of pyrrhotite, pentlandite, and chalcopyrite up to 1 mm in diameter. Magnetite exists in anhedral and infrequent euhedral blebs up to 0.5 mm in diameter and sometimes contains inclusions of sulfide.

A typical example of pinto net-textured sulfide facies is shown in Figure 7. In the optical image (Fig. 7A) the amount of sulfide is relatively small (~20–25%) and occurs almost exclusively in close association with small serpentinized olivine crystals (black) interstitial to coarse sulfide-free orthopyroxene oikocrysts (gray) (Fig. 7A). In the Ni-Cu-S map (Fig. 7B), pyrrhotite (blue), pentlandite (magenta), and chalcopyrite (green) are more or less evenly disseminated throughout the sample except within the large weakly aligned pyroxene domains (black). Nevertheless, chalcopyrite is more common on peripheries of the pyroxene oikocrysts. In the Cr-Fe-Ca map (Fig. 7C), the matrix is dominated by sulfides (green) and fine-grained olivine (black), and the pyroxene-rich domains include parts comprising mainly orthopyroxene (black), chromite (red), and ferrichromite (orange and yellow), as well as parts comprising mainly clinopyroxene (blue).

Disrupted net-textured sulfides: Disrupted net-textured sulfide (Fig. 8) is a localized but common form of net-textured sulfide (~30% of net texture) containing ~6 to 13% S and 15 to 32 wt % sulfide. It appears as localized <20-m zones predominantly between 500- to 900-m depth in the deposit, although it is not restricted to that area. Disrupted net texture appears as highly irregular serpentinized olivine cumulate patches surrounded by interstitial sulfide, directly adjacent to sulfide-poor regions of talc-carbonate–altered pyroxenite. In drill hole NOT-09-053, a large, 50-cm-zone of pyroxenite cuts through disrupted net, with a very similar style of contact with olivine cumulates on its edges (App. Fig. A6). The first disrupted net-texture section (857.1–857.7 m) contains olivine cumulate and interstitial sulfide with highly irregular 2- to 10-cm zones of pyroxenite throughout. The central pyroxenite (857.7–858.2 m) contains minimal mineralization and relict olivine, with a gradual transition back to leopard net texture (858.2–858.4 m). In disrupted net-textured sulfide, the most abundant sulfide mineral is pyrrhotite, with varying amounts of pentlandite (0.1–0.5 mm) and chalcopyrite (0.1–3 mm). Magnetite occurs sporadically in both barren and mineralized areas, occurring up to 0.5 mm in diameter and ranging from anhedral to euhedral. Minor chromite is also present.

The contacts between the cumulate and pyroxenite zones are very irregular and wispy, with some serpentinized olivine crystals/chains isolated from their respective cumulates and seen floating in the pyroxenite (Fig. 8A). The barren pyroxenite zones contain frequent fresh olivine, which can be up to 4 mm in diameter, but almost all olivine in the cumulates containing interstitial sulfide is serpentinized (App. Fig. A4C, D).

A typical example of disrupted net texture containing ~25 wt % sulfides is shown in Figure 8. The optical image (Fig. 8A) indicates that sulfide is much more abundant in leopard net-textured domains containing abundant finegrained serpentinized olivine (black) and much less abundant in areas containing abundant pyroxene (gray) of the pyroxenite. In the Ni-Cu-S map (Fig. 8B), there is a zonation away from the pyroxenite from (1) pyrrhotite (blue) ≈ pentlandite (magenta) >> chalcopyrite (green) through (2) chalcopyrite >> pyrrhotite ≈ pentlandite to (3) pyrrhotite ≈ pentlandite > chalcopyrite. In the Cr-Fe-Ca map (Fig. 8C), the distinction between the pyroxenitic material (blue), which contain patches of residual olivine (black), magnetite (yellow), and chromite (red), and the surrounding net-textured sulfides (green; all sulfide facies) is more distinct.

Patchy net-textured sulfides: Patchy net-textured sulfide (App. Fig. A4E, F) is a moderately common sulfide texture (~15% of all net texture), which occurs adjacent to leopard net-textured and disrupted net-textured sulfides and contains ~4 to 10% S and ~10 to 25 wt % sulfide. It appears as peridotite with 50–60% 1- to 3-mm-diameter olivine and 40 to 50% interstitial altered pyroxene, with isolated patches of sulfide that do not visually connect on the two-dimensional surface of the half core. Patchy net-textured sulfides differ from leopard net texture because all olivine crystals are uniform in size, and interstitial sulfide occurs irregularly (see Fig. 3). Patchy net texture differs from disrupted net because no patches of barren pyroxenite occur. Sulfides are predominantly pyrrhotite and contain varying amounts of pentlandite (0.1–1 mm) and chalcopyrite (0.1–4 mm). Magnetite can be commonly found between some silicate minerals but also as indiscriminate euhedral-subhedral minerals up to 1 mm in diameter.

Semimassive to massive sulfides: Semimassive sulfide, defined as massive sulfide containing silicate/oxide inclusions (Fig. 9, left side of the photo) is very uncommon in the Eagle’s Nest deposit (~6% mineralization) but contains ~20 to 31% S and ~50 to 80 wt % sulfide. It occurs in both the lower and upper parts of the deposit, although most commonly along the footwall. Semimassive can generally be described as massive sulfide containing ~10 to 20% silicate inclusions. The nonsulfide component is normally composed of gabbro or tonalite xenoliths derived from the country rocks and peridotite anteliths or autoliths (also known as “cognate xenoliths” derived from this or other phases of the intrusive system). Although much less common, sulfide-bearing inclusions also occur within this facies and are normally composed of leopard net-textured sulfides. Semimassive sulfides contain no olivine and pyroxene crystals, except those within ultramafic inclusions. In thin section, the sulfide portion of semimassive sulfides appears similar to massive sulfide, although with more chalcopyrite.

Massive sulfide: Massive sulfide (Fig. 9, right side of the photo) is highly localized within Eagle’s Nest and contains ~>32% S and ~>80 wt % sulfide (up to 20 m thick and 40 m strike length). It occurs predominantly within two ~100- × 50-m pockets at 100- and 450-m depths. In all other areas it occurs as 0.5- to 10-m lenses or as crosscutting veinlets (30– 50 mm thickness). Massive sulfide is yellow-brass in color, with 50 to 75% pyrrhotite, 20 to 25% pentlandite (2–20 mm), and 0.5 to 9% chalcopyrite (5–100 mm). Pentlandite often occurs as distinctly reflective, roughly circular crystals, and chalcopyrite occurs as irregularly distributed grains throughout and as concentrations along contacts with host rocks, in fractures, or in secondary veining.

