Abstracts No.34: SGEG 2nd National Meeting, 1993, Armidale

Geological Society of Australia

ABSTRACTS Number 34

SPECIALIST GROUP IN ECONOMIC GEOLOGY 2nd. National Meeting Armidale, 4th-5th February, 1993.

Abstracts for the 2nd. National Meeting of the Specialist Group in Economic Geology of the Geological Society of Australia.

4th-5th February, 1993 University of New England, Armidale

Organising

committee: P.K. Seccombe,

Sponsors:

Society of Economic

JM

Parr, M.T. Jones, P.G.

Geoiogists,

IGCP,

Flood.

PanContinentai

PREFACE

Welcome to Armidale! We hope that this, the Second National Meeting of SGEG proves as successful and as worthwhile as the first. The main themes of the technical program reflect suggestions made by the National Key Centres in Economic Geology and should prove to be of interest to many workers in ore deposit research and exploration from a wide variety of fields representing industry, government organisations and the universities. We are very pleased to be able to sponsor two keynote speakers to address the principal themes of the conference - Dick Sillitoe, London-based consultant, will lead off the session on felsic magmatism and ore deposits and Ken McClay from the University of London will provide the keynote lecture on the topic of microstructures as timing evidence In the genesis of stratabound base metal ores. We are sure that these two talks will be very well received. Unfortunately, a third keynote speaker. Dave London from the University of Oklahoma had to withdraw from the program due to health reasons. However, it is pleasing that Dave London's area of expertise, rare metal ores, will be represented by talks in the general sessions. We have a very full program, for which we thank all of the contributing authors. Most offers of oral presentations have been accommodated, however some authors have been asked to present their papers as posters to avoid the prospect of parallel technical sessions. Oral and poster presentations are collectively published in this volume. Many thanks go to Peter Flood and the NEO '93 team at Armidale, which have given the SGEG executive the organisational base to stage this conference. Our job has been made very much easier as a result. However, apart from the benefit of logistics, we believe that a 'back-to-back' program for the two meetings has many advantages, because those who attend both conferences have the opportunity to exchange ideas and views with another group of specialists and can compare problems from a particular region with broader issues and concepts. We hope that you agree! Of course, those participating fully in both conferences have the prospect of weathering five packed days of technical presentations! Finally, I would like to thank Joanna Parr for her hard work and help in preparing for this conference and in keeping the SGEG executive running in Newcastle for the past two years. SGEG now moves to Perth and we wish the Key Centre at the University of Western Australia well in furthering the group and its profile.

Phil Seccombe Chairman, Specialist Group in Economic Geology

PROGRAMME OF ORAL PRESENTATIONS AND POSTER DISPLAYS

THURSDAY. 4TH FEBRUARY 8.50-9.00 Welcoming ceremony

Hydrothermal mineral deposits associated with felsic magmatism: 9.00-9.50 keynote address: R.H. Slilitoe: Gold deposits in intrusion-centred systems. 9.50-10.10 G.D. Carman: Transitional magmatic to epithermal mineralisation at the Lihir Island gold deposit, PNG: Implications for the evolution of ore-forming fluids from fertile magmas. 10.10-10.30 K.C. Lawrie: Gold localisation in the inner contact aureole of the Tabletop Granite, Pine Creek Goldfield, N.T. 10.30-10.50 D.R. Cooke, M.S. Bloom and I. Cartwright: The Acupan South porphyry copper-gold prospect, Baguio District, Philippines: relationship to epithermal mineralisation. 10.50-11.20 COFFEE 11.20-11.40 J.R. Ridley, D.I. Groves, E.J. Mikucki and N.J.McNaughton: Indirect evidence for a granitoid fluid source of Archean lode-gold deposits. 11.40-12.00 N.J. McNaughton, 8. Sheppard and N.M. Goellnicht: Understanding the nature of proximal-distal mineralization in thermal aureole gold deposits in the Proterozoic of northern Australia using lead isotopes. 12.00-12.20 A. Andrew: 6^®0-exchange and evolution of magmatic hydrothermal ore fluid 12.20-12.40 P. Kitto: Oxygen isotopes as an indicator of large-scale hydrothermal paleoflow and economic Sn mineralisation, Renison Mine, western Tasmania. 12.40-1.40 LUNCH 1.40-2.00 J. Zhiyu and C. Minyang: Stable isotope evidence for two sources of mineralisation at the Daboashan polymetallic deposit, southeastern China. 2.00-2.20 S.M. Rowins, D.I. Groves, N.J. McNaughton, P.E. Brown, R.L. McLeod and D. Hall: Evidence of unusually carbonic and reduced ore fluids in the Late Proterozoic Seventeen Mile Hill porphyry copper-style deposit, Telfer district, Western Australia.

2.20-2.40 N.M. Goellnicht, D.I. Groves and N.J. McNaughton: The role of Late Proterozoic fractionated granitoids In the genesis of polymetalllc mineralisation in the Telfer district, W.A. 2.40-3.00 K.G. McQueen and C. Perkins: Granitoid-related gold mineralisation In the Braidwood Granodlorlte, southeastern N.S.W. 3.00-3.20 T.M. Leach and G.J. Corbett: Porphyry-related carbonate-base metalgold systems: The transition between epithermal and porphyry environments. 3.20-3.50 TEA 3.50-4.10 T. Zhou: Zoning of the Fengsan Dong skarn copper-molybdenum deposit, China: shrink or prograde? 4.10-4.30 P.J. Williams and M. Heinemann: Maramungee: A Proterozoic Zn "skarn" in the Cloncurry Metamorphic Terrain. Is there a magmatic connection? 4.30-4.50 N.J. McNaughton, P.J. Pollard, J. Stacey, D.l. Groves and R.G. Taylor: An extreme high heat-producing Sn-W granite from the Bushveld complex: evidence for a long-lived hydrothermal system. 4.50-5.10 G.J. Davidson and H.L. Paterson: Oak Dam East: A prodigious, uraniumbearing, massive iron-oxide body on the Stuart Shelf 5.10-5.30 D.J. Whitford, R. Sharpe and B. Gemmell: Origin of barite from the Hellyer VHMS deposit, Tasmania: a Sr Isotopic study. 5.30-5.50 S.K. Matthal and R.W. Henley: Structurally controlled fluid pathways, fluid mixing and gold precipitation in Intrusive-related hydrothermal systems.

FRIDAY, 5TH FEBRUARY Macro- and micro- structural relationships in stratabound ore deposits: 9.00-9.50 keynote address: K.R. McClay: Deformation texture In stratiform massive sulphide deposits. 9.50-10.10 W.G. Perkins: Timing of replacement processes In Mount Isa copper and lead-zinc orebodles using stmctural and microstructural relationships. 10.10-10.30 R.K. Valenta and A. Wilson: The effects of deformation and alteration on metal zoning patterns in the Hilton Mine, Mount Isa. 10.30-10.50 G. Broadbent and S. McKnight: Microstructures of ore minerals from the Century deposit, northern Queensland.

10.50-11.20 COFFEE 11.20-11.40 N. Phillips: Microstructures for timing ore introduction: Caution 11.40-12.00 J. Parr: Micro-textural and geochemical evidence for hydrothermal venting at the Pinnacles deposit, western New South Wales. 12.00-12.20 S. Bodon: Genetic aspects of the Currawong Zn-Cu-Pb(-Au) massive sulfide deposit, Benambra, northeastern Victoria. 12.20-12.40 B.S.E. Mapani and C.J.L. Wilson: Gold mineralization at Magdala Mine, Stawell, western Victoria. 12.40-1.40 LUNCH General topics and rare metals associated with felsic magmatism: 1.40-2.00 P.J. Pollard, R.P. Taylor and R.G. Taylor: Extreme chemical fractionation and late stage fluid evolution associated with Ta-Nb-Li mineralization at the Yichun mine, south China. 2.00-2.20 T. Han and P.L.F. Collins: Rare metal mineralisation during magmatichydrothermal transition in the Greenbushes Pegmatite western Australia. 2.20-2.40 C.A. Heinrich, C.G. Ryan and T. P. Mernagh: Ore metals In magmatic brine and Vapour: New evidence from PIXIE microanalysis of fluid inclusions 2.40-3.00 D.C. McPhail: The behaviour of iron in high-temperature chloride brines. 3.20-3.50 TEA 3.50-4.10 H. Bresser and R. Myers: Timing of brine migration in the Lawn Hill Platform: Evidence from the Lawn Hill Mineral Field. 4.10-4.30 G-Y. Dong and G. Morrison: Character of Adularia in epithermal veins, Queensland. 4.30-4.50 R.A. Binns and J.C. Eames: Sea-floor hydrothermal deposits of Pual Ridge, eastern Manus Basin, Papua New Guinea. 4.50-5.20 D.L. Huston, S.H. Sie, G.F. Suter and C.G. Ryan: Compositional studies of pyrite from volcanic-hosted massive sulphide deposits In eastern Australia. 5.20-5.40 G.R. Carr, J.A. Dean, G.W. Morrison and D. Suppel: Precise isotopic fingerprinting of fertile hydrothermal events - Exploration applications and examples from the Lachlan and Thompson Fold Belts systems.

POSTERS: A.L Bainbridge, G.J. Corbett and T.M. Leach: The Nena High sulphidation epithermal Cu/Au system, Frieda River Copper, Papua New Guinea: The application of new ideas. P.L. Blevin: Halogen (F, CI) content of biotite from I- and S-type granites of eastern Australia: possible metallogenic implications. I.M. Hart: Hematite-barite alteration in the Owen Conglomerate, North Lyell, Tasmania. D.L. Huston, M. Power and R.R. Large: Laser ablation analysis of sulphur isotopes: an analytical technique now available in Australia. J.T. Knight, J.R. Ridley. D.I. Groves and R. Napier: The genesis of high temperature Archaean lode-gold mineralization in the Coolgardie Goldfield, western Australia: the role of granitoids. T.P. Mernagh and W.K. Witt: The role of methane in gold mineralisation in the Menzies-Kambalda area, eastern Goldfields, W.A. V.J. Ojala, J.R. Ridley, D.I. Groves and G.C. Hall: Granny Smith: an example of a granitoid-hosted Archaean lode-gold deposit. M.J. Roach: Geology and geophysics of the Lilse-Golconda Goldfield, northeast Tasmania. R.G. Skirrow and J.L. Walshe: The West Peko Au-Cu-Bi deposit, Tennant Creek, NT: Fluid inclusion constraints on ore fluid chemistry.

6180.EXCHANGE AND EVOLUTION OF MAGMATIC HYDROTHERMAL ORE FLUID Anita S. Andrew* CSIRO Division of Exploration Geoscience, North Ryde 2113 The oxygen isotope fractionation between minerals and water have been modelled to detemnine the oxygen isotope composition and to identify the source of the ore fluids in SKIOL"^'"®^''^®^ intrusive complexes. The first example, an intrusive complex at Fifield (NSW). IS formed from water-poor magma with oxygen isotope fractionations indicative of isotope exchange between minerals only. The second example, from the Bushveld Complex (Republic of South Africa), and the third example, from the breccia pipe at Kidston (North Queensland), formed from relatively water-rich magmas and oxygen isotope fractionations indicate exchange with a cooling magmatic hydrothermal fluid. In the second example the magmatic system was closed to externally derived fluids whereas this is more problematical in the third example, where the ore fluid fomied by explosive devolatilization of a water-rich magma. Platinum-group-element mineralization is hosted by pyroxenites and peridotites within zoned Alaskan-type intrusive complexes at Fifield (Johan et al. 1989). The Owendale Intrusive Complex and adjacent Tout Intrusive Complex comprise peridotite, dunite, clinopyroxenite. homblendite. monzogabbro. monzodiorite and monzonite. Magmatic layering is variably developed in most rock types except dunite and peridotite. Pd-poor and S-poor mineralization is predominantly in cross-cutting clinopyroxene-rich veins and pipes called P-units viflthin olivine-bearing pyroxenite. biotite pyroxenite, pyroxenite and homblendite The veins are commonly monomineralic with textural zoning detemiined by crystal size and the ennchment of biotite and magnetite in adjacent host rocks. Oxygen isotope fractionation between clinopyroxene. plagioclase and magnetite indicate high temperature crystallization. 5I80 values of clinopyroxene and magnetite from least altered pyroxenites and P-units are 5.5 and 4.1 permil V-SMOW, respectively, suggesting temperatures as high as 1300»C from a mantle-derived magma with 6 'OO of 5.7. All other values plot within the triangular fields defined by these primary values on 6-6 plots, demonstrating closed-system subsolidus exchange to <670°C in a vvater-poor environment. In pyroxenites and P-units. exchange of ' S Q has enriched clinopyroxene and biotite and depleted magnetite. In the more felsic rocks, feldspar has been ennched and clinopyroxene and magnetite have become depleted in 180. Tin^ungsten-rare earth element mineralization at Zaaiplaats (Pollard et al.. 1991). is hosted by the Bobbejaankop and Lease Granites - miarolitic. brick-red alkali feldspar granites which fom at high levels within the Lebowa Suite of the Bushveld Complex. Pervasive hydrothermal alteration of the granites resulted in replacement of original biotite with chlonte. senate, carbonate and Ti oxides, development of coarse vein and patch perthite and antiperthite in alkali feldspar, and precipitation of hydrothemial minerals in miarolitic Mvites. Cassiterite occurs as a cavity filling and replacement mineral in tabular, sub honzontal zones and in shallowly-plunging and branching pipe systems in the granites Disseminated cassiterite mineralization exhibits no relationship to fracture systems and hydrothermal fluids appear to have evolved essentially in situ during crystallization to produce miarolitic cavities. Oxygen isotope fractionations between coexisting quartz and alkali feldspar are reversed indicating disequilibrium. On a 6-^ plot quartz-feldspar pairs are consistent v ^ feldspar and quartz exchanging with an evolving magmatic hydrothemial fluid; the rate of exchange of

oxygerisotor^^^^^

t T

r i p e r ^ L t 200»C Th^n'^^ ^ichTe b ^^JasSel

fluid inclusion temperature estimates and hydrothermal fluid decreases from 6 at >500»C to

rnl?!*

trajectories on the 5-5 plot of serlcite. chlorite

^

The Kidston breccia pipe (Baker and Andrew. 1991) is a gold-rich sub volcanic breccia dykes which intrude Proterozoic metamorphic and Siluro-Devonian granitoid host rocks The

I

pta°e:

^ Permo-Carbo^Ss^'amXl^^^^^

maamLc h S Z S f o

S

and are all associated with magmatic and/or

by K a T y L ' " " " ' "

TSTJ!^ brecciation resulted from multiple buildup of high S)To (>500'C) highfy saline (>40 wt% NaCI equiv) magmatic fluid (5I80 = 8. 5D = ^ J L c? ^ ^ ^ ^ ^ crystallizing rhyolite stocks. Crenulate quartz layers w^ttiin S d rocks resulted'fromryclic buildup and release of fluid pressures within the crystallizing rhyolite The droo in fluid ( l o J s C b a S an7resuSam which condensed to a liquid with similar salinity. i r i f n i l ® ? n ! f K ? " M i n e r a l i z a t i o n deposited from a liquid with low salinity (5-10 wt% NaCI equiv) and magmatic isotopic signature (5I80 = 4-8. 6D = -50 to 340-500°C. The characteristics are similar to liquid inclusions within prebreccia mineralization and the fluid is interpreted to have fomied similariy by condensation of vapour produced by boiling highly saline magmatic fluid at depth. The alteration associated with this fluid (quartz-epidote-sulfide cavity fill) is locally overprinted by a seconda^ biotite (-magnetite-pyrrhotite) alteration associated with the introduction into higher levels of the pipe of the highly saline, higher temperature magmatic fluid. The latestage postbrecaa mineralization hosting economic-grade gold within sheeted veins and late cavities was deposited between 380 and 170«C. Fluid inclusion salinities (2-10 wt% NaCI equiv) and calculated isotopic composition of the fluid (S^Sq = 2-8, 6D = -90 to -70) suggest no significant quantities of lower salinity. 18o/D-depleted meteoric fluids were introduced into the system at this stage. The stable isotope composition of hydrothermal ore fluids from three styles of magmatic ore deposit demonstrate the importance of exchange processes within the cooling magmas The resultant oxygen isotopic composition of the fluid can be demonstrated to evolve as a consequence of the exchange. References Baker. E.M. and Andrew. A.S. 1991. Economic Geology 86, 810-830. Johan. Z.. Ohnenstetter. M.. Slansky. E.. Barron. L.M., and Suppel. D. 1989. Mineralogy and Petrology 40, 289-309. Pollard, P.J., Andrew. A.S. and Taylor. R.G. 1991. Economic Geology 86,121-141.

THE NENA HIGH SULPHIDATION EPITHERMAL CU/AU SYSTEM, FRIEDA RIVER COPPER, PAPUA NEW GUINEA: THE APPLICATION OF NEW IDEAS Bainbridge A L* (1), Corbett G J (2) & Leach T M (3) 1. AMB Exploration 5 Key Court Noosa Heads 4567 2. Corbett Geological Services, 29 Carr St North Sydney 2060 3. CMS New Zealand Ltd., 11 Blake St., Auckland 1, Exploration at Frieda River up to 1983 inferred a porphyry copper resource of 860 Mt @ 0.47% Cu and 0.31 g/t Au within the Koki and Horse-lvaal deposits and 32 Mt @ 2.35% Cu and 0.58 g/t Au within the Nena high sulphidation deposit, located 6 km north-east of the porphyry deposits (Hall et al 1990). An increase in the understanding of high sulphidation Cu/Au mineralization and the relationship to buried porphyry copper deposits, in particular Lepanto, Philippines (Garcia 1990) and Wafi, PNG (Leach and Erceg 1990, Erceg et al 1991), during the gold boom of the 1980's, has facilitated a re-evaluation of the Nena mineralization. The Nena Prospect occurs on the margin of the Frieda River porphyry copper intrusive system in what is inferred to represent a down faulted block exposing both epithermal and porphyry copper mineralization at roughly the same elevation. The NW trending Frieda Fault, formed as a splay fault from the more regional EW trending Fiak-Leonard Schultz Thrust", localises the porphyry copper intrusive centre and, an inferred dextral rotation, has imparted a dilational character to the set of structures parallel to the Frieda Fault. The dilational structures which host the Nena and North Debom Prospects are defined as a 3-4 km wide series of NW trending silica and silica-alunite ridges which extend for over 10 km from the Horse-lvaal porphyry copper deposits. The classical high sulphidation system at Nena occurs as NW oriented concentric bulbous alteration zones consisting of a central mineralized vughy silica, grading outward to a barren quartz-alunite zone (locally sulphur-bearing) and rimmed by thin zones of pyrophyllite-dickite-kaolinite, interlayered illite-smectite and carbonate-gypsum-chlorite. The alteration is interpreted to have formed from acid leaching by an initial vapourrich magmatic fluid phase (White 1990) which migrated laterally in th enorth to south direction along dilational feeder structures. The gradation from broad central zones of vughy silica outward to quartz-alunite alteration is postulated to have formed as a response to the progressive cooling and neutralization of this acidic fluid through rock reaction; whereas peripheral thin clay zones are suggestive of rapidly changing fluid physio-chemistry upon mixing with circulating meteoricdominated fluids. The alteration shows a preference for the permeable volcaniclastic units within a sequence interlayered with lavas. Copper and gold mineralization are associated with a later, predominantly liquid, magmatically derived fluid which has utilised the same feeder structures as the volatile phase, and brecciated the earlier competent vughy silica. Fractures, breccias and open leached vughs have been sealed by initial multiple phases of pyrite. Copper mineralization occurs as late stage enargite (luzonite) deposition in cavities and fractures in the pyrite, in places intergrown, and locally rhythmically banded with barite. Intense brecciation and local fluidised breccias accompany high grade copper mineralization within the central vughy silica zones; whereas more fracture controlled sulphide deposition results in low grade Cu-mineralization in the peripheral quartz-alunite zones. Primary gold mineralization is postulated to occur as tellurides, possibly as submicroscopic inclusions in sulphides. Supergene

leaching has resulted in an oxidised gold zone, overlying a zone of supergene covellite-chalcocite enrichment. Initial fluid studies on barite associated with copper mineralization indicates that the mineralized fluid was two phase, relatively hot (>300-350 deg C) and moderately saline (>9-10 wt % equivalent NaCI). Mineralization responded to rapid cooling upon mixing with a low temperature (<150-200 deg C), dilute (<1-2 wt % NaCI) liquid. High sulphidation systems in the South-Pacific region may occur as: i) high temperature, barren, alunite-silica-andalusite caps to porphyry systems, (Sillitoe & Gappe 1984); eg adjacent to the Horse-lvaal at Frieda. ii) relatively low temperature gold-rich systems which exhibit broad alteration zonations with gold deposited in mixing zones; eg Wafi (Leach and Erceg 1990). ill) moderate temperature gold-copper systems with bulbous structurally controlled vughy silica and quartz-alunite assemblages; thin peripheral clay zones; and central silica-alunite mineralized zones; eg Lepanto (Garcia 1990) Nena has many characteristics similar to Lepanto and Wafi in which dilational structures focus fluid flow, providing an ideal plumbing system to tap magmatic fluids from depth. Fluid flow vectors at Nena are given by: the gradation in the morphology of the copper-bearing structures from subvertical feeder zones to concentric outflow pipe-like features, changes in alteration mineralogy, the distribution and characteristics of silica-alunite ridges. The emphasis of present exploration is to upgrade the current Nena resource, evaluate the presence of a buried porphyry copper source to Nena (similar to the FSE at Lepanto of Rafferty's at Wafi), and to explore for other high sulphidation deposits beneath oxidised, gold-bearing, silica-alunite ridges elsewhere in Nena-North Debom region. Kind permission for publication from the Frieda Copper Partners is acknowledged. REFERENCES CITED Hall R J, Britten R M & Henry D D 1990. Frieda River copper gold deposits, in Geology of the Mineral Deposits of Australia and Papua New Guinea, Ed. F E Hughs pp 1709-1715 Aust. I. M. M. Leach T and Erceg M M 1990. The Wafi high sulphidation epithermal gold deposit, Papua New Guinea Proc Pacific Rim Congress 90, pp 451-456. AIMM Erceg M M, Craighead G A, Halfpenny R and P J Lewis 1991 The exploration history, geology and metallurgy of a high sulphidation epithermal gold deposit at Wafi River, Papua New Guinea AII\/IM, PNG Geology, Exploration and Mining Conference Rabaul PNG pp 58-65 Sillitoe R H and Gappe I M 1984 Philippine porphyry copper deposits: Geologic settings and characteristics CCOP technical report 89 pp. Garcia J S 1990 Geology and mineralization characteristics of the Mankayan Mineral District, Benguet, Philippines. Third Symposium on Deep Crustal Fluids - . Geol Survey of Japan pp 17-26 White N (1990) High sulphidation epithermal gold deposits: characteristics and model for their origin. Third Symposium on Deep Crustal Fluids. Geol Survey of Japan pp 5-14.

SEA-FLOOR HYDROTHERMAL DEPOSITS OF PUAL RIDGE, EASTERN MANUS BASIN, PAPUA NEW GUINEA

R.A. Binns and J.C. Eames

CSIRO Division of Exploration Geoscience, North Ryde, NSW

Three sites of hydrothermal activity have been detected on Pual Ridge, a "Y"-shaped felsic volcanic edifice lying between two major transform faults in the Eastern Manus Basin at an extensional rift zone (precursor of sea-floor spreading) within earlier island arc crust. The underlying dacites, with some andesites and rhyodacites, form a calc-alkaline fractionation series with arc geochemical affinities (Si02, 58-74%; Mg#,41-17; K/Na atom ratio, 0.19-0.28; lowTi,Nb; nCe/nYb, 1.2; Eu*, 1). The more extensively-studied PACMANUS site lies 1650-1675m deep at a bathymetric high on the stem of the "Y", close to a zone of probable cross-faulting. Chimneys to 4m or more high and sulfide mounds up to 300m long extend discontinuously for some 3km along a crestal zone about 500m wide (Fig. 1), overlying dacite lava outcrops and banks of dacitic hyaloclastite. Dark bluish grey deposits and white mats also occur, presumed to be Mn-oxides and bacterial growths respectively: some of the former are aligned along crevices in apparently altered dacite. Faunal concentrations, shimmering water, and white smoke denote active hydrothermal venting in places. Samples of anhydrite and massive sulfide recovered from a PACMANUS chimney have chalcopyrite, with rare bomite inclusions, as the dominant sulfide. In the massive sample this is altered marginally to tennantite, in turn overgrown by sphalerite with rare galena. Analyses of anhydrite-rich and massive samples respectively include:- Cu; 10.5, 30.3%: Zn; 0.34, 1.4%: Pb; 95, 355 ppm: Mo; 118, 360 ppm: In; 31, 104 ppm: Ag; 19, 42 ppm: Au; 2.1, 10.0 ppm: U; 330, 380 ppb: As; 62, 2440 ppm: Sb; 45, 320 ppm. Rare earth elements are highly fractionated (nCe/nYb; 8, 52), with pronounced positive Eu anomalies (Eu*; 11,28). Rare samples of altered vesicular dacite dredged at PACMANUS are composed mainly of white mica. Relative to parent dacite glass (Fig. 2) these are distinctly depleted in Mo, Si, Fe, Sr, Na, Ca, Cs, Co, Mn, and Li, and enriched in K, Mg, Zn, Ba, Tl, Pb, Cr, Bi, U (4.9 ppm), Sb, Cu (640 ppm). As (20 ppm), Au (48 ppb), and S (4250 ppm). Ti is depleted by some 30% relative to Al, Zr, La, Ce and P, which cluster at a ratio of 1.6 (Fig. 2) and imply substantial volume reduction during alteration if they were immobile. Heavy rare earth elements are depleted some 40% relative to LREE. These results, apparently the first ever on pristine sea-floor alteration of felsic volcanic rocks, imply problems for some "traditional" lithogeochemical interpretations applied to ancient VMS environments.

An isolated chimney and several dark mounds were observed on dacite outcrops at a second site 6.5 km NE of PACMANUS (about 1750m depth near the crest of the eastern arm of the Pual Ridge "Y"). Fragments of vesicular dacite with altered surfaces were recovered from a nearby cliff The third site, with a fauna-encrusted spire and dark patches on dacite outcrops, occurs 10km NE of PACMANUS, at about 2000m depth, on the lower eastern flank of the western arm of Pual Ridge. Besides fresh dacites, many manganiferous fragments were dredged. These display cm-thick layers of sandy dacite hyaloclastite (or fine talus), overlying semi-consolidated hemipelagic ooze. Volcanic clasts at the former sea-water interface are coated by dull black Mn oxide, below which the dacitic sand is cemented by lustrous black Mn oxide passing downwards into soft brownish Fe-rich material. Veinlets of flinty to soft Fe oxide cut the underlying ooze and merge into the lower ferruginous hyaloclastite cement. The cementing deposit contains little Si, and Mn/Fe is variable (0.2 to 10, cf 0.07 for underlying vein). Manganiferous variants are enriched in Na. Anomalously high contents occur of P, S (to 3%), Li, V, Cr, Ni, Cu (to 200 ppm), Zn, As (to 500 ppm), Sr, Mo, Ag (to 200 ppm), Sb, Ba, Au (to 30 ppb), Pb (to 180 ppm), and U (to 140 ppm). The ferruginous vein material is rich in CaO (7.7%) and P205 (18%), and has a similar range of anomalous trace elements to the cements (including As, 1400 ppm; Ag, 270 ppm) except that Zn, Cu, Pb and Au enrichments are lacking. These samples appear to be the first knovm modem examples where the transitionfi-omfeeder to immediately subsurface deposit to true exhalite can be examined. When compared with data for alteration at PACMANUS, enriched lithophile elements in such low-T hydrothermal deposits might derive fi'om the immediately underlying volcanics, but the chalcophile elements have a deeper source. Elevated metal contents occur in hemipelagic ooze at the PACMANUS site (e.g. Zn up to 8700 ppm) and lower levels (e.g. above 120 ppm Zn) define a pronounced halo extending at least 15 km under the particulate plume mapped near 1600m in the seawater column above PACMANUS. The felsic volcanic-hosted PACMANUS deposit is the closest modem analogue yet known for a major category of ancient VHMS ore deposit. The observations above are based on comparatively few samples and are accordingly preliminary, but they help interpret phenomena encountered in land-based exploration for fossil equivalents. They will be refined and expanded when more samples are recovered during cruises and submersible dives planned at the site in 1993-94.

3'43'S

3*44'S

ISIMI-E

15r40'E ••• Hydrothermal deposits yir Abundant biota

f

Camera track Fault zone

Fig. 1. Bathymetry ofthePAC^MNUShyd^othe^mal site on Pual Ridge, showing distribution of hydrothermal deposits encountered during bottom camera tows. Occurrences of abundant biota probably indicate further ofF-track deposits.

<D

100

O CO

o

JO (0 (D -o <D k.

0)

HH <

Depleted

Fig. 2. Geochemical comparison of fresh and altered dacite at the PACMANUS site. Enriched and depleted elements are shaded assuming A1 as the immobile element, but this is as yet unsubstantiated. Missing REE conform to the pattern shown by La-EuLu.

