Fluid-inclusion characteristics of hydrothermal Cu–Ni–PGE ...abyss.elte.hu/users/molnar/cikkek/Chem Geol 1999b.pdf · The high-temperature hydrothermal fluids were driven by

Ž .Chemical Geology 154 1999 279–301

Fluid-inclusion characteristics of hydrothermal Cu–Ni–PGEveins in granitic and metavolcanic rocks at the contact of the

Little Stobie deposit, Sudbury, Canada

Ferenc Molnar a,b,), David H. Watkinson b, John O. Everest bá Department of Mineralogy, EotÕos L. UniÕersity, Budapest 1088, Hungary¨ ¨

b Department of Earth Sciences, Ottawa–Carleton Geoscience Centre, Carleton UniÕersity, Ottawa, Canada K1S 5B6

Abstract

Cu- and precious-metal-enriched massive ore and veins occur at contacts of two Fe–Ni–Cu sulfide orebodies of the LittleStobie Mine, Sudbury, with metavolcanic rocks of the Elsie Mountain Formation, and the Murray granite. Veins containchalcopyrite, pyrrhotite, pentlandite, platinum-group minerals, quartz, carbonate, chlorite, amphibole and other minerals. Agranitic dike, resulting from partial re-melting of the Murray Granite, cuts back into the Sudbury Igneous Complex nearLittle Stobie Mine, and contains barren veinlets with similar mineralogy but no sulfides. Quartz from ore veins contains fluid

Žinclusions that were trapped during several stages. Early high-temperature fluids at least 180–2708C in orebody 1 and.280–3508C in orebody 2 were extremely saline and occur as polyphase, isolated and fracture-controlled inclusions with

halite, sylvite, Fe–Mn-, as well as Pb–Ba-chloride daughter minerals and other unknown solids. Late secondary aqueousinclusions are not chloride-saturated; they were trapped at a minimum of 80–1508C. Their microthermometric behavior maybe modeled in the CaCl –NaCl–H O system with salinity of 21–27 CaCl equiv. wt.%. Very late secondary inclusions2 2 2

Ž .have either Ca-rich saline, or very dilute about 1 equiv. wt.% NaCl compositions and homogenization temperatures ofŽ .200–3008C. With the exception of these very late secondary inclusions, the association of CO – CH -rich inclusions with2 4

aqueous ones was usually observed. The density of these carbonic fluids decreased from early to late stages. Microthermo-metric data from barren veins are fundamentally different from those of early inclusions from orebodies; this implies thatthese heavy-metal-rich fluids were responsible for ore deposition in veins. The minimum pressure of entrapment for early

Ž .fluids was 1800–2200 bars. Late Ca-rich brines were trapped at lower minimum pressure 200–900 bars . High-pressuredata are in agreement with the estimated minimum paleodepth of crystallization of South Range ores. Later fluids wereprobably trapped during the uplift of orebodies and their host rocks. Comparison of data to other Cu–PGE–Au-rich ore ofthe Sudbury Structure suggests that the presence of CO -rich fluids in South Range deposits and their absence in North2

Range deposits may be related to different metamorphic histories. The high-temperature hydrothermal fluids were driven bythe heat of the Sudbury Igneous Complex; these very saline fluids interacted with primary magmatic ores, remobilizedmetals and redeposited them along convenient structures such as fracture zones and breccias in and along various units nearthe footwall contact. The identification of such highly saline fluid inclusions with high heavy-metal content may be useful in

) Corresponding author. Department of Mineralogy, Eotvos L. University, Muzeum Krt 4rA, Budapest 1088, Hungary. E-mail:¨ ¨ ´[email protected]

0009-2541r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved.Ž .PII: S0009-2541 98 00136-3

( )F. Molnar et al.rChemical Geology 154 1999 279–301´280

the exploration for Cu–PGE–Au-enriched, footwall vein-type ores in the Sudbury Structure. q 1999 Elsevier Science B.V.All rights reserved.

Keywords: Sudbury; Hydrothermal Ni–Cu–PGE ores; Heavy-metal rich fluid inclusions

1. Introduction

Ž .Magmatic Ni–Cu–platinum group element PGEores result from liquid immiscibility between sulfide

Ž .and silicate magmas Naldrett, 1989 ; however, it isclear that hydrothermal processes also play a role in

Ž .PGE and base-metal especially Cu enrichment inŽsome magmatic ore deposits Ballhaus and Stumpfl,

1986; Johan and Watkinson, 1987; Nyman et al.,1990; Watkinson, 1990; Farrow, 1994; Molnar et al.,´

.1997 .Ž .The Sudbury Igneous Complex SIC in northern

Ontario contains many Fe–Ni–Cu sulfide depositsalong the contact of the layered magmatic body withfootwall rocks. The magmatic-segregation origin ofthese massive sulfide orebodies in the Sublayer andFootwall Breccia to the Sudbury Igneous Complex is

Ž .clear Naldrett, 1989 . Some deposits, enriched in Cuand precious metals near the footwall contacts, maybe explained by fractional crystallization of sulfide

Ž .magma Ebel and Naldrett, 1996 . However, thereare also Cu–PGE–Au-enriched deposits in the foot-

Žwall rocks for example Strathcona Deep CopperZone, McCreedy West Cu–Ni veins along the NorthRange and Lindsley 4b deposit along the South

.Range that have hydrothermal wall-rock alterationselvages, and saline fluid inclusions occur in both

Žsilicate and sulfide minerals Farrow, 1994; Watkin-.son, 1994; Molnar et al., 1997 .´

In addition to the massive Sublayer ores, theLittle Stobie Mine along the South Range of the

Ž .Sudbury Structure Fig. 1 contains vein-type oreswith hydrothermal alteration of enclosing footwallrocks and enhanced PGE content. The purpose ofthis study is to outline the temperature–pressure–composition characteristics of fluids trapped in inclu-sions in quartz from these veins at and near thecontact of the Little Stobie orebodies with graniticand metavolcanic rocks, and in barren veins in theadjacent granitic rocks. These data are used to evalu-ate the possible role of fluids in enrichment of PGEin vein-type ores hosted by the footwall units of theSudbury Igneous Complex.

2. Geology, mineralogy and sampling

The Sudbury Igneous Complex is located in ameteorite impact-related elliptical structure along theboundary of the Archean gneissic rocks of the Supe-rior Province and Proterozoic metasedimentary–metavolcanic and granitic basement rocks of theSouthern Province of the Canadian Shield. The lay-ered sequence of the SIC is composed of norite,gabbro and granophyre from bottom to top, and is

Žemplaced in impact-related breccia Onaping Forma-.tion and subsequently deposited sedimentary rocks

Ž .Onwatin and Chelmsford Formations . Radiometricdata indicate a 1.85-Ga age of the emplacement of

Ž .the igneous rocks Krogh et al., 1984 . Disseminatedor massive Fe–Ni–Cu sulfide ores of the SIC occurin a discontinuous ‘sublayer’ around the basal con-tact of the norite with the basement rocks. Most

Žmagmatic deposits occur in breccias Contact Sub-.layer, Footwall Breccia, Offset Sublayer character-

ized by igneous matrices with mafic–ultramafic andfelsic clasts. Footwall-type deposits are located in thebasement units along both the North and South

Ž .Ranges Morrison et al., 1994 . These are vein-liketo massive bodies in brecciated zones of the footwallrocks with Cu and precious metal enrichment com-pared to ores at the basal contact of the SIC.

The North Range of the Sudbury Structure isessentially unmetamorphosed; however, the SouthRange shows evidences of faulting, folding as wellas metamorphism locally up to the amphibolite fa-

Žcies Bennett et al., 1992; Cowan and Schwerdtner,.1994 . Metamorphism and deformation of the South

Range may be related to the Penokean OrogenyŽ .1.7–1.9 Ga; Card et al., 1984 or a slightly younger

Žcontractional orogeny 1.44–1.45 Ga; Fueten and.Redmont, 1997 .

The local geology and setting of the Little StobieŽ .Mine, Sudbury Structure Fig. 1 , are described by

Ž .Davis 1984 . Geochemical data from Little StobieŽ .Mine were presented by Hoffman et al. 1979 . Two

Ž .orebodies in differing settings Fig. 1 comprise themine: Orebody 1 is a typical Contact Sublayer de-

( )F. Molnar et al.rChemical Geology 154 1999 279–301´ 281

Ž .Fig. 1. Generalized underground plan and sections of orebodies at the Little Stobie Mine after Davis, 1984 . Samples from the mine arefrom the contact zone of the No. 1 Orebody at the 2000-ft level and No. 2 Orebody at the 980-ft level ramp and the 1600-ft level stope.


posit at the contact of SIC norite with Murray Gran-Ž . Ž .ite 2.4 Ga and metavolcanic rocks greenstone of

Žthe Elsie Mountain Formation about 2.5 Ga; Krogh.et al., 1984 . Orebody 2, perpendicular to the attitude

of the Sublayer, is entirely in the footwall rocks,enclosed by breccia at the greenstone–Murray Gran-ite contact. The latter orebody is enriched in precious

Žmetals by a factor of at least two Hoffman et al.,. Ž .1979 . Davis 1984 pointed out that distal stringers

to each orebody are characteristically enriched inchalcopyrite and precious metals with respect to theorebody in general.

Sulfide-bearing vein samples near the footwallcontact with the Murray granite were collected inareas of Cu-enrichment of Sublayer ore from ore-

Žbody No. 1 LS92-1 and LS92-2 from the 2000-ft.level; Fig. 1 . The textural evidence of Cu-enrich-

ment is manifested by the replacement of pyrrhotiteŽby chalcopyrite in the Sublayer ore. Veins 1–10 cm

.thick hosted by the granite are characterized byxenomorphic quartz grains intergrown with sulfides,predominantly chalcopyrite. Along the contacts ofquartz and sulfide grains, chlorite–epidote–amphi-bole alteration assemblages are present. Graniteshows silicification and K-feldspar metasomatic al-teration along the veinlets.