In thin section, pyrrhotite contains fractures that often host anhedral pentlandite and chalcopyrite, and <0.5-mm flames of pentlandite occur perpendicular to these fractures. Pyrrhotite is pinkish brown in color and contains euhedral-subhedral magnetite ranging from 0.5 to 3 mm. In thin section, pentlandite occurs as large pale-yellow eyes that range from 1 to 20 mm in diameter and show higher relief than pyrrhotite and can exist with multiple minerals forming clusters or as smaller eyes distributed regularly throughout the sample (right side of Fig. 9A; see semimassive sulfide section above). Chalcopyrite occurs as irregular patches throughout massive sulfide 1–5 cm in diameter. All massive sulfides also contain irregularly distributed magnetite (0.5–5 mm). There are often small (<100 μm) PGMs associated with massive sulfide as well, composed mostly of Pd, Bi, and Te. In some massive sulfide samples, small ductile deformation structures seen predominantly in alignment of chalcopyrite and pentlandite, although this is not common.

A contact between semimassive and massive sulfide facies is shown in Figure 9. The amount of sulfide varies from ~60% in semimassive sulfides (left side of Fig. 9A) to ~95% in massive sulfides (right side of Fig. 9A). The semimassive sulfide domain contains abundant inclusions of altered, recrystallized, and partially melted gabbro. In the Ni-Cu-S map (Fig. 9B), it is clear that the semimassive sulfides are composed of pyrrhotite (blue) >> pentlandite (magenta) ≈ chalcopyrite (cyan) and that the massive sulfides are composed of pyrrhotite > pentlandite ≈ chalcopyrite. Pentlandite and chalcopyrite are finer grained and more dispersed (as patches and along pyrrhotite grain boundaries) in semimassive sulfides and coarser grained and more segregated in massive sulfides. In the Cr-Fe-Ca map (Fig. 9C), pyrrhotite is bright green, and pentlandite and chalcopyrite are both dark green. The sulfides minerals display a combination of coarse granular pentlandite (at contacts with chalcopyrite) and thin (sub-mm) loops of exsolved pentlandite (commonly in semimassive sulfides around pyrrhotite grain boundaries). The inclusions contain a Ca-rich phase (dark blue)—likely altered clinopyroxene, based on the habit—and a Cr-Fe-Ca–poor phase (black)—likely altered (albitic) plagioclase. Ferrichromite (orange) occurs as isolated crystals and as rims on inclusions.

A contact between semimassive/massive sulfide vein and leopard net-textured sulfide facies is shown in Appendix Figure A7. The amount of sulfide in the semimassive vein is ~60% (central part of App. Fig. A7A). The semimassive sulfide domain contains abundant inclusions of a Ca-rich phase. In the Ni-Cu-S map (App. Fig. A7B), it is clear that the semimassive sulfides are composed of pyrrhotite (blue) >> pentlandite (magenta) ≈ chalcopyrite (cyan) and that there is abundant chalcopyrite (green) along the contact between semimassive and leopard net-textured sulfide. Pentlandite and chalcopyrite are in large patches in semimassive sulfides and more segregated in massive sulfides. In the Cr-Fe-Ca map (App. Fig. A7C), pyrrhotite is bright green, and pentlandite and chalcopyrite are both dark green. The Ca-rich phases stand out from the Fe-rich (green) sulfides. The sulfides minerals display a combination of coarse granular pentlandite (at contacts with chalcopyrite) and thin (sub-mm) loops of exsolved pentlandite (commonly in semimassive sulfides around pyrrhotite grain boundaries).

Contacts between sulfide textures: The vast majority of contacts between sulfide textures are gradational over >30 cm, with three exceptions. The first is within disrupted net texture, where larger zones of apparently crosscutting pyroxenite often exhibit irregular but abrupt contacts with net-textured sulfide (Fig. 8; App. Fig. A6). The second is between massive sulfide and silicate lithologies (App. Fig. A5B-D), which are extremely sharp and sometimes bordered by skeletal chromites/ferrichromites. The third is between contacts of massive sulfide/crosscutting massive sulfide veins with all lithologies (App. Fig. A7). These veins typically cause chalcopyrite to bleed into existing interstitial sulfide within net-textured sulfide, significantly increasing the sulfide/olivine ratios proximal to the vein contact and similar to soft-walled veins described in Barnes et al. (2017).

Ranges and averages of chemical compositions of sulfide minerals are given in Table 2. All mineral compositions are available in Zuccarelli et al. (in press).

Pyrrhotite (Fe_1−xS) is the most common sulfide mineral in the Eagle’s Nest deposit. In core, it appears light brownish pink and is magnetic. Pyrrhotite is the most abundant sulfide in the interstitial mineralization between silicate minerals in all net-textured sulfide ores, and most blebs within disseminated sulfide are also pyrrhotite.

Pentlandite (Fe,Ni)₉S₈ is the main nickel-bearing sulfide mineral in the Eagle’s Nest deposit. In core, it appears bright brown-beige, noticeably more yellow than surrounding pyrrhotite. It exists as either coarse anhedral eyes (5–20 mm), small anhedral crystals within pyrrhotite (0.1–0.5 mm), or exsolution flames in fractures within pyrrhotite (<0.1 mm diam). Small anhedral crystals of pentlandite occur in all the net-textured sulfides, with some occurring in disseminated sulfides as well, and in some of the coarser eyes there is clear alteration of the pentlandite in the form of Co-rich fractures and stringers.

Chalcopyrite (CuFeS₂) is the major copper sulfide mineral in the Eagle’s Nest deposit and exists as large brassy yellow patches (>60 mm) within massive sulfide ore. It also occurs as smaller crystals (0.5–1 mm) within pentlandite and as remobilized veinlets in net-textured sulfides. Chalcopyrite is common adjacent to contacts between host rock and deposit rock, and where massive sulfide contacts other sulfide textures.

Most of the PGMs in the Eagle’s Nest deposit appear to be merenskyite (Pd,Pt)(Te,Bi)₂ and michenerite (PdBiTe), occurring as small crystals (<20 μm) within massive sulfide and net-textured sulfide (App. Fig. A8). These are relatively common, with at least one to two grains visible in each sample, although several uncommon examples of PGMs in massive sulfides are up to ~110 μm in diameter. Other minor minerals include electrum (Ag-Au) in some altered samples of net texture, occurring as small <10-μm grains among patches of amphibole alteration. Most PGMs are slightly zoned, but one sample of massive sulfide contained an ~40-μm concentrically zoned euhedral mineral with a Pt-bearing core, a Pd-bearing inner zone, and an As-bearing outer rim.

Whole-rock geochemical compositions of the host rocks and ores are available in Zuccarelli et al. (in press) and plotted in Figures 10–14. Where indicated, data from this study are presented in comparison with Noront assay data to provide an accurate overview of the deposit.