GENETIC ASPECTS OF THE CURRAWONG Zn-Cu-Pb(-Au) MASSIVE SULFIDE DEPOSIT, BENAMBRA, NORTHEASTERN VICTORIA. S. B. Bodon VIEPS, Department of Earth Sciences, Monash University, Clayton, Victoria, 3168. broadly divided into three groups: i) massive to banded ores containing > 70% sulfide, ii) disseminated ores containing 5% to 70% sulfide and iii) late overprinting veins and breccias that are infilled with possible remobilized base-metal sulfides during deformation. Three generations of deformation have been recognized in the host stratigraphy and can be correlated to deformation textures observed within the ore, based on consistent kinematic geometries. This is unequivocal evidence that mineralization pre-dates the earliest known deformation observed within the host stratigraphy. The internal stratigraphy of the deposit tends to be erratic with the development of ore fades, changing laterally over tens of metres; however, there is a broad consistent vertical zonation of sphalerite-stilpnomelane and chalcopyrite-chlorite stringer mineralization in the footwall that grades upward into pyrite-magnetite which then grades into a zone of intermingled massive pyrite and disseminated sphalerite-galenaarsenopyrite ores that are associated with small intervals of hydrothermally altered host stratigraphy. Metal zonation characteristics also tend to be erratic except for lens 2 which displays an increasing Cu/Cu+Zn ratio (Cu ratio) from the hangingwall to the footwall. Gold displays two associations: i) Au-As-Zn association occurs throughout lens 1, and occurs toward the top of lens 2 and, ii) Au-BiCu association typically occurs in the footwall of lens 2 and is associated

The Late Silurian Currawong Zn-Cu-Pb(-Au) massive sulfide deposit is located in the northeastern highlands of Victoria, southeastern Australia. Since the discovery of the deposit in 1979, there has been debate over the origin of the mineralization. Two schools of thought have developed over the past decade: i) the deposit is representative of a volcanichosted massive sulfide deposit and is syngenetic, or ii) it is a structurally controlled epigenetic replacement-type deposit that is localized within a highly strained zone. This contribution aims to resolve the timing of mineralization relative to the deposition of host stratigraphy and deformation, and to propose a genetic model, based on metal zonation and textural studies, resolving the syngenetic versus epigenetic controversy. Delineation of the latter aspects of the deposit has an obvious relevance to exploration within the area. The deposit is comprised of two stratabound concordant massive sulfide lenses (lenses 1 and 2) that have been dissected by late subvertical faulting. No discordant stockwork feeder system exists. Host stratigraphy to the deposit is dominated by a Late Silurian volcano-sedimentary package of rocks (Gibsons Folly Formation). In the vicinity of mineralization the host sequence is composed of sub-aqueous rhyodacitic to andesitic breccias and coherent volcanics, and equivalent (temporally and spatially) epiclastic rocks that are contained in and interfinger with a monotonous succession of turbiditic sediments. Mineralization can be

8

with chalcopyrite-chlorite stringer mineralization. However, there are a number of occurrences where there is no elemental or mineralogical association, suggesting that at least some of the gold was either remobilized or introduced during deformation. The limited primary (predeformational) textures that are preserved within the deposit are dominated by mesoscopic replacement textures of host stratigraphy and microscopic replacement textures of various minerals. Primary depositional textures such as botryoidal and framboidal textures are best preserved in pyrite and arsenopyrite; however, they do not signify evidence for an exhalative depositional origin for massive ore types, as they also occur in disseminated ores that were clearly formed by the replacement of host stratigraphy. Taking into account the internal stratigraphy, metal zonation characteristics, primary textures, and in addition the paragenesis of the ores and limited stratigraphic reconstruction, the deposit is interpreted to be subseafloor replacement in origin and not exhalative or a structurally controlled epigenetic-type. The sub-seafloor replacement model proposes that hydrothermal fluid was selectively focussed upward through a syndepositional fault related to a Late Silurian rifting event, and where the hydrothermal fluid encountered porous permeable strata, it infiltrated laterally through the unconsolidated rock (i.e. fluid migration was tightly stratigraphically controlled). The geometry and distribution of ore types in the Currawong deposit is characteristic of both sheet- and layered-type deposits in Australia (based on the styles of Australian volcanic-hosted massive sulfide deposits after Large, 1992). The genesis of the deposit is consistent with the replacement model (after Large, 1992) for the genesis of sheet-

type volcanic-hosted massive sulfide deposits. The Currawong deposit, however, contains abundant stilpnomelane that is enigmatic for volcanic-hosted massive sulfide deposits in Australia. Instead, the deposit bears a closer resemblance, mineralogically, to Archean "Norandatype" Cu-Zn volcanic-hosted massive sulfide deposits (Franklin, 1990) of the Abitibi subprovince in Canada, especially the Mattagami (Costa, et. al., 1983; Roberts and Reardon, 1978) and Phelps Dodge (Kranidiotis and MacLean, 1987) deposits, where the occurrence of stilpnomelane has been attributed to a high potassium content of the ore-forming hydrothermal fluid.

References. Costa. U. R.. Barnett, R. L., and Kerrich, R.. 1983. The Mattagami Lake Mine Archean Zn-Cu sulfide deposit, Quebec: hydrothermal coprecipitatton of talc and sulfides in a sea-floor brine pool evidence from geochemistry. I S o / i e o , and mineral chemistry: Econ. Geol., v. 78. p. 1144-1203. Franklin. J. M.. 1990. Volcanic-associated massive sulphide deposits, in Ho, S. E., Robert. F.. and Groves. D. I., eds.. Gold and Base-metal Mineralization in the Abitibi Subprovince, Canada, with Special Emphasis on the Quebec Segment: short course notes. Geology Department (Key Centre) and University Extension, The University of Westem Australia, publ. no. 24. p. 211-241. Kranidiotis, P.. and MacLean. W. H.. 1987. Systematics of chlorite alteration at the Phelps Dodge massive sulfide deposit. Matagami. Quebec: Econ. Geol., v. 82. p. 1898-1911. Large, R. R., 1992. Australian volcanic-hosted nfiassive sulfide deposits: features, styles, and genetic models: Econ. Geol., v. 87. p. 471-510. Roberts. R. G.. and Reardon. E. J.. 1978. Alteration and ore-forming processes at Mattagami Lake Mine. Quebec: Canadian Journal of Earth Sciences, v. 15. p. 1-21.

TIMING OF BRINE MIGRATION IN THE LAWN HILL PLATFORM: EVIDENCE FROM THE LAWN HILL MINERAL FIELD Hugh Bresser and Russell Myers Key Centre in Economic Geology Geology Department James Cook University The Lawn Hill Mineral Field is located some 240 kilometres north west of Mt. Isa. It consists of around 40 small vein deposits of lead» zinc and silver covering an area of around 150km^. The field was discovered in 1887 and explored periodically since that time. The largest single deposit became the Silver King Mine and produced around 4500 tonnes of lead and 2000 grams of silver between 1897 and 1967. A study of the detailed paragenesis at two deposits, Silver King and Watson's Lode, and a more cursory examination of several other deposits has shown that the mineralising sequence is remarkably similar, suggesting that all experienced roughly the same history of structural movement and fluid migration. In Silver King and Watson's Lode there are three m^or episodes of sphalerite deposition. Fluid inclusion data collected from sphalerite m Silver King and Watson's Lode show that the sphalerite was deposited from brines whose salinity was approximately 17wt% NaCl equivalent. Eutectic meltuig temperatures hidicate a Ca/Na ratio of approximately one ui these fluids. Homogenisation temperatures range from 110-140°C. Sulphur isotope data from the three generations of sphalerite at Watson's Lode give values ranging from +24 to +29, evolving toward heavier compositions with time. Preliminary results from Silver King suggest that the sulphur isotopes will have similar values and may show a similar trend. In summary, the fluids responsible for the mineralisation in the Lawn Hill Mineral Field were similar to diagenetic basmal brines typically associated with Mississippi Valley Type deposits. Structurally, the m^ority Lawn Hill vem deposits occur in a domain of north-south trending folds. In fact, most occur within the same doubly-plunging anticlinal fold. In this fold the mineralisation occurs in north-east trending ri^t-lateral wrench faults on the limbs of the fold. These north east trending faults have strike lengths of less than five kilometres and are found ahnost exclusively ui the area of north-south foldmg. The orientation and sense of motion on these structures uidicates that they have formed ni response to the same compressional forces as the larger host fold. In the cases of Watson's Lode and SUver King, mineralisation is localised by the development of dilational jogs in these structures. The vein mineralisation found in the Lawn Hill Mineral Field may not be economically significant in it's ownright,however, these venis show conclusively that hot mineraUsuig brines were present and moving during the deformation of the Lawn Hill Platform.

10

MICROSTRUCTURES OF ORE MINERALS FROM THE CENTURY DEPOSIT, NORTHERN QUEENSLAND Graeme Broadbent' and Stafford McKnight^

t CRA Exploration Pty.Ballarat Ltd University College. 2 School of Mining, Geology and Materials, The Century zinc-lead deposit was discovered in April, 1990, by CRA Exploration Pty Ltd. It lies approximately 250 km NNW of Mt Isa in NW Queensland and is hosted by sikstones and shales of the mid-Proterozoic age Lawn Hill Formation. Two major mineralization styles are present in the deposit. The first is hosted by thinly bedded to laminated carbonate rich siltstones. These display various degrees of development of stylolitic solution fabrics and post carbonate replacements by sphalerite and galena, either as disseminations or infilling microfi-actures. This style of mineralization is relatively low grade and does not constitute a large proportion of the deposit. The second, more important style of mineralization is hosted by finely laminated black shales. Sphalerite occurs asfinelaminae, irregular masses and veinlets, with the vast majority occurring as bedding parallel laminae. Galena occurs as disseminated euhedral grains to 200 micron size, asfinegrained laminae sub-parallel to bedding, as irregular replacive masses and in veinlets. Volumetrically the bulk of galena is in transgressive styles. The sequence which hosts the mineralization is gently folded but evidence for a penetrative fabric is difficult to obtain due to the low metamorphic grade of the deposit. There appears to be a relationship between early compactional and later brittle disruption fabrics. This has complicated the interpretation of evidence for the timing of mineralization. Optical examination of etched polished thin-sections and transmission electron microscopy (TEM) of Ar+ ion beam thinned foils of Century sphalerites indicate that the primary crystallization textures of this mineral have been preserved in much of the ore body. Submicron grainsized polycrystalline sphalerite shows a variable but commonly intimate association with bitumen, which exhibits a maturity ranging fi-om epi-impsonite to mesoimpsonite. TEM and optical observation of the bitumen shows that the transition to mesoimpsonite within the one sample has been influenced in places by the local micro-environment. Galena displays a range of features fi-om obviously unmodified replacement textures through to a fully dynamically recrystallized fabric. The dominant processes during deformation have been the early development of subgrains and consequent destruction of primary growth features, followed by progressive subgrain size reduction and rotation resulting in recrystallization.

11

TRANSITIONAL MAGMATIC TO EPITHERMAL MINERALISATION AT THE LIHIR ISLAND GOLD DEPOSIT, PNG: IMPLICATIONS FOR THE EVOLUTION OF ORE-FORMING FLUIDS FROM FERTILE MAGMAS Graham D. Carman VIEPS, Department of Earth Sciences, Monash University, Victoria The Ladolam gold deposit, comprising two separate orebodies (Minifie and Lienetz), is located on the island of Lihir, New Ireland Province, northeastern Papua New Guinea. Ladolam is one of the largest "epithermal" gold deposits in the world, and one of the world's largest undeveloped gold resources. Lihir island is part of the Pliocene to recent Tabar - to - Feni island chain, and is comprised of potassium - rich silica undersaturated volcanics. These rocks have been incorporated within an ankaramite - alkali basalt - trachyte suite (Wallace etai, 1983). The Ladolam deposit is situated within a 'caldera', an elliptical - shaped structure partly breached by the sea with escarpment dimensions of 5.5 km by 3.5 km. Host rocks are comprised of ubiquitous discordant breccias which display complex textural relationships and intense metasomatic alteration. Based largely on these criteria, the caldera is interpreted to represent the near -surface expression of a magmatichydrothermal breccia pipe. Such a structure is believed to have formed in response to release of volatile - rich fluids from the top of a magma chamber (cf. Sillitoe, 1985). Mineralisation at the Minifie deposit is subdivided into three broad stages. Stage I was associated with fluids of "magmatic" origin, and developed an alteration 'core' of biotite - anhydrite orthoclase - pyrite ± sericite ± albite ± carbonate ± chalcopyrite ± molybdenite spatially associated with monzonite breccias. Many fluid inclusions in anhydrite veins contain halite and other salts, and are similar to inclusions reported from potassic zones in porphyry Cu deposits (cf. Nash, 1976). "Transitional" stage II 'carbonate microbreccia' is recognised outward from the inner core, and is associated

with pervasive carbonate - chlorite pyrite - Kspar - albite - sericite alteration in 'outer' heterolithic breccias. Inclusion fluids in carbonate mainly homogenise between 230° and 280°C, contain visible CO2, and are variably saline (from 2 to =30 wt. % NaCI equiv.). Coexisting liquid and vapour-rich inclusions in some samples are indicative of boiling conditions. Stage II at Minifie is currently correlated with the widespread brecciation and explosive activity that accompanied caldera formation. Stage III "epithermal" gold mineralisation at Minifie is related to intense adularia - pyrite alteration in hydrothermal breccias. Refractory gold is contained in disseminated arsenical pyrite and marcasite within breccia matrix and as replacements of primaiy Fe - bearing minerals. Preliminary fluid inclusion data obtained from adularia overgrowths indicate homogenisation temperatures between 180® and 230°C and salinities below 5 wt.% NaCI equiv. Late stage vein breccias with quartz carbonate - adularia - sulphide anhydrite matrix occur below the ore breccia, overprinting biotite - altered intrusive breccias. These display typical low - sulphidation epithermal textures (cf. White and Hedenquist, 1990), and locally carry ore grade. The late siliceous vein - breccias were deposited by relatively cool (<200°C) and dilute (1 to 5 wt.% NaCI equiv.) fluids which locally boiled. Sulphides comprise abundant pyrite with minor marcasite, arsenopyrite, tetrahedrite - tennantite, chalcopyrite, galena and sphalerite. Silicification and mixed-layer clays are dominant wallrock alteration effects surrounding the silicic breccias. Advanced argillic alteration is a nearsurface oxidation phenomenon and is most intense around modern day hotspring discharges. 12

for the intense hydrothermal brecciation and alteration in the ore breccias at Minifie may be a direct consequence of late magma emplacement. It is clear, however, that the multitude of overprinting relationships at Minifie are indicative of a complex interaction between magmatic, phreato-magmatic and hydrothermal events.

The large mushroom - shaped ore zone at Minifie is comprised of crosscutting clast - rich breccias, lying immediately above potassically - altered rocks. These breccias are extremely heterolithic, and their clast - rich nature suggests elutriation of fines. The latter feature is atypical of hydrothermal eruption breccias described in other epithermal systems (eg. Hedenquist and Henley, 1985). Heterolithic ore breccia at Minifie is interpreted to represent near - surface infill- of hydrothermal eruption vents. The occurrence of pyroclastic rocks as fragments within high grade ore breccia indicate that mineralisation and brecciation post - date phreatomagmatic maar - type volcanism. Limited K-Ar dating indicates that early 'porphyry' mineralisation formed less than 1 million years ago, while the main gold mineralization event occurred around 0.3 M.A. (Moyle et al., 1990). A large explosion, or a series of explosions associated with the emplacement of high - level Pleistocene trachytes into the upper portion of the Ladolam breccia pipe, in conjunction with unroofing of the system, seems the most likely mechanism to explain the superposition of low - temperature epithermal mineralisation directly above biotite - altered intrusive breccias. Early magmatic - hydrothermal mineralisation at Ladolam was undoubtedly an important precursor to eventual ore deposition. Upgrading of large volumes of gold - enriched magmatic hydrothermal mineralisation has already been identified by Sillitoe(1989) as potentially fundamental in the generation of giant -sized gold deposits. Two other megadeposits in the western Pacific, Porgera (Richards, 1992) and Baguio (Cooke et al., 1992), exhibit close spatial and possibly genetic associations with early magmatic hydrothermal events. The role of late - stage trachytes emplaced into the upper portions of the pipe with respect to gold mineralisation is at this stage unclear. It seems likely that high thermal gradients responsible

References Cooke, D. R.. Bloom, M. S. and Cartwright, I., 1992. The Acupan South porphyry coppergold prospect, Baguio District, Philippines: relationship to epKhermal mineralisation [abs]: National meeting, SGEG, Geol. Soc. Aus. (this volume). Hedenquist. J. W. and Henley. R. W., 1985. Hydrothermal eruptions in the Waiotapu geothermal system, New Zealand: Their origin, associated breccias and relation to precious metal mineralization: Econ. Geol., v. 80, p 1640-1668. Moyle, A. J.. Doyle. B. J.. Hoogvliet. H. and Ware. A. R. 1990. Ladolam gold deposit, Lihir Island: in Hughes. F. E.. Geology of the mineral deposits of Australia and Papua New Guinea, The Australasian Institute of Mining and Metallurgy, Melbourne. 1793-1805. Nash. T. J. 1976. Fluid inclusion petrology - Data from porphyry copper deposits and applications to exploration: U.S Geol. Surv. Prof. Paper 907-D. 16p. Richards, J. P. 1992. Magmatic-epithermal transitions in alkalic systems: Porgera gold deposit, Papua New Guinea: Geology, v. 20, p. 547-550. Sillitoe. R. H. 1985. Ore-related breccias in volcanoplutonic arcs: Econ. Geol.. v. 80. p. 1467-1514. Sillitoe, R. H. 1989. Gold deposits in western Pacific island arcs: The magmatic connection: Econ. Geol. Monograph 6, p. 274-291. White, N. C. and Hedenquist, J. W. 1990. Epithermal environments and styles of mineralization: variations and their causes, and guidelines for exploratton in Hedenquist. J. W.. White. N.C.. and Siddeley, eds., Epithennal gold mineralization of the CircumPacific: Geology, geochemistry, origin and exploration I: J. Geochem. Expl., v. 36, p. 445-474.

13

PRECISE ISOTOPIC FINGERPRINTING OF FERTILE HYDROTHERMAL EVENTS EXPLORATION APPLICATIONS AND EXAMPLES FROM THE LACHLAN AND THOMSON FOLD BELT SYSTEMS Graham R. Carr\ Judith A. Dean\ Gregg W. Morrison^ and David Suppel^ ^CSIRO Division of Exploration Geoscience. North Ryde, NSW ^Klondike Exploration. Townsville, OLD ^New South Wales Department of Minerals and Energy, NSW A key factor in successful exploration for base metal and gold mineralization is the ability to made confident decisions regarding the potential of a given prospect area at all phases of project evaluation, from selection of ground to definition of drilling targets. The use of isotopic techniques, particularly Pb isotopes, has become increasingly important over the past decade in defining the metallogenic associations of exploration samples and thus their likelihood of representing economic mineralization. However, in many regions the usefulness of this information is limited by an incomplete understanding of the nature and age of hydrothermal activity, the styles of mineralization present and their isotopic compositions, or "fingerprints". The aim of a current research project within the Division of Exploration Geoscience is to define isotopic, geochemical and geophysical signatures of each hydrothermal event, and the various stages of those events, that are likely to be represented in the major metallogenic provinces of Australia. Comparisons of the early Pb isotope database of the Lachlan and Thomson Fold Belt Systems (LFB-TFB) with the current "intelligence base" shows how more precise definitions can be made of exploration potential when the fingerprints are integrated with a more complete understanding of the age, tectonic context and metallogenic association of known mineralization. In the early database, a large number of Ordovician-to-Devonian deposits including Balcooma, Thalanga, Captains Flat, Woodlawn, Benambra, Cobar and Elura defined a single massive sulfide signature for the LFB-TFB which was surprisingly similar over both provinces despite the broad geographic distribution and the assumed = 100 Ma age range. Isotopic variations that were apparent could not be satisfactorily explained and there was only a very limited understanding of the Pb isotopic fingerprints of epigenetic mineralization associated with essentially coeval granite intrusion. Since then, collaborative studies with James Cook University, the University of Wollongong, the NSW Department of Minerals and Energy and mining companies CRAE, NBH-Peko, BHP, Newcrest and Homestake have provided a framework for the definition of a large number of Pb isotopic signatures, each of which can be associated with a mineralizing event within a specific province. The important signatures have been defined below. 1. 2. 3. 4. 5.

Cambro-Ordovician VHMS signature - Seventy Mile Range Group, and Mt Windsor Sub-Province Ordovician cratonic signature (Girilambone signature) Ordovician low-|Li signature (Molong Arc - shoshonite signature) Ordovician-to-Devonian granite vein signature - Lolworth-Ravenswood Province Silurian-to-Devonian high Th/U granite vein signature - Georgetown-Coen Province

14

6. 7. 8. 9. 10. 11. 12.

Silurian-to-Devonian crustal Th/U granite vein granite signature - Georgetown-Coen Province Silurian VHMS signature - LFB NSW and Victoria Devonian deep vein granite signature - LFB NSW Devonian sediment-hosted massive sulfide signature - Cobar Trough Permo-Carboniferous breccia pipe/epithermal signature - Lolworth-Ravenswood Province Permo-Carboniferous breccia pipe/epithermal signature - Georgetown-Coen Province Permo-Carboniferous intrusion'related signature - LFB NSW

These signatures, although providing more precise discriminators of the nature of hydrothermal activity in each of the provinces, are, as yet, not comprehensive. For example, although individual deposits in each of signatures 1, 3, 7 and 8 have very homogeneous isotopic compositions there are small, though significant isotopic differences between deposits. The variation between deposits does not appear to define regional trends and in some instances two or more discrete data populations can be defined in a single deposit. Thus it seems unlikely that variability is in response to regional source rock heterogeneities in Pb isotopic composition but rather is caused by the interaction of more than one hydrothermal system, or a single hydrothermal system whose isotopic composition changed over its period of activity. If the latter is the case, then it should be possible to assign signatures to the stages of the development of hydrothermal activity in a province and define precisely that stage most likely to have deposited major amounts of sulfides. Current research to define these signatures is based on detailed paragenetic studies, a well defined age framework and includes multi-isotopic techniques (Pb, Sr, S).

15

THE ACUPAN SOUTH PORPHYRY COPPER-GOLD PROSPECT, BAGUlO DISTRICT, PHILIPPINES: RELATIONSHIP TO EPITHERMAL MINERALISATION. D.R. Cooke^*. M.S. Bloom^ and I. Cartwright^ ^ CODES Key Centre, University of Tasmania 2Department of Earth Sciences, Monash University, Victoria Porphyry copper-style mineralisation occurs on the southern margins of a large adulariasericite vein system at the Acupan gold mine (Cooke and Bloom, 1990). Porphyry-style mineralisation is intimately associated with the Pliocene Ampucao dacite porphyry (ADR), a small subvolcanic stock that has intruded Miocene andesitic lavas of the Zig-Zag Formation. Porphyry-related mineralisation and K-silicate alteration is also recognised in the Pliocene Virac granodiorite (the principal host rock for epithermal vein mineralisation). High temperature (mode: 500X; range: 350^ to >600''C) hypersaline fluids (>34 eq. wt. % NaCI) and low salinity vapour were released during initial pressure quenching of the crystallising ADP. Inner K-silicate (orthoclase and biotite) and outer propylitic alteration assemblages formed around stage I quartz magnetite - anhydrite ± orthoclase ± biotite ± sericite ± pyrite ± chalcopyrite veins. Stage Ha quartz - anhydrite - chlorite - chalcopyrite - rutile ± pyrite veins formed in association with sericiteclay-chlorite (SCC) alteration, as cooling high temperature (mode: 450X; range: 300^->600X) hypersaline fluids (>31 eq. wt. % NaCI) and low salinity vapour migrated away from the ADP. Oxygen isotopic analyses of quartz from stage I and lla veins indicates a magmatic-hydrothermal origin for both vein stages. values calculated for waters from stages I and lla are the highest recorded from the Baguio district (5.5 to 6.7 %o), and are clearly distinct from younger meteoricdominated epithermal veins (-6.2 to -0.8 %o) and modem thermal waters (-11.5 to -2.2 %o\ Fig. 1). The fluid inclusion and stable isotopic evidence for deposition of stage lla veins from high temperature magmatic-hydrothermal fluids casts doubt on SCC alteration representing an equilibrium mineral assemblage. The volume of clays in SCC alteration halos from Acupan South was invariably too small for identification via XRD analyses. However, Sillitoe and Gappe (1984) noted that kaolinite and/or illite are the dominant clay fractions in SCC alteration zones from other Philippine porphyry systems. These clay minerals are unlikely to form at temperatures much higher than 300'^C (eg. Henley and Ellis, 1983), so it is likely that clays in the SCC alteration assemblage at Acupan represent a

retrograde overprint of higher temperature sericite-chlorite alteration. Later episodes of porphyry-related vein mineralisation are associated with sericitic alteration. These veinlets were deposited from low salinity, moderate temperature fluids (1 - 2 eq. wt. % NaCI; 260^ - 345X). Stage lib veinlets are characterised by anhydrite ± pyrite ± quartz ± sericite ± calcite ± chlorite, whereas stage lie veins are dominated by calcite ± chalcopyrite ± pyrite ± anhydrite ± serfcite ± quartz ± sphalerite ± galena. Stages lib and lie are believed to represent ingress of meteoric fluids into the magmatic-hydrothermal domain during waning porphyry-related magmatic activity. By the criteria of Sillitoe (1989), two adulariasericite vein systems in the Baguio district (Acupan and Antamok) are classifiable as 'major* gold deposits («200 tonnes Au per deposit), and the Baguio district as a whole can be classified a •mega* gold district (< 400 tonnes Au), with the majority of gold production coming from epithermal veins. The only other Western Pacific •epithermal' systems that meet the criteria of mega gold deposits/district are the Lienetz and Minifi deposits at Ladolam, Lihir Island and the Porgera deposit, PNG (Sillitoe, 1989). As for Acupan, epithermal mineralisation at both Ladolam and Porgera overprints older porphyry copper-style mineralisation (Carman, 1993; Richards, 1992). Other smaller epithermal deposits of the Western Pacific do not exhibit this spatial association between adularia-sericite veins and porphyry copper-gold mineralisation. There is no unambiguous fluid inclusion or stable isotopic evidence for a magmatichydrothermal contribution to the epithermal fluids at Acupan, even though epithermal mineralisation is believed to have formed no more than 1 to 2 Ma after the porphyry event (Cooke and Bloom, 1990; Cooke 1991). Although a direct genetic relationship between porphyry copper and epithermal mineralisation at Acupan cannot be unequivocally demonstrated, there is little doubt that the ADP remained an excellent potential source of gold for the epithermal system, because any Au-undersaturated fluids that passed through the ADP's alteration zones during epithermalrelated meteoric convection had the potential to leach gold from the porphyry system.

16

600

500-

400"

Magmatic water Porphyry-related vein quartz

Epithermal vein quartz

300-

\

•

Stage I (A)

A Stage lla (A) O Epithermal (A) A Epithermal (B)

200"

Modern thermal waters

•

Thermal waters (B)

•

Thermal waters (C)

100

-15

-10

-5

6^0

0

10

Figure 1: vs. temperature plot for vein quartz and modern thermal waters of the Baguio district, Philippines. Temperatures for quartz samples are the average primary fluid inclusion homogenisation temperatures for that sample. The temperature of the modem thermal waters were measured directly at the discharge site. The lower temperature margin of the magmatic water 'box' is uncertain; it has therefore been drawn as a dashed line on this diagram. Sources of data: (A) - this study; (B) - Sawkins et al., 1979; (C) - JICA, 1984.

Sillitoe (1989) speculated that '^upgrading of magmatic-hydrothermal preconcentrations may be a fundamental factor in the generation of giant epithermal gold deposits**. We observe that (1) there is an anonymously large quantity of gold dispersed through the adularia-sericite vein system at Acupan in comparison to many other epithermal vein systems; and (2) there is older copper-gold mineralisation in the Acupan South porphyry system which is proximal to the epithermal vein system. These observations support (but do not prove) Sillitoe's speculation.

Cooke, D.R., 1991. Styles and controls of mineralisatbn, Acupan gold mine, Baguio district, Philippines. PhD Thesis (unpubl.), Monash University, Melbourne, Vic: 396 p. Henley, R.W. and Ellis, A.J., 1983. Geothermal systems ancient and modern: a geochemical review. Earth Science Review., 19:1-50. JICA, 1984. Report on Acupan-ltogon Geothermal Development. Second Phase Survey. Japan International Cooperation Agency, 123 p. (unpubl.). Rfchards, J.P., 1992. Magmatic-epithermal transitfons in alkalic systems: Porgera gold deposit, Papua New Guinea. Geok>gy, 20:547-550. Sawkins, F.J., O'Neill, J.R. and Thompson, J.M., 1979. Fluid inclusk)n and geochemical studies of vein gold deposits, Baguk) District, Philippines. Econ. Geol., 74:1420-1434. Sillitoe, R.H., 1989. Gold deposits in western Pacific island arcs: The magmatic connection. Econ. Geol. Mon., 6:266-283. Sillitoe, R.H. and Gappe, I.M., 1984. Philippine porphyry copper deposits: geologic setting and characteristics. CCOP Tech. Pub. 14,89 p.

Rel^rencBs Carman, Q.D., 1993. Transitional magmatic to epithermal mineralisation at the Lihir Island gold deposit, PNG: Implications for the evolution of oreforming fluids from fertile magmas, [abs]: National meeting, SGEG, Geol. Soc. Aus. (this volume). Cooke, D. R., and Bloom, M. S., 1990. Epithermal and subjacent porphyry style mineralization, Acupan, Baguio District, Philippines: a fluid inclusion and paragenetic study. J. Geochem. Expl., 35:297-340.

