The Endako low-F granodiorite-type porphyry Mo deposit is hosted by the Triassic to Eocene Endako batholith, which comprises five temporally distinct plutonic suites, only one of which is mineralized. Pre-mineralization suites range in composition from diorite to granodiorite. The synmineralization Jurassic–Cretaceous François Lake suite includes two granodiorite- to monzogranite-bearing subsuites. Postmineralization phases include the Eocene Sam Ross Creek monzogranite. The batholith spans a silica range of 44–80 wt.% and consists of metaluminous to slightly peraluminous, low- to high-K, I-type granitoids; the Sam Ross Creek phase is an A-type granite. Positive εNd(T) values (+1.1 to +7.2) indicate derivation predominately from juvenile source materials, but with variable input from an older crustal component. Evidence suggests generation of older plutonic suites in a juvenile arc-type setting and younger K-rich felsic suites via recycling of juvenile arc crust without significant mantle-derived contributions. Three distinct Mo-deposition events in the Endako camp are linked to repeated generations of oxidized, highly evolved monzogranitic phases (pre-ore dykes, aplitic Nithi and Casey intrusions) belonging to both François Lake subsuites. Late pre-ore dykes with “Casey-like” geochemical signatures, along with massive unmineralized Casey intrusions near the Endako deposit, could reflect repeated injections from an underlying magma chamber that remained molten during the youngest Mo-deposition event. A genetic link may exist between the Sam Ross Creek phase, a pluton with Climaxtype granite characteristics, and Eocene kaolinite alteration in the Endako deposit. Also, potential exists for Eocene-age Climax-type Mo mineralization within the Endako mining camp.
Porphyry molybdenum deposits have been classified into (i) an alkalic–calcic (and alkalic) granite type, exemplified by the Climax deposit; and (ii) a calc-alkaline granodiorite type, an example being the Endako deposit (Mutschler et al. 1981; Westra and Keith 1981). Research over the last 20 years on porphyry molybdenum deposits, and their cogenetic granitoid suites, has focused almost exclusively on Climaxtype deposits (e.g., Carten et al. 1988, 1993; Theodore and Menzie 1984; Keith et al. 1993 and references therein), because these deposits exhibit both higher grades and larger tonnages than granodiorite-type deposits. Comparable geological, geochemical, or isotopic research has not been published on granitoid suites associated with granodiorite-type molybdenum deposits. Such studies are particularly relevant to answering questions involving the petrogenesis of such granitoid suites and the metals they host (cf. Blevin and Chappell 1995; Blevin et al. 1996; Christiansen and Keith 1996) and the tectonic settings in which they are generated (cf. Westra and Keith 1981; Sillitoe 1996).
Within the Cordillera, the Endako deposit of central British Columbia (Fig. 1) represents both the oldest (ca. 145 Ma) porphyry molybdenum deposit (Villeneuve et al. 2001) and the largest mined granodiorite-type molybdenum deposit (Keith et al. 1993). The Endako deposit is hosted by the northwesterly trending Triassic to Eocene, composite Endako batholith, which comprises at least five temporally distinct plutonic suites–episodes, all but one of which is weakly mineralized (Whalen and Struik 1997; Anderson et al. 1998a, 1998b; Struik and Whalen 1998; Whalen et al. 1998; Villeneuve et al. 2001). Herein, we focus mainly on the Jura-Cretaceous François Lake plutonic suite, with which Mo mineralization is spatially and likely genetically related. To document the evolutionary history of the Momineralized magmatic system, the Late Triassic to Middle Jurassic plutonic suites are also described briefly and their geochemical features compared to those of the François Lake suite. In-depth coverage of the geochemical and isotopic features of the Endako batholith, with emphasis on the older plutonic suites, will be presented elsewhere.
Geology of the Endako batholith
The composite Endako batholith forms a wide swath extending broadly southeast from south of Babine Lake to north of François Lake, spanning the Fort Fraser and Nechako River map areas (National Topographic System map sheets 93 F and 93 K) (Fig. 1). The various intrusive units of the batholith were initially grouped into the Topley Intrusions (Hanson and Phemister Thomas 1929). In the early 1970s, the division and distribution of the calc-alkaline plutonic suites surrounding and hosting the Endako deposit were extensively mapped (Carr 1965, 1966; Dawson 1972; Dawson and Kimura 1972; Kimura et al. 1976; Bysouth and Wong 1995). Carter (1982) utilized existing data to redefine and rename the granodiorite- to monzogranite-dominated plutonic suite related to porphyry Mo mineralization, as the François Lake suite. Kimura et al. (1976) interpreted the batholith as consisting of deeper level mafic intrusions, located along the margins, broadly changing to felsic epizonal plutons along a central axis. It is within this central axis that the Endako molybdenum deposit is located.
Recent geological mapping and geochemical studies (Whalen and Struik 1997; Anderson et al. 1998a, 1998b, 1998c; Struik and Whalen 1998; Whalen et al. 1998), together with geochronological studies (Villeneuve et al. 2001), have further refined the geology of the Endako batholith and documented a 75 Ma time frame for its emplacement. It has been subdivided into three main Mesozoic suites and a number of subsuites (groupings of likely cogenetic intrusions), each of which was subdivided into various phases (lithological units) and subphases (cosanguineous subunits). Brief descriptions of these, in addition to some post-Jurassic phases, are given in the following sections (from old to young).
Late Triassic (220–215? Ma) Stern Creek plutonic suite
Originally termed the Boer plutonic suite (Anderson et al. 1998a; Whalen et al. 1998), this suite has been renamed to reflect new geochronological data. These intrusions, which occur along the northern and northeastern part of the batholith, are commonly mesocratic, foliated, equigranular hornblende ± biotite diorite to granodiorite with local Kfeldspar phenocrysts and significant variation in texture and composition on the outcrop scale. Foliation varies from strong mineral alignment to development of a gneissic texture. A U–Pb zircon age of 219.3 ± 0.4 Ma obtained from the west end of Fraser Lake (Fig. 1) is interpreted as representative of the crystallization age for the Stern Creek suite (Villeneuve et al. 2001).
