We present new results and review existing data highlighting different aspects of the genetic relationship between partial melting and hydrothermal mineralizing processes in the contact aureole of the Sudbury Igneous Complex. At the basal contact of the igneous complex, in the footwall breccia, crystallized partial melt pods and veins, referred to as footwall granophyres are abundant and intrude all rock types including the breccia matrix, as well as massive Ni-Cu sulfide ore. The final crystallization of these melts was accompanied by the segregation of high salinity fluids dominantly in the temperature range of 450° to 550°C, as revealed by Ti-in-quartz thermometry and studies on primary fluid inclusions. In mineralized parts of the footwall breccia there is clear evidence that interaction of partial melts, exsolved fluids, and preexisting magmatic sulfides occurred. In the deeper footwall of the Sudbury Igneous Complex, similar footwall granophyres occur as networks of veins intruding impact brecciated (Sudbury breccia) country rocks and have an intimate spatial association with hydrothermal, low sulfide assemblages highly enriched in Pt, Pd, and Au (with a unique assemblage of PGM, including malyshevite and lisiguangite).

It is suggested that partial melting processes were widespread in the contact and proximal footwall environment of the igneous complex and they were important in providing high salinity magmatic fluids to a hydrothermal system responsible for redistribution of base and precious metals. We describe observations which may evidence the migration of partial melts from the contact into the deeper footwall. Furthermore, we suggest a cogenetic model for the precipitation of sulfide-hydrous silicate assemblages in close spatial association to footwall granophyres.

We emphasize that the use of footwall granophyres in mineral exploration for Cu-Ni-PGE ores in the footwall of the Sudbury Igneous Complex is clearly justified. Mapping of footwall granophyre vein networks highlights areas that were proximal to the contact environment and point out structural pathways that may have been used by syn- or postgenetic mineralizing fluids in a given area. Widespread partial melting appears to have occurred in the footwall of other large mafic-ultramafic complexes (such as Bushveld and Duluth) and similar genetic relationships of partial melting and ore-forming hydrothermal processes may have possibly existed also in these systems.

Crystallization of the superheated impact melt sheet within the 1.85 Ga Sudbury impact structure to form the layered sequence of the Sudbury Igneous Complex provided the heat source for various processes within its footwall that had a significant effect on formation of world-class Ni-Cu-PGE ores. These processes include the following: (1) an assimilation process that consumed several hundred meters of footwall and resulted in the formation of the so-called “contact sublayer,” one of the main hosts of contact Ni-Cu-PGE sulfides (Prevec et al., 2000; Prevec and Cawthorn, 2002), (2) the development of an 1- to 2-km-wide contact metamorphic aureole (Coats and Snajdr, 1984; Dressler, 1984; Hanley and Mungall, 2003; Boast and Spray, 2006), (3) localized partial melting of both mafic and felsic units up to several hundred meters below the Sudbury Igneous Complex/footwall contact (Péntek et al., 2009), (4) the formation of the so-called “footwall breccia,” another important host of contact deposits, through a combination of partial melting and thermal recrystallization (e.g., Coats and Snajdr, 1984; Lakomy, 1990; McCormick et al., 2002a), and (5) circulation of hydrothermal fluids, which had a great significance in the final distribution of metals especially in the footwall of the Sudbury Igneous Complex (e.g., Farrow and Watkinson, 1992; Li and Naldrett, 1993; Jago et al., 1994; Watkinson, 1999; Molnár et al., 1997, 1999, 2001; Farrow et al., 2005; Hanley et al., 2005; Péntek et al., 2008).

Although the exact role of hydrothermal fluids in the formation of footwall Cu-Ni-PGE occurrences up to 2 km away from the Sudbury Igneous Complex/footwall contact is debated and was certainly not the same within different deposits, there is a general agreement that they re-mobilized metals from magmatic ores (see reviews of this topic by Watkinson, 1999; Molnár et al., 2001; Molnár and Watkinson, 2001; Farrow and Lightfoot, 2002; Hanley et al., 2005; Ames and Farrow, 2007; Péntek et al., 2008). In contrast to the formation of typical “sharp-walled” Cu-rich massive sulfide veins, the hydrothermal origin of low sulfide-type footwall occurrences is generally accepted (Farrow et al., 2005; Péntek et al., 2008; Lesher et al., 2009; Tuba et al., 2010; White et al., 2011). These mineralized zones are exceptionally enriched in Pt, Pd, and Au (up to tens of ppm), despite consisting only of disseminations, and stringers of sulfides accompanied by hydrous silicate assemblages. A few authors reported signs of hydrothermal activity within mineralized parts of the sublayer and footwall breccia, which are important evidence for a model suggesting the remobilization of metals from the contact environment into the deeper footwall (Farrow and Watkinson, 1996; McCormick and McDonald, 1999; Molnár et al., 2001; McCormick et al., 2002b). The source of fluids that interacted with contact sulfides, as well as those involved in formation of footwall Cu-Ni-PGE occurrences, still remains speculative (e.g., Marshall et al., 1999). Molnár et al. (2001) described a fluid exsolution process from granitic veins, derived from partial melting within mineralized footwall breccia. They suggested that magmatic brines segregating from these melts may have been major components of the hydrothermal system in the footwall of the Sudbury Igneous Complex.

In this paper we present new results from two footwall Cu-Ni-PGE occurrences (Wisner South zone and Frost Amy Lake zone) as part of an ongoing effort to unravel the spatial and genetic relationships of partial melting, melt segregation, and hydrothermal mineralizing processes (Molnár et al., 2001; Péntek et al., 2009; Hanley et al., 2011). Reviewing recent results and incorporating this new data, we demonstrate that coeval migration of felsic partial melts and mineralizing hydrothermal fluids into the footwall occurred in certain areas and suggest a model describing the role of partial melting in providing fluids to the hydrothermal system in the contact aureole of the Sudbury Igneous Complex.

The Sudbury Igneous Complex is a 30 × 60 km, ~2.5-km-thick, elliptical body that crystallized from the melt sheet derived by fusion of impacted target rocks within the 1.85 Ga Sudbury impact structure. The superheated impact melt sheet underwent magmatic differentiation to form the layered sequence of the Sudbury Igneous Complex, which comprises a basal norite overlain by quartz gabbro and granophyre (Ivanov and Deutsch, 1999; Prevec and Cawthorn, 2002; Therriault et al., 2002; Zieg and Marsh, 2005). The shock metamorphosed and impact brecciated footwall of the Sudbury Igneous Complex is dominated by the Levack Gneiss Complex and the Cartier batholith of the Archean Superior province in the North and East Ranges (Fig. 1). The orthogneisses of the Levack Gneiss Complex yield protolith ages of 2.71 Ga, whereas granulite facies metamorphism of the complex was dated at 2.65 Ga (Krogh et al., 1984; James et al., 1991). The Cartier batholith is believed to have originated by partial melting accompanying this metamorphism (Meldrum et al., 1997) and is composed of granodiorite to monzogranite. The gneissic and granitoid rocks were later intruded by mafic units of the East Bull Lake suite (2.49–2.47 Ga: Krogh et al., 1984), the Matachewan-type dike swarm (2.47–2.45 Ga: Heaman, 1997), and the 2.2 Ga Nipissing suite (Corfu and Andrews, 1986). In the Sudbury impact event, the footwall rocks underwent brecciation, forming pseudotachylite referred to as Sudbury breccia, occurring up to at least 80-km distance from the base of the Sudbury Igneous Complex. These millimeter-wide veins to hundreds of meters wide zones dominantly contain fragments of the local footwall in a fine-grained, recrystallized matrix. Wide zones of Sudbury breccia were important conduits of Sudbury Igneous Complex-related magmatic and hydro-thermal mineralizing processes (Rousell et al., 2003).

