Mineralium Deposita

, Volume 40, Issue 1, pp 76–99

The development of volcanic hosted massive sulfide and barite–gold orebodies on Wetar Island, Indonesia

Authors

    • School of Ocean and Earth Science, Southampton Oceanography CentreUniversity of Southampton
    • WRG3 Sidings Court
  • Stephen Roberts
    • School of Ocean and Earth Science, Southampton Oceanography CentreUniversity of Southampton
  • Richard J. Herrington
    • Department of MineralogyNatural History Museum
  • Adrian J. Boyce
    • Isotope Geosciences UnitSUERC
  • Ray Burgess
    • Department of Earth SciencesUniversity of Manchester
Article

DOI: 10.1007/s00126-005-0468-x

Cite this article as:
Scotney, P.M., Roberts, S., Herrington, R.J. et al. Miner Deposita (2005) 40: 76. doi:10.1007/s00126-005-0468-x

Abstract

Wetar Island is composed of Neogene volcanic rocks and minor oceanic sediments and forms part of the Inner Banda Arc. The island preserves precious metal-rich volcanogenic massive sulfide and barite deposits, which produced approximately 17 metric tonnes of gold. The polymetallic massive sulfides are dominantly pyrite (locally arsenian), with minor chalcopyrite which are cut by late fractures infilled with covellite, chalcocite, tennantite–tetrahedrite, enargite, bornite and Fe-poor sphalerite. Barite orebodies are developed on the flanks and locally overly the massive sulfides. These orebodies comprise friable barite and minor sulfides, cemented by a series of complex arsenates, oxides, hydroxides and sulfate, with gold present as <10 μm free grains. Linear and pipe-like structures comprising barite and iron-oxides beneath the barite deposits are interpreted as feeder structures to the barite mineralization. Hydrothermal alteration around the orebodies is zoned and dominated by illite–kaolinite–smectite assemblages; however, local alunite and pyrophyllite are indicative of late acidic, oxidizing hydrothermal fluids proximal to mineralization. Altered footwall volcanic rocks give an illite K–Ar age of 4.7±0.16 Ma and a 40Ar/39Ar age of 4.93±0.21 Ma. Fluid inclusion data suggest that hydrothermal fluid temperatures were around 250–270°C, showed no evidence of boiling, with a mean salinity of 3.2 wt% equivalent NaCl. The δ34S composition of sulfides ranges between +3.3‰ and +11.7‰ and suggests a significant contribution of sulfur from the underlying volcanic edifice. The δ34S barite data vary between +22.4‰ and +31.0‰, close to Miocene seawater sulfate. Whole rock 87Sr/86Sr analyses of unaltered volcanic rocks (0.70748–0.71106) reflect contributions from subducted continental material in their source region. The 87Sr/86Sr barite data (0.7076–0.7088) indicate a dominant Miocene seawater component to the hydrothermal system. The mineral deposits formed on the flanks of a volcanic edifice at depths of ~2 km. Spectacular sulfide mounds showing talus textures are localized onto faults, which provided the main pathways for high-temperature hydrothermal fluids and the development of associated stockworks. The orebodies were covered and preserved by post-mineralization chert, gypsum, Globigerina-bearing limestone, lahars, subaqueous debris flows and pyroclastics rocks.

Keywords

GoldBariteMassive sulfideBanda ArcSubmarine hydrothermal systems

Introduction

Spectacular examples of asymmetric massive sulfide mounds, flanked by barite deposits, are preserved on Wetar Island, Indonesia. The orebodies occur marginal to hydrothermally altered volcanic rocks, which overly ocean floor basalts, and are preserved beneath post-mineralization Globigeriena-bearing limestones, lahars, subaqueous debris flows and pyroclastic rocks (Sewell and Wheatley 1994; Herrington and First 1996; Scotney et al. 1999). Associated hydrothermal alteration has argillic characteristics and the sulfide mounds contain pyrite, chalcopyrite, tennantite, covellite and low-Fe sphalerite. Mineralization occurred at around 4.7 Ma (Herrington and First 1996) and, unusual for a volcanogenic massive sulfide (VMS) system, only the precious metal-bearing barite resource was exploited. As a result, a subrecent VMS system is exceptionally preserved and exposed within open pits and associated drill core. Ore at Kali Kuning contained 1.9 Mt at 4.6 g/t gold, 151 g/t silver and 60% barite, with 2.2 Mt at 4.0 g/t gold, 146 g/t silver and 40% barite at Lerokis (Abadi 1996).

Volcanogenic massive sulfide systems are often significant repositories of gold and silver (see Hannington et al. 1999 for review). Various factors are recognized to play an important role in the gold enrichment. These include the tectonic setting of the deposits, which in turn influences the nature of the igneous basement, and the physical and chemical characteristics of the hydrothermal fluids, in particular temperature, salinity and oxygen fugacity (Hannington et al. 1999). The gold bearing characteristics of the sulfide assemblage and the argillic nature of the alteration at Wetar Island, has led to speculation that the hydrothermal fluids responsible for the gold mineralization contained a significant contribution of magmatic volatiles (Sillitoe et al. 1996). This paper describes the mineralization and alteration preserved on Wetar, and the results of fluid inclusion and stable and radiogenic isotope studies. These new data provide a better understanding of the nature and origin of the hydrothermal mineralizing system. Furthermore, as a relatively young system of Miocene age, the Wetar deposits provide an ideal opportunity to link observations from active systems on the ocean floor with a system only recently incorporated into the geological record.

Geological setting

Wetar Island forms part of the Inner Banda Arc, an array of active and inactive volcanic islands surrounding the Banda Sea, which are the result of the arc-continent collision of the NNE moving (75 mma−1) Indian–Australian plate beneath the Eurasian plate (Audley-Charles 1986; Masson et al. 1991) (Fig. 1). This zone of plate contact lies along the Java Trench to the west and continues into the Timor Trough. Seismic refraction surveys indicate that the Timor–Tanimbar Trough (1,200 km in length, up to 70 km wide and 2–3 km deep) is presently underlain by continental lithosphere, varying between 31 km and 40 km in thickness from west to east (Audley-Charles 1986; Masson et al. 1991).
Fig. 1

Location map of Wetar Island, Indonesia showing principal tectonic and volcanic features of the Banda Arc. After Hamilton (1979), Varekamp et al. (1989), Breen and Silver (1989) and Masson et al. (1991). The Banda Arc is divided into an Outer “non-volcanic” and Inner “volcanic” arc. The extent of the Australian continental crust is shown within the inset

The Outer Banda Arc is dominantly non-volcanic in origin, with Timor, to the south of Wetar preserving an accretionary prism and central collision complex, which was accreted onto the front of the Australian continental plate. Richardson and Blundell (1996) proposed that a substantial part of the collision complex consists of a micro-continental fragment that lay some considerable distance to the north of the Australian continental margin, and which collided with the subduction zone at approximately 8 Ma. Seismic velocity and gravity modelling suggests that the collision complex across the Timor profile is 37–60 km thick and 135–160 km wide (Woodside et al. 1989; Richardson and Blundell 1996). The frontal portion of this collision complex consists of a number of high-angle thrusts imbricated from the subducting Australian continental margin (Hughes et al. 1996) (Fig. 2). Uplift rates in both the accretionary zone and associated islands of the inner volcanic arc are high, with eastern parts of Timor presently situated 3 km above sea level (Snyder et al. 1996). Microfaunal and palaeobathymetry studies, on the islands of Timor, Buru, Seram and Kai, show that continent-arc collision has produced episodic uplift of the outer islands at rates of between 500 mm ka−1 in a million years, to 5,000 mm ka−1 in some hundreds of thousands of years (De Smet et al. 1989). These results indicate that uplift rates differ greatly along the arc, with some islands experiencing long episodes of submergence intermittent with rapid pulses of uplift during the Pliocene–Quaternary.
Fig. 2

Simplified cross-section of the Banda Orogen based on geophysical data from Masson et al. (1991), Richardson and Blundell (1996) and Snyder et al. (1996). Timor was formerly an outlier of Australian continental crust and is now trapped between the Inner Banda Arc and the Australian continental margin. The collision zone is dominated by shallow, southward dipping faults, which have accommodated crustal shortening and thickening within the zone of collision

A complex zone of normal and strike-slip faulting offsets the Inner Banda Arc between the islands of Alor and Wetar, to a distance of approximately 50 km (Fig. 1) (Masson et al. 1991). Recent GPS measurements, seismicity data and seismic reflection profiles suggest that the Wetar Thrust, located at the northern edge of the inactive segment of the Inner volcanic Banda Arc, accommodates the majority of the present day 75 mma−1 convergence, between the Australian margin and the Banda Arc (Silver et al. 1983; McCaffrey 1988; Genrich et al. 1994). This thrust may represent the site of incipient arc reversal, due to the increased difficulty in subducting the buoyant Australian continental plate post-arrival of the Australian continental margin with the collision zone at approximately 2.4 Ma (Richardson and Blundell 1996).