Ž .The sulfide bearing veins thickness up to 20 cmcut metavolcanic rocks of the Elsie Mountain Forma-tion along the contact of Orebody 2. These veins

Žwere sampled at the 980-ft level ramp LS-2 and. ŽLS-6 and the 1600-ft level stope LS-1, LS-3 and.LS-4 . Xenomorphic quartz grains as large as several

centimeters are enclosed in or intergrown with sul-fides. Quartz is often associated with ankerite, Fe-rich

Ž .and Cl-bearing up to 2.45 wt.% Cl amphibole, andchlorite. In LS-6, zoned tourmaline is also abundant.The enclosing metavolcanic rocks have chlorite–actinolite alteration along the veins.

In addition to chalcopyrite, pyrrhotite and pent-landite, the veins also contain sphalerite, galena,michenerite, merenskite, kotulskite and bismuth–tel-lurides. The quartz–sulfide textural relationshipssuggest that quartz was precipitated earlier than andsynchronously with sulfides.

Barren veins were sampled from outcrops ofŽ .granitic dikes 1.85 Ga, Krogh et al., 1984 cutting

Ž .quartz-rich norite, LS-S, 1 km north of the mineŽand quartz gabbro LS-R, from a roadcut 2.5 km

.northeast of the mine . These 2–5 cm thick veinshave interlocking grains of xenomorphic quartz withK-feldspar, chlorite, sericite and calcite. No sulfideminerals were found in this paragenesis. The precipi-tation of barren veins post-dates the intrusion of theSudbury Igneous Complex; however, the sulfide-rich

Žveins occur exclusively below the norite the lower-.most unit of the SIC and they do not cut the

granophyre, quartz gabbro or norite.

3. Analytical methods

The fluid-inclusion study was carried out using aU.S.G.S. apparatus at Carleton University and on a

Ž .Chaixmeca microthermochamber Eotvos Lorand U. ,¨ ¨ ´both calibrated using phase changes in pure chemi-cals and synthetic fluid inclusions. Reproducibility ofthese measurements was 0.18C at and below 08C,and 28C at 3748C. Analyses were carried out on50–150 mm-thick, doubly polished sections of trans-parent minerals. Due to the complicated freezingbehavior of inclusions, temperature cycling was usedduring the low temperature measurements. To studythe chemical composition of daughter minerals influid inclusions, quartz grains were broken and im-mediately carbon-coated, then placed in the JeolSEM, and analyzed using the Link eXL L24 instru-

Ž .ment Carleton U. . The details of analytical condi-Ž .tions were described by Farrow et al. 1994 .

4. Fluid inclusion petrography

Quartz is characterized by abundant and relativelylarge fluid inclusions, whereas carbonates containonly a few populations of very small inclusionsunsuitable for detailed microscopic and microther-mometric investigations. Quartz contains four main,and several sub-types of fluid inclusions classifiedaccording to their phase proportions observed at

Ž .room temperature Fig. 2 .Type I: monophase aqueous inclusions. The ab-

sence of vapor at room temperature is related toŽeither metastability they occur with Type IIA aque-

.ous inclusions and developed vapor during coolingŽor to necking down of other inclusions these latter

inclusions may sometimes be characterized by an.elongate tail and very irregular, angular shape .


Fig. 2. Types, abundance and mode of occurrence of fluid inclusions. Laq: aqueous liquid, V: vapor, LCO2: CO liquid, VCO2: CO vapor,2 2Ž .SH: unknown daughter mineral, DF: degree of filling for aqueous inclusions vol. liquidrvol. inclusion , DFC: degree of filling for carbonic

Ž . Ž . Ž .inclusions vol. CO phasesrvol. inclusion , N.O.: not observed, P: primary single , S: secondary hosted by fractures .2

Type II: two-phase inclusions containing an aque-ous phase and vapor. Based on the volume of thevapor bubble two subclasses were distinguished:

ŽType IIA: inclusions with 5–10 vol.% vapor Fig.

.3a , and Type IIB inclusions with 20–30 vol.%vapor.

Type III: polyphase inclusions containing aqueousqvaporqhalite phases with or without additional


Fig. 3. Photomicrographs of the most common type of inclusions. A—Type IIA, secondary inclusion, sample LS-3; B—Type IIIB,secondary inclusion with small sylvite in addition to halite daughter mineral, sample LS-92-2; C—Type IIIC, single inclusion, sample LS-1.Note the difference in refractive index of daughter minerals; D—Type IIIE, secondary inclusion, sample LS-6. Note that halite is round inthe presence of Fe–Mn chloride daughter mineral. E—Type IVA, secondary CO inclusion without visible aqueous phase, sample LS-3; F2

—Type IVC secondary inclusion with halite in the aqueous phase, sample LS-3. The length of scale bar is 10 mm on each photomicrograph.


solid phases. Vapor usually occupies about 10–20vol.% of the inclusion. Most solid phases dissolvedduring heating and are daughter minerals.

The chemical composition of solid phases wasdetermined by SEM-EDS analyses carried out onopened fluid inclusions. Some phases which dis-played anomalously low Cl-content in comparison tothe sum of their cations may have experienced Cl-loss

Ž .during analysis Farrow, 1994 or partial oxidationafter opening. The analytical results are semi-quanti-

Ž .tative because a analyses were made on unevenŽ .surfaces of very small minerals in holes, b solid

phases may be covered by a thin film of materialquenched from liquid.

It is not possible to determine whether the SEM-EDS analyses were performed on isolated or frac-ture-related inclusions due to the lack of three-di-mensional information. The analyzed phases areshown on Fig. 4. Na, K, Fe, Mn, Pb, Ba and Cu werethe major metallic elements found in the solids. Insome minerals, Ca and Mg were also detected. Theanion is Cl in most cases, but S is more abundant

Žthan Cl in three minerals from LS-6 phases 15, 19.and 20 on Fig. 4D and E . The sulfur-rich minerals

also contain Fe and Cu and are probably chalcopy-rite. Opaque minerals in polyphase inclusions werenot observed under the polarizing microscope andtheir unusual presence found during SEM studiessuggests that they are accidental and not daughterminerals.

In Cl-rich minerals, where Fe is the most abun-dant element, Mn is also present or vice versa. Pband Ba occur together with Pb predominant. Na andK chlorides are usually poor in other elements. Ca isubiquitously present as a minor constituent at con-centrations below 2 wt.%; higher concentrations oc-

Ž .curred only in Fe-chlorides of LS-6 Fig. 4C and D .Some Ca-rich chlorides are round grains on the

Ž .inclusion wall Fig. 4C, grain 14 possibly due todeposition during opening of the inclusion. Mg oc-curs sporadically, mostly in Ca-bearing chlorides.

ŽNaCl usually has a perfect cubic habit Fig. 3B–.CFig. 4F in most Type III inclusions. Sylvite is

always anhedral. Pb-rich chloride often formsbipyramidal or apparently rhombohedral weakly

Ž .anisotropic crystals Fig. 3DFig. 4H . PbCl has2

orthorhombic–pseudohexagonal symmetry andŽBaCl has the same crystal structure Donnay and2

.Ondik, 1973 . The Fe–Mn rich chlorides in Type IIIinclusions have variable morphology and weak ani-sotropy. However, Fig. 3DFig. 4A and I clearlyshow their common acicular, hexagonal–prismatic

Ž .appearance. Pure FeCl lawrencite and MnCl have2 2Ž .trigonal symmetry Donnay and Ondik, 1973 , and

their common occurrence in the same daughter min-eral is compatible with this structural similarity. Thepresence of other cations in these phases may berelated to deposition of chlorides on the surfaces ofdaughter minerals due to the evaporation of liquidduring the opening of inclusions. In addition to thedaughter minerals listed above an unknown solidphase composed of a fibrous, or granular mass ofslightly anisotropic crystallites with intermediate re-

Ž .fractive index SH phase was also observed in someinclusions with Fe–Mn and Pb–Ba chlorides.

In inclusions with acicular Fe–Mn-chloridedaughter minerals, and rarely in inclusions withbipyramidal Pb–Ba chloride phases, a round isotropicphase with refractive index and yellowish coloursimilar to halite cubes of other Type III inclusions

Ž .was found Fig. 3D . Round halite was observedŽ .during the SEM study Fig. 4G, grain 25 , and in

petrographic study of unopened inclusions, confirm-ing that this habit did not exclusively result frompressure-release on fracturing of quartz. Moreover,the crystal habit of halite may also be a function of

Žthe composition of the parent solution e.g., in thepresence of PbCl , octahedral faces appear on halite2

.cubes ; thus the modification of crystal morphologyis expected in strongly saline heavy-metal-rich solu-tions. It was also observed that upon cooling afterdissolution of daughter minerals, the previously roundhalite re-precipitated as cubes.

From the polarizing microscope and SEM study,five subclasses of Type III inclusions were deter-mined according to their daughter mineral associa-

Ž .tions see also Figs. 2 and 3B–D :Ž .Type IIIA: liquidqvaporqhalite cubic ,

Ž .Type IIIB: liquid q vapor q halite cubic qŽ .sylvite anhedral ,

ŽType IIIC: liquid q vapor q halite cubic or. Žround qPb-Ba chloride anhedral to bipyramidal

or apparently rhombohedral morphology, slightly.anisotropic, high refractive index .