Nonmineralized peridotites and pyroxenites in the Eagle’s Nest dike vary between ~22–38% and ~21–34% MgO (Fig. 10), respectively, and plot along the same trends as barren host rocks at Black Thor (Carson et al., 2015). For other elements, Si-Ti-Ca increases with decreasing Mg except for samples containing significant orthopyroxene (which plot at lower Si-Al-Ca; Fig. 10A-D; Si and Ti are not shown). The same trends are observed for the silicate components of disseminated and net-textured mineralization. Cr-Fe (Fig. 10B, C) broadly decreases with decreasing Mg, and analyzed olivine plots along the trend of Fo_82–86, whereas most samples plot along the cotectic olivine-chromite cumulate trend, but some plot below it. Importantly, most of the trends from mineralized rocks project to olivine compositions and equilibrium liquids that are more magnesian than any of the analyzed olivines (Fig. 10C).

All rocks in the dike are enriched in highly incompatible lithophile elements (HILEs; Cs-U-Th-light rare earth elements) relative to Nb-Ta-Ti and moderately incompatible lithophile elements (MILE; Zr-Hf-middle rare earth elements-Y-heavy rare earth elements) (App. Fig. A9A). Two pyroxenite samples are significantly enriched in Li-Rb-Ba.

In terms of Nb/Tb versus Th/Yb (App. Fig. A9B, C), the data range between 0.1–1 Th/Yb and 0.9–2 Nb/Yb, extending between normal mid-ocean ridge basalt (N-MORB) and enriched mid-ocean ridge basalt (E-MORB) toward Archean upper continental crust. The enrichment in Th (and other HILEs) suggests up to 30 to 40% crustal contamination.

Metals versus S: Ni, Cu, and Fe increase with increasing S content (Fig. 11A-C), indicating they are housed within sulfides. Palladium-platinum increases only weakly with increasing S, indicating that it is housed in—or closely associated with—phases other than sulfides and/or that it has been mobile during metamorphism (Fig. 11D, E).

Elemental variations: The abundances of metals in the different ore facies in the Noront assay database and in the ore subfacies analyzed in this study are presented as box plots (Tukey plot mode in ioGAS v. 7.0) in Figure 12.

Sulfur: The Noront assays (Fig. 12A) and breakdown of net-textured sulfides (Fig. 12B) have already been described above (App. Fig. A1); the box plots in Figure 12 provide details on outliers as well as statistical details of wt % S for each texture.

Nickel: Massive sulfides contain ~5 to 10% Ni, semimassive sulfides contain ~0.5 to 8% Ni, net-textured sulfides contain ~0.5 to 3% Ni, and disseminated sulfides contain ~<1% Ni (Fig. 12C). In general, leopard net texture contains more Ni (~1.5–3%) than other net textures (~0.5–2.5%), but overall net textures cannot be distinguished by Ni content (Fig. 12D).

Copper: Massive sulfides contain ~0.1 to 10% Cu, semimassive sulfides contain ~0.1 to 12%, net-textured sulfides contain between ~0.1 and ~2.5%, and disseminated sulfides contain <0.5% (Fig. 12E). Inclusion net-textured sulfides contain the most Cu by far (~0.25–4.75%), which correlates with the preference of chalcopyrite to form on the edges of clasts. All other net textures contain ~0.1 to 1.75% Cu (Fig. 12F).

Iron: There is a gradational overlap in wt % Fe (Fig. 12G) from disseminated sulfide (~2.5–12.5%) and net-textured sulfides (~10–23%), net-textured sulfides to semimassive sulfides (~20–45%), and massive sulfide (~27–50%). All net textures strongly overlap with Fe %, which means this cannot be used to distinguish them from each other (Fig. 12H).

Calcium: Disseminated sulfide by far has the widest range (~0.1–7% Ca), whereas massive sulfide has the narrowest range (0.1–0.5% Ca) (Fig. 12I). Leopard net-textured sulfide (~0.1–0.2% Ca) contains significantly less Ca than all other net textures (~0.2–1.9% Ca), allowing it to be distinguished from other textures on the basis of chemistry (Fig. 12J).

Palladium and platinum: Massive sulfide contains the widest range of Pd (~0.1–23 ppm), whereas disseminated sulfide contains the least (~0.1–2 ppm) (Fig. 12K). There are no significant variations in Pd content among the net-textured sulfide subfacies, although leopard- net texture has the highest consistent Pd content (~2.5–5 ppm versus ~0.1–5 ppm for all other textures) (Fig. 12L). This is a general indicator that Pd is associated with sulfide. Platinum contents are not distinguishable between all major textures, with the general range between ~0.1 and 5.25 ppm (Fig. 12M, N).

Metal tenors: Metals were recalculated to 100% sulfide to compare samples containing different abundances of sulfides and to establish metal modal mineral abundances. This was done using the method of Naldrett (1981) in which Cu is allocated to stoichiometric chalcopyrite (34.62% Cu, 30.43% Fe, 34.94% S), Ni is allocated to pentlandite (34.15% Ni, 31.06% Fe, 33.72% S, equivalent to Fe₄Ni₅S₈) after correction for 0.2% Ni in olivine, and the remaining S is allocated to pyrrhotite (61.05% Fe, 38.94% S, equivalent to Fe_0.9S). These values are based on analyzed mineral compositions from this deposit (Table 2)

Ni₁₀₀ values are mainly 6 to 10% (Fig. 13), ranging to higher and lower values for some disseminated and inclusion net samples. Within the subfacies of net textures, Ni₁₀₀ values are relatively consistent (~6–8.5%) with the exception of inclusion net-textured sulfide (~3–7.5%) (Fig. 13). Based on Noront Resources assays, most samples range from ~0.1 to 12% Cu₁₀₀ (Fig. 13) with uncommon semimassive and some inclusion-net sulfides containing higher Cu₁₀₀ values. Within the subfacies of net-textured sulfides, leopard net texture (~1–6%), pinto net texture (~0.5–3%), and disrupted net texture (~1–4%) have similar Cu₁₀₀ compositions; inclusion net texture (~1–17%) and patchy net texture (~2–9%) have far broader ranges (Fig. 13).

Based on these tenors, massive sulfide contains ~12.5 to 28% pentlandite and ~0.1 to 28% chalcopyrite. Semimassive sulfides contain ~2 to 25% pentlandite and ~0.1 to 33% chalcopyrite, net-textured sulfides contain ~2.5 to 8% pentlandite and ~0.1 to 8% chalcopyrite, and disseminated sulfides contain ~0.1 to 2.5% pentlandite and ~<2% chalcopyrite. When breaking down net-textured sulfides into the five subfacies, leopard net contains ~4 to 8% and other net textures contain ~1.5 to 7% pentlandite. Inclusion net texture by far contains the largest range of chalcopyrite (~1–14%); all other textures contain ~0.5 to 5%.