Materials and methods

Geological information was obtained from the open-pit mines, outcrops and exploration drill-core. Polished thin sections were examined in both reflected and transmitted light. Standard XRD and FTIR methods were employed to characterize alteration mineralogical assemblages on 135 samples.

Mineral chemistry was determined by electron microprobe analysis using a Cameca SX50 at the Natural History Museum, London. Operating conditions were 15 kV (accelerating voltage), 20 nA (beam current) and count time of 20 s.

Doubly polished fluid inclusion chips were prepared to a thickness of 100 μm. Microthermometric analyses were completed on a Linkam TMS600 stage calibrated against a pure H2O–CO2 inclusion at low temperatures, and checked daily against internal standards. Accuracy is estimated at ±0.1°C for low-temperature phase changes (−100°C to 0°C) with a precision ±0.1°C and ±2°C, for homogenization measurements between 100°C and 400°C, respectively.

Sulfides were prepared for conventional isotopic analysis at SUERC by standard heavy liquid, magnetic, diamond micro-drilling and hand picking techniques. Barite was prepared by micro-drilling. In both cases, around 5–10 mg was used for isotopic analysis. Minor contamination by non-S-bearing phases was tolerated, and has no effect on the final data. Sulfides were analyzed by standard techniques (Robinson and Kusakabe, 1975) in which SO2 gas was liberated by combusting the sulfides with excess Cu2O at 1,075°C, in vacuo. Barite analyses were performed by the technique of Coleman and Moore (1978), in which SO2 gas is liberated by combustion with excess Cu2O and silica, at 1,125°C. Liberated gases were analyzed on a VG Isotech SIRA II mass spectrometer, and standard corrections applied to raw δ66SO2 values to produce true δ34S. The standards employed were the international standards NBS-123 and IAEA-S-3, and the SUERC standard CP-1. These gave δ34S values of +17.1, −31 and −4.6‰, respectively, with 1σ reproducibility better than ±0.2‰. Data are reported in δ34S notation as per mil (‰) variations from the Vienna Cañon Diablo Troilite (V-CDT) standard. Selected barite concentrates for sulfate oxygen analyses were carefully cleaned by washing in Aqua Regia, and thorough rinsing in deionized water. Oxygen was then extracted following the method of Hall et al. (1991). The evolved CO2 was analyzed on a VG Sira 10 mass spectrometer, with all results reported in standard delta notation as ‰ variations relative to the V-SMOW international standard. Replicate analyses of the NBS-127 BaSO4 standard during these analyses gave +9.6±0.3‰.

Sr isotopes were measured at Southampton Oceanography Center on a seven-collector VG Sector 54 mass spectrometer with a separable-filament source. Isotope ratios were determined as the average of >100 ratios by measuring ion intensities in multidynamic collection mode and fractionation corrected by normalization to 86Sr/88Sr = 0.1194. Measured values for standard NBS SRM-987 were 87Sr/86Sr = 0.710242 ±13 (2 SD, n=42).

Stepped heating Ar/Ar data for biotite grains and illite separates (<2 μm), were analyzed at the University of Manchester, UK with analytical techniques following that of Kendrick et al. (2001). Samples of the syeno-granite and dacite were disaggregated by light crushing, and individual biotite grains (2–5 mm in length) were hand picked and cleaned in deionized water. Due to the fine-grained nature of the illite sample it was expected that there would be significant recoil loss of 39Ar during irradiation, therefore, this sample was vacuum encapsulated in a quartz vial prior to irradiation. The recoil 39Ar gas was extracted using an ultra-violet wavelength laser to drill into the tube. The recoil 39Ar amounted to only 5% of the total 39Ar released from the sample and was recombined in the total age calculation.

Geology of Wetar Island

Wetar Island measures approximately 110 km by 40 km and is composed entirely of Neogene volcanic rocks and minor oceanic sediments (Sewell and Wheatley 1994). Submarine, basaltic–andesites, with local pillows, form the volcanic basement to the island (Fig. 3). The basaltic–andesites are intruded by rhyo-dacite domes (Ruxton 1989) and overlain by dacitic lavas, tuffs and breccias, debris flows and lahar deposits (Fig. 4). Reef limestones are evident around the perimeter of the island at varying heights.
Fig. 3

Simplified geological map of Wetar Island, after Nash and Snodin (1992) and Farmer and Clifford (1993). The map shows the principal mineralized areas, structural lineaments, and key geological units

Fig. 4

Stratigraphic column and summary of tectonic and geochronological data from the Kali Kuning, Lerokis and Meron areas of Wetar Island. Age constraints are: (1) Scotney 2002 (Ar/Ar); (2) Herrington 1993 (K/Ar); (3) Abbott and Chamalaun 1981 (K/Ar); (4) Herrington 1993, a biostratigraphic age for a post-mineralization Globigerina bearing limestone from the Kali Kuning deposit. The Zanclian and Messinian aged strata are locally referred to as the mine sequence

Radiometric dating, largely K–Ar, of the volcanic assemblages suggests that the basement volcanic rocks, to the south of the island, were extruded around 12 Ma with overlying dacites, diorites and basaltic–andesites deposited between 7.78 Ma and 3.03 Ma (Abbott and Chamalaun 1981). A Globigerina-bearing limestone outcrops on Wetar Island, which locally overlies basaltic–andesitic volcanism and mineralization hosted by calc-alkaline andesites to rhyodacitic flows. This limestone yields a biostratigraphic age of between 5.2 Ma and 3.9 Ma (Herrington 1993), and based on the ratio of planktonic:benthic assemblages (Table 1), is likely to have formed in up to 2,000 m of water and possibly deeper (J. Murray, personal communication). Uplift rates for Wetar based on these parameters are 420–570 mm ka−1, which are consistent with the lower rates calculated for Timor in the outer arc (De Smet et al. 1989; Audley-Charles 1986b). Based on these uplift rates, the localized areas of mineralization on the Wetar Island edifice would have emerged from the Banda Sea between 0.5 Ma and 0.4 Ma. However, the central spine of Wetar (1,500 m, present height) would have emerged around ~3 Ma based on the current height differentials, and likely uplift rates.
Table 1

Planktonic and Benthic foraminifera assemblages identified within post-mineralization limestones at Kali Kuning

Planktonic assemblage includes

Benthic assemblage includes

Orbulia universa

Cibicidoides

Globigerinoides conglobatus

Globocassidulina

Globigerinoides sacculifer

Favocassidulina

Globorotalia menardiform group

Fontbotia

Sphaeroidinella dehiscens

Pleurostomella

 

Stilostomella

 

Uvigerina

Planktonic:Benthic ratio = 99:1

Subaerial lahars generated along the central spine of Wetar developed into extensive subaqueous debris flows. These units, locally up to 250 m in thickness, covered mineralized areas and infilled topographic depressions. After collision of the Australian margin with the Outer Banda Arc, at approximately 2.4 Ma (Richardson and Blundell 1996) uplift may have been substantially increased. In particular, the development of the Wetar thrust and subsequent back-arc thrusting may have aided in the rapid exhumation of the Wetar volcanic edifice.

Geology of the sulfide deposits

Volcanic and structural setting of the deposits

The economic deposits of Kali Kuning and Lerokis zones 4 and 5 are located towards the central and northern part of the island (Fig. 3). These deposits, and the majority of other recognized mineralized zones, which lie within the central part of the island, are bound to the west and east by extensive NW–SE and NE–SW trending faults. The base of the volcanic stratigraphy comprises fine-grained basaltic–andesitic flows, which dip around 10° towards 028°. Overlying the basal vesicular basalts and basaltic–andesites is a >450 m thick sequence of altered volcanic rocks, locally termed the mine sequence (Fig. 4). At the base of this sequence green, chloritic altered, vesicular pillow lavas are well preserved. Up section andesitic to rhyodacite flow units and local breccias are preserved, and these are the host rocks to the mineralization. Unconformably overlying this sequence are a series of post-mineralization lahars and debris flows, which appear geomorphologically controlled by the palaeotopography. Local hydrothermally altered dykes cross-cut the mine sequence, with clear evidence of post-mineralization dykes restricted to unaltered E–W striking andesitic dykes, which cut lahars and debris flows within coastal exposures.

The deposits are discordant to the local stratigraphy and are associated with faults. The Lerokis zone 5 mound developed at the intersection of a northwest and westerly trending structure and the Kali Kuning sulfide mound is located along a northwesterly trending fault.

