mesothermal quartz stuff - ansdell, kyser

Economic Geology Vol. 87, 1992, pp. 1496-1524

Mesothermal Gold Mineralization in a Proterozoic Greenstone Belt:

Western Flin Flon Domain, Saskatchewan, Canada

KEVIN M. ANSDELL AND T. KU•TIS KYSEB

Department of Geological Sciences, University of Saskatchewan, Saskatoon, Saskatchewan, Canada S7N OWO

Abstract

The Flin Flon domain, Trans-Hudson orogen, Canada, is an example of a Proterozoic greenstone belt that hosts a large number of small mesothermal gold occurrences. The greenstone belt consists of tholeiitic to calc-alkaline volcanic rocks (Amisk Group; •' 1900-1880 Ma) unconformably overlain by molasse-type sedimentary rocks (Missi Group; 1850-1840 Ma). The supracrustal rocks are intruded by gabbroic to granitic rocks ranging in age from synvolcanic to late tectonic (1890-1835 Ma). Metamorphic grade varies from prehnite-pumpellyite facies to amphibolite facies, and peak thermal conditions generally were attained locally during granitoid intrusion. Many of the shear zones and the dominant regional foliation developed during ductile deformation coeval with peak metamorphism. These shear zones were reactivated and mineralized under brittle-ductile conditions during postpeak metamorphic uplift.

Mesothermal gold mineralization is hosted in quartz veins that developed at jogs or zones of competency contrasts along brittle-ductile shear zones, in all lithologies in the region. Alteration envelopes, which consist of quartz-carbonate-chloritekalbite-muscovite-pyrite, are usually less than a few meters wide and overprint regional metamorphic assemblages. The veins consist of milky-colored quartz and may also contain tourmaline, ankerite, chlorite, and muscovite. Pyrite and arsenopyrite are the dominant sulfides, and gold grade is proportional to the modal abundance of sulfides.

The dominant fluids associated with gold mineralization were H20-CO2-NaC1 (0.6-14.7 wt % NaC1 equiv) in composition, with generally uniform H20/CO• phase ratios in a given plane or zone but variable phase ratios between zones. Oxygen isotope mineral pairs indicate temperature of mineralization of between 360 ø and 420øC, which, when combined with fluid inclusion density estimates, indicate that most of the gold mineralization occurred at a pressure of about 2 kbars. Rapid pressure release in dilatant zones with related phase separation and change in fluid composition and properties, and local fluid-wall rock interaction, are the likely gold-precipitation mechanisms.

The O, H, S, C, and Sr isotope compositions of hydrothermal minerals in the shear zones indicate that the mineralizing fluids interacted extensively with Proterozoic metamorphic and igneous rocks similar in composition to those presently exposed at the surface, at high temperatures and at low water/rock ratios. The O and H isotope compositions are compatible with formation of the fluids from devolatilization reactions during prograde metamorphism at depth. The O isotope composition of barren quartz veins varies with the host rocks indicating that these fluids were either locally derived or interacted extensively with the immediate host rocks.

An Rb-Sr isochron age of 1760 _+ 9 Ma on tourmaline and muscovite from the Rio deposit is similar to 4øAr/39r plateau ages for muscovites from the premetamorphic Laurel Lake deposit in the Flin Flon domain. This suggests that fluid advection through the Rio deposit was contemporaneous with the thermal event that affected Laurel Lake muscovites. These ages, along with an age from the Tartan Lake mesothermal deposit (1791 _ 4 Ma) in the Flin Flon domain, indicate that fluid flow related to hydrothermal activity along the shear zones in the western Flin Flon domain occurred periodically over a period of about 30 Ma.

These Proterozoic gold occurrences are similar to Archean mesothermal gold deposits in terms of geologic, structural, and tectonic setting, alteration and vein mineralogy, and fluid composition, pressure, and temperature. However, the limited extent of alteration, the low gold content, the isotopic systematics, and the lack of either S-type granites or mantle-derived lamprophyres suggest that the shear zones hosting the western Flin Flon domain occurrences sampled limited quantities of fluids and, thus, were not transcrustal features capable of hosting giant gold deposits.

0361-0128/92/1377/1496-2953.00 1496

PROTEROZOIC MESOTHERMAL GOLD, FLIN FLON DOMAIN, CANADA 1497

Introduction

Tin; geology, structure, geochemistry, timing, and tectonic setting of mesothermal gold mineralization has been the focus of intensive research over recent

years. Efforts have concentrated on understanding the genesis of the large Archean gold deposits in the Canadian Shield (Card et al., 1989, and references therein) and Western Australian Shield (Groves and Phillips, 1987, and references therein), although information derived from the study of Proterozoic (Rye and Rye, 1974; Kyser et al., 1986; Ibrahim and Kyser, 1991), Paleozoic (Sandiford and Keays, 1986; Kontak et al., 1990), Mesozoic (Bohlke and Kistler, 1986; Nesbitt et al., 1986; Goldfarb et al., 1988), and Cenozoic (Curti, 1987) gold deposits has placed further constraints on the development of an all-encom- passing model for the formation of mesothermal gold deposits.

The Trans-Hudson orogen in northern Saskatche- wan and Manitoba includes two Proterozoic green-

stone belts, the La Ronge and Flin Flon domains (Fig. 1), which host a large number of gold occurrences (e.g., Coorobe, 1984). Gold has been mined intermittently from shear zone-hosted deposits in the Flin Flon domain since about 1930, although production and reserves only total about 30 metric tons (t) of gold with the majority of that tonnage (19 t of Au) being derived from the Nor-Acme mine, Manitoba (Coombe, 1984; Bailes et al., 1987). This study fo- cuses on selected mesothermal gold occurrences in the western Flin Flon domain, Saskatchewan (Fig. 2), which contain reserves of about 2.5 metric tons of gold. The geologic characteristics of these occurrences will be described, in addition to fluid inclusion and isotopic constraints on the composition and age of the vein-forming fluids. A brief comparison will be made with Archean greenstone-hosted mesothermal gold deposits to determine if there are any significant differences between these highly productive terranes and the relatively unproductive western Flin Flon greenstone belt.

•o ø w RAE PROVINCE

LAKE ATHABASCA CREE LAKE ZONE 96øw (HEARNE PROVINCE) • 6øø N

:BAY

ETER • PHANEROZOIC WATHAMAN COVER p BATHOLITH

ROTI'ENSTON E

REINDEER ZONE

•2oOw I

PHANEROZOIC •eøøw COVER

I

I sASK. LNn. OB• ONTARIO U.S.A.

Figure 2

KISSEYNEW

SUPERIOR PROVINCE

100 km

FIc. 1. Simplified map of the lithotectonic domains of the early Proterozoic TransHudson orogen, and adjacent Archean Rae, Hearne, and Superior provinces, northern Saskatchewan and Manitoba (modified after Stauffer, 1984; Hoffman, 1988). The Reindeer zone consists of juvenile, arc-related early Protero- zoic domains, whereas the Cree Lake zone consists of portions of the Hearne province that underwent thermotectonism during the TransHudson orogen. The boundaries between domains (thin solid line) generally conform to lithological, structural, or metamorphic changes and are sometimes defined by ductile shear zones (thick solid line). Abbreviations: C-S BZ: Churchill-Superior boundary zone, NZ = Needle Falls-Parker Lake shear zone, STZ = Snowbird tectonic zone, SWZ = Sturgeon Weir shear zone, SZ: Stanley shear zone, TZ -- Tabbernor fault zone. The location of the region is shown as a stippled box in the inset. These units are unconformably overlain by rocks of the Athabasca basin and of the Phanero- zoic cover. The location of Figure 2, the area of this study, is outlined.

1498 K. M. ANSDELL AND T. K. KYSER

.................. KISSEYNEW DOMAIN

ß . .

.......... Ralne-

.......... Walke

102o00 /

l P5 ROSS LAKE .............. FAULT.

Black ....... Diamond ' 'ß

AMISK LAKE

o 5

..... Lake

Lake

',. ,;,;,;,;:;:;:;,;

•THAPAPUSKOI LAKE

PHANEROZOIC Ordovician dolomites

PROTEROZOIC

•--• Post-P3 intrusions

,,• Pre-to syn-P3 intrusions

• Feldspar porphyry

Boundary Intrusions

• iorite, gabbro

• Pre-tectonlc Intrusions

*...'../.:.**., ,:, ** :• Missi Group

:r--:-• Amlsk Group

• Shear zones

ß P4 Embury • Lake anticline

ß Gold occurrences

FIe. 2. Geologic map of the western Flin Flon domain, showing the locations of epigenetic gold occurrences. Those referred to in the text are named, except the Laurel Lake deposit, which is described in detail elsewhere (Ansdell and Kyser, 1991a). Other gold occurrences are located in Figure 3. The intrusions of known age (Ansdell and Kyser, 1991b) are labeled in italics and the ages given in Table 1. Figure modified after Byers and Dahlstrom (1954), Byers et al. (1965), Coombe (1984), and Stauffer (1984). P = deformation phase.

Regional Geology The Flin Flon domain is one of the lithotectonic

elements of the Reindeer zone, Trans-Hudson orogen (Fig. 1), that were involved in major nappe em- placement and probable crustal thickening attending collision between the Archean Superior, Hearne, and Rae provinces (Bickford et al., 1990). The exposed Flin Flon domain is about 250 km long and about 40 km wide, although there is geophysical evidence to indicate that this domain extends southward for hundreds of kilometers underneath the Phanerozoic cover (Green et al., 1985). The Flin Flon domain is bounded to the east by the Archean Superior province, to the north by amphibolite- and granulite- grade gneisses of the Proterozoic Kisseynew domain, to the west by the Sturgeon Weir shear zone and the Hanson Lake block, and is unconformably overlain to

the south by Ordovician dolomites (Fig. 1). This study concentrates on the western part of the Flin Flon domain in Saskatchewan; accordingly, the de- scription of the geology is specific to this area (Fig. 2). The absolute ages of rock types in the area are summarized in Table 1.

Supracrustal rocks

The oldest rocks in the domain, termed the Amisk Group (Bruce, 1918), are a sequence of tholeiitic to calc-alkaline volcanic and volcaniclastic rocks, which vary in composition from basaltic to rhyolitic. Major and trace element geochemistry indicate that the volcanic rocks represent a complex mixture of midocean ridge, intra-oceanic island-arc, and back-arc extru- sive rocks (Stauffer et al., 1975; Gaskarth and Par- slow, 1987; Bailes and Syme, 1989; Thom et al.,


T.•s•_• 1. Absolute Ages of Zircons from Various Rock Types in the Western Flin Flon Domain

Rock type Method Age (Ma +__ 2a) Reference

Volcanic rocks

Amisk Group U-Pb 1886 +__ 2 1 U-Pb 1926_•0 2

Sedimentary rocks Missi Formation Pb-Pb 1854 +__ 13 to 3

2529 _+ 20 Intrusive rocks

4+32 Cliff Lake U-Pb 187 -•5 1 Annabel Lake Pb-Pb 1860 +__ 6 4

Reynard Lake Pb-Pb 1853 +_ 8 4 Missi Island Pb-Pb 1848 +__ 11 4 Boot Lake Pb-Pb 1842 +_ 13 4 Phantom Lake Pb-Pb 1840 _+ 7 4

Neagle Lake Pb-Pb 1837 +_ 5 4 Phantom Lake dike Pb-Pb 1834 + 13 4

References: I = Gordon et al. (1990), 2 = Syme et al. (1991), 3 = Ansdell et al. (1991), 4 = Ansdell and Kyser (1991b)

1990). Nd isotope systematics suggest that the magmas were derived by partial melting of a relatively depleted mantle source (Chauvel et al., 1987) that was slightly contaminated by fluids derived by dehydration of subducting oceanic crust (Thom et al., 1990).