Recent modeling suggests that the cooling impact melt sheet consumed several hundred meters of its footwall (Prevec and Cawthorn, 2002) through an assimilation process that was aided by the impact-brecciated nature of the basement (Boast and Spray, 2006). The contact sublayer, which occurs at the base of the Sudbury Igneous Complex, is interpreted as a mixing zone of this downward eroding melt system (e.g., Prevec et al., 2000; McCormick et al., 2002a). This unit is made up of a magmatic-textured matrix of noritic composition containing inclusions of adjacent footwall rocks and exotic mafic to ultramafic units (Naldrett et al., 1984; Lightfoot et al., 1997). The contact sublayer is underlain by the footwall breccia, which mostly contains locally derived clasts in an igneous to metamorphic matrix of dioritic to granitic composition (Dressler, 1984; Lakomy, 1990; McCormick et al., 2002a). World-class Ni-Cu-(PGE) ores are particularly prevalent in “embayments” (troughlike depressions of the Sudbury Igneous Complex/footwall contact), where these two units are thickened. The formation of these disseminated to massive sulfide deposits can be described by magmatic sulfide-liquid fractionation processes with enrichment of Cu and precious metals toward the footwall (Hawley, 1962; Kullerud et al., 1969; Naldrett, 1984; Ebel and Naldrett, 1996; Mungall et al., 2005). In contrast, the origin of Cu-Ni-PGE ores occurring up to 1 km away from the contact within the footwall has been debated. Magmatic sulfide fractionation and hydrothermal processes have been suggested for their formation with the significance of one or the other process being different from deposit to deposit (e.g., Farrow and Watkinson, 1992; Li et al., 1992; Li and Naldrett, 1993; Jago et al., 1994; Morrison et al., 1994; Farrow and Watkinson, 1996; Watkinson, 1999; Molnár et al., 2001; Farrow and Lightfoot, 2002; Hanley et al., 2005; Ames and Farrow, 2007; Péntek et al., 2008; Lesher et al., 2009). Although most studies have focused on the vein-type footwall deposits consisting of swarms of sharp-walled chalcopyrite-rich sulfide veins, recently the so-called “low sulfide” style of mineralization has also received much attention (Farrow et al., 2005; Péntek et al., 2008; Tuba et al., 2010, White et al., 2011). These systems consist of sulfide stringers and disseminations (<5 modal % total sulfide), accompanied by hydrous silicate (epidote, actinolite, chlorite, etc.) assemblages exceptionally enriched in precious metals (Pt, Pd, Au) and are interpreted to have originated entirely through the remobilization of metals from primary magmatic ores by saline fluids. The first such system put into production (in 2005) was the McCreedy West PM deposit, which had typical grades of 0.3 wt % Ni, 1.0 wt % Cu, and over 5 g/t Pt + Pd + Au (Farrow et al., 2005).

Below the zone of complete assimilation by the cooling Sudbury Igneous Complex, several hundred meters of its contact aureole have undergone partial melting, as described in detail by Péntek et al. (2009). We documented in situ partial melting features and so-called “footwall granophyres” (Molnár et al., 2001), representing segregated and crystallized partial melts from numerous localities within the footwall along the North and East Ranges. Our observations confirmed the thermal modeling of Prevec and Cawthorn (2002) and revealed that partial melting of both felsic and mafic rocks was widespread within the contact aureole of the Sudbury Igneous Complex up to distances of at least 500 m from the contact. Earlier studies established a contact aureole of 1-to 2-km widths that can be divided into three metamorphic facies: pyroxene hornfels (up to 200–350 m), hornblende hornfels (up to 600–1,100 m), and albite-epidote hornfels (up to 1,200–2,000 m) in order of increasing distance from the Sudbury Igneous Complex (Dressler, 1984; Coats and Snajdr, 1984; Hanley and Mungall, 2003; Boast and Spray, 2006).

A genetic link between partial melting, melt segregation, and Cu-Ni-PGE mineralizing hydrothermal processes was first proposed by Molnár et al. (2001) and further evidence for this was provided by Péntek et al. (2009). Molnár et al. (2001) documented the mineralogy and fluid inclusion characteristics of footwall granophyres occurring in mineralized footwall breccia. The authors discovered primary fluid inclusions in quartz filled miarolitic cavities of footwall granophyres, which contained brines very similar to those that have been found to play a significant role in the origin of some major footwall Cu-Ni-PGE deposits. This evidence led the authors to propose that fluids exsolving from crystallizing Sudbury Igneous Complex-related partial melts may have had significance in formation of ore deposits in the footwall. Péntek et al. (2009) described similar structural orientation of footwall granophyres and hydrothermal sulfide veins deep within the footwall. Hanley et al. (2011) suggested that the emplacement of footwall granophyres at the Barnet occurrence created favorable sites for later deposition of sulfides.

The Sudbury structure underwent gravitationally driven crater modifications (Spray, 1997), as well as deformation due to the ongoing NW-SE–directed compressional Penokean orogeny (1.89–1.83 Ga: Sims et al., 1989) even before the complete cooling of the igneous complex. Further deformation of the structure, especially in the South Range, is believed to have occurred during later tectonic events: the 1.7 to 1.6 Ga Mazatzal (Bailey et al., 2004), the ~1.45 Ga Chieflakien (Szentpéteri, 2009) and the ~1.0 Ga Grenville orogenies. Important characteristics of the Sudbury structure that are the results of deformation include the following: (1) the present ellipsoidal shape of the Sudbury structure, (2) NW-trending strike-slip faults, which offset the Sudbury Igneous Complex on the North Range, (3) the present dip magnitude (30°–40° to the S-SE in the North Range and 70°–75° to the W-SW in the East Range) of the Sudbury Igneous Complex-footwall contact (Rousell, 1984; Cowan and Schwerdtner, 1994; Riller, 2005).