ŽType IIID: liquid q vapor q halite cubic or.round qPb-Ba chlorideqsylvite. SH phase with


Ž .Fig. 4. SEM photographs of opened fluid inclusions. A—sample LS-6, 1: Fe–Mn– Na–Ca –Cl, 2: calcite, 3: quartz; B—sample LS-6, 4:Ž . Ž . Ž . Ž . Ž .Na– Pb–Mn–Ca–K –Cl, 5: Mn–Fe–K– Na–Ca –Cl, 6: Pb–Ba–K –Ca –Cl, 7: Mn– K–Fe–Ca–Na –Cl, 8: Pb–Ba–K– Na,Ca –Cl, 9:Ž . Ž . Ž . Ž .Mn– K–Fe–Ca–Na –Cl; C—sample LS-6, 10: Fe–Mn– Na–Ca –Cl, 11: Fe–Ca– Mn–Na–Mg–K –Cl, 12: Fe–Ca– Mn–Na–Mg–K ,

Ž . Ž . Ž . Ž . Ž13: Fe–Ca– Mn–Mg–K–Na –Cl, 14: Fe–Ca– Na–Mn–Mg–K –Cl; D—sample LS-6, 15: Fe–Cu– Ca –S– Cl , 16: Fe–Ca– Mn–Na–. Ž . Ž . Ž .K –Cl, 17: Fe–Na–Ca– Mn–Pb–K –Cl, 18: quartz; E—sample LS-6, 19: Cu–Fe–S– Cl , 20: Cu–Fe– Pb–Ca–Mg –S–Cl; F—sample

Ž . Ž . Ž . Ž . Ž . Ž .LS-1, 21: Na–Cl– S , G—sample LS-1, 22: Pb–Ba–K– Na –Cl– S , 23: Na– Pb–Ba–K –Cl, 24: K– Na–Pb –Cl, 25: Na–K– Pb –Cl; HŽ . Ž . Ž .—sample LS-4, 26: Pb–K–Cl; I—sample LS-4, 27: Fe– Mn–Cu–Na–Ca –Cl, 28: Fe– Mn–Na–Ca –Cl, 29: Fe–Na–Mn– Ba–Pb–Ca –Cl,

Ž .30: Na– Fe–Pb–Ba–Mn–Ca –Cl. The background mineral in all cases is quartz. The order of elements displays their relative abundanceafter substraction of Si and Al from the host mineral. Concentrations of elements in brackets are less than 2 wt.%. Numbers on scale bars arein micrometers.


unknown composition is also present in someinclusions.

Ž .Type IIIE: liquidqvaporqhalite round qFe–ŽMn chloride acicular, slightly anisotropic, inter-

.mediate refractive index Pb–Ba chloride qsylviteqSH.A very few fracture-hosted inclusions of Type

IIIA, IIIB and IIIC from orebody No. 1 also containŽ .a red-brown flake probably hematite .

Type IV: carbon dioxide-rich inclusions. Accord-ing to the volume of the CO -phase and the presence2

of other phases, the following subtypes were distin-Ž .guished Fig. 2 :

Type IVA: CO -inclusions without visible aque-2Ž .ous phase Fig. 3E .

Type IVB: inclusions with aqueous liquid andliquid CO . The phase ratios are very variable2

even along the same fracture.Type IVC: inclusions with aqueous liquidq liquid

Ž .CO qhalite Fig. 3F . In a few inclusions other2

daughter minerals also occur. Phase ratios are alsovariable.Some Type IV inclusions may contain gas within

liquid CO at room temperature depending on its2

density.The shapes of the inclusion types are highly

variable. Type IIA and Type IV inclusions mostcommonly have round, negative crystal shapes. Theirgreatest diameter is most typically less than 20 mm;single Type IV inclusions may be larger. Type IIIinclusions are usually large, attaining 50–80 mm,and they are most commonly angular to round, some-times with almost negative crystal shapes.

Almost all types of inclusions occur in all samplesŽ .Fig. 2 . Type I and IIA inclusions are ubiquitous inall samples and they occur in fractures cutting quartzgrain boundaries. Type IIB inclusions are also sec-ondary, but they were observed only from barren

Ž .veins LS-S and LS-R , and from orebody No. 2Ž .LS-3 and LS-6 . In barren veins, they occur infractures cutting the trails of Type IIA inclusions.The shape of Type IIB inclusions is usually moreirregular than Type IIA inclusions.

Type IIIA inclusions occur in all samples, mostcommonly along fractures cutting quartz grainboundaries. In some Type IIA inclusion trails, theseinclusions also occur sporadically. Thus some TypeIIA inclusions probably have metastable phase com-

positions due to the lack of halite nucleation. Thisinterpretation is also supported by the precipitationof halite in a few IIA inclusions during freezing.Isolated Type IIIA inclusions were also observed and

Ž .they may be primary Roedder, 1984 . The mode ofoccurrence of Type IIIB and IIIC is identical to TypeIIIA inclusions, but their abundance is less in mostsamples. These inclusions with differing phase com-positions may occur along the same fractures, andthe absence of solid phases other than halite may berelated to their metastable phase composition at roomtemperature.

Type IIID and IIIE inclusions occur most com-monly in isolation, and Type IIIE inclusions areessentially restricted to orebody No. 2, although afew were found in LS92-1 from orebody No. 1.However, they also occur in fractures, but some ofthese trails terminate in quartz grains. Crosscuttingrelationships of trails of Type IIA and Type IIIinclusions suggest that the two-phase inclusions weretrapped later.

Ž .Type IV CO -rich inclusions are less abundant2Ž .Fig. 2 , especially in orebody No. 1. In most sam-ples they are isolated as well as fracture-hostedinclusions. However, in LS-6 and LS-S, Type IVCoccurs only as isolated inclusions. Type IVA inclu-sions were found only from orebody No. 2.

CO -rich inclusions with high variability in the2

volume of the CO phase occur together with Type2

II or Type III inclusions in fractures. This suggeststhat the observed volume ratios are the result ofinhomogeneous trapping and an unmixed CO rich2

Žphase existed at the time of entrapment Diamond,.1994 . Alternatively, the variation in the phase com-

position may also result from necking down; how-ever, petrographic observations do not support this.The presence of halite in some CO -rich inclusions2

may also reflect the heterogeneity; miscibility ofCO with highly saline solutions is very restricted2

Ževen at high T and P Bowers and Helgeson, 1983;.Trommsdorff and Skippen, 1986 .

The petrographic observations support the follow-ing chronology of inclusion entrapment:1. Isolated Type III and isolated Type IVA, B, and

C inclusions.2. Type III and Type IVA, B and C inclusions in the

same fractures.Ž .3. Type IIA, Type I metastable or necked down


and some Type IVA and IVB inclusions alongfractures.

4. Type IIB inclusions in fractures.

5. Microthermometry

5.1. Heating runs

Fig. 5A shows the frequency distribution of ho-mogenization temperatures measured in Type II in-clusions; statistical evaluation of these data is inTable 1. In agreement with the observations on thedegree of filling, Type IIB inclusions usually have

Ž .higher homogenization temperatures 190–3008CŽ .than Type IIA inclusions 60–2108C .

The vapor phase of Type III inclusions homoge-nized at a lower temperature than the dissolution ofthe last solid phase in most runs. During homoge-nization of polyphase inclusions the heating rate inthe temperature range of daughter mineral dissolu-tion was about 2–38Crmin. The dissolution ofdaughter minerals was reproducible with "58C con-fidence in verification measurements. The re-precipi-tation of daughter minerals was usually observedafter cooling down to room temperature or severalhours after their homogenization; halite was alwaysthe first phase to nucleate. The total homogenizationof the relatively large fluid inclusions was notachieved in many runs due to leaking or decrepita-tion after homogenization of the vapor phase.

In Type IIIB, D and E inclusions, sylvite mostcommonly dissolved between 50 and 1008C during

Ž .heating Fig. 6 . Homogenization of Pb–Ba-chloridetook place in two temperature ranges: at about 1008

and above 2008C. The unnamed SH phase in TypeIIID inclusions dissolved at various temperatures.Acicular Fe–Mn chloride phases homogenized be-tween 150 and 2008C in Type IIIE inclusions with aslight increase of dissolution temperature towards

Ž .low total homogenization Th halite temperatures.The observed variation in the dissolution temperatureof various daughter minerals probably reflects theeffect of bulk composition on the solubility of thechloride phases or the occurrence of double salt

Ž .formation Linke, 1965; Roedder, 1984 . The lasthomogenizing phase was halite in the vast majorityof inclusions.

Halite homogenization temperatures from frac-Ž .ture-hosted Type III inclusions Fig. 5B group into

Ž .two main ranges 220–3708C and 100–2108C . Thehigh-temperature range is characteristic of both ore-bodies and barren veins, but the lower halite dissolu-tion temperatures occur mostly in samples from ore-bodies.

Ž .Data from isolated primary inclusions differ be-Ž .tween the two orebodies Fig. 5C ; those from ore-

body No. 2 have much higher halite homogenizationŽ .temperatures 280–3508 than those from orebody

Ž .No. 1 180–2708C . Data from barren veins are inboth temperature ranges.

Ž .The high Th liquid 180–2808C —high Th haliteŽ .240–3508C data population occurs uniquely in sin-gle and fracture hosted inclusions of orebody No. 2Ž .population ‘3’ of Fig. 7B . The other data of ore-body No. 2 are from fracture-hosted inclusions. Pop-

Ž .ulation ‘2’ Fig. 7B is mainly composed of TypeŽ .IIIA inclusions. A similar population ‘2’ in Fig. 7A

may be recognized from orebody No. 1. However,most isolated and fracture-hosted inclusions of this

Ž .orebody form population ‘1’ Fig. 7A . This popula-tion with Th liquid data of 140–2108C and Th halitedata 170–3008C differs from data of Type III inclu-

Žsions in orebody No. 2 and in the barren veins Fig..7C , although some overlap occurs. Most inclusions

from barren veins show similar Th liquid–Th halitedata pairs to inclusion population ‘4’ detected inorebody No. 2.

