223

2009 Society of Economic Geologists, Inc. Economic Geology, v. 104, pp.
223248
Stable Isotope Constraints on Ore Formation at the San Rafael Tin-Copper Deposit, Southeast Peru
THOMAS WAGNER,
Institute of Isotope Geochemistry and Mineral Resources, ETH Zurich, NW F 82.4, Clausiusstrasse 25, CH-8092 Zrich, Switzerland
MICHAEL S. J. MLYNARCZYK,
Rathdowney Resources Ltd., Killaderry, Daingean, County Offaly, Ireland
ANTHONY E. WILLIAMS-JONES,
Department of Earth and Planetary Sciences, McGill University, 3450 University Street, Montral QC H3A 2A7, Canada
AND
ADRIAN J. BOYCE
Scottish Universities Environmental Research Centre (SUERC), East Kilbride, Glasgow, G75 0QF, Scotland, United Kingdom
Abstract
The San Rafael tin-copper deposit in the Eastern Cordillera of the Peruvian Central Andes is the worlds largest hydrothermal tin lode, with a total resource of about 1 million metric tons metal, at an average grade of 4.7 wt percent Sn. The mineralization is of the cassiterite-sulfide type and occurs in a vertically extensive veinbreccia system centered on a shallow-level, late Oligocene granitoid stock. The tin ores form cassiterite-quartzchloritebearing veins and breccias hosted by several large fault-jogs at depth in the lode. By contrast, the copper ores, which contain disseminated acicular cassiterite, are localized in the upper part of the system. Both ore types are associated with a very distinctive, strong chloritic alteration, which was preceded by intense sericitization, tourmalinization, and tourmaline veining. The 34S values of the sulfides range between 2 and 6 per mil, and vary very little with location in the deposit. This indicates that the hydrothermal system was large, with a relatively homogeneous source of sulfur, likely of magmatic origin. This is confirmed by stability relationships of ore minerals, which indicate that the ore fluids were initially reducing. Microthermometric studies of fluid inclusions in cassiterite, quartz, tourmaline, and fluorite show that the fluids responsible for the early, barren stage were hot, hypersaline brines (380540C, 3462% NaCl equiv), whereas the ore-stage fluids had moderate to low salinity (021 wt % NaCl equiv), and were of moderate temperature (290380C). In addition to the marked dilution of the ore fluids with evolution of the hydrothermal system, they became progressively more oxidizing, as inferred by the local association of minor hematite with cassiterite and the ubiquitous replacement of pyrrhotite by pyrite and marcasite. The 18O values of the fluid decreased systematically with time, as indicated by the 18O values of different generations of tourmaline, cassiterite, and quartz. This evolution was paralleled by an increase in the D values of the fluid, inferred from the D values of tourmaline and chlorite. This trend is consistent with mixing of the ore fluids with a cooler fluid that had substantially lower 18O, and cannot be explained by fluid boiling. Based on structural evidence for an opening of the vein system and a transition from lithostatic to hydrostatic conditions at the onset of mineralization, we infer that ore deposition was caused by an influx of hot groundwater of meteoric origin which mixed repeatedly with tin-bearing magmatic brines. The oxidation, dilution, cooling, and acid neutralization of the ore fluids destabilized chloride complexes of tin and triggered the large-scale precipitation of cassiterite.
Introduction HYDROTHERMAL TIN LODES represent a major class of graniterelated tin deposits characterized by very high grade (about 15 wt % Sn) and high tonnage, which makes them an ideal target in the exploration for economically viable primary tin resources (Taylor, 1979; Menzie et al., 1988). However, despite their having been mined since antiquity and being well described in the literature, there is still considerable uncertainty about key issues related to the genesis of these deposits, such as the source of the tin, the origin of the ore fluids, and the mechanism(s) of ore deposition. Although it is generally believed that the bulk of the metal is derived directly from an evolved granitic melt by being partitioned into a magmatic
Corresponding
author: e-mail: thomas.wagner@erdw.ethz.ch
aqueous phase, the leaching of crystallized magmatic minerals by later fluids may be another significant metal source (Taylor, 1979; Lehmann, 1990). In addition, some metal could have been derived from the sequences of metasedimentary rocks, which host the granitic plutons (Williamson et al., 2000; LeBoutillier et al., 2002). A variety of ore fluids have been invoked, with emphasis on primary magmatic fluids, but also including meteoric waters, basinal brines (formation waters), metamorphic fluids, and their mixtures. With regard to the meteoric fluids, the timing of their circulation and interaction with magmatic fluids may be relatively early or occur only in the waning stages of the hydrothermal system (Collins, 1981; Primmer, 1985; Sun and Eadington, 1987; Alderton and Harmon, 1991; Sheppard, 1994; Linnen and Williams-Jones, 1995; Wilkinson et al., 1995; Smith et al.,
0361-0128/09/3809/223-26
223
224
WAGNER ET AL.
1996; Walshe et al., 1996; Jackson et al., 2000). Finally, the following mechanisms have been proposed for the precipitation of cassiterite: cooling, pressure decrease, wall-rock alteration, boiling, and fluid mixing (Kelly and Turneaure, 1970; Eadington, 1985; Heinrich and Eadington, 1986; Heinrich, 1990, 1995; Linnen and Williams-Jones, 1995). Among these, mixing between magmatic and meteoric waters is thought to be the most effective means of producing the unusually high ore grades of some deposits (Heinrich, 1990). However, this hypothesis has not been properly tested because the contribution of meteoric water to hydrothermal systems depositing lode tin mineralization has been poorly documented. San Rafael, the worlds richest primary Sn-Cu deposit (Kontak and Clark, 2002; Mlynarczyk et al., 2003), which is located in the high Andes of the Eastern Cordillera of southeastern Peru and is currently exploited by MINSUR S.A., is an ideal natural laboratory in which to investigate the various issues of lode tin genesis raised above. With a total resource of about 1 million metric tons (Mt) Sn, and an average grade of 4.7 wt percent Sn, the San Rafael lode, which has been exposed by mining activity over a vertical extent of more than 1,300 m, allows an unprecedented opportunity to advance our understanding of the formation of this important deposit type. In addition, the deposit is very young (about 2224 Ma; Clark et al., 1983; Kontak et al., 1987; Kontak and Clark,
2002) and, therefore, was not affected by any later geologic processes. Following a study of the deposit geology (Mlynarczyk et al., 2003) and its alteration (Mlynarczyk and WilliamsJones, 2006; Mlynarczyk et al., 2009, submitted), a comprehensive stable isotope (S, O, H, C) investigation of ore and gangue minerals and a preliminary fluid inclusion study were performed to evaluate the source of the fluids and metals, as well as the timing and the mechanism of ore deposition. Our data indicate that ore deposition was caused by an influx of hot groundwaters of meteoric origin and their mixing with the tin-bearing magmatic brines. Geologic Setting Regional geology The San Rafael deposit lies in the northwesternmost extension of the Central Andean Sn-W(-Ag) metallogenic province (e.g., Turneaure, 1960a, b; Kelly and Turneaure, 1970; McBride et al., 1983; Mlynarczyk and Williams-Jones, 2005) and crops out on the glaciated slopes of the Cordillera Carabaya, which forms part of the Eastern Cordillera of southern Peru (Fig. 1). The geology of the region is dominated by a thick sequence of Lower Paleozoic, marine clastic metasedimentary rocks, which experienced extensive crustal deformation in the Cenozoic. The unexposed Precambrian
N
PERU
PACIFIC OCEAN
Lima
SAN RAFAEL
VIC R TO IA
MA RIA NO
E RG JO
MARIA
AR
VI
IA
CE
GU
NT E
IL
NE
LA
SA
LE
DES S ANUANO R E P
Meta-sediments (Upper Ordovician): Phyllites (Sanda Fm.) Quartzites (Sanda Fm.) Slates and hornfelses (Sanda Fm.)