Ni₁₀₀ decreases with increasing Cu₁₀₀ (Fig. 13A, B) in the Noront data set and the inclusion net-textured samples in this study. Pt₁₀₀ and Pd₁₀₀ correlate poorly with Cu₁₀₀ (Fig. 13C, D). The calculated R factor trends from Mungall et al. (2010) are plotted for reference. Where the textures of the samples are known, there are no systematic differences between the compositions of different sulfide textural facies.

Mantle-normalized metal variations: The Ni-Cu-(PGE) mantle-normalized metal abundances of average Eagle’s Nest ore types and the ranges for Kambalda and Raglan ores are given in Figure 14. The amount of each element in each major ore type (massive, semimassive, net-textured, disseminated) is based on the median from Noront Resources assays, divided by the amount of each element in primitive mantle (Taylor and McLennan, 1985). In a general sense and with the exception of disseminated sulfides, Eagle’s Nest ores follow the same enrichment-depletion pattern as Kambalda and Raglan ores. Eagle’s Nest ores (massive, semimassive, net-textured sulfides) are consistently slightly more enriched than Kambalda and Raglan ores in Pd, Cu, Pt, and Rh but similar to Raglan and Kambalda in terms of Ru, Os, and Co.

Sulfur isotopes: Eagle’s Nest ores range narrowly between –0.3 and 1.2‰ δ³⁴S (Table 3). This is within the –3 to 2‰ range of sulfides in the Black Thor intrusion (N. Farhangi, pers. commun., 2019), within the –2 to –0.5‰ range of Black Label sulfides, and within the range of some Blue Jay data but far outside the range of Blue Jay outliers (–15 to 0.5‰). The Eagle’s Nest values are close to but slightly heavier than the 0.1 ± 0.5‰ range for MORB (Sakai et al., 1984), but also within the range of sulfidic iron formation in the footwall of the Black Thor intrusion (–1 to 1.5‰; H.J.E. Carson, pers. commun., 2019). There are no significant differences between different sulfide textures, suggesting that all S in the Eagle’s Nest deposit had a similar source.

Although the Eagle’s Nest dike is presently a subvertical flattened pipe (Fig. 2), the following features suggest that it was originally emplaced as a subhorizontal blade-shaped dike that was subsequently rotated ~90° to the southeast:

1. All of the contacts and layering in the Esker intrusive complex, including the Black Thor and Double Eagle intrusions and all of the stratiform chromitite horizons, dip subvertically or steeply to the northwest and exhibit textures and geochemical variations consistent with them younging to the east-southeast but emplaced subhorizontally and, therefore, rotated into the present position (Mungall et al., 2010; M.G. Tuchscherer, unpub. report, 2010; Carson et al., 2015; Houlé et al., 2020; Metsaranta and Houlé, 2020).

2. The SE-grading massive\net\disseminated sulfide segregation profile in the upper embayment of the Eagle’s Nest deposit has been attributed to gravitational segregation in all other deposits of this type (e.g., Naldrett, 1966; Ewers and Hudson, 1972; Usselman et al., 1979; see reviews by Lesher, 1989; Naldrett, 2004; Barnes, S.-J., and Light-foot, 2005; Barnes, S.J., et al., 2017), also consistent with the mineralization younging to the east-southeast but emplaced subhorizontally and, therefore, rotated into the present position.

3. The southeast gradation from peridotite (olivine-pyroxene cumulates) to pyroxenite (pyroxene cumulates) in the Eagle’s Nest dike is not a diagnostic indicator of younging to the southeast, as some intrusions exhibit reverse gradations (e.g., Perseverance: Barnes et al., 1995), but the consistency of this gradation along the exposed length of the dike is consistent with the intrusion having been originally emplaced subhorizontally (Zuccarelli et al., 2018).

Rotation of the Esker intrusive complex and the Eagle’s Nest dike 90° back to the northwest would restore all of the igneous layering to a gravitationally stable subhorizontal orientation and leave the Eagle’s Nest dike ~250 m beneath the Double Eagle intrusion (Laudadio, 2019; Zuccarelli et al., 2019). The geometry would then be that of a subhorizontal blade-shaped dike (Mungall et al., 2010) with mineralization along the keel. Asymmetrically differentiated mineralized blade-shaped dikes have also been identified in the Expo intrusive suite in the Cape Smith belt (Mungall, 2007), at Savannah in Western Australia (Barnes and Mungall, 2018), and in several localities in China (Lu et al., 2019) (see review by Lesher, 2019). Blade-shaped dikes appear to form only in areas where density contrasts in the country rocks limit upward ascent and where a radial stress field favors linear rather than planar magma emplacement (e.g., Rubin and Pollard, 1987; Bolchover and Lister, 1999).

The irregular lower contact of the dike may represent the original morphology of the intrusion, but it may also have been enhanced by preferential thermomechanical erosion below the high-density, low-viscosity molten sulfides (e.g., Groves et al., 1986; Lesher, 1989; Williams et al., 1998). Barnes and Mungall (2018) suggest a dynamic mechanism whereby sulfide liquid derived from assimilation of country rock at higher levels accumulates gravitationally along the lower edge of the blade and partially drives the downward propagation of the dike during its emplacement.

Connection to other deposits in Esker intrusive complex: The lithologies, mineralogy, mineral chemistry, and geochemistry of the Eagle’s Nest dike are similar to those of the Double Eagle intrusion (Azar, 2010; Mungall et al., 2010) and the Black Thor intrusion (Carson et al., 2015), suggesting that they are petrogenetically related. However, the Eagle’s Nest dike contains predominantly peridotite, pyroxenite, and sulfides with sparse chromitite inclusions, whereas the Double Eagle and Black Thor intrusions contain dunite, peridotite, gabbros, and chromitites with negligible (Double Eagle) or minor (Black Thor) sulfides (Farhangi et al., 2013).

The paucity of adcumulate dunite and chromite mineralization (other than anteliths) in the Eagle’s Nest dike and the relatively small amounts of Fe-Ni-Cu sulfides in the Double Eagle intrusion suggest that the Eagle’s Nest dike may not be the feeder to the Double Eagle intrusion (Houlé et al., 2020; see discussion by Laudadio et al., in press). If so, this suggests that the Double Eagle intrusion was fed from an unexposed or eroded part of the plumbing system and that the Eagle’s Nest dike fed another unexposed or eroded intrusion. More drilling is needed to test these possibilities, but geologic, geochemical, and petrological inconsistencies between the compositions of putative feeder dikes and sills and overlying volcanic rocks have also been observed at Thompson (Lesher et al., 2001), Norilsk (Latypov, 2002), Raglan (Lesher, 2007; McKevitt et al., 2019), and other areas (see review by Lesher, 2019), highlighting the complex nature of magmatic plumbing systems.

