Evolution of ore-forming fluids in the Bukovik-Kadiica porphyry Cu deposit, Republic of Macedonia

The Bukovik-Kadiica minerali(cid:93)ed system is hosted by Tertiary dacitic and andesitic volcanic rocks (cid:90)hich have intruded the basement of (cid:51)alaeo(cid:93)oic schists in the (cid:54)erbo-Macedonian Massif of east ern Macedonia. The latest geological e(cid:91)ploration has con(cid:191)rmed


Abstract
The Bukovik-Kadiica minerali ed system is hosted by Tertiary dacitic and andesitic volcanic rocks hich have intruded the basement of alaeo oic schists in the erbo-Macedonian Massif of eastern Macedonia. The latest geological e ploration has con rmed that this porphyry copper minerali ation is characteri ed by a dominance of chalcocite and covellite ith associated chalcopyrite, emplectite, and bornite, ith the highest grades in a one of supergene enrichment. ilici cation is the dominant alteration ithin the minerali ed system, hilst ones of potassic, phyllic, argillic, propylitic, and advanced argillic alteration are also present. ilici cation and sulphide minerali ation are located in stock orks in altered dacite and andesite breccia. issolution of primary sulphides and chemical leaching are evident in the ones of o idation, dominated by limonite breccia. The main copper minerali ation has a vertical e tent bet een and m. luid inclusion studies of minerali ed uart veins have identi ed three separate groups of fluids: saline inclusions hich homogeni e at -C and have a salinity of -t aCl e uiv., vapourdominated inclusions hich homogeni e at -C and have a salinity of -t aCl e uiv., and more dilute, t o-phase vapour li uid inclusions that homogeni e at -C and have a salinity of -t aCl e uiv.

REGIONAL GEOLOGICAL SETTING
The Bukovik-Kadiica deposit is located in the Serbo-Macedonian Massif (SMM), which was described by DIMITRIJEVIC (1959) as a separate geotectonic and lithostratigraphic unit in the southern part of the Balkan Peninsula. The SMM is a crystalline basement complex boarded by the Vardar Zone to the west and by the Rhodope Massif to the east (Figs. 1 and 2). The SMM consists mainly of Palaeozoic gneisses and schists (KARAMATA, 1974a,b;KOCKEL et al., 1975, GRUBIĆ, 1980DU MUR DŽANOV et al., 2005), that have been intensely folded and faulted (BURCHFIEL et al., 2008a;ZAGORCHEV et al., 2008;ROBERTSON et al., 2009). The SMM represents a wedge of continental crust developed as part of the Morava-Rhodope zone at the margin of the Tethys with Eurasia. It is thrust westwards over the Vardar zone and east wards (Morava unit) over the Strouma unit. Precambrian (Up per Archaean to Proterozoic) and Vendian-Cambrian complexes (DIMITRIJEVIC, 1995;ZAGORCHEV & MILOVA NO-VIC, 2006) are unconformably overlain by Lower Palaeozoic, Per mian, Triassic, and Upper Cretaceous rocks. The Precambrian Ograzhdenian complex is composed of amphibolite-facies polymetamorphic and polydeformational rocks: gneisses, micaschists and amphibolites. It contains lenses of ultrabasic and basic rocks (serpentinites, harzburgites, lherzolites, norites), some of them transformed into eclogites. Migmatites and metagranites are present, as well. Recent data point toward tectonometamorphic amalgamation of very old (Late Archaean to Early Proterozoic) oceanic crust with Precambrian continental crust, a major Cado-mian tectonometamorphic event, Ordovician metagranites, and Late Hercynian post-metamorphic granites (ZAGORCHEV et al., 2008;ZAGORCHEV et al., 2015). A Vendian-Cambrian green schist-facies complex (Vlasina Group, Frolosh Formation) has a thickness estimated between 1 and 4-5 km, and composition made of chlorite and actinolite schists, metasandstones, metaconglomerates, and metadiabases. Metabasic rocks (gabbros) are also present. Zircon data indicate an age of c. 560-570 Ma (GRAF, 2001). BOYANOV et al. (1989) and DABOVSKI et al. (2002) considered the Alpine evolution of the region as a sequence of opening and closing of epicontinental basins with the formation of separate collisional orogens at the north-eastern side of the active plate (Tethys) margin, whereas the Tethys ocean itself evolved by the opening of several seaways between comparatively stable zones of similar lithologies. Although the closure of the Vardar ocean occurred in the late Mid Jurassic to Late Jurassic times, these isopic zones continued their evolution throughout the whole remaining part of the Mesozoic and most of the Cenozoic, with the youngest marine sediments of Pliocene age (ZAGORCHEV et al., 2008;DUMURDŽANOV et al., 2004) (Fig. 1).