FIG. 1. Geologic map of the western part of the San Rafael district (modified after Arenas, 1980). 0361-0128/98/000/000-00 $6.00
ELEN A
RM
RA
C AT A LI N A
FA E
Granitoids (Late Oligocene):

Coarse-grained granitoids Fine-grained granitoids Dyke of granitoid porphyry Tourmalinized granite
PE DR O
500 m
Structures: Mineralized veins
Major faults Access to the mine
224
STABLE ISOTOPE CONSTRAINTSSAN RAFAEL SN-CU DEPOSIT, SE PERU
225
gneissic basement is covered by a 10- to 15-km-thick sequence of Ordovician to Siluro-Devonian metapelites and metapsammites (the San Jos, Sanda, and Ananea Formations), followed by about 3 to 4 km of late Paleozoic psammites and carbonates (Mississippian Ambo Group, Pennsylvanian Tarma Group, and Permian Copacabana Group) and 3 km of mid-Permian to Triassic red beds and intercalated volcanics (Mitu Group). These rocks are, in turn, overlain by roughly 1 km of Cretaceous psammites and carbonates (Cotacucho Group) and 800 m of Miocene-Pliocene felsic ignimbrites and red beds (Crucero Supergroup). A detailed review of the geology of the area is provided by Laubacher (1978), Clark et al. (1990), and Sandeman et al. (1996). Granitic and granodioritic plutons (some of which host significant Sn-W-base metal mineralization) were emplaced between the late Devonian to early Carboniferous (Kontak and Clark, 1988) and the late Tertiary (Kontak et al., 1987; Kontak and Clark, 2002). Examples of these are the large PermoTriassic granitic batholiths, which host the Sarita polymetallic deposit, and the middle- to late-Tertiary granitic stocks, which host the San Rafael tin-copper and Palca XI tungsten deposits (Clark et al., 1983; Kontak and Clark, 1988; Kontak et al., 1990). The peraluminous Tertiary granitic intrusions are thought to originate from deep-crustal partial melting of metasediments related to magmatic underplating (Sandeman et al., 1995; Kontak and Clark, 2002). In addition to granitoids, there are also subordinate peralkaline syenites and gabbro-diorites. However, most rocks of mafic composition are volcanic and form localized occurrences of shoshonite, minette, absarokite, and high K calc-alkaline basalt (Kontak et al., 1986; Sandeman et al., 1995). Deposit geology The bulk of the tin in the San Rafael-Quenamari mining district is hosted by a single, large vein-breccia system called the San Rafael lode. This lode is part of an array of a few dozen subparallel veins that host Sn-(W)-Cu-Zn-Pb-Ag mineralization. Typically, the veins are planar, 500 to 3,500 m long, have a northwestern strike and dip moderately to steeply to the northeast. They are centered on a small (about 15 km2), high-level, late Oligocene granitoid pluton and straddle the contact with the surrounding Lower Paleozoic metasedimentary rocks, which consist dominantly of Ordovician phyllites and quartzites of the Sanda Formation (Fig. 1). The pluton is a polyphase intrusion composed of coarse- to medium-grained K-feldspar megacrystic, cordierite-biotite(garnet-sillimanite) granite, leucogranite and granodiorite, with local minor enclaves of tonalite and quartz diorite. The magma was strongly peraluminous, S-type in character, and clearly formed due to the partial melting of metasediments (Mlynarczyk et al., 2003). However, it was not geochemically evolved, and the different igneous phases do not represent a fractional crystallization series (Dolejs et al., 2009, submitted). The pluton is elongated to the northeast and corresponds to a prominent topographic high (Nevado Quenamari, 5,300 m), but is only exposed locally, although its horizontal extent is clearly evident from the distribution of the overlying contact metamorphic rocks (hornfelses and slates). The available geochronological data (Kontak and Clark, 2002) demonstrate that the pluton was emplaced at 24.6 0.2 (U-Pb zircon
0361-0128/98/000/000-00 $6.00
age) to 24.7 0.2 Ma (U-Pb monazite age). Ar/Ar data show that the upper part of the stock had cooled down to the closure temperature of biotite (about 300350C: Harrison et al., 1985) at about 23.7 0.2 Ma. The main Sn-Cu mineralization appears to be slightly younger than this cooling age, based on Ar/Ar ages ranging between 22.7 0.7 and 21.9 0.2 Ma, obtained from hydrothermal adularia and muscovite (Clark et al., 1983; Kontak et al., 1987; Kontak and Clark, 2002). The San Rafael lode crops out on the eastern flanks of the glacier-capped Quenamari mountain, at an altitude between 4,500 and 5,100 m, and has a length of over 3 km. It has been exposed by mining over a vertical extent of about 1,300 m (from 5,100 to 3,800 m above sea level). Current mining activity is restricted to the lower 500 m of the deposit. The upper part of the lode is hosted by metasedimentary rocks, attains a width of up to 2 m, and contains appreciable chalcopyrite and subordinate needle-tin cassiterite. By contrast, the lower part of the lode is hosted by granitic rocks, opens into a series of subvertical shoots up to 50 m wide, and contains a large volume of high-grade cassiterite ore (Fig. 2). The ore shoots are composed of sets of veins that have thicknesses in the range of 5 to 30 cm. Quartz and chlorite are the principal gangue minerals, and are common in both the upper and lower parts of the deposit. The lode has an average strike of 330, dips 40 to 75NE, and is texturally quite complex, having undergone multiple episodes of vein reopening and brecciation. Mlynarczyk et al. (2003), who provided a detailed description of the deposit geology, proposed that the large ore shoots at depth in the lode represent fault jogs formed during sinistral-normal strike-slip movement along the plane of the lode, which was synchronous with circulation of the ore fluids. Alteration and Vein Paragenesis A detailed campaign of core-logging and underground mapping in the San Rafael lode outlined as many as 15 distinct vein types, which were arranged chronologically, based on crosscutting relationships (Mlynarczyk et al., 2009, submitted). The mineralogy of these veins and the associated wall-rock alteration are summarized briefly below, within the framework of four principal paragenetic stages, adopted from a subdivision originally proposed by Palma (1981). (I) Early, barren tourmaline stage The early, barren stage was initiated by pervasive sericitic alteration, followed by strong tourmalinization, which produced wide alteration halos around early tourmaline-quartz veins and tourmaline stringers. These early veins, which cut through large parts of the granitic pluton, are typically sealed and lack cassiterite. Locally, however, they host abundant arsenopyrite. The tourmaline is mainly dravite and Fe-rich dravite, but other varieties (foitite, Mg foitite) were also observed. A detailed description of the mineralogy, chemistry, and temporal relationships of the different varieties of vein and alteration tourmaline can be found in Mlynarczyk and Williams-Jones (2006). In addition to the tourmaline-rich veins, laterally extensive (up to several meters wide) veins of hydrothermal microbreccia are very common. These are composed of strongly
225
226
WAGNER ET AL.
SE
Nevado Quenamari
NW
5000 m
R165 R709 R725 R726
R708 R729
R428 R429 R137 R65 R404 R573 R511 R28 R27 R55 R57 R399 D3-27 D3-39 D3-69 A13 R17 R113 R119 R123
4500 m
Copper stopes Tin orebodies
500m
Ore Shoot
Contact Orebody
4000 m a.s.l.
FIG. 2. Longitudinal section through the San Rafael lode, showing the location of hand samples collected during this study. The location of drill core samples (elevations below about 4,300 m) is indicated in Tables 1 to 3.