Parental magma and contamination: The parental magma for Eagle’s Nest and other parts of the Esker intrusive complex has previously been estimated to have contained ~22% MgO and ~12% FeO_t (Azar, 2010; Mungall et al., 2010; Carson et al., 2015), which would have been in equilibrium with ~Fo₉₀ olivine. However, Laarman (2014) reported olivine up to Fo₉₄ in the Black Label zone of the associated Black Thor intrusion, Carson (pers. commun., 2019) has noted that the compositions of large numbers of Black Thor dunites/peridotites require that they contained (prior to serpentinization) olivine up to Fo₉₄, and many of the higher-Mg, lower-Fe peridotites (barren and recalculated sulfide-free) analyzed in this study also appear to have contained olivine with a composition up to Fo₉₄ (see Fig. 10).

Importantly, a small but significant number of Eagle’s Nest rocks have Cr contents that are consistent with accumulation of olivine and less than cotectic proportions of chromite (Fig. 10B, C), also consistent with a more magnesian liquid (Murck and Campbell, 1986; Lesher and Stone, 1996). Together, this suggests that the parental magma contained ~28% MgO and ~10% FeO_t, making it a high-Mg komatiite (H.J.E. Carson, pers. commun., 2019). The maximum Fo_82–86 analyzed in this study and the maximum Fo_82–86 analyzed by Mungall et al. (2010) can be attributed to the most magnesian olivine not being analyzed (very few are preserved), fractional crystallization ± crustal contamination, and/or reequilibration with trapped liquid (see Barnes, 1986; Cawthorn and Barry, 1992; and discussion by Mungall et al., 2010).

The enrichments in HILEs relative to MILEs (App. Fig. A9) with negative Nb-Ta-(Ti) anomalies and the Th/Yb versus Nb/Yb diagram are consistent with up to ~30 to 40% crustal contamination of a mantle-derived magma (App. Fig. A9B, C), significantly more than the 13% inferred by Mungall et al. (2010). Because all rocks are contaminated, the majority of the contamination most likely occurred below/upstream rather than within the current conduit (see discussion by Lesher and Arndt, 1995; Lesher et al., 2001).

Minerals and alteration: Like most other komatiite-associated Ni-Cu-(PGE) deposits (see reviews by Lesher, 1989; Lesher and Keays, 2002; Barnes, 2006; Arndt et al., 2008; Lesher and Barnes, 2009), the Eagle’s Nest dike contains mainly olivine-rich cumulate rocks. In many of these deposits, olivine exhibits highly elongate, embayed, or branching crescumulate textures, and in some deposits relict igneous olivine compositions vary systematically with stratigraphic height (e.g., Kambalda: Lesher, 1989; Raglan: Lesher, 2007), indicating that olivine crystallized in situ rather than representing transported phenocrysts. This requires the host units to represent dynamic conduits in which olivine crystallized and accumulated (Lesher et al., 1984; Lesher, 1989; see review by Arndt et al., 2008).

The scatter of Cs-Rb-K-Na and Ba-Sr-Ca (App. Fig. A9A) in most peridotites and many pyroxenites suggests that they have been mobile during serpentinization. In contrast, the well-defined negative correlations between Mg and Ti-Al-Si (Fig. 10A) suggest that they have been relatively immobile, which is consistent with evaluations of alteration in other greenschist facies komatiitic rocks (see reviews by Lesher and Stone, 1996; Arndt et al., 2008). The presence of multiple cumulate phases (olivine-orthopyroxene, olivine-clinopyroxene, olivine-sulfide) and variable amounts of trapped silicate liquid hamper evaluations of the mobility of Si-Mg-Fe-Mn, but by analogy with studies of other serpentinized but texturally well-preserved komatiitic rocks we can assume that they have been only slightly mobile (e.g., Lesher and Stone, 1996; Arndt et al., 2008).

Late pyroxenite: The late pyroxenite phase that disrupted the net-textured mineralization in the Eagle’s Nest dike is similar to, but more olivine rich than, the barren late websterite phase that invaded the lower part of the Black Thor intrusion, including the Black Label Cr deposit (Spath et al., 2015; Spath, 2017). Establishing the composition of the magma it crystallized from is hampered by the cumulate nature of the pyroxenite, the absence of unaltered orthopyroxene, and the paucity of unaltered clinopyroxene, but the lower abundance of olivine (App. Fig. A2) indicates that it was less magnesian and more silica rich than the magma from which the peridotites in the Eagle’s Nest dike and Double Eagle intrusion crystallized.

The textural relationships between silicate minerals and sulfides at Eagle’s Nest provide insights into the genetic and temporal relationships between silicates and sulfides. Olivine exhibits subhedral crystal forms and could not have crystallized from a sulfide liquid, suggesting that olivine crystallized from silicate melt and was later infiltrated by sulfide melt. There are abundant clasts of olivine cumulate in inclusion net-textured sulfides with sulfide in between and breaking up individual olivine (Fig. 6; App. Fig. A3G, H), suggesting that these formed before sulfides. Sulfides sometimes crosscut olivine, but in most if not all cases these appear to be sulfides mobilized into or replacing magnetite in fractures generated during serpentinization of olivine. There are rare occurrences of pockets of sulfide in olivine crystals; this may be attributed to remobilized sulfide exploiting serpentinizing fractures in olivine or may be a result of sulfide adjacent to olivine being cut so some sulfide is left on the olivine surface, appearing to be inside the olivine.

The abundance of olivine ± sulfide cumulate rocks in the Eagle’s Nest dike requires that (1) olivine and molten sulfide crystallized/exsolved and accumulated together in the dike during emplacement without physical segregation from one another or that (2) one or both were physically transported into and accumulated in the dike independently of one another and subsequently mixed. The former is unlikely, because olivine and sulfide crystallize/exsolve along a cotectic, forming rocks containing ~60× more olivine than sulfide (e.g., Duke, 1986; Barnes, 2007). The latter is more likely, but the relative transportability of less dense olivine and denser sulfide melt droplets depends on their sizes (or effective sizes in the case of slugs or pseudoslugs: Lesher, 2019); the density, viscosity, and velocity of the magma (e.g., Lesher and Groves, 1986; de Bremond d’Ars et al., 2001; Robertson et al., 2015); and the orientation of transport (e.g., Lesher, 2017, 2019). Coarser olivine and finer sulfide can be transported together, but they would be expected to become segregated from one another if not hydrodynamically equivalent.

The presence of a well-defined massive\net\disseminated sulfide segregation profile in the upper embayment and a comparatively irregular distribution of net and massive sulfide in the lower embayment may be explained by Eagle’s Nest being emplaced at a shallow angle. The upper embayment could represent a less dynamic area in front of or behind the topographical high where ore could undergo gravitational segregations, while the lower embayment may represent a more dynamic area further away from the topographical high; however, complex fluid dynamic studies are beyond the scope of this contribution.