17

OAK DAM EAST: A PRODIGIOUS, URANIUM-BEARING, MASSIVE IRON-OXIDE BODY ON THE STUART SHELF Garry J. D a v i d s o n & Hamish L Paterson^ I.Centre for Ore Deposit & Exploration Studies, University of Tasmania. Hobart, Tasmania, 7005 2. Western Mining Corporation Ltd., P.O. Box 114, Daw Park, South Australia, 5041 The Oak Dam East magnetic anomaly and corresponding 3 - 4 mgal gravity anomaly occur on the Stuart Shelf, 39 km west of Lake Torrens in South Australia. Thirteen drillholes have identified a very large iron-oxide body ( over 500 million tonnes) at or immediately beneath the unconformity between flat-lying Adelaidean cover — consisting of 400 to 500 m of Arcoona Quartzite, Tregolana Shale, and Pandurra Formation — and a deformed/metamorphosed granitic basement. The iron oxide mass formed a large monadnock on the Proterozoic land-surface prior to Pandurra Formation deposition. Basement is also transected by 0.5 - 5 m wide dykes of undeformed coarse to pegmatitic alkali granite (altered in places but generally fresh), and medium-grained diorite (very altered). The body occurs along a regional north-south discontinuity termed the Pernatty Culmination, a feature best defined by a rapid westward increase in depthto-basement, thought to represent a major regional fault. Oak Dam East consists mainly of massive hematite and interstitial quartz, with typical Fe values of 41 to 56% (58.6 to 80.0 wt. % Fe203)- Fragments of gneissic granite and minor schist occur throughout, displaying variable degrees of replacement from the margins inward. Some fragments have been totally assimilated but have preserved the granofelsic texture of the precursor; in others, ribboned or equant quartz are the only preserved relicts. Many fragments have not been severely hematitised, but are generally altered to the sheet silicate which dominates the phyllic alteration in the local matrix. Monazite concentrates around the margins of such clasts. Early hematite consisted of delicately laminated colloform bands with no consistent orientation; a massive "steely" hematite pervasively replaces disseminated magnetite and most of the early colloform material. Chalcopyrite and pitchblende are concentrated in a layer sub-parallel to the oxidesulphide transition, approximately 50-100 m into the body. Sulphide above this was leached during pre-Pandurra Formation Proterozoic weathering, whereas the metalenriched layer is attributed to primary hydrothermal activity, not supergene processes. Chalcopyrite in the enriched layer commonly replaces disseminated pyrite along fractures. Pyrite and chalcopyrite are isotopically unequilibrated, with S^^s for chalcopyrite (-14 to -7 %o) being 4 to 6%o lighter than replaced pyrite (-6 to 0%o). One explanation for this isotopic change during apparent pseudomorphous replacement is that exchange occurred with a fluid relatively enriched in sulphate. Petrographic studies indicate an early history of mesothermal to epithermal brecciation of the basement with subsequent growth of high temperature hydrothermal mineral phases (still poorly understood), including anhydrite. Fluid inclusion studies indicate this part of the history involved phase separation and highly saline fluids, which were probably granite-derived. Magnetite, pyrite, apatite and quartz were deposited subsequently, followed by botryoidal hematite-quartz-chalcedony ± monazite. During the latter period, a near-horizontal boundary between phyllic (chlorite-dominated) and

18

argillic (sericite-illite-hydromuscovite) alteration was formed, attributed to boiling in the sub-surface. High-grade U-Cu mineralisation occurs locally at this boundary. A model is proposed for the origin of Oak Dam East, in which a convective fluid cell formed in a basement breccia-zone above a crystallising pluton, pluming hot, saline, magmatic/meteoric water along the Pernatty Culmination. Magnetite, pyrite and quartz were deposited as the dominant mineralogy near the base, with hematite-quartz towards the top of the system, growing in botryoidal masses around granite-breccia fragments from more oxidised meteoric water, Hydro-magmatic brecciation continued episodically during deposition, but was not sufficiently active to move volcanics or tuff fragments down into the sub-surface breccia zone, as has been the case at Olympic Dam, located 63 km to the northwest. The most active brecciation occurred along narrow fracture zones during rapid "gas-streaming" events, the results of phase separation at depth. Boiling of fluid to nearly the top of the existing body produced widespread chalcedony, and K-sllicates above 630 m. Late in the development of the iron-oxide body, an upwelling, boiling, Fe-Cu-U-bearing fluid mixed along a horizontal boundary with oxidised, low-temperature meteoric water, depositing oxidised metal assemblages and sheet silicates. Fluctuations in the level of the groundwater base can account for the width and concentration of various metals. As the upwelling plume collapsed, cooler oxidised meteoric waters percolated or were drawn lower into the oxide body, converting magnetite to hematite. This model is a direct analogue of processes which form reduced, disseminated gold deposits in the epithermal zones of collisional plate settings.

19

CHARACTER OF ADULARIA IN EPITHERMAL VEINS, QUEENSLAND. Guo-yi Dong and Gregg Morrison Geology Department, James Cook University, Townsville, Queensland. The examination of approximately 80 samples from 12 occurrences and operating mines in Queensland has led to the recognition of four types of adularia, based on the form and size of the individual crystals. Very often, aggregates of different types of adularia give distinctive morphologies which are readily identified at hand specimen scale. I.Crystalllne rhombic: Adularia crystals in this group are usually anhedral or subhedral with rhombic terminations, commonly 2 - 4 mm in size. Typically, adularia crystals are dispersed through a matrix of finer grained quartz, giving an overall spotty appearance in hand specimen, II. MIcrocrystalllne rhombic: Individual crystals in this type can only be identified under the microscope due to their extremely small crystal size (<0.2mm) and generally have a perfect rhombic form. Aggregates of adularia crystals of this type, associated with chalcedony or microcrystalline quartz, usually have a moss - like appearance in hand specimen. III. Tabular: The individual crystals are generally fine - grained (0.5 - 2mm), and lath - shaped. Adularia of this type usually occurs as groups of parallel or subparallel crystals oriented perpendicular to the vein walls In crustiform banded veins. IV. Pseudo-aclcular: This texture shows a needle - like appearance of adularia in hand specimen. Under the microscope, each needle actually consists of several small (0.1 - 0.5mm) elongated or ragged adularia crystals with a poorly defined crystal habit. It has been suggested that this texture is formed after replacement of carbonate. Further investigation of the structural state of adularia, using X-ray diffraction techniques, has revealed that two groups can be identified in terms of AI/SI disordering. The first group corresponds to the crystalline rhombic adularia type, which has relatively low Al/Si disordering with 2ti of most samples above 0.84. As an extreme case, two samples from an epithermal vein showed their crystal structure close to that of maximum microcline with tio of 0.94 and 0.95, respectively. The second group includes microcrystalline rhombic and tabular adularia types, which commonly show medium - high Al/Si disordering with 2ti values ranging from 0.64 to 0.78. There is no fresh samples of pseudo-acicular adularia so far for XRD analyses because of extensive sericite or clay alteration.

20

A review of available literature indicates that the structural state of most unaltered adularia in epithermal veins is mainly controlled by the process of primary crystal growth, and extensive interaction with late fluids is responsible for advanced ordering in the structure of altered adularia. Considering the process of crystal growth, high nucleation and crystallisation rates at low temperatures are essential for the formation of metastable disordered adularia. which is grown and preserved in the stability field of maximum microcline. In epithermal environments, boiling, characterised by loss of volatiies, increase in pH and cooling, Is probably the most likely mechanism which favours formation of metastable disordered adularia. In general, crystalline rhombic adularia may form in the bottom zone of a boiling system where boiling initiates at a relatively high temperature and the fluid is slightly supersaturated with respect to adularia and quartz. Relatively slow changing conditions and deep position in vein systems are responsible for relatively low Al/Si disordered phases and general lack of sericitic alteration in crystalline rhombic adularia's crystals. In contrast, microcrystalline rhombic and tabular adularia crystals are characterised by a higher degree of Al/Si disordering in their structure, implying rapid nucleation and crystallisation conditions which are most likely induced by violent boiling of hydrothermal fluids. Acicular adularia pseudomorphs, presumably after carbonate, are suggested to form at more or less the same conditions as microcrystalline rhombic adularia. As boiling is protracted further, the effect of drop in temperatures becomes dominant over loss of CO2. Carbonate, precipitated due to rapid loss of CO2 during early boiling, will be later dissolved and replaced by adularia and quartz. Assaying of character samples suggests that microcrystalline rhombic and pseudo-acicular adularia types are good indicators of high gold grades. This is consistent with the conventional hypothesis that gold usually precipitates when boiling is protracted and effect of H2S loss becomes dominant over pH change.

21

THE ROLE OF LATE PROTEROZOIC FRACTIONATED GRANITOIDS IN THE GENESIS OF POLYMETALLIC MINERALISATION IN THE TELFER DISTRICT, WA N.M. Goellnicht

D.I. Groves ^ NJ. McNaughton ^

1 Placer Exploration, GPO Box 558, Brisbane, Q.4001, Australia 2 Key Centre for Strategic Mineral Deposits, University of Western Australia, Nedlands, W.A. 6009, Australia Polymetallic mineralisation in the Telfer district is hosted by Late Proterozoic, low-grade, metasedimentary sequences of the Yeneena Group (-1000-750 Ma) in the Paterson Province. Epigenetic Au-Cu and base metal mineralisation formed during the late stages of the Paterson Orogeny (700-600 Ma), broadly contemporaneous with the intrusion, at about 680 to 620 Ma, of late- to post-tectonic, highly fractionated, I-type granitoids into the Yeneena Group (Goellnicht et al., 1989, submitted). Two lines of evidence suggest the role of granitoids as ore fluid and heat sources in the formation of spatially associated mineralisation: (i) the field, petrologic, geochemical and source characteristics of the granitoids; and (ii) Pb isotope studies on the ore sulphides and potential metal sources. The granitoids range from biotite monzogranites, through syenogranites, to alkali-feldspar granites (Goellnicht et al., 1991). They were emplaced at shallow crustal levels (4-6km), and their contact metamorphic aureoles are between 1 and 2 km wide. The granitoids can be subdivided into three suites, each differing in their initial Pb isotopic ratio. The Mt Crofton Suite comprises apparently undeformed, more fractionated (> 71 wt.% Si02, Rb/Sr ranges up to 20.5), magnetite-series granitoids, whereas the Minyari Suite and the 0*Callaghans Suite comprise variably deformed, less fractionated, ilmenite-series granitoids, which contain a higher proportion of biotite compared to the Mt Crofton Suite (Goellnicht et al., 1991). The granitoids are highly fractionated, metaluminous and calc-alkalic. They have high Si02, K2O, Rb and LREE, relatively high NajO, U and Th, and low MgO, CaO, Ti02 and P2O5. Trends on both major- and trace-element Harker variation diagrams and REE plots are strongly controlled by fractional crystallisation of accessory minerals, K-feldspar and plagioclase (Goellnicht et al., 1991). Most of the granitoids, in particular the Mt Crofton Granite Complex, have unusually high radiothermal-heat production (McNaughton and Goellnicht, 1990). On a global scale, these features are common to granitoids associated with a variety of mineralisation styles. On a common Pb-Pb diagram, Pb isotopic data for sulphides in various styles of deposits in the Telfer district form a linear array between the initial ratio of the Mt Crofton Suite (and possibly the 0*Callaghans Granite) and the field for host sedimentary sequences. Proximal deposits preserve more of a magmatic-Pb component, whereas distal deposits contain dominantly host-rock derived Pb. This implies the Pb in the regional deposits is a mixture of magmatic- and host rock-derived Pb, in a proportion broadly related to the distance of the deposits from the source granitoid (Goellnicht et al, submitted). This evidence of derivation of Pb from the granitoids, combined with indications of penecontemporaneous timing of the granitoids and mineralisation, and evidence for 22

relatively high-temperature, complex, highly saline ore-fluids in a relatively lowmetamorphic grade setting, suggests at least partial ore-fluid derivation from the granitoids (Goellnicht 1992; Goellnicht et al., submitted). The highly fractionated, I-type granitoids of the Telfer district, in particular the more oxidised, magnetite-series granitoids of the Mt Crofton Suite are, thus, interpreted to play an important role as solute, fluid, and heat sources in a zoned hydrothermal system involving the mixing of magmatic fluids with a host rock component, either through mixing with a basinal fluid or by reaction with the host rocks, ultimately forming Au, Cu, W and/or base-metal mineralisation (Goellnicht 1992; Goellnicht it al., submitted). References: Goellnicht N.M., 1992. Late Proterozoic fractionated granitoids and their role in the genesis of gold and base-metal mineralisation in the Telfer district. Western Australia. Unpubl. Phd Thesis, University of Western Australia. Goellnicht, N.M., Groves, D.I., McNaughton, N.J. and Dimo, G., 1989. An epigenetic origin for the Telfer gold deposit. In Keays, R.R., Ramsay, W.R.H. & Groves, D.I. (eds). The Geology of Gold Deposits: The Perspective in 1988. Econ. Geol. Monograph 6: 151-167. Goellnicht, N.M., Groves, D.I. and McNaughton, N.J., 1991. Late Proterozoic fractionated granitoids of the mineralised Telfer area, Paterson Province, Western Australia. Precamb. Res. 51: 375-391. Goellnicht, N.M., Groves, D.I. and McNaughton, N.J., submitted. Lead isotope evidence for the role of Late Proterozoic fractionated granitoids in the genesis of polymetallic mineralisation in the Telfer district. Western Australia. Mineralium Deposita. McNaughton, N.J. and Goellnicht N.M., 1990 The age and radiothermal properties of the Mount Crofton Granite, Telfer area. Western Australia. Aust. J.Earth Sci. 37: 103-106.

23

RARE METAL MINERALISATION DURING MAGMATIC HYDROTHERMAL TRANSITION IN THE GREENBUSHES PEGMATITE, WESTERN AUSTRALU Thein Hani and Peter L.F. Collins School of Applied Geology. Cwrtin University. Kent Street. Bentley. WA 6102 ^ Present address: Department of Geological Survey and Mineral Exploration. Ministry of Mines. Myanmar

The Greenbushes pegmatite district is one of the world's largest rare-metal pegmatite deposits. The Sn-Ta-Nb-Li-bearing Archaean pegmatite was emplaced within the crustalscale Donnybrook-Bridgetown shear zone in middle to upper amphibolite facies metamorphic rocks of the Balingup Metamorphic Belt in the southwest Yilgam Craton, Westem Australia (Bettenay et al, 1986; Partington, 1990). Deformation fabrics in the pegmatite are the same as in its host rocks indicating syntectonic emplacement (Partington, 1990), though some smaller pegmatite bodies marginal to the main pegmatite are apparentiy only mildly deformed The Greenbushes pegmatites have no obvious parental granitoid, and the main pegmatite (ca. 2530 Ma) clearly is younger than nearby syntectonic granitoids (ca. 2580-2610 Ma) (Partington, 1990). The main Greenbushes pegmatite has been subdivided into four major primary mineralogical zones (contact, lithium, K-feldspar and albite zones), with a number of smaller subsidiaty compositional zones (eg, tourmaline-rich zones, quartz-feldspar zones). The macroscopic zonation is not concentric and is the reverse sequence to that normally found in zoned pegmatites (Bettenay et al, 1986), though paragenetic relationships between mineralogical zones in the Greenbushes pegmatite have not been fully established. The contact zone is a fine-grained aplitic unit in which cassiterite and tantalite are preferentially enriched at contacts with mafic-ultramafic wall rocks. The lithium zone is dominated by Fe-poor spodumene, quartz and feldspar. Some of the spodumene in this zone exhibits a symplectic texture of vermicular intergrowth of quartz and spodumene which may be after the high temperature, metastable, silica-saturated Li-Al-silicate phases virgilite or p-spodumene (London, 1984). The K-feldspar zone is dominated by microcline, perthite and quartz. Most cassiterite and tantalite are confined to the main albite zone which consists of albite, quartz and tourmaline, though much of the Sn and Ta-Nb in this zone was introduced during late stage metasomatic alteration that also affected K-feldspar and lithium zones. Much of the history of crystallization of the Greenbushes pegmatite and its evolutionfroma magma to a hydrothermalfluidis preserved influidinclusions in britde minerals, particularly tourmaline which is present in all mineralogical zones and crystallized throughout all stages in the evolution of the pegmatite and does not appear to have been adversely strained during deformation (which may have caused stretching or leaking of inclusions). Fluid inclusions preserved in tourmaline, quartz, beryl, albite, cassiterite and spodumene include single phase liquid or vapour inclusions, two phase C02-(CH4) and H2()-NaCl inclusions, three phase H20-NaCl-C02±(solids), and 'melt'-H20-NaCl±C02 inclusions. Early-formed blue tourmaline contains rare C02-(CH4) inclusions, but is devoid of aqueous inclusions, whereas later brown tourmaline contains H20-NaCl, H20-NaCl-C02 and 'melt*-H20Naa±C02 inclusions. Afinalphase of green tourmaline overgrowth generally is devoid of fluid inclusions. The coexistence of aqueous liquid-dominant and 'melt'-dominant inclusions indicates exsolution of aqueous and carbonic aqueous phases from the m a ^ a and the presence of a heterogeneous fluid at high temperatures and pressures. The primary aqueous fluid is of low-moderate salinity (4-14 mass % NaCl) and is markedly differentfroma later(?) highly saline (25 mass %) CaCl2-rich external fluid, preserved in H20-CaCl2 inclusions in vein 24

quartz in the host rocks. Oxygen isotope data are consistent with a magmatic source for the aqueous phase, but the hydrogen isotopic composition is markedly depleted in deuterium, possibly due to mixing or post-crystallization exchange with an extemal deuterium-depleted water. Crystallization of the Greenbushes Pegmatite, as determined from fluid inclusion petrography and microthermometry, commenced with tourmaline from an homogeneous hydrous melt at temperatures above -850®C. Aqueous and carbonic aqueous phases of lowmoderate salinity were exsolvedfromthe magnia on cooling below 800®-850®C, forming a heterogenousfluidfromwhich much of the remainder of the pegmatite crystallised. Albite, cassiterite and tantalite in albite zones and tourmaline zones, and a ^i-Al-silicate)spodumene-quartz assemblage in lithium zones crystallisedfroma magmatic-hydrothemial fluid at temperatures of the order of 700-850®C. The hanging wall K-feldspar zone crystallized at 680-750®C and apparentiy was coeval or transitional with the hanging wall lithium zone. Although most minerals in the pegmatite had crystallised by thetimethe fluid had cooled to --650®C, a third stage of Sn-Ta-Nb minei^isation accompanied 'low' temperature (550®-650®C) metasomatic(?) alteration of the primaiy crystallization phases. During emplacement and crystallization of all phases of the pegmatite the pressure spears to have been constant at 5-5.5 kbar, which is consistent with the prevailing metamorphic conditions, and indicating a deptii in excess of 12 km. The fluid inclusion mterpreted pressure and temperature conditions also are consistent with the mineralogical assemblages. For example, the spodumene-quartz symplectic intergrowtii (after(?) virgilite or pspodumene) in the lithium zone, combined with the absence of petalite, is consistent with a crystallization pressure >4.1 kbar and temperature >680 (London, 1984); and tiie presence of orthoclase inclusions exsolved in perthite indicates temperatures up to 700 for a pressure of 5 kbar for the K-feldspar zone. The Greenbushes pegmatite was emplaced at higher pressures (and greater depths) and crystallised at higher temperatures than most otiier rare metal pegmatites, with Sn and Ta-Nb oxides and Li-Al-silicates crystallising from tiie earliest stages of exsolution of a hydrothermal phase, with subsequent enrichment of Sn-Ta-Nb during infiltration of the pegmatite by a late-stage aqueous fluid. References Bettenay. L.F. Groves, D.I. & Partington, G.A., 1986. Proc 7th lAGOD Symp., 401-408. London, D. 1984. American Mineralogist. 69:995-1004 Partington, G.A., 1990. Economic.Geology. 85: 437-456.

25

HEMATITE-BARITE ALTERATION IN THE OWEN CONGLOMERATE, NORTH LYELL, TASMANIA I. M. Hart CODES Key Centre, University of Tasmania The Cambrian Mt Read Volcanics are the principal host to Cu-Pb-Zn mineralisation in western Tasmania (Fig. 1), hosting significant Cambrian VHMS style mineralisation at Rosebery, Hellyer and Que River. This belt of mafic to felsic volcanics occurs along the western margin of the Precambrian Tyennan Block, with Cambrian Dundas Group sediments occupying a trough between the volcanics and the Rocky Cape region to the west. In the Mt Lyell area, Late Cambrian to Ordovician siliciclastic and carbonate sediments of the Denison and Gordon Groups overlie the volcanics unconformably. Hydrothermal alteration and Cu mineralisatton extends up into the Ordovician, sediments raising questions about the timing, extent and nature of hydrothermal activity at Mt Lyell. The rocks of the Central Volcanic Complex (CVC) dominate the stratigraphy within the Mt Lyell mine area, and are dominated by acid lavas and pyroclastics of rhyolitic-dacitic composition. The eastern margin of the CVC is faulted against the overlying Denison Group by the Great Lyell Fault (GLF)(Fig. 1). The Ordovician Denison Group is con^sed of the Lower Owen Conglomerate (quartz conglomerates and quartz sandstones), the hematitto Middle Owen Congtomerate and the Upper Owen, 'Pioneer Beds'. The Pioneer Beds consist of medium-coarse grained, chromite rich sandstones and conglomerates that have disconformable and locally unconformable relationships with the underlying units. The Ordovician Gordon Group conformably overlies the Pioneer Beds. The Mt Lyell Cu field consists of 17 pyriticcopper orebodies contained within an intense zone of sericite - quartz - chlorite - pyrite alteration. The North Lyell orebody is one of the richest within the field, having produced 5mt of ore with 5.3% Cu, 33 g/t Ag and 0.4 g/l Au. The main ore zone at North Lyell is tocated near the intersection of the GLF and the WNW trending North Lyell Fault. Mineralisation occurs primarily as coarse grained masses of bornite ± chalcopyrite ± covellite, within bodies of pyritic cherty quartz. Minor tennantite, galena, digenite, mawsonite, molybdenite, sphalerite, linnaeite, enargite, stromeyerite and rutile are associated with mineralisation (Hills, 1990). Although the CVC is the principal host to mineralisatton at North Lyell, bornite also occurs in the adjacent Owen Conglomerate (across the GLF), with some bornite masses showing relict pebble

outlines on fresh faces. Alteration around North Lyell consists of an inner quartz-chloritebarite-pyrophyllite zone which grades outwards to a quartz-chlorite-sericite assemblage. There is a strong host rock control on secondary Fe-bearing minerals in both assemblages, with pyrite occurring in the altered volcanics, and hematite in the conglomerate. Alteration/weathering in the Gordon Limestone is characterised by kaolinite quartz; no cartx)nate is preserved. The timing of alteration/mineralisation at North Lyell is reflected in the extent and style of alteration observed within the overlying siliciclastte Owen Conglomerate and Gordon Group limestones. Intense hematisation of the Owen Conglomerate (whole rock Fe203 up to 85%), has occurred in areas adjacent to the GLF, near the North Lyell and Lyell Tharsis pits, with some units showing relict conglomeratic pebble structures defined by hematite. Increased phosphorous and barium values are also associated with intense hematisation. These values decrease away from the main zone of alteration, with barite virtually disappearing from the alteration package, and hematite becoming less conspicuous. Al and K are found to increase outwards from the main altered areas, reflecting an increasing mk^a component in the alteration assemblages away from the fault. As hydrothermal fluids progressed through the North Lyell system, their compositions were modified by interaction with the various host lithologies, producing distinct alteration assemblages within the surrounding units. Geochemical analyses of altered Owen Conglomerate around North Lyell reveals three assemblages that define a geochemical signature for the alteration fluids: sericitic (Al203+K20+Cr+Rb), phosphate-hematite (Fe203+P205+La), and baritic assemblages (Ba+ Sr). The sericitic and phosphatehematite assemblages have also been detected within the Upper Owen Conglomerate. The change in alteration styles from the volcanics through to the Owen Conglomerate reflects the transition from a weakly acidic, reduced, sulphide-rich environment, to acidic, highly oxidised conditions within the conglomerate. Oxidation of the fluW occurred primarily at the contact of the CVC with the Owen Conglomerate, as demonstrated by intense hematisation in the Owen

26

Post-Devonian cover rocks Late Devonian Granites 11,1,»[ Ordovictan Gordon Group limestone and c m Siluro-Devonian Eldon Grp siliciclastics F ^ l

Upper Cambrian-Lowef Ordovician Owen Conglomerate and correlates

CAMBRIAN NW of Henty Fault

m

Dundas Gp and related sequences

_ p M

Central Volcanic Complex

SE of Henty Fault Tyndall Gp and related sequences r-T-ra Central Volcanic Complex y — j Western Vole. Sequence

A j Unassigned Cambrian sequences Ultramalic Mafic Complexes Success Creek Group

PRECAMBRIAN Metaquartzites etc. of Tyennan Region

10 km

Figure 1: General geology of central-western Tasmania (after Corbett et al., 1989).

Conglomerate adjacent to the GLF. The presence of minor aluminium and iron phosphates replacing euhedral pyrite within the highly altered Owen conglomerate, together with the abundance of pyrophyllite and hematite, provides evidence for the passage of acidic, oxidised fluids in the vicinity of the GLF. Away from the GLF, the transition to sericite-hematite alteration within the Owen Conglomerate indicates progressive neutralisation and oxidation of acidic alteration fluids by reaction with the conglomerate. Structural and isotopic evidence suggest a Cambrian volcanogenic origin for many of the Mt Lyell deposits (eg. Cox, 1981; Gulson and Porritt. 1987), however, mineralisation within the Ordovician rocks precludes this as the sole mineralising event at North Lyell. Many authors (eg. Hills, 1990; Solomon et al, 1987) have postulated that a Devonian hydrothermal event (during the Tabberabberan Orogeny) remobilised preexisting Cambrian Cu mineralisation and formed zones of bornite enrichment at structural intersections. This appears to have occurred where major conduits (eg. the intersection between the GLF and the North Lyell Fault) have channelled low pH reduced hydrothermal fluids into the overlying, oxidised Owen Conglomerate. The systematic

variations in alteration intensity observed around Nth. Lyell are believed to broadly define fluid pathways within the Owen Conglomerate away from the fault. This project would not have been possible without the logistic and technical support of the Mt Lyell Mining and Railway Company Ltd including access to maps , company reports and drill core and in particular, Chief Geologist, Murray Flitcroft is thanked for his ready assistance with invaluable discussions on local geology and interpretation. REFERENCES COX. S.F. 1981, The stratigraphic and structural setting of the Mt. Lyell volcanic hosted sulphide deposits. Tasmania. Econ. Geol., 76; 231 - 245 GULSON. B.L and PORRITT, P.M.. 1987, Base metal exploration of the Mt Read Volcanics. Western Tasmania: R. II Lead isotope signatures and genetic implications. Econ. Geol. 82:291-307 HILLS, P.B.. 1990. Mt Lyell copper-gold-silver deposits, in Geology of the mineral deposits of Australia and Papua New Guinea, ed. F.E. HUGHES, AusIMM, Monogr. 14,2 :1257-1266 SOLOMON. M.. VOKES. F.M. and WALSHE. J.L, 1987, Chemical remobilization of volcanichosted sulphide deposits at Rosebery and Mt. Lyell, Tasmania. Ore Geol. Reviews, 2:173-190

27

For "Specialist Group in Economic Geology' meeting, Armidale, 2-6 February 1993

Ore Metals in Magmatic Brine and Vapour: New Evidence from PIXE Microanalysis of Fluid inclusions Christoph A. Heinrich"', Christopher G. RyanS, and Terry P. Mernaghi Australian Geological Survey Organisation, Canberra, 2 Commmonwealth Scientific and Industrial Research Organisation, North Ryde Layer-yield modelling of Proton Induced X-Ray Emission (PIXE) spectra has been used for non-destructive analysis of individual fluid inclusions within a granite-hosted quartz (- cassiterite) vein from the Yankee Lode, Mole Granite (New England; Heinrich et al„ in press). Preliminary PIXE data have also been obtained from the Kidston breccia-hosted gold deposit (North Queensland). Both samples contain inclusions of magmatic brine and vapour which may have coexisted at temperatures in excess of SOOOQ and pressures above SOObar. The brine inclusions from Yankee Lode have an estimated composition (in weight percent): NaCI 20% , KCI 7%, CaCl2 0.9%, MnCl2 4%, FeCl2 14%, ZnCl2 1% plus (in

ppm by weight) Co 300, Cu 900, As 500, Br 400, Rb 1700, Sn -400, Cs 2000, Pb 3000. Cation ratios in this brine compared with trace element concentrations in the source granite closely match experimental data on equilibrium metal distribution between chloride-bearing aqueous fluids and silicate melts. A thermodynamic mass transfer model based on this analysis and published geological and isotopic observations predicts the following mineral deposition sequence as a result of progressive fluid-rock equilibration and mixing of the brine with cooler low-salinity meteoric fluids: cassiterite + arsenopyrite --> chalcopyrite —> galena ± sphalerite ± chalcopyrite ± Fesulfides (Heinrich and Ryan, 1992). This closely matches the paragenetic sequence and regional zonation observed in New England and in many other granite-related tin and base metal provinces. Low-salinity vapour inclusions coexisting in the same quartz sample form the Yankee Lode contain higher concentrations of CO2 and sulfur and about 1% (!) Cu. Concentration ratios of copper to all other first-row transition metals (Mn, Fe, Zn, and also Pb) are 10-100 times higher in the vapour compared with the brine inclusions. Complexes with reduced sulfur may be responsible for the preferential partitioning of copper into the low-salinity fluid. This surprising result indicates that liquid - vapour separation and preferential vapour transport of copper may be a major mechanism of base and precious metal segregation in magmatic-hydrothermal ore systems. If sulfur complexing is responsible for preferential partitioning of copper into the magmatic vapour phase, then gold is predicted to show a similar, but probably more extreme behaviour (Heinrich et al., in press). The Kidston breccia-pipe in North Queensland is presently the best-studied example of a major granite-related gold deposit that may have formed from a magmatic vapour. Baker and Andrew (1991) concluded from detailed geologic, fluid inclusion and stable isotope studies at Kidston,

28

that gold was precipitated at a relatively late stage in the paragenesis of the breccia system, from a low-salinity liquid that originated by cooling and condensation of a magmatic vapour phase. The high-temperature magmatic vapour precursor to this ore fluid is preserved in paragenetically early veins, including crenulate quartz in banded rhyolite known as 'brain rock'. This quartz contains C02-bearing vapour inclusions as well as high-salinity brine inclusions, similar to the sample from the Yankee lode except that both inclusion types consistently contain large tetrahedral daugter crystals of chalcopyrite. First qualitative PIXE data from a sample of 'brain rock' from Wise's Hill (Baker and Andrew, 1991) indicate that the vapour inclusions consistently contain higher concentrations of Cu (on a wt% basis) but much lower concentrations of Fe, Mn,

E

lOV.