Total homogenization of Type IVB and Type IVCinclusions was never achieved due to decrepitationabove 200–2508C.

5.2. Freezing data and composition of inclusions

5.2.1. Type II inclusionsFreezing data for Type II inclusions are summa-

rized in Table 1. Type IIA inclusions froze at abouty808C. At the temperature of freezing, inclusionsbecame ‘hazy’ brown. Eutectic melting with clearlyvisible liquid occurred between y68 and y528C.Above the eutectic temperature the remaining solidswere still ‘hazy’ but further rapid melting and subse-quent ‘clearing’ of color occurred between y50 andy408C in many inclusions. This ‘clearing’ was re-producible with about "28C confidence. In oneinclusion, temperature cycling in this range caused


ŽFig. 5. Frequency distribution diagram of total homogenization temperatures Th liquid for Type II inclusions and Th halite for type III.inclusions determined in aqueous inclusions.

the growth of a small, birefringent phase whichmelted at y458C. In inclusions with very low ice-melting temperatures, the ‘clearing’ was not ob-served and these inclusions had large amounts ofliquid with a few round ice grains after eutecticmelting between y55 and y658C.

The melting of ice in Type IIA inclusions tookplace between y50 and y108C with average valuesbetween y22.3 and y31.78C in samples from ore-body No. 1, between y31.4 and y34.68C in sam-ples from orebody No. 2 and y28.78C in samples

Ž .from barren veins Table 1 .


Table 1Microthermometric data for Type II inclusions

Ž . Ž . Ž . Ž .Sample Inclusion Th 8C Te 8C Tc 8C Tm ice 8C Salinity meantype n Mean s.d. n Mean s.d. n Mean s.d. n Mean s.d.

LS-1 IIA 9 136.3 25.1 – – – – – – – – – –Ž .LS-2 IIA 12 102.5 27.7 3 y56.0 3.6 3 y44.0 2.7 7 y22.3 10.8 21.7 CaŽ .LS-3 IIA 23 102.8 26.4 3 y52.5 2.5 – – – 4 y24.7 0.5 22.8 CaŽ .IIB 17 267.6 23.3 – – – – – – 6 y0.7 0.7 1.22 NaŽ .LS-4 IIA 15 162.1 52.6 8 y59.7 4.0 6 y48.9 2.1 13 y31.7 5.4 25.4 CaŽ .LS-6 IIA 11 125.1 43.5 7 y58.9 5.9 7 y44.4 2.9 7 y23.4 4.5 22.2 CaŽ .LS92-1 IIA 11 110.1 22.2 2 y62.5 – 1 y40.0 – 2 y31.4 – 25.3 CaŽ .LS92-2 IIA 7 125.6 28.5 4 y59.3 2.8 3 y48.7 2.3 3 y34.6 1.3 26.4 Ca

LS-S IIA 4 119.1 26.6 – – – – – – – – – –Ž .LS-R IIA 5 144.1 39.5 4 y65.5 1.6 4 y43.9 1.9 4 y28.7 3.1 24.4 CaŽ .IIB 5 208.1 8.7 2 y63.2 – 2 y48.3 – 2 y28.9 – 24.4 Ca

Th—homogenization temperature of the vapor phase; Te—eutectic melting; Tc—cotectic melting of hydrohalite; Tm ice—meltingŽ . Ž .temperature of ice; n—number of measurements; s.d.—standard deviation; Ca salinity in CaCl equiv. wt.%; Na salinity in NaCl equiv.2

wt.%.

The freezing behavior of Type IIA inclusions maybe interpreted in the NaCl–CaCl –H O system2 2ŽCrawford, 1981; Shepherd et al., 1985; Haynes,1985; Vanko et al., 1988; Oakes et al., 1990;

.Schiffries, 1990 . The eutectic temperature of theŽternary system is y528C where antarticite CaCl P2

.6H O melts in the presence of hydrohalite, ice,2

liquid and vapor; the observed lower temperatures

Ž . Ž .Fig. 6. Dissolution temperatures of various solid phases Th solid as a function of halite melting temperatures Th halite in Type IIIinclusions.


Ž . ŽFig. 7. Distribution of halite dissolution temperatures Th halite as a function of homogenization temperatures of the vapor phase Th.liquid . The differences among various populations may be related to the differences in chemical composition of inclusion liquids

influencing halite dissolution temperatures.

may correspond to metastable melting or recrystal-Žlization of the CaCl P4H O phase Davis et al.,2 2

.1990 . Above the eutectic temperature, hydrohalite,ice, liquid and vapor may coexist and if the lastmelting phase is ice, hydrohalite melting tempera-

Ž .tures are related to the NaClr NaClqCaCl ratio2

of liquid. Melting of a birefringent phase at y458Cin one inclusion and the reproducible ‘clearing’ withincrease in volume of liquid between y40 andy508C may correspond to cotectic melting of hydro-

Ž .halite. Thus the NaClr NaClqCaCl ratios for2

these inclusions are less than 0.1 and the total salin-


Žity may be calculated as CaCl equiv. wt.% Oakes2.et al., 1990 . The calculated average salinity values

of Type IIA inclusions are 21–27 CaCl equiv. wt.%2Ž .Table 1 .

In three freezing runs on LS-3, different meltingpaths were observed in Type IIA inclusions associ-ated with Type IIIA inclusions along the same frac-ture. Eutectic melting took place in the same temper-ature range as in other Type IIA inclusions, but smallbirefringent phases remained after ice melting be-tween y21.8 and y23.88C. Temperature cyclingclose to final melting of these phases resulted insingle birefringent crystals with monoclinic habitindicative of hydrohalite. The reproducible meltingof these phases was between y6.9 and y13.98C butit was not accompanied by precipitation of halite.This melting path may also be explained in theNaCl–CaCl –H O system and it suggests that the2 2

composition of these inclusions is within the hydro-halite stability field, or that halite failed to nucleate

Žduring the melting of hydrohalite Vanko et al.,.1988 . This latter statement is supported by the

presence of halite in other inclusions associated withthese peculiar ones along the same fracture. Thusthese inclusions show metastable phase assemblagesat room temperature and their salinity cannot bedetermined exactly.

Type IIB inclusions from LS-R had similar freez-Ž .ing behavior to Type IIA inclusions Table 1 . How-

ever in LS-3, Type IIB inclusions froze betweeny40 and y508C, initially melted above y108Cwith ice rapidly melting close to 08C. This high-melt-ing temperature indicates that these inclusions havevery dilute fluids with about 1 NaCl equiv. wt.%salinity. Dilute KCl–H O, alkali–sulfate or alkali–2

bicarbonate–carbonate solutions may have this kindŽ .of freezing behavior Shepherd et al., 1985 .

5.2.2. Type III inclusionsType III inclusions had very variable freezing

behavior. In the vast majority, freezing was notachieved even by supercooling to y1808C. Thisbehavior is not uncommon in highly saline inclu-sions due to the sluggish nature of ice nucleationrelated to high viscosity of saturated liquidsŽ .Shepherd et al., 1985 . Some fracture-hosted TypeIIIA inclusions from LS92-2 developed small aggre-gates of acicular isotropic phases with low refractive

index at the apices of halite during cooling. Thefreezing of these inclusions was also never achieved,and the newly formed phases melted in the same

Ž .temperature range at about 608C as did sylvite inType IIIB inclusions of this sample. This observationalso suggests that the phase composition is metastablein some Type IIIA inclusions at room temperature.

When freezing was observed, it took place atabout y80 to y908C in Type IIIA, IIIB, IIIC andsome Type IIID inclusions. The formation of ice wasa relatively slow process starting from one point ofthe inclusion wall in some cases. After ice formation,the inclusion color became dark brown. The appear-ance of liquid was detected between y54 andy788C, with subsequent formation of hydrohalite insome cases. However, halite remained virtually in-tact in many inclusions, although some hydrohalitedeposition was finally achieved by temperature cy-cling. Complete reaction of halite to hydrohalite wasnever achieved. Ice melting took place between

Ž . Žy20.8 and y458C Fig. 8 , and metastable non-re-.producible hydrohalite melting took place above

0.18C. In inclusions with hydrohalite, rapid super-cooling from a temperature close to hydrohalite melt-ing resulted in the sudden appearance of ice at aboutthe eutectic temperature.

Although SEM-EDS analyses proved that fluidsof Type III inclusions are enriched in heavy-metals,freezing may also be explained by phase transitionsknown in the NaCl–CaCl –H O system. The first2 2

melting temperatures are lower than the eutectictemperature of this system which may be related tometastable melting andror recrystallization relatedto the presence of the metastable CaCl P4H O2 2

phase.Metastable behavior and the very complex com-

position of fluids did not permit the exact determina-tion of salinity. Taking into account the halite melt-ing temperatures, for inclusions in which hydrohalitewas present at the melting of ice, salinity valuesŽ .Vanko et al., 1988 are between 49.4 and 39.3

Ž .NaClqCaCl wt.% with NaClr NaClqCaCl ra-2 2

tios at about 0.5 for the lower values and between0.65 and 1 for the higher values. However, thesesalinity data are only estimates due to the presenceof other chloride species. Halite melting after ho-mogenization of vapor also complicates the determi-

Ž .nation of exact salinity Bodnar and Vityk, 1994 .