The S isotope compositions of Eagle’s Nest ores (Table 3) are consistent with the S being derived from (1) a mantle-derived magma, (2) an unfractionated crustal source such as the iron formation in the country rocks, or (3) some combination of the two. The much smaller variations in the strongly mineralized Eagle’s Nest dike and the much greater variations in the weakly mineralized Black Thor intrusion are not consistent with a control by the relative masses of magma and sulfide (see Lesher and Burnham, 2001; Ripley and Li, 2003), suggesting multiple isotopically different crustal sources.

The amount of sulfide in the Eagle’s Nest dike, which is estimated to be ~11% (weighted bulk sulfide content calculated from S data for mineralized and barren rocks in the Noront database), is 35 times greater than the ~0.3% S that can dissolve in a komatiitic liquid (e.g., Shima and Naldrett, 1975; Li and Ripley, 2009; Fortin et al., 2015; Smythe et al., 2017). Although solubility is reduced by cooling and/or contamination (see review by Naldrett, 2004), those processes would induce crystallization of 60–100 times more olivine than sulfide (Duke, 1986; Barnes, 2007), resulting in disseminated, not net-textured, mineralization. This suggests that Fe-Ni-Cu sulfides at Eagle’s Nest formed by partial melting of an S-rich horizon and that immiscible sulfides were left behind in the dike (as inferred for Thompson: Lesher et al., 2001; Expo-Ungava: Mungall, 2007) or that they were transported into their current location from upstream in the magmatic plumbing system (as inferred for Kambalda: Lesher and Campbell, 1993; Raglan: Lesher, 2007). Given the abundance of sulfide in the Eagle’s Nest dike, the latter process seems more likely, although we cannot discount a contribution by the former process.

Though many models for the genesis of Ni-Cu-(PGE) deposits involve upward transport (e.g., Lightfoot and Evans-Lamswood, 2015; Robertson et al., 2015), Lesher (2017, 2019) has argued that most sulfides do not appear to have been transported vertically and that most appear to have formed at more or less the same stratigraphic level at which they are localized. This is consistent with the inferred subhorizontal orientation of the Eagle’s Nest dike and the presence of some sulfide-bearing lithologies at the same stratigraphic level in the McFaulds Lake greenstone belt.

The distribution of S contents in the Noront assay database indicates that the majority of the samples are subeconomic rocks with <2% S and <5 wt % sulfide and net-textured ores with 5 to 13% S and 25 to 33 wt % sulfide. The maximum amount of sulfide in Eagle’s Nest net-textured mineralization is similar to Jinchuan (up to 35%: deWaal et al., 2004; Tonnelier, 2010) but much less than what is observed at Alexo (up to 50%: Houlé et al., 2011), Kambalda (up to 60%: Ewers and Hudson, 1972), and Raglan (up to 70%: Lesher, 2007). The reason for Eagle’s Nest net-textured sulfides not extending to higher sulfide contents appears to be related to the olivine being packed more tightly than in other deposits.

The greater abundance of massive and semimassive sulfide mineralization at Kambalda and Raglan is consistent with the sulfide having formed early—prior to accumulation of the majority of the olivine in the host units. The lower abundance of massive and semimassive sulfide mineralization at Eagle’s Nest and Jinchuan is consistent with the sulfide melt having been introduced later, allowing it to accumulate with and infiltrate olivine.

The positive correlations between Fe-Ni and S (Fig. 11) confirm that these minerals are housed in sulfides. Copper correlates less well, which is the case in many Ni-Cu-PGE deposits. The weaker correlation between Pd and S is consistent with the majority of the Pd being housed in sulfides or exsolved from phases associated with sulfides, with the remainder in PGMs. The scatter of Pt (also Au and Ag, not shown) is most likely attributed to them being housed in PGMs or alloys and heterogeneously distributed on the scale larger than the analyzed samples. The greater variation of Pd compared to Pt may represent a greater mobility of Pd in magmatic/metamorphic hydrothermal fluids than Pt, which has different solubilities in S-rich and Cl-rich fluids (see review by Hanley, 2005).

The lower Ca content of leopard net texture (~0.1–0.25%) relative to the other net textures (~0.30–2.75%; Fig. 10D) reflects the greater abundance of olivine and lower abundance of clinopyroxene, permitting this texture to be identified from other net textures geochemically. The negative correlation between MgO and both Al₂O₃ and CaO (Fig. 10) is an indicator that where more magnesian-rich (olivine) rocks exist, fewer less primitive (CaO-altered pyroxenes) and non-ultramafic-mafic (Al₂O₃) rocks exist.

The very similar Ni tenors (Fig. 13B) across all textures indicate that all of the sulfide equilibrated with magmas of similar compositions at similar magma/sulfide/olivine ratios (see Lesher and Burnham, 2001). The slight differences in Cu tenors (Fig. 13) indicate that mobilization of chalcopyrite has a moderate impact on tenors. The fact that massive sulfide contains the most pentlandite but similar chalcopyrite to semimassive sulfide is an indicator that pentlandite is significantly less mobile than chalcopyrite and that late-stage veins may be Cu fractionates of massive sulfide. The negative correlation between Ni₁₀₀ and Cu₁₀₀ (Fig. 13) indicates that the more Ni rich the sulfide is, the less Cu it contains.

The Ni-Cu-(PGE) mantle-normalized metal abundances indicate magmatic enrichment and contamination in all presented metals (Fig. 14; Pd-Cu-Pt-Rh-Ru-Ni-Ir-Os-Co), with massive and semimassive sulfide at Eagle’s Nest containing equal to or higher values than Kambalda and Raglan.

Although massive sulfide and disseminated sulfide are present in Eagle’s Nest, the numerous subfacies of net-textured sulfide, particularly those not seen in other deposits, require a more in-depth analysis. There are several existing models for the generation of net-textured sulfides as a whole:

1. The flow segregation model of Hudson (1972) attributes the formation of net texture to dynamic gravitational segregation of massive sulfide melt, olivine + sulfide melt, and olivine + silicate melt ± small sulfide droplets during lava/magma emplacement. It may also apply at Kambalda and other orebodies with simple massive, net, or disseminated ore profiles, and it may also explain some of the more complex ore profiles at Alexo or Raglan as multiple pulses.

2. The billiard ball model of Naldrett (1973) attributes the formation of net texture to static gravitational segregation (from base to top) of dense sulfide melt (analogous to mercury), closely packed olivine with interstitial sulfide melt (analogous to closely packed billiard balls with interstitial mercury), and olivine and interstitial silicate melt with disseminated sulfides trapped between olivine crystals (analogous to billiard balls + water ± mercury). This model was modified by Usselman et al. (1979) to allow for solidification of the lower layer of massive sulfide melt in situations where the overlying column of olivine was too thick to preserve any massive sulfide. It may apply to Kambalda and other ore horizons with simple massive, net, or disseminated ore profiles but cannot explain the ore profiles at Alexo or Raglan where the ore profiles are much more complex (Lesher, 2007; Houlé et al., 2011).

