The Evolution of a Porphyry Cu-Au Deposit, Based on LA-ICP-MS Analysis of Fluid Inclusions: Bajo de...

0361-0128/01/3206/1743-32 $6.00 1743

IntroductionHYDROTHERMAL fluid transport associated with shallowcrustal magmatism is a key process contributing to the for-mation of major metal resources. Among the magmatic-hy-drothermal ore deposits, porphyry Cu-Au-Mo deposits arethe economically most important today and have been the

subject of intensive exploration and research over the lastdecades. The formation of porphyry-style and other mag-matic-hydrothermal orebodies requires a subtle interplay be-tween large-scale tectonics, magma generation, and emplace-ment in the upper crust, advective flow of hydrothermal fluidof diverse provenance, and complex chemical reactions be-tween fluids, precipitating ore minerals, and host rocks un-dergoing hydrothermal alteration. The complexity of theseprocesses demands an integrated approach of geologic, geo-chemical, and experimental investigations (cf. Hedenquistand Richards, 1998).

The Evolution of a Porphyry Cu-Au Deposit, Based on LA-ICP-MS Analysis of Fluid Inclusions: Bajo de la Alumbrera, Argentina

THOMAS ULRICH,,* DETLEF GNTHER,**ETH Zrich, Isotope Geochemistry and Mineral Resources, Department of Earth Sciences, ETH Zentrum NO, 8092 Zrich, Switzerland

AND CHRISTOPH A. HEINRICHETH Zrich, Isotope Geochemistry and Mineral Resources, Department of Earth Sciences, ETH Zentrum NO, 8092 Zrich, Switzerland,

and Faculty of Mathematics and Natural Sciences, University of Zrich, Ramistrasse 17, 8006 Zrich, Switzerland

AbstractThe chemical and physical evolution of magmatic to hydrothermal processes in the porphyry Cu-Au deposit

of Bajo de la Alumbrera (northwestern Argentina) has been reconstructed with a quantitative fluid inclusionstudy. Fluid inclusion petrography, microthermometry, and single inclusion microanalysis by Excimer laser ab-lation-inductively coupled plasma-mass spectrometry (LA-ICP-MS) are combined to determine the evolutionof pressure, temperature, and ore metal concentrations (including Cu and Au) in the fluids. Complementaryhydrogen and oxygen isotope analyses are used to further constrain the water sources in the evolving system.The combined data provide a new level of insight into the mechanisms of metal sourcing and ore mineral pre-cipitation in a porphyry-style magmatic-hydrothermal system.

Based on previously reported observations of the igneous geology, alteration geochemistry, and veining his-tory of the subvolcanic porphyries at Alumbrera, the distribution of fluid inclusion types in space and time isdocumented. Six major inclusion types are distinguished. The highest temperature brine inclusions (up to750C; P >1 kbar) are mainly recorded in barren quartz magnetite veins in the core of the alteration system.These polyphase brine inclusions (halite sylvite + multiple opaque and transparent daughter crystals) are in-terpreted as the most primitive magmatic fluid recorded at the level of the deposit. They are of moderatelyhigh salinity (5060 wt % NaCl equiv) dominated by NaCl, KCl, and FeCl2, and contain on average 0.33 wtpercent Cu and 0.55 ppm Au. Upon cooling and decompression, these saline liquids exsolve a vapor phase,which is preferentially enriched in Cu relative to its main salt components but probably plays a minor role inthe formation of this particular deposit because of the inferred small mass fraction of vapor. Cooling and de-compression from the highest initial P-T conditions down to about 450C causes magnetite K silicate alter-ation but no saturation in Au or Cu sulfides, as recorded by continually high ore metal concentrations in thefluid inclusions. Coprecipitation of Cu and Au as chalcopyrite and native gold ( some early bornite) occursover a narrow range of decreasing fluid temperature. With cooling from ~400 to 305C, the Cu concentrationin the brine drops by about one order of magnitude to less than ~0.07 wt percent, without a proportional de-crease in major salt components. Ore mineral precipitation extracts ~85 percent of the Cu and Au from thefluid. It is associated with potassic alteration, as shown by a concomitant decrease in the K/Na ratio of the cool-ing magmatic brine and by an increase in its Ba and Sr concentrations (elements which are probably liberatedin the destruction of calcic igneous minerals).

The fluid chemical data demonstrate that the metal ratios in this and probably many other porphyry-styleore deposits are primarily controlled by the magmatic source of the ore brines. On the other hand, the final hy-pogene ore grade of the deposit is controlled by the efficiency of ore mineral precipitation. At Alumbrera,metal extraction is governed by the efficiency of cooling a high flux of magmatic fluid within a small rock vol-ume. Dilution of residual magmatic fluids, as recorded by aqueous fluid inclusions of decreasing salinities andtemperatures below 295C, follows after the main stage of copper introduction and is associated with feldspar-destructive (phyllic) alteration. Geometric relationships, fluid analyses, and stable isotope data together indi-cate that phyllic alteration results from postmineralization hydrothermal activity involving minor mixing be-tween meteoric water, residual brine, and a waning input of magmatic vapor.

Economic GeologyVol. 96, 2001, pp. 17431774

Corresponding author: e-mail, [email protected] *Present address: Geology Department, The National University of Aus-

tralia, 0200 Canberra, Australia.**Present address: Laboratory for Inorganic Chemistry, ETH Zentrum

CAB, 8092 Zrich, Switzerland.

The key role of hydrothermal fluids in the formation of oredeposits associated with shallow intrusions was recognizedvery early (e.g., Lindgren, 1905), and numerous studies onfluid inclusions have demonstrated many common featuresamong porphyry copper deposits worldwide (e.g., Roedder,1971; Nash, 1976; Chivas and Wilkins, 1977; Etminan, 1977;Eastoe, 1978; Wilson et al., 1980; Weisbrod, 1981; Preece andBeane, 1982; Reynolds and Beane, 1985; Quan et al., 1987;John, 1989; Bodnar and Cline, 1991; Cline and Bodnar, 1994;Beane and Bodnar, 1995). High-temperature fluid inclusionswith very high salinity are generally associated with low-den-sity vapor inclusions and indicate a magmatic contribution tothe metal-transporting fluids (Bodnar, 1995). Stable isotopedata of igneous phases, alteration minerals, and fluid inclu-sions have shown that magmatic as well as meteoric watersare involved in porphyry copper systems (e.g., Taylor, 1974;Batchelder, 1977; Solomon, 1983; Bowman et al., 1987; Dilleset al., 1992; Oreskes and Einaudi, 1992; Rye, 1993; Sheets etal., 1996; Hezarkhani and Williams-Jones, 1998). Fluids fromboth sources may be essential for ore formation (Henley andMcNabb, 1978), but the present weight of geologic, experi-mental, and fluid chemical evidence indicates that the metalsare introduced dominantly by magmatic fluids (Candela andHolland, 1986; Heinrich et al., 1992; Bodnar, 1995; Cline,1995; Shepherd and Chenery, 1995; Hedenquist andRichards, 1998; Ulrich et al., 1999). However, the sources andthe geochemical role especially of low- to intermediate-salin-ity fluids of variable density remain ambiguous (Shinoharaand Hedenquist, 1997; Heinrich et al., 1999).

Quantitative information about the major and trace ele-ment composition of the metal-transporting hydrothermalfluids has been a limiting factor in reconstructing the mag-matic-hydrothermal enrichment process in time and space.Petrographic observations of ore mineral daughter crystalsprecipitated inside fluid inclusions (Roedder, 1984, see refer-

ences therein) and more recent microanalytical data (Dia-mond et al., 1990; Boiron et al., 1991; Heinrich et al., 1992;Mavrogenes et al., 1995; Damman et al., 1996) show thatmagmatic fluid inclusions can contain very high ore metalconcentrations. With the recent development of laser abla-tion-inductively coupled plasma-mass spectrometry (LA-ICP-MS; Gnther et al., 1998; Audtat et al., 2000a, b), a sin-gle inclusion analytical technique of sufficient sensitivity,quantitative accuracy, and routine applicability has becomeavailable for extensive studies of texturally complex samples(Audtat et al., 1998). In this paper we report the first studyof a world-class magmatic-hydrothermal ore deposit, whereLA-ICP-MS analyses of some 300 individual fluid inclusionsin 27 samples are integrated with conventional fluid inclusiontechniques and other geochemical data within a good geo-logic framework, as reported in Ulrich and Heinrich (2001).The aim of this study is to quantify the behavior of ore metalsand other elements involved in hydrothermal mass transferby high-temperature brines, vaporlike fluids, and low-salinityaqueous solutions. By combination with stable isotope datawe determine the sources of fluids and metals independentlyand define the mass transfer and the dominant causes formetal precipitation in the orebody.

Geologic SettingThe Cu-Au deposit of Bajo de la Alumbrera in northwest-

ern Argentina (lat 2730' S, long 6660' W) is hosted by sev-eral high K dacite porphyries that intruded into andesiticflows and pyroclastic rocks of the Faralln Negro VolcanicComplex. From crosscutting relationships and petrographicobservations, several porphyries can be distinguished (Table1) that form a composite intrusive complex (Ulrich and Hein-rich, 2001; J. M. Proffett, writ. commun.). The Ar/Ar dates ofbiotite from the earliest mineralized P2 Porphyry (7.10

1744 ULRICH ET AL.

0361-0128/98/000/000-00 $6.00 1744

TABLE 1. Relative Timing Relationships between the Porphyry Intrusions at Bajo de la Alumbrera

Porphyry Timing relationships Alteration/mineralization

Los Amarillos Porphyry Cut by Early P3 and Northwest Porphyries Intensely feldspar destructive; barren to weakly mineralized

NE Porphyry Cut by Early P3; no other timing relationships Moderate potassic, chlorite epidote; weakly to moderately mineralized

P2 Porphyry Cut by Early P3 Porphyry and all younger porphyries; Quartz-magnetite, strong potassic; strongly mineralized associated with purple quartz veins

Early P3 Porphyry Cuts veins of P2 Porphyry; is cut by Late P3 and Partly quartz-magnetite, strong potassic; moderately to younger porphyries strongly mineralized

Quartz-Eye Porphyry Cut by Late P3; contains most of the purple quartz veins Intensely feldspar destructive; weakly to moderatelymineralized

Late P3 Porphyries Cuts Early P3 Porphyry; is intruded by later porphyries Weak-moderate potassic, weak feldspar destructive; weakly(Campamiento and mineralized North Porphyries)

Northwest Porphyry dikes Intrudes P2 and Early P3 Porphyries Weak potassic; barren

Post-Mineral Porphyry Cuts Late P3 Porphyries Commonly unaltered, slightly feldspar destructive; barren

Summary of the geologic relationships among the different porphyry intrusions; for a complete description of the geology see Ulrich and Heinrich (2001)and J.M. Proffett (unpub. data)

0.13) and a postore dike (6.83 0.07 Ma) constrain the lateMiocene age of mineralization (Sasso and Clark, 1998). TheP2 and the Early P3 Porphyries are the two distinct intrusionshosting the bulk of the economic copper-gold mineralization,which was introduced by at least two separate magmatic-hy-drothermal fluid pulses into a dense network of quartz mag-netite-chalcopyrite veinlets. The slightly later Quartz-EyePorphyry is overprinted by a stockwork of characteristic pinkquartz veins that are associated with intense feldspar-destruc-tive alteration in outcrops. These veins contain only minorbornite and chalcopyrite. At least four subsequent stages ofminor intrusion are weakly altered and essentially unmineral-ized, and all have much lower vein densities than the P2, theEarly P3, and the Quartz Eye Porphyries (Ulrich and Hein-rich, 2001; J. M. Proffett, writ. commun.; Table 1).