Massive sulfides

Two well-preserved polymetallic sulfide mounds at Kali Kuning and Lerokis zone 5 have exposed dimensions of ~150×100×70 m and ~120×90×30 m, respectively (Fig. 5a–d). Pre-mining, no sulfide mounds were exposed at the surface. In plan, the mounds are broadly arcuate. The sulfide mounds are blocky in appearance, with clasts of massive pyrite ranging in diameter from a few centimeter up to boulders some 30 cm across (Fig. 6). Talus and redeposited sulfides occur marginal to the mounds where matrix supported angular fragments of massive sulfide are held in a fine-grained sulfide mud. Minor evidence for seafloor reworking is evident at Lerokis zone 5, where a 30 cm zone of interbedded sulfide and volcaniclastic material overlies the mound. Chert, gypsum and globigerina-bearing limestone overlie the Kali Kuning sulfide mound. At the margin of the sulfide mounds, fine-grained poorly consolidated granular pyrite marks the contact zone (0.2–2.5 m) between the sulfide mound and the associated barite deposits. The mineralogy of the massive sulfide mounds is dominated by pyrite, accounting for >98% of all sulfides present with minor amounts of chalcopyrite and sphalerite. Typical of seafloor sulfides, the pyrite and chalcopyrite often show “porous” textures as well as collomorphic growth zones up to 3 mm across (Fig. 7a). The collomorphic pyrite tends to nucleate on and around euhedral pyrite grains (Fig. 7b). This texture appears most frequently at the margins and upper parts of the sulfide mounds. Chalcopyrite frequently rims and locally replaces pyrite (Fig. 7c) and is more apparent at the margins of the sulfide mounds and particularly at the base of mounds and in the underlying footwall. Occasional banding of pyrite and chalcopyrite is evident on a centimeter scale. A later fracture network permeates the pyritic mounds, with a sulfide assemblage dominated by covellite, Fe-poor sphalerite and lesser amounts of tennantite and tetrahedrite and tabular barite laths (Fig. 7d). Overall, typical sulfide abundance within the mounds are pyrite >> chalcopyrite > sphalerite > covellite/marcasite/tennantite/tetrahedrite and bornite. No sulfide mound is evident at the Lerokis zone 4 deposit despite drilling beneath the barite mineralization. All three deposits are surrounded by extensive gossanous material.
Fig. 5

a View north–northwest of the Kali Kuning (KK3) deposit. The irregular nature of the sulfide mound is clear; pre-mining, the entire mound was covered by baritic ore. Hydrothermal alteration is evident around the deposit, post-mineralization lahars / debris-flows are shown. b Exposed sulfide mound at Kali Kuning. Height of the sulfide mound is approximately 60 m. c Lerokis zone 5 situated at a topographic height of 550 m on a prominent ridge (view approximately north). The host depression for the barite ore deposit is evident. Conformable, post-mineralization volcaniclastic-sediment overlies the massive sulphide mound. Pre-mined and remediated zones 1, 2 and 3 are also shown. d Exposed sulfide mound at Lerokis zone 5, height of the sulfide mound above the pit floor is approximately 15 m. Extensive gossanous material surrounds both deposits

Fig. 6

Blocky sulfide talus at the base of the Lerokis zone 5 sulfide mound

Fig. 7

Photomicrographs of polished sulfide sections: py pyrite; cpy chalcopyrite; sp sphalerite; ba barite; ten tennantite; cov covellite; si silica. ae field of view = 5 mm, f field of view = 2.5 mm. a Collomorphic pyrite (sample 097056, Lerokis zone 5) within the massive pyritic sulfide mound. b Euhedral pyrite cores overgrown by collomorphic pyrite (sample 097009). c Chalcopyrite replacement of pyrite (sample 097059). d Fracture-fill sulfide assemblage (sample 097016). e Disseminated pyrite within an altered volcanic clast in the Lerokis zone 5 footwall, rimmed by pyrite and a fracture-fill assemblage of sphalerite and covellite (sample 097122). f Detail of fracture-fill sulfide assemblages in e, showing variation in tennantite composition

Stockwork zones

The pyritic mounds at Kali Kuning and Lerokis zone 5 are underlain by stockwork zones which reach to a depth of >210 m below the Lerokis zone 5 deposit and >150 m below Kali Kuning. The stockwork zone is hosted by hydrothermally altered, locally vesicular, silicified volcanic rocks. Brecciated, angular volcanic footwall clasts are rimmed by sulfides up to 4 mm in thickness and also contain disseminated sulfides (Fig. 7e). The stockwork veins range from <1 mm up to 4 cm in width and contain pyrite, chalcopyrite and sphalerite. Intense vein and disseminated mineralization occurs within an extremely silicified zone immediately beneath the sulfide mound at Lerokis zone 5 and appears to be related to a fault intersection. Disseminated pyrite is also abundant in this zone and is generally euhedral to subhedral (usually <1 mm). Gold was observed in association with a zone of intense clay (illite) alteration in the stockwork beneath the Lerokis zone 5 sulfide mound (Herrington 1993). A late sphalerite, tennantite and covellite-rich fracture network (Fig. 7e, f) locally permeates the stockwork zone.

Barite deposits

The mined gold-rich barite deposits have been termed barite sands by Sewell and Wheatley (1994) and Scotney et al. (1999) due to their unconsolidated friable nature. At Kali Kuning and Lerokis zones 4 and 5 the deposits contain on average ~60–70% barite, and up to 90% barite where more massive. The barite consists of a friable mass of barite crystals, which shows variable degrees of cementation and colour variation, such that chaotic bedding with evidence of slumping are locally defined. Individual barite crystals typically range in size between 2 mm and 4 mm, up to a maximum of approximately 7 mm in length. Barite crystals typically show euhedral, rectangular, rhombohedral and polyhedral forms (Fig. 8a). Cross-bedding within the barite is reported by Sewell and Wheatley (1994), suggesting the local reworking of barite on the seafloor; however, these features were largely evident pre-mining and within the upper parts of the deposits. At Kali Kuning, barite overlies and locally surrounds the sulfide mound. The contact zone between the massive sulfide and barite deposit is gradational, with a zone of granular pyrite, clay and barite up to 2 m thick.
Fig. 8

a SEM image of rectangular, rhombohedral and polyhedral interlocking barite crystals from the Kali Kuning ore zone (sample 096879). b Relationship of aresendescloizite (ad), arsenbrackebuscite (ab) and mimetite (m) as cementing phases within the Lerokis zone 4 barite ore zone. c Secondary nature of anglesite (an) pseudomorphing galena and cementing barite laths. d Barite lath rimmed by jarosite (ja) and subsequently rimmed / cemented by karibibite (k) and arsenbrackebuscite (ab)

The matrix of the poorly consolidated barite largely comprises collomorphic Fe-oxide, traces of clay material and arsenates (Table 2), with increased cementation towards the footwall contact. Arsendescloizite, arsenbrackebuscite and mimetite are all found intergrown, cementing tabular barite (Fig. 8b). Anglesite is observed pseudomorphing galena and as a cementing phase to tabular laths of barite and brecciated fragments (Fig. 8c). Arsenbrackebuscite occurs as rhombic crystals (<20 μm) showing compositional zoning (Fig. 8b,d), with cores depleted in As and Pb and enriched in Sb and S compared to the margins; this phase also occurs as amorphous cement. Euhedral barite laths are cemented by jarosite and a complex intergrowth of karibibite and arsenbrackebuscite (Fig. 8d). Gold occurs in the baritic ore unit as free grains and in one instance attached to a barite lath (Fig. 9). An association between gold and Fe-oxides is also noted by Herrington (1993). The Hg bearing phase moschellandsbergite occurs at the stratigraphically highest position of the Kali Kuning barite ore zone near to the overlying chert and limestone cover, whereas, the Ag-bearing mimetite phase along with the majority of the complex arsenates tend to occur at the base of the Lerokis zone 4 barite deposit; which coincides with the highest Au and Ag grades reported towards the base of other ore zones, e.g. Lerokis zone 1 and Kali Kuning (Sewell and Wheatley 1994).
Table 2

Barite ore cementing and inclusion phases identified at Wetar Island

Minerals identified cementing barite

Minerals identified included within barite

Anglesite: PbSO4

Sphalerite: ZnS

Romerite: Fe2+ Fe23+ (SO4)4·14H2O

Galena: PbS

Silica/Jasper: SiO2

Chalcopyrite: CuFeS2

Mimetite: Pb5(AsO4)3Cl

Argentite: Ag2S

Moschellandsbergite: Ag2Hg3

Acanthite: Ag2S

Jarosite: KFe33+ (SO4)2(OH)6

Pyrite: FeS2

Fe-oxides

Bournonite: PbCuSbS3

Arsendescloizite: PbZn(OH)(AsO4)

Bornite: Cu5FeS4

Arsenbrackebuscite: Pb2(Fe,Zn)(AsO4)2·H2O

Jordanite: Pb14(AsSb)6S23

Karibibite: Fe23+ As43+ O9

Boulangerite: Pb5Sb4S11

Plumbojarosite: PbFe63+ (SO4)4(OH)12

Geochronite: Pb14(SbAs)6S23

Perroudite: (Hg5Ag4S5Cl4)

Stibio-Luzonite: Cu3SbS4

Clorargyrite: AgCl

Enargite: Cu3AsS4

Cerussite: PbCO3

 

Dusserite: BaFe3(AsO4)2(OH)·5H20

 

Cinnabar: HgS

 

Copiapite: Fe2+ Fe3+4(SO4)6(OH)2·20H2O

 

Romerite: Fe2+ Fe3+2(SO4)4·14H2O

 
Fig. 9

Anhedral, <20 μm gold grains adhering to barite laths

Variable amounts of sulfide phases, predominantly pyrite and galena are included within the barite laths (see Table 2). Included phases range from euhedral to anhedral in form and the size of inclusions are generally <50 μm in diameter. Significantly more sulfides are included within barite laths from the barite orebodies compared to barite inclusions from within the sulfide mound and stockwork structures.