Unconformably overlying the Amisk Group is a sequence of coarse clastic fluvial sedimentary rocks, termed the Missi FormatiOn, which are interpreted as alluvial fan and braided stream molasse deposits (Stauffer, 1990). A well-developed paleoregolith is present in pre-Missi rocks below the unconformity indicating extensive pre-Missi subaerial weathering. The youngest detritral zircon from Missi metasedimentary rocks in the Flirt Flon basin has an age of 1854 +_ 13 Ma (Table 1). The Phantom Lake pluton, dated at 1840 +_ 7 Ma (Table 1), crosscuts one of the Boundary intrusions (Syme and Forester, 1977), which intrude Missi metasedimentary rocks. Missi Formation deposition in the Flin Flon basin thus is limited to between 1854 and 1840 Ma (Arisdell et al., 1991). Intrusive rocks

A wide variety of intrusive bodies ranging in composition from gabbroic to granitic, and ranging in age from synvolcanic to posttectonic, crosscut these supracrustal rocks (Fig. 2). The plutons named in Fig- ure 2 have been dated using zircons and range in age from 1874 to 1834 Ma (Table 1). The geochemistry and Nd isotope compositions of these plutons suggest that they are volcanic island-arc granitoids and were not derived by partial melting of an older, more radiogenic source (Arisdell and Kyser, 1990). Granitoid plutons from the eastern Flirt Flon, La Ronge, Glen- hie Lake, and Hanson Lake domains in the Trans-

Hudson orogen have similar ages (Van Schmus et al., 1987; Delaney et al., 1988; Gordon et al., 1990), indicating a major period of intrusive activity related to cratonization of this part of the North American shield (Hoffman, 1988).

Deformation

All the rocks in the western Flin Flon domain have

been variably deformed by up to five deformation events, although the intensity of deformation observed in many areas is a complex function of competency contrasts, metamorphic grade, and heteroge- neous strain. The morphology and orientation of structures developed during these events have been described in detail by Stauffer and Mukherjee (1971), Bailes and Syme (1989), Wilcox (1990), and Fedorowich et al. (1991). The first two phases of deformation, the pre-Missi (phase 1), and the first post- Missi (phase 2), resulted in folds with no apparent attendant axial planar foliation. The dominant pene- trative foliation and north-south ductile shear zones

developed during the third phase of deformation (phase 3). These shear zones were reactivated under brittle-ductile conditions during the development of the Embury Lake fold (phase 4) (Fedorowich et al., 1991). Locally, phase 3 foliation was intensely folded during phase 4. The Ross Lake fault system (phase 5) crosscuts the Ernbury Lake fold and has a dominantly brittle character.

Metamorphism

Metamorphic grade in the western Flin Flon domain increases from prehnite-pumpellyite grade in the Flirt Flon area (Bailes and Syme, 1989; Digel and Gordon, 1991), to amphibolite grade northward to the contact with the Kisseynew domain (Ashton et al., 1987), westward toward the Hanson Lake block (Ashton, 1990), and eastward toward Snow Lake, Manitoba (Aggarwal and Nesbitt, 1987). Peak regional metamorphism is considered to be broadly coeval with the development of foliation during phases 3 and 4. Isograds are offset by phase 5 fault zones (Bailes and Syme, 1989; Digel and Gordon, 1991). The regional metamorphic gradient in the Flirt Flon region is complicated by amphibolite-grade contact metamorphism around many of the plutons (Longiaru, 1980; Digel and Gordon, 1991). Longiaru (1980) suggested that the contact aureole around the Reynard Lake pluton represents a local increase in the thermal gradient during peak metamorphism, whereas Digel and Gordon (1991) identified prehnite-pumpellyite veins that crosscut the contact aureole of the Phantom Lake granite dike, suggesting that low-grade regional metamorphism postdates the intrusion of this pluton.


Gold Mineralization

The locations of epigenetic gold occurrences in the western Flin Flon domain are shown in Figures 2 and 3. Representative surface, underground, or drill core samples of veins and alteration envelopes were col- lected from the named gold occurrences with the ob- jective of characterizing their mineralogical and geochemical characteristics. All have features typical of mesothermal gold deposits, with the exception of the Laurel Lake deposit (Fig. 2). This is a premetamorphic deposit with elevated silver and base metal contents reminiscent of an epithermal gold deposit and has been described in detail elsewhere (Ansdell and Kyser, 1991a).

Mineralized shear zones are generally oriented parallel to the main phase 3 ductile shear zones in the region (Fig. 2). Gold mineralization is specifically associated with sulfides in tensional quartz veins, some of which are folded during later ductile deformation. Although the timing of all the gold occurrences rela-

tive to each other is difficult to constrain, the alteration halos around the quartz veins overprint peak metamorphic minerals. This, and the character of the quartz veins, suggests that veining and alteration occurred during brittle-ductile deformation as the region cooled and was uplifted after peak regional metamorphism. Fedorowich et al. (1991) showed that the mineralized quartz veins at Tartan Lake, Manitoba, developed coevally with reactivation of the earlier, ductile Tartan Lake shear zone during the formation of the phase 4 Erabury Lake fold (Fig. 2). In many cases, mineralized quartz veins are located at jogs in shear zones (e.g., Rio) or at sites of competency contrasts (e.g., metasedimentary rock-marie dike contact at Graham) but can occur in all lithologies in the region (e.g., marie volcanic rocks--Newcor; conglom- erates-Graham; diorite--Henning-Maloney; granite--IMC-Cain B). The geologic characteristics of the occurrences referred to in this study are summarized in Table 2, except for the Rio deposit, the largest mesothermal gold occurrence in the Saskatche-

Henning-•, Mal•ey 7%/•,

Nofih ß Shear Zone I

ß •) ß ......

PHANTOM LAKE

McMillan

Phantom,

• Phantom Lake granite

.•...•::• Boot Lake g ranodiorit e quartz- monzodiorile

quartz- cliorite

I Boundary Intrusions

• Amisk Group

Gold ß occurrences

Shear zones Lithelogical contacts

Fig. 3. Geologic map of the Phantom Lake area, showing the location of shear zones and gold occurrences which are part of this study (modified after Galley and Franklin, 1987; Thomas, 1989). The ages of some of the rocks are summarized in Table 1.


wan portion of the Flin Flon domain, which is described in detail below.

Rio deposit

The Rio deposit is one of a number of deposits spatially associated with shear zones in the vicinity of the Boot Lake-Phantom Lake intrusive complex (Fig. 3). Gold mineralization is located within a zone of in-

tense hydrothermal alteration and veining along the Rio fault, a shear zone exhibiting ductile and brittle characteristics, which crosscuts all lithologies in the area (Pearson, 1984; Galley and Franklin, 1987; Thomas, 1989). Published reserves are 220,000 tons at 0.14 oz of Au/ton (Northern Miner, July 18, 1988).

The Rio fault transects a sequence of westerly fac- ing, sporadically porphyritic, massive and pillowed basalts and debris flows (Fig. 4A). This volcanic sequence is in turn cut by diorites and granodiorites of the Boot Lake pluton. The youngest intrusion in the area, the Phantom Lake granite (Fig. 3), is repre-

• Phantom Lake granite dyke

.. ;.:...::.....:•::.. ........ a.. .• .......... .•'• 1•3otLake ß . ".':.'•....'•..:..'..'...?...;(:.:• ...... •... * •,:,•,**,2,:.:, granodiofite :':: ' diorite Amisk Group

i..'..• Basaltic

Lithological Shear • • contact zone

A

N

20 m

WC ZONE ALTERATION

Underground • workings r•x,,v. xx,x.• • • • •, PHANTOM LAKE

.,i• GRANITE DYKE

AMISK G•'• ZONE

BASALTS • ' • AJ ZONE VEINS AND ALTERATION

F]c. 4. A. Surface geology at the Rio deposit (based on Pear- son, 1984). B. Geologic plan of the 215 • level, Rio deposit, showing the three different ore associations, the WC, AJ, and Phantom Lake granite dike zones.

sented in the vicinity of the Rio deposit by an irregularly shaped microcline-porphyritic dike (Fig. 4A) having an age of 1834 _ 13 Ma (Ansdell and Kyser, 1991 b). The alteration related to gold mineralization overprints this dike indicating that the gold mineralization postdates the main magmatic events in the area. Intense alteration and deformation are concen-

trated in an area up to 60 m wide where basalts occur on either side of the fault, and there is a change in strike of the Rio fault (Fig. 4A). The Rio fault dips approximately 75 ø to the northwest. Three different alteration and vein associations have been identified

underground and have been named the AJ, WC, and Phantom Lake granite dike zones (Fig. 4B). In all three types of mineralization the gold grade is correlated with modal pyrite abundance; gold occurs as inclusions in pyrite and along pyrite grain boundaries (Pearson, 1984; Fig. 5A). The average composition of the native gold is Au•9Ag•0Hg•.

AJ zone mineralization consists of a set of branch- ing and en echelon quartz veins and associated alteration up to 2 m wide that strike approximately north- south and dip to the northwest at greater than or equal to 60 ø (Fig. 4B). The earliest veins are anker- itc-quartz stringers that have been folded and boudinaged during later ductile deformation. Later veins parallel the foliation and consist of either massive but irregular quartz-ankerite _+ tourmaline _ chlorite _ pyrite veins (Fig. 5B) or banded veins with a similar mineralogy although with textures indicative of a crack and seal mechanism. Pyrite is irregularly dis- tributed within both of the later quartz vein types and gold grade is correlated with pyrite content. Visi- ble hydrothermal alteration is associated with the quartz veins but is not always symmetrically disposed about them. The width and intensity of alteration varies but generally both increase proximal to vein inter- sections. The alteration closest to a particular vein often has a brecciated appearance and consists of quartz-ankerite-albite-pyrite-chlorite (Fig. 5C). This alteration is generally in sharp contact with an outer, buff-colored zone ofankerite, quartz, chlorite and pyrite, which is frequently cut by chlorite-ankerite-pyrite veinlets. Dark, apparently unaltered mafic host rocks between alteration zones consist of chlorite-

plagioclase-epidote-calcite-muscovite-pyrite and are commonly cut by calcite-quartz veinlets. This assemblage postdates the greenschist facies assemblage of actinolite-epidote-chlorite-plagioclase-quartz _+ biotite generally observed in metabasaltic rocks in the region (Digel and Gordon, 1991), and thus, indicates that mineralization occurred after peak metamorphism.

Gold mineralization in the WC zone is associated with lensoid bodies of intense alteration that trend

about 025ø; it overprints mafic volcanic rocks and the granite (Fig. 4B). The alteration mineralogy is

PROTEROZOIC MESOTHERMAL GOLD, FLIH FLON DOMAIN, CANADA 1503

/ Pyr /

FIC. 5. Photomicrographs of veins and alteration at the Rio deposit. A. Native gold along pyrite grain boundaries, WC mineralization, 400' level. Scale bar = 0.2 ram. B. Quartz-pyrite-tourmaline-ankerite vein, AJ raineralization, 135' level. Scale bar = 0.5 min. C. Alteration consisting of ankerite, albite, and quartz, AJ raineralization, 215' level. Scale bar = 0.2 mm. D. Quartz~pyrite-muscovite-ankerite vein cutting altered Phantom Lake granite, 360' level. Scale bar = 0.5 mm. Ank = ankerite, Muse = muscovite, Pyr = pyrite, Qz = quartz, Tour = tourmaline.

simple and consists of an up to a 20-m-wide zone of fine-grained quartz, ankerite, albite, and pyrite, with irregularly developed quartz-pyrite veins.