The alteration sequence is characterized by an early quartz-magnetite assemblage in the center of the deposit. This alter-ation is concentrated on the P2 Porphyry, but a secondweaker phase of this alteration also affects parts of the EarlyP3 Porphyry. The adjacent potassic alteration (K feldspar +biotite magnetite) is most intensely developed in the min-eralized porphyries but affects in a less intense form (sec-ondary biotite) some of the later intrusions, such as theNorthwest Porphyry dikes. This alteration is surrounded by apropylitic zone at the fringe of the deposit, overprintingmainly the andesitic volcanic rocks and some porphyry dikes.Concentric zones of quartz-magnetite, potassic, and propy-litic alteration of the mineralized stock and all later intrusionsare overprinted by feldspar-destructive alteration, which ad-vanced from the top of the deposit downward along steepcrosscutting structures and lobes. Mineralization is domi-nated by chalcopyrite and intimately intergrown gold. Oremineral precipitation is associated with potassic alteration andoccurs as disseminations in the porphyries and proximal an-desitic volcanic rocks and as texturally late phases in quartz magnetite veinlets. Economic Cu and Au concentrationsoccur in a high-grade ore shell above and around a barrencore of highly quartz-magnetite-altered but sulfide-poorporphyries.

Sample Material for Fluid Inclusion and Isotope StudiesSamples with veins and phenocrysts for petrographic, fluid

inclusion, and stable isotope studies were taken from originalsurface outcrops and the developing mine, but most sampleswere collected from diamond drill core (Appendices 1 and 2).The selection of samples was based on core logging and sur-face mapping of the main alteration zones (quartz-magnetite,potassic, propylitic, and feldspar-destructive alteration) andthe distribution of high Cu-Au grades. The vertical samplingextends to a depth of 550 m below the premining landsurface.

Vein densities and crosscutting relationships were mostclearly observed in the P2, Early P3, and the Quartz-Eye Por-phyries. Fluid inclusion and isotope data were derived fromsamples of the major vein and alteration types in these intru-sions. The Early P3 Porphyry is intimately associated with thesecond major pulse of mineralization and potassic and quartz-magnetite alteration. It provided the preferred and some ofthe best fluid inclusion material. Exceptionally well preservedfluid inclusions were also found in a 5- to 8-cm-thick vuggy

vein of coarse-grained purple quartz from a block of Quartz-Eye Porphyry in the southeastern part of the open pit (sam-ple BLA97). Euhedral crystals up to 1 cm contain inclusionassemblages in a particularly clear time sequence. Biotite in-clusions indicate that this quartz crystallized during potassicalteration. Pyrite and possibly anhydrite (now gypsum) weredeposited later in the vugs, associated with feldspar-destruc-tive overprint of the wall rock.

Fluid inclusion assemblages in the small quartz veinlets re-lated to feldspar-destructive and chlorite-epidote alterationwere inadequate for analysis. Suitable inclusions probably rep-resenting the feldspar-destructive stage were observed on sec-ondary trails in strongly feldspar destructively overprintedsamples, even though the quartz host in these samples belongsto veins of the potassic stage. Secondary trails are very com-mon in these samples and cut trails of primary or pseudosec-ondary inclusions preserved from the potassic veining stage.

For a reconnaissance stable isotope study, selected miner-als were separated from veins and whole-rock samples to ob-tain an enriched mineral fraction of 90 percent purity ormore. Fluid inclusion samples were also used for isotopicanalysis, as far as possible. Quartz phenocryst separates of theleast altered intrusion of the Northwest Porphyry were ana-lyzed to obtain an isotopic signature representative for themagmas. Coexisting quartz and magnetite from the quartz-magnetite alteration zone were used to estimate both the18O composition and the temperature conditions of the ear-liest alteration stage at Alumbrera. To estimate the oxygenisotope composition of fluids associated with the main alter-ation zones, we analyzed the quartz veins related to potassicalteration in three samples of the Early P3 and Quartz-EyePorphyries and quartz separates from late veinlets and matrixquartz associated with feldspar-destructive alteration. Sec-ondary biotite was analyzed for 18O in three wall-rock an-desites with intense potassic alteration, and two albite sam-ples were analyzed to represent the chlorite-epidote zone.

Hydrogen isotopes were determined on biotite of potassi-cally altered rocks, on chlorite of the chlorite-epidote zone,and sericite and/or clay fractions from samples with feldspar-destructive overprint. Hydrogen isotopes of fluid inclusionswere analyzed after thermal extraction of inclusion waterfrom veins related to the quartz-magnetite, the potassic, andthe feldspar-destructive alteration.

To obtain an estimate of the local meteoric water at thetime of hydrothermal activity, we analyzed quartz and fluidinclusion extracts from the post-sulfide stage of an epithermalvein of the Faralln Negro gold mine, located 15 km north-west from Alumbrera along the same structural trend(Malvicini and Llambias, 1963).

Petrography and Distribution of Fluid Inclusions

Fluid inclusion types

Petrography and all subsequent single inclusion measure-ments strictly concentrated on fluid inclusion assemblages,i.e., closely associated groups or trails of inclusions with visu-ally identical phase ratios and similar shape (Goldstein andReynolds, 1994). Six types of fluid inclusions were distin-guished according to the number, the nature, and the volumeproportions of phases present at room temperature (Fig. 1).

EVOLUTION OF PORPHYRY Cu-Au DEPOSIT, BAJO DE LA ALUMBRERA, ARGENTINA 1745

0361-0128/98/000/000-00 $6.00 1745

1746 ULRICH ET AL.

0361-0128/98/000/000-00 $6.00 1746

FIG. 1. Photomicrographs of inclusion types and relative timing among assemblages. A. Polyphase brine inclusion withhematite and three unidentified daughter phases. B. Opaque-bearing brine inclusion characterized by one or two opaquedaughter crystals. C. Simple brine inclusions containing only a halite daughter crystal, which has a similar or slightly largersize than the vapor bubble. D. Two-phase aqueous inclusion (325 m) with a larger or similar proportion of liquid thanvapor. E. High-density vapor inclusions, commonly with a chalcopyrite crystal (1545 m). F. Low-density vapor inclusionscontaining more than 80 percent vapor. G. A low-density vapor trail is cut by a similar trail with small high-density inclusions(sample BLA97). H. Fluid inclusion relationship in a phenocryst that is crosscut by a quartz veinlet (stippled line shows thecontact). The small polyphase and opaque-bearing brine inclusions (I) occur exclusively in the phenocryst (see also Fig. 2A).J. and K. Trails of small polyphase brine inclusions are cut by later low-density vapor trails.

polyphase brine opaque-bearing brine brine

high-density low-density

aqueous

hematite

halite halite

halite

Low-densityvapor inclusions

High-densityvapor inclusions

chalcopyrite

Vein quartz



Polyphasebrine inclusions

Polyphasebrine inclusions

Quartz phenocryst

20m 50m

20m50m

20m 20m20m

20m 20m20m A B C

FED

G H

I

J K

Three inclusion types are halite saturated, generally haverounded isometric or negative crystal shapes, and are typically20 to 30 m in size. The polyphase brine inclusions contain adense packing of halite and up to five other transparent andopaque daughter phases (Fig. 1A). Sylvite crystals, alwayssmaller than the halite crystal, were only identified with cer-tainty in a small number of inclusions, as the only isotropiccrystal other than halite. Metallic phases including hematite(red, platy) and chalcopyrite (opaque, triangular) are invari-ably present in polyphase brine inclusions. The size of mea-sured inclusions ranges from 10 to 50 m. Some of thesehighly saline inclusions show minor variations in occurrenceand nature of daughter crystals even among clearly coeval in-clusions in an assemblage, probably due to non-nucleation ofinternal precipitates (non-nucleation; Roedder, 1984). Theopaque-bearing brine inclusions have only a halite crystal plusone or two opaque phases, one of which can be hematite. In-clusion size ranges from 5 to 40 m (Fig. 1B). The simplebrine inclusions are between 5 and 35 m large and contain ahalite cube as the only daughter crystal (Fig. 1C).

Aqueous inclusions have a high filling ratio (avg 80% liquidand 20% vapor) and occur as irregular-shaped to rounded in-clusions (Fig. 1D) in texturally late quartz crystals or as sec-ondary inclusions in earlier quartz. They rarely are larger than15 m.

Vapor-rich inclusions were divided, according to their fill-ing ratio, into high-density vapor inclusions (6090% vapor;Fig. 1E) and low-density vapor inclusions (generally >90%vapor; Fig 1F). Both types commonly reach 25 m in size andhave slightly rounded to perfect negative crystal shapes. Liq-uid occupying less than 20 percent is generally invisible dueto internal reflection of the inclusion walls, except in particu-larly large inclusions near protrusions. High-density vapor in-clusions often contain a small chalcopyrite crystal (Fig. 1E).No daughter crystals have been observed in low-densityinclusions.