tourmalinized rock flour and subordinate quartz, generally occur in series, and contain the bulk of the tourmaline in the deposit. Commonly, they are crosscut by a dense swarm of younger quartz veins and are associated with subordinate silicification which, locally, could have been coeval with tourmaline precipitation. Both the tourmaline-rich veins and microbreccia veins occur throughout the vertical extent of the lode, although they are more abundant in its central and upper parts. Petrographic observations indicate that reopening of the early tourmaline veins was quite common, and that there were many episodes of brecciation. It is also apparent that some of the major tourmaline-bearing veins, which clearly follow the lode along strike and were responsible for establishing its initial structural framework, subsequently reopened during the ore stage (Mlynarczyk et al., 2003). The principal chemical changes associated with the strong sericite-tourmaline alteration were a nearly complete removal of alkalis and alkaliearth elements (Na, K, Ca, Ba, Rb, Sr, Cs, Li, and Cl) from the wall rock and a marked addition of B, Mg, and Fe, and possibly Sn, W, and In, though the latter could represent a late overprint (Mlynarczyk et al., 2009, submitted). (II) Main cassiterite stage The initiation of the ore stage was related to a transition from a relatively closed to an open vein system, as all ore and later vein types are characterized by textures consistent with filling of open fractures. The earliest phase of the ore stage is marked by a rare type of tourmaline-cassiterite-chlorite-arsenopyrite vein, in which an obvious but continuous transition in mineral chemistry is observed. An early variety of orangecolored dravite was progressively replaced by green-colored schorl (a strongly Fe-rich tourmaline), and was accompanied by simultaneous precipitation of cassiterite and chlorite, followed by massive arsenopyrite (Mlynarczyk and WilliamsJones, 2006). The majority of the veins and breccias of the main tin ore stage, however, are characterized by a simple mineralogy that
0361-0128/98/000/000-00 $6.00
consists of cassiterite, chlorite, and quartz (Fig. 3a). Cassiterite occurs mainly as massive, dark-brown, locally botryoidal layers (up to 7 cm wide) or as beige-colored, collomorphic wood tin. Chlorite is dark green and moderately to strongly enriched in iron, corresponding compositionally to ripidolite and daphnite (Mlynarczyk et al., 2009, submitted). Locally, the cassiterite-chlorite-quartz veins also contain minor proportions of Fe-rich wolframite or arsenopyrite (Fig. 3b), which is the essential sulfide mineral associated with this stage. The larger, high-grade veins are complexly banded, with multiple generations of alternating quartz, cassiterite, and chlorite, that commonly record a consistent temporal sequence of mineralization that can be distinguished on the basis of crystal morphology and color (Fig. 4). Some of these veins have quartz-rich centers with a high proportion of sulfides (e.g., chalopyrite, arsenopyrite, pyrite) that were formed during the later sulfide stage, as deduced from crosscutting vein relationships. Finally, traces of hematite are associated with chlorite and cassiterite in the brecciated ores at about an elevation of 4,400 m. No tourmaline occurs in the cassiteritechlorite-quartz veins, but hairline veinlets of a late, Fe-rich variety of dark-blue or dark-green tourmaline crosscut the early tourmaline veins and microbreccias of the barren stage. The ubiquitous cassiterite-chlorite-quartz veins are surrounded by wide halos of strong chloritic alteration and display evidence of frequent reopening (crack-and-seal textures) and cockscomb textures, producing a characteristic recurrent alternation of cassiterite, chlorite, and quartz layers. The bulk of the tin ore is restricted to the lower, granitehosted part of the San Rafael lode, especially the wide, subvertical fault-jogs, which form elongated sinuous zones up to 50 m wide. The highest ore grades (up to 45 wt % Sn) are associated with several-meter-wide breccia zones, which either contain fragments of cassiterite veins cemented by chlorite and quartz, or fragments of chloritized wall rock cemented by cassiterite and quartz. Very high grade cassiterite mineralization also occurs in the major (12 m wide) footwall and
226
227
chl qz cp tou cas cas chl sp
apy qz
cas
5.5cm
2.5cm
sp cp qz qz
cp
3.5cm
3.5cm
FIG. 3. Photographs of representative ore hand specimens. a. Example of typical tin ore from a fault-jog in the lower part of the deposit. Layers of massive, dark-brown cassiterite (cas) overgrown on fragments of early quartz (qz)-tourmaline (tou) breccia and chloritized (chl) wall rock (sample SAR-R464, contact orebody, level 4350 m). b. Complex, open space-filling; quartz-cassiterite-sulfide vein (the arrow indicates the direction of growth). Near the strongly chloritized margin of the vein there are alternating layers of cassiterite, chlorite, and quartz. The center of the vein consists of large quartz crystals capped by arsenopyrite (apy), with minor grains of sphalerite (sp) and chalcopyrite (cp) (sample SAR-R511, ore shoot, level 4310 m). c. Typical copper ore from the upper part of the deposit, composed dominantly of chalcopyrite, with subordinate quartz and pyrite (sample SAR-R710, San Rafael lode, level 4820 m). d. Broken fragments of sphalerite and chalcopyrite, cemented by quartz, in the central part of a quartz-chalcopyrite vein (sample SAR-R722, San Rafael lode, level 4730 m).
hanging-wall veins of the lode structure, as well as in locations where veins are deflected, branch, or intersect. It should be emphasized that the bulk of the tin is contained in the veins, veinlets, and breccias, as opposed to being disseminated in the chloritized wall rock (Mlynarczyk et al., 2003, 2009, submitted). The entire structure of the San Rafael lode is enveloped by a 2- to 10-m-wide zone of very strong, texturally destructive, chloritic alteration, which affected the host rocks during this and the following stage. Chloritization is the dominant type of alteration in the deposit, affecting its entire vertical extent and strongly overprinting all earlier alteration types. Locally, chloritization is associated with silicification, but the latter may have been developed late. Studies of the whole-rock chemistry indicate that chloritization leached most of the content of alkali and alkali-earth elements from the granitic wall rocks (Na, K, Ca, Ba, Rb, Sr, Cs, and F, Cl, S) and introduced large amounts of Fe, H2O, Mn, Sn, W, In, and Mg (Mlynarczyk et al., 2009).
0361-0128/98/000/000-00 $6.00
(III) Sulfide stage The bulk of the sulfide ore is restricted to the middle and upper parts of the San Rafael lode, and is most strongly concentrated where the lode straddles the intrusion-slate contact and within the slates. The sulfides clearly postdated cassiterite deposition, and are characterized by a great mineralogical diversity. The dominant sulfide mineral is chalcopyrite (Fig. 3c), which is associated with subordinate amounts of needle-tin cassiterite and stannite, yielding a low grade (0.51 wt %) tin ore. Other sulfides, which locally are very abundant, are pyrrhotite, arsenopyrite, pyrite, sphalerite, and galena. In addition, hematite, marcasite, bismuthinite, and native bismuth have also been observed. The main gangue minerals are quartz, chlorite, and fluorite. The minerals that formed during the sulfide stage occur either in distinct quartz-sulfide veins (commonly with a marked predominance of either quartz or sulfides) or as a late filling (or reopening) of stage II cassiterite-quartz-chlorite veins. The structural orientation of
227
228
WAGNER ET AL.
3 cm
qz8
qz 9.8
py 10.8
cas5
cp
2.7 9.7 qz 10.1 3.0 3.3 cas 2.2
qz7
inside large vugs in massive pyrrhotite ore, and siderite veinlets cutting across pyrrhotite, that most of the pyrite, marcasite, and siderite originated by hypogene replacement of pyrrhotite. Mineralogical and textural observations suggest that the sulfide ores of San Rafael formed at comparatively high temperatures, as indicated by the widespread presence of welldeveloped sphalerite stars in chalcopyrite (Wiggins and Craig, 1980; Kojima and Sugaki, 1984; Sugaki et al., 1990) and common sphalerite blebs enclosed in other sulfides. Examples of some of these ore textures are shown in Figure 5. However, although specific paragenetic sequences can be determined for individual samples, these parageneses generally cannot be extended to the entire deposit, as most sulfides (aside from pyrite and marcasite) occur in multiple generations and are broadly coeval. (IV) Late, barren quartz-carbonate stage The last stage of veining is represented by numerous barren quartz carbonate (calcite, siderite) veins with trace amounts of chalcopyrite and sphalerite but no cassiterite. They have narrow, strongly chloritized margins. These veins crosscut all other types of veins and breccias, locally producing dense stockworks, and are commonly quite thick (0.21 m), implying massive precipitation of quartz in the waning stages of the hydrothermal system. On this basis, the bulk of the silicification observed in the deposit is tentatively assigned to this stage. In addition to the four hypogene stages, a fifth, supergene stage is well developed in the uppermost, copper-rich part of the lode. This resulted in the alteration of the primary ores to Fe oxides and hydroxides (limonite) and secondary Cu minerals such as covellite and chalcocite. No secondary enrichment of tin minerals was observed. Fluid inclusions We carried out a reconnaissance fluid inclusion study of the San Rafael deposit in order to provide important constraints on the temperature and composition of the ore-forming fluids. The minerals studied were quartz, cassiterite, tourmaline, and fluorite, for which over two dozen fluid inclusion sections were made. Microthermometric measurements were performed at McGill University on a Linkam THMSG-600 fluid inclusion stage, using synthetic fluid inclusions (SYNFLINC) as temperature standards (calibration temperatures: 56.6, 21.2, 10.7, 0.0, and 374.1C). Heating rates were 0.1C/min when phase transitions were approached. Errors were 0.2 and 5C for final ice melting and homogenization temperatures, respectively. The total salinity of the inclusion fluids (expressed as wt % NaCl equiv) was calculated using the equation of Bodnar (1993) for low- and moderatesalinity inclusions, and the equation of Sterner et al. (1988) for hypersaline inclusions. Fluid inclusion densities and isochores were calculated using the computer package FLUIDS (Bakker, 2003). Fluid inclusion petrography The studied samples contain abundant fluid inclusions, ranging in diameter between 2 and 200 m, many of which clearly represent different generations. Representative fluid
cas4 4.6 qz6 qz5
cas3
cas
3.3
qz4
qz
10.6 2.9
qz3 qz2 cas2
chl 10.6 3.4 R 119
qz1
FIG. 4. Photograph of a large, banded cassiterite-quartz-chlorite-chalcopyrite vein (sample R 119, contact orebody, level 4310 m). Also shown are the locations of crystals analyzed for their oxygen isotope composition. The arrow indicates the direction of vein growth; different generations of quartz and cassiterite are identified by labels at the left side of the figure. Note that the oldest cassiterite generation (cas1) is not shown in the photograph, but is present in the sample. The minerals analyzed are quartz (white label), cassiterite (yellow), chlorite (green) and wolframite (pink).