3. The gravity percolation model of Barnes et al. (2017) attributes the formation of net texture to downward migration of denser sulfide melt through a network of closely packed olivine, where the degree of migration is controlled by the size of the sulfide droplets, the size of the pore spaces between olivine crystals, and the ability of sulfide melt to displace the silicate melt (see Mungall and Su, 2005; Chung and Mungall, 2009).

All three models are consistent with the broadly upward segregation of discontinuous massive, net-textured, and disseminated mineralization at Eagle’s Nest, but none explain (1) the bimodal olivine crystals and clasts in leopard net-textured sulfides, (2) the different aggregate and antelith populations in the different sulfide textures (leopard net textured versus inclusion net textured versus disseminated sulfides), or (3) the transgressive nature of disrupted net-textured sulfides.

Leopard net texture: This contains fine-grained subhedral olivine phenocrysts and coarse-grained ellipsoidal serpentinized olivine-(chromite) aggregates. Similar aggregates are present at Sakatti (Brownscombe et al., 2015) and Nova (Barnes et al., 2020), and their origin is not well understood. The difference in the mineralogy of the phenocrysts and aggregates at Eagle’s Nest suggests that the former crystallized from a higher-Mg magma that was saturated only in olivine, whereas the latter crystallized from a lower-Mg magma that was saturated in olivine-chromite (Murck and Campbell, 1986; Barnes, 1998). They could be autoliths that crystallized upstream in the magma conduit and were transported into the dike, but this does not explain their more or less even distribution. Crystallization in situ, like pyroxene oikocrysts in other types of leopard net-textured sulfides (Barnes et al., 2017) better explains their distribution but requires the magma to crystallize olivine-chromite in the aggregates and olivine in the remainder of the network. Regardless of their origin, the absence of sulfides (other than late-stage veinlets formed during serpentinization) inside the phenocrysts or aggregates suggests that the sulfides percolated into the olivine crystal network after they accumulated.

Pinto net-textured sulfide: This texture has not yet—to our knowledge—been described in other deposits. It comprises a network of olivine net-textured sulfide with large talc ± serpentine ± anthophyllite-altered orthopyroxene oikocrysts and appears to have formed from a more siliceous magma (see review of phase equilibria in Arndt et al., 2008). Because this texture is uncommon and is sporadically distributed, it may reflect incomplete homogenization of contaminants (see discussion by Lesher and Arndt, 1995) rather than an influx of a distinct, more contaminated magma pulse.

Inclusion net-textured sulfide: This texture has been described at Nova by Barnes et al. (2020) but appears to be rare to absent in other deposits. It contains sporadic and unevenly distributed coarse, ellipsoidal peridotite and pyroxenite anteliths, which were likely incorporated upstream in the plumbing system. Subsequent sulfide infiltration of the anteliths may have caused some smaller olivine cumulate fragments to break off and be preserved in the net-textured sulfide surrounding the anteliths.

Patchy net-textured sulfide: Patchy net texture has been described at Raglan (Bazilevskaya, 2009) and Jinchuan (Tonnelier, 2010) and appears to represent localized transition zones between barren peridotite and more common leopard net-textured sulfide. Because olivine is tightly packed throughout patchy net-textured samples, this implies a mechanical explanation for the patchiness of the sulfide.

Disrupted net-textured sulfide: This texture contains patches of leopard net-textured mineralization with crosscutting pyroxenite (Fig. 8). The diffuse, irregular contacts suggest that the pyroxenite was emplaced before the net-textured mineralization had solidified, and the preservation of relict igneous olivine in the barren pyroxenite suggests that it was refractory. The absence of sulfide in the pyroxenite or additional sulfide adjacent to the pyroxenite suggests that it was dissolved. The restriction of disrupted net-textured mineralization to the central parts of the dike, not the upper or lower parts, suggests that this part of the dike was rheologically more susceptible to being intruded.

The nature of the invading phase and the mechanism of replacement of net-textured mineralization are not clear, but the process appears to have involved reactions of the form

and dissolution of sulfides. It may have involved one or more of the following processes:

1. Reaction with an internally or externally derived supercritical Si-Ca–rich fluid phase, as proposed by Barnes et al. (2016) for similar rocks at the Ntaka deposit in Tanzania: Such a process is consistent with the highly irregular contacts and gradational nature of the pyroxenite zones that have invaded leopard net-textured sulfides, but this model does not explain the restriction of the transgressive pyroxenite zones to net-textured sulfide zones.

2. Reaction with an internally derived residual silicate melt: This process is supported by the presence of rare barren pyroxenite crosscutting barren peridotite, but it is not clear whether an internally derived residual liquid would be out of equilibrium enough to replace silicates and sulfides.

3. Reaction with an externally derived but petrogenetically related silicate melt, similar to the process proposed by Spath (2017) for more orthopyroxene ± chromite-rich rocks in the Blue Jay (formerly AT-12) keel of the Black Thor intrusion: Evidence for this includes the large scale (up to 40 cm) of some of the transgressive pyroxenite (App. Fig. A6) and the preservation of relict igneous olivine in disrupted net-textured mineralization, which is consistent with it being refractory during emplacement of the pyroxenite.

The above constraints suggest that the emplacement and crystallization of the Eagle’s Nest dike and the formation of the Ni-Cu-(PGE) mineralization occurred as follows (Fig. 15).

T0 (not shown): This stage features generation of sulfide-undersaturated high-Mg komatiitic magma, most likely related to a mantle plume (see, e.g., Sproule et al., 2002; Herzberg et al., 2007) and ascent near the northern margin of the North Caribou superterrane, possibly related to rifting along that margin (Stott et al., 2010; Houlé et al., 2020).

T1: Intrusion of the first pulse of komatiitic magma, as a blade-shaped dike below the contact between less dense underlying granitoids and overlying more dense mafic-dominated volcanic rocks, formed barren olivine-poor margins.

T2: This stage features incorporation of S-rich country rock, probably from footwall iron formations upstream from the dike (e.g., Carson et al., 2015), and generation of immiscible Fe-(Cu) sulfide xenomelts that were upgraded to Fe-Ni-Cu sulfide melts during transport.

T3: Deposition of olivine phenocrysts and/or olivine primocrysts through in situ fractional accumulation (e.g., Lesher, 1989) formed a network of ortho- to mesocumulate olivine (65–85% olivine, 15–35% silicate melt) and minor anteliths/xenoliths.

T4: Deposition of sulfide melts onto the network of cumulus olivine, displacing less dense silicate melts, formed leopard net-textured mineralization along the length of the dike.

T5: This includes density-driven segregation of massive sulfide melts into embayments along the base of the dike, which formed during intrusion but were likely enlarged by thermomechanical erosion by sulfides (Groves et al., 1986; Williams et al., 1998).