O

10

10

15

x - r a y energy (keV)

Zn and Pb compared with the associated brine inclusions (see Figure). Short of direct analysis for Au in the fluid inclusions, these results provide the most direct confirmation that metal segregation by brine/vapour separation may be an essential, but so far not widely recognised process of selective metal enrichment. This may be of importance to the formation of many magmatic-hydrothermal gold deposits in the Southwest Pacific region, and may also play a role in the genesis of so-called 'metamorphic' gold deposits such as those in the Archaean Yilgarn Block of Western Australia. References Baker E. M., and Andrew, A. S.. 1991, Geologic, fluid inclusion, and stable isotope studies of the goldbearing breccia pipe at Kidston, Queensland, Australia: Econ, GeoL 86, 810—830. Heinrich CA, Ryan CG (1992) Mineral paragenesis and regional zonation of granite-related Sn-As-Cu-PbZn deposits: a chemical model for the Mole Granite district (Australia) based on PIXE fluid inclusion analyses. /nY. K. Kharakaand A S. Maest, (eds.) Fluid—Rock Interaction, Balkema, 1583-1587. Heinrich CA, Ryan CG, Mernagh TP, and Eadington PJ (in press) Segregation of ore metals t^etween magmatic brine and vapor: a fluid inclusion study using PIXE microanalysis. Econ. Geol. 87(6).

29

LASER ABLATION ANALYSIS OF SULPHUR ISOTOPES: AN ANALYTICAL TECHNIQUE NOW AVAILABLE IN AUSTRALIA D.L Huston"*\ M. Power^ and R.R. Large^ ^Centre for Ore Deposit and Exploration Studies, University of Tasmania Central Science Laboratory, University of Tasmania Laser ablation extractton has recently been applied in North America and Europe to the extraction of gasses from minerals which can then be analysed using conventional stable isotope mass spectrometry. This technique has been applied most successfully to the analysis of sulphur isotopes at the University of Wisconsin and the Scottish Universities Research and Reactor Centre. The Centre for Ore Deposit and Exploration Studies in conjunction with the Central Science Laboratory at the University of Tasmania have developed a laser ablation system for the analysis of sulphur isotopes in Australia. This communication reports initial results and calibration of the system. SYSTEM DESIGN Figure 1 illustrates the design of the University of Tasmania's laser ablation extraction and analytical

system for sulphur isotopes. A18 watt NdiYAG laser (1064 nm) is focussed and reflected such that it impinges vertically upon a sample which is located within a glass sample chamber. The sample chamber has two inlets: one allows pumping of the chamber and the introduction of oxygen gas, whereas the other allows the passage of oxygen and sulphur dioxide gas directly to an automatic clean-up line on a VG Sira 10 mass spectrometer. The sample and ablation are monitored from above using a zoom microscope with a video attachment. ANALYTICAL PROCEEDURE The samples consist of 150 ^m thick, 10 mm x 10 mm polished chips; these are placed on a glass pedestal in the sample chamber for analysis. The sample chamber is evacuated, a set amount of oxygen (-25 torr) is introduced, and the sample

Video Camera

turning mirror sample

electronic crosshair generator Monitor Olympus stereo zoom microscope focussing electrically timed shutter ^ element Quantronix model 117 CW TEM.. Nd YAG Laser fused silica window glass sample VG Isoprep Automatic high/low vacuum gas cleaning system

\

cold finger on VG Sira 10 Series II Mass Spectrometer

V," flexible x-y-z translator steel tubing ' flexible tubing Baritron pressure guage

HjO trap '(-100°C) 'irani pressure guage <:{> high/low vacuum

liquid Nj, trap (-ISOOC)

manifold =£>0, Oj aliquot

Figure 1. Design of the University of Tasmania's laser ablation sulphur isotope analytical facility. 30

Table 1. Fracttonation factors and precisions (one sigma) for laser ablation sulphur isotope analysis at the University of Tasmania (150-250 ^im holes). Rough surfaces

Polished surfaces

Pyrite Chalcopyrite Sphalerite Galena Pyrrhotite

Fractionation factor (%o)

Precison

6.75 3.80 0.19 3.90 2.92

0.15 0.33 0.41 0.40 0.38

(%,)

n 4 12 20 6 8

Fractionation factor (%o)

Precision

n

4.63 2.77

0.46 0.66

6 12

4.82

0.50

7

(%o)

chamber is isolated. The sample is lased for 1 -2 s to samples which were inferred to be homogeneous produce sulphur dioxide. based on conventional analyses. These analyses Sulphur dioxide and excess oxygen are introduced were used to determine the sanrple fractionation into the automatic clean-up line on the mass factors presented In Table 1. From these and earlier spectrometer, where water is collected at -10O'^C and experiments, the following observations were made: sulphur dioxide is collected in a liquid nitrogen trap. • Heating of the sample chamber and transfer is Although collection occurs over a set 10 minutes, required to facilitate sample transfer. only about 60% of the sulphur dioxide is collected • Varying the oxygen pressure from 10 to 50 torr did owing to the slow diffusion of sulphur dioxide in an not effect the results significantly. oxygen atmosphere. Alternative collection techniques • Varying the ablation period from 1 to 3 s had only are presently being tried to improve this efficiency. minor effects of the results. After collection, the excess oxygen gas is pumped • Although only 60% of the produced sulphur dioxide away, the sample is transferred to the cold finger on is transferred and analysed, automation of the the inlet to the mass spectrometer, and the sample is collection, clean-up and analytical procedures has analyzed. The collection and analytical process has allowed reproducible results. been automated in a process that requires --40 • Ablation of rough surfaces resulted in larger minutes per analysis. fractionation factors than ablation of polished surfaces. • Larger holes (>250 |im) yielded larger fractionation RESULTS factors and poorer precisions than smaller holes (150-250 tim). Ablation generally resulted in holes that penetrated through the chip with diameters between 70 and 500 • One 150-250 ^im hole yielded sufficient sulphur dioxide for analysis of pyrite, sphalerite, pyrrhotite ^m. Holes in chalcopyrite and sphalerite tended to and chalcopyrite. Multiple holes are required for have relatively constant diameters, whereas the the analysis of galena. diameters for pyrite, pyrrhotite and galena tended to be more variable. Ablation of sphalerite and galena For an oxygen pressure of 25 torr and a 2 s ablation produced white (ZnO) and brown (PbO) powders, time, 150-250 ^im holes on polished surfaces yielded respectively. The power required for ablation varied a precision (one sigma) of between 0.3 and 0.4%o. as follows: pyrite > sphalerite > pyrrhotite > This precision compares well with precisions reported chalcopyrite > galena. The power required also varied by other workers (Crowe et al., 1990, Geochim. with the quality of polish; highly polished samples Cosmochim. Acta, v. 54, p. 2075-2095; Kelley and Fallick, 1990, Geochim. Cosmochim. Acta, v. 54, p. required higher power for ablation. Table 1 illustrates the results of experiments 883-888). Fractionation factors for analyses on conducted from May to July 1992 using automated polished surfaces vary from 0.19%ofor sphalerite to sample freeze down, clean-up and analysis. During 6.75%ofor pyrite. Ablation of rough surfaces increases this period over 180 analyses were undertaken on these by up to 2.78%o. pyrite, sphalerite, pyrrhotite, chalcopyrite and galena

31

COMPOSITIONAL STUDIES OF PYRITE FROM VOLCANIC-HOSTED MASSIVE SULPHIDE DEPOSITS IN EASTERN AUSTRALIA D.L Huston"^^ S.H. Sie^, G.F. Suter^ and C.G. Ryan^

''centre fo^ Ore Deposit and Exploration Studies, University of Tasmania CSIRO Division of Exploration Geoscience, Sydney

Although not an economically important mineral in volcanic-hosted massive sulphide (VHMS) deposits, the ubiquitous occurrence and "garbage can" nature of pyrite allows its trace element composition to be used as a guide to the mechanisms of ore formation. The proton microprobe (pixeprobe) has been used to determine pyrite compositions in samples from three eastern Australian VHMS deposits: Mt. Chalmers, Queensland; Rosebery, Tasmania; and Dry River South, Queensland.

hydrothermally recrystallised pyrite at Mt. Chalmers and metamorphically recrystallised pyrite at Rosebery and Dry River South generally contain values below the detection limit (15-20 ppm). Pyrite grains have higher maximum values of groups (1) and (2) in the nearly undeformed Mt Chalmers deposit than the greenschist grade Rosebery deposit, which contains higher maximum values than the amphibolite grade Dry River South deposit (Table 1). These data suggest that metamorphic recrystallisation not only "cleans" pyrite ANALYTICAL METHOD of tiny inclusions, but it also anneals non-stoichiometric lattice substitutions. At the Rosebery deposit, this Polished thin section were used for pixeprobe process has caused the expulsion of Au from pyrite analysis. As the penetration range of 3 MeJ protons to form electrum along pyrite grain boundaries. in sulphide minerals is 20-40 ^im, grains greater than Group (3) elements are uneffectedby hydrothemial 50 |im were analyzed to minimize the possibility of and metamorphic recrystallisation. At Mt. Chalmers, analysis of underlying grains. The larger grain size the highest Se values occur in the base of the also allowed the use of 20-30 ^im beam spots; beam AilpN currents (to 10 nA) were limited by the count rate in 8#(PP«) the detector. Analyses were carried out for a 5-6 m-C • charge with a 200 ^m or a 300 ^im Al filter to absorb most of the FeK lines. Spectra were analyzed using the GeoPIXE method. Zn-Pb-rlch CiHfch RESULTS AND DISCUSSION 20 Table 1 summarises the total variation in the 163 analyses undertaken. Elements for which concentrations were determined can be divided into three groups: (1) elements present as the result of i i.. inclusions of other minerals (Cu, Zn, Pb, Ba, Ag, Sb 1 and Bi), (2) elements present as non-stoichiometric substitutions in the pyrite lattice (As, Tl, Mo, Au, In and Cd), and (3) elements that substitute for Fe (Co and Ni) or S (Se and Te) in the pyrite lattice. 10 • Groups (1) and (2) elements (as exemplified by As) in hydrothermally recrystallised pyrite are elevated in the upper part of the massive sulphide lens relative to the lower part and the stringer zone of the Mt. Chalmers deposit (Fig. 1). Rapid quenching of pyrite nat the top of the mound allows the incorporation of non-stoichiometric substitutions. Lower in the mound V v|Volcanic$ands««ments Massive sulfide and in the stringer zone, (re)crystallisation occurs Cliloritic siltstone Altered volcanic rocks, more slowly, which prevents the incorporation of, or ^ ^showing stringers expels, group (1) and (2) elements from the pyrite. These processes also effect the concentration of Au in pyrite. Pyrite with habits (colloform or spongy) Figure 1. Diagrams showing variation with depth and suggestive of rapid crystallisation have the highest lithology in the concentrations of As and Se in pyrite concentrations of Au (to 200 ppm), whereas from the Mt Chalmers deposit.

32

Table 1. Variations in the content of trace elements (ppm) in pyrite grains from the Mt. Chalmers, Rosebery and Dry River South deposits, eastern Australia. Element

Mt. Chalnners (n = 105 from 15 sections)

Group 1

Cu Zn Pb Ba Ag Sb Bi

Group 2

As Tl Mo Au In Cd

Group 3

Co Ni Se Te

Rosebery (n = 28 from 6 sections)

Dry River South (n = 30 from 6 sections)

Low

High

Low

High

Low

High

55 <10 <15 <35 <4 <10 <10

45200 9430 86600 187000 460 7290 2300

<30 <20 <10 <35 <6 <10 <10

3340 8090 20300 130 375 710 90

25 <10 <15 <30 <5 <10 <15

6190 580 6040 190 46 25 660

<7 <15 <4 <15 <7 <6

46700 7830 580 210 320 22

<10 <15 <4 <20 <8 <7

11600 955 1530 50

16 <10 <4 <15 <7 <6

5750

<50 <40 <6 <15

2080

<60 <50 <6 <10

<50 <45 <6 <10

590 300 420 25

—

200 65

massive sulphide lens intothe Cu-rich stringer zones. Se values in pyrite from the upper part of the massive sulphide lens and from the footwall alteration zone below the stringer zone are substantially lower (Fig. 1). A similar distribution occurs at the Dry River South deposit. These distributions may be caused by the ascentof a hydrothermalfluid containing magmatically derived Se through the Cu-rich stringer zone into the basal portion of the massive sulphide lens. At the top of the lens, this fluid is diluted by seawater which contains minimal quantities of Se. CONCLUSIONS • High Cu, Zn, Pb, Ag, Bi and Ba contents in pyrite result from inclusions of chalcopyrite, sphalerite, galena, tetrahedrite and/or barite. • As, Mo, Au and Tl occur as non-stoichiometric lattice substitutions.

33

—

52 —

90 15 20

—

25 25 — —

The highest trace element content in pyrite results from rapid crystallization. Colloform and spongy pyrite have higher trace element contents than coarse-grained pyrite, and coarse-grained pyrite grains from the massive sulphide lens contains higher trace element levels than those in the footwall stringer zone. Metanrx)rphic and hydrothermal recrystallisation "cleans"pyrite of inclusions and non-stoichiometric substitutions formed during rapid precipitation. Au values are highest in As-rich, colloform pyrite grains from Mt. Chalmers. Most Rosebery pyrite grains contain Au values below the detection limit (15-20 ppm). The increase in Se values in the Mt. Chalmers Curich stringer zone in consistent with a hydrothermal fluid that contains magmatic Se mixing with seawater.

Oxygen isotopes as an indicator of large-scale hydrothermal paieoflow and economic Sn mineralisation, Renison Mine, western Tasmania. Paul A. Kitto Centre for Ore Deposit and Exploration Studies, University of Tasmania. Carbonate replacement stages of mineralisation at the Renison Tin Mine are intimately associated with extensional, brittle deformation structures that developed during the forceful emplacement of the Devonian Pine Hill Granite (PHG) (Kitto and Berry, 1991). The Federal Bassett Fault (FBF) and associated fault structures acted as conduits that focussed Sn bearing fluids from the underlying northwest trending PHG ridge into host dolostones of the mine sequence (Kitto and Berry, 1992; Bajwah et. al. ,in press). In the immediate mine area, metal accumulation calculations along a two kilometre strike length of the FBF highlight the presence of two major dilational zones associated with economic Sn mineralisation, i.e., the Federal and Ren-Deep orebodies. An overlapping metal zonation is recognised outward from the Sn rich dilational zones through to low grade Pb-Zn mineralisation. The 3^®0qtz values for main stage mineralisation (quartz-pyrrhotitecassiterite) show a systematic increase from +11.2 %»to +17.1 %o

within the FBF stucture above the PHG (Figure 1). Fluid inclusion results, along the FBF, show systematic decreases in homogenisation temperatures that mimic the oxygen isotope data. Systematic increases in d^®Oqtz away from the FBF have also been recognised on the listric extensional Transverse Faults that developed off convex west flexures within the footwall of the FBF (Kitto, 1992). The Transverse Faults acted as feeders for the hydrothermal fluids responsible for formation of carbonate replacement up-dip orebodies. Isotopic variations in 8^®0qtz suggest a gradual cooling of magmatic hydrothermal fluids as they ascended along fault structures, prior to reaction with the mine sequence dolostones. Contoured 3^®0qtz values and fluid inclusion results tor the FBF clearly define two major dilational zones associated with fluid focusing and economic Sn mineralisation at Renison.

References BAJWAH, Z. U.; WHITE. A. J.; KWAK, J. A. P.; PRICE. R. C. In prep. The Renison Granite, western Tasmania: A petrologlcal and geochenfiical investigation of hydrothennal alteration. KITTO, P. A.; BERRY, R. F. 1991. A history of brittle deformation and related mineralisation at Renison tin mine, Western Tasmania. Rec. Bur. Miner. Res. Geo!. Geophys. Aust. 1990/95. (SGEG Ore Fluids Conference, Canberra). KITTO, P. A.; BERRY, R. F. 1992. The structural controls on mineralisation at the Renison tin mine. Westem Tasmania. Abstr. geol. soc. Aust. 32:70-71. KITTO, P. A.; 1992. The geological and structural controls on mineralisation at the Renison tin mine.

Bull. geol. Surv. Tasm. 70:97-117.

34

Figure

Contours for 3 ^Oqiz from the main stage of qtz.-po.-cass. mineralisation, Renison Tin Mine.

THE GENESIS OF HIGH-TEMPERATURE ARCHAEAN LODE-GOLD MINERALIZATION IN THE COOLGARDIE GOLDFIELD, WESTERN AUSTRALIA: THE ROLE OF GRANITOIDS. J.T. Knightl*, J.R. Ridleyl, D.I. Groves^, R.Napier2. 1 Key Centre for Strategic Mineral Deposits, University of Western Australia, Nedlands, W.A. 6009, Australia. 2 The Department of Earth Sciences, University of Leeds, Leeds, LS2 9JT, The U.K. The Coolgardie Goldfield, located at the eastern margin of the ca. 2700-2690 Ma volcano-sedimentary sequence of the Norseman-Wiluna Belt, Western Australia, covers approximately 800 km^, was discovered in 1892, and has since produced over 50t of gold. The regional geology consists of an arcuate series of complexly deformed maficultramafic rocks, with minor sedimentary rocks and felsic volcaniclastic rocks. The sequence has been intruded by a suite of felsic and mafic porphyries, and is bounded to the west by the Bali and Calooli Monzogranites, to the east by the Mungari Granite, and to the south and south-east by unnamed granitoid bodies. The country rocks at Coolgardie have been metamorphosed to the lowamphibolite facies. In mafic rocks, the metamorphic assemblage consists of homblende+calcic plagioclase (An>35)-filmenite±titanite±quartz±epidote, and in ultramafic rocks the assemblage is tremolite-fchloriteiforsterite ±talc± carbonate ± anthophyllite. Amphibole-plagioclase geothermometry (Spear, 1980; Blundy and Holland, 1990) indicates that, to the east, peak metamo^hic temperatures reached 480±50®C. The metamorphic grade of the greenstones increases westwards towards the syntectonic Bali Monzogranite, with temperatures of 520±50°C attained in the centre of the Goldfield, and peak metamorphic temperatures of 545±50®C proximal to the granitoid contact. Archaean lode-gold deposits in the Coolgardie Goldfield are located in laminated quartz reefs sited in brittle-ductile shear zones, along sheared porphyry-ultramafic rock contacts, and in quartz vein-sets. The structural style, domainal distribution with respect to the western granitoid-greenstone contact, and fabric-defining mineralogy of the structures which host gold mineralization suggest that they formed during a progressive deformation event, during which heterogeneous stress patterns developed in response to either the anomalous geometry of the syntectonic granitoids, or their intrusion into the greenstones. Two different styles of gold-related walkock alteration can be recognised at Coolgardie. On the basis of recent classifications (Witt, 1991; Mueller and Groves, 1991), the gold deposits can be subdivided into: (i) a high-temperature group characterised by the presence of gamet+hornblende+plagioclaseH-calcite±biotite±Kfeldspar, with pyrrhotite the main sulphide mineral and no arsenopyrite, and (ii) a medium-temperature group characterised by calcic amphibole+biotite+plagioclase-fcalcite, with abundant arsenopyrite and pyrrhotite. Amphibole-plagioclase and arsenopyrite geothermometry indicate that the medium-temperature group of deposits formed at 480±50°C in the east, increasing to 510-520±50°C in the central and western parts of the Coolgardie area. Gamet-biotite and amphibole-plagioclase geothermometry (Hodges and Spear, 1982; Spear, 1980) undertaken on equilibrium assemblages of the hightemperature group indicates that they formed at 520-560t25'^C at the western margin of the Goldfield. Sphalerite and amphibole-plagioclase geobarometry (Hutchison and Scott, 1981; Blundy and Holland, 1990), and critical metamorphic mineral assemblages at Nepean, in the south of the Goldfield (Groves et al., 1975), indicate that both styles of gold-related wallrock alteration formed at pressures of 3-4kbars. The spatial distribution of the two styles of gold-related walkock-alteration is broadly controlled by distance from the syntectonic granitic plutons to the west of the belt. The high-temperature group are located more proximal to the granitoid contact than the medium-temperature group. An isograd, which divides the occurrence of the two alteration styles closely parallels the plan-view contact between the greenstones and the

36

Bali Monzogranite. This pluton, and other syntectonic granitoids in the Coolgardie Goldfield, are all strongly magnetic, magnetite-bearing plutons, with magnetic susceptibilities of 6-11 S.I. xlO"3 units. In contrast to the metamorphic evidence for two distinct styles of alteration, reconnaissance fluid inclusion and stable isotope studies indicate that gold mineralization was deposited by a homogeneous H20-C02-CH4-NaCl fluid, characterised by a narrow range of ore-fluid values (7-10 %oSMOW) and 5D values characteristic of either magmatic or metamo^hic waters. There is no clear relationship between die metamorphic grade of a gold deposit, and therefore its plan-view distancefromthe Bali Monzogranite, and its stable isotopic signature andfluidcomposition. In addition, Pb and Sr isotope studies of gold-related galena and scheelitefromdeposits in the Coolgardie Goldfield indicate gold-camp-scale homogeneity of the Pb- and Sr-source to the fluid. The data from this study and others (Mueller et al., 1991; McNaughton et al., in review) suggest that gold mineralization in the > 150km district from Mt Pleasant to Kambalda, including Kalgoorlie and Coolgardie, was deposited from a homogeneousfluid,characteristic of a large-scale hydrothermal system an order of magnitude larger than the Coolgardie Goldfield. The granitoids at Coolgardie, through the mechanics of their intrusion and/or their anomalous geometry with respect to regional stressfields,have clearly played a role in the development of heterogeneous gold-bearing structures, and, together with regionalscale metamorphism, appear to be responsible for the distribution of walkock alteration types in the Goldfield. Their oxidised nature also suggests that they could be genetically related to gold mineralization, since Cameron and Carrigan (1987) have shown that Archaean gold mineralization in Canada is preferentially associated with oxidised (magnetite-bearing) granitic intrusions. However, the lack of obvious gradients in stable or radiogenic isotope ratios awayfromthe granitoids combined with the homogeneous Pb isotope data on a province-scale, argues against a local source, such as the granitoids actually exposed at Coolgardie, for the auriferousfluids.A deeper granitic source is not excluded by the available data.

References

Blundy, J.D. and Holland, T.J.B. 1990. Calcic amphibole equilibria and a new amphibole-plagioclase geothermometer. Contrib. Mineral. Petrol. 104: 208-224. Cameron, E.M. and Carrigan, W.J. 1987. Oxygen fugacity of Archean felsic magmas: relationship to gold mineralization. In Current Research. Part A: Geol. Surv. Can. Paper 87-lA: 281-298. Groves, D.I., Binns, R.A., Barrett, F.M.and McQueen, K.G. 1975 Sphalerite compositions from Western Australian nickel deposits, a guide to equilibria below 300^C. Econ. Geol. 70: 391-396. Hodges, K.V. and Spear, F.S. 1982. Geothermometry, geobarometry and the Al2Si05 triple point at Mt. Moosilauke, New Hampshire. Am. Mineral. 67: 1118-1134. Hutchison, M.N., and Scott, S.D. 1981. Sphalerite geobarometry in the Cu-Fe-Zn-S . system. Econ. Geol. 76: 143-153 McNaughton, N.J., Groves, D.I. and Witt, W.K. 1992. The source of lead in Archaean lode-gold deposits of the Menzies-Kdgoorlie-Kambalda region, Yilgam Block, Westem Australia. Mineralium Deposita, submitted. Mueller, A.G. and Groves, D.I. 1991. The classification of Westem Australian greenstone-hosted gold deposits according to walb-ock alteration assemblages. Ore Geol. Rev. 6: 291-331. Mueller, A.G., De Laeter, J. and Groves, D.I. 1991. Strontium isotope systematics of hydrothermal minerals from epigenetic Archean gold deposits in the Yilgam Block, Westem Australia. Econ. Geol. 86: 780-809. Spear, F.S. 1980. NaSi-CaAl exchange equilibria between plagioclase and amphibole: an empirical model. Contrib. Mineral. Petrol. 72: 33-41. Witt, W.K. 1991. Regional metamorphic controls on alteration associated with gold mineralization in the Eastern Goldfields Province, Westem Australia: implications for the timing and origin of Archean lode-gold deposits. Geology 19: 982-985.

37

GOLD LOCALISATION IN THE INNER CONTACT AUREOLE OF THE TABLETOP GRANITE, PINE CREEK GOLDFIELD, N.T. K. C. Lawrie Department of Geology, James Cook University, Queensland, 4811. The Enterprise gold mine is a lOMt deposit hosted within Lower Proterozoic metagreywackes and shales of the Pine Creek Geosyncline. Mineralisation occurs predominantly in quartz veins, within the contact aureole of the late orogenic Tabletop Granite. Structural mapping demonstrates that the auriferous veins are localised at the intersection of a transpressional shear zone and an anticlinal fold hinge within the indurated aureole. Mapping of open pit exposures has revealed a complex vein history that results from progressive reactivation of the principal shear zone, and associated mineralised splays. Multi-stage parageneses have been elucidated within each splay, and several distinctive fluid components recognised. The main open pit lies within the cordierite and biotite zones of the contact aureole. The granite truncates and is not deformed by the pre-existing regionally pervasive NWtrending upright folds and associated cleavage, and the contact metamorphic assemblages overprint the regional greenschist facies assemblages. Studies of vein paragenesis clearly indicate that throughout reactivation of the shear zone, fluids of different origins were tapped. Dominant fluid types are high temperature, high to hypersaline fluids, and lower temperature, low salinity H2O-CO2 fluids. Observations are consistent with fluid mixing and boiling, particularly during deposition of sulphiderich assemblages. However, petrological and microstructural reltionships indicate that gold occurs late in the mineralisation paragenesis, and is associated with quartz with a distinctive Au-Bi± base metals and fluid signature. The gold-bearing stage is associated with extensive retrogressive chloritisation of pre-existing homfels and higher temperature wallrock alteration assemblages. The Tabletop Granite appears to have played an important role in localising gold at the Enterprise deposit through ground preparation and in destabalising gold complexes through fluid mixing and boiling. Preliminary data indicates that it is unlikely that gold emanated directly from the pluton at its site of emplacement. It appears more likely that auriferous fluids are related to lower crustal metamorphism coincident with the underplating that gave rise to generation of the I-type magmas that characterise the Cullen Batholith, and were introduced to the site of deposition through focussing in regionally developed shear zones.