Ž .Fig. 8. Frequency distribution diagram of melting temperatures T melt of ice in Type IIIA, B, C and D inclusions and an unknown hydrateŽ .phase possibly CaCl P6H O or Fe–Mn–chloride-hydrate developed in Type IIID and IIIE inclusions.2 2

Some Type IIID and IIIE inclusions froze be-tween y30 and y608C. Halite of these inclusionsdid not show observable resorption by, or transfor-mation to, hydrohalite. After freezing, the inclusionsremained clear, and the phase formed at freezing hadyellowish–greenish anisotropy. During the formationand melting of the anisotropic phase, corrosion andredeposition of acicular Fe–Mn chloride was ob-served in some inclusions. A very small amount ofliquid was formed between y20.58C and y23.38Cin most runs, but between y30 and y358C in twoinclusions and at about y558C in one inclusion. Theanisotropic phase melted between q10 and q358CŽ .Fig. 8 , and these melting temperatures were repro-ducible. Recooling before their final melting causedregrowth of phases without deposition of ice. Duringmelting of this anisotropic phase, very small acicularcrystallites appeared in five of these inclusions, andthey dissolved between 66.5–68.28C.

The melting behavior described above cannot befully explained by the available data for chloridesystems, but an attempt may be made consideringlow-temperature behavior in the Ca-rich part of the

Ž .NaCl–CaCl –H O system. Schiffries 1990 de-2 2

scribed similar inclusions from the Bushveld Com-plex. His type-1 inclusions showed initial melting at

Žabout y358C, with melting of antarcticite CaCl P2.6H O between q22 and q298C. Melting between2

y20.5 and y23.38C in the Little Stobie samplesmay correspond to the peritectic point at y228C ofthe CaCl –NaCl–H O system, where antarcticite,2 2

halite, hydrohalite and liquid may coexist. Observedstable melting of the anisotropic phase well above08C may correspond to the melting of antarcticite.However, SEM-EDS analyses proved that Type IIIEinclusions are saturated with respect to Fe–Mn chlo-rides.

FeCl and MnCl have various hydrated phases.2 2

The eutectic temperature for FeCl –H O is y358C2 2Ž .Linke, 1965 . If the FeCl content is more than 30.42

wt.%, FeCl P6H O may persist to q12.38C. In a2 2Ž .solution with very high salt content above 38 wt.%

FeCl .4H O may be the stable phase to 76.58C. In2 2

the MnCl –H O system the eutectic temperature is2 2

y5.68C where ice, liquid and MnCl P6H O coexist2 2Ž .concentrations30.4 wt.% . This hydrate is stableto y28C, where it may convert to MnCl P4H O2 2

which is stable to 588C in high salinity fluids. Bothsystems have other hydrated phases. These examplesdemonstrate that the observed phase transitions mayalso be explained by decomposition of these hy-drates. However, the exact concentration of these


inclusions cannot be determined, due to the lack ofdata for the high-temperature solubility of halite inthese very complex solutions.

5.2.3. Type IV inclusionsType IVA as well as the CO phase of IVB and2

IVC inclusions froze at about y1008C. The meltingof solid CO occurred between y61.3 and y56.68C;2

average values are distributed between y56.6 andŽ .y59.18C Table 2 . The Tmelt CO data for single2

IVA inclusions and for those which are associatedwith Type III inclusions along fractures are lowerthan data for CO inclusions occurring together with2

Type IIA inclusions from orebody No. 2, and forIVB and IVC inclusions from barren veins.

Ž .Regarding the Th CO L–V L values, Type2

IVA isolated inclusions also form a discrete group incomparison to Type IVA, IVB, and IVC inclusions

Ž .occurring in fractures Table 2 . These latter inclu-Ž .sions have higher Th CO L–V L values. Two2

Type IVA inclusions from LS-3 associated withType IIA secondary inclusions revealed critical ho-mogenization during the shrinkage of vapor phase atabout q298C. Type IVB and IVC inclusions frombarren veins are characterized by relatively high Th

Ž .CO L–V L data; these high values occur only in a2

very few inclusions from samples of orebody No. 2Ž .Table 2 .

The slight depression of Tmelt CO data from2

y56.68C indicates that these inclusions contain addi-Ž .tional constituents Burruss, 1981 . Qualitative Ra-

Žman studies St. Kliment Ohridski U., Sofia, Mi-.crodil 28 equipment powered by a 30 mW Ar laser

on these inclusions revealed only methane, in addi-tion to CO . Using the V–X diagrams of the CO –2 2

Ž .CH system Thiery et al., 1994 , the freezing data4

for single Type IVA inclusions correspond to XCH 4

s0.05–0.11 with molar volume of 46–50 cm3rmoleŽ .Table 2 . For fracture-hosted IVA inclusions associ-ated with polyphase inclusions these data are XCH 4

s0.01–0.16 and Õs48–71 cm3rmole; and forthose which are associated with Type IIA inclusions,they are X s0–0.19 and Õs49–55 cm3rmoleCH 4

in samples from orebody No. 2. Two inclusions inthis latter group with critical homogenization lower

Ž .than 31.18C the critical temperature of CO had2

Tmelt CO at y56.68C, thus the deviation from the2

critical temperature of CO is probably related to the2

uncertainty of the observations during critical ho-mogenization and not to the presence of other

Table 2Microthermometric data for Type IV inclusions

3Ž . Ž . Ž . Ž .Sample Type of Description n Tmelt CO 8C Tm 8C Th 8C X Õ cm rmole2 CO clath. CO CH2 2 4Ž . Ž .inclusion mean"s.d. mean"s.d. mean"s.d. min.ymax. min.ymax.

Ore-body No. 2Ž .LS-1 IVA single P 2 y59.0 – y3.0 0.08–0.1 48–49

IVA in fractures with 20 y58.2"1.2 – 9.7"7.2 0.01–0.16 48–71type III inclusions

IVB in fractures with 2 y59.1 y8.5 8.3 0.09 54–57type III inclusions

IVC in fractures with 2 y58.9 y9.2 12.8 0.03–0.16 53–66type III inclusions

LS-3 IVA in fractures with 10 y57.2"1.4 – 6.8"4.9 0.0–0.19 49–55type IIA inclusions

Ž .IVA in fractures with 2 y56.6 – 28.9 C 0.0 94type IIA inclusions

Ž .LS-4 IVA single P 8 y59.0"0.7 – y8.4"4.7 0.05–0.11 46–50

Barren ÕeinsŽ .LS-S IVC single P 4 y57.8"0.2 y12.5"0.3 19.0"9.6 0.04–0.05 50–84Ž .LS-R IVB single P 1 y57.0 y12.4 25.1 0.02 64Ž .IVC single P 1 y56.8 y11.1 26.4 0.01 65

n—Number of measurements; Tmelt CO —melting temperature of CO ; Tm —melting temperature of CO clathrate; Th —ho-2 2 CO clath. 2 CO2 2Ž .mogenization temperature of the CO phase; X —mole fraction of CH ; Õ—molar volume of the carbonic phase; C —critical2 CH 44

homogenization.


volatiles. These inclusions are considered to be pureŽCO fluids with critical molar volume 942

3 .cm rmole .The CO -rich phases of Type IVB and IVC inclu-2

sions associated with Type III inclusions from ore-body No. 2 have compositions between those ofsingle and fracture-hosted Type IVA inclusions. TheCO -rich phase of Type IVB and IVC inclusions2

Žfrom barren veins has very similar composition with3 .X s0.01–0.05 and Õs50–84 cm rmole toCH 4

Type IVA inclusions associated with Type IIA inclu-sions from the ore veins.

The determination of an eutectic melting tempera-ture of the aqueous phase in Type IVB and IVCinclusions was not possible due to melting andclathrate formation involving the CO phase. Aver-2

age CO -clathrate melting temperatures group be-2Ž .tween y8.5 and y12.7 C Table 2 corresponding

to the very high salinity of the aqueous phase. Theexact salinity cannot be calculated from these datadue to the possibly heavy-metal rich composition ofthe aqueous phase and presence of methane in thecarbonic phase. Tm CO clathrate at y10 C corre-2

sponds to 24.3 NaCl equiv. wt.% concentration ofthe aqueous phase in the NaCl–CO –H O system or2 2

22.6 CaCl equiv. wt.% in the CaCl –CO –H O2 2 2 2Ž .system Darling, 1991; Bakker et al., 1996 . The

actual salinity of these inclusions may be higher dueto the effect of methane content on clathrate meltingtemperatures.

6. Estimation of trapping conditions for fluid in-clusions

Trapping of different inclusion generations tookplace in the immiscibility field of CO -rich and2

aqueous solutions, as CO -bearing inclusions with2

variable carbonic liquidraqueous liquid volume ra-tios are associated with synchronously trapped aque-ous inclusions. Inclusions trapping end member im-miscible fluids should homogenize by dissolution ofeither phase at the same temperature, and the pres-sure in inclusions at the homogenization temperatureequals the trapping pressure. In P–T space, theintersection of isochores for coevally trapped endmember fluids is located at this unique point, which

in turn is the same as the intersection of their solvusŽ .curves Diamond, 1994 .

Primary magmatic ores of the South Range wereŽformed at a minimum depth of about 8 km Souch et

.al., 1969; Hoffman et al., 1979 and were uplifted totheir present position by about 5–6 km. Thus theminimum paleodepth of fluid migration processesrecorded by fluid inclusions was 2–8 km. This depthcorresponds to 550–2200 bars lithostatic pressureusing 2.74 gmrcm3 average density for the rocks of

Žthe SIC and the Sudbury Basin density data are.from Mcgrath and Broome, 1994 .