Longitudinal zoning of mineral associations is well developed in the NW-SE direction . The most important types of mineralization are: • Skarn mineralization related to intercalations of marble in Precambrian crystalline schist within the the Serbo-Macedonian massif. Mineralization occurs in lenticular and stratabound ore bodies (e.g. the Sasa Pb-Zn deposit). • Hydrothermal stockwork mineralization developed along fault structures (e.g. Baltašnica Pb-Zn-Cu deposit). • Hydrothermal veins emplaced in different lithological settings (common but mostly small and non-economic). The Bukovik-Kadiica area is composed mainly of Riphean-Cambrian and Palaeozoic metadiabases and schists (a diabasephyllite complex, Vlasina complex) that have experienced greenschist facies metamorphism and display a well-developed schistosity. Near Berovo they directly overlie amphibolite-facies gneisses (KARAMATA, 1974A;KOCKEL et al., 1975, GRUBIĆ, 1980DUMURDŽANOV et al., 2005).
North of Belo Brdo towards Kriva Buka and Pančarevo, granitoids dominate. They are represented by granites, granite porphyries, and granodiorites. These igneous rocks build up a belt along the Macedonian-Bulgarian border. To the west, they subsided into the Pehčevo-Delčevo graben, and are covered with Tertiary (Palaeogene and Neogene) sediments. The oldest rocks present are diorites, and they are cross-cut by granite-porphyries and granodiorites. They are all thought to be of Palaeozoic age (STOJANOV et al., 1995). The granite-porphyries are a marginal facies of the Hercynian Delčevo granites. Their age is proven by the presence of granite fragments within the younger Permian and Triassic conglomerates. The granitoids are cross-cut by diabase dykes near Pančarevo. The granitoids are hydrothermally altered, most probably due to post-magmatic hydrothermal processes associated with the Tertiary volcanic rocks.
Permian sediments (sands and claystone) lie transgressively over metagabbro-diabases and granodiorites. Palaeogene products are represented by Eocene sediments that, in accordance with their lithological features and superposition of layers, can be distinguished into grey conglomerate facies (conglomerate, breccia, sand, clay, and marl) and flysch facies (yellow sand with occasional intercalations of silty clays and microconglomerates with rhythmic alternation).

MAIN GEOLOGICAL FEATURES OF THE DEPOSIT
Tertiary volcanism in the Bukovik-Kadiica area is represented by the Bukovik volcanic dome (1722 m) and small subvolcanic bodies of dacite and rhyodacite at Belo Brdo and Kadiica (1932 m) covering an area of about 4 km 2 . Small dacite bodies and dykes were discovered in Bulgarian territory within a NW-SE oriented zone, as well as a cryptodome with subvolcanic breccias at the intersection of faults of N-S and NE-SW strikes (HARKOVSKA, 1984). The Kadiica subvolcanic intrusive centre is one of several Neogene dacitic plugs that have intruded a variety of metamorphic rocks of the SMM. Several of these intrusives are associated with quartz vein stockwork development and widespread hydrothermal alteration. The subvolcanic intrusive complex is poorly exposed on Bukovic Hill (1722 m asl) and consists of flow-banded dacite and massive fine crystalline dacite that have been intruded by a coarser dacite phase, a quartz-, biotite-, and plagioclasephyric dacite porphyry, and a postulated second dacite porphyry at depth. Near-surface pyroclastics and dacitic autobreccias are exposed along the southwest side of Bukovic Hill. According to HARKOVSKA (1984), HARKOVSKA et al. (1989), and STO-JANOV et al. (1995), the Kadiica intrusive complex has been dated as between 35 and 27 Ma (Oligocene; whole rock, K/Ar method). An extensive area of brecciation consists predominantly of tectonic breccias associated with a northwest-southeast-striking major fault zone. Late stage phreatic or phreato-magmatic breccias are intercepted at shallow levels in exploration drill holes (Fig. 1).
The subvolcanic intrusive complex has been intensely altered and is cut by numerous quartz, quartz-sulphide, and sulphide veins and veinlets. Surface alteration is dominated by silicification with abundant limonite as disseminations and in veins with lesser amounts of kaolinite, sericite, and local alunite. The petrography of selected samples shows the presence of a near-surface advanced argillic overprint characterized by alunite with local diaspore and andalusite. Alteration and metal zoning suggest the presence of a deeper dacite porphyry phase that was responsible for the formation of the large stockwork exposed at Bukovic Hill. Alteration and weak copper mineralization are hosted by an Oligocene dacitic volcanic complex (TASEV, 2010), which has intruded into Palaeozoic sediments, andesites, and gabbros ( Fig. 4).
In particular, because the life span and thermal evolution of the hydrothermal system plays a major role in the understanding of the genesis of a particular hydrothermal ore deposit, this study has attempted to establish some constraints on the thermal evolution of the magmatic-hydrothermal system at the Kadiica porphyry copper deposit using fluid inclusion analysis.