the veins formed during the sulfide stage is the same as that of stage II veins, suggesting a common mechanism of formation. Moreover, the sulfide stage veins are associated with the same, pervasive chloritic alteration as the stage II veins (Mlynarczyk et al., 2003). In addition, the lode hosts breccias, which either contain sulfide fragments (Fig. 3d) or are cemented by sulfides. It also hosts large (1030 cm) vugs filled by pyrite, marcasite, and rare hematite. We infer from the occurrence of cubes of pyrite (several centimeters in diameter)
0361-0128/98/000/000-00 $6.00
228
229
cas chl qz cas cp qz
chl
a
py cp
500 m
580 m
sph po py cas stn chl cas apy
cp sp py po
460 m
c
mc
chl
300 m
d
py
po py cp cp sp
qz sp
apy
gn py
580 m
f
cp
200 m
sp stn sp qz cp py qz
80 m
300 m
FIG. 5. Photomicrographs in reflected light showing representative textures of ore assemblages. a. Alternating layers of chlorite (chl), cassiterite (cas) and quartz (qz), typical of the rich tin ores hosted by the fault-jogs in the deeper parts of the deposit (sample R119, vein selvage, contact orebody, level 4310 m). b. Aggregates of needle-tin cassiterite crystals (seen in cross section), intergrown with chlorite, quartz, and chalcopyrite (cp) in ore from the upper part of the deposit (sample R725, San Rafael lode, level 4730 m). c. Needle-tin cassiterite, rimmed and crosscut by stannite (stn), associated with chlorite, chalcopyrite, and pyrite (py), which contains inclusions of sphalerite (sp) and pyrrhotite (po) (sample R119, center of the vein, contact orebody, level 4310 m). d. Euhedral arsenopyrite (apy) overgrown by chlorite, chalcopyrite, and cassiterite. The host chalcopyrite contains abundant blebs of pyrrhotite (locally replaced by pyrite) and veinlets of sphalerite (sample R55, ore shoot, level 4330 m). e. Arsenopyrite replaced by sphalerite and galena (gn), and overgrown by chalcopyrite and pyrrhotite. Pyrrhotite is partly replaced by pyrite and marcasite (mc) (sample A1-1, orebody 150, level 4100 m). f. Sphalerite with abundant chalcopyrite inclusions (chalcopyrite disease), veined by coarse-grained chalcopyrite, quartz, and pyrite. The latter sulfides contain fine sphalerite inclusions (sample R28, ore shoot, level 4330 m). g. Sphalerite rimmed by stannite, enclosed in quartz, and neighboring chalcopyrite and pyrite (sample R119, center of the vein, contact orebody, level 4310 m). h. Euhedral crystals of late pyrite in small vugs hosted by chalcopyrite, and crosscut by late sphalerite veinlets (copper ore, characteristic of the upper levels of the deposit. sample R710, San Rafael lode, level 4820 m). 0361-0128/98/000/000-00 $6.00
229
230
WAGNER ET AL.
inclusion assemblages are shown in Fig. 6. The vast majority of the inclusions are liquid-vapor (LV) and liquid-vapor-solid (LVS) types, with the latter type commonly containing more than one solid. On the basis of EDS analyses of opened LVS inclusions, these solids are (1) a ubiquitous, highly birefringent mineral, which in some inclusions forms bundles and could be a hydrous phyllosilicate (chlorite or phengitic sericite-illite), (2) aggregates of radiating birefringent crystals, (3) transparent, platy solids, some of which are birefringent (probable Fe-Al-Na-Ca-bearing chlorides and carbonates), (4) single tiny specks of an equant opaque mineral (likely chalcopyrite), (5) cubes of halite (in hypersaline inclusions), (6) relatively uncommon, finely prismatic crystals of cassiterite (needle-tin variety), (7) anhedral grains of cassiterite (occurring also as solid inclusions), and (8) bladed crystals of calcite (in late quartz-calcite veins). With the exception of halite, these minerals do not redissolve on heating and likely represent accidently trapped crystals. In addition to the solidbearing inclusions, vapor-rich and vapor-only fluid inclusions are locally present, but these are restricted to the early, barren tourmaline stage (I). No aqueous-carbonic inclusions were observed in the minerals studied and preliminary gas chromatographic analyses of inclusion fluids indicate very low proportions of CO2 and CH4. Because multiple generations of secondary inclusions are very abundant in most samples and their trails are locally very dense, it appears possible that some of the original primary fluid inclusions were subsequently refilled. In addition, some of the secondary fractures in the host minerals (especially quartz) have healed very well, producing a scattered distribution of those inclusions, which texturally appear isolated but
are in fact secondary in origin. As a result, clusters of fluid inclusions displaying highly variable salinities, likely unrelated to primary entrapment, are quite common. We have addressed these issues by a particularly careful selection of primary and pseudosecondary fluid inclusions, where we adhered strictly to the criteria for these fluid inclusion types established by Roedder (1984) and Goldstein and Reynolds (1994). Moreover, in establishing the final data set reported in this paper, we restricted ourselves to fluid inclusion assemblages, i.e., groups of texturally well-constrained fluid inclusions that yielded highly consistent microthermometric results. Microthermometric results Several fluid inclusion assemblages (in the sense of Goldstein and Reynolds, 1994) in each sample were analyzed microthermometrically using the cycling technique method for determination of Tmice and Th (Haynes, 1985). The preliminary results of fluid inclusion microthermometry are plotted on a homogenization temperature (Th) versus salinity diagram (Fig. 7). Although NaCl is the dominant solute, cassiteritehosted primary and pseudosecondary inclusions contain appreciable CaCl2 and lesser amounts of KCl. This was inferred from the microthermometric measurements (temperature of initial ice melting between 60 and 50C, temperature of hydrohalite dissolution close to 30C) and, more directly, by EDS analysis of fluid inclusion decrepitate mounds. In addition, FeCl2 was also present in the ore fluids, as demonstrated by comparatively low but consistent amounts of FeCl2 detected in the analyzed decrepitates. Stage I quartz- and tourmaline-hosted inclusions have a salinity range of 34 to 62 wt percent NaCl equiv (with most
30 m
30 m
25 m
50 m
FIG. 6. Photomicrographs of representative fluid inclusion assemblages from a complex cassiterite-quartz-sulfide vein (sample R-119, contact orebody, level 4310 m). a. Cluster of primary fluid inclusions in stage II quartz. b. Primary fluid inclusions in stage II cassiterite developed along growth zones. c. Large primary fluid inclusion in stage II quartz. d. Densely distributed very low salinity secondary fluid inclusions in stage IV quartz. 0361-0128/98/000/000-00 $6.00
230
231
600
Homogenization temperature (C)
500
LVH inclusions (homogenize by halite dissolution) LVH inclusions (homogenize by vapor disappearance) Primary / pseudosecondary LV inclusions Stage I quartz / tourmaline Secondary LV inclusions
i o cb ilin g
400
200
Stage IV quartz
Stage II quartz
100 0 10 20 30 40 50 60 70
necking down
Total salinity (wt.% NaCl eq.)