Geology of the studied low sulfide Cu-Ni-PGE occurrences in the footwall

The geology of the Wisner and Frost Lake areas was described in detail by Péntek et al. (2008, 2009) and will only be summarized here. The South zone low sulfide Cu-Ni-PGE occurrence in the Wisner area is situated in a SE-NW–trending (~120–300-m-wide), intensively brecciated zone and is roughly 500 m northwest of the Rapid River Ni-Cu-PGE deposit, the nearest known contact type ore in the area (Fig. 1A). The western trench at the South zone exposes brecciated, medium- to coarse-grained Cartier quartz monzonite to granodiorite, which contains inclusions of Levack gneiss and is cut by granitic pegmatites of the Cartier magmatism. Massive Sudbury breccia occurs as a SE-NW–trending zone in the northern part of the trench, whereas in the other parts the granitoid is only cut by centimeter-sized veins of Sudbury breccia trending either parallel or subperpendicular to the massive breccia zone.

The low sulfide Cu-Ni-PGE mineralization of the Amy Lake zone is located in the Amy Lake breccia belt of the Frost Lake (East Range), an up to 100-m-wide Sudbury breccia body parallel to the basal contact of the Sudbury Igneous Complex (Fig. 1B). This belt is also host to the Capre 3000 footwall deposit (Davis, 2007) and connects with the contact where the Capre contact deposit is located. The Amy Lake zone exposes massive, mostly matrix-supported Sudbury breccia, which contains a variety of clast types. Most common are fragments of fine-grained gabbroic rocks which are accompanied by Levack gneisses and migmatites, as well as Cartier granitoids. The Sudbury breccia shows evidence of intense ductile deformation: centimeter-sized clasts are all sheared and strongly elongated in a NNE-SSW orientation.

Bulk rock samples were analyzed for major elements, as well as 50 base metals and trace elements (including Pt, Pd, Au), using inductively coupled plasma-mass spectrometry (ICP-MS) and inductively coupled plasma-atomic emission spectrometry (ICP-AES) by ALS Chemex Ltd. Details of these methods can be found on the ALS Chemex website (www.alsglobal.com).

Mineral chemical data were acquired using a Camebax MBX electron microprobe by wavelength dispersive analysis at Carleton University, Ottawa (Canada), at 15kV and 15 nA; counting times of 15 to 20 s or 40,000 counts were employed for each element except for F (40 s) and Ni (40 s).

Titanium concentrations in quartz were analyzed with Laser ablation ICP-MS method at the laboratory of the Hungarian Academy of Sciences, Institute for Isotopes (Budapest, Hungary) in order to better constrain formation temperatures of footwall granophyres. We used a double-focusing magnetic sector ICP mass spectrometer (ELEMENT2, Thermo Electron Corp., Bremen, Germany) equipped with a New Wave UP-213 laser ablation system. A beam diameter of 60 μm was used for pre-ablation in order to avoid contamination deriving from the surface of quartz grains. Analyses were carried out with a beam diameter of 20 and 60 μm. Where the size of the analyzed grains was large enough, ablation was carried out along single lines or with raster method in order to check the homogeneity of the material. NIST-612 synthetic glass standard was used for calibration using the values of Pearce et al. (1997) and the evaluation of this data was performed with the software Pepita by Dunkl et al. (2008).

Six element PGE (Pt, Pd, Os, Ir, Ru, Rh) concentrations were determined at Labtium Oy (Espoo, Finland), using an NiS fire preconcentration, Te-coprecipitation, and determination with ICP-MS. For details see the Labtium Oy website (www.labtium.fi/en). Detection limits of this analysis were 2 ppb (Ru), 1 ppb (Os, Rh, Pd), and 0.1 ppb (Ir and Pt).

The South zone low sulfide Cu-Ni-PGE occurrence (Wisner, North Range)

The footwall granophyres occur mainly as “rootless” (meaning that their source can not be established) veins and dikes a few mm to 10 cm in width (Fig. 2A) having a dominant strike of NW-SE (~140°–320°) (Fig. 3). In all cases, they are aplitic, pink colored, and contain up to centimeter-sized miarolitic cavities. Some veins and dikes are relatively straight, but they tend to be curvilinear were they intrude Sudbury breccia, which reflects a transition between brittle and ductile conditions at emplacement. Mineralogy of footwall granophyre dikes is dominated by granophyric, micrographic, and micro poikilitic quartz-feldspar intergrowths, in which feldspar is generally almost pure albite (An00-06) and K-feldspar (Or95-100). In the groundmass these rocks contain acicular opaque inclusions made up of rutile, ilmenite, and titanite—similar to those described from the footwall granophyre of the contact environment (Péntek, 2009). Miarolitic cavities are filled by euhedral quartz and feldspar (of compositions similar to those in the intergrowths) accompanied by epidote and titanite. Additional details about footwall granophyre occurrences in the South zone are described by Péntek et al. (2009). The low sulfide Cu-Ni-PGE mineralization of the Wisner South zone area has been described and interpreted to be hydrothermal in origin (Péntek et al., 2008). It consists of disseminations and veinlets of sulfides (+ hydrous silicates); however, PGE-rich sulfide assemblages occur also in close spatial relationship to the footwall granophyre veins and dikes. We distinguished the following: (1) sulfide-epidote patches within footwall granophyre (sample S-803 in Table 1), (2) sulfide-epidote-actinolite veinlets in the immediate continuation of footwall granophyre (sample S-804 in Table 1), (3) a magnetite-epidote-sulfide vein in the immediate continuation of footwall granophyre (sample SZE-06 in Table 1), and (4) massive sulfide veins in the continuation of footwall granophyre.

The first two types are described together because of their similar mineralogy and mineral chemistry. Sulfide-epidote patches are common in footwall granophyre, occurring always in their central parts and within miarolitic cavities, and are absent from the outer parts and surrounding host rocks (Figs. 2C, 3D, 4). These up to 3-cm patches replace the quartz-feldspar groundmass of footwall granophyre. Sulfides (chalcopyrite and millerite) are either rimmed by epidote (XEp = 0.88-1.19) or are in contact with the groundmass. The hydrothermal veinlets in the immediate continuation of footwall granophyre are characterized by the same minerals, with the addition of intergrown euhedral actinolite. Two important characteristics of this actinolite are the consistently high NiO contents of up to 0.85 wt % and elevated F contents of up to 0.36 wt %. The veinlets generally occur in the northwestern extension of footwall granophyre and terminate within 1 m of the gradational, replacive transition (Fig. 3B). Platinum-group minerals (PGM) and other trace metal minerals are very abundant in both assemblages, especially along the contact of chalcopyrite with epidote (included in both minerals) or the quartz-feldspar groundmass of footwall granophyre. They mostly occur as composite, isometric, or irregular grains up to 120 μm in diameter.

A unique magnetite-epidote-sulfide vein (SZE-06) occurs in the immediate continuation of a set of several parallel-running, up to 1-cm-wide partial melt veinlets (Fig. 3B, Table 1). From the gradational transition from the footwall granophyre veinlets, this 2-cm-wide hydrothermal vein terminates after continuing approximately 50 cm into the footwall. It dominantly consists of magnetite pseudomorphs after platy hematite (mushketovite) up to 1 cm in length containing inclusions of chalcopyrite, PGM, and associated trace metal minerals. Euhedral epidote with strong chemical zonation (XEp = 0.77–1.24) is intergrown with this magnetite, and both are corroded by later chalcopyrite and millerite. A second generation of euhedral magnetite occurs as inclusions in the sulfide minerals accompanied by a precious mineral assemblage similar to that in the platy magnetite. Small amounts of primary euhedral pyrite and secondary polydymite are also part of the assemblage, whereas actinolite and chlorite are concentrated along the margins of the vein.