Early to Middle Jurassic (181–160 Ma) Stag Lake plutonic suite
The compositionally heterogeneous (Fig. 2) Stag Lake plutonic suite, which defines the western, northeastern, and eastern batholith margins (Fig. 1), consists of sequentially intruded biotite–hornblende diorite and gabbro, and hornblende–biotite quartz monzodiorite and granodiorite; biotite quartz monzonite to monzogranite may represent cogenetic felsic phases. The rocks are commonly mesocratic and xenolith-, clinopyroxene-, and titanite-bearing. The suite is undeformed, except near contact zones between mafic and intermediate phases, where mineral and xenolith foliation is developed.
The early dioritic to gabbroic Boer phase (Fig. 1) yielded a U–Pb zircon crystallization age of 181.0 ± 1.0 Ma, and 40Ar/39Ar mineral ages on later phases span an age range of 173–160 Ma (Villeneuve et al. 2001). Based on field relationships (Struik et al. 1997), the Stag Lake plutonic suite includes phases younger than those that have been radiometrically dated.
Late Jurassic (159–154 Ma) and Jura-Cretaceous (148–145 Ma) subsuites of the François Lake suite
The Jurassic François Lake biotite monzogranite to granodiorite phases, which underlie the northeastern part of the Endako batholith (Fig. 1), serve as host rock to Mo mineralization in the Endako area. The suite has been subdivided into two subsuites on the basis of composition (Anderson et al. 1998a) and age (Villeneuve et al. 1997; Villeneuve et al. 2001). The Glenannan subsuite, consisting of the Glennanan and Nithi phases, with their various subphases, displays a range of compositions from biotite and hornblende–biotite monzogranite to granodiorite (Fig. 2), is generally medium to coarse grained, and is located in the north-central to northwestern parts of the batholith. The Endako subsuite includes the Endako phase, its François subphase, the Casey phase, and a set of pre-ore felsic dykes. This suite generally consists of medium- to fine-grained monzogranite to granodiorite units (Fig. 2). The François Lake plutonic suite records a complex, protracted history that includes emplacement, solidification, locally intense veining, felsic dyke injection, Mo and pyrite mineralization and alteration, and late dyke intrusion, fracturing, and jointing.
The Glennanan phase forms a large body to the northeast of the Endako deposit. Although variable in texture, it is generally massive, unaltered, coarse to very coarse grained, K-feldspar-porphyritic, biotite (±hornblende) granodiorite to monzogranite (Struik et al. 1997; Whalen et al. 1998). The Tatin Lake subphase crops out over a significant area north of the Endako River valley, fringing the Glenannan phase. It consists mostly of beige to pink, fine- to medium-grained, equigranular to K-feldspar-subporphyritic, biotite monzogranite, but also includes beige to grey, fine- to medium-grained, equigranular, hornblende–biotite and biotite–hornblende granodiorite. The Tatin Lake subphase was previously mapped by Kimura et al. (1980) as part of the Casey phase. Observed contacts of the Tatin Lake subphase and Glenannan phase included both sharp intrusive contacts, the Tatin Lake subphase being younger, and gradational contacts, with the Glenannan phase becoming less coarse grained, less porphyritic, and more felsic over a distance of less than 0.25 km (Whalen et al. 1998). Biotite from a Glenannan phase sample collected from just northeast of the Endako deposit yielded a 40Ar/39Ar age of 157.2 ± 1.5 Ma (Villeneuve et al. 2001). However, another sample collected 20 km farther north from proximal to the Hanson Lake phase yielded concordant 40Ar/39Ar hornblende and biotite ages of 139.6 ± 1.5 and 138.8 ± 1.4 Ma (M.E. Villeneuve, unpublished data, 2000), suggesting that either the Glenannan phase is composite, or thermal flux from the younger (ca. 127 Ma) Hanson Lake phase disturbed the Ar systematics of this sample.
The other major phase in the Glenannan subsuite, the Nithi phase, includes two textural variants exposed on and adjacent to Nithi Mountain, which host and are apparently cogenetic with the significant Nithi porphyry Mo showing (L’Heureux and Anderson 1997). The seriate variant is a medium-grained, alkali feldspar porphyritic, biotite monzogranite. The aplitic variant, a pink to brown, fine- to medium-grained, aplitic biotite monzogranite, was included in the Casey phase in previous mapping (L’Heureux and Anderson 1997). This was reevaluated based on molybdenite hosted by this phase yielding a Re–Os age of ca.154 Ma (Selby and Creaser 2001). Biotite from a seriate Nithi phase sample, collected distal from its contact with the aplitic Nithi variant, gave a 40Ar/39Ar age estimate of 154.5 ± 1.9 Ma. Although considered to be a minimum age, this age is also considered to be a reasonable representation of the crystallization age (Villeneuve et al. 2001).
The Endako phase forms a northwest–southeast elongate body in the south-central portion of batholith. This phase consists of coarse-grained, dark pink to orange, biotite– hornblende granodiorite to monzogranite (Fig. 2), subporphyritic with distinctive orange K-feldspar (0.1–1 cm). Except within the Endako Mo deposit, it is remarkably fresh, with unaltered mafic minerals and a paucity of veining. The François subphase of the Endako phase, interpreted as intruding the Endako phase (Bysouth and Wong 1995), is exposed adjacent to the Endako phase north of François Lake. It consists of rusty brown to red, medium-grained, equigranular, biotite (±hornblende) granodiorite to monzogranite. Although relatively equigranular compared to the Endako phase, where it is unaltered, it contains the same distinctive orange K-feldspar. This, coupled with apparently gradational contacts with the Endako phase, suggests that it is an equigranular variant of the Endako phase (Whalen et al. 1998). Biotite 40Ar/39Ar ages from the younger Endako subsuite samples overlap: 148.4 ± 1.5 Ma for the Endako phase and 147.9 ± 1.5 Ma for the François subphase (Villeneuve et al. 2001), results which further substantiate the François as an Endako subphase.