Barite also occurs as disseminated laths within stockwork zones (up to 1 cm), as vug fill within sulfide mounds, associated with the fracture-fill sulfide assemblage and as coatings on external sulfide mound faces. Vein barite occurs both proximal and distal to mineralization with individual crystals up to 5 cm in long dimension.

Iron oxide–barite pipe structures

Distinct iron oxide–barite structures are preserved at Kali Kuning and Lerokis zones 4 and 5. These structures are pipe-like in appearance and, at Lerokis zone 4 (Fig. 10), are 1–1.5 m high and ~1.5 m in diameter and are located towards the base and central part of the mined barite deposit. At Lerokis zone 5 two irregular structures (1 m ×12 m) are present (Fig. 5d) and at Kali Kuning two iron oxide–barite pipe features are observed, one at the base of the mined barite ore zone, the other directly underlying the chert and limestone cover rocks. All these structures comprise abundant anastomosing layers of barite separated by layers of iron oxides, (goethite and limonite). Locally, incorporated clasts of vesicular and highly altered volcanic rocks are noted, which were probably derived from the volcanic footwall.
Fig. 10

Examples of Fe-oxide barite pipe structures, approximately 1.5 m in height, preserved at the base of the Lerokis zone 4 barite ore deposit

Mineral paragenesis

The mineralization on Wetar can be subdivided into paragenetic stages based on observations from the three major components of the systems: stockwork zone, sulfide mounds and barite deposits (Fig. 11).
Fig. 11

Paragenetic sequence for the mineralization events at the Wetar Island ore deposits

Stage I

Pyrite (+ chalcopyrite + sphalerite + marcasite). Forming the massive sulfide mounds, and associated stockworks.

Stage Ia

Barite as poorly consolidated barite ore deposits, typically with inclusions of mound sulfides, flanking the massive sulfide mounds. Gold is evidently associated with barite precipitation. Notably, sulfide inclusions in the proposed feeder structures to the barite deposits only contain inclusions of sulfides typically reported for the massive sulfide event, e.g. pyrite, sphalerite, chalcopyrite, with no evidence of sulfosalts as inclusions.

Stage II

Multiple fracture events reactivate the stockwork zone and the sulfide mounds. These veins show chalcopyrite/sphalerite intergrowth, Fe-poor sphalerite, Zn-poor tennantite, Zn-rich tennantite ± tetrahedrite, covellite, bornite, digenite and barite. Tabular barite laths up to 7 mm long are associated with this event.

Stage III

Late barite veins with individual crystals up to 5 cm developed both proximal and distal to mineralization and coating external faces of the sulfide mounds.

Stage IV

Cementation of the barite deposit by oxides, arsenates and sulfates other than barite (see Table 2) and partly through the local oxidation of sulfides. The relationship of this event to stage III is uncertain.

Alteration

Intense zones of alteration are present within the footwall to Kali Kuning, and Lerokis zones 4 and 5. At the Kali Kuning orebody, the immediate footwall to the sulfide mound and barite mineralization exhibits intense silicification, including microcrystalline silica and cristobalite, and a clay alteration assemblage of illite, pyrophyllite, kaolinite and minor alunite. This alteration gives way laterally and vertically to an illite–smectite assemblage, which passes outwards to zones of chloritic alteration (Fig. 12a). At Lerokis zone 5, a similar zonation is observed, with the most intense alteration associated with the stockwork that underlies the sulfide mound. Beneath the massive sulfide mound is a ~25 m thick zone of silicification, microcrystalline silica, and extensive argillic alteration, illite, pyrophyllite and kaolinite. Beneath this zone the predominant alteration is chloritic. At Lerokis zone 4 (Fig. 12b), which lacks an exposed sulfide body, the immediate footwall to the barite deposit is a zone of quartz–illite rich alteration and microcrystalline silica. Residual silica and local jarosite veins are observed (Fig. 12c, d). This alteration passes outwards into a zone of quartz–illite, illite–smectite and finally chloritic alteration. Intense silicification is observed within the footwall adjacent to all the barite deposits. More peripheral footwall zones to the barite deposits show predominantly chloritic alteration, which is well exposed in road cuts away from the mineralization.
Fig. 12

a Illite–smectite and chloritic alteration zonation around the Kali Kuning ore following removal of the barite. b Zoned alteration sequence around the Lerokis zone 4 orebody, major phases are shown determined by XRD, FTIR and initially by PIMA: qtz microcrystalline quartz; ill illite; ja jarosite; goe goetite; chl chlorite; sm smectite; kao kaolinite; al alunite. The rivers S. Koreng, S. Kelapa Tiga and the mine road are shown. Located at the centre of the ore zone are the Fe oxide/barite pipe structures. c Sample of residual silica collected from the centre of the barite ore zone footwall at Lerokis zone 4. d Jarosite veining at Lerokis zone 4

The alteration around the Wetar mineralization is thus extensive, intensive and zoned. The dominant alteration assemblages comprise silicification proximal to mineralization, moving outwards and downwards to illite–smectite and finally to distal chlorite. Although spatially restricted, the presence of pyrophyllite, kaolinite and local alunite suggest that low pH fluids at moderate temperatures have contributed to the alteration in the immediate footwall to the sulfide mounds and the barite ore deposits (Reyes 1990).

Mineral identification

Pyrite

Pyrite is ubiquitous in all associations except barite ores and the associated pipe structures where iron is generally present as oxides. Stage I pyrite, forms the bulk of the sulfide mound contains a distinct arsenian pyrite. This pyrite often shows complex grains or collomorphic banded zones (Fig. 13a, b) and may contain up to 6.7 wt% As (Table 3). The arsenian cores to complex grains often contain idiomorphic inclusions of barite. Overgrowth pyrite is characteristic of the later part of stage I, immediately pre-dating or coincident with main stage chalcopyrite. This later pyrite (stage Ia) is characteristically arsenic-poor.
Fig. 13

Back-scattered electron microphotographs of As zonation within massive pyrite samples from the Lerokis zone 5 sulfide mound. Point locations correspond to the microprobe traverse, the graphs illustrate the variation of As along each sample, a sample 097009; b sample 099903

Table 3

Representative analysis of Wetar Island mineral phases from the Kali Kuning and Lerokis orebodies, Indonesia

 

py

py

ten

bor

sph

sph

tet

ten

bour

ga

cpy

arg

jor

py

 

Mim

ars

Stage

I

Ia

Ia

Ia

Ia

II

II

II

Ia

Ia

Ia

Ia

Ia

Ia

IV

IV

 

Sulfide phases

Sulfide inclusion phases within barite

 

Cementing phases

S

48.01

52.60

31.23

30.46

32.61

32.29

25.16

26.56

19.36

13.74

34.32

12.00

17.94

51.64

Cl

2.09

0.20

Ag

0.19

0.05

0.10

0.01

0.06

0.26

80.34

0.02

SO2

0.65

0.46

As

6.71

0.19

15.91

0.07

7.02

17.79

1.83

0.07

0.10

0.06

11.12

1.42

FeO

0.25

14.02

Ba

0.46

0.12

0.16

0.79

0.87

2.63

0.42

0.67

ZnO

0.13

0.07

Cd

0.07

0.48

As2O5

22.60

30.67

Cu

0.03

0.20

45.80

62.21

0.23

1.56

38.56

41.24

12.78

0.02

34.62

0.06

0.03

0.23

SeO2

0.03

0.02

Fe

44.37

46.48

0.04

7.15

6.12

0.10

2.20

0.11

0.05

29.07

1.26

44.86

Ag2O

0.79

0.15

Hg

0.03

0.06

0.14

0.65

PbO

73.05

50.65

Pb

41.08

84.37

69.48

Sb2O3

0.09

0.04

Sb

3.95

0.03

0.19

19.87

3.10

22.57

1.31

0.12

0.64

0.04

BaO

0.04

Se

0.04

0.38

0.13

0.03

0.03

0.04

0.05

0.06

0.03

0.01

1.20

0.05

0.04

HgO

0.00

Zn

0.04

0.16

58.28

64.98

5.22

8.30

0.16

0.07

0.01

0.18

Bi2O3

0.15

0.33

Total Wt%

99.20

100.04

97.34

100.43

97.48

99.71

98.07

97.16

97.92

100.73

99.30

98.21

99.68

99.08

Total compound%

99.85

96.60

Sulfide phases (wt %): py pyrite; sph sphalerite; tet tetrahedrite; ten tennantite; bour bournonite; ga galena; cpy chalcopyrite; arg argentite; jor jordanite; bor bornite

Cementing phases (compound %): mim mimetite; ars arsendescloizite

Chalcopyrite

Chalcopyrite is associated with the overgrowth pyrite and locally replaces the earlier stage I pyrite. Compositional variation in the chalcopyrite is negligible with respect to paragenetic association. Chalcopyrite often occurs intergrown with later stage II barite.