The available data for the H O–NaCl–CO sys-2 2

tem show that the solubility of CO in aqueous2

saline liquid or solubility of saline liquid in CO is2

very restricted even at relatively high temperature asŽsalinity increases in this pressure range Bowers and

.Helgeson, 1983 . Interpolating these data for theextremely saline fluids of this study, the aqueous partof the immiscible system may contain so little CO2

or the CO -rich phase may dissolve so little aqueous2Žsolution that these phases may be undetected or

overlooked due to the deposition of various chloride.hydrate phases in inclusions of this study in inclu-

sions trapped at a temperature about 300 C. This ismore pronounced at lower temperatures. The obser-vation of the total homogenization of the CO -poor2

Žsaline inclusions is not a problem ‘normal’ homoge-nization by the disappearance of vapor bubble or

.halite in these inclusion pairs, but detection ofŽhomogenization of CO -rich inclusions i.e., disap-2

pearance of a very thin aqueous film along the wall.of inclusion is improbable.

Assuming that Type IVA inclusions trapped thecarbonic endmember and Type IIA and Type IIIinclusions in association with them trapped the aque-ous endmember of the immiscible fluid system, mi-crothermometric data can be used for the estimationof the minimum trapping pressure. This minimumpressure is determined by that pressure which corre-sponds to the homogenization temperature data forType IIA and Type III inclusions along the isochoresof coevally trapped Type IVA inclusions.

Fig. 9 shows isochores for CO -rich inclusions.2

Most typical halite homogenization temperatures forisolated Type III inclusions from orebody No. 2correspond to 1800–2200 bars minimum pressurealong the isochore of isolated Type IVA inclusions.


Fig. 9. Estimation of minimum trapping conditions of fluid inclusions from orebody No. 2. The minimum trapping temperatures are equal tothe total homogenization temperatures. Minimum pressure of trapping is estimated by the pressure corresponding to the homogenizationtemperatures of aqueous inclusions along the isochore of coevally trapped Type IVA, CO -inclusions. Isochores for single and2

fracture-hosted Type IVA inclusions from orebody No. 2 were calculated for their representative compositions using equations forŽ .‘Holloway’s soup’ and the FLINCOR program Holloway, 1981; Brown, 1989 . Data of CO isochores: A, Tmelt CO sy598C; Th2 2

Ž . ŽCO sy58C associated with sigle Type III inclusions ; B, Tmelt CO sy58.58C, Th CO s4.58C associated with high-temperature2 2 2. ŽType III inclusions in fractures ; C, Tmelt CO sy56.88C, Th CO s88C associated with low-temperature Type III and Type IIA2 2

. 3 Žinclusions in fractures ; D, isochore for CO inclusions with critical molar volume of 94 cm rmole associated with Type IIA inclusions in2.fractures .

Data for the same type, but fracture-hosted inclu-sions are 1400–2000 and 700–1200 bars. The mini-mum trapping pressure corresponding to the homog-

Ženization temperatures of Type IIA inclusions sam-.ple LS-3 is 200–900 bars.

The range of minimum trapping pressures esti-mated for single Type III and IVA inclusions in-cludes the upper minimum pressure limit calculatedfrom the paleodepth data. The pressure data for verylate, Type IIA secondary inclusions cluster aroundthe lower minimum pressure limit calculated from

Ž . Ž .the paleodepth data Fig. 9 . Marshall et al. 1999also showed that inclusions with similar microther-

mometric data to Type IIa inclusions of this studyŽwere trapped under low pressure conditions at about

.950 bars in veins postdating the deposition of Cu–Ni–PGE ores.

7. Discussion

High-temperature fluids trapped as polyphase in-Žclusions in quartz associated with sulfides inclusion

.population ‘1’ and ‘3’ on Fig. 7 exhibit differentmicrothermometric behavior in comparison with bar-ren veins in the vicinity of the Little Stobie Mine.The differences seen on Fig. 7 may reflect the differ-


ent chemical composition of inclusions, becausehalite solubility is strongly influenced by the pres-

Ž .ence of other cations Linke, 1965; Roedder, 1984 .However, there are differences in these data not onlybetween orebodies and barren veins, but betweenveins at contacts of the two orebodies. Since theseveins are hosted by different rock types at the twoorebodies, the observed differences may be the resultof water–rock interaction. Late fluids trapped assecondary fluid inclusions are present in both ore-

Ž .body contact zones population ‘2’ on Fig. 7 , and inŽorebody veins and barren veins population ‘4’ on

.Fig. 4 and Type II inclusions . Therefore, fluids thatmay be responsible for sulfide-vein formation werethose trapped as high-temperature, polyphase inclu-

Žsions in samples from the orebodies populations ‘1’.and ‘3’ on Fig. 7 . Minimum trapping temperatures

reflected by the Th halite data for these inclusionsare between 180 and 3508C. This range is in agree-ment with the stable isotope thermometric dataŽ .230–3408C from amphibole and epidote in associa-tion with quartz of similar veins from the Sudbury

Ž .Structure Marshall et al., 1999 . Later, low-tempera-ture fluids trapped in secondary inclusions were pre-sent regionally, and they probably did not play a rolein the deposition of PGE enriched ore in veins.

Ž .Marshall et al. 1999 showed that similar low-tem-perature inclusions of late veins postdating the depo-sition of Ni–Cu–PGE ores are related to neotectonicŽ .5–13 Ma fluid mobilization.

Fluid-inclusion microthermometric data indicatethat some minerals in vein-type, Cu- and PGE-richdeposits in the North Range footwall were deposited

Ž .from intermediate-temperature 130–4408C , highlyŽsaline fluids Farrow and Watkinson, 1992; Li and

. Ž .Naldrett, 1993; Farrow, 1994 . Molnar et al. 1997´have shown that similar fluids at similar tempera-tures were involved in disseminations and veins ofCu–PGE–Au ore enrichment at the contact of theLindsley 4b Mine of the South Range. SEM-EDSanalyses on opened inclusions from quartz and sul-fides from these deposits also revealed substantial

Ž .amounts of heavy-metal phases Pb, Fe, Mn, Ba .North Range fluid inclusion data differ from those ofLindsley and Little Stobie only in the absence ofCO -rich inclusions. In contrast to the North Range,2

South Range rocks were more strongly affected byŽthe Penokean Orogeny 1.7–1.9 Ga; Card et al.,

.1984 that was penecontemporaneous with the em-Ž .placement of the SIC 1.85 Ga; Krogh et al., 1984 .

Therefore the presence of carbonic fluids along theSouth Range may indicate the incorporation of meta-morphic fluids into the hydrothermal system.

The results of this study and data for other veintype ore show that hydrothermal activity involvinghighly saline and heavy-metal rich fluids was notuncommon in the footwall of the SIC. Areas offootwall rocks affected by these highly saline hotfluids are characterized by consistent wall-rock alter-ation assemblages with Cl-bearing amphibole, epi-dote, chlorite, stilpnomelane, K-feldspar and similarvein-filling mineral parageneses with PGE mineralsŽ .Farrow, 1994; Watkinson, 1994 . These similaritiessuggest that hydrothermal processes responsible forwall rock alteration and deposition of vein-type orehad a common origin.

ŽThe relatively high temperatures up to 350–.4008C in all hydrothermal environments examined

in the Sudbury footwall units suggest that an appro-priate heat source for the circulation of fluids ex-isted. This source can not be of metamorphic originrelated to the Penokean Orogeny that affected theSouth Range because North Range deposits alsodisplay similarly high temperatures. Moreover, Cl-bearing alteration minerals in contact with sulfidesare not typical for metamorphosed magmatic–sulfide

Ž .deposits of South Range Watkinson, 1994 , whichalso suggests that the hydrothermal activity was notexclusively generated by metamorphism. Therefore,it is reasonable to relate the circulation of hot fluidsto the thermal effect of the large volume and compli-cated intrusive activity of the SIC.

The thermal effect of the SIC is expressed as acontact metamorphic aureole, 1–1.5 km wide, pre-

Ž .served along the North Range Dressler, 1984 . Aswell, near the Little Stobie Mine, a granite-dike

Ž .swarm that cuts back into the SIC Fig. 1 is inter-preted to be the result of partial remelting of the

ŽMurray Granite by the SIC Peredery and Morrison,.1984 . Recent studies on the radiometric age of the

hydrothermal alteration of the Onaping Formationthat was invaded by the SIC also demonstrate thatwidespread hydrothermal activity took place in con-

Ž .nection with SIC emplacement Ames et al., 1998 .These data imply that heat originating from the SICpenetratively affected the country rocks and could


have caused fluids to circulate. Convenient channelsfor fluid circulation were the brecciated zones along

Žthe contacts of the footwall units e.g., the green-.stone–granite contact at the Little Stobie Mine or

the fracture systems developed due to the emplace-Ž .ment of Sudbury Igneous Complex 1.85 Ga and

subsequent faulting and thrusting accompanying theŽpenecontemporaneous Penokean Orogeny 1.9–1.7

.Ga .The source of these fluids is not clear. Fluid

inclusion data from PGE-enriched sulfide pods orŽveins of the Bushveld Complex Ballhaus and

.Stumpfl, 1986 and from a shear-zone related PGEŽ .deposit at New Rambler, WY Nyman et al., 1990

show that saline fluids similar to those found in theLittle Stobie Mine were also present during theformation of these deposits. They are considered tobe the result of fluid activity related to late stages of

Žmagmatic activity in the Bushveld orthomagmatic–.deuteric enrichment of PGE in pegmatoids or fluid

mobilization along shear zones at New Rambler.Ž .Farrow and Watkinson 1996 assumed that deposi-

tion of Cu-rich veins from the North Range of theSudbury Structure was related to a magmatic–hydro-thermal system, where saline fluids of possibly for-mational brine origin were circulated by the heat ofthe SIC.