The Casey phase is exposed only south of the Endako River valley, bordering the Endako phase on its northeast side. Texturally and compositionally similar plutonic rocks north of this valley, which were previously included by Kimura et al. (1980) within this phase, were interpreted by Whalen et al. (1998) as a subunit of the Glenannan phase (Tatin Lake subphase) (Fig. 2). Also, as noted above, lithologically similar plutonic rocks exposed around Nithi Mountain, previously included in the Casey phase, have been, based on new Re–Os data (D. Selby, personal communication, 2000), included within the Nithi phase. The Casey phase consists of fine- to medium-grained, dark pink, granophyric biotite monzogranite or aplitic monzogranite (Fig. 2). Small miarolitic cavities are relatively common. Portions of this unit contain dark pink K-feldspar and ovoid quartz-eye phenocrysts. Biotite is commonly oxidized or chloritized. A sample of the Casey phase collected just northeast of the Endako deposit gives a U–Pb zircon age of 145.1 ± 0.2 Ma, the interpreted age of crystallization (Villeneuve et al. 2001).
The Endako batholith is intruded by two sets of dykes. One set, common within the Endako deposit, is crosscut by Mo-bearing fine-grained quartz veins and therefore intruded prior to mineralization (pre-ore dykes). These dykes are pink felsic and mafic poor; some are aphanitic with quartz and Kfeldspar phenocrysts, whereas others lack phenocrysts and are fine-grained granophyric. A pre-ore dyke sample collected from within the Endako deposit provided a minimum age on the Endako phase and a maximum age on mineralization. It yielded a zircon U–Pb age of 147.4 ± 0.6 Ma and a thermally reset biotite 40Ar/39Ar age of 144.7 ± 1.4 Ma (Villeneuve et al. 2001). The second dyke set comprises Eocene (Grainger et al. 2001) mafic dykes, which are also common in the Endako deposit.
Post-Jurassic phases occur as small (<5 km), scattered stocks within the Endako batholith. The Hanson Lake phase was formerly interpreted by Whalen et al. (1998), based on lithological similarities and geochemistry, as a subphase of the Glenannan phase. However, it has yielded overlapping 40Ar/39Ar hornblende and biotite ages of 126.1 ± 1.6 and 129.3 ± 1.3 Ma (M.E. Villeneuve, unpublished data, 2000), indicating an Early Cretaceous age of emplacement. The Hanson Lake phase, which forms an ovoid body on the northern end of the Glenannan phase (Fig. 1), consists of grey to white, coarse- to medium-grained, feldspar-porphyritic hornblende–biotite granodiorite to quartz monzonite. In places, it has a fine-grained matrix, in which hornblende and quartz are also phenocrysts.
The Fraser Lake suite forms compositionally similar small stocks along the northwestern, northeastern, and northern margins of the Triassic–Jurassic suites. The Fraser Lake stock, which occurs just south of the town of Fraser Lake, comprises a medium-grained, equigranular, biotite monzogranite. It has yielded an 40Ar/39Ar biotite and U–Pb zircon age of ca. 112 Ma (M.E. Villeneuve, unpublished data, 2000).
The Sam Ross Creek phase is located about 5 km northwest of the Endako deposit; related stocks are scattered to the south and west, south of François Lake. The rocks are dark red to purple, medium- to coarse-grained, strongly miarolitic biotite monzogranite. Miarolitic cavities range from 0.5 to 3 cm in size and are lined by euhedral quartz, feldspar, and rare fluorite and molybdenite. Pervasive kaolinite alteration of constituent feldspar is typical, and granophyric textures common. A hill just to the west of the Sam Ross Creek pluton is underlain by dark red brown, white plagioclase-phyric (0.1–0.3 cm), rhyolitic rocks that, based on their lack of obvious flow banding or bedding and distinctive geochemical features, represent a higher level equivalent of the Sam Ross Creek pluton. The suite intrudes coeval Eocene Ootsa Lake Group felsic rocks and yielded an Eocene U–Pb age of 50.6 ± 0.3 Ma (Villeneuve et al. 2001). Previously, based on its appearance and distribution, the Sam Ross Creek pluton was interpreted as a higher level and (or) water-oversaturated equivalent of the François subphase, which itself was interpreted as bearing a similar relationship to the Endako phase (Whalen et al. 1998).
Geology of porphyry Mo mineralization hosted by the Endako batholith
Significant porphyry Mo mineralization within the Endako batholith occurs in two locations, the Endako deposit and the Nithi Mountain area (Lefebure and Hoy 1996), of which the former has been the subject of much more detailed study. The Endako deposit is hosted by the Endako phase and is associated with two distinct quartz–Mo-bearing vein types and three alteration events (Bysouth and Wong 1995; Kimura et al. 1976; Selby et al. 2000). Although minor primary Mo mineralization is associated with K-feldspar alteration along stockwork veins (Bysouth and Wong 1995; Kimura et al. 1976), the majority of the ore is associated with ribbon veins bordered by sericite alteration (Selby et al. 2000). Ribbon veins at Endako were brecciated by Eocene-age deformation, with kaolinite alteration, along with calcite veining, overprinting earlier ore-related sericite alteration (Selby et al. 2000). Based on the spatial relationship within the orebody between Eocene mafic dykes, faults, and calcite veining and kaolinite in the Endako deposit and surrounding area, a Tertiary age has been suggested for the kaolinite alteration (Selby et al. 2000).