Sphalerite

Sphalerite is associated with the chalcopyrite and as a later phase with sulfosalts and late barite. Earlier chalcopyrite-associated sphalerite is generally opaque to translucent in thin section, reflecting an elevated iron content of up to 6 wt% Fe. Later sulfosalt-associated sphalerite (stage II) is generally iron-poor, often less than 0.1 wt% Fe (Table 3)

Sulfosalts

Tennantite is a minor accessory of the chalcopyrite–sphalerite stage of mineralization and is more abundant in stage II cross-cutting veins containing covellite, digenite, bornite and barite. In the earlier stage (chalcopyrite + sphalerite, Fig. 7d), the tennantite is generally zinc-poor with between 0 wt% Zn and 1 wt% Zn, ~31 wt% S, ~17 wt% As and 0.1–3.9 wt% Sb. In the later association (+covellite), both zinc-poor and zincian tennantite are found (Fig. 7f). Zn-rich tennantite contains up to 8.3 wt% Zn, ~27 wt% S and ~19 wt% As and between 0.4 wt% and 6.5 wt% Sb. Zincian tetrahedrite is often seen as inclusions in Fe-poor sphalerite. Bournonite and jordanite are common as tiny inclusions in larger barite laths.

Minor sulfides

Covellite and digenite are common in the stage II veins and bornite plus argentite are common inclusion phases in barite laths. The covellite from stage II cross-cutting veins in the sulfide mound and stockworks contain between 0.2 wt% and 1.3 wt% arsenic.

Barite

Barite appears as a ubiquitous phase throughout the main sulfide mound, sulfide talus breccias, barite ores, stage II stockwork veins and stage III veins. The mineral chemistry of the barites is variable with both Sr-poor and Sr-enriched types present in samples from the stockwork and the ore.

Barite cementing phases

The phases cementing the barite ores are largely iron oxides, hydroxides, complex arsenates and sulfate phases, many of which are amorphous. Distinct phases analyzed include mimetite, arsenbrackebuscite, arsendescloizite, anglesite and moschellandsbergite. Mimetite contains significant silver, along with the minor argentite as inclusions in barite, and was probably the source for the silver recovered in the vat-leach processing of the barite ore. Along with cinnabar, metacinnabarite, perroudite and tiemannite (De Roever 1991), moschellandsbergite is a major host for mercury in the barite ore, the presence of which was discovered during early gold production at Wetar and subsequently recovered during processing (Sewell and Wheatley 1994).

Fluid inclusion studies

A representative suite of barite samples, encompassing all stages of ore paragenesis, were analyzed using microthermometric techniques. These included samples from the stockwork zone, barite ore, late vein barite and iron oxide–barite feeder structures to the barite ore. The majority of analyzed samples (7 of 11) were collected from the Lerokis zone 5 deposit. Optical examination indicates that >95% of all barite hosted inclusions are associated with secondary and pseudo-secondary inclusion trails (Fig. 14a), approximately 5% appear primary. The solitary bi-phase primary inclusions are two phase (L + V) and range in size from 12 μm to 25 μm; these are described type 1A. Type 1B are also two-phase liquid plus small amounts of vapour inclusions (L + V), which occur as pseudo-secondary and secondary trails. Recognition of a few mono-phase (L) inclusions (by first and final ice-melt phase changes) are described as Type 2. Inclusions are concentrated towards crystal margins particularly in stockwork associated samples (Fig. 14b).
Fig. 14

ab Photomicrographs of barite hosted fluid inclusions from Wetar Island. a Pseudo-secondary and secondary trails in samples 097084, decrepitated inclusions are also evident (d). b Abundant 2 phase inclusions in sample 097106, a pronounced reduction in number of inclusions is evident toward the crystal core

Microthermometry results

For all Type 1 fluid inclusions, the degree of fill (F) ranges from 0.6 to 0.95, typically around 0.9–0.95, no daughter phases are observed and many inclusions show evidence of decrepitation or leakage (Fig. 15a). Homogenization of inclusions (Th) is consistently L + V to L and occurs between 144°C and 314°C with a mode of 238°C (n=206) (Table 4, Fig. 15b). First ice-melt temperatures (Tfm) occur between −20°C and −24.4°C (n=35), with final ice melting (TMice) between −0.9°C and −3.1°C (mode −1.6); corresponding to salinities of 1.6–5.1 wt% equivalent NaCl (Bodnar, 1993) (Fig. 15c and d). This suggests that salinities span the salinity of seawater (~3.2 wt% equivalent NaCl).
Fig. 15

Fluid inclusion microthermometric data from Wetar Island barite samples

Table 4

Isotopic and microthermometric data obtained for the Wetar samples

Sample

Material analyzed

δ34S (CDT)

δ18O (SMOW)

87Sr/86Sr

±2 SE

Th (°C)

TFm (°C)

Tmice (°C)

NaCl (wt%)

Description

Location

Range

Av

n

Range

Av

n

Range

Av

n

111943

Barite

22.78

 

0.70764

7

196–262

235

16

−22.1–22.8

−22.5

4

−1.4–2.3

−1.9

15

2.5–3.8

Barite asso. with sulfides

KK

100134

Barite

24

             

Barite asso. with sulfides

KK

096877

Barite

24.23

             

Barite ore

KK

096878

Barite

26.98

             

Barite ore

KK

096879

Barite

23.65

10.9

0.70773

7

209–289

242

21

−21–22.5

−21.9

4

−1–1.9

−1.6

17

1.7–3.2

Barite ore

KK

096889

Barite

22.39

             

Barite ore

KK

100133

Barite

9.1

             

Barite ore

KK

100103

Gypsum

12.8

13.3

0.70909

7

          

Post-mineralization cover

KK

096890

Pyrite

11.55

             

Granular pyrite

KK

096858

Pyrite

8.55

             

Granular pyrite

KK

096879

Pyrite

8.44

             

Sulfide mound

KK

111943

Pyrite

9.52

             

Sulfide mound

KK

111943

Chalcopyrite

10.25

             

Sulfide mound

KK

100134

Pyrite

11.69

             

Sulfide mound

KK

096857

Pyrite

10.37

             

Sulfide mound

KK

097228

Pyrite

10.78

             

Sulfide mound

KK

111942

Pyrite

8.47

             

Sulfide mound

KK

100144

Native sulfur

15.8

             

Vein from the footwall

KK

097333

Pyrite

9.89

             

Vein/disseminated pyrite

KK

111979

Barite

25.18

6.5

0.70827

           

Barite ore

L4

111931

Barite

26.3

7.7

0.70774

8

189–267

228

17

−21.1

1

−1.1–2.1

−1.6

6

1.9–3.6

Barite ore

L4

096948

Barite

24.6

8.7/9.4

0.70801

8

197–309

240

20

−23

1

−0.9–2.4

−1.6

16

1.6–4.1

Baritic feeder structure

L4

099917

Barite

24.2

8.4

0.70829

7

183–237

220

15

−20.4–21.8

−21.1

3

−1.6–2.6

−2.2

12

2.8–4.3

Barite asso. with sulfides

L5

099939

Barite

23.9

6.8

0.70836

8

199–269

233

18

−2.6

1

−1.2–1.8

−1.6

17

2.1–3

Barite ore

L5

099938

Barite

31

11.3

0.70786

8

144–212

194

15

−22–24.4

−22.7

11

−1.4–2.7

−2.2

22

2.5–4.5

Barite vein, distal to sulfide mound

L5

099933

Barite

27.8

10.5

0.7088

8

          

Barite vein, proximal to sulfide mound

L5

099923

Barite

24.5

7.8

0.7083

8

          