Massive sulfide veins up to 5 cm in width occur in the central part of the western trench in close relationship with footwall granophyres (Fig. 3C). Although a transition like that described for the veins above could not be observed, the sulfide veins were emplaced with the same NW-SE–trending orientation within 1 to 2 m of footwall granophyres. The sulfide veins have a groundmass of chalcopyrite that contains up to centimeter-sized inclusions of octahedral pyrite and an epidote-rich alteration selvage with disseminated chalcopyrite.

Amy Lake zone low sulfide footwall Cu-Ni-PGE occurrence (Frost, East Range)

A close spatial relationship of PGE-bearing assemblages and footwall granophyres also occurs in the Amy Lake zone (Fig. 1B). Footwall granophyres of the Southern trench (Péntek et al., 2009) occur as irregular bodies and pods in the matrix of the breccia, or as “flame-like” intrusions in mafic clasts (Fig. 5; samples 804274 and 804278 in Table 1); they do not form straight veins or dikes. The footwall granophyre bodies are extremely sheared and some are cut by the Sudbury breccia matrix, which indicates that the matrix was ductile at the time of their emplacement. The footwall granophyres are texturally and mineralogically much more diverse than those from Wisner, as they are coarse-grained and usually made up of different quartz-feldspar textures (graphic, poikilitic, euhedral) on a macroscopic scale. Some are also rich in up to centimeter-sized euhedral crystals of amphibole (Fig. 6A, B). Miarolitic cavities occur in all footwall granophyre bodies and are larger in diameter (up to 5 cm) than those from Wisner. They are dominantly filled by euhedral quartz, feldspar, epidote, and amphibole accompanied by varying proportions of titanite, zircon, and apatite.

Precious metal-bearing low-sulfide assemblages occur in close spatial relationship to footwall granophyres either in the form of (1) sulfide disseminations and patches within footwall granophyre, especially in their miarolitic cavities (samples 804273 and 804280 in Table 1), or as (2) hydrothermal veins in the immediate continuation of footwall granophyre (samples 804275 and 502694 in Table 1). Chalcopyrite, millerite, and minor galena within the melt bodies form individual grains of several tens of μm included in quartz, feldspar, and amphibole of miarolitic cavities, or as centimeter-sized replacive patches corroding the same minerals (Fig. 6A, C). Sulfide patches occur again in the central parts of the melt bodies and are completely absent from the immediate host rocks. PGM are individual isometric grains up to 20 μm in diameter.

A gradual transition from footwall granophyre veins to hydrothermal veins was also observed at this locality, best exemplified by samples 804273 to 804275 (Figs. 5A, 6B, D; Table 1). In this case, an offshoot from a footwall granophyre pod (containing both feldspar- and amphibole-rich parts) terminates after roughly 2 m. The proportion of granophyric quartz-feldspar gradually decreases until the vein is entirely composed of hydrothermal amphibole, quartz, and albite. Biotite flakes (up to 1 mm) are also present along the contact of the vein with the gabbroic host rock. Other hydrothermal veins graduating from footwall granophyre (like sample 502694) contain abundant coarse-grained, euhedral epidote. Amphiboles in the sulfide-bearing footwall granophyres are extremely zoned having cores of magnesio-hornblende, or ferrimagnesio-hornblende followed by compositions ranging from actinolitic hornblende to ferro-actinolite, actinolite, or ferri-tremolite (according to the amphibole classification by Leake et al., 1997). Hydrothermal veins graduating from footwall granophyre pods and veins are dominantly composed of actinolite with occasional actinolitic hornblende cores. An important characteristic of these amphiboles is their high NiO contents (up to 0.5 wt %). Halogen contents are usually low, or slightly elevated (up to 0.28 wt % F and 0.26 wt % Cl).

Platinum-group and silver minerals in sulfide assemblages related to footwall granophyres

At the Wisner South zone, sulfide-bearing assemblages associated with footwall granophyres host an assemblage of numerous PGM and associated silver minerals. Most abundant are grains of malyshevite (CuPdBiS3) and lisiguangite (CuPtBiS3), which usually occur as zoned grains with Pt and Pd proportions ranging between the two minerals (Fig. 7; Table 2). To our knowledge, this is the first description of a chemical continuity between malyshevite and lisiguangite (despite their different crystal structures). These PGM generally occur as irregular grains, either individually in chalcopyrite or as rims around other PGM (Fig. 8C-F). The mückeite component (CuNiBiS3) is usually low, but Pd is substituted by up to 2.8 wt % Ni in a few grains.

Other PGM are dominantly tellurides, with compositions most abundant between merenskyite (PdTe2) and moncheite (PtTe2). They usually have atomic Pt/(Pt+Pd) between 0.41 and 0.82, but a few grains also plot into the merenskyite apex, or slightly toward the melonite (NiTe2) endmember. These compositions overlap with those described from disseminated sulfide assemblages not related to footwall granophyres (Péntek et al., 2008). Tellurium is substituted by 10 to 16 wt % Bi in all grains, apart from those that are the most Pd rich and which have much lower or higher Bi contents (4–5 and 25 wt %, respectively). Kotulskite (PdTe) and rare sobolevskite (PdBi) have 0.39 to 0.88 at. % Te (Table 2). These PGM occur as individual grains or intergrown with or rimmed by malyshevite, lisiguangite, and silver minerals (Fig. 8C-F).

Two Ag-Pd tellurides were also found in these samples, but their atomic proportions do not correspond to any known PGM. Among Ag minerals, most common are grains with compositions along the bohdanowiczite (AgBiSe2) and matildite (AgBiS2) join. Minerals of the acantite-empressite series also occur, including hessite and compositions close to aguilarite (Ag4SeS) and cervelleite (Ag4TeS), as well as Ag4TeSe. A few grains appear to represent the sulfosalt berryite (Pb3(Ag,Cu)5 Bi7S16), with 7 to 10 wt % Se substituting for S.

Chalcopyrite in the miarolitic cavity of a melt body (804280) from Frost Amy Lake zone hosts mostly individual grains of merenskyite (PdTe2) (Fig. 8G), which are completely devoid of Pt, but contain 20 to 21 wt % Bi substituting for Te. These were accompanied by grains of michenerite (PdBiTe), kotulskite (PdTe) and temagamite (Pd3HgTe3) (Table 2). Kotulskite is Ag-, Hg- and Bi-bearing, whereas temagamite contains significant amounts of Ag and Bi.