FIG. 7. Homogenization temperature versus salinity of fluid inclusions from San Rafael ores. The host minerals investigated were quartz, fluorite, cassiterite, and tourmaline. Abbreviations: L = liquid, V = vapor, H = halite.
data clustering around 4156 wt %) and homogenize by disappearance of either halite or vapor in the range 240 to 535C, but mainly between 340 and 475C. Based on textural evidence, it is inferred that the inclusions, which have a very high salinity and homogenization temperature, formed due to local boiling of the early fluids, although necking down after a phase change could also explain the highly variable phase ratios observed within fluid inclusion assemblages. The data obtained are in broad agreement with those of Kontak and Clark (2002). Stage II, cassiterite-hosted, primary and pseudosecondary fluid inclusions were measured in several generations of cassiterite (cas1 to cas3, Fig. 4), and have a very narrow range in salinity (1721 wt % NaCl equiv) and Th (354361C). Quartz-hosted inclusions from the same stage exhibit consistently lower temperature and salinity, which vary depending on the generation of the host quartz (1116 wt % NaCl equiv and 265305C for qz1 and qz2, and 2.55.5 wt % NaCl equiv and 265295C for the later qz3 and qz4, Fig. 4). The salinity of the fluids which precipitated stage II quartz is likely to have spanned the entire range between 2.5 and 16 wt percent NaCl equiv, as suggested by the behavior of fluids trapped in secondary inclusions. However, the most remarkable feature of the quartz-hosted inclusions from this stage is a population of inclusions of very low salinity (Tmice as high as 1.3C), which are quite similar in composition and temperature to the fluids that precipitated the barren quartz from stage IV (Fig. 7). In addition, the character of the ore veins, in which bands of cassiterite, chlorite, and quartz repetitively alternate, and the consistently lower homogenization temperature and salinity of the fluid inclusions trapped in quartz
0361-0128/98/000/000-00 $6.00
231
ha
bands relative to those in cassiterite layers indicate marked, periodic fluctuations in the composition of the fluids circulating in the vein system. Primary inclusions in stage III quartz and fluorite are characterized by a salinity in the range of 0.513 wt % NaCl equiv (mostly 36 wt %) and a Th in the range of 265340C. Significantly, primary and pseudosecondary inclusions in quartz associated with chalcopyrite in the upper zone of the deposit (sample R709, elev 4,820 m) have, on average, a Th lower by about 40C than inclusions hosted by a very similar variety of quartz associated with the chalcopyrite-rich centers of deeper-seated stage II-III veins (sample R119, elev 4,310 m). Another important observation is the lower Th limit of about 265C consistently displayed by all (primary-secondary) stage II and stage III inclusions. Fluid inclusions hosted by stage IV quartz have the lowest salinity (02.5 wt % NaCl equiv) and homogenization temperature (235265C). When the microthermometric data for all four paragenetic stages are considered jointly (Fig. 7), it is evident that the sets of fluid temperature and salinity recorded for each stage form a broad linear trend of salinity with homogenization temperature. Quartz-hosted primary-pseudosecondary stage II inclusions form two populations at lower and salinity than stage II cassiterite- (the main cassiterite stage) hosted inclusions, whereas stage III quartz-fluorite-hosted inclusions have higher Th values than stage II quartz-hosted inclusions. Overall, the fluid inclusion data point to repeated fluid mixing during the evolution of the San Rafael deposit, with the hotter and more saline fluids of the main cassiterite stage more closely reflecting the composition of the primary ore fluid.
lit
sa
300
tu
ra
tio n
cu
rv
Stage III quartz / fluorite
Stage II cassiterite
is ep
od
232
WAGNER ET AL.
Mineral Geothermometry Mlynarczyk et al. (2009) have tested two empirical chlorite thermometers (Cathelineau, 1988; Jowett, 1991) on different samples of alteration and vein chlorite from stage II of the paragenesis. The chlorite temperatures obtained are in the range of 335415C. The thermometer of Jowett (1991) yields values about 515C higher than those from that of Cathelineau (1988), and the values for alteration chlorite are, on average, roughly 50C higher than those for vein chlorite. The theoretical geothermometer of Walshe (1986) yields temperatures in the range 295 to 320C for the alteration chlorite (Mlynarczyk et al., 2009). These temperatures are consistently below the Th values of most stage II fluid inclusions. We have not further considered these results, because the thermodynamic model of Walshe (1986) has been shown to be inconsistent with more recent experimental and theoretical studies (Holland et al., 1998; Vidal et al., 2001; Aja, 2002; Aja and Dyar, 2002;). Overall, the results suggest that chlorite (and, by inference, the associated cassiterite) formed in the temperature range 350420C. The stage III sulfide ores commonly contain microscopic overgrowths of stannite on sphalerite (Fig. 5g), as well as less common intergrowths between these two minerals. It has been shown that the equilibrium compositions of sphalerite and stannite can be used successfully as a geothermometer (Nekrasov et al., 1976, 1979; Shimitzu and Shikazono, 1985; Bortnikov et al., 1990). Carefully selected sphalerite-stannite intergrowths in representative ore samples were analyzed by electron microprobe at McGill University. Representative results and average data are given in Table 1. The equation of Nakamura and Shima (1982) was used to calculate the formation temperature of the intergrowths, which is interpreted to represent the lower temperature limit of stage III ore formation. The resulting temperatures are generally in the range 270 to 290C, i.e., they are slightly lower than fluid inclusion homogenization temperatures from this stage. Stable Isotope Studies In order to test our hypothesis that fluid mixing was the dominant ore-forming process, we performed a comprehensive stable isotope study on the main ore and gangue minerals. Representative samples from the principal mineralization stages (stages IIV) of the San Rafael deposit were selected
for sulfur, oxygen, hydrogen, and carbon isotope analysis. The samples cover a wide range of textures of hydrothermal sulfides, oxides, and silicates, and permit characterization of the isotopic evolution of the system along the paragenetic sequence. In addition to samples that cover the different mineralization stages, we also performed selected small-scale studies of individual veins employing a high sample density to resolve isotopic effects at the micro- and mesoscales. Analytical techniques Separates of hydrothermal carbonate (siderite, calcite), oxide, and silicate minerals (quartz, cassiterite, wolframite, chlorite, tourmaline) were prepared by careful handpicking under a binocular microscope, followed by cleaning in doubly distilled water. Sulfide minerals (pyrite, pyrrhotite, arsenopyrite, chalcopyrite, sphalerite, galena) were analyzed by in situ laser combustion from standard polished blocks. Whole-rock oxygen isotope compositions were analyzed from rock powders prepared by standard crushing and milling techniques. All mineral separates and whole-rock samples were dried at 80C for at least 6 hours prior to isotopic analysis. The sulfur, oxygen, and carbon isotope analysis was conducted in the stable isotope lab at SUERC (East Kilbride, UK), whereas hydrogen isotope analysis was carried out at the isotope laboratory of the Department of Geological Sciences at Queens University (Kingston, Canada). Sulfide minerals were combusted in an oxygen atmosphere using a SPECTRON LASERS 902Q CW Nd:YAG laser (1 W power); typical spot diameters in this study were 200 to 400 m. Laser extraction was followed by cryogenic purification of the SO2 gas and subsequent on-line mass-spectrometric analysis. Details of the laser extraction technique, calibration, and correction procedures are provided by Kelley and Fallick (1990), Kelley et al. (1992) and Wagner et al. (2002). Laser calibration data for arsenopyrite were taken from the study of Wagner et al. (2004). Reproducibility of the analytical results, and mass spectrometer calibration, were monitored through replicate measurements of international standards NBS-123 (34SV-CDT: 17.1) and IAEA-S-3 (34SV-CDT: 31.0), as well as the internal lab standard CP-1 (34SV-CDT: 4.6). The analytical precision (1) was about 0.2 per mil. All sulfur isotope compositions are reported in standard delta notation relative to V-CDT.