Baritic feeder structure

L5

097106

Barite

27.5

7.7

0.70807

8

217–302

255

24

−20–21.8

−20.7

5

−1.8–3.1

−2.1

17

2.5–5.1

Disseminated stockwork barite

L5

097106

Pyrite

3.33

             

Disseminated stockwork sulfides

L5

097104

Sphalerite

6.1

             

Disseminated stockwork sulfides

L5

097098

Pyrite

6.7

             

Disseminated stockwork sulfides

L5

111997

Pyrite

7.25

             

Pyrite vein, distal to sulfide mound

L5

099902

Pyrite

6.92

             

Sulfide mound

L5

099917

Pyrite

6.26

             

Sulfide mound

L5

097219

Pyrite

8.4

             

Vein/disseminated pyrite

L5

097122

Barite

28.1

8.6

0.70836

 

186–314

269

20

−22.1–22.5

−22.4

4

−1.9–2.8

−2.2

16

3.2–4.1

Barite asso. with cov-spha-ten

L5

097059

Barite

22.7

5.8

0.70809

8

          

Barite asso. with sulfides

L5

097024

Barite

24.6

11.3

0.70873

 

210–288

245

18

−1.6–2.4

−2.1

11

2.8–4.1

Barite asso. with sulfides

L5

097023

Barite

24.43

             

Barite asso. with sulfides

L5

097098

Barite

25.04

8.9

0.70836

           

Disseminated stockwork barite

L5

097084

Barite

27.02

8.6

0.70854

8

187–286

230

22

−21.4

1

−1–2.1

−1.8

14

1.7–3.6

Disseminated stockwork barite

L5

097084

Pyrite

9.03

             

Disseminated stockwork sulfides

L5

097084

Sphalerite

4.26

             

Disseminated stockwork sulfides

L5

097122

Pyrite

8.02

             

Stockwork sulfides

L5

097122

Cov-spha-ten

6.91

             

Cov-spha-ten stockwork sulfides

L5

097041

Pyrite

6.56

             

Stockwork sulfides

L5

097024

Pyrite

8.57

             

Stockwork sulfides

L5

097059

Pyrite

8.01

             

Sulfide Mound

L5

097059

Chalcopyrite

7.44

             

Sulfide mound

L5

097023

Pyrite

7.52

             

Sulfide mound

L5

097009

Pyrite

7.4

             

Sulfide mound

L5

096920

Whole-rock

  

0.70773

17

          

Coastal basalt sample

096921

Whole-rock

  

0.70773

18

          

Coastal basalt sample

096922

Whole-rock

  

0.70775

17

          

Coastal basalt sample

096884

Whole-rock

  

0.70834

21

          

Altered volcanic rock

KK

096168

Whole-rock

  

0.71166

20

          

Altered post-mineralization dacite

Meron

096167

Whole-rock

  

0.71107

12

          

Post-mineralization dacite

Meron

111914

Whole-rock

  

0.70747

17

          

Altered volcanic rock

KK

097106

Whole-rock

  

0.70772

7

          

Brecciated stockwork volcanic

L5

100129

Whole-rock

  

0.70789

9

          

Altered, vesicular footwall volcanic

KK

111935

Whole-rock

  

0.70819

23

          

Altered volcanic rock

L4

111999

Whole-rock

  

0.70803

21

          

Syeno-granite

Meron

111994

Whole-rock

  

0.70781

7

          

Basaltic–andesite

KK

096926

Whole-rock

  

0.70748

9

          

Basaltic–andesite

KK Kali Kuning; L4 Lerokis Zone 4; L5 Lerokis Zone 5; cov-spha-ten covellite, sphalerite, tennantite composite sample

Although the petrographic data allows the inclusions to be subdivided into Type 1A, 1B and Type 2 inclusions, they show similar Th and salinity ranges throughout the paragenesis, which suggest no significant evolution of the hydrothermal fluid throughout the barite crystallization and that a single-phase fluid was responsible for the precipitation of barite within the system. However, the variable salinities above and below that of seawater suggest that super-critical phase separation may have occurred deeper in the system (Delaney and Cosens 1982; Bischoff and Pitzer 1985). The Globigerina-bearing limestone caps the mineralization at Kali Kuning and preserves fauna, including Favocassidulina, and benthic:planktonic foraminifera ratios, which suggest deposition occurred at least 2,000 m below sea-level and possibly considerably deeper (J. Murray, personal communication). These results contradict those of Sewell and Wheatley (1994), who suggested that the mineralization occurred in water depths of <600 m. Based on a minimum depth estimate of 2,250 m and a maximum estimate of 3,000 m, the confining pressure equates to a homogenization temperature correction of +23°C and +40°C (Bodnar and Vityk 1994). If the maximum correction is applied, the Wetar Island average trapping temperature will be 275°C. The fluid inclusion data suggest that the hydrothermal fluids responsible for the barite deposition throughout the Wetar hydrothermal systems were similar in salinity and temperature to those currently venting as white smoker fluids on active vent sites at mid-ocean ridges (Rona et al. 1993).

Stable isotopes

δ34S

δ34S analyses of 56 samples, comprising 29 sulfides, 25 sulphates, 1 gypsum and 1 native sulfur from the deposits of Kali Kuning, and Lerokis zones 4 and 5, significantly expand on earlier data and ranges reported by de Ronde (1995). Sulfides were collected from all stages of mineral paragenesis and sulfates were recovered from barite bodies, within the mound sulfides and from stockwork zones (Table 4).

Sulfides

The δ34S values of pyrite within the massive sulfides range from 6.2‰ to 11.7‰, stockwork sulfides range between 3.3‰ and 9.8‰ (Fig. 16 and Table 4). The poorly consolidated granular pyrite from the margins of the mounds, show a similar range of δ34S to the massive sulfides of between 8.5‰ and 11.7‰. No significant variations are observed for the different sulfide phases, sphalerite, chalcopyrite and a mixture of sphalerite, tennantite and covellite. The similarity of the sulfide data from both the underlying stockwork and sulfide mounds suggests a common source of sulfur at each deposit. The mean value of δ34S for this study is 8‰, close to δ34S values of +5 and +7‰ reported for Indonesian arc lavas by De Hoog et al. (2001). It is suggested that arc-related lavas are enriched in δ34S as a result of the significant incorporation of reduced seawater sulfate during petrogenesis. In general, disseminated sulfide in the stockwork system has a similar range of δ34S to the mineralization (mound and vein stockwork, Table 4). This consistency of δ34S suggests a genetic correlation between the volcanic sulfide and mineralization. Anomalously heavy δ34S is a common occurrence in these settings (e.g. Arribas 1995).
Fig. 16

Comparison of sulfur isotope data from the Kali Kuning, Lerokis zone 4 and 5 sulfide barite deposits

Sulfates

The δ34S of sulfates recovered from the Barite deposits of Wetar varies between 22.4‰ and 26.9‰, with the heaviest δ34S values, up to 31‰, recorded from the stage III barite veins distal to the ore deposits (Fig. 16). Ohmoto et al. (1983) reported similar δ34S barite values from Kuroko deposits, whereas, Goodfellow and Franklin (1993) report barite δ34S values from the Bent Hill system, which are considerably less than seawater, by as much as 10‰. The Wetar sulfate data reflect or are heavier than the predicted Miocene seawater sulfate value of δ34S 22‰ (Claypool et al. 1980). Ohmoto et al. (1983) suggested that mixtures of seawater sulfate and hydrothermal fluid sulfate, with equilibrium δ34S values of between 29‰ and 34‰ between 280°C and 200°C, could account for the values observed in Kuroko systems. However, only limited mixing of seawater and hydrothermal fluid would return the δ34S sulfate values to seawater, as modern vent fluids contain extremely low levels of dissolved sulfate (Scott 1997). Chiba et al. (1998) suggested that increases in δ34S observed for anhydrite in the TAG mound is due to the partial reduction of seawater sulfate by ferrous iron in the hydrothermal fluid. Given that the formation of barite requires the mixing of seawater and hydrothermal fluid a similar interpretation is favored. The heaviest δ34S sulfate values are reported from the distal stage III barite veins and suggest that seawater and hydrothermal fluid mixtures are subject to closed system reduction, highlighted by the peripheral vein at +31‰ (Table 4, sample 099938). Notably, coexisting pyrite and barite pairs commonly give unrealistic isotope equilibrium temperatures (usually higher), which corroborates our observations that the pyrite and barite is seldom co-precipitated.

Goodfellow and Franklin (1993) account for the relatively light δ34S data at Bent Hill, by the mixing of barium in vent fluids with sulfate formed by the oxidation of pre-existing sulfides, or H2S from the hydrothermal fluid, either within chimney structures or in the underlying sulfide mound. There is no evidence for this at Wetar. However, the gypsum and native sulfur δ34S values, suggest oxidation of excess H2S in the hydrothermal fluid and local oxidation of sulfides may be important in their formation.