Relative timing of fluid entrapment

The sequence of fluid entrapment was inferred from thedistribution of inclusion assemblages in veins of different par-agenetic stages (Ulrich and Heinrich, 2001), from petro-graphic observations of crosscutting trails of inclusions alongindividual fractures (Fig. 1G, J, K), or rare primary inclusionsor pseudosecondary trails that can be attributed to specificgrowth zones in euhedral vein crystals (Fig. 1H, I; see Roed-der, 1984; Goldstein and Reynolds, 1994). However, the com-monly random distribution and superimposition of numerousinclusions in most samples made the determination of relativetiming between inclusion assemblages very difficult.

Veins related to quartz-magnetite and potassic alterationphases of the P2 and Early P3 Porphyries have indistinguish-able inclusion populations, and we infer (but cannot unam-biguously prove) that the fluid history of the two main miner-alizing pulses was similar. Rare crosscutting relationshipsbetween inclusion trails indicate that the polyphase brine in-clusions are earlier than the other brine inclusion types, andthat trails containing only polyphase brine inclusions are ear-lier than trails containing simultaneously trapped polyphaseand vapor inclusions. No direct crosscutting relationships in-volving simple brine inclusions were observed, but these are

considered later than the polyphase and opaque-bearingbrine inclusions, because of their generally lower homoge-nization temperature in individual samples.

Vapor inclusions occur as trails and clusters and low-, aswell as high-density, types were found together withpolyphase brine inclusions in boiling assemblages. These aresingle trails containing vapor- as well as liquid-rich inclusions,trapped from a two-phase fluid in one microfracture. The for-mer coexistence of brine and vapor is unequivocally demon-strated by such boiling trails, but polyphase brine and vaporinclusions are more commonly randomly distributed in largenumbers in the core of quartz crystals. The possible coevaltrapping of such assemblages is difficult to prove. Some as-semblages of high-density vapor inclusions in phenocrystsand veins show early and probably primary character, whereaslow-density vapor inclusions generally occur on sharp sec-ondary trails cutting through entire grains, or in parageneti-cally late veins, indicating a later stage of entrapment.

Some samples from the Quartz-Eye Porphyry have pre-served a particularly clear record of fluid entrapment, and in-clusion assemblages from sample BLA97 (Fig. 2B) havetherefore been singled out in the subsequent data presenta-tion. Quartz phenocrysts contain a generation of small (10m) polyphase brine inclusions with one or more transparentdaughter crystals and a prominent red hematite flake. Thesame type of inclusion was also found together with rare high-density vapor inclusions (Fig. 1E) in a primary growth zone inthe core of the euhedral quartz crystal in the vuggy vein. Thecrystal records a complex fluid history starting with primaryor early pseudosecondary polyphase brine inclusions withoutcoexisting vapor, followed by later polyphase brine inclusionswith high-density vapor inclusions. Subsequent pseudosec-ondary polyphase brine plus high- and low-density vapor in-clusions are associated with biotite precipitation, before thecrystal is rimmed by a thin final growth zone containing ir-regular aqueous and low-density vapor inclusions. The latterfluids slightly predate or are coeval with the precipitation ofpyrite in the residual vug and probably with feldspar-destruc-tive alteration of the wall rock.

Aqueous inclusions are generally late relative to all otherliquid-rich inclusions and to all phases of quartz magnetitestockwork veining. They were found in late anhydrite veinsand as secondary inclusions in earlier quartz veins occurringin samples with an intense feldspar-destructive overprint.Here they commonly occur together with low-density vaporinclusions but never in clear boiling assemblages.

Deposit-scale distribution of fluid inclusions

The distribution of the different inclusion types on thescale of the deposit was studied in an attempt to delineate therelationships between fluids, alteration, and ore deposition.The relative abundance of inclusions was visually estimatedfor each type of inclusion in a total of 38 polished sections(Table 2). The schematic section (Fig. 3) shows the ore-gradedistribution and alteration zones in section 47, with additionalsample and alteration data projected from neighboringsections (see Ulrich and Heinrich, 2001; J.M. Proffett, writ.commun.).

Halite-saturated inclusions generally exceed 50 percent ofthe inclusions in samples from the central part of the deposit.


0361-0128/98/000/000-00 $6.00 1747

The samples with the largest numbers of polyphase andopaque-bearing brine inclusions are associated with the low-grade core and the quartz-magnetite, as well as the potassic,alteration. Simple brine inclusions are not very abundant andshow no association with the alteration zones nor any correla-tion with ore grades.

Toward the fringe of the deposit, polyphase and opaque-bearing brine inclusions diminish and vapor-rich inclusions,invariably associated with aqueous inclusions, become pre-dominant. This zoning is correlated with the feldspar-de-structive overprint of some of these samples. Aqueous inclu-sions throughout the deposit are secondary and clearly belongto the incipient feldspar-destructive overprint in many sam-ples. Petrography and abundance distribution of the differentinclusion types possibly indicate a trend of decreasing salinityfrom the center to the periphery of the deposit and from earlyto late fluid stages in any one sample. The salinity and tem-perature distribution is further quantified in the followingsection on the basis of microthermometric measurements.

Microthermometry

Experimental approach and evaluation of measurements

Heating and freezing experiments were conducted on aLinkam THMS600 stage for homogenization temperature upto 600C and on a Linkam TS1500 stage for temperatures

above 600C. The precision for freezing runs is about 0.1Cand for heating runs 2C at 500C. The uncertainty for high-temperature measurements with the Linkam TS1500 stagemay be up to 20C. Apparent salinities are reported in weightpercent NaCl equivalent, based on the halite solubility equa-tion for halite-saturated inclusions and on the final ice melt-ing for halite-undersaturated inclusions (Bodnar and Vityk,1994). Minor CO2 detected by laser Raman microprobe insome of the high-density vapor inclusions may cause errors inthe calculation of salinities, but clathrates were generally ab-sent or minor in abundance at the temperature of final icemelting (see Table 3 for abbreviations).

Microthermometric results reported in Tables 4, 5, 6, and allfigures strictly refer to fluid inclusion assemblages. The salinitywithin a single assemblage is quite uniform (typically 3 wt %NaCl equiv) but varies greatly between different assemblages.

Freezing measurements and heating experiments up to thetemperature of halite dissolution were carried out before LA-ICP-MS analysis to obtain the apparent salinities needed forinternal standardization of elemental analyses (see below).Final homogenization above 600C led to stretching and de-crepitation of some inclusions. Limited high-temperaturehomogenization data were therefore measured with inclu-sions remaining in the same assemblage after LA-ICP-MS oron similar assemblages from other parts of the same section.Average homogenization temperatures of these were then

1748 ULRICH ET AL.

0361-0128/98/000/000-00 $6.00 1748

biotite

pyrite

opaque-bearingbrine inclusions

primary and pseudo-secondary? polyphasebrine inclusions

high-densityvapor inclusions

polyphasebrine inclusions


small polyphase andopaque-bearing brineinclusions,somehigh-densityvapor inclusions

growth zone with irregularsmall aqueous andvapor-rich inclusions

low-and high-densityvapor inclusions

low-densityvapor inclusions

1 mm

vein quartz

phenocryst

magnetite

aqueousinclusions


secondary polyphasebrine inclusions

polyphase brine andhigh-density vaporinclusions

low-densityvapor inclusions andpolyphase brine inclusions

primary? polyphase andopaque-bearing brineinclusions with hematite

polyphase andhigh-densityvapor inclusions

high-densityvapor inclusions and opaque-bearing brine inclusions

heterogeneoustrapping of vapor-richand polyphasebrine inclusions


polyphase brineinclusions andlow-densityvapor inclusions

1 mm

Sample 45-61.3 S Sample BLA 1/97 A, II

FIG. 2. A. Sketch of a phenocryst cut by a veinlet (sample 45-61.3 S). Small, hematite-bearing inclusions (smallest dots)are restricted to the phenocryst, whereas later generations of polyphase and opaque-bearing brine inclusions also occur inthe vein. B. Simplified drawing of crystals from sample BLA97 showing the relative timing relationships of saline and vaporinclusions. The polyphase and opaque-bearing brine inclusions occur in the core, whereas the vapor inclusions are on trailsand commonly along the outer zone of the crystal.


0361-0128/98/000/000-00 $6.00 1749

Tabl

e 2.

Dis

trib

utio

n of

Flu

id I

nclu

sion

s in

Sam

ples

Use

d in

Thi

s St

udy

Vapo

r in

clus

ions

Sam

ple

Dep

thO

re g

rade

sO

paqu

eH

igh-

Low

-ID

(m

bel

ow

Cu

Au

Poly

phas

ebe

arin

gB

rine

Aqu

eous

dens

ityde

nsity

(Fig

. 3)

Dri

ll ho

lesu

rfac

e)(w

t %)

(ppm

)Po

rphy

ry ty

pe a

nd a

ltera

tion

(%)

(%)

(%)

(%)

(%)

(%)

AK

51-6

1.1/

652

40.

020.

03Q

uart

z-m

agne

tite,

P2

Porp

hyry

1025

3510

20A

M51

-61.

1/11

547

0.01

0.02

Qua

rtz-

mag

netit

e, P

2 Po

rphy

ry10

2560

5J

47-5

5.3/

1014

71.

401.

95Q

uart

z-m

agne

tite,

P2

Porp

hyry

405

510

40A

J51

-61.

1/5

523

0.02

0.03

Qua

rtz-

mag

netit

e w

ith c

alci

te fi

lling

, P2

Porp

hyry

2525

1010

30A

45-6

1.3/

1342

81.

402.

43Po

tass

ic w

ith q

uart

z-m

agne

tite,

P2

Porp

hyry

B35

105

4010

H47

-55.

3/7

110

0.50

0.78

Pota

ssic

, P2

Porp

hyry

5015

35I

47-5

5.3/

914

61.

401.

95Po

tass

ic, P

2 Po

rphy

ry25

105

1050

AF

51-5

2.2

/920

40.

010.

02Po

tass

ic, P

2 Po

rphy

ry60

55

525

AG

51-5

2.2

/11

211

0.04

0.09

Pota

ssic

, P2

Porp

hyry

505

1515

510

AH

51-5

2.2/

1321

20.

040.

09Po

tass

ic, P

2 Po

rphy

ryB

5020

1515

AI

51-5

2.2/

1422

40.

110.

11Po

tass

ic, P

2 Po

rphy

ry45

1510

2010

D47

-53/

4822

60.

560.

40Po

tass

ic, P

2 Po

rphy

ry10

255

305

25M

49-5

0.1/

435

70.

090.