The Nithi Mountain area, the site of about a dozen British Columbia Ministry of Energy, Mines and Petroleum Resources MINFILE localities which are about evenly distributed between the seriate and aplitic Nithi phase variants (Bailey et al. 1995), has been recently studied by L’Heureux and Anderson (1997). They found that the seriate Nithi phase exhibits intense clay alteration near and southwest of Nithi Mountain; propylitic alteration is common in all phases. East-northeast-trending Mo-bearing veins and later veining, aplite intrusion, jointing, and fracture formation record late-stage events in emplacement of the aplitic Nithi phase; all are localized to within 3 km of its contact with the seriate Nithi phase.
Although the primary molybdenite mineralization is tied to the latest stages of Jurassic plutonism, the actual source of the Mo has remained an enigma. At the Endako deposit, Bysouth and Wong (1995) call upon an unrecognized igneous event that postdated the Casey and Endako – François Lake phases and formation of the South Boundary Fault located south of the Endako deposit. Mo mineralization of the Endako phase and the intruding felsic pre-ore dykes clearly shows that hydrothermally driven mineralization postdated these phases. Two hydrothermal biotite 40Ar/39Ar age ages of 143.7 ± 1.5 and 145.2 ± 1.5 Ma are interpreted by Villeneuve et al. (2001) as thermally reset ages that reflect the sericite alteration plus Mo-mineralization event, not the earlier K-feldspar alteration event, during which the biotite formed. Recent Re–Os molybdenite dating of ribbontextured veins at the Endako deposit yielded two distinct ages, ca. 148 and ca. 145 Ma (Selby and Creaser 2000, 2001). In addition, molybdenite from the Nithi showing yielded a Re–Os age of ca. 154 Ma (Selby and Creaser 2001). Although primary Mo mineralization is clearly tied to the terminal stages of batholith formation, this new geochronological evidence indicates multiple mineralizing stages–events.
Alteration and major elements
Table 1 contains representative analyses of phases making up the François Lake plutonic suite plus younger suites– phases from the Endako batholith, and the compete data set is available in Struik and colleagues.2 Analyses in Table 1 were selected to include those on which Nd isotopic analyses were carried out (Table 2). Analytical techniques are described in the Appendix.
The aplitic Nithi phase and the Sam Ross Creek phase exhibit petrographic evidence for deuteric and (or) hydrothermal alteration. Comparison of samples of the Endako phase collected proximal and distal to mineralization at the Endako deposit indicates that even for the more mobile alkali elements (K, Na, and Rb) hydrothermal alteration has not changed the main geochemical characteristics described herein. Although the freshest available samples were chosen for geochemical analyses, some effects of alteration, such as the slightly peraluminous character (i.e., aluminium index ≈ 1.1) of some samples (e.g., a few Stag Lake suite and aplitic Nithi and Casey phase samples) (Fig. 3b), could not be avoided.
Trends defined by suites on modal quartz–K-feldspar– plagioclase (QKP) diagrams can be used to identify magmatic associations (Lameyre 1987) (see insert in Fig. 2). On this basis, the Endako batholith mainly follows a calc-alkalic trend. A calc-alkaline affinity is also indicated by an AFM (Al2O3–FeO–MgO) plot (not shown). The batholith spans a significant compositional range (SiO2 = 44–80 wt.%) and includes low-, medium-, and high-K components (Fig. 3a), with the post-Middle Jurassic François Lake and Sam Ross Creek suites being the most evolved (i.e., highest K2O and SiO2). All but the felsic (>70% wt.% SiO2) phases include amphibole-bearing portions; intermediate to felsic components are sodic (Na2O > 3.0 wt.%) and metaluminous to slightly peraluminous (<0.1) (Fig. 3b).
Endako batholith samples exhibit a large range in trace elemental concentrations (Fig. 4), as would be predicted from the large silica range they span (Fig. 3a). All suites exhibit negative Nb and Ti anomalies; the magnitude and direction of Ba, Sr, and Eu anomalies are variable. By “anomalies,” we refer to positive or negative spikes for an element with respect to adjacent elements in the normalized pattern. In general, the François Lake suite is enriched in light lithophile elements (LILE) relative to the Late Triassic to Middle Jurassic Stern Creek and Stag Lake suites (Figs. 4a, 4b); in addition to LILE, the Eocene Sam Ross Creek phase is significantly enriched in high field strength elements (HFSE) relative to all other suites (Fig. 4g). The Stern Creek suite and lower silica Stag Lake suite samples, like those from the Boer phase, are less depleted in heavy rare earth elements (HREE) than François Lake suite samples.