38

PORPHYRY-RELATED CARBONATE BASE METAL GOLD SYSTEMS : THE TRANSITION BETWEEN EPITHERMAL AND PORPHYRY ENVIRONMENTS Leach TM (1) and Corbett GJ (2) 1) CMS New Zealand Ltd. 11 Blake St., Ponsonby, Auckland 1, New Zealand 2) Corbett Geological Services, 29 Carr St, Waverton, NSW 2060 The classification of porphyry-related carbonate base metal gold deposits, groups a number of significant low sulphidation SW Pacific Rim gold systems which are neither shallow level, precious metal , quartz vein hosted epithermal ( adularia-sericite ) nor deep seated, Cu/Au porphyry in character. Handley and Bradshaw ( 1986 ) and Sillitoe ( 1988 ) alluded to the existence of this class of deposits in emphasising the magmatic association and noting an overlap between the epithermal and porphyry environments in relation to Porgera. Characteristics of the porphyiy-related carbonate-base metal gold systems are: i) gold mineralisation is closely associated with iron and base metal sulphides and carbonate deposition within vein through to breccia settings ii) commonly hosted within, or at the margins of, intrusive rocks ill) high base metal contents typically as Zn>Pb>Cu iv) carbonates exhibit a wide range in chemistry; with Fe-, Mn-, Mg-, and Cacarbonates having distinct spatial zonations v) carbonate-base metal-gold deposition may be preceded either by distal porphyry-related quartz stockwork or epithermal/mesothermal quartz ± adularia ± sericite veining depending on the depth of the system. vl) mineralising fluids are transition between those of low temperature dilute epithermal systems and high temperature, saline porphyry systems Some significant SW Pacific Rim gold deposits which are classified within this porphyry-related carbonate base metal type are : in Indonesia, Kelian ( >4 M oz Au ); in Papua New Guinea, Porgera mineralisation types A, B and E ( >6 M oz Au), Mt. Kare, the Morobe Goldfield group of deposits ( past production with alluvial 3.7 Moz Au) including Wau, Edie Creek, Kerimenge (1.8 M oz Au), Hidden Valley (2.4 M oz Au ), as well as Busai and Kulumadau on Woodlark Island and Maniape at Kainantu; in the Solomon Islands, Gold Ridge. This mineralisation is distinct from the structurally controlled type D roscoelite-bearing and locally bonanza gold grade mineralisation described from Porgera ( Handley and Bradshaw 1986 ) as post carbonate-base metal; and which at Mt. Kare is interpreted to be contemporaneous with the more widespread carbonate-base metal gold event. Mineralisation forms within these systems through the mixing of hot, gaseous, relatively saline mineralised fluids from buried porphyry bodies at depth, with cool, dilute condensate or ground waters from near surface environments. Thus, major structures such as the graben bounding structures of the Bulolo Graben, Morobe Goldfied, provide an important control to the setting of these deposits. Regional structures may also control the setting of diatreme breccia intrusive bodies which commonly predate the formation of these deposits. As phreatomagmatic explosive events, derived from the same intrusive source, these breccias are not in themselves mineralised but provide valuable ground preparation for the following mineralised fluids. Local dilational settings within subsidiary structures may facilitate the formation

39

of banded, in places crustiform, veins which, in a manner similar to adularia-sericite epithermal gold systems, may account for bonanza gold grades. Mineral zonations of the carbonate-base metal gold system with time and space reflect the initiation of degassing of a porphyry source and the gradual cooling and influx of surficial waters. A similar paragenetic sequence which is evident for many deposits may be represented as : 1. An initial quartz event which varies from porphyry-related quartz stockwork veining in deep systems, to quartz-adularia-sericite veining in shallower environments, depending on the level within a hydrothermal system 2. The carbonate base metal event is the main phase of gold deposition by the progressive mixing of magmatic fluids with ground waters. In places gold mineralisation partially predates deposition of carbonates and base metals, elsewhere they are totally contemporaneous. The resultant differing mineral assemblages which reflect the distance from the porphyry source, as exemplified at Kelian ( Van Leeuwen et al, 1990 ) and present with variations in all other deposits, are : DISTAL to the porphyry source or COOL Vein Tvpe quartz>carbonate carbonate>quartz carbonate

Carbonate Tvpe Fe (siderite) Mn (rhodochrosite) MnMg (kutnahorite) MgCa (dolomite) CaMg (Mg-calcite)

Base Mstals Zn>Pb>Cu

Ca (Calcite)

Cu^Zn^Pb

Sphalerite Zn>Fe

Fe-Suiphides pyrite-marcasite pyrite

Fe>Zn

pyrite-pyrrhotite

PROXIMAL to porphyry source or HOT 3. Late stage either dominated by surficial fluids with kaolin, interlayered clay, gypsum, quartz deposition; or dominated by deep fluids with calcite deposition. Locally gold mineralisation may persist into the initial stages. Economic gold mineralisation occurs where optimum mixing conditions were present, generally within the Mn-Mg carbonate zone and in structurally favourable environments;, whereas bonanza grades occur where sudden quenching of the upflowing hottest mineralised fluids have taken place ( eg at Woodlark). Identification of feeder structures and zonations in the above mineralogy and mineral chemistry can provide vectors which point towards significantly mineralised portions of a porphyryrelated carbonate base metal gold system. Furthermore, projecting the fluid flow direction down dip can identify a possible location of the porphyry source. Thus a careful analysis of these systems represents a valuable exploration tool. Handley G.A. and Bradshaw P.M.D. 1986. The Porgera Gold Deposit, Papua New GuineaJa Proc. Gold '86 Symposium Toronto pp 416-424 Sillitoe R.H. 1989. Gold Deposits in the Western Pacific Arcs: The Magmatic Connection Jn The Geology of Gold Deposits: The Perspective in 1988, Econ Geol Mono 6 p.274-290 Van Leeuwen, T.M., Leach T.M., Hawke A.A. & Hawke M.M. 1990. The Kelian disseminated gold deposit. East Kalimantan, Indonesia. Jour Geochem Explor 35 pp1-61

40

GOLD MINERALIZATION AT MAGDALA MINE, ST AWELL, WESTERN VICTORIA. By

Benjamin S.E. Mapani and Christopher J.L. Wilson. School of Earth Sciences, University of Melbourne, Parkville 3052, Victoria. Abstract The sediments that host the gold bearing laminated quartz veins in the Magdala gold mine, are affected by six discrete deformation episodes Di-De, the first three deformations are related to regional events that predate mineralization. The earlier deformations produced the Magdala, Extended, and Crown cross anticlines that control the localization of subsequent shearing and location of the gold mineralization. Shearing that relates to the formation of the gold mineralizing systems, took place during D4, in which shear zones and quartz veins were developed. The Stawell granite is believed to have intruded at about this time, probably supplying thefluidswhich contributed to the formation of the quartz veins and the source of heat for driving a convective cell that facilitated mineralization. D5 a brittle-ductile phase, and associated with episodic and synchronous faulting led to the development of polymictic breccias on pre-existing shear surfaces and is partly responsible for the remobilization of the mineralization. D6 was a faulting event constrained by the intrusion of late monchiquite and furchite dykes. The structural environment at Magdala can be divided into two zones, the Central Lode which is a reverse thrust with W to S W over E to NE transport and the Scotchmans Fault Zone(SFZ) which is also a reverse fault zone with a NW over SE transport. The SFZ truncates the Central Lode faults implying a superposition of different stress regimes. The principal paleostress orientations were calculated for both Central Lode and the SFZ as follows: Central Lode: al plunges 24® 224°, a2 plunges 20° 312° and a3 plunges 32°-^l 15° while the SFZ paleostress regime is defined by al 21°^281°, c2 56°-^140°anda3 21°-^21° In Central Lode the quartz veins have been deformed such that they no longer retain their original texture and morphology, but now have developed a mylonitic texture, and a

41

sigmoidal shear zone foliation. Discrete S-C bands exist within the quartz veins and host sediments, such that strain can be qualitatively mapped. Shortening estimates in Central Lode have yielded values of 78%-85%. In the SFZ the 30-40cm thick quartz veins are faulted, but still retain their original laminated texture and morphology, and are less deformed than the host graphitic schists. Smaller 5cm thick quartz veins which occur in the graphitic schist have been reduced in grain size forming smaller 2-5cm shear bands and rotated into a shear zone sigmoidal foliation defined by layer parallel graphite, quartz, muscovite, chlorite and minor clinozoisite within 0.6-2m thick shear zones. Ore zonation is observed in Central Lode, in which the pyrrhotite-pyritearsenopyrite mineralization is concentrated on the shear surfaces that form S-C bands and on the discrete through going principal slip surfaces, which are continuous over several tens of metres. In the SFZ no ore zonation is observed, and the principal sulphides are pyrite-arsenopyrite-±chalcopyrite, found as 2-5mm euhedral crystals restricted to the quartz veins or on the host rock-quartz vein fault contacts. Pyrrhotite is present as anhedral deformed grains in the host sediments and on fault surfaces. Specimensfromthe SFZ indicate that these rocks were deformed first by dislocation creep and then were overprinted by pressure solution associated with D5. The pressure solution overprint led to the formation of low temperature copper and iron sulphides such as tetrahedrite, bornite and marcasite at the expense arsenopyrite, pyrite and pyrrhotite. This contributed to the relocation of gold within sulphide lattices and the release offreegold. Thefreegold was redeposited in cracks of the deformed minerals and in grain boundaries of sulphides and occasionally silicates. In the Central lode this overprint has similar effects as in the SFZ, however there is also the formation of stilpnomelane rims around ankerite porphyroblasts.

42

DEFORMATION T E X T U R E S IN STRATIFORM MASSIVE SULPHIDE DEPOSITS K.R. McClay, Department of Geology, Royal Holloway, University of London, Egham, Surrey, TW20 OEX UK. Many Zn-Pb (Barite) and precious metal massive sulphide deposits occur in deformed and metamorphosed terranes. Primary depositional features and relationships are commonly significantly modified or obliterated by later deformation and metamorphism. Mineralogical transformations, textural and microstructural changes, recrystaliisation and remobilisation may alter the character and morphologies of these massive sulphide deposits. Deciphering these changes is essential for the erection of genetic and exploration models and for the timing of ore formation. The macro and microstructural textures of massive sulphide deposits from the Canadian Cordillera, Greenland and Alaska are used to illustrate the response of Fe, Zn, Pb and Cu sulphides to deformation and

Major stratiform massive sulphide deposits of the northern Cordillera (after McClay 1991). The Zn-Pb (ibarite) deposits of the northern Canadian Cordillera range in age from Cambrian, Ordovician, Silurian and Mid-Late Devonian (McClay 1991). These stratiform deposits occur in the Palaeozoic Selwyn Basin - Kechika Trough

43

(Fig. 1), which developed by episodic extension within the carbonate platform of the passive margin of ancestral North America. The massive sulphide deposits formed in extensional sub-basins dominated by fine-grained clastic sedimentation. During the Late Jurassic to Mid Cretaceous these basins and the stratiform ore deposits were subjected to sub-greenschist fades to low greenschist fades regional metamorphism and fold / thrust deformation as they were Incorporated into the western part of the Rocky Mountains-f\/lackenzle Mountains foreland fold and thrust belt. Deposits described In this paper Include - the Anvil District (Faro and Vangorda orebodies), the MacMillan Pass area (the Tom deposit), the Howards Pass deposit, the Cirque deposit, the Driftplle Creek deposit and the Greens Creek orebody, Alaska (Fig. 1). The deposits have been mapped at surface and underground where available. Samples from surface, underground and diamond drill holes have been studied using transmitted and reflected light microscopy, scanning electron microscopy and electron microprobe analysis. The dominant ore mineralogy Is sphalerite, galena, pyrite and In some cases significant barlte is present. Minor sulphides may include marcasite, pyrrhotite, chalcopyrite and silver sulphosalts. The host lithologies are typically fine-grained, carbonaceous black shales and siltstones. Laminae and bands of fine-grained framboldal to spheroidal pyrite are commonly concentrated in the host sediments in the vicinity of the Zn-Pb deposits. Bedded, banded and disseminated barite beds characterise the mineralised stratigraphy in the Gataga and MacMillan Pass areas. The deposits are typically tabular to lensoidal sulphide and barite bodies. The ores vary from massive and banded sulphides ± barite, to disseminated and laminated sulphides in fine-grained black shales and siltstones. Macroscopic folds and foliations are developed where the deposits have been strongly deformed (Anvil District, MacMillan Pass - Tom deposit; the Gataga deposits, the Cirque deposit, and In the Greens Creek orebody). Disruption and modification by ductile shear zones and faulting is common. Detailed microstructural studies have shown that primary depositional and early diagenetic textures, particularly in the more refractory sulphides, pyrite ( Gu Llanxing and McClay 1992) and sphalerite, can be recognised in deposits that have enjoyed sub-greenschist to greenschist fades metamorphism and deformation. Galena, pyrrhotite, barite and chalcopyrite are typically totally recrystalllsed and / or partly remobilised such that primary depositional textures are rarely preserved. Coarse-grained (vent proximal?) pyrite in massive sulphide ore fades exhibits colloform banding, radial growth and zoned growth textures whereas the laminated ore fades characteristically show spheroidal and framboldal pyrite and microcrystallite pyrite grains. Coarse-grained sphalerite shows sector growth zoning and growth twinning whereas fine-grained sphalerite occurs as Isolated grains and atoll Intergrowths with pyrite spheroids. Deformation microstructures formed at sub- to low greenschist fades conditions are characterised by cataclastic flow and brecciation In coarse-grained massive pyritic sulphides, and by folding, recrystallisation and cleavage development in fine-grained sulphides and barite. Galena is usually totally recrystalllsed commonly with preferred shape and crystallographic orientations. Strong pressure solution fabrics are found In barite, in fine-grained laminated pyrite layers and In mixed sulphide - carbonate + silicate lithologies. Stylolitic pressure solution seams are enriched In sphalerite and galena. Velning with remobilised galena and barite quartz/carbonates Is common. Marked local transposition is 44

found with significant remobilisation of galena, barite and fine-grained pyrite. Chalcopyrite is commonly remobilised into fractures within coarse pyrite grains and as pressure shadows around pyrite porphyroblasts. Pyrrhotite is usually recrystallised and displays deformation lamellae and kinking. At upper greenschist to lower amphibolite facies most of the sulphides and barite are totally recrystallised with marked increase in grain sizes and the development of foam and porphyroblastic textures. Pyrite in particular develops strong grain growth and metablastic textures. Preferred shape orientations have been observed in pyrites (Faro orebody, Anvil District). Preservation of primary depositional or early diagenetic features are rare but spheroidal and framboidal pyrite features have been observed encased in coarse-grained pyrite. At these higher grades microstructures are dominated by recrystallisation and grain growth textures. Veining and remobilisation of galena, chalcopyrite and in places sphalerite attests to the importance of fluids during deformation. Thermal annealing post the peak of deformation produces grain growth and equant / foam textures in most of the sulphides. Microstructural studies indicate that most of the sulphide deposits studied were formed as massive to disseminated lensoidal bodies deposited at or near the same time as the enclosing clastic sediments but have undergone significant textural modification by deformation and metamorphism. In some deposits important remobilisation occurs and this may significantly alter the metal distributions within the deposit. The textural and microstructural characteristics of the major ore forming sulphides in these deposits are critically evaluated. The significance for depositional models and for exploration are discussed. References

Gu Lianxing and McClay, K.R. 1992. Pyrite deformation in stratiform lead-zinc deposits of the Canadian Cordillera. Mineralium Deposita, 27,169- 181. McClay, K.R. 1991. Deformation of stratiform Zn-Pb (barite) deposits in the northern Canadian Cordillera. Ore Geology Reviews, 6, 435-462.

45

AN EXTREME HIGH HEAT-PRODUCING Sn-W GRANITE FROM THE BUSHVELD COMPLEX: EVIDENCE FOR A LONG-LIVED HYDROTHERMAL SYSTEM

N.J. McNaughton"1, P.J. Pollard2, J. Stacey^ D.I. Groves'^ and R.G. Taylor2

1. Key Centre for Strategic Mineral Deposits, Department of Geology. University of Western Australia, Nedlands, W.A. 6009, Australia 2.

Department of Geology and Key Centre in Economic Geology, P.O. James Cook University of North Queensland, Townsville, Old. 4811, Australia

Felsic magmatic systems are responsible for a range of mineralization styles at the magmatic and near-solidus stages, but these rarely persist for longer than 10^-10^ years after emplacement. High heat-producing (HHP) granites, however, may maintain temperatures of >200''C "indefinitely" (e.g. 10^ years; McNaughton et al., in press) if their U-Th-K contents are sufficiently high and they are blanketed by a few kilometres of insulating or heat-producing cover (e.g. Fehn et al., 1978). In such systems, it has proved difficult to estimate the duration of cooling because isotopic resetting and blocking temperature effects lead to equivocal conclusions. In order to obtain a better understanding of the duration of hydrothermal alteration within HHP granites, it is necessary to: (i) reliably estimate heat-production within the granite since emplacement, and (ii) identify a reliable method of determining ages of both granite emplacement and the time of uplift to near-surface environments, where cooling below isotopic blocking temperatures may be induced.

To address both these problems, the U-Pb isotopic decay scheme was employed to study an extreme HHP granite. Zircon U-Pb studies by SHRIMP ion-microprobe indicate that the Bobbejaankop Granite in the Zaaiplaats area of the northern Bushveld Complex was emplaced at 2059+2/-3 Ma, which is broadly synchronous with layered mafic-ultramafic rocks and other granites of the Complex. Tin-tungsten mineralization at Zaaiplaats is related to high-temperature (>400^0) magmatic-hydrothermal fluids (Pollard et al., 1991a; 1991b), but alteration, including secondary hydrothermal-mineral growth, occurred at temperatures down to 200''C. Fluid inclusion and stable isotopic studies suggest the system remained largely closed to external non-magmatic fluids during cooling to this temperature.

The current radiothermal heat production of the Bobbejaankop Granite from the zone of disseminated Sn-W mineralization, unmineralized granite and the related Lease Granite are estimated from U-Th-K contents and average 8.2,8.6 and 11.5 |iW/m3, respectively (McNaughton et al., in press). This is markedly higher than the <6 jiW/m3 recorded for the well-documented HHP granites of SW England (Lee et al., 1987).

From Pb-Pb isotopic studies of whole-rocks and an estimate of the initial isotopic composition, pre-weathering heat-production of the mineralized and unmineralized Bobbejaankop Granite is calculated to be 30.3 and 25.3 |iW/m3, respectively. These extreme values were operating prior to weathering, which caused cryptic but major U-loss and lesser Th-loss in othen/vise fresh granite sampled from the open cut and diamond drill core. Pb-Pb isotope studies of this type have only been undertaken on a few HHP granites, but suggest most Precambrian HHP granites may show 25-75% U-loss and lesser Th-loss due to recent weathering. It follows that the heat-producing properties of many granites may be significantly underestimated by analysing U-Th-K contents without a Pb-isotopic study.

46

The duration of hydrothermal activity in granites of the Zaaiplaats area is estimated from the age of emplacement at 2059 Ma, and the age derived from Pb-Pb whole-rock isochrons for various sampling domains within the area. The Pb-Pb isochron ages represent the time when all samples within a particular suite or domain had the same initial isotopic composition, and thereafter evolved independently to their modem Pb isotopic composition: i.e. the time of uplift-induced cooling below the U-Pb mobility blocking temperature for the granite samples. All granite suites and groups yield imprecise but consistent isochron ages around 1100 Ma, which also corresponds to the minimum zircon 207pb/206pb ages recorded by SHRIMP on zircons with extensive corrosion, alteration and infilling with secondary hydrothermal minerals (McNaughton and Pollard, submitted). Initial isotopic ratios are compatible with essentially closed-system isotopic evolution within specific domains of the granites from emplacement to ca. 1100 Ma. For example, the 80 m thick sheet of disseminated Sn-W mineralization appears to have evolved independently from an overlying granite sheet some 50 m above the top of the mineralized zone. In summary, the granites at Zaaiplaats have extreme HHP-characteristics and show secondary mineral growth and infilling of primary porosity which persisted for a billion years at temperatures >200''C. As HHP granites are also implicated in the genesis of major Au(-Cu)deposits such as at Telfer and in the Pine Creek Inlier, Northern Territory, their importance to ore deposit research may dwarf their megre abundance, particularly in Precambrian rocks where Pb-isotopic studies are most applicable. Acknowledgements:

The authors acknowledge the support of Zaaiplaats Tin Mining Company Ltd., Ian

Williams and Bill Compston (ANU), and JCU, UWA, ANU, CUT, ARC and MERIWA. References: Fehn, U., Cathles, L.M. and Holland, H.D., 1978, Hydrothermal convection and uranium deposits in abnormally radioactive plutons: Econ. Geol. v. 73, p. 1556-1566. Lee, M.K., Brown, G.C., Webb, P. C., Wheildon, J. and Rollin, K.E., 1987, Heat-flow, heat production and thermo-tectonic setting in mainland U.K.: J. Geol. Soc. v. 144, p. 35-42. Ludwig, K.R. and Silver, L.T., 1977, Lead isotope inhomogeneity in Precambrian igneous feldspars. Geochim. et Cosmochim. Acta v. 41, p. 1457-1472. McNaughton, N.J. and Pollard, P.J., submitted. Mineral lead isotope systematics within high-heat producing granites: Chem. Geol. McNaughton, N.J., Pollard, P.J., Groves, D.I., Taylor, R.G., in press, A long-lived hydrothermal in Bushveld granites at the Zaaiplaats tin mine: lead isotope evidence: Econ. Geol. Pollard, P.J., Andrew, A.S. and Taylor, R.G., 1991a, Fluid inclusion and stable isotope evidence for interaction between granites and magmatic-hydrothermal fluids during formation of disseminated and pipe-style mineralization at the Zaaiplaats tin mine: Econ. Geol. v. 86, p. 121-141. Pollard, P.J., Taylor, R.G., Taylor, R.P. and Groves, D.I., 1991b, Petrographic and geochemical evolution of pervasively altered Bushveld granites at the Zaaiplaats tin mine: Econ Geol. v. 86, p. 1401-1433.

47

UNDERSTANDING THE NATURE OF PROXIMAL-DISTAL MINERALIZATION IN THERMAL-AUREOLE GOLD DEPOSITS IN THE PROTEROZOIC OF NORTHERN AUSTRALIA USING LEAD ISOTOPES

N.J. McNaughton*, 8. Sheppard and N.M. Goellnicht

Key Centre for Strategic Mineral Deposits, University of Western Australia, Nedlands WA 6009

Exploration for thermal-aureole gold deposits in the Proterozoic rocks of northern Australia has intensified in recent years. The spatial association of economic deposits with felsic magmatism has long been recognised, and some form of genetic association has been demonstrated at both Telfer, Western Australia (Goellnicht et al., 1989; submitted) and in the Pine Creek Inlier, Norther Territory (Wall, 1990; Wall and Taylor, 1990; Sheppard and McNaughton, submitted). The spatial and genetic association between mineralization and granites does not necessarily identify the source of gold and ore-related metals. Ore metals could derive from the granite, be scavenged from wallrocks by fluid-rock interaction around the granitic heat source, or could derive from a deeper source.

Studies into metal sources are increasingly looking to isotopic data to identify the origin of solutes in ore fluids, and hence constrain genetic models of ore formation. Lead isotopic studies are particularly applicable to the study of metal sources in Precambrian ore deposits, due to the unique dual U-Pb decay schemes and favourable half-lives. Ore-related minerals, such as galena and pyrite, have U/Pb«0 and preserve the Pb isotopic composition at the time of mineralization. Comparison of these ore fluid compositions with coeval Pb in potential source rocks offers a powerful tool to constrain Pb sources, and hence genetic models of ore formation, particularly for granite systems which typically show isotopic contrasts with hostrocks by virtue of their different evolutionary paths.

Lead isotopic studies of Proterozoic Au deposits at Tom's Gully in the Pine Creek Inlier of the Northern Teritory show that deposits proximal to spatially associated granites do not derive all of their Pb from either the granite or the host-rocks to mineralisation. Some ore fluid Pb is most likely derived from deeper parts of the thermal aureole, either from the lower stratigraphy or older Archaean basement

(e.g. Sheppard and

McNaughton, submitted). Large scale (i.e. >5 km vertical extent) hydrothermal circulation around a heat source is necessary. In this respect, the vertical scale of the hydrothermal system and the source of Pb is similar to that proposed for Archaean lode-gold deposits hosted by greenstones in Western Australia (McNaughton et al., 1990; 1992).

In the Telfer district of the Paterson Province, there is both proximal (<1 km from granites) and distal (> 1 km) Au-Cu (± other metals) mineralization.

Distal mineralization, typically stratiform to stratabound reefs,

stockworks and sheeted veins, has initial Pb isotope compositions intermediate between those of coeval granite and host-rock Pb. Proximal mineralization, typically porphyry-style Cu ± Au and polymetallic skarns, shows greater evidence of a magmatic Pb signature, partiuclarly in vein-style, rather than replacement-style mineralization. In this district, there is clearly a magmatic input from l-type granites, with an increasing contribution from host-rock Pb due to fluid: rock interaction along fluid conduits away from the granites

48

(Goellnicht et al., 1989; submitted). The Pb isotopic data for deposits in the Teller district and the Pine Creek Inlier, in combination with experience from Archaean terranes. provide important constraints on the genesis of thermal-aureole gold deposits. In all cases. Pb from the spatially associated granite is. at best, a minor component of Pb in the more economic deposits. Coeval host-rock Pb in the case of Telfer, or Pb from deeper parts of the thermal aureole in the case of Tom's Gully, dominates the Pb budget. The combination of major structures to act as ore-fluid conduits and, a granitic heat ± metals source appear to be the necessary ingredients to form a major Proterozoic thermal aureole-style of gold deposit. For some deposits, for example Tom's Gully, there may also be a Au-Pb contribution from non-magmatic source rocks at depth, and in this respect these deposits have some similarities to Archaean lode-gold deposits.

References:

Goellnicht. N.M.. Groves, D.I. and McNaughton, N.J., submitted. Lead isotope evidence for the role of Late Proterozoic fractionated granitoids in the genesis of polymetallic mineralization in the Telfer district. Western Australia. Mineralium Deposita. Goellnicht. N.M., Groves, D.I., McNaughton, N.J. and Dimo, G., 1989. An epigenetic origin for the Telfer gold deposit. Econ. Geol. Monograph 6, 151-167. McNaughton, N.J., Cassidy, K.F., Dahl, N., Groves, D.I., Perring, C.S. and Sang, J.H., 1990. Constraints on genesis of primary gold deposits: Lead isotope studies. In Ho, S.E., Groves, D.I. & Bennett, J.M. (eds), Gold deposits of the Archaean Yilgarn Block, Western Australia: Nature, genesis and exploration guides. Geol. Dept & Extension, Univ. West. Aust., Publ. 20, 226-236. McNaughton, N.J., Cassidy, K.F., Dahl, N., de Laeter, J.R., Golding, S.D., D.I. Groves, Ho, S.E., Mueller, A.G., Perring, C.S., Sang, J.H. and Turner, J.V., 1992. The source of ore components in lode-gold deposits of the Yilgarn Block, Western Australia. In: Glover, J.E. and Ho, S.E. (eds.). The Archaean: Terrains, Crustal Processes and Metallogeny, Geol. Dept & Univ. Extension, Univ. West. Aust., Publ. 22, 351-363. Sheppard. S. and McNaughton, N.J., submitted. Thermal-aureole gold mineralisation at Tom's Gully, Australia: isotopic and physico-chemical composition of ore fluids, and gold depositional controls. Mineralium Deposita. Wall. VJ.. 1990. Fluids and metamorphism. Unpubl. PhD thesis, Monash University, Victoria. Wall, V.J. and Taylor, J.R., 1990. Granite emplacement and temporally associated gold mineralization. Proceedings 10th AGC, Geol Soc. Aust., Abstr. 25, 264-265.

49

THE BEHAVIOUR OF IRON IN HIGH-TEMPERATURE CHLORIDE BRINES D.C. McPhail 1 VIEPS Department of Earth Sciences, Monash University Clayton, Victoria 3168. Iron has perhaps the most poorly-understood behaviour of the major rock-forming elements in hydrothermal fluids. The dissolution, transport and deposition of iron and iron-bearing minerals are critically dependent on the equilibrium behaviour of aqueous iron complexes, and can also influence, directly and indirectly, the transport and deposition of other elements in ore-forming environments. The purpose of this abstract is to present experimental data on hightemperature (350®C to 500°C) aqueous iron behaviour and highlight some of the aspects and controls on the behaviour of iron in subcritical and supercritical hydrothermal chloride brines.

0.5(10-4)m to 4.0(10-4)m at 350®C and 1.1(10-4)m to 1.4(10-2)m at SOO^C, as KCI concentration increased from 0.1 to 4.0m in both cases. The data obtained at 350®C were interpreted to show an average ligation number of approximately 2 (2 CI atoms for every Fe atom) and probably indicate a combination of two or more iron-chloride complexes. More than one interpretation of the iron-chloride complexing is possible; however, the stabilities of ironchloride complexes were interpreted to be much higher than suggested in previous studies. Provisional thermodynamic properties were derived by assuming two of the most likely iron-chloride complexes individually and fitting their properties to the experimental data: log K350 = 7.0 (FeCI+ = Fe2+ + CI") and log K350 = -9.1 (FeCl2(aq) = Fe2+ + 2CI-). While these values are up to 4 log units lower than previous estimates from experimental and theoretical studies, they compare well with a value derived from measured iron concentrations in natural geothermal waters: log K300 = -7.0 for FeCI+ dissociation (Spycher and Reed, 1989 - Econ. Geol., 84, 328-359).

The solubility of the mineral assemblage Magnetite - K-feldspar Biotite - Quartz in KCI - MCI solutions was measured at 350®C, vapoursaturated pressure (subcritical) and 500°C, 1 kbar (supercritical), where the ligand concentration was varied between 0.1 and 4.0 molal (m) KCI (for reference, seawater is approximately 0.6m NaCI). Redox conditions were controlled by the equilibrium: Magnetite + K-feldspar + H2O = Annite (in biotite) + 0.502

The dramatic increase in the measured iron concentrations in the 500®C - 1 kbar experiments, more than two orders of magnitude, is best explained by the presence of a highorder iron-chloride complex, such as FeCU^- or perhaps K2FeCl4(aq). This contradicts current assumptions and interpretations of limited

and pH was controlled by the presence of the minerals, an estimated activity of water, and mass balances on potassium, chlorine and charge. Measured aqueous iron concentrations increased from

50

experimental data that neutral, intermediate-order iron-chloride complexes, such as FeCl2(aq), dominate in supercritical brines. The existence of high-order complexes means the transport of iron is greatly enhanced with increased salt concentration, even in near-neutral pH fluids. Even more importantly, a change in complexing from FeCU^(K2FeCl4(aq)?) to a lower-order complex like FeCl2(aq) with decreasing temperature will result in the precipitation of iron-bearing minerals. The magnitude of the precipitation may be spectacular.

precipitation of minerals would also affect some of the important chemical variables in the fluid phase, such as acidity (pH) and redox (ao2)- Finally, the amounts of fluid and therefore the time required for formation of iron-rich ore deposits may be considerably shortened. There are many ore deposits where iron transport and deposition are important as part of either primary ore deposition or as precursors to subsequent hydrothermal ore deposition. A number of Australian ore deposits exhibit massive amounts of hydrothermal iron minerals, for example, Tennant Creek, Olympic Dam, and Starra, where temperatures greater than 350°C and saline fluids are indicated. Iron transport and deposition are also important in porphyry copper deposits and the formation of mid-ocean ridge hydrothermal systems (black smokers), which also show large amounts of iron transport and deposition and occur at temperatures of 350X and greater, also with saline fluids. The ideas presented here may help to explain at least some of the chemical mechanisms and processes that operate in such environments.