04Po

tass

ic, P

2 Po

rphy

ryB

5515

1515

N49

-50.

1/8

371

0.05

0.05

Pota

ssic

, P2

Porp

hyry

4010

2030

AL

51-6

1.1/

853

70.

010.

07Po

tass

ic, P

2 Po

rphy

ry25

1045

20

AE

50.3

-59.

4/8

146

0.74

1.28

Qua

rtz-

mag

netit

e, E

arly

P3

55

525

60F

47-5

3/72

413

0.51

0.75

Qua

rtz-

mag

netit

e, E

arly

P3

1030

1050

R49

-52/

322

60.

050.

14Q

uart

z-m

agne

tite,

Ear

ly P

3B

1530

1510

1020

V49

-52/

1933

40.

100.

06Po

tass

ic w

ith q

uart

z-m

agne

tite,

Ear

ly P

35

1015

6010

B47

-53/

2511

80.

060.

10Po

tass

ic, E

arly

P3

B40

2025

510

C47

-53/

3012

80.

100.

17Po

tass

ic, E

arly

P3

555

205

15E

47-5

3/62

311

0.30

0.23

Pota

ssic

, Ear

ly P

310

4030

155

T49

-52/

1026

00.

080.

06Po

tass

ic, E

arly

P3

1050

515

515

W49

-52/

2133

70.

050.

05Po

tass

ic, E

arly

P3

B15

1045

525

L49

-50.

1/1

354

0.05

0.02

Pota

ssic

, Ear

ly P

3B

1525

355

515

Y49

-61.

4/16

277

0.58

0.80

Fel

dspa

r de

stru

ctiv

ely

over

prin

ted,

Ear

ly P

330

510

55Z

49-6

1.4/

1928

31.

401.

94F

elds

par

dest

ruct

ivel

y ov

erpr

inte

d, E

arly

P3

515

1010

60A

D50

.3-5

9.4/

689

0.72

2.12

Fel

dspa

r de

stru

ctiv

ely

over

prin

ted,

Ear

ly P

310

55

525

50A

A49

.2-4

6.3/

427

80.

800.

53F

elds

par

dest

ruct

ivel

y ov

erpr

inte

d, E

arly

P3

2020

55

2525

AB

49.2

-46.

3/5

279

0.80

0.53

Fel

dspa

r de

stru

ctiv

ely

over

prin

ted,

Ear

ly P

320

2060

AC

49.2

-46.

3/8

292

0.77

0.71

Fel

dspa

r de

stru

ctiv

ely

over

prin

ted,

Ear

ly P

315

1020

55S

49-5

2/9

242

0.06

0.05

Slig

htly

feld

spar

des

truc

tivel

y ov

erpr

inte

d, E

arly

P3

B

3025

105

1020

U49

-52/

1431

50.

140.

16Sl

ight

ly fe

ldsp

ar d

estr

uctiv

ely

over

prin

ted,

Ear

ly P

35

1015

505

15O

49-5

1/1

182

0.94

1.55

Slig

htly

feld

spar

des

truc

tivel

y ov

erpr

inte

d, E

arly

P3

6010

30P

49-5

1/3

184

0.85

1.50

Slig

htly

feld

spar

des

truc

tivel

y ov

erpr

inte

d, E

arly

P3

1015

530

40Q

49-5

1/5

192

0.91

1.48

Slig

htly

feld

spar

des

truc

tivel

y ov

erpr

inte

d, E

arly

P3

355

2040

AO

BL

A 9

7Su

rfac

eF

elds

par

dest

ruct

ivel

y ov

erpr

inte

d,

Qua

rtz-

Eye

Por

phyr

yB

5015

520

10A

NB

LA

18

Surf

ace

Qua

rtz

phen

os in

feld

spar

des

truc

tivel

y ov

erpr

inte

d,

Qua

rtz-

Eye

Por

phyr

y10

55

550

25

Sele

cted

sam

ples

that

wer

e us

ed to

del

inea

te th

e di

stri

butio

n of

the

diff

eren

t inc

lusi

on ty

pes

in th

e de

posi

t; th

e ab

unda

nce

of e

ach

incl

usio

n ty

pe w

as e

stim

ated

opt

ical

ly fr

om d

oubl

e-po

lishe

d se

c-tio

ns; s

ampl

e ID

ref

ers

to th

e fr

eque

ncy

diag

ram

s of

Fig

ure

3; B

= s

ampl

e w

ith b

oilin

g as

sem

blag

e

assumed for the analyzed inclusions. Measuring several as-semblages of identical petrographic appearance within onesection generally showed consistent homogenization temper-atures, indicating that this assumption is valid.

Results

The highest temperature was measured in an assemblage ofopaque-bearing brine inclusions that homogenize to the liq-uid at 771 16C (assemblage 92, Table 4). In general, high-est homogenization temperatures of polyphase brine and

opaque-bearing brine inclusions are around 710 to 730C(Figs. 4A, 5F, G). Among the simple brine inclusions, only theassemblage of sample BLA97 homogenized above 600C, at671 6C (assemblage 104). Two high-temperature boilingassemblages (assemblages 114, 115) showed consistent ho-mogenization of the polyphase brine inclusions at 720C, andthis value was adopted for assemblages 113 and 117, whichdid not homogenize below the upper temperature limit of theheating stage (600C). Homogenization of the polyphase,opaque-bearing, and brine inclusion assemblages occurs ei-ther by halite melting or to liquid (Tm(halite) > Th(L) or Th(L)).

Polyphase brine inclusions: The data for all polyphase brineinclusions in Figure 5A and F show a wide range in homoge-nization temperature and salinity values with a main mode at400 to 500C and a consistent high-temperature mode above700C, which mainly represents the assemblages of sampleBLA97. The assemblage with the lowest Th(L) homogenizes at432 6C and has a salinity of 47 1 wt percent NaCl equiv,whereas the assemblage (62) with the highest salinity (71 1wt % NaCl equiv) homogenizes by halite dissolution (Tm(halite)> Th(L) = 580 3C). Within some assemblages, inclusions

1750 ULRICH ET AL.

0361-0128/98/000/000-00 $6.00 1750

TABLE 3. Nomenclature of Fluid Inclusion Measurements

Th(L) Final homogenization temperature to liquidTm(halite) > Th Final homogenization by halite dissolutionTh(V) Final homogenization to vaporTm(ice) Last ice melting temperatureTm(CO2) Melting temperature of CO2Tm(halite) Melting temperature of halite

TABLE 4. Microthermometry Data and Pressure Estimates

Salinity (wt % Th(L) and Th(halite) Basis for P Assemblage Sample NaCl equiv) (C) P (bars) estimate

21 49-52 19A 3 0.3 240 0.3 50 LV curve4 46-59QZE 7.1 4.2 242 2.6 50 LV curve

20 49-52 14 1.4 0.3 271 20.7 80 LV curve25 51-61.1 II,5/1 17.7 0.1 292 10.0 80 LV curve

110 49-52 3/3 35 0.5 305 5.0 80 Boiling curve100 49-52, 19B/3 41.4 3.9 368 39.6 2810/1501 Halite disappearance101 49-52, 19b/3 IV 42.1 3.7 387 43.2 150 LV curve106 45-61.3, 13/1 43.9 2.8 397 6.5 180 Boiling curve108 49-50,1, 1/8 41 4.7 400 14.1 200 Boiling curve111 49-52 3/3 48.4 4.0 419 26.2 250 Boiling curve109 49-50,1, 4/6 46 3.8 421 1.0 220 Boiling curve73 47-53, 62/1 44 0.3 423 22.5 220 LV curve

112 49-52,21/3 48.4 6.0 448 3.5 270 Boiling curve54 49-51, 5/5 50.9 4.3 463 19.0 320 LV curve

116 BLA1/97A I,2 52.7 3.0 470 14.1 320 Boiling curve58 51-61.1, 5/ 3 49.9 1.8 471 18.4 340 LV curve59 51-61.1, 5 IV 50 0.6 499 12.6 380 LV curve

107 47-53/25B 61.7 2.5 530 0.3 370 Boiling curve57 51-61.1 5 I 45.1 3.3 540 69.1 530 LV curve62 BLA1/97B,2 71.1 0.2 580 3.0 1135/3801 Halite disappearance47 47-53 25B 61.4 2.9 6652 620 LV curve

104 BLA1/97B I,1 36.9 2.8 671 6.0 1100 LV curve113 51-52,2, 13 52 0.4 7203 1050 Boiling curve117 BLA1/97B,2 65.5 2.1 7203 680 Boiling curve115 BLA1/97A,1 63.8 1.9 7202 750 Boiling curve114 BLA1/97A,1 58 1.1 7202 900 Boiling curve61 BLA1/97B2 66.1 3.6 7222 680 LV curve67 BLA IV 97, 3/1 64 1.6 7223 750 LV curve66 Avg BLA IV 97, 3 62 1.2 7223 800 LV curve68 BLA IV 97, 2/6 56 4.0 7223 950 LV curve93 BLA2/97 B2 53 1.5 7302 1080 LV curve91 BLA 1/97B II,1 36.9 2.8 7303 1280 LV curve92 BLA2/97B2 46.5 3.0 771 16.3 1320 LV curve

Microthermometric data and pressure estimate of boiling assemblages and single-phase assemblages that are shown in Figure 6 and which were subse-quently analyzed by LA-ICP-MS; sorting of the table is based on increasing homogenization temperatures (roughly corresponding to a sequences of late toearly assemblages); pressures for nonboiling assemblages are minimum pressures

1 Pressure calculated according to Bodnar (1994) for halite homogenizing inclusions and the assumed pressure, respectively (see text for explanation)2 Single analysis due to decrepitation of inclusions from different assemblages3 Extrapolated values from measurements of similar but different assemblages (see text for explanation)


0361-0128/98/000/000-00 $6.00 1751

47-51.2

47-5347-55

47-55.347-62

ENEWSW49-.2-46.3 49-50.1

49-52

49-51

51-52.2

50.3-59.4

49-61.4

51-61.1

45-61.4

Deposit-scale distribution of fluid inclusion types

m a.s.l. GypsumFault

Ste

ves

Fault

potassic high-gradeore zone

low-gradecore

coexisting polyphasebrine andvapor inclusions

polyphasebrine incl.

opaque-bearingbrine incl.

aqueous incl. brine incl.

low-densityvapor incl.

high-densityvapor incl.