Extended element normalized patterns for Glenannan subsuite phases and the Early Cretaceous Hanson Lake phase are presented in Figs. 4c and 4d. Spatially well separated samples of the Glenannan phase exhibit very similar patterns which are indistinguishable from Early Cretaceous Hanson Lake phase patterns. Tatin Lake subphase samples include granodioritic compositions (Figs. 2, 3a) which overlap closely Glenannan phase patterns and monzogranitic compositions which exhibit higher LILE content, much more pronounced negative Ba, Sr, and Eu anomalies, and overall lower contents of the normalized elements between La and Er. Although seriate Nithi phase patterns are less tightly clustered than those of the Glenannan phase, they mainly parallel patterns exhibited by that phase. Patterns of aplitic Nithi phase samples resemble those of monzogranitic Tatin Lake subphase and Casey phase samples (Figs. 4d, 4f). Extended element normalized patterns for Endako subsuite phases, the Fraser Lake suite, and the Sam Ross Creek phase are presented in Figs. 4e, 4f, 4g. Normalized patterns of eight samples from the Endako phase and three from its François subphase are remarkably tightly clustered and indistinguishable from each other. Relative to the Endako phase, Casey phase samples are enriched in LILE and strongly depleted in Ba and Sr, and the elements between La and Er. The Casey phase patterns closely resemble those of felsic samples from the Tatin Lake subphase and the aplitic Nithi phase (Fig. 4d). Two felsic pre-ore dyke samples exhibit contrasting patterns: (i) an older type, dated at 147.4 ± 0.6 Ma (Villeneuve et al. 2001), that essentially parallels the François subphase pattern, but at slightly lower concentrations of the elements between La and Sm; and (ii) a probable younger type whose pattern is indistinguishable from those exhibited by a number of Casey phase samples, including the sample dated at 145.1 ± 0.2 Ma (Villeneuve et al. 2001). Two Fraser Lake suite patterns (Fig. 4g) are very similar, but differ in elemental abundances to the right of U, with the lowest silica content (67 wt.%) sample (Fig. 3a) containing higher elemental concentrations. Fraser Lake suite patterns bracket the François phase patterns, except the former contains higher U and Th, similar to Glenannan subsuite phases, such as the Glenannan (Fig. 4c). The Eocene Sam Ross Creek phase (Fig. 4g) differs from the Endako subsuite in being remarkably enriched in the normalized elements to the right of Nb, particularly the HFSE.
Variations in elements which are strongly partitioned into feldspars (i.e., Eu, Ba, Sr, and Rb) are illustrated in Fig. 5. Within the François Lake suite, Ba and Sr contents decrease with magnitude of negative Eu anomaly, whereas the opposite is true for Rb. The most evolved compositions, that is those samples with the lowest Sr and Ba contents, highest Rb contents, and most negative Eu anomalies, belong to the Casey, Tatin Lake, aplitic Nithi, and Sam Ross Creek phases plus the younger or “Casey-like” pre-ore dyke.
Nd isotopic data, expressed as values of epsilon Nd at the time of formation (εNd(T)), from a sample set representative of various plutonic suites making up the Endako batholith are plotted verus time in Fig. 6. Isotopic compositions from the François Lake plutonic suites are presented in Table 2. All suites exhibit consistently positive εNd(T) values: Stern Creek suite (+3.9 to +6.0), Stag Lake suite (+3.1 to +7.2), François Lake suite – Glenannan subsuite (+5.2 to +5.9), Endako subsuite (+4.3 to +4.7), Fraser Lake suite (+5.5), and Sam Ross Creek pluton (+1.1).
Exclusively positive εNd(T) values indicate that Endako batholith granitoid suites were derived predominately from a reservoir with a history of light rare earth element (LREE) depletion, such as the depleted mantle (DM), which during this period had an εNd(T) range of +7.3 to +9.8 (Fig. 6). However, only one Stag Lake suite sample overlaps, within analytical uncertainty, the DM range, and variations in εNd(T) values within and between suites suggest variable input from a more LREE enriched older crustal component. This is reflected in depleted mantle model ages (TDM) for the François Lake suite which are significantly older than radiometric ages of crystallization (Villeneuve et al. 2001) (Table 2). As well, the evolution curve for the Stern Creek suite only overlaps the lower end of the εNd(T) range shown by the Stag Lake suite and ranges to less positive values than those exhibited by the François Lake and Fraser Lake suites (Fig. 6). This suggests that the younger suites received greater contributions from a juvenile or DM-like component than did the Stern Creek suite. However, the Stag Lake suite exhibits an εNd(T) range whose evolution with time overlaps the isotopic range in the younger François Lake and Fraser Lake suites. In contrast, the Sam Ross Creek sample evolution curve lies at less positive εNd(T) values than all other samples, dictating derivation from a distinct more LREEenriched source or a different source component mixture than the other suites.
Granitoid classification and tectono-magmatic affinity
Based on the presence of amphibole-bearing mafic to intermediate phases, sodic (Na2O > 3.0 wt.%) and metaluminous to slightly peraluminous (< 1.1) (Fig. 3b) compositions, and a lack of quartz enrichment characteristic of S-type granites (Fig. 2), granitoids making up the Endako batholith represent, by definition, infracrustal or I-type granitoids (Chappell and White 1992). On tectono-magmatic discrimination diagrams (Fig. 7a), all but one phase in the Endako batholith plot within the field of volcanic-arc or subduction-related (VAG) magmatism; the exception, the Eocene Sam Ross Creek phase, plots in the field of withinplate granites (WPG). The Sam Ross Creek phase is indicated to have an A-type granite affinity by the discrimination diagrams of Whalen et al. (1987) (not shown). The presence of well-defined negative Nb anomalies (Fig. 4) over a silica range of 44–80 wt.% substantiates formation of the Late Triassic to Jura-Cretaceous plutonic suites making up the Endako batholith in an arc-type setting or via recycling of crust formed in such a setting. Similar negative Nb anomalies within the Eocene Sam Ross Creek phase likely reflect derivation from older arc crust. However, based on the εNd(T) value (Fig. 6), this phase would not be derived via simple remelting of the granulitic lower crustal restites of pre-Eocene Endako batholith suites (cf. Whalen et al. 1987).
Petrogenetic processes and granitoid source materials
Overall, changes in the ratios of Sr, Ba, Rb, and Eu/Eu* within the Endako batholith parallel the mineral vector trends (Fig. 5). Trends within the Stern Creek and Stag Lake suites appear to be compatible with fractionation of pyroxene and plagioclase. Trends within granodioritic François Lake suite phases (Glenannan, Endako–François, seriate Nithi) look compatible with plagioclase plus alkali feldspar removal, whereas within monzogranitic phases (younger pre-ore dykes, Casey, some Tatin Lake, aplitic Nithi, and Sam Ross Creek) fractionation of alkali feldspar probably played a predominant role. Depletion in middle rare earth elements (REE) and Zr in these most evolved samples indicate fractionation of a REE-enriched accessory phase, possibly zircon (Figs. 4d, 4f, 4g). Magma evolution from granodioritic to monzogranitic compositions via similar fractional crystallization processes appears to have been repeated in the François Lake suite in both the Glenannan and Endako subsuites.