Formation temperature of footwall granophyre on the basis of the titanium-in-quartz thermometry

The nearly endmember compositions of K-feldspar and albite, and the actinolitic character of amphiboles in most of the footwall granophyre samples rules out the use of conventional thermometric methods usually applied to granitic rocks (e.g., amphibole-plagioclase and ternary feldspar solution thermometry) (Péntek, 2010; Péntek et al., 2009). The temperature of the partial melting process may be estimated using the zircon saturation thermometry of Watson and Harris (1983): resulting temperatures are 727 ± 21°C for Wisner South and 757 ± 29°C for the Amy Lake zone (Péntek et al., 2009). Hornblende and plagioclase crystallization in some footwall granophyres from the Craig deposit occurred at ~750°C (Molnár et al., 2001).

We used the titanium-in-quartz thermometer (Wark and Watson, 2006) to determine the crystallization temperature of footwall granophyres. This method is based on the Ti concentration in quartz crystallized in equilibrium with rutile (aTiO2 = 1), although it can also be applied to rutile-absent systems, such that assuming aTiO2= 1 will result in a minimum temperature of crystallization. In the presence of rutile, the temperature estimates with this method are usually within ±5°C, whereas for rutile-absent rocks a ± 0.2 error in aTiO2 estimation will still yield temperatures off by no more than 50°C (Wark and Watson, 2006). Recent experimental work revealed that there is a significant pressure effect on the incorporation of titanium into quartz, and thus a new equation was suggested, which takes this effect into consideration (Thomas et al., 2010).

Since the footwall granophyres from Wisner South zone contain rutile, ilmenite, and titanite, these rocks have aTiO2 = 1. All the other samples contain titanite and/or ilmenite; for these we can assume aTiO2 between 0.5 and 1 and the crystallization temperatures lie between those calculated for these two values. The pressure in the contact and proximal footwall environment has been estimated at 1.5 ± 0.5 kbar (Molnár et al., 2001; Péntek et al., 2009), in accordance with suggested depth of erosion in the Sudbury structure (Dressler, 1984). Hence, we calculated Ti-in-quartz temperatures for aTiO2 = 1 and 0.5 using pressures of 1, 1.5 and 2 kbars. In Table 3 we report average temperature results for both aTiO2 values at 1.5 kbars, with the effects of changing pressure and the error in Ti analysis calculated into ± temperatures; a simplified histogram is shown in Figure 9. As a comparison, in this figure and Table 3 we also show results of footwall granophyre samples from two localities right along the Sudbury Igneous Complex-footwall contact (from footwall breccia at Craig and Wisner, Rapid River deposit) (Fig. 1).

Quartz in footwall granophyre veins associated with sulfides from Wisner South zone yields crystallization temperatures from 576° ± 26° down to 439° ± 15°C. Because of the very fine grained vermicular intergrowth of quartz and feldspar in the first crystallized granophyric parts, we could not obtain the highest crystallization temperatures of footwall granophyre. However, in accordance with relative timing suggested by petrographic observations, micrographic and subhedral (micropoikilitic) intergrowths consistently show higher temperatures than coarse-grained subhedral to euhedral grains in miarolites (Fig. 10). In the miarolitic cavities of most footwall granophyre samples, the concentration of Ti was below the detection limit; thus, only a maximum temperature of crystallization could be obtained. The analysis of such miarolitic quartz crystals was in some instances further complicated by a completely inhomogeneous distribution of Ti. Thermometry on quartz in the footwall granophyre from Frost gives final crystallization temperatures between 404° and 505°C, whereas the amphibole-granophyre vein yields average temperatures of 435° to 485°C.

The Ti-in-quartz thermometry gives a well-defined range of final quartz crystallization temperatures in footwall granophyres (approx. 450°–550°C), which is in good agreement with the presence of actinolite, epidote, and titanite, as well as almost pure endmember feldspar crystals intergrown with quartz in the miarolitic cavities. Primary fluid inclusions in the same quartz preserve the fluid segregating from the crystallizing partial melts and generally yield total homogenization temperatures by dissolution of halite or other unknown daughter minerals in the same 450° to 550°C temperature range (Molnár et al., 2001; Péntek, 2009). Recently, Hanley et al. (2011) also used the p-sensitive Ti-in-quartz thermometer to calculate quartz crystallization temperatures in the granophyric portions of two footwall granophyre samples. In accordance with relative timing of crystallization, these granophyric portions show higher crystallization temperatures (617° ± 18° and 643° ± 26°C) than the final quartz crystallization temperatures obtained by our study.

Footwall granophyres in the basal contact units of the Sudbury Igneous Complex

Footwall granophyres are abundant within the basal contact units of the Sudbury Igneous Complex (footwall breccia and sublayer) and postdate the emplacement of magmatic Ni-Cu-PGE sulfides (Molnár et al., 2001; Péntek et al., 2009; Hanley et al., 2011). They have been interpreted to represent segregated and crystallized partial melts that formed by melting of footwall rocks due to the contact heat of the crystallizing Sudbury Igneous Complex melt sheet. However, Hanley et al. (2011) proposed that they may also be injected silicate residues from the melt sheet itself.

Recent models suggest that cooling of the superheated impact melt sheet induced assimilation of up to 800 m of footwall (Prevec and Cawthorn, 2002). As such, the sublayer and the upper part of the footwall breccia may be interpreted as a mixing zone and thus a transition exists between the basal Sudbury Igneous Complex and the brecciated footwall (Prevec et al., 2000; McCormick et al., 2002a). Several studies suggest that in the upper part of the footwall breccia the matrix was melted, whereas the matrix in the lower parts was thermally recrystallized and partially melted (e.g., Lakomy, 1990; McCormick et al., 2002a). The change from igneous-textured to metamorphic-textured breccia matrix, as well as the change from dioritic to granitic composition of the matrix toward the footwall, is in accordance with this model. The emplacement of footwall breccia dikes into the overlying main mass norite, as well as the underlying footwall, was described by some authors (Dressler, 1984; Lakomy, 1990, McCormick et al., 2002a) and these may be regarded as segregated melts derived by partial melting within the footwall breccia.

In our study of Sudbury Igneous Complex-related partial melting, we have shown that the bulk rock composition of such footwall breccia dikes is very similar to footwall granophyres formed in the deeper footwall, with lower SiO2 contents as the main difference (Péntek et al., 2009). This is in accordance with partial melting occurring only within more felsic rocks as temperature decreases away from the Sudbury Igneous Complex. The higher temperature of partial melting within the contact environment is also demonstrated by the much more mafic character (e.g., abundant pyroxene) of footwall granophyre in sublayer and footwall breccia in comparison to those occurring deeper into the footwall (Péntek, 2009; Péntek et al., 2009). Even felsic footwall granophyres in footwall breccia show characteristics of a higher melting temperature: the higher concentrations of Zr and P are good indicators because their solubility increases with increasing temperature and decreasing SiO2 contents (e.g., Watson and Harrison, 1983; Harrison and Watson, 1984). The thermometer based on the solubility of Zr yields partial melting temperatures of 914°C for a footwall granophyre sample from footwall breccia in contrast to 727° ± 21°C for the more distal Wisner South zone footwall granophyre samples (Péntek, 2009).