2600

2400

2300

2200

strong K-feldspar-biotite alterationmoderate andweak K-feldspar-biotite alterationsecondarybiotite alterationweak secondarybiotite alteration inlate porphyries

late feldspar-destructivealteration

quartz-magnetitealteration

chlorite-epidotealteration

Cu-grade contour0.5 wt%approximateboundaries ofmain porphyriesagainst surroundingandesites

average abundance

AA

AB

AC

O

P

Q

AOAN

AJ

B

C

HAD

AE

Z

I

J

Y

AF

AI

D

N

EW

U

S

A

F

AM

AL

R

AH

AG

L

AK

M

qtz-magalteration

13 samples22 samples 16 samples9 samples 16 samples

?

?

?

feldspar-destructive

2500

?

?

FIG. 3. Schematic illustration of section 47through the deposit, with alteration and ore-gradezones from Ulrich and Heinrich (2001) and pie di-agrams showing the relative abundance of the in-clusion types in single samples. Samples with boil-ing assemblages are indicated by white centers,and ticks show the coexistence of low- or high-den-sity vapor inclusions with polyphase brine inclu-sions. Key shows average overall abundance of thedifferent inclusion types in the low-grade core (22samples), quartz-magnetite alteration (9), potassicalteration (16), the high-grade ore zone (16), andthe feldspar destructively altered parts of the de-posit (13). Letters in the pie diagrams correlatewith sample ID of Table 2.

100

200

300

400

500

800100 200 300 400 500 600 700

Boilingassemblages

105

57

78

93

115/117

61/6465

66/67

6970

7172

114

104

92

89

84

90

56

5859 96

48 50

107

91

113

5376102

108/11197

75110

98

88

77

74

100

87

82

83

106

103

6051

101112

109/52

94

73

55

85

62

79

116/54

68

47

49

8663

95

99

80

B

900

81

Tm h

alite

(C

)

Th L (C)

100 200 300 400 500 600 700 800 900

Aqueous inclusions

High density vapor inclusions

Brine inclusions

Opaque-bearingbrine inclusions

Polyphase brine inclusions

calcite

anhydrite

0

10

20

30

40

50

60

70

80

anhydrite

7172

6970

113

91105 104

5662

64

65-68

92

7858

57

79

60

89

4847

50

49

93114

115/117

90

84

43-46

116

5376

59 96

80

88102

108/11197

75110

98

103 85106

101 52

81-83/871165155

86

95/54

107

39

37

34

30

3828/29

41/423640

33

6

31/32

12

11/18

725

26

27

5

5a

21714

161

3 8

4

20

22

2413

21

9

10/15/19

23

63

7477

94

99

100 109

A

61

Sal

inity

(wt%

NaC

l equ

iv.)

Th L or by halite melting (C)

112

FIG. 4. A. Salinity vs. final homogenization temperature (to liquidor by halite melting) for all inclusion types of Alumbrera. Enclosinglines represent single assemblages from different samples. Not alltemperatures shown above 600C have been explicitly measured,some are adopted from similar assemblages (see Table 4) and slightlyshifted for clearer presentation. Stippled lines connect coexistingvapor and liquid inclusions of boiling assemblages. The gap in salin-ity around 25 to 35 wt percent NaCl equiv may result from non-nu-cleation of halite (see Roedder, 1984) or may be the artifact of devi-ations of the real fluids from the NaCl-H2O model system used forthe salinity estimate (Vityk and Bodnar, 1998). B. Halite dissolutiontemperatures as a function of vapor-liquid homogenization temper-ature to liquid for the halite-saturated inclusion types. Inclusionsplotting above the line undergo final homogenization by halite dis-solution, whereas inclusions below the line homogenize to liquid.Numbers next to the labels refer to the measured assemblages.

homogenize both to liquid and by halite disappearance over asmall temperature interval (Fig. 4B). The exceptionally highhomogenization temperatures could indicate some H2O lossdue to postentrapment reequilibration (Sterner et al., 1995;Audtat and Gnther, 1999), but no textural evidence for sucha process was observed in this sample. Even these extremevalues are probably real fluid temperatures, comparable tothose reported by Eastoe (1978) for the Panguna porphyrydeposit, Bougainville.

Melting temperatures of the daughter crystals other thanhalite were observed at 100 to 150C for the first unknownphase, which is rounded, weakly birefringent, and has a rela-tively low relief. After initial melting it reprecipitates oncooling to 90C as an irregular aggregate with strong bire-fringence. The second daughter phase is orthorhombic,weakly birefringent with a moderate relief, and it melts be-tween 200 to 250C. The third phase is small, isometric,round, and birefringent (erythrosiderite?; Wilson et al., 1980)and it melts at 320 to 370C. A small cubic daughter crystal

(sylvite?) melts at 460 to 490C always before halite (e.g., as-semblage 67). Such high melting temperatures are rarely re-ported for sylvite but were previously found in the Endeavour26 North porphyry Cu-Au deposit at Parkes, Australia (Hei-thersay and Walshe, 1995), implying very high K/Na ratios upto 1.1 by weight (Roedder, 1971; Cloke and Kesler, 1979).

In polyphase brine and opaque-bearing brine inclusions,the opaque phases never dissolve even up to 800C. The com-mon presence of hematite daughter crystals in hematite-freerocks indicates H2 diffusion out of the inclusions (Mavro-genes and Bodnar, 1994). Rare, birefringent rectangulardaughter crystals (anhydrite?) also did not dissolve to at least600C, which also may be related to postentrapment oxida-tion of sulfur.

Opaque-bearing brine inclusion: These assemblages aresimilar in salinity and homogenization temperatures topolyphase brine inclusions (Figs. 4A and 5B, G). Apart fromcommon assemblages homogenizing to liquid, there are alsoassemblages in which inclusions homogenize by halite disso-

1752 ULRICH ET AL.

0361-0128/98/000/000-00 $6.00 1752

TABLE 5. Summary of the Microthermometric Data and Estimated Pressures of the Different Inclusion Types

Description of assemblages Microthermometry data Pressure estimate Deposit-scale distribution

Polyphase brine assemblages: Highest avg Th (L) ~722C Highest P(min) = 950 bars Abundant in low-grade core, Randomly distributed inclusions Lowest Th(L) = 430C Lowest P(min) = 350 bars quartz-magnetite and potassic of primary character; secondary Salinity range 40 to 71 wt alteration trails in phenocrystals and veins % NaCl equiv are cut by boiling assemblages; assemblages homogenizing by halite melting are later than those homogenized to liquid

Boiling assemblages: Commonly Highest avg Th(L) ~720C Highest P(min) = 1,050 bars Low-grade core and high-grade ore secondary trails; coexistence of Lowest Th(L) = 300C Lowest P(min) = 80 bars zone, from present surface to at least polyphase and vapor-rich Salinity range 35 to 65 wt 430 m deeper inclusions % NaCl equiv

Opaque-bearing brine Highest avg Th(L) ~771C Highest P(min) = 1,320 bars Dominant in low-grade zone inclusions: Randomly distributed (one assemblage 92) Lowest P(min) = 220 bars high-grade ore zone inclusions of primary character in Lowest Th(L) = 360Cphenocrystals; assemblages Salinity range 37 to 63 wt homogenizing by halite % NaCl equiv melting are later

Brine assemblages: Highest avg Th(L) ~671C Highest P(min) = 1,100 bars Subordinate in low-grade as well as Only small assemblages, (one assemblage 104) Lowest P(min) = 150 bars in high-grade zone trails or randomly distributed Lowest Th (Tm (halite)) = 336C

Salinity range 36 to 46 wt % NaCl equiv

Aqueous assemblages: Highest avg Th(L) = 328C P(min) 600C Highest P(min) = 760 bars High-grade ore zone, feldspar-Secondary trails cutting polyphase Lowest Th(V) ~370C Lowest P(min) = 280 bars destructive overprint, minor in brine trails; randomly distributed Salinity range 1 to 19 wt low-grade zone inclusion of primary character % NaCl equiv

Low-density vapor assemblages: Feldspar-destructive overprint, Primary on late growth zones peripheral areas of the depositrelated to feldspar-destructive alteration; secondary trails cut polyphase brine trails


0361-0128/98/000/000-00 $6.00 1753

TAB

LE

6. E

lem

ent C

once

ntra

tions

of S

elec

ted

Ele

men

ts in

Flu

id I

nclu

sion

s (a

naly

zed

by L

A-I

CP-

MS)

Ass

em-

Type

of

Tm

(hal

ite)

Th(

L)

Salin

ity w

t %

Li

Na

K

Mn

Fe

Cu

Zn

As

Rb

Sr

Mo

Ag

Sn

Cs

Ba

La

Ce

W

Tl

Pb

Au

blag

ein

clus

ion

Petr

ogra

phic

des

crip

tion

or T

m(ic

e)(

C)

NaC

l equ

iv(p

pm)

(wt%

)(w

t%)

(wt%

)(w

t%)

(wt%

)(w

t%)

(ppm

)(p

pm)

(ppm

)(p

pm)

(ppm

)(p

pm)

(ppm

)(p

pm)

(ppm

)(p

pm)

(ppm

)(p

pm)

(ppm

)(p

pm)

92O

paqu

e-be

arin

g Q

uart

z-E

ye P

orph

yry,

str

ong

390.

577

1.5

46.5

120

10.6

5.30

1.10

8.60

0.01

0.60

5038

060

140

4058

020

9015

02

6

0037

.980

5.00

7.10

1.30

13.2

00.

030.

60n.

a.24

060

n.a.

10n.

a.20

490

6

0053

.39.

200.

0240

1754 ULRICH ET AL.

0361-0128/98/000/000-00 $6.00 1754

57Po

lyph

ase

P2 P

orph

yry,

cle

ar q

uart

z in

vug

37

6.5

540.

845

.111

06.

5010

.20

3.20

10.5

00.

320.

6010

390

5080

n.a.

n.a.

20

60n.

a.5

4n.

a.15

00

brin

e in

clus

ions

with

late

cal

cite

, sec

onda

ry tr

ail

31

.4

69.1

3.

3

20

0.56

0.

69

0.45

0.

96

0.16

0.

10

3

60

9

30

5

10

4

2

100

59Po

lyph

ase

P2 P

orph

yry,

cle

ar q

uart

z in

vug

43

9.0

499.

051

.919

.80

0.10

14

0


0361-0128/98/000/000-00 $6.00 1755

35H

igh-

dens

ity

Ear

ly P

3 Po

rphy

ry, m

oder

ate

-6.3

39

5.0

9.5

720

3.00

1.60

0.15

2.9

3.00

0.