The Phantom Lake granite dike zone consists of tension-gash quartz veins hosted in a dike. The quartz veins trend approximately north-south and are concentrated along the margins of the dike or are parallel to the axial plane of minor folds. The veins do not continue into the surrounding marie volcanic rocks. Quartz is the dominant vein mineral, whereas pyrite, muscovite, and ankerite, where they occur, are intergrown along the vein margins (Fig. 5D). The quartz exhibits evidence of later strain, namely undulose extinction and subgrain development, and sporadically the veins are boudinaged and offset by ohio- rite-filled fractures implying that deformation and hydrothermal alteration continued after the main period of vein formation. Around the quartz veins, the host granite is bleached and consists of quartz, albite, muscovite, carbonate, pyrite, and minor he-

matite and stilpnomelane. The mineralogy of a dike apparently unaffected by hydrothermal alteration is quartz - microcline - plagioclase - biotite - amphibole - sphene-maguetite, and so the dominant mineralogical changes involve albitization, sericitization, car- bonatization, and pyritization. Changes in the chemical composition of the Phantom Lake granite dike during hydrothermal alteration are evident from the composition of altered samples from the Rio deposit (Table 3). Increases in Na and Ca, and decreases in K, Rb, Sr, and Ba in samples associated with gold mineralization are related to the breakdown of the igneous feldspars and the formation of hydrothermal albite, as evidenced by the petrographic relationships. Loss on ignition also increases, which is related to CO2 addition and precipitation of carbonate. Although the alteration halos in the Phantom Lake granite dike zone are small in size, the alteration mineralogy and geochemical changes around the quartz veins are very similar to those seen around other mesothermal


TABLE 3. Major and Trace Element Compositions of Samples of Phantom Lake Granite Proximal to the Rio

Mesothermal Gold Occurrence

Sample no. 449 • 179 • 3113 2094

SiO• 67.9 64.9 58.3 50.4 TiO• 0.36 0.67 0.68 0.54 AI•O3 15.6 16.1 17.7 14.5 Fe•O3 2.57 3.91 2.85 5.95 MnO 0.04 0.04 0.07 0.12

MgO 1.19 1.47 1.31 2.66 CaO 1.58 2.68 3.98 7.37

Na•O 4.74 5.2 9.59 8.06 K20 4.5 3.0 0.77 0.36 P•O5 0.15 0.27 0.31 0.29 L.O.I. 1.31 1.23 4.85 8.62 Total 100.3 99.9 100.5 99.0 Cr 27 25 33 24 Rb 76 68 29 23 Sr 1,480 1,700 537 608 Zr 128 121 172 119

Ba 1,680 1,620 241 104

• Center of Phantom Lake granite 2 Phantom Lake granite dike, within 10 m of Rio fault 3 Apparently unaltered Phantom Lake granite dike, Rio 300'

level

4 Altered Phantom Lake granite dike in PLG zone, Rio 215' level Major elements in weight percent; trace elements in ppm

gold veins, e.g., the Oriental mesothermal gold deposit, Alleghany, California (Bohlke, 1989).

Fluid Inclusions

All the quartz veins examined exhibit evidence of deformation (e.g., undulose extinction and subgrain development), and petrographic relationships between fluid inclusions are complex. The inclusions typically are small, with the largest inclusions being 20 •m in diameter, but most are less than 10 •m in length. Although attempts were made to classify inclusions as primary, pseudosecondary, or secondary using the criteria outlined by Roedder (1979), it is likely that early inclusions have undergone necking and related volume changes, and leakage during later deformation. In this study, it is assumed that primary and pseudosecondary inclusions are considered to provide information on some of the earlier fluids passing through the shear zone systems although not nec- essarily the actual fluid from which the quartz precipi- tated.

Primary, pseudosecondary, and most of the secondary inclusions are CO2 bearing and can be divided into H20-CO•-NaC1 (Fig. 6A and C) and CO• inclusions (Fig. 6B). Inclusions that contain a carbonic fluid may have a thin film of water around the inclusion rim which is difficult to discern optically. The volume percent CO• estimated for CO•-bearing inclusions in all occurrences is usually constant within a given zone or plane, with few exceptions (e.g., Rio AJ

zone, samples 196-1-2-1 to 5, Table 4). Differences between adjacent zones or planes are marked, although the time relationships between these regions are unclear. All quartz veins are strained and so volume changes and undetected leakage may explain these variations in volume percent CO2. Two-phase aqueous inclusions generally occur along secondary planes (Fig. 6D) and these healed fractures are sometimes observed offsetting earlier CO•-bearing planes of secondary inclusions. However, there are some aqueous inclusions at the Newcor deposit that appear to be early, because they are located away from an obvious healed fracture. Thermometric measurements were made on all inclusions that could be con-

fidently classified and were large enough to permit easy observation of phase changes (Table 4).

Measurements of the melting point oof CO• (Tmco) fall within the range -56.6 ø to -57.8 C (Fig. 7AI, indicating that these inclusions contain less than 10 mole percent CH 4 equiv in the carbonic phase (e.g., Swanenberg, 1979). However, some primary inclusions from the Newcor occurrence exhibit Tmco• as low as -63.2øC (Fig. 7A), most likely due to substantial amounts of either CH 4 or N•. As most of the CO•- bearing inclusions have Tmc o close to -56.6øC indicating only minor amounts •f either CH 4 or N2, the salinities calculated using Tmc,athrate (Collins, 1979) should be a good approximation of the salinity of the inclusion. Clathrate-melting temperatures (Tm½lathrate) from primary and pseudosecondary inclusions that could be measured range from 0.8 ø to 10.1øC (Table 4, Fig. 7B). Fluid inclusions from the Henning-Ma- loney occurrence exhibit the greatest range (0.8 ø- 8.8øC), whereas samples from the Rio and Graham deposits show much more restricted ranges. Salini- ties calculated are 4.0 wt percent NaC1 equiv for the Rio deposit, 0.6 to 5.0 wt percent NaC1 equiv for the Graham occurrence, 2.4 to 14.7 wt percent NaC1 equiv for Henning-Maloney, and 12.5 wt percent NaC1 equiv for IMC-Cain B. Only one of the Tm½l•thr•t½ values obtained from the Rio deposit could be used to calculate salinity because the others did not melt in the presence of CO• liquid and gas (Table 4). How- ever, the Tm•l•th•at½ data from other fluid inclusions at the Rio deposit are all greater than 8.0øC, and so the salinities must be less than about 6 wt percent NaC1 equiv (Collins, 1979, fig. 1).

The homogenization temperatures of CO• (Thco•) in primary and pseudosecondary inclusions range from -18.6 to 30.1øC (Table 4, Fig. 7C). Most of these inclusions homogenize to the liquid phase, although two exhibited homogenization to the vapor phase and three exhibited critical behavior. The wid- est range is shown by inclusions from the AJ zone at the Rio deposit, and one sample (196) has a zone in which fluid inclusions have Tmc o that range from -8.1 ø to +21.3øC (Table 4). CO•'-•oearing inclusions


FIG. 6. Photomicrographs of representative fluid inclusions. A. Primary CO2-H20 and CO2 inclusions. AJ vein, Rio deposit, 215' level. Scale bar -- 75 ym. B. Primary, one-phase CO2 inclusion. Monarch occurrence. Scale bar -- 75 ym. C. Two-phase CO•-H•O early secondary inclusions. Graham occurrence. Scale bar = 75 urn. D. Secondary aqueous inclusions. North Shear Zone occurrence. Scale bar = 75 urn.

at the Newcor occurrence also have variably low Thco• (Fig. 7C), and this correlates with the higher concentrations of CH4 or No. in these inclusions.

Most of the inclusions are COo. rich and decrepitate prior to bulk homogenization. Decrepitation temperatures at the Rio deposit range from 174 ø to 282øC, whereas COo.-rich inclusions at the North Shear Zone deposit decrepitate above 400øC (Table 4). The only COo.-rich inclusions to homogenize are those that homogenize to the vapor phase (220ø-329øC; Table 4). Although trapping temperatures cannot be inferred from these measurements, their bulk compositions can still be determined using the technique described by Burruss (1981). The bulk densities and mole fraction COo. and Ho.O are variable (Table 4), although most primary and pseudosecondary inclusions have densities ranging from 0.80 to 1.01 g/co.

Primary aqueous inclusions at Newcor have low eutectic melting temperatures (Fig. 8A) as well as low final melting temperatures (Fig. 8B), but no salinities are calculated because of the uncertainty in the actual composition of these inclusions. Aqueous

secondary inclusions from other occurrences have ice-melting temperatures (Tin.) ranging from -12.2 o to -0.2øC (Fig. 8B). Salinitle• calculated from the equation of Potter et al. (1978) range from 16.2 to 0.4 wt percent NaCI equiv.

All aqueous inclusions homogenize to the liquid phase at temperatures in the interval of 108 ø to 500øC, although the range exhibited by inclusions in a given plane is much more restricted (Table 4). How- ever, homogenization temperatures exhibited by primary aqueous inclusions at Henning-Maloney and Newcor (284ø-500øC) are generally higher than those observed in secondary aqueous inclusions (108ø-330øC; Fig. 8C). This temperature difference indicates that the secondary fluid inclusions record lower temperature fluids infiltrating through the sheared rocks, most likely during uplift of the area after gold mineralization.

Stable Isotope Geochemistry

The $xsO values of quartz from veins in the gold occurrences range from 0.0 to 13.0 per mil, although


•1111 ii•l • i, ill ]1


48

0 36

._

'• 28 ß

24

2o

16

12

$

4

0

A

I RIO - AJ ZONE :.ij[] RIO - PLG ZONE •;• NEWCOR ]HENNING-MALONEY ]NORTH SHEAR ZONE I IMC - CAIN B --"]GRAHAM I MONARCH m GOLDEN CROSS

8

0 I

0 2 4 6 8 10

Tm Clath(OC)

18 -

16

14

12 -

o

.c_ 8

Z

4

2

o •, , -24 -t6 -8 0 8 16 24 32

-64 -63 -62 -61 -60 -59 -58 -57 -56 Th CO 2 (oc) Tm CO 2

FIC. 7. Microthermometric measurements from early (primary and pseudosecondary) CO2-bearing fluid inclusions for auriferous veins shown in Figures 2, 3, and 4. A. Melting point of CO2. B. Melting point ofclathrate. C. Homogenization temperature of CO•. In all cases there is no obvious compositional difference between different zones or planes of inclusions from a given gold occurrence.

the range of values seen in specific gold deposits is more restricted (e.g., Rio, 9.9-10.6%0; Graham, 11.3-12.2%0; Table 5). These data are comparable to those of other auriferous quartz vein samples from the Flirt Flon domain, which have •180 values of between 10.6 and 14.3 per mil (Kyser et al., 1986; Fe- dorowich et al. 1991) and fall within the range exhibited by quartz from mesothermal gold deposits of all ages (Kerrich, 1989). The •180 values of quartz from barren veins (Fig. 9) range from 5.6 to 15.5 per mil and overlap the data from auriferous quartz veins (Fig. 9), indicating that the oxygen isotope composition of quartz is not a definitive exploration criterion in this area as was suggested by Kyser et al. (1986). However, there appears to be a relationship between the oxygen isotope composition of barren quartz veins, and the oxygen isotope composition of their host rock (Fig. 9). Barren veins hosted by intrusive rocks have lower quartz •180 values than barren veins hosted by the relatively 1SO-rich Missi sedimentary rocks. Barren veins hosted by Amisk volcanic rocks usually have intermediate •180 values (10.7- 12.5%0) but range up to 15.5 per rail.

Tourmaline and chlorite from the AJ mineralization at the Rio deposit have •180 values of 7.4 and 3.8 per mil, respectively (Table 5). These minerals are in textural equilibrium with quartz and give quartz- tourmaline and quartz-chlorite oxygen isotope tem-

peratures of 400 ø _+ 30 ø and 395 ø _+ 30øC, respectively, if equilibrium is assumed. Muscovite from a Phantom Lake granite dike zone vein has an oxygen isotope composition of 7.5 per mil, which yields an oxygen isotope equilibration temperature with coexisting quartz of 360 ø _+ 30øC (Table 5). Chlorite at the Graham deposit is in oxygen isotope equilibrium with coexisting quartz at 420 ø _+ 30øC. These temperatures suggest that vein formation at the Rio and Gra- ham deposits occurred at temperatures between about 360 ø and 420øC.

The hydrogen isotope compositions of the hydrous phases discussed above are reported in Table 5, as well as the calculated oxygen and hydrogen isotope compositions of the fluids in equilibrium with the vein minerals based on the temperatures calculated from oxygen isotopes. The Rio AJ and Phantom Lake granite dike veins formed from a fluid with a •180 value of between 5.5 and 6.2 per mil, and a •D value of between -34 and -50 per mil (Fig. 10). The Gra- ham vein fluid is slightly enriched in 180, having a •180 value of 8.6 per rail and a •D value of -37 per mil (Fig. 10). All these values are typical of fluids that have become 180 rich as a result of extensive fluid- rock interaction.