Published granitoid Nd isotopic data from the western Stikinia terrane, adjacent accreted arc terranes to the west (Alexander and Wrangellia), and the Coast Plutonic Complex (see Fig. 6b) suggest they, like Endako batholith suites, were mainly derived from isotopically juvenile sources. This terrane-scale feature is most easily interpreted as mainly reflecting the nature of the crust underlying these terranes (Samson et al. 1989; 1990), such that both potential crustal and mantle derived contributions to granitoid magmas would be isotopically juvenile with respect to Nd. Earlier Stern Creek and Stag Lake suites include prominent gabbro–diorite phases that can only reflect significant direct mantle input. These older suites would be expected to represent products of assimilation and fractional crystallization (AFC) processes (cf. DePaolo 1981) between depleted mantle and relatively juvenile crustal components.
The younger felsic LILE-enriched François Lake plutonic suite, in which cogenetic mafic phases are lacking, could in large part reflect remelting or recycling of juvenile arc crust (see Fig. 6b) without significant direct depleted mantle derived contributions. Roberts and Clemens (1993) suggest that generation of high-K granitoid suites, like the high-K Endako suites (Fig. 3a), is only possible via recycling of preexisting continental crust. This model involves fluidabsent partial melting of hydrated, mafic to intermediate, transitional to high-K calc-alkaline metaigneous protoliths in the lower to middle crust via heat supplied from underplated or intraplated mafic magmas.
Tectono-metallogenic affinity of Endako granitoid suites and porphyry Mo deposits
The Endako deposit has been previously classified as a calc-alkaline granodiorite-type molybdenum deposit (Westra and Keith 1981; Mutschler et al. 1981). Data presented herein support this, including calc-alkalic affinity and prominence of granodioritic to monzogranitic compositions (Fig. 2), a potassium range at 57.5% silica of 1–2.5 (Fig. 3a), and F and Sn analyses (Table 1). Except for three Endako phase samples collected within the Endako mine (e.g., WXBC97-57 in Table 1), all François Lake suite samples contain <600 ppm F (Fig. 8), at the lower end of the range of 500–1500 ppm indicated for this deposit type by Westra and Keith (1981). In contrast, three of five Sam Ross Creek phase samples contain elevated F contents (1000–1500 ppm) similar to the Endako phase open-pit samples. Except for the samples collected proximal to the mine, Sn contents of all samples are <3 ppm.
Blevin and Chappell (1995) and Blevin et al. (1996) have shown that ore element ratios (Sn/W/Cu/Mo, etc.) of graniterelated mineralization are a function of relative oxidation state (whole-rock Fe2O3/FeO) and degree of fractionation (whole-rock Rb/Sr) within associated granite suites. Keith et al. (1993) also emphasize high fo2 as the most critical factor in making Mo behave as a highly incompatible element within a fractionating granitic magma. In the Rb/Sr versus Fe2O3/FeO plot (Fig. 9), (i) aplitic Nithi, some Tatin Lake, Casey, pre-ore dyke, and Sam Ross Creek samples are more fractionated, i.e., higher silica and Rb/Sr, than samples from other phases, and scatter over the W, Mo, and Sn granite fields; and (ii) almost all samples are quite oxidized (Fe2O3/FeO > 0.1) but, with the exception of the Sam Ross Creek phase, not as oxidized as the field for Mo granites. The more fractionated character of these samples with high Rb/Sr is further substantiated in Fig. 5 by their low Ba and large negative Eu anomalies.
Christiansen and Keith (1996) concluded that ore metal concentration within felsic magmas is a function of many factors, including source composition, magmatic processes, volatile fugacities, and tectonic setting. They employed the Y + Nb versus Rb diagram of Pearce et al. (1984), with ore metals superimposed (Fig. 7b), to summarize compositional– tectonic relationships in mineralized granites. The aplitic Nithi, younger pre-ore dyke, and Casey phases plot in Fig. 7b in the area of Mo-mineralized I-type granites, equivalent to the calc-alkaline granodiorite porphyry Mo type of Westra and Keith (1981) and Mutschler et al. (1981) (Fig. 7b). The Sam Ross Creek phase plots in the area of Mo-mineralized A-type granite, equivalent to the alkaline granite Climax porphyry Mo type.
Westra and Keith (1981) considered porphyry molybdenum deposits to be fundamentally subduction-related, with the calc-alkaline type forming in the continental magmatic arc proximal to the subduction zone and the alkali-calcic and alkalic types forming increasingly more continentward, where back-arc spreading is significant. Within a subduction-zone setting, granitoid magmas may have a significant crustal residence time during which processes, such as hybridization with repeated injections of mantle-derived magma and crustal assimilation via AFC or melting– assimilation–storage–homogenization (MASH) processes, can forestall magma evolution by fractional crystallization and reduce the opportunity for development of incompatible element (e.g., Mo) ores (Christiansen and Keith 1996). In the Endako batholith, the François Lake plutonic suite consists of multiple pulses of chemically fractionated (Figs. 2⇑⇑–5) and isotopically juvenile (Fig. 6) granitoids with a paucity of contemporaneous mafic intrusive rocks. These features are permissive of a relatively long crustal residence time for magma fractionation during which hybridization by new mantle inputs or crustal assimilation were not significant processes, conditions highly favourable for Mo deposit development. An additional factor favourable to the Momineralization process may have been the Endako batholith being the focus of episodic plutonism for ∼170 Ma (Villeneuve et al. 2001).