Among the various footwall granophyre types described from footwall breccia occurrences by Péntek (2009), one abundant variety is especially important in the interpretation of observations described in this paper. These pink, granophyric footwall veins within the footwall breccia have modal compositions (and mineral chemistry) very similar to the footwall granophyres observed in the deeper footwall at Wisner South zone. Apart from very rare biotite and epidote filling miarolitic cavities, they are devoid of mafic minerals and are made up almost entirely of quartz-feldspar inter-growths. Rutile, ilmenite, and titanite are important constituents of granophyric melt bodies from both environments. We may assume that these footwall granophyres in the footwall breccia represent the following: (1) partial melts formed by the melting of more felsic rocks, either simultaneously with the more mafic footwall granophyre types, or at a later, lower temperature stage, formed by wet melting; or (2) late, fractionated melts of originally less felsic partial melts. The latter interpretation may be indicated by felsic, K-rich melt inclusions, which occur in a mafic footwall granophyre (Péntek, 2009). Whichever is the case, the resemblance of these footwall granophyres to those occurring in the deeper footwall suggests that such melts could have been the material that segregated and intruded into the deeper footwall to form the seemingly rootless footwall granophyre vein systems exposed at Wisner South zone.

Exsolution of fluids from crystallizing partial melts and interaction with magmatic sulfides

Several studies have recognized that halogen-rich fluids may have played an important role in the evolution of not only the footwall Cu-Ni-PGE mineralization but also the contact magmatic Ni-Cu ores. Farrow and Watkinson (1996) described pervasive hydrothermal alteration assemblages and related saline aqueous fluids from the footwall breccia of the Epidote zone (Fraser mine). McCormick et al. (2002b) revealed chlorine geochemical halos within mineralized parts of the footwall breccia and sublayer (Strathcona embayment), whereas McCormick and McDonald (1999) and Péntek (2009) reported different generations of amphibole from mineralized footwall breccia, including F- and Cl-rich varieties. Some of these hydrous silicates appear to be syngenetic with the sulfide assemblages, whereas others are clearly replacive.

Although the ductile vs. brittle behavior of rocks depends on numerous factors (Fournier, 1999), at the high temperatures associated with partial melting it is reasonable to assume that the footwall breccia and significant parts of the footwall would deform in a ductile manner. The Sudbury breccia zones would have been ductile during partial melting, whereas stockwork of footwall granophyre in granitic and gneissic units suggests that these rocks were able to deform in a brittle way if strain rates were sufficiently high and melt pressure temporarily exceeded lithostatic pressure (Péntek et al., 2009). Under ductile conditions, large amounts of external fluids (i.e., heated formational brines of the basement) may not have entered the contact environment, as the permeability of such a ductile zone would be very low. Thus, at least in the early stages of footwall breccia evolution, fluids derived locally by exsolution from crystallizing partial melts may have been of major importance in the interaction with magmatic sulfides. Additional fluids may have segregated from the sulfide melts themselves, as proposed by some authors (McCormick et al., 2002a; Hanley et al., 2005).

Footwall granophyres in general show evidence that a fluid phase exsolved during their crystallization (e.g., miarolitic cavities, primary fluid inclusions). Abundant F-rich apatite (F/(F + Cl)= 0.80–0.91) and amphibole (1.0 to 1.8 wt % F) in the footwall granophyre (Molnár et al., 2001; Péntek, 2009) suggest that this exsolved fluid phase was saline. It is well established that fluorapatites in (both mafic and felsic) magmatic systems indicate the segregation of a Cl-rich fluid from the crystallizing melt (Boudreau and McCallum, 1989; Piccoli and Candela, 1994). The Ti-in-quartz thermometer suggests that final quartz crystallization in the presence of an exsolved fluid phase has taken place at temperatures of approx. 450° to 550°C, in perfect accordance with the usual range of total homogenization temperatures of primary high salinity (~50 wt % NaCl equiv) fluid inclusions within quartz crystals of such footwall granophyre (Molnár et al., 2001; Péntek, 2009; Hanley et al., 2011). Although individual partial melt veins are in most cases relatively small, the partial melting process was so widespread that fluids exsolving from these melts were of significant volume. Similar fluorapatite and F-rich amphibole were also described from the footwall breccia matrix (McCormick and McDonald, 1999), which suggests that exsolution of Cl-rich fluids may have occurred from the crystallization of this breccia matrix as well.

We suggest that due to cooling of the system, ductile conditions where progressively replaced by brittle behavior, which enabled the influx of fluids derived from the deeper footwall. These fluids from different sources (e.g., fluids segregated from partial melts in the footwall, heated basinal brines, metamorphic fluids) were able to percolate through the contact environment, causing alteration and further interaction with sulfides.

The presence of partial melts and magmatic fluids in the ductile stage and the presence of different fluids in the brittle stage must be taken into consideration in the evolution of contact magmatic sulfides. Evidence for the interaction between the partial melts, hydrothermal fluids, and the sulfides (Molnár et al., 2001; Péntek, 2009) supports the idea that fluids were responsible for redistributing metals within the contact environment and into the footwall (e.g., Farrow and Watkinson, 1996; Molnár et al., 1997, 1999; Watkinson, 1999). The solubility of base and precious metals (including Pt and Pd) in hydrothermal fluids and their preferred transportation as chloride complexes is generally well accepted and will not be discussed further here (see reviews of this topic in relation to Sudbury fluids by Farrow and Watkinson, 1996; Molnár et al., 2001; Hanley, 2005; Péntek et al., 2008). Hydrothermal veins containing Pt and Pd in the ppm range have been described from several contact orebodies and such veins were found to host primary fluid inclusions of high salinity brines (e.g., Molnár et al., 1999, 2001). This demonstrates the capability of these brines to transport significant concentrations of metals released through the interaction with preexisting sulfides.

Partial melting and subsequent exsolution of saline fluids from the crystallizing melts was widespread in the footwall breccia and some parts of the deeper footwall (Péntek et al., 2009); thus, the interaction of these melts and fluids with magmatic sulfides may, in our opinion, have been of significant enough volume to play an important role in redistribution of metals and thus formation of low-sulfide mineralization.

Injection of partial melts and mineralizing hydrothermal fluids into the deeper footwall of the Sudbury Igneous Complex

The examples of the close spatial association of footwall granophyre and low sulfide Cu-Ni-PGE in two footwall environments documented in this study provide important evidence for a genetic relationship between these two processes. Any genetic model explaining the association of the sulfides and the footwall granophyre must take into account the following important textural characteristics: (1) the sulfides occur either with or without associated hydrous silicates, enclosed by or in continuation of footwall granophyre, having a gradational replacive contact with them, (2) there is no sulfide precipitation and significant hydrothermal alteration within surrounding country rocks (at Wisner South zone), and (3) PGM and associated trace metal minerals preferentially occur along the margin of sulfide blebs and footwall granophyre groundmass or hydrous silicate rims.