Pyrites from the Rio and Graham deposits exhibit a restricted range in sulfur isotope composition from 2.8 to 5.5 per mil (Table 5). These values fall within


-80 -70 -60 -50 -40 -30 -20

Te (oc)

-36 -32 -28 -24 -20 -16 -12 -8 -4 0

Tmlce(OC)

10" PRIMARY INCLUSIONS

NEWCOR 8

•) HENNING-MALONEY

'•4

0 I I I I I I I I

100 150 200 250 300 350 400 450 500

Th (oc)

FIG. 8. Microthermometric measurements from all aqueous fluid inclusions. All inclusions are secondary, except those from the Newcor and Henning-Maloney occurrences as indicated. Sym- bols as in Figure 7, except for the primary inclusions. A. Eutectic melting temperature. B. Melting point of ice. C. Homogenization temperature.

the range of I to 6 per mil exhibited by mesothermal gold deposits of all ages (Kerrich, 1989) and are similar to the range from 0.4 to 3.1 per mil reported by Kyser et al. (1986) and Fedorowich et al. (1991) for pyrite and chalcopyrite from other mesothermal gold occurrences in the western Flin Flon domain.

Ankerites from the Rio, Graham, Henning-Ma- loney, and Robinson Creek occurrences exhibit a restricted range in •]aC values of between -3.8 and -7.3 per mil (Table 5), which is similar to the range exhibited by carbonates from Archean mesothermal gold deposits (Kerrich, 1989). The •]80 values of the ankerites are more variable (Table 5) and quartz-ankerite oxygen isotope equilibration temperatures calculated using the fractionation factor of Matthews and Katz (1977) are not concordant with those calculated using silicate mineral pairs. Variable •80 values of carbonates are common in mesothermal gold deposits and can be attributed to partial recrystallization and reequilibration of oxygen isotopes in carbonates during later mineral-fluid interaction (Kyser et al., 1986; Kerrich, 1987).

Timing of Gold Mineralization

Mesothermal gold occurrences in the western Flin Flon domain are hosted in shear zones which crosscut

all lithologies in the area. A maximum age of 1834 _ 13 Ma for gold mineralization at the Rio deposit is provided by the age of the Phantom Lake granite and associated dikes (Ansdell and Kyser, 1991b), which are the youngest rock units in the area. A Phantom Lake granite dike is overprinted by hydrothermal alteration and gold mineralization at the Rio deposit. An approximate age for gold mineralization at the Rio deposit is provided by an Rb-Sr mineral isochron con- structed using tourmaline (sample 42) from an AJ zone vein (Fig. 5C) and muscovite (sample 428) from a Phantom Lake granite dike zone vein (Fig. 5F). The phases were deposited from hydrothermal fluids of very similar temperatures and stable isotope compositions, implying that they were broadly contemporaneous, and yield an age of 1760 _ 9 Ma and an initial •?Sr/•6Sr ratio of 0.7022 (Table 6, Fig. 11A).

An age of about 1760 Ma for the fluids associated with gold mineralization at the Rio deposit obtained from Rb-Sr systematics is similar to 4øAr/a9Ar plateau ages obtained from muscovites associated with the Laurel Lake Au-Ag deposit (Fig. 2); namely, 1746 _ 10 Ma for fine-grained muscovite in the alteration zone (Table 7, Fig. 11B) and 1753 _ 10 Ma for coarse-grained vein muscovite (Table 7, Fig. 11C). The Laurel Lake deposit predates peak regional metamorphism (Ansdell and Kyser, 1991a) and is cut by shear zones that are similar in orientation and rela-

tive timing to those that host mesothermal gold mineralization in the region. The oxygen and hydrogen isotope compositions of muscovite in mineralized veins and alteration zones indicate at least partial preserva- tion of the original, premetamorphic stable isotope signature (Ansdell and Kyser, 1991a), and so any recrystallization at Laurel Lake probably did not occur in the presence of a pervasive fluid phase but rather under relatively dry conditions. The plateau ages at Laurel Lake are thus likely to represent the timing at which the muscovites cooled through their closure temperature for Ar (about 350øC, Dallmeyer and Keppie, 1987) during the thermal event associated with fluid movement along the regional shear zones. Thus, the Rb-Sr isochron age of 1760 _ 9 Ma obtained from the Rio deposit represents the time of mineral growth in veins from an auriferous fluid that was nearly contemporaneous with the thermal event that affects muscovite Ar-Ar systematics at the Laurel Lake deposit. Fedorowich et al. (1991) also reported an 4øAr/agAr plateau age of 1791 +_ 4 Ma obtained from hydrothermal muscovite at the Tartan Lake gold deposit, Manitoba. The Tartan Lake and Rio deposits were both formed from similar hydrothermal fluids with oxygen isotope equilibration temperatures of between 360 ø and 400øC (Table 5; Fedoro- wich et al., 1991). Together, these ages suggest that fluid flow and associated thermal affects related to

hydrothermal activity along the shear zones in the


.%


AMISK VOLCANIC ROCKS Basa•t•'andesites

Rhyohtes

MISSI SEDIMENTARY ROCKS

ß Conglomerate

INTRUSIVE ROCKS Ullramafic • • Granite

6 i

4 I

2 M

2 •

i I I I

6 7 8 9

BARREN VEINS

HOST ROCKS

:;•B AMISK VOLCANIC ROCKS

SEDIMENTARY ROCKS

.,.r• INTRuSIvE ROCKS

0 m

GOLD-BEARING QUARTZ VEINS

:"':• RIO D GRAHAM M ARTAN LAKE I OTHER VEINS

I I I I

10 11 12 13

518 0 WR (permil)

I i I

14 15 16

I I I I 5 6 7 8 9 10 11 12 13 14 15 16

(• 18 O qtz (per mil) FIc. 9. /5]sO values of supracrustal and intrusive whole rocks, and barren and gold-bearing quartz veins

in the western Flin Flon domain. Data from this study, Kyser et al. (1986), Aggarwal and Longstaffe (1987), and Fedorowich et al. (1991). Barren tension-gash veins, which are classified on the basis of their host rock, have b]So values of barren quartz that correlate with the b]sO values of surrounding host rock.

western Flin Flon domain occurred periodically over a period of tens of millions of years.

Discussion of Results

Constraints on fluid and solute sources

The calculated oxygen and hydrogen isotope compositions of the fluids in equilibrium with vein minerals at the Rio and Graham occurrences, at the temperatures calculated using oxygen isotope mineral pairs, are similar to those of fluids that have formed from or interacted with metamorphic or igneous rocks at high temperatures and under conditions of low water/rock ratios (Fig. 10). The involvement of

magmatic waters (fluids exsolved from magmas) is considered unlikely in any of the western Flin Flon domain gold occurrences, because all the observed intrusive rocks in the region predate gold mineralization by at least 40 m.y. and fluid properties are dissim- ilar to well-documented magmatic hydrothermal fluids. The partial overlap between the isotopic composition of the fluids in the Rio deposit and typical magmatic water most likely indicates that these fluids interacted at low water/rock ratios with igneous rocks. The high b180 value of the fluid involved in the Graham deposit may indicate greater interaction with 1SO-rich sedimentary rocks of the Missi Group which host the deposit. Low-latitude meteoric waters with bD and blSo values of about -40 and -5 per mil,


-4O

-8O

-120

•_••wSEA Archcan ATER mesothermal Au

deposit waters

,L • ! Waters O•

• Metamorphic

• (300 - 600 o C)

I Tartan Lake

Magmatic Waters

-'0 6 ' 1•) ' 2b ' 3b • 0 H 20 (per mil)

FIc. 10. Calculated oxygen and hydrogen isotope compositions of water involved in the formation of the Rio and Graham gold occurrences. The range of values calculated for other TransHud- son orogen mesothermal gold deposits are also shown: the Tartan Lake deposit, Flin Flon domain (Fedorowich et al., 1991); the Star Lake deposit, La Ronge domain (Ibrahim and Kyser, 1991). The isotopic compositions of meteoric waters, seawater, and values typical of magmatic and metamorphic waters are shown for reference (Taylor, 1974; Kyser, 1987). The typical range of values exhibited by Archcan lode gold deposits is taken from Kerrich (1987).

respectively, could evolve to the calculated isotopic compositions for the Rio and Graham hydrothermal fluids by interaction with typical crustal rocks, but only under conditions of very low water/rock ratios. Under these conditions, the chemical and isotopic compositions of fluids derived directly from metamorphic dehydration will become similar to those from extreme interaction between extraneous water and metamorphic or igneous rocks. Carbon and sulfur are significant components of the hydrothermal fluids, as indicated by the high CO2 content of early fluid inclusions and by the ubiquitous presence of carbonates and sulfides in the veins and surrounding alteration halos. The •i•3C values of carbonates and the •i34S values of pyrites from occurrences in the western Flin Flon domain range from -4.5 to -7.3 per mil and 2.8 to 5.5 per mil, respectively. The carbon isotope data alone cannot differentiate between a magmatic source or an average crustal source, but they do indicate that the bulk of the carbon was not derived from a reduced carbon reservoir (•il•C between -15 and -35%0) or a source dominated by ma- rine carbonates (•il•C between -2 and +4%0) (Hoefs, 1987; Ohmoto and Rye, 1979). The •i•4S values of sulfides in the western Flin Flon domain are similar

to typical sulfides in igneous rocks (Ohmoto and Rye, 1979), although they cannot be used to discriminate between sulfur derived directly from magmas and

that derived by later high-temperature dissolution or desulfidation of magmatic sulfide minerals.

Tourmaline is commonly associated with mesothermal gold mineralization, is resistant to retrograde alteration, and has an extremely low Rb/Sr ratio, which makes it an ideal indicator of the strontium isotope composition of the hydrothermal fluid and the source of the strontium (e.g., King and Kerrich, 1989). A tourmaline sample from the Rio deposit has an Rb/Sr ratio of 0.008 and an 87Sr/86Sr ratio of 0.70œ791 +_ 36 (Table 6). This is similar to the 87Sr/S6Sr isotope compositions of tourmalines from late veins at the Laurel Lake deposit (Fig. 2; Ansdell and Kyser, 1991a) and to initial S7Sr/S6Sr ratios obtained from Amisk Group volcanic rocks and granitoids (Mukherjee et al., 1971; Watters and Armstrong, 1985); this implies that the Sr was likely derived by leaching of Protero- zoic island-arc volcanic rocks and granitoids, similar in age and strontium isotope composition to those exposed at surface.

In summary, the isotopic compositions of O, H, C, S, and Sr in the hydrothermal fluids indicate that the fluids have interacted with crustal rocks similar to

those presently exposed in the Flin Flon area under conditions of low water/rock ratios. The temporal relationship between known igneous activity in the Flin Flon area and gold mineralization indicates that a magmatic source for the auriferous fluids is unlikely. Interaction between surface-derived waters and Flin

Flon domain rocks could produce the observed isotopic compositions; however, the percolation of surface-derived waters to depths in excess of 15 km can be discounted from field evidence for lithostatic fluid

pressures and would require a large, as yet unidenti- fied, heat source to drive fluid convection. The isotopic systematics are best explained by deriving the fluids and solutes by devolatilization and dissolution reactions during regional metamorphism of rocks similar in age and composition to those presently exposed on surface.

Depositional conditions Oxygen isotope equilibration temperatures have

been calculated for quartz and chlorite, quartz and muscovite, and quartz and tourmaline in textural equilibrium, and provide estimates of the temperatures of formation of vein minerals (Table 5). These minerals are in textural equilibrium with pyrite and gold, indicating that the gold-bearing veins at the Rio deposit formed at temperatures between about 360 ø and 400øC; the veins at the Graham occurrence formed at about 420øC. Fluid inclusion bulk homogenization temperatures cannot be used to provide minimum vein formation temperatures, because most of the early CO2-rich inclusions decrepitate be- fore homogenization. The measured decrepitation temperatures (Table 4), which are lower than the


A

0.85 428 musc

0.80 t Ag• 0.75

•'• In'dial •Sr/•Sr

0.65

0 I 2 3 4 5 6

• Rb/80 Sr

B

C

lOOO

lOO lO

1775

1750

1725

1700

1675

--

. '1 Plateau age = 1746 :t10 Ma

- 267 Laurel Lake muscovite

I I ] I [ I f I [ 0 20 40 60 80

% 39Ar cumulative release lOO

1800 •-- 297 Laurel Lake 7

1760

1740

I • I • I • I • I , I 0 • 40 •0 • 100

% 39 Ar cumulative release

Fie. 11. A. Rb-Sr mineral isochron for tourmaline and muscovite from the Rio deposit, which provides an estimate of the age of mineralization and the isotopic composition of the source of Sr in the gold-bearing hydrothermal system. B. 4øAr/39Ar apparent age spectra for fine-grained muscovite in altered quartz- feldspar porphyry at the Laurel Lake Au-Ag deposit (Ansdell and Kyser, 1991a). C. 4øAr?gAr apparent age spectra for vein muscovite from the Laurel Lake deposit. Isotopic data and calculated ages for B and C are from Table 7.

bulk homogenization temperatures, are also lower than the calculated oxygen isotope equilibration temperatures.