HYDROTHERMAL ALTERATION
Hydrothermal alterations of the host rocks (dacite, trachydacite, and granodiorite) within the Bukovik-Kadiica deposit are widespread and distinctive and they reflect the spatial and temporal evolution of hydrothermal fluids. The overall form of the alteration system has a pin-like shape and distinctive alteration types overlap each other like layers of an onion (Fig. 4). The alterations and the porphyry Cu mineralization show an intimate genetic relationship. The overprint of older alteration zones by younger alterations reveals a sequential evolution of hydrothermal fluids and indicates that the alteration processes had a dynamic and complex character. The study and description of deep hypogene hydrothermal features have been complicated by a relatively deeply  penetrating advanced argillic overprint and by subsequent supergene alteration that modified both the deep and shallow hydrothermal features.
The distribution of hydrothermal alterations at the Bukovik-Kadiica porphyry Cu deposit differs from the common alteration zonation, with a potassic core, a phyllic alteration halo and a wider peripheral propylitic zone (e.g., LOWELL & GUILBERT, 1970;RUSK et al., 2008). In contrast, this deposit is characte rized by distant zones of potassic, phyllic, argillic, propylitic, and an advanced argillic character as well as by the presence of limonitization.
Potassic alteration is represented by the mineral assemblage composed of K-feldspar, quartz, biotite and/or magnetite, amphibole, and anhydrite. Magmatic biotite has been partly replaced by a Mg-rich variety associated with rutile. This is the most widespread alteration type in the Bukovik-Kadiica ore district, located just above the granodiorite intrusion, but the biotite halo occupies a much wider area. Usually potassic alteration is overprinted by lower temperature hydrothermal alteration assemblages. In the mineralized porphyry rocks and surrounding dacite-trachydacite, K-feldspar occurs as irregular replacements of the igneous matrix while in more pervasively altered samples it replaced the phenocrysts and destroyed the original igneous texture. Similar fea-tures have been recognized in porphyry Cu deposits elsewhere (e.g. LI et al., 2013).
Phyllic alteration overprints the potassic alteration and at places is very intensive. It may form a wide halo around the mineralization, although it is also closely associated with the Cu mineralization itself. Its vertical extensions significantly exceed the horizontal dimensions. This type of alteration is characterized by the presence of quartz, sericite, and disseminated pyrite (Fig. 5a). Sericite is present as aggregates, while orthoclase and biotite are extensively altered to muscovite and partly to secondary biotite; primary biotite is altered to sericite (Fig. 5b), and sericite has replaced plagioclase phenocrysts (Fig. 5c). This type of alteration contains a significant amount of pyrite that is directly associated with the alteration process. Phyllic alteration overlaps with the central parts of the mineralized zone (red zone shown in Fig. 4), while the argillic alteration zone characterized by kaolinite replacement progresses outwards. Similar features have been recognized in other porphyry Cu deposits such as Buchim (ČIFLIGANEC, 1993), Borov Dol (GJORGJEVIC et al., 1975, Morenci, Ajo and Bisbee (NASH, 1976;MISRA, 2000).
Argillic alteration is intensive, but distal from the granodiorite intrusion, and is characterized by newly formed kaolinite at the expense of plagioclase. The primary textures are partly or completely obscured, although quartz phenocrysts can still be recognized. Primary plagioclase was replaced by kaolinite (usually closer to the intrusion area, Fig. 5d), while montmorillonite occurs at higher levels (where hypogene mineralization is almost insignificant).
The propylitic alteration zone is the dominating feature of the great majority of porphyry Cu deposits, for example at Plavica in the Republic of Macedonia (IVANOV & DENKOVSKI, 1980;STOJANOV, 1980) as well as in deposits in the Collahausi district-Rosario, Ujina, and Quebrada Blanca, then the Los Loros, Los Pelambres, Bajo de la Alumbrera, Chuiquicamata deposits (all in Chile), (TAYLOR, 1935;SILLITOE, 1973;URQUETA et al., 2009). At the Bukovik-Kadiica deposit this alteration type is represented by a mixture of epidote, smectite chlorite, calcite, talc, and kaolinite (Figures 5e, 5f). Spatially, propylitic alteration characterizes the marginal parts of the hydrothermal system where fluid/rock ratios were the lowest. Epidotization and chloritization were mainly associated with weak porphyry mineralization.
The uppermost part of the Bukovik-Kadiica deposit is characterized by an advanced argillic alteration zone predominantly composed of quartz-, kaolinite, pyrite (±limonite), rutile and alunite (Fig. 5g). The mineral assemblage suggests extremely low pH conditions and it is a common assemblage for shallow parts of porphyry Cu deposits worldwide (e.g., SILLITOE, 1973;GUS-TAFSON & HUNT, 1975;BRIMHALL, 1979;THOMPSON et al., 1986;LINDHORST & COOK, 1990;WORMALD & PRICE, 1990). In some places, the surface of the advanced argillic alteration zone is covered by a very thin iron cap.