δ18O

The sulfate oxygen data vary between δ18O of 5.8‰ and 11.3‰ (Table 4), with a mode at 9‰. As the barite δ18O data are shifted to values both higher and lower than the composition of seawater, the sulfate oxygen data suggests isotopic exchange has occurred. Nevertheless, the δ18O mode of 9‰ coincides with the δ18O value of seawater sulfate and suggests that the bulk of barite precipitated in equilibrium with seawater.

The δ34S and δ18O stable isotope data suggest the predominant source of sulfur in the sulfides was derived from the volcanic rocks in the basement, with the sulfur in the sulfate largely derived from seawater and reaffirms the importance of hydrothermal fluid and seawater mixing in the formation of VMS systems (Teagle et al. 1998a, 1998b; Roberts et al. 2003).

87Sr/86Sr

Whole-rock 87Sr/86Sr analyses were completed on variably altered host volcanic rocks, and barite separates from the massive barite bodies, fractures within the sulfide mounds and iron oxide-barite pipe structures. The barite data range between 87Sr/86Sr 0.7076 and 0.7088 with no systematic variation according to setting (Fig. 17). The 87Sr/86Sr ratios of the barites are considered to reflect mixtures between Miocene seawater (87Sr/86Sr 0.70849, Farrell et al. (1978)) and hydrothermal fluid. However, the majority of the data plot close to the value of Miocene seawater suggesting a significant contribution of seawater to the barite formation.
Fig. 17

Histogram summarizing the range in 87Sr/86Sr for Wetar Island sulfate and whole rock data. Data from Herrington (1996) is included. s syeno-granite; d dacite

Notably, one analysis shows a 87Sr/86Sr ratio of 0.70644 (Herrington 1996), from the stockwork of an undeveloped mineral prospect, Batu Kapal (Fig. 3). This value suggests that locally the hydothermal end member value may be less than 0.70644, and also indicates that less radiogenic basement may be involved in the source of Sr to this hydrothermal fluid.

The whole-rock data of variably altered basement show 87Sr/86Sr values between 0.7074 and 0.7116. These values are typically more radiogenic than MORB, and are consistent with data reported from surrounding islands (Whitford et al. 1977; Margaritz et al. 1978; Varekamp et al. 1989; Vroon et al. 1993, 2001). The more radiogenic values of 0.71165 and 0.71106, from dacitic flows, suggest a significant contribution of continental crust and or sediment in the generation of these post-mineralization magmas (Vroon et al. 2001; Elburg et al. 2002). Values of up to 0.72227 are reported by McCulloch et al. (1982), however, the nature of these samples is uncertain.

The unaltered volcanic samples tend to show 87Sr/86Sr ratios between 0.70748 and 0.70781 and Sr values >160 ppm (160–388); which are substantially changed during alteration. The progressively altered samples show a significant decrease in whole-rock Sr concentration to <60 ppm with a concomitant increase in the 87Sr/86Sr value from 0.70746 to 0.70833. The most highly altered samples show Sr concentrations and 87Sr/86Sr signatures that suggest that the samples have undergone complete isotopic exchange with seawater (Fig. 18).
Fig. 18

87Sr/86Sr v ppm Sr from variably altered Wetar volcanic rocks. Increased alteration coincides with lower ppm Sr and more radiogenic 87Sr/86Sr ratios, reflecting increased interaction with Miocene seawater. The syeno-granite and post-mineralization dacites are omitted from this figure

Ar–Ar age determinations

Three samples were chosen for Ar–Ar age determination in order to better constrain the age of mineralization and volcanic events of the area. Samples of the following intrusive and volcanic rocks were analyzed: (1) a syeno–grantite intrusion (sample no. 111999) collected from the Kali Lurang river in the Meron area; (2) a post-mineralization capping dacite flow (sample no. 097167) from Meron; and (3) a fine-grained illite (sample no. 056896, <2 μm size fraction) collected from the hydrothermally altered footwall volcanic rocks at Lerokis at a depth of 26 m below the mineralization. Herrington (1993) reported a K/Ar age of 4.7±0.16 Ma for the illite sample.

Ar–Ar data are given in Table 5 and shown as apparent age spectrum diagrams in Fig. 19. Errors are quoted at the one standard deviation level and include the uncertainty in monitor age (Hb3gr 1072±11 Ma). An isotope correlation diagram of 39Ar/40Ar versus 36Ar/40Ar (not shown) reveals that the trapped 40Ar/36Ar ratios are lower than the present day atmospheric ratio (295.5) constrained most precisely by the syeno–granite sample to be 284±3. Previously, Nagy et al. (1999) determined similar anomalously low 40Ar/36Ar ratios from dacite flows and attributed them to either a minor contaminant at m/z 36 in the mass spectrometer, or fractionated atmospheric argon within the samples. The age obtained from the illite sample of 4.93±0.21 Ma is within error of the K/Ar age of the same sample (4.7±0.16 Ma) reported by Herrington (1993). The age of the syeno–granite intrusion and dacite flow are 4.73±0.16 Ma and 2.39±0.14 Ma, respectively (Fig. 19).
Table 5

Stepped heating Ar/Ar data for biotite grains and illite separates (<2 μm), Wetar Island, Indonesia. Amounts of Cl and K obtained from measured 38ArCl and 39ArK using the Hb3 gr monitor and the parameters of α = 0.542±0.01, β = 4.37±0.03 and J = 0.017093±0.000026

Sample No.

Location

Sample type

Weight (mg)

Cl (ppm)

K (Wt%)

40Ar*×10−6 cc/g

Age (Ma)

111999

Meron area

Biotite

12

2,288±6

7.35±0.01

1.35±0.4

4.73±0.16

097167

Meron

Biotite

11

9,404±7.1

8.73±0.01

0.75±1.2

2.39±0.14

056896

Lerokis

Illite

9.1

13.88±0.90

7.90±0.01

1.38±0.05

4.93±0.21

Fig. 19

Age versus cumulative 39Ar released during stepped heating for the Wetar Island samples

The age data indicates that the spatially related syeno–granite intrusion (proximal to the Meron prospect) and the mineralization are the same age and therefore implies that this intrusive event supplied heat to the hydrothermal system. The Ar–Ar age of the illite is within error of the previously published conventional K–Ar age (Herrington 1993) and indicates mineralization of the Lerokis deposit occurred between 4.7 Ma and 4.9 Ma. The age of the post-mineralization dacite flow indicates that volcanism continued at least as recently as 2.4 Ma, which is the proposed age for the collision and accretion of the Australian continental margin with the Outer Banda Arc in the region of Timor (Richardson and Blundell 1996). The sample may record a period of extension within the Inner Banda Arc as a direct result of compressional tectonics to the south in the Outer Arc. The age of the dacite flow also indicates that the debris flow that overlies the Meron deposit also occurred post-2.4 Ma. This new age data coupled with the Sr data confirms an increasingly contaminated source region under Wetar Island, progressively modified by subducted continental material (SCM) related to tectonic events further to the south.

Evolution of the hydrothermal system

The data collected suggest that the Wetar massive sulfide and barite deposits were formed on the flanks of a volcanic edifice during the development of the Inner Banda Arc. Observations of the volcanic stratigraphy and tectonics suggest the Wetar edifice initially formed around 12 Ma due to extensive rifting and associated volcanism within oceanic crust. The mineralization is associated with bimodal volcanism, on a basement of basalts and basaltic–andesite, which most likely formed around 5 Ma, given the dates of the overlying mine sequence. The major sulfide mounds show talus textures and are localized on faults, which provide the main pathway for high temperature hydrothermal fluids and the development of associated stockworks (Fig. 20a). Within the massive sulfide mound much of the pyrite is arsenian (up to 6.7 wt%), and given the established relationship between arsenic and gold content of pyrite (Cook and Chryssoulis 1990; Cline 2001; Pals et al. 2003) may represent an initial reservoir for Au subsequently remobilized by later hydrothermal fluids responsible for the barite–gold ore. The pyrite δ34S data suggest that the sulfur is sourced from basement arc volcanic rocks, modified by a subduction zone component, which is also suggested by the whole rock 87Sr/86Sr data. The slightly elevated δ34S values, compared to arc values, indicates a component of reduced seawater sulfate during pyrite precipitation. The hydrothermal fluids responsible for sulfide precipitation produce a well zoned, intensive alteration sequence (Fig. 12b) with illite–smectite centered on the mineralization and chlorite alteration deeper and distal to the ore zones. The heat source driving the hydrothermal convection is most likely intrusive syeno–granite bodies at depth, which from Ar/Ar dating are known to be coeval with mineralization.
Fig. 20