An estimate of the trapping pressure can be made using isochores determined from the fluid inclusions and the temperatures determined by oxygen isotope geothermometry (Fig. 12). CO2-rich inclusions are prevalent in the AJ veins at the Rio deposit and at the Graham occurrence, and so isochores for pure CO2 are used to limit the pressure conditions during vein formation. Primary inclusions in the Phantom Lake

T/•BL•. 6. Rb-Sr Isotope Data for Tourmaline (tour) and Muscovite (musc) from the Rio Deposit

Sample no. Rb (ppm) Sr (ppm) S7Rb/86Sr 87Sr/S6Sr

42 tour 5.2 651.00 0.0233 0.702791 +_ 36 428 musc 163.2 86.60 5.5251 0.842000 _+ 85

Samples are described in Table 4

granite dike zone at the Rio deposit contain a visible aqueous phase; accordingly, bulk densities have been calculated using an estimated volume percent CO•. Precise volume measurement is difficult, so estimated densities are subject to a large uncertainty. The constraints provided by the densities of the inclusions in conjunction with temperatures obtained from oxygen isotopes suggest a pressure of mineralization of approximately 1.5 to 3 kbars (Fig. 12), equivalent to a depth of about 6 to 12 km assuming lithostatic pressure. The density of CO• in primary CO•-rich fluid inclusions at the Monarch and Golden Cross occurrences is similar to that seen at the Gra-

ham occurrence, suggesting these three occurrences formed at about the same depth in the crust. The Henning-Maloney, North Shear Zone, and IMC-Cain B occurrences generally have inclusions with lower density CO• than the adjacent Rio and Newcor deposits (Table 4, Fig. 7C) implying that the former occurrences may have developed at a lower fluid pressure than the latter deposits.


TABLE 7. 4øAr-a9Ar Analytical Data for Incremental Heating Experiments

Temperature a7Ar?gArl a6Ar/a9Arl a•Ar a 40Ar. 4 Apparent age 5 (øC) 4øAr?•Arx I (X100) (X100) 4øAr*/a•Ara (%) (%) K/Ca (Ma)

Sample 267, muscovite, Laurel Lake alteration zone, 90-50 gg, 0.0999 g, J = 0.012029

450 123.5 6.024 5.453 107.4 0.38 86.94 9 1496 _ 20 525 129.6 7.660 1.816 124.2 0.39 95.84 7 1649 _ 10 600 133.3 103.9 0.9633 130.6 1.01 97.90 0.5 1703 _ 8 675 136.7 72.74 0.2525 136.0 2.98 99.48 0.7 1749 _ 8 725 136.4 1.564 0.0683 136.1 4.39 99.83 30 1750 _ 16 775 136.3 0.1851 0.0380 136.2 5.75 99.90 300 1750 _ 14

825 135.6 0.0478 0.0531 135.4 11.27 99.87 1,000 1744 _ 16 875 136.7 3.116 0.0445 136.6 12.18 99.89 20 1753 _ 12 925 135.6 3.157 0.0592 135.4 21.22 99.85 20 1744 _ 4 975 135.4 2.902 0.0969 135.1 19.87 99.77 20 1741 _ 8 1025 136.3 3.103 0.0711 136.1 16.52 99.83 20 1750 _ 4 Fuse 138.5 0.3048 0.2037 137.8 4.03 99.55 200 1764 _ 10

Total gas age 1745 Plateau age (675-1025øC) 1746 _ 10

Sample 297, muscovite, Laurel Lake vein, 70-150 gm, 0.1123 g, J = 0.012168

500 142.8 0.8385 2.084 136.6 0.70 95.67 60 1766 ñ 4 600 145.2 0.2416 0.3942 144.0 1.64 99.18 200 1827 ñ 4 675 138.0 2.921 0.0868 137.8 4.09 99.80 20 1776 ñ 6 750 135.9 0.2997 0.0672 135.7 6.36 99.84 200 1759 ñ 6 800 135.0 0.1745 0.0490 134.8 9.13 99.87 300 1752 ñ 6 840 135.0 1.4455 0.0660 134.7 12.90 99.84 40 1751 ñ 6 870 134.8 0.1584 0.0665 134.6 9.15 99.84 300 1750 ñ 6 900 135.0 0.4341 0.1182 134.6 4.91 99.72 100 1750 ñ 4 930 134.7 0.2785 0.0749 134.4 12.10 99.82 200 1749 ñ 4 960 134.8 0.1524 0.1098 134.5 12.02 99.74 300 1749 ñ 6

990 135.8 0.0427 0.1700 135.2 8.54 99.61 1,000 1755 ñ 4 1020 135.7 0.0019 0.1078 135.4 9.00 99.75 >2,000 1756 ñ 4 Fuse 136.3 0.2094 0.1690 135.8 9.48 99.61 300 1760 ñ 4

Total gas age 1755 PlYeau age (750øC-•se) 1750 ñ 10

i Corrected for line blanks of atmospheric Ar composition (1 X 10 -14 moles 4øAr for T < 1,200øC and 2 x 10 -14 moles > 1,200øC) and 37Ar decay

g Correction factors used (a6Ar?7Ar)ca = 2.70 x 10 -4, (3•Ar?7Ar)ca = 6.51 X 10 -4, and (4øAr/agAr)x = 0.039 a Percent to total 39Ar released by fraction 4 Percent of radiogenic 4øAr in each fraction 5 Corrected for error in J value: •, = 5.543 x 10-1ø; monitor used was MMhb-I amphibole standard with an age of 518.9 Ma; errors

reported at 2• level

Gold grade is generally correlated with sulfide content, although gold may occur as inclusions in sulo fides, along grain boundaries, or in fractures, suggesting that initial gold deposition may be related to sulfide deposition. However, some of the gold may postdate sulfide deposition or represent remobiliza- tion of early gold during continued deformation and fluid events. In low-salinity, high-temperature mesothermal gold systems gold is transported as a reduced sulfur species (AuHS ø or Au(HS) -2) (Seward, 1989; Romberger, 1991). The O, H, S, and C isotope compositions and temperatures of the hydrothermal fluids are uniform, there are no known substantial compositional differences between early or late CO2- bearing fluid inclusions in any of the occurrences examined, and gold occurrences are found in all litholo-

gies in the western Flin Flon domain, implying that dilution, wall-rock sulfidation, oxidation, or lowering temperature are not important mechanisms for desta- bilizing reduced sulfur complexes. However, spo- radic variations in H•O-CO• phase ratios and the bulk fluid compositions and pressure and temperature conditions (Fig. 12) do suggest that phase separation may have occurred, at least intermittently, during the development of the gold occurrences. The un- mixing of CO• and H•O, and the loss of CO2 to the adjacent wall rocks to form carbonate, results in par- titioning of H•S into the vapor phase, which destabi- lizes any Au-S complex, or in an increase in pH in the aqueous fluid. Romberger (1986) suggests that with an increase in pH, pyrite and gold may coprecipitate as auriferous pyrite. Later recrystallization of the py-


Rio (AJ zone) Rio (PLG Zone)

\ - ß 2 -

" A B

0 I I I I I I I 100 200 300 400 500

Temperature (o C)

FIG. 12. Pressure-temperature conditions of mineralization at the Rio (AJ and PLG•Phantom Lake granite zones) and Graham occurrences. The COz isochores for densities of 0.75, 0.78, 0.90, and 0.98 g/cc and the COz-HzO-NaC1 isochore for a bulk density of 0.81 g/cc •e calculated using the program of Nicholls and Crawford (1985). Solvi in the H•O-COz-NaC1 system (A- H•OsoCO•s0, B-H•O49.s-CO•s0-NaC0.•) are taken from Bowers and Helgeson (1983). Temperatures are constrained using oxygen isotope geothermometry, and bulk densities and composition of inclusions are from Table 4. Note that the HzO-CO• solvus moves to higher temperatures with increasing salinity. If the temperature of mineralization is at a higher temperature than the solvus, then no HzO-COz phase sep•ation should occur.

rite allows the formation of discrete gold grains as inclusions within pyrite and also along grain boundaries.

Controls on the oxygen isotope composition of quartz veins

In the western Flin Flon domain the oxygen isotope compositions of mineralized quartz veins range from 9.9 to 14.4 per mil (Fig. 9), although the range of values seen in specific gold occurrences or areas is much more restricted (Table 5). This variation could be the result of differences in the temperature of the quartz-depositing fluid from area to area, variations in the isotopic composition of the fluid as a result of interactions with source or conduit rocks of different

isotopic compositions, different fluid/rock ratios, or some combination of the above. A range of temperature of about 200øC would be required if the regional variations are solely the result of differences in temperature, assuming a uniform b•sO fluid. The equilibrium temperatures calculated using oxygen isotope pairs are 365 ø to 400øC for the Rio deposit and 420øC for the Graham deposit. Fedorowich et al.

(1991) obtained a temperature of 370 ø _ 40øC for hydrothermal fluids at the Tartan Lake gold occurrence using oxygen isotope geothermometry. Within analytical error, the temperatures for the three occurrences are similar. Thus, temperature variations are an unlikely explanation for the regional variations in the isotopic composition of quartz veins.

The gold occurrences are spatially associated with shear zones which acted as fluid conduits. An end-

member scenario is that the isotopic composition of the fluids passing along these shear zones is controlled primarily by the isotopic composition of the source area, assuming low water/rock ratios in the source area and a high water/rock ratio in the shear zones. Kerrich (1989) proposed this mechanism for fluids passing along the Destor-Porcupine and Larder Lake breaks in the Abitibi belt, Canada, which pro- duced quartz that is not in isotopic equilibrium with the host rocks. However, there is a general correla- tion between blSO values of quartz veins and host rocks in the Foothills metamorphic belt, California (Bohlke and Kistler, 1986), and Charters Towers, Queensland (Peters and Golding, 1989), indicating that the isotopic composition of the quartz veins in these areas is dominantly controlled by interaction with the host rock. Barren veins from the Flin Flon domain have very limited alteration halos, suggesting that the vein fluids were approximately in chemical equilibrium with the wall rock, whereas many of the mesothermal gold occurrences in the Flin Flon area have larger, albeit limited, alteration halos indicative of a fluid not in equilibrium with the immediate wall rock. However, the bxso values of quartz from both auriferous and nonauriferous veins do vary with the isotopic composition of their host rocks (Fig. 9), which indicates that all fluids in the Flin Flon area must have interacted with rocks similar to those that

host the deposits. The relatively small alteration halos in most of the Flin Flon occurrences also imply that only limited quantities of fluid permeated the shear zones so that the rocks in the immediate envi-

rons of the shear zone controlled the isotopic and chemical composition of the fluids. Such low effec- tive water/rock ratios in the shear zones may have been a critical factor in determining why gold mineralization in the Flin Flon area is limited.