MINERALIZATION
The mineralized area (4 km 2 ) consists of numerous dacite and rhyolite-dacite bodies and dykes that extruded through the SMM basement of metamorphic rocks, represented mainly by chlorite-sericite schists, amphibolite and gneiss in the lowest parts ( Figure 6).
Emplacement and subsequent cooling of the complex resulted in intensive fracturing and brecciation. Mineralized bo dies have different lens-like morphologies, with a NE-SW orientation, mostly due to differences in permeability of the host struc tures.
The Bukovik-Kadiica deposit consists of of three distinctive mineralized zones: • a primary sulphide zone or hypogene mineralization • a secondary Cu sulphide enrichment zone or cementation zone • an oxidation zone Data collected from the drill cores revealed that a hypogene zone occurs at depths between 64 to 420 m, and has a variable thickness of 10 to 183 m. The Cu content ranges from 0.007 up to 0.13%. In the deeper parts, the mineralization occurs in veins which mutually intersect. The most enriched stockworks were found at greater depths (> 200 m) in dacites. So far, significant hypogene mineralization has not been confirmed.
The secondary Cu sulphide enrichment (cementation) zone ( Figure 7) occurs at depths from 60 to 260 m below the surface, with an average thickness of 70 m, and reflects the level of ground water during deposition of the secondary mineralization. The horizontal dimensions of this zone are 1200 m x 700 m. It contains 70 Mt of ore with an average grade of 0.22% Cu (B+C1 category by the Macedonian Law of Mineral Resources and appropriate rulebook). Gold and silver contents appear to be low (Au < 0.6 g/t; Ag up to 250 g/t, but usually <50 g/t). These ore reserves are economically exploitable and their calculation was based on 0.15% Cu cut-off grade (SERAFIMOVSKI, 2012). The secondary Cu sulphide enrichment zone consists predominantly of chalcocite, pyrite, molybdenite, chalcocite, bornite, covellite, dige nite, tetrahedrite-tennantite series minerals, as well as arsenosul vanite, colusite, and mawsonite (TASEV, 2010). The Cu distribution clearly defines the leached zone and an enriched copper zone with chalcocite precipitated far below the base of the oxidation zone. The chalcocite/covellite ratio changes vertically from the chalcocite predominant shallow parts to the covellite predominant deeper parts of the zone. However, the Cu distribution in the deeper part within the majority of drilled holes does not show a clear enrichment trend. Figure 8 illustrates the complete paragenetic sequence at the Bukovik-Kadiica deposit (Fig. 8).
The paragenetic sequence distinguishes an early magmatic mineral assemblage composed of rutile, magnetite, haematite and pyrite. However, the majority of the pyrite was deposited simultaneously with the propylitic alteration.
The oxidation zone is fully developed and represented by limonite-silica masses. Magnetite, chalcopyrite, rutile, haematite, and calcite have also been registered. Copper concentrations in this zone ranges from 0.003-0.12% Cu while the thickness (from the surface) varies from 20-130 m. Copper was intensively mobilized and carried out of this zone, which is a common feature of pyrite-bearing mineralizations. Oxidation of pyrite signifi-cantly decreases pH values and promotes mobilization of all metals, including Cu. A carbonate rich lithology buffers the pH value of descending ground waters and allows deposition of malachite and azurite.

FLUID INCLUSION METHODOLOGY
The mineralization in the Bukovik-Kadiica area has been studied using a variety of chemical and mineralogical techniques. Intensive exploration activity was carried out during 2011 and 2012 (including 16,500 m of drill core); and based on these findings reserves of 70 Mt at a grade of 0.216% Cu were estimated. At the same time as the exploration proceeded, detailed mineralogical and geochemical studies were performed (TASEV, 2010). More detailed descriptions of the mineralization, focusing on the fluid inclusions in mineralized quartz samples are the target of the research (drill holes 09, 10, 11 complemented with those from 01 and 08). A fluid inclusion study was conducted on 23 samples of doubly polished, transparent plates of quartz, 150 μm thick, with numerous separate fluid inclusions (5-40 μm in size). Quartz was selected from veinlets in dacites and fluid inclusions in calcites were also occasionally analysed (TASEV, 2010). Quartz samples were taken from stockworks which were exposed along a surface section over more than 100 m and from drill cores. Fluid inclusions were evenly distributed in the studied quartz grains and only those with strong indications of primary origin (GOLD-STEIN & REYNOLDS, 1994), were taken into account during the microthermometric studies (ROEDDER, 1984). The study was performed using Nikon and Olympus BX51 optical microscopes and in each analysed sample, at least 20 inclusions were analysed. Microthermometric data were obtained using a Linkam THMSG600 heating-freezing stage (temperature range -196 o C to +600 o C) and TMS 90 controller attached to a conventional petrographic microscope. The stage was calibrated using the Synflinc set of synthetic fluid inclusions and revealed a precision of ± 0.1°C for the freezing runs and ± 5°C for temperatures near to or higher than 360°C. Fluid inclusions with homogenization temperatures higher than 600 °C were carried out on a modified Leitz 1350 heating stage with precision of the measurement of ±5 °C.