68n.

a.70

lution or by near-simultaneous vapor and halite disappear-ance (Fig. 4B). Groups of inclusions (e.g., assemblage 91-93)homogenizing above 600C were found exclusively in sampleBLA97. The assemblage with the lowest Th(L) (363 42C)has a salinity of 38 1 wt percent NaCl equiv. The assem-blage with the highest salinity (63 4 wt % NaCl equiv) ho-mogenizes by halite dissolution at 524 25C.

Brine inclusion: Most assemblages homogenize to liquid,and their salinity is near the lower end of the salinity range ofopaque-bearing brine inclusions (Figs. 4A and 5C). Homoge-nization temperature is distinctly lower compared to thepolyphase and opaque-bearing brine inclusions, mostlyaround 340 to 400C. Assemblage 100 shows halite meltingat temperatures much higher than vapor to liquid homoge-nization temperatures, which may indicate trapping at highpressure (Bodnar, 1994; Cline and Bodnar, 1994) or someotherwise unrecognizable postentrapment modification. Theassemblage (104) with the lowest salinity (37 3 wt % NaClequiv) has the highest homogenization temperature of 671 6C, whereas the assemblage (101) with the highest salinity(42 4 wt % NaCl equiv) homogenizes at 387 43C.

Aqueous inclusions: Within a few groups of simple two-phase inclusions, hydrohalite was observed as the last meltingphase in individual inclusions (assemblages 5, 5a, 22, Fig. 4A).

This phase showed metastable behavior as described byRoedder (1984), melting above 0C (in one case up to 9.3C).The brown color of the ice in some of the aqueous inclusionsis typical for a CaCl2-bearing fluid (Shepherd et al., 1985, p109). In assemblages 4, 5, 11, and 18 of feldspar destructivelyaltered Quartz-Eye Porphyry as well as potassically altered P2Porphyry and Early P3 Porphyry, unusually large salinity vari-ations (e.g., up to 8 wt % NaCl equiv) can occur within oneassemblage of aqueous inclusions with constant filling ratiosand homogenization temperatures, suggesting mixing be-tween fluids of different salinity.

Limited microthermometry was carried out on aqueous in-clusions in calcite and anhydrite (Fig 4A, assemblages 1-3).The homogenization temperature of assemblages in anhydriteis 231 22C. The salinity of the aqueous inclusions is veryvariable and ranges from 7 5 wt percent NaCl equiv in oneassemblage to 15 9 wt percent NaCl equiv in another as-semblage. Aqueous inclusions in calcite are very rare and onlya few were suitable for heating and freezing experiments.They have very low salinity with a melting point depression of0.2 to 0.3C, which corresponds to a salinity of 0.4 to 0.5wt percent NaCl equiv. Inclusions in calcite have the lowesthomogenization temperatures measured at Alumbrera (Th(L)= 204 17C). These data are in agreement with the occur-rence of calcite in the latest stage of the vein paragenesis. Thelarger variations in homogenization temperature and salinityof aqueous inclusion assemblages in anhydrite than assem-blages in quartz is possibly due to the lower stability of anhy-drite and/or stretching of the inclusion during heating andfreezing runs.

High-density vapor inclusions: Due to the small amount ofliquid and the difficulty of seeing any phase transitions inhigh-density vapor inclusions, the determination of the ho-mogenization temperature may vary several tens or even hun-dreds (150300C) of degrees (Bodnar et al., 1985). There-fore, homogenization temperatures shown in Figures 4A and5J have a larger range and may be severely underestimated.Similarly, ice melting as a measure of apparent salinity is dif-ficult, and results vary considerably, probably because ofminor coentrapment of coexisting brine. The lowest salinitywithin one assemblage was measured at 1.6 0.7 wt percentNaCl equiv and the highest apparent salinity was 19 2 wtpercent NaCl equiv. For most boiling assemblages, the salin-ity of low- and high-density vapor inclusions had to be in-ferred from the coexisting polyphase brine inclusions, usingthe NaCl-H2O phase diagram (Sourirajan and Kennedy,1962; Urusova, 1975; Bodnar et al., 1985). High-density vaporassemblages, containing inclusions with and without opaquedaughter crystals, show the same salinity and homogenizationtemperature as assemblages where only opaque-bearingvapor inclusions occur.

In a few assemblages (31-34), clathrates were observedduring freezing runs, which indicate the presence of someCO2. Melting of CO2 occurred at Tm(CO2) = 56.1C, andclathrate melting could only be measured in one assemblageat +2.4C. The total CO2 content is probably well below 3 molpercent, the lower limit for visible liquid CO2 (Nash, 1976).

Low-density vapor inclusions: These inclusions providedno information during microthermometric experiments. Nophase changes were observed during freezing.

1756 ULRICH ET AL.

0361-0128/98/000/000-00 $6.00 1756

05

10

15

20

25

0

5

10

15

20

0

10

20

30

40

05

10

15

20

25

05

1015202530

05

1015

20

25

02468

1012

Th L and Tm halite>Th L (C)Salinity (wt% NaCl equiv.)

Opaque-bearingbrine inclusions

Brine inclusions

High-density vapor inclusions

0

5

10

15

20Polyphase brine inclusions

0 10 20 30 40 50 60 70

05

1015202530

Aqueous inclusions

A

B

C

E

F

G

H

J

ID

102030405060

800200 300 400 600 700500100

0

FIG. 5. Frequency diagrams of salinity and homogenization temperature.The modes at high homogenization temperatures of the polyphase brine andopaque-bearing brine inclusions are real and mainly from high-temperaturefluids trapped in the vein from the Quartz-Eye Porphyry (sample BLA97).

Pressure Estimation and P-T-XNaCl Evolution

Given appropriate experimental data, a unique estimate oftemperature (T) and pressure (P) conditions can be deter-mined from a boiling assemblage (e.g., Roedder and Bodnar,1980). All other inclusion data provide information aboutminimum pressures and temperatures only. The occurrenceof boiling assemblages as well as brine assemblages homoge-nizing by halite dissolution requires significant variations in Pand T along and outside the two-phase region of the multi-component fluid system. Approximate constraints on P-T-XNaCl conditions can be derived by reference to experimentsin the NaCl-H2O model system (Sourirajan and Kennedy,1962; Urusova, 1975; Bodnar et al., 1985; Fig. 6). Table 4 andFigure 6 give the pressure estimates of all boiling assem-blages and the inclusion populations that were later analyzedwith LA-ICP-MS.

The homogenization temperature and salinity of the liquidmember of the boiling assemblages are plotted as filled sym-bols in the phase diagram of the system NaCl-H2O (Fig. 6;dots for sample BLA97, squares for all other samples). Theassemblages (113115, 117) with the highest homogenizationtemperatures (~720C) indicate pressures between 1,050 and680 bars. Lower pressures (300400 bars) are represented bytwo boiling assemblages (107 and 116), and several boilingassemblages (106, 108112) indicate very low pressures of100 to 250 bars at temperatures of 300 to 450C.

Minimum pressures for liquid assemblages homogenizingby vapor disappearance to the liquid were estimated in thesingle-phase liquid field of the system NaCl-H2O and areshown in Figure 6, using open symbols with a bar extendingto higher pressure. The highest minimum pressures of >1.08to >1.32 kbars are indicated by a high-temperature brine andthree opaque-bearing brine assemblages of sample BLA97


0361-0128/98/000/000-00 $6.00 1757

1 1 0 50 100wt% NaClequiv

pre

ssu

re(b

ars)

700C

600C

550C

500C

450Ccritical curve

"liquid""vapor"

V+NaCl

100

0

200

300

400

500

600

700

800

900

1000

1100

1200

1300

1400

62

92

91

93104

61

115

116

59

57

47

10758

L+V+NaC

l

54

112

73106

108109

111

100101

25420 21

114

110

11368

sample BLA 97

other samples

31

32

33

3540

41/42

45/4644 43

400C

67

66

117

800C

boilingassemblages

91: fluid inclusionassemblage

117: boilingassemblage

minimum P estimate

FIG. 6. Phase diagram of the system NaCl-H2O (Sourirajan and Kennedy, 1962; Urusova, 1975; Bodnar et al., 1985),showing all boiling assemblages and single-phase assemblages analyzed by LA-ICP-MS. Circles describe the P-T-X variationof assemblages from sample BLA97, and squares represent other samples from the main mineralized porphyries. Filled la-bels indicate boiling assemblages (numbers underlined) and open symbols show lower limits for pressure estimates based onbrine-only assemblages. The salinity and homogenization temperature data of the vapor-rich inclusion assemblages havelarger errors due to the underestimation of homogenization temperatures and possible overestimation of the salinity due tocoentrapment of minor liquid. For example, assemblage 35 must have been trapped at pressures greater than 480 bars in thevapor field.

pre

ssur

e (b

ars)

wt% NaClequiv

(9193, 104). Slightly lower minimum pressures are esti-mated for high-temperature polyphase brine inclusions thatfall close to the 700C isotherm on which the high-tempera-ture boiling assemblages plot. In contrast, low-temperatureassemblages ( 500C) of the different inclusion types giveminimum pressures over the range of 400 to below 100 bars.

Several assemblages consistently show homogenization byhalite melting, and in some assemblages, individual inclusionshomogenize by liquid or by halite melting. Homogenizationby halite dissolution has been interpreted to indicate entrap-ment of a halite-saturated hydrothermal brine (Eastoe, 1978;Cloke and Kesler, 1979; Wilson et al., 1980), homogeneoustrapping at high pressures (Bodnar, 1994; Cline and Bodnar,1994), or postentrapment H2O loss or volume shrinkage(Sterner et al., 1988). Halite-homogenizing inclusions atAlumbrera never occur on boiling trails, which would be clearevidence for postentrapment modification (Audtat and Gn-ther, 1999). The analyzed fluid inclusion assemblages showedno petrographic evidence of postentrapment disturbance(with the possible exception of assemblage 100). We there-fore suggest that the halite-homogenizing inclusion assem-blages record the entrapment of a single-phase brine, proba-bly not halite saturated, at pressures well above theliquid-vapor phase boundary (Bodnar, 1994; Cline and Bod-nar, 1994). Minimum pressures can be derived for this type ofassemblage if the salinity is around 40 wt percent NaCl equiv,where experimental isochore data are available (Bodnar,1994). Two assemblages that homogenize by halite melting(62, 100) give unreasonably high pressures using the calcula-tion approach of Bodnar (1994). Assemblage 100, althoughwithin the salinity range of the experimental data, gives pres-sures of around 2.8 kbars due to the large difference betweenvapor homogenization and halite dissolution, but the mi-crothermometric data vary over a large range, probably indi-cating postentrapment disturbance in this case. In the follow-ing plots, this assemblage is grouped next to assemblage (101)of otherwise similar inclusions in the same sample. Assem-blage 62 indicates a minimum pressure of 1,135 bars, but theextrapolation of the experimental data to such high salinities(70 wt % NaCl equiv) and the known deviation from theNaCl-H2O model system may lead to errors in the pressureestimate. When the final homogenization temperature andthe salinity are plotted, this assemblage falls inside the halite-saturated field at ~380 bars (Fig. 6).