Discussion of Mo-mineralization petrogenesis
Re–Os molybdenum dating has established the existence of three distinct ages of molybdenum mineralization in the Endako batholith. The earlier Mo deposition age (ca. 148 Ma) at the Endako deposit overlaps 40Ar/39Ar ages obtained from the Endako phase and its François Lake subphase, and the younger molybdenum deposition age (ca. 145 Ma) overlaps the crystallization age of the Casey phase (Selby and Creaser 2000; Villeneuve et al. 2001). Deposition of molybdenum at the Nithi showing (ca. 154 Ma) (Selby and Creaser 2001) overlaps the crystallization age obtained from the seriate Nithi phase and inferred for the associated aplitic Nithi phase (formerly correlated with the Casey phase).
Geochemistry plus geochronology have established the presence of similar evolved–fractionated aplitic monzogranite phases or subphases associated with the Glenannan subsuite (Tatin Lake subphase, aplitic Nithi phase) and the Endako subsuite (pre-ore dykes, Casey phase). Their geochemically evolved features, as evidenced by Rb, Sr, Ba, and Eu/Eu* data (Fig. 5), and oxidized character (Fig. 9) are those expected of granitoid phases cogenetic with Mo mineralization (Blevin et al. 1996; Keith et al. 1993; Christiansen and Keith 1996). Both the Nithi and Casey phases are, based on Re–Os dating (Selby et al. 2000; Selby and Creaser 2000) and field relationships, contemporaneous with and arguably cogenetic with molybdenum mineralization events. The older Re–Os molybdenum deposition age at the Endako deposit may be related to late crystallizing fractionates of the Endako phase. The presence of such magmatic compositions has been indicated by our establishing that the pre-ore dykes are composite, including both older dykes with Endako phase compositions and younger dykes with Casey phase characteristics.
The contrast between the fractured, veined, and mineralized younger Casey-like pre-ore dykes within the open pit and massive, unveined, and barren Casey phase intrusions, just east and northeast of the mine (site of the ca. 145 Ma U–Pb zircon sample) (Fig. 1), might reflect the existence of an underlying magma chamber of Casey-like composition, which remained molten during the younger (ca. 145 Ma) mineralizing event. Studies of other molybdenum deposits have shown that the deep-level (>10 km) granite plutons beneath such systems are relatively unaltered, because the mineralizing fluid was evolved at the top of the system (∼3 km) well before the plutonic roots finished crystallizing (Shinohara et al. 1995).
The irregular intrusion geometries characteristic of Mo porphyries appear to form episodically (Shinohara et al. 1995). For example, at Henderson, at least 11 intrusions with similar compositions were emplaced within a narrow time span (∼5 Ma) in the same geological environment, each of which resulted in various degrees of alteration and mineralization (Carten et al. 1988). In the case of the Endako batholith, the episodic character of the plutonism is much longer lived, with at least three periods of Mo deposition being associated with the Jura-Cretaceous, calc-alkaline François Lake plutonic suite (Villeneuve et al. 2001; Selby and Creaser 2000; D. Selby, personal communication, 2000).
It is interesting to note that the Eocene Sam Ross Creek phase, which was emplaced proximal to the Endako deposit (Fig. 1), represents a high-F (Fig. 8) alkali monzogranite with close geochemical affinity to granites cogenetic with Climax-type porphyry Mo deposits (Fig. 7b). Villeneuve et al. (2001) speculated that this phase may have provided the thermal flux to partially disturb K–Ar and 40Ar/39Ar ages in the Endako deposit area. An apparent spatial relationship between Eocene mafic dykes and late-stage kaolinite alteration was noted by Selby et al. (2000), who found that fluids associated with kaolinite formation were of low salinity and temperature (190–300°C) and dominantly of meteoric origin, and their movement was focused along faults. Possibly the less fractured mafic dykes acted as impermeable barriers which channelled fluid driven by thermal flux from the Sam Ross Creek phase. Features indicative of vapour oversaturation and localized kaolinite alteration within the Sam Ross Creek phase support such an interpretation. Given the repeated previous Mo-deposition events within the Endako batholith and the geochemical characteristics of Eocene plutons, such as the Sam Ross Creek, there may also have been an Eocene Mo-mineralizing event. Of note in this context is the fact that, although the geochemistry suggests the Endako deposit to be of low-F, calc-alkaline type, fluidinclusion characteristics are of those typical of both Climax and low-F type deposits (Selby et al. 2000). The Endako molybdenum camp appears to have the unique (?) potential of including deposits of both calc-alkaline granodiorite and alkalic (Climax) types, a situation which reflects the transition within the camp from arc-type (compressional regime) to within-plate-type (extensional regime) magmatism (Anderson et al. 1998a, 1998b).
The calc-alkaline Endako batholith comprises at least five separate plutonic suites that together span a silica range of 44–80 wt.% and were emplaced episodically over a 170 Ma period (Villeneuve et al. 2001). The premineralization Triassic Stern Creek and Early to Middle Jurassic Stag Lake suites are predominated by dioritic to granodioritic phases, whereas the synmineralization Jurassic–Cretaceous François Lake suite consists of two subsuites (Glenannan and Endako) consisting of monzogranitic to granodioritic intrusions. Postmineralization suites are also mainly felsic and include the Eocene monzogranitic Sam Ross Creek phase. Included within the batholith are low-, medium-, and high-K intrusive phases, all of metaluminous to sightly peraluminous composition. On extended element normalized plots, all suites exhibit negative Nb and Ti anomalies, but differ in their LILE and HFSE contents and the magnitude and direction of Ba, Sr, and Eu anomalies. Geochemistry indicates that some François Lake suite phases–subphases (aplitic Nithi, Tatin Lake, pre-ore dykes, and Casey) and the Sam Ross Creek phase are the most evolved–fractionated. All suites–phases exhibit positive εNd(T) values (+1.1 to +7.2), indicating derivation mainly from juvenile source materials with a history of LREE depletion.