Tectonic setting of gold mineralization Mesothermal gold and associated hydrothermal al-

teration in the western Flin Flon domain postdates all observed intrusive events in the area and overprints peak regional metamorphic mineral assemblages (Fig. 13). Stable and radiogenic isotope systematics indicate that the likely source of the hydrothermal fluids is through devolatilization reactions during metamorphism. Since the alteration and veining associated with gold mineralization postdates peak re-

PROTEROZ01C MESOTHERMAL GOLD, FLIN FLON DOMAIN, CANADA 1519

1900 1880 1860 1840 1820 1800 1780 1760 1740 1720 1700

Amisk Group (1) VOLCANISM

....... Cliff Lake (1) -- Annabel Lake/Reynard Lake/Missi Island (2) PLUTONISM

-- Neagle Lake (2)

-- Phantom Lake (2) MOLASSE

Missi Formation, Flin F;on Basin (3) SEDIMENTATION

DEFORMATION

P1 _ _ _-•.•.•.p2 BuriaI P3 ••Uplift

P4

Ma

Peak regional P5 metamorphism

• Cooling

METAMORPHISM Peak thermal Low-grade '• • Laurel Lake area (4) conditions in • Ar-Ar low-grade High-grade (1) areas (2)

MESOTHERMAL GOLD -- Tartan Lake Ar-Ar (5) MINF RALIZATION

Rio Rb-Sr (4)

FIC. 13. Sequence of geologic events in the western Flin Flon domain. Solid lines represent precise ages; dashed lines represent imprecise ages or inferred age ranges. The plutons are named in Figure 2. References: (1) Gordon et al. (1990); (2) Ansdell and Kyser (1991b); (3) Ansdell et al. (1991); (4) this study; (5) Fedorowich et al. (1991).

gional metamorphism in the western Flirt Flon domain, the fluids must be derived from an external source undergoing prograde metamorphism at the time of mineralization. England and Thompson (1984) show that in a terrane undergoing crustal thickening, peak metamorphic conditions are achieved later at greater depth. Thus, continued de- gassing of amphibolite and granulite grade rocks underlying the presently exposed Flirt Flon domain may be the source of the mineralizing fluids. Lewry et al. (1990) and Bickford et al. (1990) speculate that the Flirt Flon domain was thrust over the Hanson Lake

block along the Sturgeon Weir shear zone (Fig. 1), and so a possible source of metamorphic fluids may be from medium-grade devolatilization reactions during thermal equilibration of Proterozoic rocks in the Hanson Lake block after overthrusting by the Flin Flon domain. However, the structure and geologic history of the underlying crust of the Flirt Flon domain is not known at present.

Comparisons with Archcan Mesothermal Gold Deposits

In general, mesothermal gold occurrences in the western Flin Flon domain exhibit many similarities with the giant Archcan deposits. They are all spatially associated with shear zones within the brittle-ductile transition zone, and there are local constraints on the locations of the deposits, e.g., competency contrasts,

shear zone jogs. The gold-bearing quartz veins in the western Flin Flon domain, Saskatchewan, are surrounded by alteration envelopes, albeit small, of carbonate, albite, pyrite, muscovite, quartz, and chlorite, as a result of interaction with structurally controlled, low-salinity H20-CO2 hydrothermal fluids, at about 350 ø to 400øC and 2 kbars.

One major difference is that the amount of gold discovered in the western Flirt Flon domain is small

in comparison with giant Archcan deposits. The mesothermal gold reserves total about 2.5 metric tons for the western Flin Flon domain, which is less than 1 percent of the gold production at the Timmins camp (1,700 t), Abitibi greenstone belt, and at the Golden Mile deposit (1200 t), Yilgarn block, Australia. The Tartan Lake gold deposit, western Manitoba (Fedoro- wich et al., 1991), adds only a further 5.4 metric tons of Au to the above reserve estimate for the western

Flirt Flon domain. Fyfe and Kerrich (1984) and Phil- lips et al. (1987) attempted to constrain the source volume required to produce the Au and other solutes in deposits of the Timmins and Golden Mile camps. They estimated that 1,200 km a of rock at reasonable average crustal concentrations of gold, and reasonable gold extraction and gold depositional efficien- cies would be enough to produce a giant gold camp. The western Flin Flon domain (Fig. 2) has an areal extent of about 1,200 km 2 and so would be large enough to form at least one large gold deposit. The


implication from these simple calculations is that a large mesothermal gold deposit might be present in the western Flin Flon domain. However, this would likely require focusing of most of the available mineralizing fluids into a single structural zone rather than having them dissipated along a large number of shear zones.

The giant Archean gold deposits, and many Pha- nerozoic analogues, are associated with crustal-scale terrane boundary faults (Kerrich and Wyman, 1990) that represent structural zones active during the latter stages of arc-continent, or continent-continent collision (Barley et al., 1989; Hodgson and Hamilton, 1989). The spatial association of shoshonitic lamprophyres and mesothermal gold is used as evidence for a more specific tectonic setting, namely oblique subduction in a transpressive-collisional tectonic setting (Wyman and Kerrich, 1988). The presence of mantle-derived lamprophyric rocks; radiogenic isotope data indicating that many deposits must have a solute source that includes more radiogenic, and thus older, rocks than those exposed at surface (e.g., King and Kerrich, 1989; Groves and McNaughton, 1991); and geochronological evidence linking hydrothermal activity to magmatic and metamorphic events in the lower crust or subcreted crust, especially in the Abi- tibi belt (e.g., Corfu, 1987), have been used to em- phasize the probable importance of a deep and heter- ogeneous source for the gold in Archean deposits (e.g., Colvine, 1989; Fyon et al., 1989). In the western Flin Flon domain, the absence of lamprophyres associated with the shear zones may indicate that the shear zones are not trans-crustal features, and the isotopic data presented above suggest that the fluid and solutes are derived during the prograde metamor- phisin of rocks similar in age and composition to those exposed on surface.

The extent of alteration and relatively low gold contents of the occurrences in the western Flin Flon

domain suggest that the fluid flux through these vein systems was probably small in relation to the giant Archean deposits. This may be an indication that limited quantities of fluids were sampled in the western Flin Flon domain relative to the Archean mesother-

mal gold deposits. There is, at present, no indication that there was more than one substantial hydrothermal event affecting the shear zones in the western Flin Flon domain. However, in the Abitibi belt, Claou•-Long et al. (1989) suggested that gold was introduced shortly (10-20 m.y.) after peak metamorphism at about 2680 Ma, which was then upgraded during a later hydrothermal event related to intrusion of S-type collisional granites at about 2645 Ma (Feng and Kerrich, 1991). Alternatively, Bell et al. (1989), Jemielita et al. (1990), and Wong et al. (1991) suggest that hydrothermal activity and related transport and deposition of gold occurred be-

tween 80 and 240 m.y. after peak metamorphism. A multistage concentrating process may thus be important in the development of giant mesothermal gold deposits. It is possible that the small gold occurrences in the western Flin Flon domain are similar to the less

spectacular quartz veins in the internal portions of the Abitibi greenstone belt where fluid flow was min- imal relative to that in the major terrane-bounding shear zones.

Conclusions

The Flin Flon domain is a Proterozoic example of a greenstone belt, and as in several Archean examples, hosts world-renowned volcanogenic massive sulfide deposits, e.g., Flin Flon, Snow Lake. The western part of the Flin Flon domain also hosts a large number of mesothermal gold occurrences, all of which are subeconomic at present but which exhibit some characteristics typical of giant Archean mesothermal gold deposits.

The gold-bearing systems are associated with brittle veins in brittle-ductile shear zones that developed after the peak ofmetamorphism. Veins frequently developed close to contacts between lithologies exhibiting ductility contrasts or in areas where shear zones change orientation. Vein mineralogy consists dominantly of quartz, pyrite, and ankerite, with minor tourmaline, albite, chlorite, and arsenopyrite, and is surrounded by variably developed but spatially limited alteration envelopes of quartz, ankerite, albite, pyrite, muscovite, and chlorite, the actual modal pro- portions being a function of host-rock composition and alteration intensity. The presence of alteration halos, albeit restricted in extent, indicates that the mineralizing fluids usually were not in chemical equilibrium with the host rocks, whereas barren veins formed from fluids that were probably more locally derived and more nearly in equilibrium with the surrounding rocks.

Fluid inclusion data indicate that the mineralizing fluids were dominantly CO2-H20-NaC1 in composition, although the abundance of CO2-rich inclusions is greater than that for typical Archean deposits. Fluid inclusion salinities are variable and range from 0.6 to 14.7 wt percent NaC1 equiv. Temperatures of mineralization are constrained by oxygen isotope mineral pairs to about 400øC, which when used with isochores based on fluid inclusion density estimates, suggest that the pressure of mineralization was approximately 2 kbars. Although H•O-CO2 phase ratios are relatively constant, there are areas in which the phase ratios are variable, which may indicate that fluid immiscibilit• was operating during the development of the veins. Changes in fluid chemistry, in particular an increase in pH, during phase separation and local wall-rock interaction may have resulted in destabilization of gold-sulfur complexes.

PROTEROZOIC MESOTHERMAL GOLD, FLIN FLON DOMAIN, CANADA 15 21

The hydrothermal fluid was probably derived by devolatilization during prograde metamorphism. Sul- fur, carbon, and strontium isotope analyses, though not equivocal, lend support to the derivation of solutes by leaching during metamorphism. The fluids and solutes were likely derived by prograde metamorphism of rocks similar in age and composition to those that presently host the gold occurrences.

Mesothermal gold mineralization in the western Flin Flon domain postdates 1834 ___ 13 Ma, the age of the youngest intrusive rocks in the area. An Rb-Sr mineral isochron for the Rio deposit of 1760 ___ 9 Ma is similar to 4øAr/S9Ar plateau ages for vein and alteration muscovites from the Laurel Lake Au-Ag deposit of 1753 ___ 10 and 1746 ___ 10 Ma, respectively. The age of about 1760 Ma represents the time of muscovite and tourmaline growth from an auriferous fluid at the Rio deposit which was approximately contemporaneous with the thermal event that affected the Ar-Ar systematics of the muscovites at the Laurel Lake deposit. An age of 1791 ___ 3 Ma from the Tartan Lake deposit, Manitoba (Fedorowich et al., 1991), implies that fluid flow along the shear zones, and related thermal effects, probably occurred periodically over a period of at least 30 m.y.

Although the shear zones hosting the gold occurrences were active during the latter stages of the collisional orogeny in the Flin Flon area, the lack of lam- prophyre dikes and of isotopic evidence for a fluid and solute source area distinctly different from the rocks exposed at surface suggests that these shear zones do not represent transcrustal structures. The small size of the mesothermal gold occurrences thus may be a function of the smaller rock and fluid volume sampled by each shear zone, when compared to terrane-bounding faults, such as the Larder Lake- Cadillac or Boulder-Lefroy faults, in Archean greenstone belts.

Acknowledgments CAMECO and Vista Mines are thanked for allow-

ing access to many of the gold occurrences. V. So- puck, D. Jiricka, F. McDougall, M. Koziol, K. Dies, W. Coombe, K. Zazulak, and B. Hughes are thanked for logistical support and providing much useful information in the field.

Technical assistance at the University of Saskatche- wan was provided by Z. Szczepanik and B. Nova- kovski (thin sections), R. George (electron micro- probe), D. Pezderic (stable isotopes), and A. Vuletich (radiogenic isotopes). The Ar-Ar analyses were kindly performed by K. Foland at Ohio State Univer- sity.

Financial support for this project was provided by CAMECO and the Natural Sciences and Engineering Research Council through a University-Industry Re- search Grant, and by Natural Science and Engineer-

ing Research Council operating and infrastructure grants to TKK and R. Kerrich. KMA acknowledges receipt of a University of Saskatchewan graduate scholarship.

The manuscript has benefited from early reviews by P. Barton, R. Kerrich, J. Richards, D. O'Hanley, and J. Fedorowich.

August 19, 1991; February 10, 1992 REFERENCES

Aggarwal, P. K., and Longstaffe, F. J., 1987, Oxygen-isotope geochemistry of metamorphosed massive sulfide deposits of the Flin Flon-Snow Lake belt, Manitoba: Contr. Mineralogy Petrology, v. 96, p. 314-325.

Aggarwal, P. K., and Nesbitt, B. E., 1987, Pressure and temperature conditions of metamorphism in the vicinity of three massive sulfide deposits, Flin Flon-Snow Lake belt, Manitoba: Canadian Jour. Earth Sci., v. 24, p. 2305-2315.

Ansdell, K. M., and Kyser, T. K., 1990, Temporal evolution of magmatism in a Proterozoic greenstone belt: Single-zircon Pb- evaporation data from the Flin Flon domain, Canada labs.I: Geol. Soc. America Program with Abstracts, v. 22, p. A262.

-- 1991 a, The geochemistry and fluid history of the Proterozoic Laurel Lake Au-Ag deposit, Flin Flon greenstone belt: Canadian Jour. Earth Sci., v. 28, p. 155-171.

-- 1991b, Plutonism, deformation and metamorphism in the Proterozoic Flin Flon greenstone belt, Canada: Limits on timing provided by the single-zircon Pb-evaporation technique: Geol- ogy, v. 19, p. 518-521.

Ansdell, K. M., Kyser, T. K., Stauffer, M., and Edwards, G., 1991, Detrital zircons from Missi metasedimentary rocks, Flin Flon basin: Constraints on age and provenance: Saskatchewan Geol. Survey, Saskatchewan Energy Mines Misc. Rept. 91-4, p. 157- 161.