Preliminary calculations show that more than 150 g of magnetite may precipitate from 1 kg of a 4 molal chlorine fluid (15 weight percent), when the temperature drops from 500®C to 350®C (with an attendant drop in pressure from 1 kbar to vapour-saturation), just from the change in iron-chloride complexing. This estimate was calculated assuming an oxygen fugacity buffered along the Magnetite - K-feldspar Annite equilibrium, neutral pH, and a predominance of FeCl2(aq) at 350°C. The implications of this behaviour are wide-ranging. Hightemperature brines (>400°C?) may have the capacity to carry large amounts of iron, even with low acid contents. Secondly, most of the iron would precipitate on simple cooling through the geologically-important temperature range of approximately 500®C to 350®C, ie. the precipitation of large amounts of iron oxides would be a normal consequence of cooling, and would not require other mechanisms, such as fluid mixing, wall-rock interaction or boiling. Thirdly, other iron-bearing ore and gangue minerals would be stabilized by the decreased stability of aqueous iron (eg. iron sulphides, copper-iron sulphides, arsenopyrite, biotite, chlorite, iron carbonates). The

51

GRANITOID-RELATED GOLD MINERALISATION IN THE BRAIDWOOD GRANODIORITE, SOUTHEASTERN N.S.W. K.G. McQueenl and C. Perkins 2 1. Faculty of Applied Science, University of Canberra, Belconnen, ACT. 2. Research School of Earth Sciences, Australian National University, Canberra, ACT. The Braidwood Granodiorite is a highly magnetic I-type pluton within the northern part of the Bega Batholith in southeastern New South Wales. It is a multiple intrusion formed by two separate injections of magma of very similar compositions. Both are metaluminous and unfractionated with high K, Rb, REE, Ba and Sr and Fe203/Fe0 ratios of 0.45-0.70 (Wybom and Owen, 1986). The two phases of the pluton now have meridonally trending, elongate outcrop patterns and at several places are separated by narrow screens of Ordovician metasediments. Both consist predominantly of medium-grained, equigranular, pale grey granodiorite with 15-20% mafic minerals (pyroxene, hornblende and biotite). Pyrite is a common accessory mineral, occuring as fine disseminated grains, larger aggregates and minor fracture-filling veinlets. The western phase is characterised by the presence of large prisms of primary hornblende and a greater abundance of aplite dykes, leucogranite stocks and more mafic dykes. Tliese features suggest that it was richer in water than the eastern phase. A number of vein and disseminated sulphide-gold deposits are hosted by the Braidwood Granodiorite. Deposits of this type, in the now eroded roof zone and upper portion of the intrusion, appear to have been the source for the large alluvial gold deposits developed along major drainage systems both within, and surrounding, the exposed area of the pluton. Between 1854 and 1920 more than 40 tonnes of gold were recovered from these placer deposits. The known primary gold mineralisation is mostiy concentrated in the westem phase of the Braidwood Granodiorite, particulary in the Major's Creek goldfield. Dargue's Reef is the largest of the known primary gold deposits and a detailed study of tliis deposit has provided insights into primary gold deposition in the Braidwood Granodiorite. Mineralisation at Dargue's Reef occurs in two main lodes which consist of narrow zones (0.6-9.5 m wide) of intense sericitic alteration and pyritisation, (15-30% pyrite) enclosed in areas of propylitic alteration. Deposition of barren euhedral-subhedral pyrite accompanied early-stage alteration and was followed by deposition of irregular pyrite containing numerous small inclusions of silicates, calcite, chalcopyrite, Bi-sulphosalts, galena, gold, trace tellurides, native bismuth and pyrrhotite. Separate aggregates of chalcopyrite, Bi-sulphosalts and tetrahedrite are intergrovm with the silicate alteration nninerals. Fluid inclusion data indicate that hydrothermal mineralising fluids evolved with time and that gold-bearing solutions were CO2rich, medium- to low-temperature (<350® C) and largely of late-stage magmatic origin. Fluids evolved from an early aqueous phase generated during water saturation of the crystallising pluton, possibly with some later incorporation of country rock fluid. K-Ar dates on alteration sericite from the main ore zone indicate a minimum age of 402 Ma for the alterations and suggest that mineralisation was broadly contemporaneous with emplacement of the host pluton in the earliest Devonian. Sulphur isotope ratios in pyrite from the Dargue's Reef mineralisation (d^^S -0.4 to -3.4 %o) and disseminated pyrite in the host granodiorite (B^^S +1.4 to +2.5%^?) suggest a magmatic source for the sulphur and partial reduction of SO2 from an early magmatic fluid phase in the upper part of the intrusion where some fluid escaped before complete reduction of SO2 to H2S. Lead isotope determinations on pyrite from the Dargue's Reef ore and on pyrite and K-feldspar from the host granodiorite are consistent with the granodiorite being the source of the Pb in the mineralisation. References Wybom, D. and Owen, M., 1986. 1:100 000 Geological Map Commentary, Araluen, New South Wales, Bureau of Mineral Resources, Geology and Geophysics, Canberra 44 p.

52

THE ROLE OF METHANE IN GOLD MINERALISATION IN THE MENZBESKAMBALDA AREA, EASTERN GOLDFIELDS, W.A. T.R Memagh^ and W.K Witt^ 1. Australian Geological Survey Organisation, GPO Box 378, Canberra ACT, 2601 2. Geological Survey of Western Australia, 100 Plain Street, Perth WA, 6000 Gold deposits in the Menzies-Kambalda area are characterised by carbonation and potassic alteration of the host rocks. Alteration assemblages vary systematically with regional metamorphic grade, up to and including upper amphibolite facies rocks (Witt, 1991). The distribution of metamorphic grade is complex, but amphibolite facies metamorphic rocks and high temperature alteration assemblages are commonly developed in broad thermal aureoles around late syntectonic granite intnisions. The spatial and temporal relationships indicate that regional metamorphic gradients existed during the mineralisation event, and that mineralisation and regional metamorphism were broadly contemporaneous. The composition of fluids associated with metamorphism and mineralisation was investigated by examining fluid inclusions (mainly in quartz) from selected deposits in the Menzies-Kambalda area which are estimated to have formed at temperatures ranging from 300' to 600'C. The inclusions were analysed by microthermometry and with the laser Raman microprobe. Low salinity HgO/COg fluids are observed in deposits forming at temperatures up to 500®C. Some inclusions contained 100% COg, but CH^ (±N2) was also present in the carbonic phase of most inclusions and some even contained 100% CH^. In the St. Albans deposit which formed at approximately 600''C, only inclusions in epidote were large enough for analysis by the above methods. Both vapour-rich (>50 vol.% vapour) and H^O-rich (<30 vol.% vapour) were observed. Only CH^ was detected in the vapour phase of both types of inclxisions and other gases, if present, were below detection limits. One generation of HgO-rich inclusions had high salinities and contained up to 5 daughter crystals, one of which was identified as calcite from its Raman spectrum. The close spatial association between compositionally distinct inclusions, and the range of intermediate compositions, suggests mixing of COg-rich fluids with CH^-richfluids.At Hawkins Find (a low temperature deposit) CO^-rich fluids occur in albite deposited at the margins of vughs, whereas CH^ also occurs in inclusions in quartz forming the centres of the vughs. These temporal relationships could not be demonstrated at other deposits, and broad contemporaneity between the two fluids seems likely. Previous fluid inclusion studies of sub-amphibolite grade deposits in the Yilgarn Block (summarised by Ho et aZ., 1990) have indicated that the ore-bearing fluids are characterised by low salinity, aqueoiis fluids containing variable amounts of COg (averaging about 25 to 30 wt %). CH^-rich fluids have been identified at Lancefield but the presence of CH^ has only been noted in a few other deposits (e.g. the Golden Mile, and Mount Charlotte). Therefore, the widespread occurrence of CH^ in the deposits selected for this study is of interest. CH^(±N2)-rich fluids may have been generated by devolatilisation reactions in the upper crust, or by interaction between metamorphic

53

fluids and the carbonaceous black shales which are common as interflow sedimentary units in the greenstones of the Menzies-Kambalda area. The ubiquitous presence of methane in all the gold deposits examined in the MenziesKambalda area has important implications for the mechanisms of gold deposition in this region. The fluid incliision studies provide evidence for mixing of CO^-rich and CH^-rich fluids which would lead to a reduction of the oxidation state of the ore bearing fluids. Furthermore, comparisons with the H^O - CO^ - NaCl system show that the addition of even small amounts of CH^ greatly enlarges the field of immiscibility at any temperature and this could also act as a trigger for gold deposition. REFERENCES Ho, S.E., Groves, D.I. and Bennett, J.M. (Editors), 1990 Gold Deposits of the Archaean Yilgarn Block, Western Australia: Nature, Genesis and Exploration Guides, Geology Department and University Extension, The University of Western Australia, Publication 20, 407 pp. Witt, W.K., 1991 Regional metamorphic controls on alteration associated with gold mineralization in the Eastern Goldfields province. Western Australia: Implications for the timing and origin of Archean lode-gold deposits. Geology, vol 19, pp. 982-985.

54

GRANNY SMITH: AN EXAMPLE OF A GRANITOID-HOSTED ARCHAEAN LODE-GOLD DEPOSIT V.J. Ojalal, J.R. Ridley^, D.I. Groves^ and G.C. Hal|2 "•Key Centre for Teaching and Research in Strategic Mineral Deposits, Geology Department, The University of Western Australia, Nedlands WA 6009 Australia 2piacer Exploration, PO BOX 764, Cloverdale WA 6105 Australia The Granny Smith gold deposits are situated 250 km NE of Kalgoorlie and 20 km S of Laverton at lat. 28°48' S, long. 122°25' E in the LavertonLeonora area of the North Eastern Goldfields Province of the Archaean Yilgarn Block, Western Australia. Mine production commenced in 1990 after 10 years of exploration, and the development plan of the mine was based on proven and probable reserves of 21 Mt with an average grade of 1.7 g/t in three deposits. The Granny Smith gold deposits, which are hosted by both granitic and sedimentary rocks, formed late in the structural history of the Yilgarn Block at a high crustal level in a largely brittle structural regime. The deposits are partly hosted by a composite, elongate (about 2 km x 5 km) and zoned calc-alkaline granitoid pluton, which has porphyritic and more mafic margins, and which has intruded into a structural corridor between zones of different structural orientations in the adjacent greenstone belt. The composition of the intrusive rocks varies from diorite, commonly containing clinopyroxene, to hornblende-biotite granodiorite which forms the major part of the Intrusion. Pegmatitic aplite veins represent the most fractionated part. Mineral composition and geochemical features (presence of hornblende + magnetite + titanite; a wide range in Si02 and metaluminous compositions)show similarities to Phanerozoic l-type granitoids. Biotite crystallization before hornblende, together with clinopyroxene, suggest that the intrusion crystallized from a hot, dry magma that would have been capable of intruding to a high crustal level. The present erosion level is close to the roof of pluton. In places it has a thin cap of hornfelsed sedimentary rocks, and it is surrounded by a 200 - 300 m wide contact metamorphic aureole. As a consequence, the outcrop pattern of the granitoid is irregular. Gold mineralization is located along a N-S striking deformation zone that partly follows the contact between granitoid and sedimentary rocks. In the granitoid, the gold mineralization is in a conjugate network of thin carbonate-quartz breccia veins and their alteration halos. The main sulphide phase is pyrite; other ore minerals present are pyrrhotite, chalcopyrite, sphalerite, galena, arsenopyrite and tellurides. Distal alteration is typified of the breakdown of clinopyroxene and amphibole and the development of sericite ± biotite ± chlorite ± hematite alteration, whereas and the proximal alteration zone is defined by more intense carbonate-sericite-silica-pyrite+rutile alteration (bleaching). Small movements along veins are common. In contrast, veins and faults in the sedimentary rocks are parallel to bedding. Importantly, conjugate vein orientations within the granitoid indicate that the local stress field was heterogeneous and controlled by the shape of the granitoid contact. The greatest variations in vein and implied stress orientations occur in the zones where the contact is most irregular. These are also the areas of best

55

mineralization. Fluid flow was thus focused in a regional-scale, low meanstress region created by the geometry of the granitoid intrusion. Its irregular contact caused deposit-scale variations in fluid flow and resulted in heterogeneous gold grades along the contact zone. Fluid inclusions interpretated to be trapped during the gold mineralization are low salinity (0 - 8 eq.wt% NaCI) C02-H20-NaCI±CH4 inclusions with variable liquid vapour ratio, and they occur as single inclusions, clusters, internal trails and cross-cutting trails. Liquid - vapour homogenization temperatures of these inclusions vary from 150 °C to 460 °C (mean 325 °C). C02-H20-NaCI±CH4 inclusions in the quartz veins from the proximal alteration zone have higher CO2 densities and lower salinities than inclusions in the quartz veins from the distal alteration zones. Nahcolite (Ts = 80 - 250 °C, mean 160 °C) is the most common daughter mineral; other carbonate minerals are also identified but they are rare. High salinity aqueous inclusions (> 20 eq.wt% NaCI, Thtot 90°C - 250°C, mean 150°C) and CH4-inclusions (Th < -81 °C) occur in quartz veins as cross-cutting secondary trails. Available data, particularly the high level emplacent of the granitoid but the CO2- and CH4-rich nature of the fluids trapped in the proximal quartz

veins, suggest that the granitic magma was not source of the ore fluid. Rather, the localization of the mineralization in and around the granitoid was due to stress heterogeneities due to its anomalous geometry within the enclocing greenstone sequence.

56

MICRO-TEXTURAL AND GEOCHEMICAL EVIDENCE FOR HYDROTHERMAL VENTING AT THE PINNACLES DEPOSIT, WESTERN NEW SOUTH WALES.

Joanna Parr* Department of Geology, University of Newcastle, Callaghan, Newcastle, NSW 2308, Australia. Stratabound, sediment-hosted, Broken Hill-type Pb-Zn-Ag deposits have commonly undergone multiple episodes of deformation and metamorphism. At Broken Hill itself, for example, the orebodies have been subject to at least three phases of folding and cross-cutting shear zones (Laing et al. 1978) which, together with granulite fades prograde metamorphism and amphibolite fades retrograde metamorphism, caused recrystallisation of the sulphide minerals and extensive remobilisation of the ores into favourable structural sites such as fold hinges (Plimer 1984). Consequently, sulphide minerals have re-equilibrated and recrystallised to form coarse grained massive orebodies which do not exhibit any primary depositional structures. The Pinnacles deposit, 15 km southwest of Broken Hill, is the second largest Broken Hill-type deposit in the Willyama Supergroup after Broken Hill itself. The Pinnacles deposit has also undergone high grade metamorphism and deformation: prograde metamorphism attained two pyroxene-granulite fades and retrograde metamorphism is represented by kyanite-bearing mylonitic shear zones. Four episodes of folding can be recognised in the area prior to the cross-cutting mylonitic shear zones. As a result, the Pinnacles orebodies have also been remobilised and concentrated in fold hinges and sulphide minerals have also extensively recrystallised to form coarser grained rocks. However, an early generation of pyrite has survived the metamorphism and deformation and is present in the sulphide orebodies as small fragments of concentrically banded pyrite which are locally interleaved with Fe oxides. The banded texture is well preserved and is found both in the galena-rich Pb lode and in the more siliceous sphalerite-rich Zn lode. Textural evidence indicates that the banding pre-dates the remobilisation of other sulphide minerals and is most consistent with formation as growth banding during hydrothermal activity, probably as colloform bands. The preservation of primary depositional textures in pyrite have been previously recorded to upper greenschist, possibly lower amphibolite fades metamorphism (McClay & Ellis, 1983). At metamorphic grades greater than these, annealing of the pyrite should occur. At the Pinnacles deposit, however, well preserved primary textures indicate that in certain circumstances pyrite grains can remain stable up to granulite fades metamorphism. The presence of pyrite that exhibits syndepositional growth banding has several important implications for the genesis of the Pinnacles deposit. Firstly, the

57

presence of colloform growth banding is indicative of hydrothermal fluid flow through open conduits, possibly some sort of hydrothermal vent, although, the lack of hydrothermal breccias or any other form of hydrothermal veining, suggests that this activity was not explosive. Secondly, the dominance of primary pyrite over pyrrhotite, in contrast to the Brol<en Hill orebody, suggests that the environment of precipitation was more oxidising at the Pinnacles than at Broken Hill. Other

significant

differences

between the

Pinnacles

and

Broken

Hill

orebodies include As-enrichment and Mn-depletion at the Pinnacles relative to Broken Hill and the local association of large quartz-magnetite units with the Pinnacles deposit. A possible stratigraphic

interpretation

might be that the

Pinnacles deposit represents a proximal hydrothermal vent in a partially oxidised (shallower?) environment, whereas the larger Broken Hill orebodies represent sulphide deposition in a distal, more reduced, (deeper?) basin, possibly sourced by the same hydrothermal centre.

References McClay, K.R. & Ellis. P.G. 1983. Min. Mag., 47: 527-538. Laing, W.P., Marjoribanks, R.W. & Rutland, R.W.R. 1978. Econ. Geol., 73: m a i l 36. Plimer, I.R. 1984. Austr. Jour. Earth ScL, 31: 379-402.

58

Timing of replacement processes in Mount Isa copper and lead-zinc orebodies using structural and microstructural relationships. W. G. Perkins MIM Exploration Pty. Ltd. Mount Isa, Q' Id. Application of structural and microstructural techniques to copper ore genesis at Mount Isa began in 1975 in collaboration with T. Bell of James Cook University. More detailed microstructural examination was subsequently completed by C. Swager. The results of these projects were published as Perkins(1984), Swager(1985), Swager et. al(1987) and Bell et. al. (1988). Re-examination of Mount Isa lead-zinc using these techniques and comparison with Hilton, Mount Novit, and HYC, was begun in 1989 and is continuing. This later work has completely changed the interpretation of lead-zinc, and it and other projects aimed at quantifying the copper system have modified previously published interpretations. These structural techniques have been applied to both copper and lead-zinc orebodies, in conjunction with mapping alteration profiles on selected sections, and, for lead-zinc, building on the sedimentological constraints provided by Neudert(1984). A consistent sequence of structures beginning with a bedding-parallel cleavage and localized folding, followed by small-scale thrusts, are folded by the main event(D3), with a variably developed cleavage. The early cleavage is reactivated during the main fold event. Sets of extension fractures are mostly younger than the small-scale thrusts and are associated with dextral shear on bedding. Dilational structures on bedding are developed during subsequent shortening, and these are symmetrically formed on both the long and short limbs of the Mount Isa Fold. Later deformation is concentrated towards the Paroo-Basement Fault system, where the earlier cleavages are more strongly developed, rotated and folded, in particular when associated with the Buck Quartz thrust. Analysis of the development of the mineralizing system involves relating the growth of neoformed non-sulphide and sulphide assemblages to these folds, cleavages and vein structures. Critical sites containing well-developed structural features in conjunction with variably advanced mineralization, in particular those with well-developed mineralization fronts, have been mapped and oriented samples taken. In many areas, the development of criteria to establish whether mineral growth is a diagenetic process or the result of alteration(defined here as being younger than deposition of the Mount Isa Group and the Paroo-Basement Fault), 59

can be quite difficult. At HYC, early small-scale thrusts are ubiquitous in the mineralized zones, but the two cleavages observed are only weakly developed. Combined with the fine-grained nature of sulphides, this makes it hard to unequivocally establish mineralization timing . A protracted alteration history is interpreted for both Mount Isa and Hilton, particularly with respect to carbonate growth. Dolomite grows replacively to form veins which are folded by the early cleavage, deformed dolomites are overgrown by new dolomites in zones of higher strain, and continue to replace folds and overgrow cleavages resulting in near bedding-parallel breccias in the lead-zinc areas, and cross-cutting breccias of the "silica-dolomite". Silicification partly overprints this dolomite at successive stages in lead-zinc areas, and forms a zone of total silicification of dolomite as a core to the "silica-dolomite" body. The economic sulphides everywhere consistently replace the alteration minerals, and appear to have a paragenetic sequence of sphalerite, galena and chalcopyrite. Sulphides are only locally deformed, not by pervasive ductile structures, but associated with late cross-cutting faults. The relationship between chalcopyrite and quartz is quite critical in understanding the controls on chalcopyrite deposition. Both these minerals very commonly deposit in sites previously occupied by coarse-grained dolomite. Rather than co-deposition of quartz and chalcopyrite(Perkins, 1984), a corrosion and replacement relationship is indicated by the sweeping of inclusions from single quartz grains around the margins of chalcopyrite(Swager, 1985, Heinrich et. al. 1988). In order to explain the cross-cutting relationships in stratiform deposits including Mount Isa lead-zinc, deposit-modifying concepts of recrystallization and remobilization have been commonly used. The nature of sulphide-gangue boundaries indicates that they are essentially unmodified from their initial replacement relationships, and that deformation, recrystallization or dissolution have not affected them. Fine-grained bedding-parallel sulphides appear to be deposited during the same event as adjacent breccia and vein-hosted sulphides. The evolution of interpretations using these techniques at Mount Isa shows that it is essential to understand the paragenetic and timing relationships of gangue and sulphide, to provide the constraints on possible chemical reactions involved in their deposition. They indicate that models involving fluid-rock reaction are much more likely to apply to Mount Isa deposits than fluid mixing models involving local basinal brines. Such interpretations have major implications for both sediment-hosted and volcanichosted base metal deposits elsewhere. 60

REFERENCES: Bell, T.H., Perkins, W.G. and Swager, C.P., 1988, Structural controls on development and localization of syntectonic mineralization at Mount Isa, Queensland:Econ.Geol., v.83, p.69-85. Heinrich, C.A., Andrew,A.S., Wilkins.R.W.T., and Patterson.DJ., 1989, A fluid inclusion and stable isotope study of synmetamorphic copper ore formation at Mount Isa .Australia. Econ. Geol. v.84, p529-550. Neudert, M.K., 1983, A depositional model for the Upper Mount Isa Group and implications for ore formation. PhD thesis Aust. Nat. Univ. published by Sedcon Pubis. Perkins, W.G.,1984, Mount Isa silica-dolomite and copper orebodies:The result of a syntectonic hydrothermal alteration system:Econ.Geol., V.79, p.601-637. Swager, C.P., 1985, Syndeformational carbonate-replacement model for the copper mineralization at Mount Isa, north-west Queensland:A microstructural study: Econ.Geol., v.80, p.107-125. Swager, C.P., Perkins, W.G., and Knights, J., 1987. Strata-bound phyllosilicate zones associated with syntectonic Cu orebodies, Mount Isa, Queensland:Aust.Jl Earth Sci., v.34, p.463-476.

61

MICROSTRUCTURES FOR TIMING ORE INTRODUCTION: CAUTION Neil Phillips Key Centre In Economic Geology James Cook University Townsville Qld Australia The current debate on the timing of metal introduction into several large Australian deposits (e.g. Mt Isa, Big Bell, Roseberry) has much in common with the debate a half century ago between proponents of "syngenetic" and "hydrothermal" origins: different methods seemingly give different timing results. Today's dilemma comes from the contradiction between evidence and interpretations based on microstructural work, and evidence from other geological observations including fluid evolution, geochemistry, deposit distribution and mineralogy. There is little doubt that microstructural studies have shown that ore minerals such as sulphides and native metals overprint late folds and fabrics in several deposits (e.g. Dj at Mt Isa). The first stage of interpreting this observation is to infer that such grains reached their final resting place during or after the specific deformation event. The second stage of interpretation is to infer that age constraints can be placed on the whole orebody fi-om these microstructural observations on individual grains: the validity of this second step is less clear. The consequence of following these stages of interpretation is to give answers that do not rest easily with conclusions drawn from independent aspects of the ore body geology, so there is some reason to question the second stage of the interpretation. Higher grade metamorphic domains are particularly prone to equivocal evidence supporting alternative timing hypotheses. These areas typically display several periods of deformation, and ore minerals overprinting the fabrics are very common. It is of some concern that microstructural studies have not identified deposits where NO ore minerals overprint these fabrics. Instead, the microstructural approach leads one towards believing that most or all ore deposits in metamorphic terrains are syn or post the metamorphism. This conclusion would seem most unlikely as similar metal deposits form in terrains that have not undergone high grade metamorphism. One way to interpret the "late" microstructures of the ore minerals is to note the crystal structure and low melting point of sulphides and native metals, link these to the viability of diffusion and mobility of these minerals during the elevated temperatures of metamorphism, and expect a different behaviour for ore minerals from the average silicate minerals. Laboratory experiments show quite convincingly the ease of resetting minerals in many ore systems (e.g. Cu - Ag - S). Base metals as an example The Mt Isa - Cloncurry area represents one of the major world zinc provinces, maybe the major one. Some deposits are at quite low metamorphic grade with minor deformation (e.g. McArthur River, Century) whereas others are within greenschist fades metamorphic domains with a multiple deformational history (e.g. Mt Isa Pb-Zn, Hilton, Dugald River). Even the highest metamorphic grade domains contain zinc-bearing deposits and prospects (e.g. Cannington, Maramungee, Pegmont). 62

This distribution does not support an essential role for medium to high grade metamoiphism in the introduction of zinc mineralization into this province. Before accepting a syn/post D3 - post peak metamorphic timing for zinc at Mt Isa on the basis of microstructures, it is important to also consider Mt Isa within the context of the zinc province around it, and also the types of fluids possibly responsible for its origin. Archaean sold as an example Archaean gold deposits can be used to illustrate the same dilemma that arises from microstructural and petrological evidence. Big Bell is a large gold deposits in an upper amphibolite facies domain of the Yilgam Block: microstructural studies have demonstrated gold grains overprinting high grade and retrograde fabrics in the schistose and gneissic rocks within the deposit. Reference to the earlier mentioned laboratory experiments on ore mineral systems would suggest that timing evidence might be best preserved in the silicate (gangue) mineralogy that hosts the ore at Big Bell and forms a surrounding alteration envelope. In contrast, the sulphides and native metals would be expected to reset more easily. The alteration envelope includes a major potassium alteration zone now represented by muscovite, Kfeldspar and biotite rich rocks that have been present during peak metamorphism (based on their high grade mineral assemblages, e.g. sillimanite - K-feldspar - cordierite garnet). This potassium alteration is a feature of most major greenstone gold deposits (excq)tions being some in banded iron formation where there is no alumina to stabilize My K-bearing mineral). The most reasonable suggestion from Big Bell is that the deposit is surrounded by a typical K-rich alteration envelope, that envelope has undergone peak metamorphism, and the alteration is most likely part of the gold event. A corollary of the microstructural interpretation of the gold at Big Bell being retrograde, however, would be to dissociate genetically the gold event from its spatially related Kalteration: this is of some concern when the gold and K-alteration are part of one event in sub-amphibolite facies deposits and there is no independent evidence to separate them at Big Bell. A different slant on the microstructural approach to gold deposits in high grade metamorphic domains is to look at the gold itself. In greenschist facies deposits, gold is typic^ly associated with margins and cracks in pyrite and other sulphides. In high grade domains significant gold occurs as dustings within silicate minerals such as hornblende, K-feldspar and clinopyroxene (e.g. Big Bell, Southern Cross Belt deposits, Karonie), where the silicate is interpreted to have overgrown the gold during metamorphism. The implications for exploration for gold in high grade domains is significant. The microstructural approach has been used to target retrograde shear zones as potential gold fluid channelways: the petrological evidence suggests these retrograde shear zones come after the gold formation. Conclusion Microstructures should be used in conjunction with all other available evidence including viable fluid sources and the regional context. The apparently contradictory conclusions coming from the microstructures are best explained in terms of preferential resetting of sulphides and native metals. Timing each grain is not necessarily a valid way to time an ore body. 63

EXTREME CHEMICAL FRACTIONATION AND LATE-STAGE FLUID EVOLUTION ASSOCIATED WITH Ta-Nb-LI MINERALIZATION AT THE YICHUN MINE, SOUTH CHINA. P.J. Pollard'*, R.P. Taylor^ and R.G. Taylor' 1. Department of Geology and Key Centre in Economic Geology, James Cook University, Townsville 4811, Australia. 2. Ottawa-Carleton Geoscience Centre, Department of Earth Sciences, Carleton University, Ottawa K1S 5B6, Canada. The Yichun Ta-Nb-Li mine is situated within the Yashan Batholith, which is composed dominantly of two-mica granites which are intruded by sheets of Limica granite and topaz-lepidolite granite emplaced in the roof zone of the batholith. The two-mica granites contain quartz, alkali feldspar, plagioclase, biotite (protolithionite) and muscovite (Li-muscovite-zinnwaldite solid solution), as well as garnet, apatite, ilmenite, zircon, monazite and columbite. The Li-mica granite is similar to the two-mica granites but lacks biotite and garnet. The topaz-lepidolite granite is composed dominantly of albite, lepidolite and quartz, with K-feldspar, topaz, amblygonite-montebrasite, and accessory Mn-columbite, microlite, Ta-rich cassiterite, apatite, monazite, ilmenite, zircon and pollucite. Chemical fractionation within the Yashan granites results in moderate to extreme enrichments in AljOg, NajO, F, Li, P, Ta, Sn, Rb and Cs, and depletions in SiOj, TiOj, FejOg, MgO, CaO, Ba, Sr, Zr and REE. Highly compatible (eg. Ti, Fe, Mg) and incompatible (eg. Ta) elements show continuous variation through the whole granite suite, suggesting a continuous evolution from two-mica to topaz-lepidolite granite. However, several elements including Al, F, Li, P, Rb and Cs show markedly discontinuous trends expressed by large enrichments in the topaz-lepidolite granite relative to the Li-mica granite. Discontinuities in chemical trends occur when the concentration of Li, P and F can no longer be buffered by minerals crystallizing from the melt (eg. apatite and Li-muscovite-zinnwaldite). Melt concentrations therefore rise sharply until a new mineral which can accommodate these elements begins to crystallize (eg. amblygonite-montebrasite, lepidolite, pollucite). For example, when molar CaO(3.2*P205) in the melt approaches zero, it no longer contains sufficient calcium to buffer the phosphorus concentration by apatite crystallization. In the Li-mica granite this occurs at approximately 0.2 wt.% P2O5 and the phosphorus concentration rises until amblygonite-montebrasite crystallizes (minimum 0.43 wt.% PjOg in the topaz-lepidolite granite). A similar situation occurs when Limuscovite-zinnwaldite can no longer accomodate sufficient Li, F and presumably Rb and Cs, and their concentrations rise until lepidolite saturation occurs (in addition to amblygonite-montebrasite and pollucite saturation). Li, P and F all tend to associate with Al in aluminosilicate melts, which accounts for the dramatic increase in AI2O3 in the topaz-lepidolite granite. Li acts as a charge balancing cation that allows Al to enter network forming sites, P forms

64

stable species with Al and perhaps allcalis, while F may form stable Na-AI-F complexes (eg. London, 1987). The combined effects of these melt interactions are to increase the activity of SiOj and promote quartz crystallization, produce a shift toward albite-rich compositions, and provide favourable sites for incorporation of elements such as Rb, Cs and Sn in the melt. The traditional A/CNK plot suggests that the Yashan granites become increasingly peraluminous with fractionation. However, the alumina saturation index {ASI = molar AljOa/KzO + NajO+O.SCaO + LijO + RbjO + CSzOdondon, 1992) indicates that the topaz-lepidolite granite becomes progressively more alkaline, due mainly to the large increase in LijO. Late-stage fluid saturation gave rise to an alkaline fluid with 1 for K, Rb, Cs, Li, F, Ba, Sr, Ti, Zr, Nb, W, Th and REE. This fluid phase precipitated a mica-rich "greisen" assemblage in the apex of the intrusion. Evidence from the Yashan granites indicates that topaz-lepidolite granites are derived by extreme chemical fractionation of two-mica granites, rather than by small degrees of fusion of residual crustal sources (cf. Manning and Hill, 1990). Extreme fractionation results from a lack of buffering of the melt composition by liquidus minerals due to the diminishing availability of elements such as Ca, M g and Fe for the formation of apatite and Fe-Li-micas. The topaz-lepidolite granite has a similar bulk composition to Macusani glasses, some rare-element pegmatites, and other P-enriched topaz-Li-mica granites. The low B concentration and lack of aluminosilicate minerals in the early phases of the Yashan Batholith may reflect derivation of the magma from a lessargillaceous source than is sometimes proposed for these types of rocks (eg. Pichavant et al., 1988). References London, D., 1987. Internal differentiation of rare-element pegmatites: Effects of boron, phosphorus and fluorine. Geochim. Cosmochim. Acta v. 51, p. 403-420. London, D., 1992. Phosphorus in S-type magmas: The PjOg content of feldspars from peraluminous granites, pegmatites and rhyolites. Am. Mineral, v. 77, p. 126-145. Manning, D.A.C. and Hill, P.I., 1990. The petrogenetic and metallogenetic significance of topaz granite from the southwest England orefield. In Stein, H.J. and Hannah, J.L. (eds.) Ore-bearing granite systems: petrogenesis and mineralizing processes. Geol. Soc. America. Spec. Paper 246, p. 51-69. Pichavant, M., Kontak, D.J., Briqueu, L., Herrera, J.V. and Clark, A.H., 1988. The Miocene-Pliocene Macusani Volcanics, SE Peru: II. Geochemistry and origin of a felsic peraluminous magma. Contrib. Mineral. Petrol, v. 100, p. 325-338.