Ž .Frape and Fritz 1982 suggested that modern,highly saline Ca-rich fluid in the Sudbury Structurewas related to equilibration of formational fluid withArchean metamorphic rocks. The ubiquitous, sec-ondary, Type IIA inclusions rich in divalent cationsas reflected by their freezing behavior, and theirpossible trapping conditions, suggest a minimum1–4 km depth with temperatures about 100 C. If thegeothermal gradient had averaged 3 Cr100 m in thisarea of the Canadian Shield during migration ofthese fluids it is reasonable to assume that they arethe equivalents of the present-day, Ca-rich forma-tional fluids. Low-density CO -rich fluid could be2

channeled into this system along deep faults relatedto thrusting and neotectonic activity. The late CO -2

rich fluids had different characteristics than do thoseof higher density which were associated with high-temperature, more saline solutions trapped as TypeIII inclusions. This association may indicate that thehydrothermal system had mixed magmatic and meta-morphic components as is suggested by stable iso-

topic data from the mineral assemblages of vein typeŽ .ores Marshall et al., 1999 . However, the magmatic

signature of these fluids may just reflect the isotopiccomposition of heated brines from the footwall hav-ing been significantly modified during interactionwith magmatic sulfides and rocks. This interaction inthe high-temperature regime of the SIC is reflected

Žby the modification of primary ore textures e.g.,.replacement of pyrrhotite by chalcopyrite and re-

sulted in enrichment in heavy-metal cations. Due tothe increased amounts of these cations, fluids trappedin Type III high temperature inclusions have signifi-cantly different phase composition than do late sec-ondary, low-temperature inclusions.

Saline fluids with high Cl-activity are convenientagents for mobilization not only of Fe, Cu, and Pb,

Žbut for precious metals, especially Pt and Pd Wood.and Mountain, 1992; Gammons et al., 1992 . Thus,

interaction of hot, highly saline fluids with primarymagmatic ores could cause the remobilization andprecipitation of metals in convenient structures suchas stringers and veins penetrating the footwall lead-ing to the deposition of vein type ore with highprecious-metal grades.

8. Conclusions

Magmatic Fe–Ni–Cu–PGE sulfide deposition ac-companied the emplacement of the Sudbury Sub-layer, Footwall Breccia, and norite of the SIC. SomeCu–Ni–PGE–Au enrichment of typical Sudbury ore,and Cu-rich veins, in some cases with high contentsof Pt, Pd, and Au, were subsequently generated inparts of the footwall. The characteristics of fluidinclusions in quartz-bearing rocks containingpyrrhotite and chalcopyrite as major minerals sug-gest that hydrothermal processes were involved inmetal redistribution near the ore contact.

In orebodies of the Little Stobie Mine, high-tem-perature hydrothermal processes were outlined byfluid-inclusion studies. The temperature of early flu-ids was at least 280–3508C near the contact oforebody No. 2, and 180–2708C near orebody No. 1on the basis of the total homogenization tempera-tures of isolated inclusions. These processes werefollowed by several lower-temperature fluid-migra-tion stages, down to about 1008C. Compositionaldifference between early and late saline fluids is


reflected in the phase assemblages of fluid inclu-sions. High-temperature inclusions contain NaCl,KCl, Fe–Mn chloride, Pb–Ba chloride, as well asother daughter minerals, whereas late secondary in-clusions are not saturated with heavy-metal chlo-rides. Exact salinities of early fluids cannot be deter-mined due to the lack of the appropriate solubilitydata for chlorides in these multi-component systems.

ŽLate fluids were also saline 21–27 CaCl equiv.2.wt.% . All inclusion generations were associated with

unmixed CO -rich fluids; the latter contain minor2

methane. The density of these carbonic fluids de-creased from early high-temperature stages to thelate stages. The minimum pressure during the earlystages was about 2 kbars, and in the late stages

Žbelow 1 kbar. The origin of low-salinity about 1. ŽNaCl equiv. wt.% and high-temperature min. 200–

.3008C fluids trapped in secondary fluid inclusions isnot clear.

Although highly saline, high-temperature fluidswere also found in the fluid inclusions from barrenveins in the vicinity of the Little Stobie Mine, theirmicrothermometric properties differ from those atthe orebody contacts suggesting their different chem-ical composition.

The circulation of highly saline hot fluids wasprobably driven by intrusive activity of the SIC. Theorigin of CO may be related to metamorphism and2

deformation accompanying the Penokean Orogeny.The North Range was less severely affected by thisorogeny and was emplaced at shallower depth; hy-drothermal quartz there does not have CO -rich in-2

clusions detectable by microscopic methods.The interaction of hot, saline fluids with primary

magmatic ore could cause the mobilization of baseand precious metals with subsequent deposition assulfides in veins and stringers near the footwallcontact. These processes occurred not only at theLittle Stobie Mine, but at other places along theSouth and North Range. Areas where the activity ofhydrothermal fluids is revealed by the presence ofhigh-temperature, high-salinity fluids would beprospective targets in exploration for CuqPGE-enriched vein-type ores.

The low-temperature, Ca-rich brines, as well asvery dilute fluids trapped in late secondary inclu-sions did not play a role in the redistribution of baseand precious metals.

Acknowledgements

INCO kindly provided underground access to theLittle Stobie Mine, and gave permission for thepublication of these results. Special thanks for dis-cussions about this research are extended to P.C.Jones, C.E.G. Farrow, and D. Marshall. Researchexpenses for DHW were provided by NSERC grantA7874 and Carleton University. Part of this studywas also supported by the NATO–NSERC Postdoc-toral Science Fellowship to F.M. F.M. also thanksCarleton University for the excellent working condi-tions and support of travel and stay in Canada duringthe early stages of this project. Special thanks go toDr. I. Vergilov and Dr. K. Bogdanov at the St.Kliment Ohridski University, Sofia, for Raman anal-ysis of CO -rich inclusions. Reviews by J. Musgrave2

and P.E. Brown on the earlier version of this paperare highly appreciated.

References

Ames, D.E., Watkinson, D.H., Parrish, R.R., 1998. Direct datingof a regional hydrothermal system induced by the 1850 MaSudbury impact event. Geology 26, 447–450.

Bakker, R.J., Dubessy, J., Cathelineau, M., 1996. Improvementsof clathrate modeling: I. The H O–CO system with various2 2

salts. Geochimica et Cosmochimica Acta 60, 1657–1681.Ballhaus, C.G., Stumpfl, E.F., 1986. Sulfide and platinum miner-

alization in the Merensky Reef: evidence from hydrous sili-cates and fluid inclusions. Contribution to Mineralogy andPetrology 94, 193–204.

Bennett, C., Dressler, B.O., Robertson, J.A., 1992. The HuronianSupergroup and associated intrusive rocks. Ontario GeologicalSurvey, Spec. Vol. 4, Part I, pp. 549–591.

Bodnar, R.J., Vityk, M.A., 1994. Interpretation of microthermo-metric data for H O–NaCl fluid inclusions. In: De Vivo B.,2

Ž .Frezotti, M.L. Eds. , Fluid Inclusions in Minerals: MethodsŽ .and Applications: Short Course of the Working Group IMA

‘Inclusions in Minerals’. Pontignano-Siena, September 1–4,1994, pp. 117–130.

Bowers, T.S., Helgeson, H.C., 1983. Calculation of the thermody-namic and geochemical consequences of nonideal mixing inthe system H O–CO –NaCl on phase relations in geologic2 2

systems: equation of state for H O–CO –NaCl fluids at high2 2

pressures and temperatures. Geochimica et Cosmochimica Acta47, 1247–1276.

Brown, P.E., 1989. Flincor: a microcomputer program for thereduction and investigation of fluid inclusion data. AmericanMineralogist 74, 1390–1393.

Burruss, R.C., 1981. Analysis of phase equilibria in C–H–O–SŽ .fluid inclusions. In: Hollister, L.S., Crawford, M.L. Eds. ,

Short Course in Fluid Inclusions. Application to Petrology.


Mineralogical Association of Canada, Short Course Hand-book, 6, pp. 39–74.

Card, K.H., Gupta, V.K., Mcgrath, P.H., Grant, F.S., 1984. TheSudbury structure: its regional geological and geophysical

Ž .setting. In: Pye, E.G., Naldrett, A.J., Giblin, P.E. Eds. , TheGeology and Ore Deposits of the Sudbury Structure. OntarioGeological Survey, Special Vol. 1, pp. 25–43.

Cowan, E.J., Schwerdtner, W.M., 1994. Fold origin of the Sud-Ž .bury Basin. In: Lightfoot, P.C., Naldrett, A.J. Eds. , Proceed-

ings of the Sudbury-Noril’sk Symposium, Ontario GeologicalSurvey, Special Vol. 5, pp. 45–55.

Crawford, M.L., 1981. Phase equilibria in aqueous fluid inclu-Ž .sions. In: Hollister, L.S., Crawford, M.L. Eds. , Short Course

in Fluid Inclusions. Application to Petrology. MineralogicalAssociation of Canada. Short Course Handbook, 6, pp. 75–100.

Darling, R.S., 1991. An extended equation to calculate NaClcontents from final clathrate melting temperatures in H O–2

CO –NaCl fluid inclusions: implications for P – T isochore2

location. Geochimica et Cosmochimica Acta 55, 3869–3871.Davis, G.C., 1984. Little Stobie Mine: a south range contact

Ž .deposit. In: Pye, E.G., Naldrett, A.J., Giblin, P.E. Eds. , TheGeology and Ore Deposits of the Sudbury Structure. OntarioGeological Survey, Special Vol. 1, pp. 361–369.

Davis, D.W., Lowenstein, T.K., Spencer, R.J., 1990. Meltingbehavior of fluid inclusions in laboratory-grown halite crystalsin the systems NaCl–H O, NaCl–KCl–H O, NaCl–MgCl –2 2 2

H O, and NaCl–CaCl –H O. Geochimica et Cosmochimica2 2 2

Acta 54, 591–601.Diamond, L.W., 1994. Introduction to phase relations of CO –2

Ž .H O fluid inclusions. In: De Vivo, B., Frezotti, M.L. Eds. ,2

Fluid Inclusions in Minerals: Methods and Applications: ShortŽ .Course of the Working Group IMA ‘Inclusions in Minerals’.