Ashton, K. E., 1990, Geology of the Snake Rapids area, Flin Flon domain (Parts of NTS 63L-9 and-10): Saskatchewan Geol. Sur- vey, Saskatchewan Energy Mines Misc. Rept. 90-4, p. 4-12.

Ashton, K. E., Wilcox, K. H., Wheatley, K. J., Paul, D., and de Tombe, J., 1987, The boundary zone between the Flin Flon domain, Kisseynew gneisses and Hanson Lake block, northern Saskatchewan: Saskatchewan Geol. Survey, Saskatchewan En- ergy Mines Misc. Rept. 87-4, p. 131-134.

Bailes, A. H., and Syme, E. C., 1989, Geology of the Flin Flon- White Lake area: Manitoba Dept. Energy Mines, Geol. Services Branch, Geol. Rept. GR 87-1, 313 p.

Bailes, A. H., Syme, E. C., Galley, A. G., and Skirrow, R. G., 1987, Early Proterozoic volcanism, hydrothermal activity, and associated ore deposits at Flin Flon and Snow Lake, Manitoba: Sas- katoon Geol. Assoc.-Mineralog. Assoc. Canada Jt. Ann. Mtg., May 19-24, 1987, Saskatoon, Field Trip 1 Guidebook, 95 p.

Barley, M. E., Eisenlohr, B. N., Groves, D. I., Perring, C. S., and Vearncombe, J. R., 1989, Late Archean convergent margin tectonics and gold mineralization: Evidence from the Norseman- Wiluna belt: Geology, v. 17, p. 826-829.

Bell, K., Anglin, C. D., and Franklin, J. M., 1989, Sm-Nd and Rb-Sr isotope systematics of scheelites: Possible implications for the age and genesis of vein-hosted gold deposits: Geology, v. 17, p. 5OO-5O4.

Bickford, M. E., Collerson, K. D., Lewry, J. F., Van Schmus, W. R., and Chiarenzelli, J., 1990, Proterozoic collisional tectonism in the Trans-Hudson orogen, Saskatchewan: Geology, v. 18, p. 14- 18.

Bigeleisen, J., Pearlman, M. L., and Prosser, H. C., 1952, Conver- sion of hydrogenic materials to hydrogen for isotopic analysis: Anal. Chemistry, v. 24, p. 1356-1357.

Blamart, D., Pichavant, M., and Sheppard, S. M. F., 1989, Experi- mental determination of the D/H isotopic fractionation between


tourmaline and water at 600, 500øC and 3 kbar: Acad. Sci. [Paris] Comptes Rendus, v. 308(II), p. 39-44.

Bohlke, J. K., 1989, Comparison of metasomatic reactions between a common CO2-rich vein fluid and diverse wall rocks: Intensive variables, mass transfers, and Au mineralization at A1- leghany, California: Eco•. GEoid., v. 84, p. 291-327.

Bohlke, J. K., and Kistler, R. W., 1986, Rb-Sr, K-Ar, and stable isotope evidence for the ages and sources of fluid components of gold-bearing quartz veins in the northern Sierra Nevada foothills metamorphic belt, California: Ecoa. GEoid., v. 81, p. 296- 322.

Bowers, T. S., and Helgeson, H. C., 1983, Calculation of the ther- modynamic and geochemical consequences of nonideal mixing in the system H20-CO•-NaC1 on phase relations in geologic systems: Equation of state for H•O-CO•-NaC1 fluids at high pressures and temperatures: Geochim. et Cosmochim. Acta, v. 47, p. 1247-1275.

Bruce, E. L., 1918, Amisk-Athapapuskow Lake district: Canada Geol. Survey Mem. 105, 91 p.

Burruss, R. C., 1981, Analysis of fluid inclusions: Phase equilibria at constant volume: Am. Jour. Sci., v. 281, p. 1104-1126.

Byers, A. R., and Dahlstrom, C. D. A., 1954, Geology and mineral deposits of the Amisk-Wildnest Lakes area, Saskatchewan: Sas- katchewan Dept. Mineral Resources, Geol. Sci. Branch, Precam- brian Geology Div. Rept. 14, 177 p.

Byers, A. R., Kirkland, S. J. T., and Pearson, W. J., 1965, Geology and mineral deposits of the Flin Flon area, Saskatchew•,n: Sas- katchewan Dept. Mineral Resources Rept. 62, 95 p.

Card, K. D., Poulsen, K. H., and Robert, F., 1989, The Archcan Superior province of the Canadian Shield and its lode gold deposits: Eco•. G•oI•. Mo•. 6, p. 19-36.

Chauvel, C., Arndt, N. T., Kielinczuk, S., and Thom, A., 1987, Formation of Canadian 1.9 Ga old continental crust. I: Nd isotopic data: Canadian Jour. Earth Sciences, v. 24, p. 396-406.

Claou•-Long, J. C., King, R. W., and Kerrich, R., 1990, Archaean hydrothermal zircon in the Abitibi greenstone belt: Constraints on the timing of gold mineralisation: Earth Planet. Sci. Letters, v. 98, p. 109-128.

Clayton, R. N., and Mayeda, T. K., 1963, The use of bromine pen- tafiuoride in the extraction of oxygen from oxides and silicates for isotopic analysis: Geochim. et Cosmochim. Acta, v. 27, p. 43-52.

Clayton, R. N., O'Neil, J. R., and Mayeda, T. K., 1972, Oxygen isotope exchange between quartz and water: Jour. Geophys. Re- search, v. 77, p. 3057-3067.

Collins, P. L. F., 1979, Gas hydrates in CO•-bearing fluid inclusions and the use of freezing data for estimation of salinity: Eco•. G•oI•., v. 74, p. 1435-1444.

Colvine, A. C., 1989, An empirical model for the formation of Archcan gold deposits: Products of final cratonization of the Su- perior province, Canada: Eco•. G•oI•. Mo•. 6, p. 37-53.

Coombe, W., 1984, Gold in Saskatchewan: Saskatchewan Geol. Survey Open File Rept. 84-1, 134 p.

Corfu, F., 1987, Inverse age stratification in the Archaen crust of the Superior province: Evidence for intra- and subcrustal accretion from high-resolution U-Pb zircon and monazite ages: Pre- cambrian Research, v. 36, p. 259-275.

Curti, E., 1987, Lead and oxygen isotope evidence for the origin of the Monte Rosa gold lode deposits (western Alps, Italy): A comparison with Archcan lode deposits: Eco•. G•oI•., v. 82, p. 2115-2140.

Dallmeyer, R. D., and Keppie, J. D., 1987, Polyphase late Paleo- zoic tectonothermal evolution of the southwestern Meguma terrane, Nova Scotia: Evidence from 4øAr/•9Ar mineral ages: Ca- nadian Jour. Earth Sci., v. 24, p. 1242-1254.

Delaney, G., Carr, S. D., and Parrish, R. R., 1988, Two U-Pb zircon ages from eastern Glennie Lake domain, Trans-Hudson orogen, Saskatchewan: Canada Geol. Survey Paper 88-2, p. 51-58.

Digel, S., and Gordon, T. M., 1991, Prehnite-pumpellyite to am-

phibolite facies metamorphism near Flin Flon, Manitoba: Can- ada Geol. Survey Paper 91-1C, p. 165-172.

England, P. C., and Thompson, A. B., 1984, Pressure-temperature-time paths of regional metamorphism I. Heat transfer during the evolution of regions of thickened continental crust: Jour. Petrology, v. 25, p. 894-928.

Fedorowich, J., Stauffer, M., and Kerrich, R., 1991, Structural setting and fluid characteristics of the Proterozoic Tartan Lake gold deposit, Trans-Hudson orogen, northern Manitoba: Ecoa. GEoid., v. 86, p. 1434-1467.

Feng, R., and Kerrich, R. W., 1991, •2645 Ma S-type collisional granites: Their influence on gold mineralization in the Abitibi southern volcanic zone [abs.]: Geol. Assoc. Canada Program with Abstracts, v. 16, p. A36.

Foland, K. A., Linden, J. S., Laskowski, T. E., and Grant, N. K., 1984, 4øAr/•gAr dating of glauconite: Measured •"r recoil loss from well-crystallized specimens: Chem. Geology, v. 2, p. 241- 264.

Fyfe, W. S., and Kerrich, R., 1984, Gold: Natural concentration processes, in Foster, R. P., Gold, 82: Rotterdam, A. A. Balkema Pub., p. 99-127.

Fyon, J. A., Troop, D. G., Marmont, S., and Macdonald, A. J., 1989, Introduction of gold into Archcan crust, Superior province, Ontario--coupling between mantle-initiated magmatism and lower crustal thermal maturation: ECON. GEOL. MON. 6, p. 479-490.

Galley, A. G., and Franklin, J. M., 1987, Geological setting of gold, copper, tungsten and molybdenum occurrences in the Phantom Lake region, Summary of investigations 1987: Saskatoon Geol. Survey Misc. Rept. 87-4, p. 115-124.

Gaskarth, J. W., and Parslow, G. R., 1987, Proterozoic volcanism in the Flin Flon greenstone belt, east-central Saskatchewan, Canada: Geol. Soc. London Spec. Pub. 33, p. 183-200.

Goldfarb, R. J., Leach, D. L., Pickthorn, W. J., and Paterson, C. J., 1988, Origin of lode-gold deposits of the Juneau gold belt, south- eastern Alaska: Geology, v. 16, p. 440-443.

Gordon, T. M., Hunt, P. A., Bailes, A. H., and Syme, E. C., 1990, U-Pb ages from the Flin Flon and Kisseynew belts, Manitoba: Chronology of crust formation at an early Proterozoic accretionary margin: Geol. Assoc. Canada Spec. Paper 37, p. 177-200.

Green, A. G., Hajnal, Z., and Weber, W., 1985, An evolutionary model of the western Churchill province and western margin of the Superior province in Canada and the north-central United States: Tectonophysics, v. 116, p. 281-322.

Groves, D. I., and McNaughton, N.J., 1991, Archcan crustal-scale plumbing systems and the formation of subgreenschist to granulite-hosted lode-gold deposits [abs.]: Geol. Assoc. Canada Pro- gram with Abstracts, v. 16, p. A48.

Groves, D. I., and Phillips, G. N., 1987, The genesis and tectonic control on Archcan gold deposits of the Western Australian Shield: A metamorphic-replacement: Ore Geology Rev., v. 2, p. 287-322.

Hodgson, C. J., and Hamilton, J. V., 1989, Gold mineralization in the Abitibi greenstone belt: End-stage results of Archcan collisional tectonics? Eco•. G•oI•. Mo•. 6, p. 86-100.

Hoers, J., 1987, Stable isotope geochemistry, 3rd ed.: New York, Springer-Verlag, 241 p.

Hoffman, P. F., 1988, United plates of America, the birth of a craton: Early Proterozoic assembly and growth of proto-Lauren- tia: Ann. Rev. Earth Planet. Sci., v. 16, p. 543-603.

Ibrahim, M. S., and Kyser, T. K., 1991, Fluid inclusion and isotope systematics of the high-temperature Proterozoic Star Lake lode gold deposit, northern Saskatchewan, Canada: Ecoa. GEOL., v. 86, p. 1468-1490.

Jemielita, R. A., Davis, D. W., and Krogh, T. E., 1990, U-Pb evidence for Abitibi gold mineralization postdating greenstone magmatism and metamorphism: Nature, v. 346, p. 831-834.

Kerrich, R., 1987, The stable isotope geochemistry of Au-Ag vein deposits in metamorphic rocks: Mineralog. Assoc. Canada Short Course Handbook, v. 13, p. 287-336.


-- 1989, Geochemical evidence on the sources of fluids and solutes for shear zone hosted mesothermal Au deposits: Geol. As- soc. Canada Short Course Notes, v. 6, p. 129-198.

Kerrich, R., and Wyman, D., 1990, Geodynamic setting of mesothermal gold deposits: An association with accretionary tectonic regimes: Geology, v. 18, p. 882-885.

King, R. W., 1990, Tourmaline from mesothermal gold deposits of the Superior province, Canada. Textural, chemical and isotopic evidence: Unpub. Ph.D. thesis, Univ. Saskatchewan, 228 p.