TABLE 1. Composition of Sphalerite-Stannite Intergrowths and Equilibrium Temperatures Calculated Using the Geothermometer of Nakamura and Shima (1982) Sample no. R27 R119 R447 R404 A1-1 R65 Location Ore shoot 4330 m Contact orebody 4310 m Contact orebody 4270 m S contact orebody 4310 m No. 150 orebody 4100 m Ore shoot 4390 m Mineral Sphalerite Stannite Sphalerite Stannite Sphalerite Stannite Sphalerite Stannite Sphalerite Stannite Sphalerite Stannite Cu (wt%) 0.13 28.34 0.20 27.93 0.51 28.25 0.10 28.53 0.82 28.27 0.42 28.52 Fe (wt%) 11.14 14.94 11.79 15.24 11.47 14.40 12.70 13.75 13.69 14.23 11.32 13.95 Zn (wt%) 54.55 2.11 53.69 1.80 52.80 1.43 52.83 2.59 50.6 1.69 53.55 1.42 Sn (wt%) 0.04 25.56 0.01 24.99 0.02 26.51 0.01 26.85 0.00 25.4 0.07 26.32 S (wt%) 33.19 30.13 33.45 30.02 33.05 29.62 34.20 30.41 33.91 29.52 33.61 30.10 Total 99.05 101.08 99.14 99.98 97.85 100.21 99.84 102.13 99.02 99.11 98.97 100.31 XFe/XZn 0.20 7.08 0.22 8.47 0.22 10.07 0.24 5.31 0.27 8.42 0.21 9.82 T (C) 282 277 269 305 288 269
0361-0128/98/000/000-00 $6.00
232
233
Oxygen was extracted from mineral separates and wholerock powders by reacting 1 to 5 mg of sample with purified chlorine trifluoride in a laser fluorination system, based on techniques of Sharp (1990) and Mattey and Macpherson (1993). The oxygen was converted to CO2 by reaction with a heated graphite rod; the isotopic composition of the cryogenically purified CO2 gas was measured on-line with a VG PRISM III mass spectrometer. Analytical precision was controlled through replicate measurements of the internal laboratory standard SES-2 (18OV-SMOW: 10.2) during the course of the study. The latter was calibrated against international standards IAEA-NBS-28 (18OV-SMOW: 9.6) and IAEA-NBS-30 (18OV-SMOW: 5.1). Precision (1) was found to be better than 0.2 per mil for the whole dataset. Some additional tourmaline samples were analyzed by conventional techniques following the procedures of Clayton and Mayeda (1963). All O isotope data are reported relative to V-SMOW. Hydrothermal carbonates were analyzed for their C and O isotope composition using an Analytical Precision AP2003 continuous-flow mass spectrometer, equipped with an automated carbonate preparation system. About 1 mg of sample powder was placed in a 6 ml vacutainer, then sealed and loaded onto the autosampling unit. Each sample was flushed with helium, then a predetermined amount of 93 percent phosphoric acid was injected into each tube, following the procedure of McCrea (1950). The acid reaction was conducted at a temperature of 70 0.1C; reaction times were 24 hours for pure calcite samples and 120 hours for all other carbonates. After completion of the reaction, the samples were transferred to the processing system and analyzed with an AP2003 mass spectrometer. Oxygen isotope data for siderite were corrected using an acid fractionation factor calculated from Rosenbaum and Sheppard (1986). Reproducibility of the
analytical results was monitored through replicate measurements of the internal Mab2b standard (13CV-PDB: 2.48 ; 18OV-PDB: 2.40) before and after each batch of samples. Accuracy was controlled by replicate measurements of international standards IAEA-CO-1 (13CV-PDB: 2.48; 18OV-PDB: 2.44) and IAEA-NBS-19 (13CV-PDB: 1.95; 18OV-PDB: 2.20 ). External precision (1) was found to be better than 0.2 for both carbon and oxygen isotope compositions. Carbon and oxygen isotope data are reported relative to VPDB and V-SMOW, respectively. Hydrogen isotope analysis of chlorite and tourmaline was carried out by elemental analysis continuous-flow isotope ratio mass spectrometry (CF-IRMS), using a high-temperature reduction method (Sharp et al., 2001). This method involves passing the sample in an He stream through a carbonpacked furnace heated to 1450C, and releasing the water, which is then reduced to H2. Hydrogen gas is purified by passing it through a 5A molecular sieve gas chromatographic column, and subsequently analyzed with a Finnigan Mat Delta XL Plus gas source mass spectrometer. Precision (1) was found to be better than 3 per mil. The hydrogen isotope data are reported relative to V-SMOW. Sulfur isotope data A total of 64 in situ laser sulfur isotope analyses were performed on polished blocks from 20 samples, spanning both the paragenesis and the vertical extent of the deposit. The majority of the samples are from stage III (sulfide stage), where sulfides are abundant and generally several sulfide minerals are present, in contrast to stages I and II for which sulfides are scarce and are represented only by arsenopyrite. The sulfides analyzed were arsenopyrite, chalcopyrite, pyrrhotite, pyrite, sphalerite, and galena. Typically, several coexisting sulfides were analyzed in each sample (Fig. 8). The
tou 3.7
gn
qz 5.3 5.9
sp
cp
5.0 py 5.1 py cp
apy 3.6
qz 5.5
sp
5.4 qz D3-27 R 404
py
3.8 4.2 R 119

gn gn
3.9
apy
cp
apy
po
6.3
sp
5.1 2.4
gn qz sp cp
1.2 4.8 qz qz
po py sp
3.0 5.4
cp
1.4
4.8 3.7
po
apy
5.0 4.4
4.0 R 511a J 410 R 428

FIG. 8. Images showing millimeter-scale textural relationships and sulfur isotope data for representative polished sections. The average diameter of the polished blocks is 2.5 cm. (apy = arsenopyrite, cp = chalcopyrite, gn = galena, po = pyrrhotite, py = pyrite, qz = quartz, sp = sphalerite, tou = tourmaline). 0361-0128/98/000/000-00 $6.00
233
234
WAGNER ET AL. TABLE 2. Sulfur Isotope Data for the San Rafael Sn-Cu Deposit (n = 64)
Sample no.
Mineral
Textural relationships Subhedral, in early tourmaline-quartz vein (DDH no. 3 to the Mariano vein, 39 m) Subhedral, in early tourmaline-quartz vein (DDH no. 3 to the Mariano vein, 39 m) Subhedral, in very early quartz veinlet (DDH no. 3 to the Mariano vein, 27 m) Anhedral, with arsenopyrite, in early quartz veinlet (DDH no. 3 to the Mariano vein, 27 m) Subhedral, with tourmaline, in early quartz veinlet (DDH no. 3 to the Mariano vein, 27 m) Early veinlets in slate, associated with tourmaline and quartz (DDH #8, 50.5 m) Early veinlets in slate, associated with tourmaline and quartz (DDH #8, 50.5 m) Early zone, massive, in tourmaline-cassiterite vein (DDH no. 3 to the Mariano vein, 182.5 m) Late zone, massive, in tourmaline-cassiterite vein (DDH no. 3 to the Mariano vein, 182.5 m) Late zone, massive, in tourmaline-cassiterite vein (DDH no. 3 to the Mariano vein, 182.5 m) Large subhedral crystals (early zone) center of quartz-chlorite-cassiterite-arsenopyrite vein Anhedral grain, with arsenopyrite, center of quartz-chlorite-cassiterite-arsenopyrite vein Large subhedral crystals (latest zone) center of quartz-chlorite-cassiterite-arsenopyrite vein Large subhedral crystals (late zone) center of quartz-chlorite-cassiterite-arsenopyrite vein Inclusions in massive pyrite ore Coarse-grained massive ore Narrow veinlet in chalcopyrite ore Coarse-grained, massive ore Vug fillings in quartz vein Anhedral grains in quartz Rim around pyrrhotite Massive vein, with quartz Massive vein, with quartz Fine-grained, replacing pyrrhotite, in quartz vein Euhedral, replacing fine pyrite in quartz vein Subhedral, in quartz vein Subhedral, in quartz vein Coarse-grained massive ore Euhedral crystals in chalcopyrite ore Coarse-grained massive ore Euhedral crystals in massive chalcopyrite ore (San German vein, level 4650 m) Veinlets crosscutting chalcopyrite (San German vein, level 4650 m) Coarse-grained massive ore (San German vein, level 4650 m) Veinlets crosscutting chalcopyrite (San German vein, level 4650 m) Subhedral inclusions in sphalerite, in quartz-chalcopyrite-sphalerite vein Coarse-grained, overgrowing quartz, in quartz-chalcopyrite-sphalerite vein Coarse-grained, overgrowing chalcopyrite, in quartz-chalcopyrite-sphalerite vein Subhedral inclusions in sphalerite, in quartz-chalcopyrite-sphalerite vein Coarse-grained, overgrowing chalcopyrite, in quartz-chalcopyrite-sphalerite vein Coarse-grained, overgrowing chalcopyrite, in quartz-chalcopyrite-sphalerite vein Subhedral inclusions in sphalerite, in quartz-chalcopyrite-sphalerite vein Subhedral inclusions in sphalerite, in