Pontignano-Siena, September 1–4, 1994, pp. 131–158.Ž .Donnay, J.D.H., Ondik, H.M. Eds. , 1973. Crystal data determi-

native tables, Vol. 2: Inorganic compounds, 3rd edn. U.S.Dept. of Commerce, National Bureau of Standards, JointCommittee on Powder Diffraction Standards, U.S.A.

Dressler, O.B., 1984. The effects of the Sudbury Event and theintrusion of the Sudbury Igneous Complex on the footwallrocks of the Sudbury Structure. In: Pye, E.G., Naldrett A.J.,

Ž .Giblin, P.E. Eds. , The Geology and Ore Deposits of theSudbury Structure. Ontario Geological Survey, Special Vol. 1,pp. 97–136.

Ebel, D.S., Naldrett, A.J., 1996. Fractional crystallization of sul-fide ore liquids at high temperature. Economic Geology 91,607–621.

Farrow, C.E.G., 1994. Geology, alteration and the role of fluids inCu–Ni–PGE mineralization of the footwall rocks to the Sud-bury Igneous Complex, Levack and Morgan Townships, Sud-bury District, Ontario. PhD Thesis, Carleton University, Ot-tawa, Canada.

Farrow, C.E.G., Watkinson, D.H., 1992. Alteration and the role offluids in Ni, Cu and platinum-group element deposition, Sud-bury Igneous Complex Contact, Onaping–Levack area, On-tario. Mineralogy and Petrology 46, 67–83.

Farrow, C.E.G., Watkinson, D.H., Jones, P.C., 1994. Fluid inclu-sions in sulfides from North and South Range Cu–Ni–PGE

deposits, Sudbury Structure, Ontario. Economic Geology 89,647–655.

Farrow, C.E.G., Watkinson, D.H., 1996. Geochemical evolutionof the epidote zone, Fraser Mine, Sudbury, Ontario: Ni–Cu–PGE remobilization by saline fluids. Exploration and MiningGeology 5, 17–31.

Frape, S.K., Fritz, P., 1982. The chemistry and isotopic composi-tion of saline groundwaters from Sudbury basin, Ontario.Canadian Journal of Earth Sciences 19, 645–661.

Fueten, F., Redmont, D.J., 1997. Documentation of a 1450 Macontractional orogeny preserved between the 1850 Ma Sud-bury impact structure and the 1 Ga Grenville orogenic front,Ontario. Geological Society of America Bulletin 109, 268–279.

Gammons, C.H., Bloom, M.S., Yu, Y., 1992. Experimental inves-tigation of the hydrothermal geochemistry of platinum andpalladium: I. Solubility of platinum and palladium sulfideminerals in NaClrH SO solutions at 3008C. Geochimica et2 4

Cosmochimica Acta 56, 3881–3894.Haynes, F.M., 1985. Determination of fluid inclusion composi-

tions by sequential freezing. Economic Geology 80, 1436–1439.

Hoffman, E.L., Naldrett, A.J., Alcock, R.A., Hancock, R.G.V.,1979. The noble metal content of ore in Levack West andLittle Stobie Mines, Ontario. Canadian Mineralogist 17, 437–451.

Holloway, J.R., 1981. Compositions and volumes of supercriticalfluids in the Earth’s crust. In: Hollister, L.S., Crawford, M.L.Ž .Eds. , Short Course in Fluid Inclusions. Application to Petrol-ogy. Mineralogical Association of Canada. Short CourseHandbook 6, pp. 13–36.

Johan, Z., Watkinson, D.H., 1987. Phase fluide riche en Na–Cl–H–O–N et son role dans le concentration des elements groupeˆde platine et de la chromite: ‘Upper Critical Zone’, complexede Bushveld. In: Guides de prospection pour les gisements deplatinoides dans les complexes ophiolitiques et stratifies: Com-´mission des Communautes Europeennes, Rapport final, pp.´ ´61–85.

Krogh, T.E., Davis, D.W., Corfu, F., 1984. Precise U–Pb zirconand baddeleyite ages for the Sudbury Area. In: Pye, E.G.,

Ž .Naldrett, A.J., Giblin, P.E. Eds. , The Geology and OreDeposits of the Sudbury Structure. Ontario Geological Survey,Special Vol. 1, pp. 431–446.

Li, C., Naldrett, A.J., 1993. High chlorine alteration minerals andcalcium-rich brines in fluid inclusions from the StrathconaDeep Copper Zone, Sudbury, Ontario. Economic Geology 88,1780–1796.

Linke, W.F., 1965. Solubilities of Inorganic and Metal OrganicCompounds. American Chemical Society 2, Van Nostrand,1914 pp.

Marshall, D., Watkinson, D., Farrow, C., Molnar, F., Fouillac,Á.-M., 1999. Multiple fluid generations in the Sudbury ig-neous complex: fluid inclusion, Ar, O, H, Rb and Sr evidence.Chemical Geology 154, 1–19.

Mcgrath, P.H., Broome, H.J., 1994. A gravity model for theŽ .Sudbury Structure. In: Lightfoot, P.C., Naldrett, A.J. Eds. ,

Proceedings of the Sudbury-Noril’sk Symposium. Ontario Ge-ological Survey, Special Vol. 5, pp. 21–31.


Molnar, F., Watkinson, D.H., Jones, P.C., Gatter, I., 1997. Fluidínclusion evidence for hydrothermal enrichment of magmaticore at the contact zone of the 4b Ni–Cu–PGE deposit, Linds-ley Mine, Sudbury, Canada. Economic Geology 92, 674–685.

Morrison, G.G., Jago, B.C., White, T.L., 1994. Footwall mineral-ization of the Sudbury Igneous Complex. In: Lightfoot, P.C.,

Ž .Naldrett, A.J. Eds. , Proceedings of the Sudbury-Noril’skSymposium. Ontario Geological Survey, Special Vol. 5, pp.57–64.

Naldrett, A.J., 1989. Magmatic Sulfide Deposits. Oxford Univ.Press, New York, 186 pp.

Nyman, M.W., Sheets, R.W., Bodnar, R.J., 1990. Fluid inclusionevidence for the physical and chemical conditions associatedwith intermediate-temperature PGE mineralization at the NewRambler deposit, southeastern Wyoming. Canadian Mineralo-gist 28, 629–638.

Oakes, C.S., Bodnar, R.J., Simonson, J., 1990. The systemNaCl–CaCl –H O: I. The ice liquidus at 1 atm total pressure.2 2

Geochimica et Cosmochimica Acta 54, 603–610.Peredery, W.V., Morrison, G.G., 1984. Discussion on the origin

of the Sudbury Structure. In: Pye, E.G., Naldrett, A.J., Giblin,Ž .P.E. Eds. , The Geology and Ore Deposits of the Sudbury

Structure. Ontario Geological Survey, Special Vol. 1, pp.491–511.

Roedder, E., 1984. Fluid inclusions. Reviews in Mineralogy, Vol.12, 644 pp.

Schiffries, C.M., 1990. Liquid-absent aqueous fluid inclusions andphase equilibria in the system CaCl –NaCl–H O. Geochim-2 2

ica et Cosmochimica Acta 54, 611–619.Shepherd, T., Rankin, A.H., Alderton, D.H.M., 1985. A Practical

Guide to Fluid Inclusion Studies. Blackie and Son, Bishop-briggs, Glasgow, 239 pp.

Souch, B.E., Podolsky, T., Geological Staff, 1969. The sulfideores of Sudbury: their particular relationship to a distinctiveinclusion-bearing facies of the nickel irruptive. In: Wilson,

Ž .H.D.B. Ed. , Magmatic Ore Deposits, Economic GeologyMonograph 4, pp. 252–261.

Thiery, R., Van Den Kerkhof, A.M., Dubessy, J., 1994. vXproperties of CH –CO and CO –N inclusions: modelling4 2 2 2

for T -318C and P -400 bars. European Journal of Mineral-ogy 6, 753–771.

Ž .Trommsdorff, V., Skippen, G., 1986. Vapor loss ‘Boiling’ as amechanism for fluid evolution in metamorphic rocks. Contri-butions to Mineralogy and Petrology 94, 317–322.

Vanko, D.A., Bodnar, R.J., Sterner, S.M., 1988. Synthetic fluidinclusions: VIII. Vapor-saturated halite solubility in part of thesystem NaCl–CaCl –H O, with application to fluid inclu-2 2

sions from oceanic hydrothermal systems. Geochimica et Cos-mochimica Acta 52, 2451–2456.

Watkinson, D.H., 1990. Platinum-group minerals in Cu-rich sul-fides from some North American mafic–ultramafic rock com-plexes. Journal of the Mineralogical Society of Korea 6,37–51.

Watkinson, D.H., 1994. Fluid–rock interaction at contact of Lind-sley 4b Ni–Cu–PGE orebody and enclosing granitic rocks,Sudbury, Canada. Transactions of the Institution of Mining

Ž .Metallurgy Section B: Applied Earth Science 103, 121–128.Wood, S.A., Mountain, B.W., Pan, P., 1992. The aqueous geo-

chemistry of platinum, palladium and gold: recent experimen-tal constraints and re-evaluation of theoretical predictions.Canadian Mineralogist 30, 955–982.

Documents

Fluid-inclusion characteristics of hydrothermal Cu–Ni–PGE ...abyss.elte.hu/users/molnar/cikkek/Chem Geol 1999b.pdf · The high-temperature hydrothermal fluids were driven by