King, R. W., and Kerrich, R. W., 1989, Strontium isotope compositions of tourmaline from lode gold deposits of the Archcan Abi- tibi greenstone belt (Ontario-Quebec, Canada): Implications for source reservoirs: Chem. Geology, v. 79, p. 225-240.

Kontak, D. J., Smith, P. K., Kerrich, R., and Williams, P. F., 1990, Integrated model for Megurea Group lode gold deposits, Nova Scotia, Canada: Geology, v. 18, p. 238-242.

Kyser, T. K., 1987, Equilibrium fractionation factors for stable isotopes: Mineralog. Assoc. Canada Short Course Handbook, v. 13, p. 287-336.

Kyser, T. K., and O'Neil, J. R., 1984, Hydrogen isotope systematics of submarine basalts: Geochim. et Cosmochim. Acta, v. 48, p. 2123-2133.

Kyser, T. K., Janser, B. W., Wilson, M. R., and Hattie, I., 1986, Stable isotope geochemistry related to gold mineralization and exploration in the western shield, in Clark, L. A., ed., Gold in the western shield: Saskatoon, Canadian Inst. Mining Metallurgy, p. 470-498.

Lewry, J. F., Thomas, D. J., Macdonald, R., and Chiarenzelli, J., 1990, Structural relations in accreted terranes of the Trans- Hudson orogen, Saskatchewan: Telescoping in a collisional re- gime? Geol. Assoc. Canada Spec. Paper 37, p. 75-94.

Longiaru, S. J., 1980, Structure and metamorphism of the north- east Amisk Lake area, Saskatchewan: Unpub. M.Sc. thesis, Univ. Saskatchewan, 119 p.

Matthews, A., and Katz, A., 1977, Oxygen isotope fractionation during dolomitization of calcium carbonate: Geochim. et Cos- mochim. Acta, v. 41, p. 1431-1438.

McCrea, J. M., 1950, On the isotope geochemistry of carbonates and a palcotemperature scale: Jour. Chem. Physics, v. 18, p. 849-857.

Mukherjee, A. C., Stauffer, M. R., and Baadsgaard, H., 1971, The Hudsonian orogeny near Flin Flon, Manitoba: A tentative inter- pretation of Rb/Sr and K/Ar ages: Canadian Jour. Earth Sci., v. 8, p. 939-946.

Nesbitt, B. E., Murowchick, J. B., and Muehlenbachs, K., 1986, Dual origin of lode gold deposits in the Canadian Cordillera: Geology, v. 14, p. 506-509.

Nicholls, J., and Crawford, M. L., 1985, FORTRAN programs for calculation of fluid properties from microthermometric data on fluid inclusions: Computers Geosci., v. 11, p. 619-645.

Ohmoto, H., and Rye, R. O., 1979, Isotopes of sulfur and carbon, in Barnes, H. L., ed., Geochemistry of hydrothermal ore deposits: New York, Wiley Intersci., p. 684-737.

O'Neil, J. R., and Taylor, H. P., Jr., 1969, Oxygen isotope fractionation between muscovite and water: Jour. Geophys. Research, v. 74, p. 6012-6022.

Pearson, J. G., 1980, Flin Flon gold project: Saskatchewan Geol. Survey Misc. Rept. 80-4, p. 70-80.

-- 1982, Gold metallogenic studies: Amisk Lake east area: Sas- katchewan Geol. Survey Misc. Rept. 82-4, p. 64-70.

-- 1983, Gold metallogenic studies: Flin Flon-Asisk Lake area: Saskatchewan Geol. Survey Misc. Rept. 83-4, p. 67-74.

-- 1984, Gold metallogenic studies: The Rio deposit: Saskatche- wan Geol. Survey Misc. Rept. 84-4, p. 123-126.

Peters, S. G., and Golding, S. D., 1989, Geologic, fluid inclusion, and stable isotope studies of granitoid-hosted gold-bearing quartz veins, Charters Towers, northeastern Australia: Ecoa. GEOL. MON. 6, p. 260-273.

Phillips, G. N., Groves, D. I., and Brown, I. J., 1987, Source re- quirements for the Golden Mile, Kalgoorlie: Significance to the

metamorphic replacement model for Archcan gold deposits: Ca- nadian Jour. Earth Sci., v. 24, p. 1643-1651.

Potter, R. W., II, Clynne, M. A., and Brown, D. L., 1978, Freezing point depression of aqueous sodium chloride solutions: ECON. GEOL., v. 73, p. 284-285.

Rafter, T. A., 1965, Recent sulfur isotope measurements on a variety of specimens examined in New Zealand: Bull. Volcanology, v. 28, p. 3-20.

Roedder, E., 1979, Fluid inclusions as samples of ore fluids, in Barnes, H. L., ed., Geochemistry of hydrothermal ore deposits: New York, Wiley Intersci., p. 684-737.

Romberger, S. B., 1986, The solution chemistry of gold applied to the origin of hydrothermal deposits, in Clark, L. A., ed., Gold in the western shield: Saskatoon, Canadian Inst. Mining Metal- lurgy, p. 168-186.

-- 1991, Transport and deposition of gold in hydrothermal systems, in Robert, F., Sheahan, P. A., and Green, S. B., Greenstone gold and crustal evolution: Geol. Assoc. Canada, Mineral De- posits Div., Nuna Conf., Val d'Or, Canada, May 24-27, 1990, p. 61-66.

Rye, D. M., and Rye, R. O., 1974, Homestake gold mine, South Dakota: I. Stable isotope studies: ECON. GEOL., v. 69, p. 293- 317.

Sandiford, M., and Keays, R. R., 1986, Structural and tectonic constraints on the origin of gold deposits in the Ballarat slate belt, Victoria: Geol. Assoc. Canada Spec. Paper 32, p. 15-24.

Seward, T. M., 1989, The hydrothermal chemistry of gold and its implications for ore formation: Boiling and conductive cooling as examples: ECON. GEOL. MON. 6, p. 398-404.

Stauffer, M. R., 1984, Manikewan: An early Proterozoic ocean in central Canada, its igneous history and orogenic closure: Pre- cambrian Research, v. 25, p. 257-281.

-- 1990, The Missi Formation: An Aphebian molasse deposit in the Reindeer Lake zone of the Trans-Hudson orogen, Canada: Geol. Assoc. Canada Spec. Paper 37, p. 121-142.

Stauffer, M. R., and Mukherjee, A. C., 1971, Superimposed defor- mations in the Missi metasedimentary rocks near Flin Flon, Manitoba: Canadian Jour. Earth Sci., v. 8, p. 217-242.

Stauffer, M. R., Mukherjee, A. C., and Koo, J., 1975, The Amisk Group: An Aphebian(?) island arc deposit: Canadian Jour. Earth Sci., v, 12, p. 2021-2035.

Suzuoki, T., and Epstein, S., 1976, Hydrogen isotope fractionation between OH-bearing minerals and water: Geochim. et Cosmo- chim. Acta, v. 50, p. 1229-1240.

Swanenberg, H. E. C., 1979, Phase equilibria in carbonic systems, and their application to freezing studies of fluid inclusions: Contr, Mineralogy Petrology, v. 68, p. 303-306.

Syme, E. C., and Forester, R. W., 1977, Petrogenesis of the Bound- ary intrusions in the Flin Flon area of Saskatchewan and Mani- toba: Canadian Jour. Earth Sci., v. 14, p. 444-455.

Syme, E. C., Hunt, P. A., and Gordon, T. M., 1991, Two U-Pb zircon ages from the western Flin Flon belt, Trans-Hudson orogen, Manitoba: Canada Geol. Survey Paper 90-2, p. 25-34.

Taylor, H. P., Jr., 1974, The application of oxygen and hydrogen isotope studies to problems of hydrothermal alteration and ore deposition: ECON. GEOL., v. 69, p. 843-883.

Thom, A., Arndt, N. T., Chauvel, C., and Stauffer, M., 1990, Flin Flon and western La Ronge belts, Saskatchewan: Products of Proterozoic subduction-related volcanism: Geol. Assoc. Canada Spec. Paper 37, p. 163-176.

Thomas, D. J., 1989, Geology of the Douglas Lake-Phantom Lake area (part of NTS 63K-12 and -13): Saskatchewan Geol. Survey Misc. Rept. 89-4, p. 44-54.

Van Schmus, W. R., Bickford, M. E., Lewry, J. F., and Macdonald, R., 1987, U-Pb geochronology in the Trans-Hudson orogen, northern Saskatchewan, Canada: Canadian Jour. Earth Sci., v. 24, p. 407-424.

Watters, B. R., and Armstrong, R. L., 1985, Rb-Sr study of recta- volcanic rocks from the La Ronge and Flin Flon domains, northern Saskatchewan: Canadian Jour. Earth Sci., v. 22, p. 452-463.

15 2 4 K. M. ANSDELL AND T. K. K YSER

Wenner, D. B., and Taylor, H. P., Jr., 1971, Temperatures of ser- pentinization of ultramafic rocks based on 180/•60 fractionation between coexisting serpentine and magnetite: Contr. Mineral- ogy Petrology, v. 32, p. 165-185.

Wilcox, K. H., 1990, Geology of the Amisk-Welsh Lakes area, Sas- katchewan: Unpub. M.Sc. thesis, Univ. Calgary, 245 p.

Wong, L., Davis, D. W., Krogh, T. E., and Robert, F., 1991, U-Pb

zircon and rutile geochronology of Archean greenstone formation and gold mineralization in the Val d'Or region, Quebec: Earth Planet. Sci. Letters, v. 104, p. 325-336.

Wyman, D., and Kerrich, R., 1988, Alkaline magmatism, major structures, and gold deposits: Implications for greenstone belt gold metallogeny: E½ON. GEOId., v. 83, p. 454-461.

APPENDIX

Analytical

Microthermometric observations of fluid inclusions were conducted on doubly polished wafers of quartz of between 100 and 500 #m in thickness using a calibrated USGS-type gas-flow heating and cooling stage. CO2 melting and CO2 homogenization temperatures are reproducible to _+0.1øC, whereas other measurements are less precise with the maximum error being _+1.0øC. Imprecision also increases as the size of the fluid inclusions decrease.

Mineral separates were obtained using standard crushing, sieving, magnetic separation, heavy liquid, and handpicking techniques; they were checked for purity optically and by X-ray diffraction. Oxygen was quantitatively removed from silicate mineral separates using the BrF5 technique of Clayton and Mayeda (1963). Hydrogen and the water content of hydrous minerals were obtained using the techniques of Bigeleisen et al. (1952) and Kyser and O'Neil (1984). SO• was obtained from the sulfide mineral separates using the technique of Rafter (1965). CO• was extracted from ankerite by reaction with HaPO 4 (McCrea, 1950) at 50øC for 2 days. Oxygen, hydrogen, sulfur, and carbon isotope analyses were performed at the University of Saskatchewan, using conventional isotope ratio mass spectrometry. The re- suits are reported in units of per mil relative to

Techniques

Vienna standard mean ocean water for O and H (V- SMOW), Canyon Diablo troilite for S (CDT), and Peedee belemnite for C (PDB). Using these techniques the •80 value of NBS-28 quartz is 9.6 per mil, the •D value of NBS-30 biotite is -65 per mil, the •a4S value of NBS-123 sphalerite is 16.7 per mil, and •aC and •80 values of NBS-19 limestone are 1.9 and 28.6 per mil, respectively. Replicate analyses indicate that reproducibility at the 2 a confidence level is _+0.2 per mil for •sO, _+3.0 per mil for •D, _+0.2 per mil for •a4S, and _+0.1 per mil for •aC.

Tourmaline and muscovite from the Rio occurrence were leached in HC1 prior to being spiked and dissolved using an HF-HC104 mixture. Rubidium and strontium were separated by conventional cation exchange procedures, and the isotopic ratios determined on a Finnigan MAT 261 multicollector mass spectrometer run under static conditions. Replicate analyses of the NBS-987 Sr standard over the last year yielded a mean value of 0.710238 _+ 0.000027.

Ar-Ar isotope analyses were performed on muscovites from an altered quartz-feldspar porphyry, and a quartz-sulfide-muscovite vein at Laurel Lake, using irradiation, extraction, mass spectroscopic, and correction procedures described by Foland et al. (1984).

mesothermal quartz stuff - ansdell, kyser

Documents

gold grade

dilatant zones

brittleductile shear

granitic rocks

supracrustal rocks

fluid composition

dominant fluids

immediate host rocks