quartz-chalcopyrite-sphalerite vein Anhderal inclusions in quartz, in quartz-chalcopyrite vein Subhedral inclusions in quartz, in quartz-chalcopyrite vein Massive, coarse-grained ore, with quartz Veinlets crosscutting chalcopyrite Massive, in the central part of cassiterite-quartz-chlorite-chalcopyrite vein Massive, in the central part of cassiterite-quartz-chlorite-chalcopyrite vein Massive, replacing pyrrhotite, in chalcopyrite-arsenopyrite-pyrite vein Massive, replacing pyrrhotite, in chalcopyrite-arsenopyrite-pyrite vein Fine-grained, replacing pyrrhotite, in chalcopyrite-arsenopyrite-pyrite vein Euhedral crystal in chalcopyrite, in quartz-chalcopyrite-pyrrhotite-arsenopyrite vein Anhedral inclusions in chalcopyrite, in quartz-chalcopyrite-pyrrhotite-arsenopyrite vein Coarse-grained, massive ore, in quartz-chalcopyrite-pyrrhotite-arsenopyrite vein Anhedral inclusions in chalcopyrite, in quartz-chalcopyrite-pyrrhotite-arsenopyrite vein Euhedral crystal in chalcopyrite, in quartz-chalcopyrite-pyrrhotite-arsenopyrite vein Anhedral crystals in quartz-chalcopyrite-pyrite-sphalerite-galena vein Anhedral crystals in quartz-chalcopyrite-pyrite-sphalerite-galena vein Anhedral crystals in quartz-chalcopyrite-pyrite-sphalerite-galena vein Anhedral crystals in quartz-chalcopyrite-pyrite-sphalerite-galena vein Inclusions in large pyrite crystal (Jorge vein, level 4000 m) Large, blocky crystal replacing pyrrhotite (Jorge vein, level 4000 m) Coarse-grained, massive ore (Jorge vein, level 4000 m) Coarse-grained, massive ore (Jorge vein, level 4000 m)
34SV-CDT () 2.0 3.0 3.6 5.4 3.7 3.8 3.7 2.4 6.2 3.6 3.9 4.2 3.9 4.4 5.4 4.0 3.8 2.3 4.2 4.7 3.7 3.0 3.1 3.8 4.2 5.9 5.3 4.8 4.9 3.9 3.9 3.4 3.7 4.0 1.4 3.0 6.3 5.1 5.4 4.8 1.2 2.4 3.1 5.6 5.1 5.1 5.5 5.0 3.4 3.8 4.5 5.0 5.8 4.9 5.6 5.1 3.3 3.9 4.5 3.7 3.7 4.8 4.0 5.0
Early, barren stage D3-39-1 Arsenopyrite D3-39-2 Arsenopyrite D3-27-1 Arsenopyrite D3-27-2 Sphalerite D3-27-3 Arsenopyrite D8-50-1 Arsenopyrite D8-50-2 Arsenopyrite D3-182b-1 Arsenopyrite D3-182b-2 Arsenopyrite D3-182b-3 Arsenopyrite Sulfide ore stage 511b-1 Arsenopyrite 511b-2 Sphalerite 511a-1 Arsenopyrite 511a-2 Arsenopyrite A13-1 Galena A13-2 Pyrite R729-1 Pyrite R729-2 Chalcopyrite R17-1 Chalcopyrite R17-2 Pyrite R573-1 Chalcopyrite R573-2 Pyrrhotite R573-3 Pyrrhotite R404-1 Pyrite R404-2 Pyrite R404-3 Sphalerite R404-4 Galena R165-1 Chalcopyrite R165-2 Pyrite R165-3 Chalcopyrite G4-1 Arsenopyrite G4-2 Pyrrhotite G4-3 Chalcopyrite G4-4 Pyrrhotite R428-1 Galena R428-2 Chalcopyrite R428-3 Sphalerite R428-4 Galena R428-5 Sphalerite R428-6 Sphalerite R428-7 Galena R428-8 Galena R726-1 Galena R726-2 Sphalerite R726-3 Chalcopyrite R119-1 Pyrite R119-2 Chalcopyrite R119-3 Chalcopyrite R65b-1 Pyrite R65b-2 Pyrite R65b-3 Pyrite R55-1 Arsenopyrite R55-2 Pyrrhotite R55-3 Chalcopyrite R55-4 Pyrrhotite R55-5 Arsenopyrite R28a-1 Galena R28a-2 Sphalerite R28a-3 Sphalerite R28a-4 Galena J410-1 Pyrrhotite J410-2 Pyrite J410-3 Pyrrhotite J410-4 Pyrrhotite
Notes: All samples were analyzed in situ by a laser combustion system; unless indicated otherwise, the samples are from the San Rafael vein and their location is shown in Figure 2 0361-0128/98/000/000-00 $6.00
234
235
results of the analyses are listed and illustrated in Table 2 and Figure 9, respectively, and show that the sulfide ores have a relatively uniform 34S composition, ranging between 2 and 6 per mil. Although the range of 34S values is narrow, as might be expected from a deposit associated with a S-type granitoid (Ohmoto and Goldhaber, 1997), minerals within individual samples commonly display variable 34S values. For example, the 34S values of arsenopyrite in sample D3-182b vary within the same growth band from 3.6 to 6.2 per mil, whereas the 34S of galena in sample R428 varies from 1.2 to 5.1 per mil within only a few millimeters (Fig. 8; Table 2). There is, however, little variation in the 34S values of the sulfide minerals with respect to their location in the deposit. Arsenopyrite (the only sulfide that occurs throughout the entire paragenesis) displays a distinct trend of increasing 34S values from stage I (2.03.8) to stage III (3.85.1%).
Oxygen isotope data Thirty-nine oxygen isotope analyses were performed on carefully handpicked mineral separates of oxides and silicates from 17 samples, including a complexly banded, 5-cm-thick, quartz-chlorite-cassiterite-wolframite-chalcopyrite-pyrite vein, which was sampled in detail (Fig. 4). In addition, 5 whole-rock samples of granitoids (ranging from fresh to strongly chloritized) and 12 vein carbonates (see section below) were analyzed. The oxide and silicate minerals analyzed were tourmaline, quartz, chlorite, cassiterite, and wolframite, and the results are summarized in Figure 10 and listed in Table 3. The 18O values for the individual minerals are all positive and have relatively narrow ranges of 9.7 to 11.9 per mil for tourmaline, 9.7 to 14.4 per mil for quartz, 3.9 to 4.6 per mil for chlorite, 1.8 to 3.4 per mil for cassiterite, and 1.7 to 3.0 per mil for wolframite. The transition from the early, barren stage I to the tin-rich stage II is characterized by a marked decrease in the 18O values of tourmaline. Within the same vein the 18O value of early, orange-colored tourmaline is 11.7 per mil, whereas the succeeding dark-green tourmaline that is associated with abundant cassiterite and chlorite (the first to form in the paragenesis) has a 18O value of 9.7 per mil (sample D3-182, Table 3). In the subsequent stages (II and III), the 18O values of cassiterite and quartz are variable, but nevertheless provide evidence for an overall decrease of 18O with time, which for the ore vein investigated in detail corresponds to a change from 3.4 to 2.7 per mil for cassiterite, and from 10.6 to 9.8 per mil for quartz (sample R 119, Fig. 4; Table 3). The latest cassiterite in the paragenesis (stage III needle-tin cassiterite, intergrown with chalcopyrite from the upper zone of the deposit), has an even lower 18O value of 1.8 per mil (sample R725, Table 3). By contrast, quartz from the paragenetically latest stage (IV) has markedly higher 18O values, in the range of 11.3 to 14.4 per mil. Whole-rock 18O values of fresh granitic rocks range from 9.6 (biotite-granodiorite) to 10.9 per mil (leucogranite). A sequence of four slabs of granodiorite with increasing chloritic alteration was investigated in detail and shows a systematic decrease in 18O with alteration, down to 8.0 per mil for the strongly chloritized granodiorite and 4.9 per mil for the chloritite, which occupies the center of the vein (Table 3). Although a lack of complete isotopic equilibrium between coexisting quartz and cassiterite has been reported for a number of hydrothermal tin lodes (Alderton, 1989), several isotopic geothermometers were tested on the silicate-oxide assemblages (quartz, cassiterite, wolframite, chlorite) from the banded ore vein investigated in detail (sample R119, Fig. 4). Because quartz and cassiterite could not always be separated from the same growth band, we calculated oxygen isotope temperatures based on data for pairs of minerals intimately intergrown in a single band, as well as for pairs where in which the minerals were in adjacent bands. Overall, the geothermometers (quartz-cassiterite, quartz-wolframite, quartzchlorite), which are all based on carefully conducted experiments (Matsuhisa et al., 1979; Zhang et al., 1994; Cole and Ripley, 1999) give average equilibrium temperatures in the range of 370 to 450C, about 40 to 60 C higher than the fluid inclusion homogenization temperatures. The error on
8 6 4 2 0 6 4 2 0 6 4 0
Arsenopyrite a
0 1 2 3 4 5 6 7 8 9 10
b
1 2 3 4 5 6
Chalcopyrite
7 8 9 10
Frequency
2 0 6 4 2 0 6 4 2 0 6 4 2 0 0 0 0
c
1 2 3 4 5 6 7
Pyrrhotite
8 9 10
d
1 2 3 4 5 6 7 8
Pyrite
9 10
e
0 1 2 3 4 5 6 7
Sphalerite
8 9 10
f
1 2 3 4 5 6 7
Galena
8 9 10
34SV-CDT ()
FIG. 9. Histograms displaying the sulfur isotope composition of San Rafael sulfides. 0361-0128/98/000/000-00 $6.00
235
236
6 4 2 0 6 4 2 0 0 0
WAGNER ET AL.