At first, the most realistic scenario may be a hydrothermal alteration and replacement of the footwall granophyre by a later, metal-bearing fluid unrelated to the partial melt emplacement. Such an explanation was given by Hanley et al. (2011) for the spatial association of sulfides with footwall granophyre at the Barnet showing, a similar (but Ni- and PGE-poor) low sulfide occurrence in the footwall of the Levack-Onaping embayment. However, the fact that there are no hydrothermal veins, pervasive alteration zones, and/or sulfide precipitations in the immediate country rocks (up to distances of several meters) does make this explanation very problematic in our case. The preferential precipitation of sulfide patches and disseminations by a later fluid exclusively within or in continuation of footwall granophyre is also difficult to explain, considering that they do not significantly differ chemically from their host rocks. The replacive character of sulfides toward the footwall granophyre groundmass, the hydrous silicate rim around the sulfides, as well as the preferred occurrence of PGM along the edge of sulfide patches, rules out the possibility of another explanation: the sulfide blebs being “fragments” of preexisting sulfides transported by the intruding partial melt.

In our opinion, a cogenetic explanation may better explain the intimate spatial association of footwall granophyre and sulfide-hydrous silicate assemblages than the model of a later, unrelated fluid preferentially migrating along and replacing the footwall granophyre in a chemically similar host rock. There are two possibilities for a cogenetic model: the partial melt and the metal-bearing fluid were migrating contemporary or the metal-bearing fluid exsolved from the partial melts during emplacement. There is some evidence supporting a coeval existence of partial melt and hydrothermal fluid, suggesting that the final crystallization of partial melts may have occurred at much lower temperatures than it would be expected in granitic systems (Péntek et al., 2009): (1) the near endmember composition of feldspar (albite and orthoclase) in the granophyric-graphic groundmass of the footwall granophyre, (2) the low crystallization temperatures of quartz in the granophyric-graphic groundmass (up to approx. 580°–600°C in graphic: this study, approx. 600°–670°C in granophyric: Hanley et al., 2011), and (3) the presence of significant F and Li in the coexisting fluid (cf. several thousands of ppm Li in volatile phases present in ore zones: Hanley et al., 2005) lowering the solidus temperature of granitic melts. As we have seen from the geochemistry of apatite and amphibole, fluorine appears to have played a significant role in the crystallization history of the partial melts. As a comparison, at 1 kbar pressure, the solidus of the haplogranite-H2O-HF system was found to decrease to 510° to 500°C at fluorine contents of 2 to 4 wt % (Manning, 1981; Xiong et al., 1999). Although we have no information regarding the exact concentration of F in the crystallizing partial melts, fluorine enrichment of the melt due to continuous segregation of a Clrich fluid phase appears to have been significant and may have been responsible for extending the crystallization of these melts to low temperatures. Thus, in our opinion it is not unrealistic to assume a coeval migration of partial melts and mineralizing fluids. In this case, the crystallization of partial melt may have occurred at temperatures of approx. 680°, down to 550°C, followed by a hydrothermal stage (~550° to 450°C) in which crystallization of miarolitic cavities and precipitation of sulfide-hydrous silicate assemblages took place.

Alternatively, the metal-bearing hydrothermal fluid may have been exsolved from the partial melts after their emplacement. We have seen that exsolution of saline fluids accompanied the crystallization of partial melts; thus, theoretically, this would be a more plausible explanation. There is no clear evidence suggesting that partial melts may have had high base and precious metal contents that would enable the exsolution of such a metal-enriched fluid phase. Primary fluid inclusions in quartz of footwall granophyre were found to have low Cu concentrations (less than 500 ppm), although accidentally trapped chalcopyrite was observed in some inclusions (Hanley et al., 2011).

The fact that significant in situ partial melting was not observed in the South zone (Péntek et al., 2009), and the “rootless” character of the footwall granophyre veins suggest that these veins intruded into this zone. There is a significant volume of melt preserved by this vein system. The veins have a dominant strike of SE-NW (140° –320°), with the hydrothermal PGE-bearing assemblage always occurring in their northwestern extension. As the third dimension of the veins is unknown, the exact vector of melt migration cannot be established, but laterally it was from southeast toward northwest. Although a physical connection to the source is not exposed, it is reasonable to assume that the melt intruded from a hotter part of the nearby contact environment. Thus, we suggest that the partial melts and thus the metals contained in the mineralizing fluids exsolved from them (or alternatively, migrated together with them) likely originated from the contact zone exposed to the southeast (hosting the Rapid River and WD-16 Ni-Cu-PGE occurrences; Fig 1A).

We have seen that partial melts very similar to those described from the footwall of Wisner South zone formed in the footwall breccia and may in some instances have intruded the footwall (as footwall breccia dikes, described by Dressler, 1984; Lakomy, 1990; McCormick et al., 2002a). Since the permeability of the ductile footwall breccia and proximal footwall was probably too low for large-scale migration of melts and fluids between these zones, we suggest that the injection of the melt and fluid phases into the footwall may have occurred during sudden “brittle episodes,” possibly related to the tectonic reequilibration of the Sudbury structure through crater modifications. High strain rates in tectonically active environments are known to cause sudden brittle behavior of systems at magmatic temperatures and ductile conditions (Fournier, 1999). Such brittle episodes were probably short-lived and the system returned to a ductile condition immediately thereafter.

Farrow et al. (2005, p. 168) also described sulfide blebs associated with “pegmatitic and granophyric veins and patches of K-feldspar and quartz” from the McCreedy West PM deposit, and noted that these appear to have shared the same fluid pathways and may be contemporaneous with the sulfide mineralization. Although these authors do not explain the origin of these granophyric rocks, from their description it is evident that they refer to footwall granophyre. As mentioned earlier, Hanley et al. (2011) described footwall granophyres from the Barnet occurrence in association with Cu-Ni-PGE sulfide; however, in this case the sulfides were interpreted to replace the granophyric rocks. These examples demonstrate that such spatial association of PGE-bearing sulfides and crystallized partial melts may be widespread and not a local characteristics of our study areas.

We have presented and reviewed observations from footwall and contact occurrences at the Sudbury structure that highlight different aspects of the genetic relationship between partial melting and hydrothermal mineralization processes in the contact aureole of the Sudbury Igneous Complex. As revealed by our study focusing on the partial melting itself (Péntek et al., 2009), this process was widespread along the basal contact of the igneous complex up to several hundred meters into the footwall and crystallization of the derived melts was generally accompanied by the exsolution of high salinity fluids. Thus, although individual footwall granophyres appear to be small and insignificant at the local scale, they were in fact important in generating the high salinity magmatic fluids to the hydrothermal system responsible for the redistribution of magmatic sulfides. Below we suggest a model highlighting the importance of partial melting in the formation of low sulfide footwall Cu-Ni-PGE ores.

1. Cooling of the superheated impact melt sheet led to the assimilation of several hundred meters of footwall, the Contact sublayer and the upper part of the footwall breccia may be regarded as a mixing zone from this process (e.g., Prevec and Cawthorn, 2002; McCormick et al., 2002a) (Fig. 11A, B).