65

INDIRECT EVIDENCE FOR A GRANITOID FLUID SOURCE OF ARCHEAN LODE-GOLD DEPOSITS J R Ridley, DI Groves, E J Mikucki and N J McNaughton Key Centre for Strategic Mineral Deposits, The University of Western Australia, Nedlands, WA 6009. Arguments concerning the source of orefluidsof greenstone-belt Archean lode-gold deposits remain unresolved. Recent realisation that this style of deposit can form over a broad spectrum of metamorphic conditions, up to sub-magmatic temperatures (Bamicoat et al., 1991), together with isotopic and geochemical evidence for a source of fluid componentsfromoutside the greenstone belts, has prompted re-evaluation of models for the source of orefluidand solutes. The case for an orthomagmatic, granitoid source for deposits in the Yilgam Block, WA, has been strengthened by: recendy acquired geochemical and isotopic data; more thorough and precise geochronological data on both mineralisation and granitoid emplacement; improvements in the understanding offluid-phaserelations; and the analysis of wallrock alteration signatures. Thefirstclear indication of afluidsource extemal to the greenstone belts camefromPband Sr-isotope data which indicate ore-fluid Pb and Sr more radiogenic than ore-hosting mafic ultramafic greenstone rocks (McNaughton et al., 1992). There is a possible correlation of Srisotope ratios with host-rock metamorphic grade, and hence distance from a granitoid terrain. Lead isotopic signatures, however, show regional (50-100 km) homogeneity of the Pb source to the fluids, and no relationship between isotopic signatures and the nature of the hosting greenstone stratigraphy. There is a strong correlation between the temperature of mineralisation, and the peak temperature of metamorphism in the hostrocks (Witt, 1991), confirming a broadly synmetamorphictimingof mineralisation. A sequence of wallrock alteration styles that reflect the temperature of mineralisation has been recognised,fromcarbonate - sericite ± albite assemblages at low temperatures, to biotite-bearing and skam-like assemblages at deposits in amphibolite-facies terrains (see summary by Groves et al., 1992). These differences in wallrock alteration can be explained as the result of formation at different temperatures from fluids of constant or similar composition with respect to many components, but whose composition has evolved during upward movement through the crust through processes such as: intern^ fluid buffering; precipitation of vein minerals; and fluid-walkock reaction. Fluid compositions at higher-temperature deposits are therefore most likely to reflect the source rock. Data from these deposits, and compositional extrapolation to higher temperatures, shows tiiat thefluidsare likely to have been derived from a quartzo-feldspathic rock, which included a Kbearing phase, and possibly amphibole. The most likely source rock is,tiierefore,granitic. Fluid inclusion studies have shown that gold-bearing ore fluids are mixed H2O-CO2NaCl fluids with a relatively narrow range of proportions of NaCl and CO2, typically 2-8 eq. wt. %, and 0.05-0.25 mole proportion respectively. Recent experimental work on syntiietic fluid inclusions in the H2(>-C02-NaCl system (Kotelnikov & KoteFnikova, 1990; Frantz et al., 1992) shows that thefieldof immiscibility in this ternary fluid system expands with increasing temperature over the range of 500 - 800°C at a few kilobars pressure. Increasing immiscibility with temperature means that many mixedfluidshave a maximum possible temperature of formation. The solvi forfluidsof the compositions recorded in gold-bearing systems would be at a temperature of about 700°C at 3-6 kbar pressure, and hence the observed salinities are those expected if thefluidsresultedfromexsolution from a C02-bearing granitic magma. Exsolution oftiiefluidfroma magma at these conditions would explain the narrow range of compositions observed. IntiieYilgam Block, gold mineralisation is dated at 2630-2640 Ma by the three available Pb-Pb isochron ages. Most granitic activity and associated felsic volcanism in and adjacent to the greenstone belts is dated at between 2660 - 2690 Ma (Hill et al., 1992). There is, however, evidence for continuation of felsic magmatism until about 2610 Ma, with a peak in activity at 2630 Ma clear in the more deeply eroded, higher-metamorphic grade portions oftiieBlock, particularly in the Soutiiem Cross Province. There is thus an overlap betweentiieages of

66

mineralisation and of granite plutonism, particularly of plutons which probably crystallised in the middle and lower crust, in contrast to the older intrusions exposed in the greenstone belts. The new data and analyses show that the nature and composition of the ore fluid is consistent with derivation from a COa-bearing granitic magma that crystallised at pressures of a few kilobars. It is unclear, however, whether there is a specific granite source type. Most potential chemical indicators of an exact source are inconclusive. PotassiumMibidium ratios of alteration minerals imply that the fluid was not derived from a granite that had undergone extreme fractionation. Initial Sr ratios are not strongly radiogenic, showing that a long midupper crustalresidencyfor the source rocks is unlikely. Estimation of fluid oxidation state from alteration assemblages show a range of values of several oiders of magnitude of fo2, from moderately reducing to moderately oxidizing, and consistent with derivation of thefluidfrom mantle-derived igneous rocks, or typical calc-alkaline magmas. It is speculated here however, that the significant control on the generation of a goldbearingfluidmay be through the depth of crystallisation of the granite. The exact controls on gold and sulphide partitioning into a magmaticfluidare unclear. At high pressures, however, CO2 will beretainedin solution in the magma untilrelativelylate in fractionation. Equally, there will be no tendency for the fluid to become increasingly saline during fractionation (cf. Bodnar & Cline, 1990), and this may be important in allowing high contents of weak bases in the fluid. References Bamicoat AC, Fare RJ. Groves DI & McNaughton NJ, 1991. Syn-metamorphic lode-gold deposits in high-grade Archean settings. Geology 19: 921-924. Bodnar RJ & Cline JS, 1990. Microthermometric and phase behavior of magmatic-hydrothennal fluid inclusions: an analysis based on PVTX data for the system albite-H20-NaCl. PACROH III, Toronto, Program and Abstracts, p 17. Frantz JD, Popp RK & Hoering TC, 1992. The composiUonal limits of fluid immiscibility in the system H20-NaCl-C02 as determined with the use of synthetic fluid inclusions in conjunction with mass spectrometry. Chem. Geol. 98: 237-255. Groves DI, Barley ME, Bamicoat AC, Cassidy KF, Fare RJ, Hagemann SG, Ho SE, Hronsky JMA, Mikucki EJ, MueUer AG, McNaughton NJ, Perring CS, Ridley JR & Veamcombe JV, 1992. Sub-greenschist- to granulite-hosted Archaean lode-gold deposits of the Yilgam Craton: a depositional continuum from deep-sourced hydrothermal fluids in crustal-scale plumbing systems. In: The Archaean: Terrains, Processes and MetaUoeeny Geol. Dept (Key Centre) & Univ. Ext, UWA, Publ 23: 325-337. HiU RI, ChappeU BW & Campbell IH, 1992. Late Archaean granites of the southeastern Yilgam Block, Westem Australia: age, geochemistry, and origin. Trans. R. Soc. Edinburgh, Earth Sciences 83: 211-226. Kotel'nikov AR & Kotel'nikova ZA, 1990. Experimental sUidy of phase state in the system H20-C02~NaCl byte synthetic method of fluid inclusions in quartz. Geokhimiya 4: 526-537. McNaughton NJ, Cassidy KF, Dahl N, de Later JR, Golding SD, Groves DI, Ho SE, MueUer AG, Perring CS, Sang JH & Tumer JV, 1992. The source of ore components in lode-gold deposits of the Yilgam Block, Westem Australia. In: The Archaean: Terrains, Processes and Metallogeny. Geol. Dept (Key Centre) & Univ. Ext, UWA, Publ 23: 351-363. Witt WK, 1991. Regional metamorphic conu-ols on alteration associated with gold mineralization in the Eastem Goldflelds Province, Westem Australia: Implications for timing and wigin of Archean lode-cold deposits. Geology 19: 982-985.

67

EVIDENCE OF UNUSUALLY CARBONIC AND REDUCED ORE FLUIDS IN THE LATE PROTEROZOIC SEVENTEEN MILE HILL PORPHYRY COPPER-STYLE DEPOSIT, TELFER DISTRICT, WESTERN AUSTRALIA. S.M. Rowins'^'. D.I. Groves^ N.J. McNaughton"*, P.E. Brown^, R.L. McLeod^, and D. Hal|3 ^ Key Centre for Strategic Mineral Deposits, Department of Geology, The University of Western Australia, Nedlands 6009, Western Australia. 2 Department of Geology and Geophysics, University of Wisconsin, Madison, Wl 53706, U.S.A. 3 Telfer Gold Mine. Newcrest Australia Ltd., Telfer 6762, Western Australia. Intrusion of late- to post-tectonic, highly fractionated, l-type granitoids at about 680 to 620 Ma, into Late Proterozoic, low grade, metasedimentary sequences of the Upper Yeneena Group (-1000-750 Ma), produced various styles of gold and base metal mineralization in the Paterson Province including the large Telfer Au-Cu deposit (e.g., Goellnicht et al., 1989, 1991; Hall, 1989; Dimo, 1990). Like Telfer, the 17 Mile Hill Cu (+Au) deposit is situated in one of several regionally prominent, NW trending, domal structures in the district and a strong structural control exists on the localization of mineralization. Drilling has not intersected an intrusive igneous body, but the presence of a buried porphyry is inferred from (i) geophysical anomalies (Anderson, 1989) (ii) contact metamorphic textures and associated minerals (iii) style(s) of mineralization (iv) ore mineral assemblage, and (v) vertical zonation of hydrothermal alteration. A detailed petrographic, isotopic, and fluid inclusion study has been undertaken to determine the source(s) of the mineralizing fluids and establish timing relationships between the different phases of alteration and mineralization in order to further test the hypothesis that 17 Mile Hill mineralization belongs to the porphyry Cu class of deposit. Comparison of the 17 Mile Hill deposit with epigenetic Telfer mineralization reveals many similarities and supports a genetic link between the two styles of mineralization. Mineralization at 17 Mile Hill is hosted in argillaceous and quartzose siltstones and sandstones of the Isdell Formation. In outcrop, rocks display strong greisenization (quartz+sericite+tourmaline±iron oxides (after sulphides)). Discrete zones of intense silicification associated with quartz veining are also common. Examination of drillcore reveals that contact metamorphism has produced a fine-grained, weak- to moderately-foliated biotitic hornfels with a variety of metamorphic "spots" of which the most common are muscovite±quartz±tourmaline, K-feldspar±albite±quartz±tourmaline, quartz+biotite, and scapolite. The dominant styles of hydrothermal alteration are phyllic (muscovitetquartz) and potassic (biotite+Kfeldparlsericite), although both propylitic (calcite+epidote+chloritetrutilelsericite) and argillic (kaolinite +quartz) occur as well. Potassic alteration is dominant in the deeper levels of the drillcore (below -- 250 m), whereas phyllic and propylitic styles are more abundant at shallower depths. The pervasive near-surface argillic alteration is undoubtably supergene in origin, but at deeper levels (below 150 m), where associated with pyrite, it may be of hypogene origin. Hydrothermal alteration commences with the potassic stage followed by phyllic, although multiple reversals in this chronologic order are indicated by different generations of biotite, K-feldspar, and muscovite. Propylitic alteration is late and best developed at shallower depths, where it overprints contact metamorphic minerals and phyllic alteration; at deeper levels it directly replaces potassic alteration. Argillic alteration is the final replacement stage. Sulphide mineralization mainly consists of pyrite, chalcopyrite, and pyrrhotite with rare sphalerite, galena, marcasite, and wolframite. Typically, these sulphides occur with quartz, K-feldspar, carbonate (siderite and dolomite), and tourmaline in quartz veins ranging from several millimetres up to several metres across. The larger veins are commonly sub-concordant to bedding, whereas the smaller ones are markedly discordant and tend to link the more concordant mineralization to produce a stockwork-style of mineralization similar to that at the Telfer gold mine (Goellnicht, 1987; Anderson, 1989). Semi-massive to massive pyrrhotite+pyrite±chalcopyrite veins, ranging from 1 to 4 m in thickness, are common below 300 m. A potassic alteration halo surrounds most veins, and there is no simple relationship between the size of the vein and the width of the alteration halo. In intensely veined sections of drillcore, alteration envelopes may coalesce to form pervasive alteration. Sulphides occur as fine disseminations, veinlets, and aggregates replacing and/or rimming alteration clots in the potassic alteration halos surrounding mineralized veins. Sulphide mineralization is also associated, to a lesser degree, with phyllic and argillic alteration. Anderson (1989) noted that quartz-sulphide veining increases in intensity with depth down drillhole as does the ratio of pyrrhotite to pyrite and the overall proportion of chalcopyrite. These features, together with the vertical zonation of hydrothermal alteration, suggest that an underlying intrusive body is being approached. Primary and pseudosecondary fluid inclusions from quartz in mineralized veins contain very complex H20-C02-CH4-salt fluids, and a wide range of ratios of aqueous to carbonic phases. Homogenization temperatures range from 175'' up to 6 1 I X . Measurement of freezing point depression in the aqueous phase of fluid inclusions provides evidence of widely varying salinities and complex solute chemistry. Indicated salinities vary from a minimum of 7 wt.% NaCI equiv. to a maximum of 38 wt.% NaCI equiv. in halite-saturated inclusions. Some inclusions contain up to eight daughter minerals, which include halite, sylvite, Fe-bearing calcite, siderite, tourmaline, and possibly dawsonite and Ca- and Fe-chlorides. The aqueous phase of the fluid inclusions has solutes in addition to NaCI, since all require a high degree of

68

supercooling to induce freezing, display initial ice melting at temperatures well below -23X, and form distinctive brown, granular ice. Temperatures of final CO2 melting in the carbonic phase of fluid inclusions indicate abundant CH4, and Laser Raman microprobe analysis confirms that some carbonic phases are 100 mole% CH4. This CH4-rich nature of many inclusions is corroborated by final clathrate melting temperatures in excess of 10X. Evidence of phase-separation ("boiling") is lacking, and the fluid inclusion data are best explained by mixing of hot, high salinity (magmatic) fluids with cooler, lower salinity (basinal) waters, with liberation of CO2 and CH4 from underlying carbonate and carbonaceous rocks during ascent of hydrothermal fluids. Modified isochore projections for aqueous and carbonic fluid inclusions predict entrapment pressures of 2-3 kbars. values of pyrite sampled from quartz-sulphide veins and their alteration selvedges range from -2.5 to 0.2 %o CDT. This restricted range around 0 %o is suggestive of an isotopically unfractionated and homogeneous sulphur source, probably magmatic. S'^^c values of dolomite associated with quartzsulphide veins display a narrow range from -1.7 to -0.1 %o PDB, overlapping the range for marine carbonate from the host Isdell Formation. values of the vein dolomites range between 14.0 and 21.5 %o SMOW, considerably lighter than marine carbonate from the host Isdell Formation. This isotopic shift to lighter values can be explained as resulting from the interaction of water-rich fluids (magmatic) with primary marine carbonate contained in the host-rocks. The lack of light values in vein carbonates could be taken as evidence that isotopic exchange between CO2 and CH4 was kinetically inhibited, such that only carbon contained in CO2, derived from the dissolution of host-rock marine carbonate, was incorporated into the hydrothermal carbonates. Assessment of these data support the contention that 17 Mile Hill is a porphyry Cu-style deposit, although the abundance of CO2 and CH4 in the ore fluids, the high proportion of pyrrhotite in the hypogene ore assemblage, and the lack of primary hematite, magnetite, and sulphate minerals is unusual for this class of deposit. The presence of reduced ore fluids at 17 Miile Hill calls into question whether oxidized ore fluids are as important in producing this style of mineralization and its attendant alteration as is commonly believed (e.g., Hunt, 1991).

References Anderson, H.F. (1989) Integrated geological and geophysical study of the 17 Mile Hill Prospect, Telfer District, Western Australia. BSc. (Hons) Thesis, Univ. West. Aust. (unpubl.). Dimo, G. (1990) Telfer gold deposits. In Hughes, F.E. (ed.) Geology of the Mineral Deposits of Australia and Papua New Guinea, pp. 642-651. The Australian Institute of Mining and Metallurgy, Melbourne. Goellnicht, N.M. (1987) Constraints on the timing and source of gold mineralization at Main Dome, Telfer, Western Australia. BSc. (Hons) Thesis, Univ. West. Aust. (unpubl.). Goellnicht, N.M., Groves, D.I., McNaughton, N.J., and Dimo, G. (1989) An epigenetic origin for the Telfer gold deposit. In Keays, R.R., Ramsay, W.R.H., and Groves, D.I. (eds.) The Geology of Gold Deposits: The Perspective in 1988. Econ. Geol. Mon. 6: 151-167. Goellnicht, N.M., Groves. D.I., and McNaughton, N.J. (1991) Late Proterozoic fractionated granitoids of the mineralized Telfer area. Paterson Province. Western Australia. Precamb. Res. 51:375-391. Hall, D. (1989) Prospect geology and mineralization on the Telfer Joint Venture Area: an update. Newmont Australia Ltd. Rept. (unpubl.). Hunt, J.P. (1991) Porphyry copper deposits. Econ. Geol. Mon. 8: 192-206.

69

GOLD DEPOSITS IN INTRUSION-CENTRED SYSTEMS Richard H. Sillitoe 27 West Hill Park, Highgate Village, London N6 6ND, England

A variety of gold deposit types is generated in intrusion-centred hydrothermal systems developed along volcano-plutonic arcs (e.g. Sillitoe, 1991). These comprise gold-rich porphyry copper deposits as well as recently recognized porphyry gold deposits (Vila and Sillitoe, 1991) in the centres of systems. Poiphyry-type mineralization grades outwards through proximal copper-gold skarns to gold + lead-zinc-silver in distal skams and carbonatereplacement bodies and, on the fringes of some systems, to sediment-hosted (Carlin-type) gold-(arsenic-antimony-mercury) deposits (Sillitoe and Bonham, 1990). Poiphyry deposits may be transitional upwaids to gold-bearing, volcanic-hosted epithermal environments, especially those of high-sulphidation (acid-sulphate) type in which gold + copper (as enargite) is characteristic. Gold may adopt one or more of these proximal through distal positions in any individual zoned system. Furthermore, a gold concentration in one zonal position cannot be used to predict the potential of the rest of the system. Distal and/or shallow gold concenti'ations ai'e favoured by the presence of syn-mineral faults and fractures which impinge on the ore-forming stocks. The presence of permeable calcareous host lithologies also plays a key role in localizing distal gold mineralization as well as apparently contributing to the development of three of the five giant gold-rich porphyry copper deposits. Vertical dimensions of systems may be decreased dramatically by telescoping of the epithermal and poiphyry-type environments, a process also conducive to generation of giant gold accumulations. Factors influencing the favourability of an intrusion-centred system for gold are incompletely understood. Alkalinity and redox state of magmas appear to be two broad parameters of mantle or deep ciustal parentage which are influential. Keith et al. (1991) propose that calc-alkaline magmas possessing relatively low redox states because of the effects of buffering during ascent through reduced crust evolve fluids depleted in chloride and hence base metals. As a consequence, base-metal deposition is suppressed and development of sediment-hosted (Carlin-type) gold deposits may be enhanced, especially if receptive lithologies are available. However, shallow-level processes may also exert powerful controls on gold transport and precipitation because of gold's facility to complex with both chloride and bisulphide in solution in addition to forming volatile complexes at high temperatures. For example, depth of stock emplacement may be an important control on magmatic fluid composition and the timing of its release (Cline and Bodnar, 1991), and may influence the copper vs. gold content of resulting porphyry-type mineralization. At this stage, the gold endowment of an intrusive centi'e is difficult for the explorationist to predict unambiguously from rock type and chemistry alone. However, if the presence of gold can be confirmed by conventional geochemical means, then entire intrusion-centred systems may be explored effectively using what is currently known about gold deposit types and styles and metal (including metal ratio) zoning. Exploration must take account of the lai'ge sizes of intrusion-centred systems: up to 8 km in radius and at least 3 km in vertical extent.

70

References

Cline, J.S. and Bodnar, R.J., 1991. Can economic porphyry copper mineralization be generated by a typical calc-alkaline melt?: Jour. Geophys Research, 96, B 5 : 81138126. Keith, S.B., Laux, D.P., Maughan, J., Schwab, K., Ruff, S., Swan, M.M., Abbott, E. and Friberg, S., 1991. Magma series and metallogeny: A case study from Nevada and environs, jn Geology and Ore Deposits of the Great Basin. Field Trip Guidebook and Compendium, Vol. 1 (eds. R.H. Buffa and A.R. Coyner), pp. 404-493 (Geol. Soc. Nevada: Reno). Sillitoe, R.H., 1991. InU usion-related gold deposits, ni Metallogeny and Exploration of Gold (ed. R.P. Foster), pp. 165-209 (Blackie: Glasgow). Sillitoe, R.H. and Bonham, H.F., Jr., 1990. Sediment-hosted gold deposits: Distal products of magmatic-hydrothermal systems: Geology, 18: 157-161. Villa, T. and Sillitoe, R.H., 1991. Gold-rich porphyry systems in the Maricunga belt, northern Chile: Econ. Geol., 86: 1238-1260.

71

THE WEST PEKO AU-CU-BI DEPOSIT, TENNANT CREEK, NT: FLUID INCLUSION CONSTRAINTS ON ORE FLUID CHEMISTRY. Roger G. Skirrow* & John L Walshe Department of Geology, Australian National University GPO Box 4, Canberra, ACT 2601 The Tennant Creek Au-Cu-Bi deposits occur within and adjacent to epigenetic massive magnetite-hematite bodies ("ironstones'), hosted by Lower Proterozoic greywacke-siltstone-shale strata of the Warramunga Group. Ironstones developed mainly in ESE-trending, steeply dipping, reverse fault and shear zones that were active during D1 deformation of the Warramunga Group. The mineralisation clearly postdates the ironstones (Wedekind et al., 1989) and was superimposed on them during late D1 or D2 reactivation of D1 shear and fault zones (Wall & Valenta, 1990; Edwards, 1990; this study). The West Peko deposit occurs at a subsurface depth of 400-700m and is at present unexploited. It is Cu- and sulphide-rich with Au grades of up to 10 g/t (which are relatively low for Tennant Creek deposits), contains alteration minerals indicative of reduced conditions (pyrrhotite, arsenopyrite, minnesotaite, greenalite, ilmenite), and has a narrow range of (chalcopyrite, pyrite, pyrrhotite) near 2 per mil. The reduced alteration assemblage is zoned both within the ironstone around the high grade Cu-Au-Bi zone and in the sedimentary wall rocks adjacent to a 'feeder* zone. This assemblage was produced by hydrothermal alteration of pre-existing ironstone and wall rocks during syn-deformational ore formation, and there is no evidence that it is of contact metamorphic origin. Fluid inclusions of primary, pseudosecondary and secondary origin in quartz and sphalerite associated with various hydrothermal alteration assemblages through the paragenetic sequence have been investigated using microthermometry and laser Raman spectroscopy (LRS). An aqueous fluid of low to moderate salinity (3-10 eq. wt. % NaCI), 300-350°C and N2-CH4-bearing is inferred to have carried Au-Bi+Cu. In rare cases the low salinity fluid inclusions contain native Bi+Au 'accidentally' trapped solids, indicating the contemporaneity of these fluids with Bi-Au deposition. Vapour-rich fluid inclusions occur in association with the low salinity inclusions and in trails of only vapour-rich inclusions. LRS analysis and microthermometry indicate that the vapourrich fluid inclusions are composed of 45-76 mole % N2, 55-24 mole % CH4; water may also be present in small amounts but was undetectable optically in most vapourrich inclusions. The textural relationships and estimates of fluid density are consistent with an origin of the vapour-rich N2-CH4 inclusions by phase separation from the low salinity fluid at pressures of 1000-1500 bars. A highly saline Ca-Na-CI, in places halite-saturated fluid with liquid-vapour homogenisation temperatures of 130-160°C ('pressure correction' unknown) is also present throughout the West Peko deposit and its wall rocks, as well as in other parts of the field. Textural relationships indicate that this brine was present prior to, after and probably during ore deposition, and it may represent a regionally distributed basinal brine. A fluid of similar composition (11-26 eq. wt. % NaCI, Ca-rich) was trapped in quartz during earlier ironstone growth at 300-350°C, as estimated from quartz-magnetite 0-isotope and chlorite geothermometry. The microthermometric and textural data suggest that mixing occurred between the low-moderate salinity, N2-CH4-bearing aqueous ore fluid and the calcic brine. N2-CH4 immiscibility may have occurred in response to this mixing, as the solubility of N2 and CH4 would have been far lower in the more saline mixed fluid

72

('salting-out' effect). The implications of such fluid mixing for ore deposition and its role in producing the variations in alteration mineralogy across the Tennant Creek goldfield are the subject of continuing research, along with mass transfer modelling of fluid-fluid and fluid-rock reaction and stable isotope studies.

References Edwards, G.C.. Booth, S.A. & Cozens, G.J., 1990. White Devil gold deposit. In: Hughes, F.E. (ed.). Geology of mineral deposits in Australia and Papua New Guinea. Aust. Inst Mining Metall. Mon. 14: 849-856. Wall, V.J. & Valenta, R.K., 1990. Ironstone-related gold-copper mineralisation: Tennant Creek and elsewhere. In: Proceedings of the Pacific Rim 90 Congress, Vol. Ill, 855-864 (AIMM, Melbourne). Wedekind, M.R., Large, R.R. & Williams, B.T., 1989. Controls on high grade gold mineralisation at Tennant Creek, Northern Territory, Australia, Econ. Geo!. Mon. 6:168-179.

73

THE EFFECTS OF DEFORMATION AND ALTERATION ON METAL ZONING PATTERNS IN THE HILTON MINE, MOUNT ISA

R.K. Valenta and A. Wilson VIEPS, Department of Earth Sciences, Monash University, Clayton, Victoria 3168

Hilton is a giant, deformed, sediment hosted zinc-lead-silver-copper deposit which is located approximately 20 km to the north of Mount Isa. Most of the underground mining to date has concentrated on an area where the orebodies which make up the deposit are relatively thick, lead-rich, silver-rich and copper-rich. Possible explanations for this pattern include: 1) a primary zoning pattern related to the formation of the deposit: ii) a zoning pattern formed by solid-state remobilization associated with deformation; or iii) a zoning pattern related to late stage hydrothermal alteration. To test these hypotheses, we carried out a detailed study of metal zoning, macroscopic to mesoscopic deformation features, and alteration patterns in a representative section of the Hilton deposit. This study revealed the following features: a)

Upper levels of the deposit are relatively zinc-rich, whereas lower levels show an enrichment in lead, silver and copper, as well as an increase in combined metal values.

b)

There is no obvious correlation between the development of mesoscopic deformation features and metal ratios. Lead-rich and lead-poor ores show strong recrystallization and grain growth, as well as abundant brecciation and recrystallization of associated gangue minerals and host rocks. This suggests that the present zoning pattern is not simply related to large scale strain variations.

c)

Pyrite which is paragenetically associated with lead-zinc mineralization is at least partly overprinted by the early bedding-parallel foliation. This, and obvious mesoscale deformation features within the orebody suggests that the earliest lead-zinc mineralization predates deformation.

d)

The lead-silver-copper rich portion of the deposit shows a strong spatial correlation with locally penetrative chlorite-sericite and ankerite alteration. Detailed logging and magnetic susceptibility measurements also show that the high grade zone is strongly enriched in pyrrhotite relative to low grade areas. This pyrrhotite does not generally replace fine grained pyrite. Relationships of phyllosilicate alteration and ankerite alteration to cleavage suggest that these alteration types formed during deformation.