Salinities are expressed as wt% NaCl equivalent and were estimated from the melting temperatures of the last crystal of ice for two-phase fluid inclusions (BODNAR, 1993) and from halite dissolution temperatures for multiphase inclusions . Eutectic temperatures were used to estimate the overall composition of the studied fluid inclusions by comparison with published data for different salt-water systems (SHEPHERD et al., 1985). In addition, calculation of KCl and possible CaCl 2 contents was performed according to the phase diagrams of ROEDDER (1984) and VANKO et al. (1988) respectively.
The pressure of heterogeneous fluids has been determined using the method of sections of isochors and isotherms. Data from THIERRY et al. (1994) were used in the construction of isochors and estimate of pressures in inclusions rich in a gaseous mixture. Salinity and pressure were determined using FLINCOR software (BROWN, 1989). In cases where primary fluid inclusions trapped heterogeneous fluid, indicative of boiling conditions, it was not necessary to perform corrections of homogenization temperatures for pressure conditions.

FLUID INCLUSION DATA FROM THE BUKOVIK-KADIICA
Petrographic studies at room temperature distinguished several types of fluid inclusions: three phase (multiphase) fluid inclusions (liquid /chloride solution/ + vapour /bubble/ + one or more translucent /halite or halite+sylvite/ or opaque daughter crystals /. The selection of fluid types ideally matches the model-type III of NASH and THEODORE, 1971; see Fig. 9a; 9b; 9c; gas-rich fluid inclusions characterized by a thick liquid rim (sometimes with halite cube daughter crystal); and undersaturated two-phase, vapour-rich fluid inclusions.
Multiphase fluid inclusions, besides liquid and vapour, contain one or more solid phases (daughter crystals). The most common daughter crystal is halite determined by its cubic habit, isotropy as well as the same relief compared to quartz. Dark opaque minerals were also observed ( Fig. 9d) but their identification was not possible using a transmitted light microscope. However, the triangular sharp habit indicates chalcopyrite morphology. This type of inclusion commonly shows a negative quartz crystal shape (Fig. 9c). Within some samples, three-phase types of fluid inclusion (L+V+S opaque ) are associated with two-phase (L+V) inclusions (Fig. 9d). Two-phase inclusions with various ratios of liquid and vapour phases have been recorded as well, but in general, V-rich inclusions are more common than L-rich ones (Fig.  9e). Fluid inclusion assemblages composed of coexisting L-rich, V-rich and multiphase inclusions suggest an entrapment from boiling fluids. Overlapping homogenization temperatures for inclusions with various phase ratios confirm the boiling environment.
The studied fluid inclusions were divided into three main types according to their appearance at 25 °C (Fig. 10): I Three phase fluid inclusions, liquid (chloride solution) + vapour (bubble) + one or more translucent (halite or halite+sylvite) or opaque daughter crystals (almost an  ideal match to the model-type III of Nash and Theodore, 1971; see Figure 10a; 10b); II Gas-rich fluid inclusions characterized by a thick liquid rim (sometimes with a halite cube daughter crystal; see Figure 10c); and III Undersaturated two-phase, vapour-rich fluid inclusions (Figure 10d).
The microthermometric data are summarized in Table 1. Temperatures are given as calculated mean temperatures for each sample. Temperatures of salt dissolution are presented within brackets together with ice melting temperatures (T mice ).
The total homogenization of brine inclusions was recorded in a wide temperature interval from 501 to 310°C (Fig. 11a). The halite melting temperature in the range between 383 and 211°C corresponds to a salinity of 45.66-32.4 wt% NaCl equiv. (Fig.  11b). The calculated fluid density varies from 1.16 to 0.90 g/cm 3 . The pressure was calculated and estimated, in accordance to the compiled data of HAAS (1976), HAAR et al. (1984), BODNAR et al. (1985, STERNER et al. (1988), BISCHOF & PITZER (1989), KNIGHT & BODNAR (1989) and ATKINSON (2002), from inclusions of saturated brines ranged from 90 to 620 bar.
Vapour-dominated fluid inclusions homogenize into a vapour phase at 497-438 °C (Fig. 12). The final ice melting tem-  Figure 12. Summary plot of the microthermometric measurements (homogenization temperatures vs. salinities) in uid inclusions from the Bukovik-Kadiica deposit. The saturation curve of NaCl and the critical curve of the H 2 O-NaCl system are given by SOURIRAJAN &KENNEDY, 1962 andHAAS, 1976; data for the saturation curve of KCl as summarized by POTTER et al. (1977).