2. Below the zone of assimilation, partial melting occurred in the footwall breccia and up to several hundred meters into the footwall, with increased intensity in Sudbury breccia zones (Fig. 11C). The occurrence and intensity of partial melting did not change gradually away from the contact, but was mainly a function of rock composition as well as the presence or absence and composition of fluids. At the same time, zonation of contact metamorphic mineral assemblages developed (e.g., Dressler, 1984; Coats and Snajdr, 1984). Although the exact location of the brittle-ductile transition zone at this time can not be ascertained, according to analogues (Fournier, 1999) it may have been at a temperature of about 400°C, within hornblende hornfels facies conditions. Magmatic sulfides in the sublayer and footwall breccia fractionated toward the footwall and hanging wall.

3. Exsolution of high salinity fluids accompanied the crystallization of partial melts (Molnár et al., 2001; Péntek et al., 2009; this study). Because of the ductile condition of the contact and proximal footwall environment, there was no constant large-scale migration of melts and/or fluids between these zones. Significant amounts of external fluids (e.g., formational brines) could not enter the contact environment (partly also because of the extreme temperature gradient), and partial melts and fluids exsolving from the melts did not leave the closed system.

4. In mineralized parts of the footwall breccia, interaction of partial melts, exsolved fluids, and the preexisting magmatic sulfides took place (Molnár et al., 2001; Péntek, 2009). As a result of such reactions, sulfides were corroded and mobile metals were released into the fluids.

5. At sudden brittle episodes, probably related to tectonic crater modifications of the Sudbury structure, “offshoots” of large volumes of partial melt into the footwall occurred in certain areas affected by such tectonic activity (Fig. 11D). This process resulted in the emplacement of footwall granophyre vein systems and associated low sulfide Cu-Ni-PGE mineralization several hundred meters below the contact. We believe that the injection of partial melts and associated mineralizing fluids is not unique to the studied localities, but is a more common process not recognized so far (e.g., it appears to have occurred also in the McCreedy West PM deposit, Farrow et al., 2005). Alternatively, migration of partial melts and mineralizing fluids may have occurred in two separate stages, as suggested by Hanley et al. (2011), based on observations from the Barnet showing. Stockworks of rootless footwall granophyre veins have also been observed in areas (Péntek, 2009) without associated sulfides, which implies that such offshoots from the contact were only accompanied by sulfide precipitation, if the contact from which they originated contained magmatic sulfides to be remobilized. Emplacement of fractionated sulfide liquids into the footwall may also have occurred during sudden brittle episodes to form some of the “sharp-walled” vein systems. After these short-lived episodes, fluids exsolving from partial melts and from the sulfide veins themselves locally redistributed the metals (Fig. 11E). Although the system returned to ductile conditions, cooling of the Sudbury Igneous Complex resulted in gradual retreating of the brittle-ductile transition zone.

6. In a later stage, when the footwall and contact environments were already under brittle conditions, fluids from different sources circulated, causing further alteration of preexisting sulfides and redistribution of metals (Fig. 11F). The formation of hydrothermal vein stockworks cutting footwall breccia (e.g., Molnár et al., 1997, 1999, 2001; Péntek, 2009), and occurring in association with many footwall-style deposits (e.g., actinolite-, epidote-, quartz-rich veining described by Li and Naldrett, 1993; Watkinson, 1994; Molnár et al., 2001; Péntek et al., 2008; Hanley and Bray, 2009, Tuba et al., 2010) proceeded during this brittle stage. Some sharpwalled vein systems may have formed at this late stage of sulfide distribution, which would explain why some of these sulfide veins crosscut low sulfide mineralization or actinolite veining (Farrow et al., 2005; Hanley and Bray, 2009). We suggest that the emplacement of sharp-walled vein systems possibly occurred in multiple events and possibly by different processes (magmatic and hydrothermal), which may explain the controversy regarding their origin.

Although more research is needed to understand the genetic explanation for the spatial relationship of footwall granophyre veins and Cu-Ni-PGE sulfides in the deep footwall, the use of footwall granophyres in mineral exploration within the Sudbury Structure is clearly justified. Mapping the distribution of footwall granophyres and zones of in situ partial melting helps to outline areas that were located relatively close (within a few hundred meters) to the basal contact of the Sudbury Igneous Complex. This may help the recognition of footwall zones prospective for Cu-Ni-PGE mineralization even in tectonically modified parts of the aureole. It has to be emphasized, however, that partial melting was widespread within the contact aureole even in unmineralized areas, and thus the recognition of footwall granophyre does not necessarily imply the presence of nearby Cu-Ni-PGE ores. Given that footwall granophyre vein networks were emplaced along the same structures as Cu-Ni-PGE sulfides (either syn- or postgenetically), a structural analysis of these vein sets may point out pathways of mineralizing fluids in a given area.

In this study we emphasize the importance of partial melting processes of footwall units in adding significant volumes of saline fluids to a hydrothermal system driven by the heat of a cooling magmatic complex. Similar partial melting processes are known to have occurred in the footwall of other large mafic-ultramafic igneous bodies, but studies focusing on these processes and their possible genetic relationships to Ni-Cu-PGE sulfides are scarce. The best known contact aureole of a mafic body is that of the relatively small Ballachulish Complex in Scotland (e.g., papers in Voll et al., 1991; Holness and Clemens, 1999). Among large, world-class ultramafic-mafic complexes, only the partial melting processes in the footwall of the Bushveld Complex have been studied to some extent in the past decade (Harris et al., 2003; Johnson et al., 2003); to our knowledge, no genetic relationship of partial melting and hydrothermal ore-deposition has been recognized in this system yet. Partial melting is also widespread in the footwall of the Duluth Complex (Sawyer, 2002) and a close spatial association with Cu-Ni-PGE sulfides suggests that a genetic relationship between partial melting and ore-forming processes may have occurred also in this system (Molnár et al., 2011).

We gratefully acknowledge Wallbridge Mining Company Ltd. for the financial and logistical support of this study. We thank Robert Katona and Zsolt Stefánka (II-HAS) for assistance in LA-ICP-MS data acquisition, István Dunkl (University of Göttingen) for his help with the PEPITA software and Jay Thomas (Rensselaer Polytechnic Institute, Troy, NY) for discussions and for sending us parts of his, at that time, unpublished manuscript. Marco Fiorentini, an anonymus reviewer, and editors Larry Meinert and Andy Tomkins provided suggestions and corrections that significantly improved the quality of the manuscript. The paper also benefited from discussions with Doreen Ames and Jacob Hanley, as well as from comments by Nick Fenner and Bruce Jago, who read an earlier version of the manuscript. The work was funded by the Canada-Hungary Science and Technology Agreement Project CAN-02/04 to F.M. and D.H.W., the NSERC grants to D.H.W., as well as the Society of Economic Geologists Foundation Hugh E. McKinstry Student Research Awards to A.P. and Gy.T.