Table 5 summarizes the main observations of the deposit-scale distribution and the microthermometric characteristicsof the fluid inclusion assemblages. Even though it is not pos-sible to document the entire fluid history in time at any pointin the deposit, a generalized fluid evolution can be deducedon the assumption that each intrusion and its fluid pulse (in-cluding the Quartz-Eye Porphyry) followed a similar range ofP-T-XNaCl paths. Significant salinity and density variations offluids present during quartz-magnetite and potassic alterationand probably the initial stages of copper deposition can be ex-plained by a P-T path of general cooling and decompressionof one magmatic fluid. The highest pressure fluids trappedabove 600C are interpreted as the earliest inclusion assem-blages. Among these, there is a trend from nonboiling brineswith ~50 wt percent NaCl equiv toward even higher salinity,essentially following a high-temperature isotherm of the

two-phase boundary (Fig. 6). Several of these assemblagesare clear boiling trails, and the salinity variation at high pres-sure and temperature is therefore interpreted to reflectminor water loss of an initial input brine to a high-densityvapor phase. These boiling assemblages exclusively containopaque-rich polyphase brine inclusions, giving a first indica-tion of high ore metal concentrations. At the lower pressureend of this evolution, halite saturation may have occurred atleast locally, as indicated by a reversal of decreasing salinitywith decreasing pressures below 400 bars. Throughoutquartz-magnetite alteration, potassic alteration, and the asso-ciated copper ore deposition, salinities seem to be controlleddominantly by variations in P and T, in a fluid system domi-nated by magmatic liquid + vapor halite. Below 400C,many of the liquid-rich inclusions contain only one or noopaque daughter crystals, and microanalytical data presentedin the following chapter show that this later part of near-iso-baric cooling is associated with the loss of metallic compo-nents by ore metal precipitation. Large variations in salinityonly occur in late-stage, low-temperature aqueous assem-blages, at temperatures between 200 and 300C. This part ofthe fluid evolution is associated with feldspar-destructive al-teration and may involve an influx of a low-salinity fluid of ex-ternal origin mixing with diminishing amounts of magmaticvapor.

LA-ICP-MS Analysis of Fluid Compositions

Methods

After petrographic characterization of the fluid inclusiontypes and the determination of salinity and some of the ho-mogenization temperatures, the composition of selected in-clusions was analyzed by LA-ICP-MS. Representative sam-pling of the bulk composition of the now heterogeneousinclusions is achieved by stepwise increase in the diameter ofthe ablation crater, using an Excimer laser ablation systemwith imaging optics (Gnther et al., 1997, 1998). The entirecontent of the inclusion, daughter crystals as well as liquid, istransported as an aerosol into the plasma, using He as a car-rier gas. Absolute quantification of single inclusion LA-ICP-MS signals was carried out by integration of all elements in-cluding Na, correction for host mineral contributions (mostcommonly Al, Ti, Li, Ca), comparison of intensity ratios withan external NIST glass standard, and referencing the ratios tothe absolute concentration of Na estimated from microther-mometry (Gnther et al., 1997, 1998; Audtat et al., 1998;Gnther and Heinrich, 1999).

Some elements are commonly below detection or were notmeasured for every inclusion assemblage (Mg, Ca, Al, Cl, As,Ag, Mo, Sn, Sb, La, Tl, Bi, U, Au). Where available, upperconcentration limits for these elements are indicated in Table6 and Figures 7 and 8. Analysis of gold concentrations in sin-gle fluid inclusions was a major achievement of this newmethod and was conducted under optimized conditions ofthe ICP-MS, for maximum response on 197Au and a reducedelement menu (Si, Na, Au, Cu, As; Ulrich et al., 1999). Thetransient signals consistently show overlapping short peaks forCu and Au, demonstrating without any doubt that the gold iscontained within the inclusions and that it is adsorbed onto(or intergrown with) a copper sulfide daughter crystal (Ulrich

1758 ULRICH ET AL.

0361-0128/98/000/000-00 $6.00 1758


0361-0128/98/000/000-00 $6.00 1759

concentration in liquid (ppm/wt%)

Ti

Ag Bi

La

Cs

TI

Sr

Rb Ba

Cu

Zn

Mn

K Fe

NaNa

Fe K

MnCu

ZnRb

Li Ba

SrCs

TlCeBi

Ce

Tl

Ag

Sr

Mo

CsBa

Bi

Rb

Pb

Zn

Cu

Mn

K

Fe

Ba

Li

Sr

Mo

Ti

RbPb

Zn

CuMn

AgTl

LaCe

Cs

K Na

Fe

100100.11001 1100.1

1

10

100

10

0.1

100

0.1

1

TlCs

Ag

BaSr

Rb

Cu

Pb

Zn

Mn

Na

K Fe

Ce

Bi

La

Tl

Rb

Sn

Pb

Zn

Cu

Mn

K

Na Fe

B

MoCs

Li

W

Bi

SbAg

LaCe

Ti

Pb

Na

F=27

F=7

F=1

F=2

F=10

A B

C D

E F

116

Assemblage 117 115

114

112106

100100.11001 1100.1

100100.11001 1100.1 100100.11001 1100.1

100100.11001 1100.1 100100.11001 1100.1

1

10

0.1

100

10

0.1

100

1

1

10

0.1

100

10

0.1

100

1

F=8

Th L= 720CP= 900 bars






Sr

Ba

Pb

FIG. 7. Element concentrations in liquid (polyphase brine inclusions) vs. concentrations in coexisting vapor inclusions.The black line represents equal concentration in the two phases. The dashed lines connect elements with similar tendenciesto be fractionated in the liquid. F is the minimum degree of Cu fractionation based on the average value of elements clearlyenriched in the liquid (eq. 1 in text). Error bars indicate 1 of the variability within one assemblage.

1760 ULRICH ET AL.

0361-0128/98/000/000-00 $6.00 1760

1600

1400

1200

1000

800

400

200

0

600

0

10

20

30

40

50

60

70

80

BLA 97: opaque-bearing and brine inclusions

BLA 97: polyphase brine inclusions Other samples

Th

P

salinity

A

1

0.1

10

100

0.1

1

10

100

1

0.1

10

100

0.1

1

10

100

Mn

Na

PbZn

Rb

Cs

Tl

NaCl eq. B

Fe

MoSr

Cu

K

Au

Ba

Fluid inclusion assemblages

C

FIG. 8. Interpreted P-T-salinity evolution (A) and LA-ICP-MS data (B, C) of the chemical composition of fluid inclusions.Evolution sequence (from left to right) is interpreted on the basis of petrographic timing evidence, decreasing Th, and de-creasing pressures for assemblages with overlapping homogenization temperatures. The high-temperature assemblages ofsample BLA97 are further separated on the left side, based on petrographic timing evidence in a sample that probably rep-resents a separate magma and fluid pulse. Inclusion types are schematically indicated at the top. A. P-T-XNaCl data separatedaccording to relative timing relationships (BLA97) and decreasing homogenization temperature (all other samples). B. andC. The concentration variation for nonreactive and reactive elements, respectively. Stippled lines interpolate elements omit-ted from analytical runs with a reduced element menu optimized for Au. Error bars extending from the bottom of the fig-ure indicate upper limits for analyses where element concentrations are below detection. Error bars show 1 variabilitywithin an assemblage of similar inclusions.

et al., 1999). Limits of detection as low as 0.1 ppm Au in 25-m inclusions were determined, corresponding to a totalmass of about 1015g Au per inclusion (Ulrich et al., 1999).Detection limits (calculated from the standard deviation ofbackground intensity; Gnther et al., 1998, 1999) vary de-pending on inclusion volume and signal length, and many Auanalyses are close to the limit of detection. Typical uncertain-ties of 20 percent are estimated from analyses of several mi-crothermometrically identical inclusions in one assemblage,but they increase for small inclusions (

analyses of individual inclusions (open circles) in comparisonto the bulk ore ratio of the deposit (square) and ore sampleassays from section 47 (small dots, see also Ulrich and Hein-rich, 2001, fig. 10). The encircled fluid analyses have anAu/Cu ratio (0.8 ppm Au/0.7 wt % Cu = 1.2 104 by wt)identical to the bulk ore deposit (0.65 ppm Au/0.51 wt % Cu= 1.3 104; Mndez, 1997; Wall, 1997, updated 0.64 ppmAu/0.54 wt % Cu = 1.2 104; David Keough, writ. commun.,2001) and are interpreted as the closest approximation to theinput ore fluid (Ulrich et al., 1999). Many of the polyphasebrine inclusions contain similar Au contents but significantlylower Cu concentrations. Fractionation of some Cu into thevapor phase (possibly on a local scale) is the most likely ex-planation, consistent with the high Cu concentrations in manyanalyzed vapor inclusions and with their poorly determinedAu/Cu ratio of less than 5 104 by weight (calculated fromaverage limits of detection). On the other hand, some Cufractionation may have occurred into magmatic sulfide phasesin some of these samples (C. A. Heinrich, pers. commun.,2001).

Oxygen and Hydrogen IsotopesTo further constrain the fluid sources that were involved in

the formation of the Bajo de la Alumbrera deposit, a selectionof samples was analyzed for their oxygen and hydrogen iso-tope compositions. The D and 18O values of igneous biotiteand quartz of an unaltered Northwest Porphyry intrusion, and

secondary minerals including biotite, albite, sericite + minorclay minerals, quartz, chlorite, and magnetite, were analyzedto deduce the isotopic composition of the fluids associatedwith various stages of alteration. Hydrogen analyses of fluidinclusion extracts of the quartz-magnetite, potassic, andfeldspar-destructive alteration were analyzed for comparisonto the isotopic fluid signatures calculated from fluid-mineralequilibrium.