Solid phases in some of the complex fluid inclusions were also analysed using a scanning electron microscope (SEM/EDX) (TASEV, 2010). The identity of some 'daughter' minerals is confirmed as halite (minerals with cubic habit, preliminarily determined under the optical microscope) while the other defined as sylvite (although very rare, they dissolved into liquid in the range of 75-150 O C), quite similar to other porphyry copper deposits worldwide (ANTHONY et al., 1984;ROEDDER, 1984;BOD-NAR, 1995;FAN et al., 1998;XIE et al., 2006;LI et al., 2011). The salinities of fluid inclusions containing NaCl and KCl daughter minerals (where KCl dissolutes as the last phase) were estimated by establishing the temperature of solution of NaCl and KCl and referring to the phase data of LINKE (1965) and to the relevant part of the system NaCl-KCl-H 2 O (ROEDDER, 1984) as well as using cotectic boundaries given by STERNER et al (1988).

DISCUSSION
Porphyry ore deposits represent the economically most important resources of Cu and Mo and host significant reserves of Au, Ag, and many other metals (SILLITOE, 2005). Porphyry deposits form at depths of approximately 1-6 km below the palaeosurface due to the condensation of supercritical fluids derived from a crystallizing magma reservoir in the shallow crust (e.g., SEED-ORFF et al., 2005;COOKE et al., 2013). The fluid inclusion studies have been used to estimate the magmatic-hydrothermal fluid composition and P-T conditions associated with the Bukovik-Kadiica porphyry Cu mineralization. Although fluid inclusion data provide an exceptional insight into the physico-chemical properties of ore-forming fluids, multiple episodes of magmatic-hydrothermal fluxes associated with complex processes of mineral growth, mineral dissolution, fracturing and sealing in porphyry Cu systems usually result in the entrapment of numerous generations of primary and secondary fluid inclusions within annealed host mineral grains and make interpretation of the obtained data vague (COOKE et al., 2013).
Three recorded fluid inclusion types from the Bukovik-Kadiica porphyry Cu deposit reflect the evolution of ore-bearing fluids within the P-T-X space. The results of microthermometric studies of over 450 individual fluid inclusions (see Table 1) have shown that ore-bearing fluids contained dissolved chlorides of Na, Ca, and Mg ± K. The earliest inclusions belong to the twophase (L+V) inclusion type. They are characterized by high homogenization temperatures and moderate salinities. Inclusions of this type are usually considered as relicts of intermediate-density primary magmatic-hydrothermal fluids that are exsolved from the crystallizing and cooling intrusive magmatic body (e.g., LANDTWING et al., 2010;REDMOND et al., 2004;SEO et al., 2012).
The earliest, two-phase, high-temperature, moderate salini ty fluid inclusions are commonly overprinted with one or more generations of fluid inclusion assemblages consisting of coexisting high-to moderate-temperatures and low-salinity vapour-rich inclusions and multiphase high-salinity liquid-rich inclusions. This type of fluid inclusion assemblage reflects the separation of vapour from a liquid-like supercritical fluid (boiling) followed by condensation of a liquid, often of high salinity, from a supercritical fluid (e.g., BODNAR et al., 1985, HEINRICH, 2005HEIN-RICH, 2007).
Late-stage fluid inclusions are mostly two-phase (L+V) lowtemperature and low-salinity inclusions of intermediate to high density. They typically homogenize to liquid, but some vapourrich inclusions may be present that homogenize to vapour. Such assemblages are indicative of boiling of low-salinity fluids (COOKE et al., 2013).
It is noted from the measurements that in all but one of the samples the halite crystal dissolves before the vapour bubble.
As there are low-salinity vapour inclusions coexisting with high-temperature brines in the Bukovik-Kadiica porphyry copper deposit, we suggest that these saline fluids have directly exsolved by later boiling from a low-salinity fluid, rather than by direct exsolution from the magma.
The high homogenization temperatures and salinities of the saturated fluid inclusions indicate the initial existence of a dense brine, single phase fluid (0.90-1.16 g/cm 3 , Table 1), at magmatic temperatures (ROEDDER, 1992;SANTANA et al., 2011). Coexistence of vapour like (low salinity) + two phase medium salinity + very rare halite bearing fluid inclusion at the temperature between 400 and 500 o C, however, is much more convincing for the boiling and separation of the original early high temperature phase into a heterogenous system.
After physical separation from dense brine in fractured porphyry stock (due to immiscibility), only the less viscous and more buoyant vapour phase may rise to the epithermal environment (HENLEY & McNABB, 1978;HEINRICH et al., 1999). The coexistence of vapour-rich and high-salinity inclusions that homogenize within the same temperature range (Table 1) suggests that these fluids represent two immiscible fluids that evolved from the dense magmatic fluid (SANTANA et al., 2011). Density and compositional differences between the saline and low-density fluids affects the physical and chemical behaviour of these fluids in the porphyry system, in turn affecting the distribution of the precipitated minerals.