Based on these observations, it would appear that the present metal zoning pattern developed at Hilton is related to syndeformational hydrothermal alteration which overprinted pre-existing lower grade stratiform orebodies. This is not surprising, since pre-existing lead-zinc orebodies show a strong contrast in both rheology and chemistry to the surrounding rocks.

74

ORIGIN OF BARITE FROM THE HELLYER VHMS DEPOSIT, TASMANIA: A Sr ISOTOPIC STUDY

David J. Whitford\ Robina Sharpe^ and J. Bruce Gemmell^ ^CSIRO Division of Exploration Geoscience, North Ryde NSW 2113 ^ CODES, University of Tasmania, Hobart TAS Enrichment in Sr, resistance to secondary alteration, and widespread occurrence in volcanogenic massive sulfide (VMS) deposits makes barite ideally suited for tracing the origin of ore-forming fluids using Sr isotopes (Whitford et al., 1992). The low Rb/Sr in barite means ®^Sr/®®Sr ratios are likely to reflect directly its primary isotopic composition, unaffected by secondary processes. The Hellyer VMS deposit occurs in the Cambrian Mount Read Volcanics of western Tasmania and shows many of the characteristics of submarine exhalative syngenetic mineralization. Approximately 17 million tonnes of high-grade massive Pb- Zn-Cu sulfides occur in a sequence of basaltic to andesitic lavas and fragmental rocks that have undergone weak deformation and regional prehnite-pumpellyite fades metamorphism during the Devonian. The deposit is underlain by a pronounced zone of stringer mineralization, including minor barite, with characteristic siliceous, chloritic and sericitic alteration (Gemmell and Large, 1992). Overlying the massive sulfides is a barite cap up to 15 m thick which is also enriched in precious metals. ®^Sr/®®Sr ratios in barites from Hellyer range from 0.7099-0.7114. Overall, there is little consistent spatial variation in ®^Sr/®®Sr ratios within the barite blanket. However there is a suggestion of a general decrease in ®^Sr/®^Sr ratio from the bottom to the top of the blanket. The relatively high ®^Sr/®^Sr ratios measured in the barites confirm the unusual Sr Isotopic character of mineralization at Hellyer. Such high ratios are also observed in barites associated with mineralization at Que River, Rosebery and Mount Lyell (Whitford et a!., 1992). The measured ®^Sr/®^Sr ratios are higher than both the initial ®^Sr/®®Sr ratios (0.7065-0.708) measured in the local basaltic and andesitic host rocks and those inferred for Cambrian seawater (ca.0.709). Assuming that the measured ®^Sr/®®Sr ratios reflect the primary isotopic compositions at the time of mineralization, the Sr in the barites must be derived, at least in part, from a source other than the local volcanic rocks and Cambrian seawater. The most likely source is old continental crust underlying the Mount Read Volcanics. Although it is difficult to constrain unambiguously the relative proportions of Sr derived from the crustal component as opposed to the local volcanic rocks and seawater, the recognition of an old crustal source has implications for understanding both the origin of the ore metals and the nature of fluid-flow pathways. If the underlying crustal component could be shown to be a significant source of ore metals, there could be important exploration implications. References Gemmell, J.B. and Large, R.R. 1992. Economic Geology 87, 620-649. Whitford, D.J., Korsch, MJ. and Solomon. M. 1992. Economic Geology 87, 953-959.

75

MARAMUNGEE: A PROTEROZOIC ZN "SKARN" IN THE CLONCURRY METAMORPHIC TERRAIN. IS THERE A MAGMATIC CONNECTION? Patrick J. Williams^ * and Maree Heinemann^ ^ Department of Geology and Key Centre in Economic Geology, James Cook University of North Queensland, Townsville, Q4811 2 Now at Placer Exploration, GPO Box 558, Brisbane Q4000 The significant but subeconomic (1.8 million tonnes @ 4.4% Zn) Maramungee deposit is 90km SSE of Cloncurry in northwest Queensland. It lies in the Eastern Fold Belt of the Proterozoic Mount Isa Inlier within rocks of the Maronan Supergroup that also hosts the large Pb-Zn-Ag deposits at Cannington and Pegmont. Maramungee has striking similarities to orthodox skarns particularly with respect to: a) b) c)

Alteration and mineralisation localised in an embayment in the margin of a felsic pluton. Ca-Fe-Mn-Mg-rich silicate alteration assemblages developed in both the pluton and its host rocks. Paragenetic sequence involving anhydrous garnet-hedenbergite (apatite) skarn overprinted by retrograde hydrous amphibole-chlorite alteration and sulphides.

On close inspection however, its seems likely that pluton emplacement, anhydrous skarnoid alteration, and sulphide mineralisation are separate elements in a history of structural reactivation. The deposit is hosted by upper amphibolite (sillimanite-K-feldspar zone) gneisses, exceptionally iron-rich metabasites, and foliated granite. The pluton belongs to a suite of deformed l-type (magnetite-bearing) trondhjemitic granitoids that are restricted to part of the NE Selwyn Range area and reflect a distinctive magma type that is not known anywhere else in the Mount Isa Inlier. The granites were deformed under amphibolite fades conditions in association with tight upright to overturned fold structures that accord broadly to the geometries of structures formed during the regional D2 event. The elongate embayment in the margin of the Maramungee Granite that contains the deposit is formed by part of a parasitic F2 antiformal fold hinge which adjoins a prominent photolinear feature developed in both the granite and its host rocks. The latter structure initiated by fabric intensification during attenuation of the adjacent major fold limb. Mineralisation is further localised within a heterogeneous altered rock package including 1) potassic rocks (generally foliated) with large amounts of microcline and/or biotite; 2) skarns (partly discordant) including both manganoan hedenbergite + Fe-Mn-rich grossularite + quartz + apatite rocks and Ca-rich almandine ± clinopyroxene-bearing metabasites; and 3) a range of compositionally-

76

intermediate assemblages distinguished by the co-existence of Fe/Mn, Ca- and Kbearing metamorphic silicates. Graphite commonly occurs at levels of several modal percent but there is no magnetite or zincian spinel associated with the alteration and mineralisation. The most strongly altered rocks have a chemical component derived from the metasedimentary gneisses and probably reflect alteration of a unit which previously contained carbonates. This alteration would have involved reduction producing graphite coupled with differential addition of K, Fe, Mn, P, and F. The potassic and skarnoid rocks are distinctive components of a regional pattern of metasomatism that exploited the steeply-inclined D2 structures which have deformed the trondhjemltic granites. Early metasomatic fluids in equilibrium with graphite at Maramungee are unlikely to have been evolved from the magnetitebearing pluton and were therefore probably metamorphic in origin. However, the granite may well have played a significant mechanical role in development of the fluid channelways. Sulphides (mainly sphalerite, pyrrhotite and pyrite) occur in veinlets and microbreccia matrices that overprint all the earlier rock types and alteration assemblages. They are associated with the development of retrograde (hydrous) and comparatively oxidised secondary minerals including CI and Fe-rich amphiboles, chlorite, scapolite, calcite, albite, epidote, and muscovite. Hematite is present in the youngest alteration assemblages. Retrograde phase relationships suggest that sulphide deposition was induced by reduction of the infiltrating orestage fluid, initially occurred at temperatures of 450-500°C, and may have continued through a significant cooling interval. The skarnoid alteration that hosts the mineralisation at Maramungee could be substantially younger that the pluton with which it has a close spatial relationship. However, the comparatively oxidised ore fluid may have been evolved during emplacement and cooling of the late to post tectonic granites of the Williams Batholith (nearest surface expression 6km away). There is a growing body of evidence that the important Cu-Au deposits in the district post-date the emplacement of these younger plutons and that the ore fluids could have had a significant magmatic component. These fluids were highly saline and may well have transported large amounts of Pb, Zn and Ag which could have been concentrated in other environments with different physico-chemical conditions.

77

STABLE ISOTOPE EVIDENCE FOR TWO SOURCES OF MINERALISATION AT THE DABAOSHAN POLYMETALLIC DEPOSIT, SOUTHEASTERN CHINA

Jiang Zhiyu' and Chen Minyang (Research Institute of Geology for Mineral Resources, CNNC, China) ^ Present address: Department of Geology, The University of Newcastle

The Dabaoshan deposit is a large polymetallic deposit containing Mo, Cu, Pb. Zn and Fe. It is situated in the southeastern rim of the Quxian Basin and lies astride the intersection of the E-W trending Dadongshan-Guidong structural belt and the NE-trending Bejiang deep fault belt in the eastern part of the Hercynian Xiangguiyue Belt on the South China Platform. In the Dabaoshan district, the stratigraphic units exposed are low-grade metamorphic rocks comprising sandstone and shale of the Cambrian Bacun Group, interbedded sandstone, conglomerate and shale of the Early-Middle Devonian Guitou Group, limestone, marl and calcshale of the Middle Devonian Donggangling Formation, limestone and marl of the Late Devonian Tianziling Formation, sandstone, shale and intercalated limestone of the Late Devonian Maozifeng Formation and sandstone and shale of the Early Jurassic Lantang Group. During the Mesozoic era, strong intrusive activity in the district resulted in the emplacement of granodiorite porphyry, sub-dacitic porphyry, granodiorite and granite. Mo mineralisation in the Dabaoshan deposit is hosted by the granodiorite porphyry and the sub-dacitic porphyry. Cu and minor W, Pb and Zn mineralisation is hosted by skarn formed at the contact of the granodiorite, granodiorite porphyry and sub-dacitic porphyry with limestone of the Donggangling and Tianziling Formations. Stratabound siderite ore bodies and major Pb and Zn mineralisation occur within the same sedimentary units. Previous K-Ar dating of two whole rock samples from the Dabaoshan granodiorite porphyry returned dates of 97 Ma and 101 Ma respectively. These may represent minimum ages because of weak alteration resulting in Ar loss. Recent K-Ar dating of biotite from a fresh subdacitic porphyry returned an age of 143 Ma. Pb-isotope ratios were determined on 27 ore sulphides (10 samples of galena, 6 pyrite, 5 sphalerite, 4 pyrrhotite and 2 chalcopyrite) from the deposit and 2 sulphides (sphalerite and pyrite) from two anomalies in the nearby Dabaoshan deposit. Additionally, 33 analyses (4 pyrite and 29 whole rock samples) represent Donggangling and Tianziling sedimentary host rocks. Data from these two formations have typical Pb-isotope features of sedimentary rocks. They give a very narrow range of '^'Pb/'^'Pb ratios (15.485-15.817), a small range of 208pb/2^Pb (37.417-38.995), a wider range of (17.610-19.761) and lie on an almost horizontal line on a vs diagram. By comparison, the data for the ore sulphides from the deposit are characterised by a wide range of ratios (15.49116.215) and '^^Pb/'^'Pb ratios (37.990-56.39), although the range for '^^Pb/'^^Pb is large like those samples from the host strata (17.930-21.480). The data for 10 galena samples from 27 ore sulphides, however, are very consistent. Pyrite, sphalerite, pyrrhotite and chalcopyrite contain relatively more radiogenic Pb and give higher ratios compared to galena. When corrected for radiogenic Pb, these data are clearly distributed along two lines on a vs diagram (Fig. 1). Line (1) is the isochron for 143 Ma, similar to that obtained from K-Ar dating of biotite from the sub-dacitic porphyry. Line (2) is the distribution line for the data from samples of host rock in the two formations. We conclude that some Pb has a magmatic origin associated with the sub-dacitic porphyry, granodiorite porphyry or granodiorite intrusion and that some Pb also has come from a source of disseminated Pb in the sedimentary host rocks. This mixed Pb signature defines a distinct halo-like distribution around the granodiorite porphyry, sub-dacitic porphyry and granodiorite. From the magmatic 78

rocks outward, the ratios of and increase gradually, whereas the ratios of decrease gradually. They illustrate the close spatial and temporal relationship between the deposit and the intrusions. The study indicates that the source of the metals in the different styles of mineralisation can be grouped into three categories, where 1) all of the ore-forming metal elements for the ore bodies hosted by granodiorite porphyry and sub-dacitic porphyry are apparently derived from a magmatic source, 2) most of the ore-forming metal elements for the ore bodies hosted by skarn are derived from a magmatic source, however 3) only some of the ore-forming metal elements for the ore bodies hosted by limestone are derived from a magmatic source, whereas others have a remobilised source from the host sedimentary rocks. A S-isotope study was undertaken on 166 sulphides representing molybdenite (5 samples), chalcopyrite (9), pyrrhotite (13), pyrite (96), sphalerite (25) and galena (9) from the various kinds of mineralisation, 4 disseminated pyrites from host rocks of the Middle Devonian Donggangling Formation and 5 from the Late Devonian Tianziling Formation. values of ore sulphides from various kinds of mineralisation are restricted to a very narrow range (-2.4 to 4.5%o; Fig. 2). However the distributions for pyrite from both host rocks of the Middle Devonian Donggangling Formation and the Late Devonian Tianziling Formation are different to values obtained for the ore sulphides. The values of all of the pyrite from the Donggangling Formation possess large negative values ranging from -5.3 to -22.5%o. On the other hand, the values from the Tianziling Formation give large positive values (8.7 to 1 7 . 9 % o ) due to the change in the sedimentary conditions (Fig. 2). Analysis of the trend in b^'^S values for pyrite from the different styles of mineralisation indicates that the values are generally close to zero per mil where Mo mineralisation is developed in granodiorite porphyry, sub-dacitic porphyry or for Cu, W, Pb and Zn skarn mineralisation, but positive values occur in outcrops of Tianziling Formation and negative values are related to outcrops of Donggangling Formation. Like the conclusion developed for the Pb-isotope data, it is suggested that S of magmatic origin was mixed with some S from a sedimentary host rock source when the ore fluid transported and deposited ore sulphides in the host rock. The S from the magmatic events appears to be dominant in all varieties of mineralisation, whereas the S in the ore sulphides derived from the sedimentary host rocks is less than 22%o generally. The of hydrothermal S from magmatic events is near -0.1 %o. The temperature of sulphide crystallisation ranged from 490-135°C, as calculated from the values of for coexisting sulphide pairs. Sulphides from two anomalies located close to the Dabaoshan deposit were also analysed. values from the Raogukeng prospect are similar to the values obtained from Dabaoshan (range 0.1 to 5.3%o). b^'^S values for the Wushi prospect (range 12.0 to 12.6%o) are different to the Dabaoshan ore, but are similar to those for sulphides from Tianziling host rocks. Hydrogen isotope data were also obtained from fluid inclusions contained within a sample of chalcopyrite, 4 samples of sphalerite and galena from skarn Cu, skarn Pb and Zn, and stratified Pb and Zn mineralisation hosted by limestone. The values of 5D decrease over a small interval from skarn Cu ( - 1 0 1 to - 1 0 3 % o ) to skarn Pb and Zn ores ( - 1 1 0 to - 1 1 5 % o ) , to stratified Pb and Zn mineralisation ( - 1 2 3 % o ) . Oxygen isotope data of the hydrothermal fluids were reported by Lu Gouqun et al. (1985), calculated on the basis of oxygen isotope data for quartz, biotite and sericite and fluid inclusion homogenisation temperatures. On a 5D vs 6'®0 diagram, meteoric water dominates the hydrothermal system, apparently driven by magmatic activity during ore formation. The 'oxygen shift' and the change in 5D values of fluid inclusions result from oxygen exchange between meteoric water and both magmatic and sedimentary rocks (Fig. 3). Based on the Pb- and S-isotope data, the Raogukeng anomaly appears more prospective than the Wushi area for further exploration.

79

16

18

20

206pb/204pb FIG. 1. Pb-isotope data. Line (1) isochron corresponding to 143 Ma, Line (2) data for whole rock samples from Devonian sediments. 40-

0

2

4

6

8

10

12

14

16

18

FIG. 2. S-isotope data: A. Dabaoshan deposit. B. Raogukeng and Wushi prospects. C. Sedimentary pyrite (negative values - Donggangling Fm, positive values - Tianziling Fm).

0

10

5l8o%o FIG. 3. Calculated H- and 0-isotope composition of ore fluids and water-rock interaction path. 80

ZONING OF THE FENGSAN DONG SKARN COPPER-MOLYBDENUM DEPOSIT, CHINA:SHRINK OR PROGRADE? Taihe Zhou National Key Centre in Economic Geology and Department of Geology, James Cook University, Australia

Abstract The Fengsan Dong deposit is the main copper-molybdenum deposit of the Feng Shan copper mining district, and is located in the western Yangzi River Cu-Fe-(Mo) Mineralisation Belt, which is one of the main copper and iron producing regions in China. The Feng Shan granodiorite is an I-type intrusive, and was emplaced into Middle Triassic dolomitic limestone during the Yanshan period, producing a z o n ^ skam system between the intrusion and the limestone. Outwards from the granodiorite, the following six skam zones are recognised: 1. plagioclase-diopside-grossular endoskam zone; 2. red andradite garnet zone; 3. green andradite gamet-diopside zone; 4. diopside zone; 5. wollastonite marble zone; and 6. serpentinite marble. Mineralisation is also zoned outwards from the intrusion, and appears to be superimposed on the skarn zones: 1. molybdenite zone; 2. magnetite-pyrite-chalcopyrite zone; 3. (molybdenite)-chalcopyrite-bomite zone; and 4. sphalerite-galena zone. Sharp replacement fronts between skam zones are reflected in relatively constant solid solution compositions of minerals within skam zones, but apparent different compositions of minerals between zones. The sharp fronts separating each zone indicate an infiltration dominated contact metasomatic zonation. Remnant replacement textures in skam minerals suggest that for each adjacent skam zone, the outside zone formed earlier than the inside zone, which is different from the conventional theory of outwards prograde skam zonation. Detailed geochemistry, thermodynamic modelling, and fluid inclusion work further support a model of '•shrink" skam zonation. In contrast to the skam zonation, the mineralisation zonation superimposed on skam zones does not show '•shrink- characteristics, and seems to be controlled by variations in the iron contents in the skam minerals, pH, and temperature.

81

LATE ABSTRACTS

HALOGEN (F, CI) CONTENT OF BIOTITES FROM I- AND S-TYPE GRANITES OF EASTERN AUSTRALIA: POSSIBLE MET ALLOGENIC IMPLICATIONS. P. L. Blevin Department of Geology, Australian National University, Canberra, ACT, 2601, Australia Fluorine and CI are important elements within both granite (sensu lato) magmatic and hydrothermal systems. Fluorine influences melt structure and may significantly depress the granite solidus. Chlorine is important in the complexing and transport of many ore elements in magmatic-hydrothermal mineralizing systems, and high CI concentrations in magmatic and hydrothermal minerals have been suggested as a possible exploration tool (Stollery et. aL, 1971). While the controls on CI concentrations in magmas and its partitioning between melts, minerals and aqueous fluids are currently the topic of muchresearch,the possibility that F and CI abundances vary among different granite types has not been assessed. In this study biotite has been used as a sensor of F and CI enrichment in granite magmatic systems from I- and S-type granites of eastern Australia. A set of 400 biotite analyses was collected for this survey. I-type granite biotites show a considerable range in F and CI content compared to S-type biotites (Fig 1). Chlorine abundances are significandy higher in I-type than in S-type biotites (Fig 1). Many biotite samples from S-type suites contain less than Ae detection limit of 0.01 wt% CI used in this study.

9 (4-4

1

d

0

0.00 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 F content (a.f.u.) Fig 1. F vs CI contents of biotites from S- and I-type granites of eastern Australia. Halogen contents are expressed as atoms per formula unit of 22 oxygens (i.e. 44 positive charges) Fluorine abundances in biotites (corrected for Fe avoidance) increase smoothly with increasing whole-rock Rb values which increase systematically with feldspar fractionation (Blevin and Chappell, 1992; Chappell and White, 1992). Chlorine behaviour is different to that of F in that CI contents are not functions of magmatic differentiation and appear to be related to the granite source material. In contrast to I-type granite biotites, CI values in S-type granite biotites vary significandy from suite to suite. The Bundarra (New England) and Bullenbalong (Lachlan Fold Belt) Supersuite samples have high relative CI abundances compared to the Cooma Supersuite (LFB) biotites. The Koetong Supersuite (LFB) samples fall generally between these two groups. The Cooma Supersuite samples are notably poorer in CI with respect to all other S-type granite samples. Chemical and isotopic data suggest tiiat the sequence of S-type supersuites from Bundarra, 83

Bullenbalong to Cooma represent granite magmas derived from increasingly mature supracrustal sources. The decrease in CI content among the biotites of these supersuites correlates with this sequence and suggest that the CI contents of these magmas (as inferred from biotite compositions) are a function of the maturity of the source materials. During vapour-absent melting of I-type source materials, water and volatiles are primarily provided to the magma by the breakdown of amphibole and biotite (White and Chappell, 1983). Water and volatiles in S-typt magmas are provided by biotite, while muscovite will be present in addition to biotite within more mature (metapelitic) sedimentary source materials. Typical CI contents of muscovite are much less than those of biotite and amphibole. Granite magmas whose volatile inventory is derived at least in partfrommuscovite breakdown would have higher F/Cl and lower Cl/OH ratios than those derived from exclusively amphiboleand/or biotite-bearing source materials. The most likely explanation for the variation in CI enrichment between I- and S-type granites is that CI is lost in addition to Ca, Na, Sr, etc during the supracrustal weathering stage in the formation of S-type protoliths. This mechanism also explains the correlation between decreasing CI enrichment in S-type biotites with the increasing maturity (and inferred presence of muscovite) of their source materials as inferredfromchemical and isotopic criteria. Metallogenic implications The S-type ^nites of eastern Australia show a remarkable paucity in nGdneralisation except for the association of Sn and W with highlyfractionatedmembers of some suites. Many oreforming elements are complexed by CI and the efficiency by which they are partitioned into hydrothermal fluids is in part dependent on the CI content of the system (e.g. Candela, 1989). The relatively low CI abundances observed in S-type biotites compared to I-type biotites suggest that magmatic concentrations of Ci are also probably low in S-type magmas and this may go some of the way in explaining the observed metallogenic association of "chlorophile" ore elements with I-type granites rather than with S-types. References Blevin, P.L. & Chappell, B.W. 1992. Trans. R, Soc. Edinburgh: Earth Sciences 83:305316. Candela, P. A. 1989. Rev, Econ. Geol. 4: 223-233, Chappell, B.W. & White, A.J.R. 1992. Trans. R. Soc. Edinburgh: Earth Sciences 83:1-26. Stollery, G., Borcsik, M. & Holland, H.D. 1971. Econ. Geol. 66:361-367. White, A. J. R. & Chappell, B. W. 1983. Geol. Soc. Amer. Memoir 159:21-34.

84

Structurally-Controlled Fluid Pathways, Fluid Mixing and Gold Precipitation in Intrusive-Related Hydrothermal Systems

S.K. Matthai. R.W. Henley (Both at: Research School of Earth Sciences, The Australian National University, GPO Box 4, Canberra ACT 2601, Australia Internet: SKM152CSC.ANU.EDU.AU) Gold quartz vein deposits in the Proterozoic Pine Creek Inlier (northern Australia), have formed at high temperatures (> 5(X)®C) in a hydrothermal system associated with the emplacement of 1830 Ma granitoids. Gold is localized within quartz veins that occupy fracture arrays in major antiforms. Although the development of the fracture arrays was initiated during bedding-parallel flexural slip that accompanied regional deformation, gold deposition and associated vein growth occurred after folding and involved incremental crackseal processes. Proximity to carbonaceous metasediments has been a key factor localizing gold deposition in the quartz vein systems. Stable isotope data and hydrothermal alteration styles indicate that the fluids responsible for Au-transport have been sourced from metamorphic devolatilization reactions within the contact aureoles. Narrow zones of pervasive quartz-veining and disseminated cordierite indicate that the fluids were expelled from the inner aureole via structurally-controlled pathways (SCP's) which we have modelled in cross section, using the Monte Carlo method. SCFs through the predominantly non-carbonaceous metasediments result in a relatively oxidized fluid (CO2 » CH4). Fluid from SCP's that cross carbonaceous slates is rock-buffered to CH4dominated compositions. In our cross-sectional flow models the distinct SCP's join in structures similar to those that host the gold deposits. Hence fluid mixing is held responsible for the gold deposition. This is supported by thermochemical data and best explains the observed gold distribution. Calculations show that such mixing precipitates gold very effectively because it is accompanied by large changes in the hydrogen fugacity of the fluid.

85

GEOLOGY AND GEOPHYSICS OF THE LISLE-GOLCONDA GOLDFIELD, NORTHEAST TASMANIA M.J. Roach Centre for Ore Deposit and Exploration Studies, The University of Tasmania The Lisle-Golconda goldfield is situated approximately 30 km northeast of Launceston in northeastern Tasmania. Alluvial gold was discovered in 1879 in the Lisle valley and subsequently at a number of nearby locations. The Lisle-Golconda goldfield comprises a number of historical mining areas, including the Lisle, Cradle Creek, Lone Star, Golconda, Panama, Denison and Lebrina goldfields. The total production from the region was approximately 10 tonnes of gold, of which over 95% came from alluvial workings, mainly from the Lisle valley. Despite intensive prospecting prior to 1900 and recent exploration, no hardrock source for the alluvial gold has been found. Many of the known sites of gold mineralisation have a close spatial association with small cupolas of hornblende granodiorite which have intruded the regionally metamorphosed Mathinna beds, a thick sequence of Ordovician to Devonian turbidites. The cupolas occur approximately 10 km to the west of the western margin of the Devonian Scottsdale Batholith. The granodiorite is typically highly weathered, and the intrusions are marked by pronounced topographic depressions. Regional gravity and magnetic data was used to provide constraints on the gross structure of the goldfield. In particular, the form and distribution of granitic rocks beneath the Mathinna beds and the spatial relationship between granodiorite and gold mineralisation. Gravity and magnetic data suggests that the granodiorite cupolas in the Lisle area are separate intrusions, distinct from the bulk of the Scottsdale Batholith. The presence of additional subsurface granitic bodies is also inferred, most notably directly beneath the Denison goldfield. The geophysical data also indicates that the granodiorite cupolas are themselves composite bodies, consisting of both strongly magnetic and effectively non-magnetic phases. Whole rock chemical analyses of granitic rocks from the Lisle area and from the western margin of the Scottsdale batholith support the geophysical conclusions. The Lisle intrusives are both geophysically and geochemically distinct from the granodiorites of the Scottsdale batholith. The alluvial gold mined at Lisle had a number of unusual features: - fine grainsize, generally less than 0.4 mm - nuggets were rare. - gold was rarely found with attached vein quartz - the gold had a high fineness (>950) - gold was concentrated in wash immediately overlying the granodiorite surface, within sediments with a high organic carbon content or sediments

86

with manganese oxide staining. These observations suggest that much of the gold may be secondary in origin. Sections through gold grains from the Lisle valley reveal complex internal structures suggestive of in situ growth within the placer from ground waters containing elevated levels of gold in solution. The origin of the gold is still conjectural. Neither the granodiorite or the Mathinna beds have significantly elevated gold contents (<3 ppb) and there are few gold bearing quartz veins within the Mathinna beds or the Lisle granodiorite. It is likely that the gold was leached from a large mass of both Mathinna beds and granodiorite and reprecipitated in favourable geochemical sites within the placer. The topographic depressions associated with the granodiorite cupolas may have played a crucial role in enabling the leaching of a large mass of rock and providing the right environment for the reprecipitation of gold.

87

Index of Authors

A. Andrew A.L Bainbridge R.A. Binns P.L Blevin M.S. Bloom S. Bodon H. Bresser G. Broadbent P.E. Brown G.D. Carman G.R. Carr 1. Cartwright P.LF. Collins D.R. Cooke G.J. Corbett G.J. Davidson J.A. Dean G-Y. Dong J.C. Eames B. Gemmell N.M. Goellnicht D.I. Groves D. Hall G.C. Hall T. Han I.M. Hart M. Heinemann C.A. Heinrich R.W. Henley D.L. Huston P. Kitto J.T. Knight R.R. Large K.C. Lawrie T.M. Leach B.S.E. Mapani S.K. Matthai K.R. McClay S. McKnight R.L. McLeod N.J. McNaughton D.C. McPhail K.G. McQueen T.P. Mernagh E.J. Mikucki C. Minyang G. Morrison

1 3 5 83 16 8 10 11 68 12 14 16 24 16 39 18 14 20 5 75 22, 48 22. 36. 46. 55. 66. 68 68 55 24 26 76 28 85 30. 32 34 36 30 38 3, 39 41 85 43 11 68 22. 46. 48. 66. 68 50 52 28. 53 66 78 20

G.W. Morrison R. Myers R. Napier V.J. Ojala J. Parr H.L. Paterson C. Perkins W.G. Perkins N. Phillips P.J. Pollard M. Power J.R. Ridley M.J. Roach S.M. Rowins C.G. Ryan R. Sharpe S. Sheppard S.H. Sie R.H. Sillitoe R.G. Skirrow J. Stacey D. Suppel G.F. Suter R.G. Taylor R.P. Taylor R.K. Valenta J.L Walshe D.J. Whitford P.J. Williams A. Wilson C.J.L. Wilson W.K. Witt J. Zhiyu T. Zhou

14 10 36 55 57 18 52 59 62 46, 64 30 36. 55. 66 86 68 28, 32 75 48 32 70 72 46 14 32 46. 64 64 74 70 75 76 74 41 53 78 81