T6: Additional sulfide (waning of system) was deposited to form disseminated sulfides and patchy net-textured sulfides along the length of the dike.

T7: Invasion of the dike by one or more pulses of pyroxenite melt and/or Si-Ca rich fluid generated disrupted net texture.

T8: Crystallization of monosulfide solid solution (MSS) and mobilization of residual sulfide melt (now represented by chalcopyrite) formed crosscutting veins of massive sulfide that transgressed other mineralized zones.

T9 (not shown): Waning magmatism, cooling, and solidification of the Eagle’s Nest bladed dike was followed by serpentinization of olivine.

T10 (not shown): Burial, deformation (rotation, faulting, local shearing), metamorphism to greenschist facies, and metasomatism by CO₂-rich fluids produced local talc-carbonate alteration of the host rocks.

If this model is correct, then the different inclusion populations imply changes in accessibility of sources for inclusions: early olivine aggregates in leopard net and peridotite-dunite anteliths in inclusion net, followed by no inclusions in patchy net, followed by chromitite anteliths in disseminated.

1. The Eagle’s Nest Ni-Cu-(PGE) deposit is hosted by a subvertical blade-shaped dike, which appears to have been emplaced subhorizontally and rotated 90° to the southeast along with the overlying Double Eagle intrusion (in the resulting geometry, the southeast portion is the top of the dike).

2. Most of the rocks were hydrated and locally carbonated during regional greenschist metamorphic grade, but they are only locally penetratively deformed, preserving a wide range of igneous textures.

3. The barren ultramafic rocks and the silicate components of the mineralized rocks are mainly lherzolite and olivine websterite with lesser wehrlite and harzburgite and rare dunite. Relict igneous olivine compositions range Fo_82–86, but the MgO compositions of Eagle’s Nest cumulate rocks suggest that the silicate component of the mineralized rocks formed from a more magnesian-rich olivine, which was not found during this study.

4. The S isotope compositions at Eagle’s Nest are consistent with an S source derived from a mantle-derived magma, local iron formation in country rocks, or both. The very narrow variations of S isotope data at Eagle’s Nest compared with the larger variations in smaller Ni-Cu-(PGE) occurrences in the Black Thor intrusion are consistent with it being a more dynamic system in which multiple isotopically distinct S sources (represented by the wider ranges in analyzed country rocks and smaller showings in the Black Thor part of the system) were homogenized.

5. The orebody is zoned northwest to southeast (i.e., bottom to top) from discontinuous massive and semimassive sulfides, a dearth of sulfides in the 35 to 60 wt % sulfide range, and then through net-textured sulfides to disseminated sulfides, consistent with gravitational segregation in the original orientation.

6. The net-textured sulfide facies contains five subfacies: a more or less continuous zone of leopard net-textured sulfide subfacies containing discontinuous zones of inclusion net, pinto net, patchy net, and disrupted net subfacies.

7. The Eagle’s Nest dike appears to represent a dynamic magmatic conduit. The ores and host rocks appear to have formed from at least three distinct magma pulses. The first pulse was a moderate- to high-Mg komatiitic magma that was sulfide undersaturated and crystallized the sulfide-free units lherzolite, harzburgite, wehrlite, which are preserved discontinuously along the northwest margin and continuously along the southeast margin. The second phase was a variably high- to moderate-Mg komatiitic magma that was sulfide saturated and appears to have carried olivine phenocrysts and Fe-Ni-Cu sulfide droplets. After local segregation of MSS and residual sulfide liquid, all of the mineralization—regardless of texture or location in the system—has similar metal tenors and ³²S/³⁴S isotope ratios, so was likely emplaced in several semicontinuous magmatic events. Each contained different inclusion types and/or formed different aggregate/oikocryst types: gabbro xenoliths and peridotite anteliths in massive sulfide facies, coarse olivine aggregates in leopard net-textured sulfide subfacies, orthopyroxene oikocrysts in pinto net-textured sulfide subfacies, peridotite and smaller olivine aggregates/inclusions and clinopyroxene oikocrysts in inclusion net-textured sulfide subfacies, and chromitite inclusions in disseminated sulfide facies. The third phase was a sulfide-undersaturated low-Mg komatiitic magma or Si-rich fluid—similar to that present in the feeder and lower part of the Black Thor intrusion—that locally disturbed mainly the net-textured sulfides, producing the disrupted net-textured subfacies.

8. The Eagle’s Nest dike contains mainly cumulate rocks and appears to have lost most of the residual magma to unexposed or unpreserved intrusions or volcanic rocks. It contains small chromitite anteliths/xenoliths but does not appear to have been the feeder to the overlying chromite-rich, sulfide-poor Double Eagle intrusion.

9. The Eagle’s Nest Ni-Cu-(PGE) deposit contains the widest range of net-textured sulfide subfacies reported thus far, allowing for insight into the genesis of net-textured sulfides as a whole and making it one of the best places in the world to study textures of this type.

This paper is dedicated to the memory of Prof. A.J. Naldrett, who was a lifelong mentor to CML and an inspiration to MGH and NZ. The project was supported by the Targeted Geoscience Initiative program of the Geological Survey of Canada, a Natural Sciences and Engineering Research Council (NSERC)-Cliffs/Noront Collaborative Research and Development grant (446109-12) and a NSERC-Discovery grant (203171-2012) to Dr. C.M. Lesher, and a Geological Survey of Canada (GSC) Research Affiliate Program award to Natascia Zuccarelli. We are very grateful to Alan Coutts (Noront) for logistical support; Matt Deller and Geoff Heggie (Noront) for insights and discussions during core logging; Cory Exell, Rob Lyght, Tristan Megan, and Roydon Spence (Noront) for assistance during core logging and the rest of the Esker Camp crew for their hospitality; Dave Crabtree and Sandra Clarke (Geo Labs) for assistance with SEM and EPMA analyses; Ed Ripley (Indiana University) for assistance with S isotope analyses; Pedro Jugo (MERC/Harquail School of Earth Sciences) for insightful discussions; Charley Duran (IGO Ltd.) and Wouter Bleeker (GSC) for comments on an early version; and James Mungall (Carleton University), Anne-Aurélie Sappin (GSC), and an anonymous reviewer for insightful comments on the submitted version. This is Mineral Exploration Research Centre contribution MERC-2021-03 and Geological Survey of Canada/NRCan contribution 20210462.

Natascia Zuccarelli (M.Sc, P.Geo) is a project geologist currently working in surface exploration across northern Ontario. She holds a Bachelor of Science degree in geology from the University of Toronto (2016) and a Master of Science degree from Laurentian University (2020; Eagle’s Nest Ni-Cu-(PGE) deposit). Since graduating she has focused her work on the Timmins gold camp, with diverse work experiences including surface exploration, underground mine geology, and detailed 3-D modeling based on decades of historical mining.