The present study shows that just slightly above 600 o C and at a pressure above 900 bar (90 MPa), the first exsolved magmatic fluid must have been a two-phase medium-salinity fluid, L+V (Fig. 13), followed by a gradual salinity decrease during magma crystallization at low temperature, which is very similar to the data given for some other deposits (KILINC & BURNHAM, 1972;CLINE & BODNAR, 1991;SHINOHARA, 1994;CANDELA & PICCOLI, 1995;CLINE, 1995;FOURNIER, 1999;CLINE, 2003;HARRIS et al., 2005;WEBSTER AND MANDEVILLE, 2007).
Fluid exsolution from hydrous magma plays an important role in mineralization (HEDENQUIST & LOWENSTERN, 1994;WEBSTER, 1997;AUDÉTAT & PETTKE, 2003;IMAI, 2005;KAMENETSKY & KAMENETSKY, 2010). Notably early fluid exsolution is favourable for economic mineralization because it facilitates the transfer of ore-forming elements into the fluid (CANDELA & HOLLAND, 1986;ZHANG et al., 2001;LI et al., 2006). Here it is apparent that at intermediate and low pressures (< 60 MPa), and temperatures lower than 500 o C, initial low-salinity magmatic fluids gradually increased in salinity during magma crystallization as the solubility decreased with falling temperature and daughter minerals (halite, sylvite, etc.) may have nucleated. Immiscible fluids must have formed with compositions corresponding to the aforementioned Type I, Type II, and Type III. These fluids correspond to the vapour and brine phases, respectively, which formed as a result of boiling. This fluid evolution pattern is quite similar to that proposed by WILLIAMS-JONES & HEINRICH (2005) for porphyry copper deposits, and furthermore the obtained P-T-X values fit quite well into the common range for porphyry deposits elsewhere (WILKINSON, 2001).
Vapour-rich inclusions are interpreted as being coeval with multiphase brine inclusions or abundant liquid-rich inclusions, indicating that the ore-forming fluid was boiling (LI et al., 2007). The fluid inclusion study demonstrates that at some point Kadiica has experienced boiling of hydrothermal fluids. The coexistence of vapour-only and vapour-rich as well as high-salinity inclusions within the same fluid inclusion associations is considered evidence of boiling (HEDENQUIST & LOWENSTERN, 1994;HEINRICH, 2005), similar to descriptions form porphyry systems elsewhere (SOURIRAJAN & KENNEDY, 1962;ROED-DER, 1979;HEDENQUIST & LOWENSTERN, 1994;GAM-MONS & WILLIAMS-JONES, 1997;HEINRICH, 2005).
Variations in salinity indicate boiling processes, which probably led to the deposition of certain metals (GRANCEA et al., 2002). Salinity variations could also be partially attributed to the existence of explosive volcanic stages (confirmed by determined phreatomagmatic breccia at Bukovik-Kadiica; WESTRA, 2005), the fracturing of adjacent rock complexes, and similar processes.
The appearance of high-salinity inclusions, beside low-salinity ones, indicates that during particular geological periods the Bukovik-Kadiica magmatic-hydrothermal system must have generated brine fluids from deeper reservoirs (not excluding some of the afore-mentioned possibilities: salinity increase as the system evolved; boiling at certain stages or some meteoric water overprint). High-salinity fluids (fluid inclusions up to 46 wt% NaCl equiv.) probably played an important role in the polymetallic mineralization of the Bukovik-Kadiica deposit, since chloride complexes are an effective mode of metal transport (BARNES, 1979). The displayed data are very common in porphyry copper deposits. Here we would like to point out, that although some fluid inclusions have shown that Cu commonly occurs at higher concentrations in vapour-type inclusions than in coexisting brine inclusions, the transport of certain elements (Cu, Au) or compounds in low-density fluids that preferentially partitioned into the vapour phase should be excluded since recent studies confirmed that Cu concentrations in quartz-hosted fluid inclusions from magmatic-hydrothermal ore deposits do not represent pristine concentrations in the trapped fluids, but are modified by postentrapment diffusional exchange through the host quartz, where quartz-hosted fluid inclusions can diffusively loose or gain Cu after entrapment (LERCHBAUMER & AUDÉTAT, 2012;SEO & HEINRICH, 2013).
The early high-temperature fluid must have already been saturated in copper (chalcopyrite daughter crystals in FI), so that Cu precipitation probably occurred before the onset of the main alteration and mineralization event, which could explain the low copper contents in the hypogene mineralization of the Bukovik-Kadiica prospect.
The abundance of solid phases both daughter (halite and occasionally sylvite) and opaque phases (ore minerals) suggest that they were part of the original fluid, which must have been rich in ore-forming elements (LI et al., 2011). The occasional presence of daughter minerals, notably chalcopyrite, in Type III undersaturated fluid inclusions, indicates that the early derived aqueous fluid characterized by low-moderate salinity, moderate CO 2 contents, were Cu rich and responsible for the transport of Cu, Fe, and S ± Au, as suggested elsewhere (GONZALEZ-PATIDA & LEVRESSE, 2003;REDMOND et al., 2004;RUSK et al., 2008;RUSK et al., 2011). The presence of alunite suggests that one part of the sampled area is located within the outer zones of mineralization, associated with argillic alteration, which is known to occur along the margin in a porphyry system (GUILBERT & PARK, 1996), and is confirmed for the Bukovik-Kadiica deposit (TASEV, 2010).