Procedure and experimental fractionation factors

The oxygen and hydrogen isotope data are expressed rela-tive to standard mean ocean water (SMOW) in per mil (Table7). The absolute errors of the measured data are about 0.2per mil for the oxygen isotopes and about 5 per mil for thehydrogen. The mineral-fluid equilibrium temperatures forthe calculation of the isotopic fluid composition were esti-mated from the fluid inclusion homogenization temperatures(Table 7). The fractionation factors used for the oxygen iso-topes of quartz and albite were according to Bottinga andJavoy (1973) and Matsuhisa et al. (1979) for the respectivetemperatures, and for biotite the fractionation factor was es-timated by subtracting quartz-water from quartz-biotite (Javoy,1977). The hydrogen fractionation factors for biotite andsericite (muscovite) are from Suzuoki and Epstein (1976),and the chlorite fractionation factor is from Graham et al.(1987). However, the temperature estimates of around 300and 350C (from fluid inclusion data) for some of the alter-ation stages are below the temperature range of experimen-tally based fractionation data (e.g., 400800C, Suzuoki andEpstein, 1976). At low temperature (below 400C) the hy-drogen fractionation curves are not well established exceptfor kaolinite (e.g., Gilg and Sheppard, 1996). For that reason,the fractionation factors (1000 ln mineral-water) at low tempera-tures for sericite-kaolinite, chlorite, and biotite were extrapo-lated from Taylor (1997, see diagram on p. 234), and are asfollows: 23 per mil (muscovite), 15 per mil (kaolinite), 35per mil (chlorite), 35 per mil (biotite). The analyzed samplesof the feldspar-destructive alteration consist of a mixture offine-grained sericite and clay minerals, therefore, an averagefractionation factor between the value for kaolinite andsericite of 19 per mil was used. The fractionation factor forbiotite depends on Fe content (Suzuoki and Epstein, 1976).An Mg-rich composition (XFe = 0.4) was assumed for the sec-ondary biotite of the potassic alteration, and a more Fe-rich(XFe = 0.7) composition was used for the igneous biotite in theunaltered sample of the Northwest Porphyry, based on thegeneral compositional trends shown by Ague and Brimhall(1988).

Results and isotopic composition of hydrothermal fluids

The measured isotopic ratios of the minerals and the calcu-lated isotopic signature of the fluid are given in Table 7. Thevariation of the measured D values among the differentalteration minerals of the porphyry samples is small, except-ing the D values for the secondary biotite of the andesiticwall rock and particularly the igneous biotite of the North-west Porphyry that are depleted (Fig. 10A). This low Dvalue of the igneous biotite from the Northwest Porphyry istypical for a melt reservoir that has already degassed much ofits fluid (Taylor, 1986), which is in agreement with the late

1762 ULRICH ET AL.

0361-0128/98/000/000-00 $6.00 1762

0.01

0.1

1

10

0.01 0.1 1 10Au (ppm)

Au/Cu=1.2*10-4

polyphasebrine inclusions

ore samples bulk ore

Cu loss

inferredinput fluid

FIG. 9. Au vs. Cu concentration for single liquid inclusion analyses (opencircles) representing some of the earliest and highest temperature inclusionassemblages in the orebody (assemblages 6668). Small dots are whole-rockassay from section 47 through the ore deposit and the white square repre-sents the bulk ore-grade ratio of the deposit. Encircled liquid-rich inclusionanalyses match the bulk ore, whereas analyses with lower Cu concentrationsmay have lost some Cu to a vapor phase or early sulfide mineral inclusions(C. A. Heinrich, pers. commun., 2001, modified from Ulrich et al., 1999).


0361-0128/98/000/000-00 $6.00 1763

TABLE 7. Oxygen and Hydrogen Isotope Composition of Minerals, Whole-Rock Samples, and Fluid Inclusions and the Calculated D and 18O Values of the Fluid in Exchange Equilibrium with These Samples

Measured T(C) Calculated fluid1D 18O D 18O

( SMOW) ( SMOW) ( SMOW) ( SMOW)

Igneous minerals

Northwest PorphyryLP Bio -109.8 (0.8) Qtz 9.2 (0.1) 700 -65.9 (2.4) 8.5 (0.4)

Hydrothermal minerals and fluid inclusions

Porphyry quartz-magnetite zone49-60/93m Fluid incl. -65.4 (9) Qtz 10.6 (0.1) 700 -65.4 (9) 9.9 (0.9)

Mag 3.1 8.3 (0.3)

49-61.4/357m Qtz 10.5 (0.04) 700 9.9 (0.9)Mag 2.8 8.0 (0.3)

Porphyry potassic zone51-52.2/308m WR (bio) -68.3 (2.7) Qtz 8.5 (0.2) 500 -29.8 (4.7) 5.3 (0.9)

49.2-46.3/452m WR (bio) -80.1 (3.8) Qtz 9.7 (0.05) 500 -41.7 (4.7) 6.5 (0.9)

BLA 97 Qtz 11.3 (0.2) 500 8.1 (1.2)

43-47.1/195m Fluid incl. -55 (22.8) Qtz 9.2 (0.2) 500 -55 6.1 (0.9)

Porphyry feldspar-destructive zone49.2-46.3/265m WR (ser) -71.1 Qtz 10.3 (0.1) 300 -52.1 (10) 3.4 (1.8)

57.5-60/51m WR (ser) -73.3 (0.2) Qtz 10.7 (0.2) 300 -54.3 (10) 3.9 (1.8)

55.5-60/192m Fluid incl. -52 Qtz 8.3 (0.08) 300 -52 1.4 (1.8)

38-49/97m Fluid incl. -79 (19.5) Qtz 10.2 (0.01) 300 -79 3.3 (1.8)

Porphyry chlorite-epidote zoneBLA55 Chl -76.6 (0.8) Ab 10.6 (0.1) 300 -41.6 (5) 4.7 (1.7)

BLA64 Chl -78.3 Ab 9.9 (0.1) 300 -43.3 (5) 4.0 (1.7)Andesite potassic49-61.4/94m Bio -79.4 (6.7) Bio 7.4 (0.04) 350 -44.4 (5) 9.4 (0.1)

37-49/275m Bio -90.1 (1.6) Bio 6.5 (0.04) 350 -55.7 (5) 8.5 (0.1)

37-49/275m WR (bio) -83.1 (1.8) WR (bio) 6.5 (0.04) 350 -48.1 (5) 8.5 (0.1)

Andesite feldspar-destructive zone46-42.5/46m WR (ser) -69.3 (15.3) Qtz 10.7 (0.2) 300 -50.3 3.9 (1.8)

Present-day meteoric water2 -30 -5

Estimate of meteoric Fluid incl. -70 (3) Qtz 11.70 200 -70 (3) 0.1 (3.2)water of Miocenc agefrom the Faralln Negro mine

Abbreviations: bio = biotite, chl = chlorite, mag = magnetite, qtz = quartz, ser = sericite, WR = whole rockMeasured 18O and D of alteration minerals and the calculated isotopic signature of the fluids in equilibrium with the rock; the numbers in brackets of

the measured isotopic signature correspond to 1 and the numbers in brackets for the calculated signature indicates the variation of the isotopic signature fortemperatures of 50C above and below the cited temperature

1Fractionation factors used for the calculation of the isotopic fluid signature in the respective temperature interval are from Bottinga and Javoy (1973),Matsuhisa et al. (1979), and Javoy (1977) for the oxygen isotopes and from Suzuoki and Epstein (1976) and Graham et al. (1987) for the hydrogen isotopes;since at low temperatures the experimental data are rare, the fractionation factors were taken from the extrapolated curves from the diagram of Taylor (1997)

2Average of present-day measurements of meteoric water at Salta, Argentina; see text for explanation

intrusive stage of the Northwest Porphyry. The oxygen valuesfor quartz and albite range from 8 to 11 per mil for the por-phyries and are lower for the secondary biotite of the an-desites (6.57.5%). This may be due to the different closingtemperatures of biotite and quartz and the higher resistanceof quartz to later isotopic exchange during the feldspar-de-structive overprint of the system. The D value for thefeldspar-destructive overprint of the andesite has a largervariation, including poor reproducibility of the duplicateanalysis. Despite the different diffusion rates and closing

temperatures for O and H of a given mineral (Giletti, 1985,1986), we used temperature data of the fluid inclusion study,because the homogenization temperatures of the fluid inclu-sions are close to or below the estimated closing temperaturesof the analyzed minerals (Zaluski et al., 1994), except for the700C of the magmatic stage. Various processes influence theisotopic composition of minerals and the fluid in a magmatic-hydrothermal system, such as fluid exsolution (Dobson et al.,1989; Taylor, 1992), phase separation (e.g., Shmulovich et al.,1999), open- and closed-system fractionation, fluid mixing,

1764 ULRICH ET AL.

0361-0128/98/000/000-00 $6.00 1764

-120

-100

-80

-60

-40

-20

0

6 8 10 12

quartz/biotite

biotite/biotite

quartz/biotite

albite/chlorite

quartz/sericite

chlorite-epidote

feldspar-destructive

potassic (wall rock)

Northwest Porphyry

18O

Measured Dand 18O values

-90

-80

-70

-60

-50

-40

-30

-20

-10

0

-10 -5 0 5 10 15

Calculated isotopic composition of the fluid

primarymagmatic water

"present-day

meteoric water"

potassic(wall rock)350C

potassic(porphyry)500C

chlorite-epidote300C

feldspar-destructive300C

NorthwestPorphyry700C

A

B

quartz-magnetite

quartz-magnetite700C

Faralln Negro mine200C

potassic (porphyry)

Fig 10. Isotopic signatures of the fluids associated with the magmatic and the alteration stages. A. Analyses of mineralpairs that were analyzed for hydrogen and oxygen, respectively. B. Calculated isotopic signature of fluids at the indicatedtemperatures for the different alteration zones and the Northwest Porphyry. Closed symbols are calculated v

The Evolution of a Porphyry Cu-Au Deposit, Based on LA-ICP-MS Analysis of Fluid Inclusions: Bajo de...

Documents

Transcript of The Evolution of a Porphyry Cu-Au Deposit, Based on LA-ICP-MS Analysis of Fluid Inclusions: Bajo de...