Schematic evolution of the Wetar deposits: a T1, the Wetar deposits initiated as typical volcanogenic massive sulfides with a zoned footwall alteration predominantly propylitic to argillic in character (Kuroko like). b T2, the barite deposits originate as a peripheral ‘vent’ system, with fluids circulating through the sulfide mound and undergoing significant mixing of seawater. As the system evolves conductively cooled hydrothermal fluids circulate beneath the massive sulfide mound generating the alteration and reflecting the passage of more oxidized and acidic fluids. This results in the argillic to advanced argillic alteration observed. This is also the major Au-precipitation phase. c T3, the sulfide and barite system is preserved beneath limestones and lahars, prior to exhumation from the ocean floor, due to continued collision of the Australian continental margin and the Outer Banda Arc

Following massive sulfide development, barium rich fluids are discharged as white smokers, from the hydrothermal system and in particular at the margins of the sulfide mound (Fig. 20b). The distribution of barite suggests these fluids exploit many of the fracture systems previously employed to develop the massive sulfide mound. For example, barite infiltrates the base of the sulfide mound and fills any voids and fractures present. The δ34S barite data suggest that the sulfate in the barite is predominantly seawater derived, whereas, the barium is most likely derived from the destruction of feldspars, within the andesites and felsic volcanic rocks of the basement. Fluid inclusion data show that the hydrothermal fluids were at around 250–270°C, with no evidence of boiling. However, the salinities are greater and less than seawater, suggesting super-critical phase separation may have taken place, prior to egress on the seafloor. The form, location, isotope and fluid inclusion data of the iron oxide–barite structures, strongly suggest they are the palaeofluid conduits for the barite deposits. Gold is significantly enriched in the barite mineralization and is closely related to its formation and the most intense phases of alteration. The presence of high levels of arsenic in the barite ore matrix suggest that ‘zone refining’ of the initial arsenian pyrite may be important. It is questionable whether black smoker mineralization was still occurring at this time.

The limited nature of the covellite dominated assemblage, and accompanying δ34S, δ18O data, suggest that no significant contribution of magmatic volatiles was involved in the formation of the Wetar orebodies. For example, no isotopically light δ34S values were observed for sulfides or sulfates compared to Hine Hina or Conical Sea-Mount (Herzig et al. 1998) at least no contribution that could be detected beyond the copious amounts of seawater that must have circulated in the system. There is no compelling evidence that the fluids responsible for mineralization boiled at the sites of deposition, however, they may have been subject to supercritical phase separation, which is becoming an increasingly recognized phenomenon in modern vent systems.

The sulfide and barite orebodies are preserved on the seafloor by the subsequent precipitation of chert, gypsum and limestone and, perhaps most significantly, by the accumulation of lahars and debris flows (Fig. 20c). A dacite flow, within the lahars and debris flows gives an age of 2.4 Ma and records continued volcanism on the Wetar edifice post-3 Ma.

Discussion and comparisons

Wetar deposits in comparison with modern systems

The fluids responsible for barite precipitation with associated gold and sulfosalt assemblages show strong similarities to late stage hydrothermal fluids reported elsewhere in modern seafloor deposits. For example, late stage hydrothermal fluids rich in Pb, As, Sb, Ag, +Au were reported from Axial Seamount (Hannington and Scott 1988); with up to 6.7 ppm Au reported from an associated sulfosalt assemblage. Furthermore, galena, anglesite and sulfosalts of Pb, Ag, As, and Sb are frequently associated with low-T venting fluids, and the mineralogy of white smokers are typically enriched in barite, sphalerite and consistently report elevated Au values (Koski et al. 1984).

The Spire at Axial Seamount shows both high- and low-Fe sphalerite, with the higher values associated with the main stage ore formation (Hannington and Scott 1988). This is also the case for Wetar with low-Fe sphalerite restricted to the later fracture network. These data are consistent with a fluid evolution that evolved from a reducing fluid at low pH and high T (350°C) to a relatively oxidizing fluid at high pH and lower T due to cooling and mixing with seawater.

Gold is commonly associated with low-T fluids in modern hydrothermal systems. High concentrations of gold are reported in low-T Zn–Ba–SiO2 precipitates (Hannington et al. 1986; Hannington and Scott 1988; 1989). In the Zn-rich chimneys of Snake Pit, Cd, Pb, Sb, Ag and Au are considered to have directly precipitated in the Zn-sulfides, with the highest Au contents (>500 ppb) observed in the Zn-rich chimneys and massive Zn-rich sulfides at the surface of the deposit (Foquet et al. 1993). At the JADE hydrothermal field, Au enrichment correlates well with the barite content of the samples, with minute rounded Au grains observed between barite crystals (Halbach et al. 1989, 1993). The Au-rich samples also showed higher concentrations of As, Ba, Sb, SiO2 and Ag, similar to Wetar. The similarity between the Wetar gold mineralization and observations from active white smoker systems is striking, suggesting there was a significant role for such fluids in the origin of the mineralization.

Au-rich volcanogenic massive sulfides and Wetar

The mineralogy of the ore deposits at Wetar is highly analogous to that reported from back-arc spreading centers, e.g, Lau Basin, where visible gold was first documented in a white smoker chimney (Herzig et al.1993). In particular, the mineralogy and precious metal content of the Wetar deposits are strongly comparable to hydrothermal vent fields developed on island-arc or continental crust, e.g. Okinawa Trough (Halbach et al. 1989, 1993). The importance of the Au-composition of the igneous basement in the generation of Au-rich VMS is debated. Herzig and Hannington (1993) suggest that back-arc lavas are not significantly enriched in gold compared to MORB, and that a gold-enriched source is not a prerequisite to the development of gold-rich VMS systems. However, Moss et al. (2001) investigating the Manus Basin, suggest that the Au-enriched arc lavas, typically at 6 ppb compared to 1 ppb and below for MORB, may have an important influence on the Au-forming potential of the system. Although not developed directly on continental crust, the isotopic data for Wetar provide strong evidence for a significant component of continental crust and or sediments in the generation of the volcanic edifice that hosts the VMS mineralization. Using simple mass balance equations, the amount of basement required to be stripped of gold to produce the Kali Kuning deposit significantly increases from 0.5 km3 for a basement of 6 ppb Au to ~3 km3 for a basement with only 1 ppb Au. These values climb to 0.8 and 5 km3, if extraction efficiency rates more in keeping with experimental work are assigned (Moss et al. 2001). Such a dramatic reduction in the rock volumes required to generate the Au-mineralization suggests that the Au content of the volcanic basement may well play an important role in the generation of Au-rich VMS. Herzig and Hannington (1993) note that gold appears most abundant in sulfides associated with immature seafloor rifts in continental or island-arc crust, settings dominated by calc-alkaline volcanic rocks, including andesites, dacites and rhyolites. Notably, the Au-rich VMS system at Boliden is thought to have developed within calc-alkaline to dacite rocks within an island-arc located on continental crust or a thin continental margin (Vivallo and Claesson 1987; Allen et al. 1996; Billstrom and Weihed 1996). Similarly, the Au-rich Eskay Creek deposit formed within a mid-Jurassic arc of calc-alkaline to dacitic rocks that developed on an earlier Triassic arc and Palaeozoic volcanic and sedimentary rocks (Macdonald et al. 1996).

Summary and conclusions

The massive sulfide and barite–gold mineralization of Wetar Island, Indonesia, provides a significant perspective on the formation of Au-rich barite in arc situations. The mineralization at Wetar relates to arc volcanism triggered as the Banda Arc and Australian continental plate collide. Progressive stages of mineralization are recognized, dominated by the early arsenian pyrite with associated chalcopyrite and minor barite. Later sulfosalts, low-Fe sphalerite and further stages of barite and then oxide mineralization developed, leaving a complex mineralization over intensely altered footwall volcanic rocks showing argillic and local advanced argillic alteration close to mineralization. The presence of oxide-cemented barite ores and iron oxide-cemented barite ‘pipes’ are evidence of the later, oxidized hydrothermal fluid dominated by seawater sulfate. These oxidized fluids released arsenic and gold from earlier arsenian pyrite precipitating free gold with barite and complex iron-bearing arsenates with significant contained silver and mercury.

Acknowledgements

PS acknowledges the generous financial support of BHP-Billiton, The Natural History Museum and The University of Southampton. The logistical and financial support of BHP-Billiton through Chris Farmer, David Hopgood, David First and James Macdonald is particularly acknowledged. Tony Fallick, John Murray, Andy Barker, Robin Armstrong, Ernie Rutter and Damon Teagle provided key insights throughout the project. The SUERC is funded through support of the Natural Environment Research Council (NERC) and the Scottish Universities. AJB is funded by NERC support of the Isotope Community Support Facility at SUERC. We acknowledge careful review of the manuscript by Donna Sewell and an anonymous referee.

Copyright information

© Springer-Verlag 2005