The large homogenization temperature range of 310-627 o C is in accordance with a homogenization temperature range from 250 up to 550 o С (commonly boiling) which is characteristic for numerous porphyry copper deposits (MOORE & NASH, 1974;GUSTAFSON & HUNT, 1975;CHIVAS & WILKINS, 1977;MOORE & MOORE, 1979;AHMAD & ROSE, 1980;BEANE & TITLEY, 1981;KLEMM et al., 2007;KLEMM et al., 2008). Boiling curves for NaCl solutions (SHEPHERD et al., 1985) and their relationship to Th and depth, were used to calculate the palaeodepth of mineralization (~ 800 m) that suggests pressures typically between 300 and 1200 bar, which corresponds to a lithostatic load cover of 1-4 km (ROEDDER, 1984;Alumbrera: ULRICH et al., 2001;Bingham Canyon: REDMOND et al., 2004). However, several episodes of possible boiling effects, the system with wide fluctuation of pressure, even volcanic extrusion and presence of phreato-magmatic breccias etc., assured us that is not convenient to judge the palaeodepth.
Relatively high metal concentrations (from a few parts per million up to weight percentages) have been determined in vapour-rich inclusions formed under pressures of 200 to 1000 bars and temperatures of 400-650 o С (WILLIAMS- JONES & HEINRICH, 2005). High copper concentrations, in volcanic vapour/gas, typically in the percentage range have been estimated from the size of chalcopyrite "daughter" crystals in fluid inclusions (ROEDDER, 1971;EASTOE, 1978;HENLEY & McNABB, 1978).
Type III inclusions, with occasional small daughter opaque mineral (chalcopyrite) occurred at levels characterized by potassic alteration (its transitional parts to phyllic alteration). Two phase, liquid-rich inclusions occurred in the propylitic envelopelike alteration and some peripheral parts of the argillic alteration. Central parts of the phyllic alteration were dominated by type II inclusions while going up to lower temperatures and pressures, toward the upper parts of the phyllic and lower parts of the argillic alteration, type I inclusions prevailed. Although optional and despite the sometimes precluding nature of fluid inclusions in the interpretation of the origin of fluids (due to multistage trapping, overprinting etc.), these findings were consistent with those of BEANE & BODNAR (1995) and RUSK et al. (2008).

CONCLUSIONS
The extension, quantity, and quality, of the relatively newly discovered porphyry copper deposit of Bukovik-Kadiica, were defined at the beginning of 2013 after completion of detailed geological exploration by the Kadiica Metal company. Porphyry copper mineralization is of the stockwork type within the brecciated dacite-andesite stock which intruded Palaeozoic schists of the Serbo-Macedonian Massif. The major ore mineral within the Kadiica deposit is chalcocite, while within the mineral assemblage covellite, bornite, enargite, emplectite, pyrite, pyrrhotite, and chalcopyrite were also determined. These minerals are now mainly concentrated in the cementation zones. Alteration minerals, which dominate in the hydrothermally altered dacite breccia, are quartz, sericite, calcite, chlorite, kaolinite, illite and alunite with progressive silicification being the most dominant.
Fluid inclusions in quartz veins of the brecciated dacite-andesite volcanic rocks have shown that hydrothermal solutions consisted mainly of chlorides of Na, Ca, K and Mg with a wide range of salinity of 3-45 wt% equiv. The homogenization temperatures are within the range of 310 to 627 o C, which implies the development of primary sulphide associations at medium and high temperatures. The correlation of such high temperatures with the calculated pressures suggests that the studied fluids are related to magmatic fluids, which later transformed into ore-bearing hydrothermal fluids of pulsative character.
The first exsolved magmatic fluids (above 600-610 o C and pressure above 900 bar) were two-phase and medium-salinity fluids, followed by a gradual decrease in salinity during magma crystallization at intermediate and low pressures (< 90 MPa), as well as a decrease in temperatures to below 500 o C. Initial lowsalinity magmatic fluids gradully increased in salinity with magma crystallization as the solubility of solutes also decreased with falling temperature and separate crystals of daughter minerals could nucleate and grow. In this manner three immiscible fluids formed with compositions corresponding to the aforementioned Type I, Type II, and Type III. The appearance of high-salinity inclusions, beside existing low-salinity ones, indicates that in particular geological periods of the existence of the Bukovik-Kadiica magmatic-hydrothermal system there was a yield of brine fluids from deeper reservoirs although some other possibilities should not be excluded (salinity increase of the magmatic volatile phase as the system evolved; boiling only during certain stages or that there could be some meteoric water overprint).