Most plutonic suites making up the Endako batholith can be classified as I types, granitoids derived from infracrustal source materials, that have volcanic-arc type granite trace element characteristics. The Sam Ross Creek represents an A-type granite with within-plate type granite features. Premineralization suites were likely generated within a primitive to mature arc-type setting via mixing between depleted mantle derived and juvenile-arc crustal materials. The François Lake suite was probably generated via remelting or recycling of juvenile-arc crust without major depleted mantle derived contributions.
The Endako deposit has been classified as a calc-alkaline, low-F, granodiorite-type molybdenum deposit (Westra and Keith 1981; Mutschler et al. 1981), a designation which is supported by the extensive data collected in this study from the cogenetic François Lake suite. The Sam Ross Creek phase exhibits features of intrusions cogenetic with graniteor Climax-type deposits. Along with being highly geochemically evolved (e.g., high Rb/Sr, low Ba and Sr), phases of the François Lake suite and the Sam Ross Creek phase are also moderately to highly oxidized, a combination of features considered characteristic of Mo-deposit related granitoids (Keith et al. 1993; Blevin et al. 1996). A lack of associated mafic intrusive phases and an extended magmatic history consisting of multiple pulses of fractionated and isotopically juvenile magmas are also considered as features that facilitated formation of economic Mo concentrations within the François Lake suite.
Combined Re–Os molybdenum dating (Selby and Creaser 2000, 2001), U–Pb zircon and 40Ar/39Ar biotite–amphibole dating (Villeneuve et al. 2001), and granitoid geochemistry have established a clear temporal link and very likely genetic link between various François Lake suite phases and three distinct Mo-deposition events within the Endako camp. These are (i) Nithi Mountain Mo deposition at ca. 154 Ma and the aplitic Nithi phase; (ii) Endako deposit first stage ribbon vein Mo deposition at ca. 148 Ma and late fractionates of the Endako phase, as represented by the older pre-ore dykes; and (iii) Endako deposit second-stage ribbon vein Mo deposition at ca. 145 Ma and late pre-ore dykes plus the Casey phase. The third event may have occurred above (>5 km) a magma chamber from which both pre- and postmineralization injections of Casey-type composition were derived. The Eocene Sam Ross Creek phase may have provided the thermal flux to partially reset K–Ar ages in the Endako deposit area (Villeneuve et al. 2001) and could also have been responsible for the late kaolinite alteration event within the deposit. The Climax-type features of this intrusion and the already established episodic nature of magmatism plus Mo deposition with the Endako camp makes this Eocene suite of intrusions highly favourable for being cogenetic with a fourth Mo-deposition event in the area.
Glen Johnson of Endako Mines, currently with Thompson Creek Mines, generously shared with us his time and experience with the geology of and around the Endako mine. He also supplied us with one of our pre-ore dyke samples. Contribution to the project of exploration geological mapping from Placer Dome Incorporated has been instrumental in providing substantially more detail to our work than would have been possible otherwise. J.B.W. benefited from instruction in Nd isotopic laboratory techniques from R. Theriault of the Geochronology Subdivision at the Geological Survey of Canada. Constructive reviews by Dave Sinclair, Dave Selby, and Sandra Barr are gratefully acknowledged.
Appendix: Analytical techniques
Whole-rock major elements were analyzed on fused glass discs by X-ray fluorescence spectroscopy (XRF) at the Geological Survey of Canada (GSC-Ottawa). FeO was determined by dichromate titration and F contents by ion electrode at the GSC. Trace element data were obtained using a combination of techniques including XRF spectrometry on pressed powder pellets (Memorial University of Newfoundland (MUN), St. John’s, Nfld.) and inductively coupled plasma – mass spectroscopy (ICP–MS; MUN and GSC). As duplicate analyses of most samples for a range of trace elements at the GSC and MUN did not indicate, within the inherent precision of the different techniques, any interlaboratory biases, most such duplicate results were averaged. Sm/Nd isotopic separations were carried out at the GSC by the senior author. Sample powders, spiked with mixed 148Nd–149Sm and 84Sr–87Rb tracer solutions, were dissolved in a concentrated HF–HNO3 mixture. Separation of REEs was done by standard cation exchange chromatography. Separation of Sm and Nd from other REEs followed HDEHP (Di (2-ethylexyl) orthophosphoric acid) – teflon powder chromatography. Total procedure blanks were approximately 0.2 ng for Nd and Sm. Mass analysis was carried out on a MAT-261 solid source mass spectrometer in static multicollection mode for Nd and Sm. Nd isotopic compositions were normalized to 146Nd/144Nd = 0.7219. Repeated measurements of an AMES Nd solution yielded 143Nd/144Nd = 0.512194 ± 22 (2 sd). All 143Nd/144Nd ratios were corrected to LaJolla 143Nd/144Nd = 0.511860. 147Sm/144Nd are reproducible to 0.5%. In addition to reporting measured 143Nd/144Nd ratios in Table 2, Nd isotopic data are reported as epsilon values (Nd), which measures the deviation in 143Nd/144Nd between a given sample and the chondritic uniform reservoir (CHUR) at the time chosen (see DePaolo 1988 for a complete review). Depleted-mantle Nd model ages (TDM) are based on the model of DePaolo (1988).
- Received January 31, 2000.
- Accepted August 11, 2000.
- Published on the NRC Research Press Web site at http://cjes.nrc.ca on May 7, 2001.
- © 2001 NRC Canada