a
1 2 3 4 5
Tourmaline
6 7 8 9 10 11 12 13 14 15 16
b
1 2 3 4 5 6
Quartz
7 8 9 10 11 12 13 14 15 16
Frequency
6 4 2 0 6 4 2 0 6 4 2 0 0 0 0
c
1 2 3 4 5
Chlorite
6 7 8 9 10 11 12 13 14 15 16
d
1 2 3 4
Cassiterite
5 6 7 8 9 10 11 12 13 14 15 16
e
1 2 3 4
Wolframite
5 6 7 8 9 10 11 12 13 14 15 16
18OV-SMOW ()
FIG. 10. Histograms displaying the oxygen isotope composition of San Rafael silicate and oxide minerals.
the calculated equilibrium temperatures, considering the analytical uncertainty of the 18O values, is on the order of about 30C. When different stages of the vein evolution are compared, the calculated oxygen isotope temperatures (e.g., the quartz-cassiterite pair for which most data are available) show a general decrease from the early stage II assemblage (450C) to the later complex cassiterite-quartz-wolframite-chlorite assemblage (400420C). Hydrogen isotope data Four tourmaline and two chlorite samples were analyzed for their hydrogen isotope composition and the results are listed in Table 4. The D values of tourmaline lie in a narrow range between 75 and 73 per mil and are essentially identical within errors. The D values of the two chlorite samples are 88 and 81 per mil. Carbon and oxygen isotope data of carbonates The oxygen and carbon isotope data for late-stage vein carbonates (5 samples of calcite and 7 of siderite) are shown in Figure 11 and presented in Table 5. The 18O and 13C values of calcite samples are in the range of 8.6 to 14.2 per mil and of 8.5 to 3.2 per mil, respectively. The data for siderite are more variable, and are in the range of 13.8 to 23.1 per mil and 6.5 to 4.3 per mil, respectively (Fig. 11). In two samples,
0361-0128/98/000/000-00 $6.00
two successive generations of siderite were studied, but did not yield a consistent trend. Discussion Conditions of ore deposition Based on our geologic and textural observations, the hydrothermal history of the ore-forming system at San Rafael comprised two contrasting episodes. The first of these (stage I) was characterized by the formation of numerous tourmaline-quartz ( arsenopyrite) veins, veinlets, and stringers, with which sericitic and tourmaline alteration are associated. It is important to note that these structures are typically quite narrow, locally discontinuous, and invariably sealed, likely indicating pressures approaching lithostatic. Stratigraphic reconstruction based on regional geologic data (e.g., Clark et al., 1990) suggests that about 2.5 km of rock has been eroded since the emplacement of the deposit (at about 25 Ma). This translates into a lithostatic fluid pressure of 700 to 800 bars, which is consistent with pressure estimates given by Kontak and Clark (2002). Fluid inclusion data demonstrate that the fluids responsible for this early, barren stage were hot, hypersaline brines (fluid trapping temperatures of 380540C based on isochoric projection of fluid inclusion Th values to 700800 bars, salinity of 3462 wt % NaCl equiv) and the
236
STABLE ISOTOPE CONSTRAINTSSAN RAFAEL SN-CU DEPOSIT, SE PERU TABLE 3. Oxygen Isotope Data for Quartz, Cassiterite, Wolframite, Chlorite, Tourmaline, and Whole-Rock Samples from the San Rafael Sn-Cu Deposit (n = 44) Sample no. Mineral Textural relationships Tourmaline vein, euhedral, prismatic tourmaline (DDH # 5 to the San Rafael vein, 70 m) Quartz-tourmaline vein (DDH #5 to the San Rafael vein, 94 m) Quartz-tourmaline vein (DDH #5 to the San Rafael vein, 94 m) Tourmaline-quartz vein (DDH #1 to the San Rafael vein, 280 m) Quartz-tourmaline veinlet: the quartz is very early and borders the tourmaline vein (DDH no. 3 to the Mariano vein, 69 m) Tourmaline-cassiterite-chlorite-arsenopyrite vein (early, orange tourmaline) (DDH no. 3 to the Mariano vein, 182.5 m) Tourmaline-cassiterite-chlorite-arsenopyrite vein (late, green tourmaline, associated with cassiterite) (DDH no. 3 to the Mariano vein, 182.5 m) Tourmaline-cassiterite-chlorite-arsenopyrite vein (earliest cassiterite in the deposit) (DDH no. 3 to the Mariano vein, 182.5 m) Banded quartz-chlorite-cassiterite-wolframite-chalcopyrite-pyrite vein (Fig. 5) Euhedral quartz associated with chlorite, in between cassiterite bands (Qtz IB) Large, euhedral quartz crystals, preceding the deposition of massive cassiterite (Qtz IV) Small, euhedral quartz crystals, with chlorite and needle-tin cassiterite (Qtz VI) Large, euhedral quartz crystals, preceding needle-tin cassiterite deposition (Qtz VII) Large, subhedral quartz crystals, following needle-tin cassiterite deposition (Qtz VIII) Very late, barren, discordant quartz-chlorite veinlet (Qtz X) Earliest layer of massive, dark-brown cassiterite (Cas I) Earliest zone of a 7-cm-thick layer of dark-brown, massive cassiterite (Cas IIIA) Middle zone of a 7-cm-thick layer of dark-brown, massive cassiterite (Cas IIIB) Latest zone of a 7-cm-thick layer of dark-brown, massive cassiterite (Cas IIIC) Early needle-tin cassiterite (Cas IV) Late, needle-tin cassiterite, associated with chalcopyrite (Cas V) Chlorite, associated with quartz and needle-tin cassiterite Wolframite, associated with chlorite and quartz Same occurrence as R119 wol Cassiterite, quartz and chlorite-cemented breccia of chloritized wall-rock fragments Euhedral quartz preceding cassiterite deposition Layer of massive, dark-brown cassiterite Euhedral quartz following cassiterite deposition Early layers of light-colored wood-tin cassiterite (San Rafael Lode, elev 4500 m) Late layers of dark wood-tin cassiterite (San Rafael Lode, elevation 4500 m) Cassiterite-chlorite-quartz-arsenopyrite vein Large, euhedral quartz crystals from a cassiterite-chlorite-quartz-arsenopyrite vein Quartz-chlorite-cassiterite-wolframite vein (DDH no. 1 to the San Rafael vein, 219 m) Quartz-chlorite-cassiterite-wolframite vein (DDH no. 1 to the San Rafael vein, 219 m) Needle tin cassiterite-chalcopyrite-quartz vein (tin ore from the upper zone) Anhedral quartz from a large quartz-sphalerite vein Subhedral quartz from a major quartz-chalcopyrite-fluorite-siderite vein Late-stage, very large crystal of euhedral vug quartz (core) Late-stage,very large crystal of euhedral vug quartz (outer zone) Late-stage, barren, very large quartz vein Quartz-calcite fill of a reopened tourmaline vein (DDH no. 5 to the Mariano vein, 155 m) Fresh alkali-feldspar cordierite-biotite leucogranite (Ramp, elev 3930 m) Fresh biotite-granodiorite (DDH no. 1 to the Mariano vein, 61 m) Mildly altered biotite-granodiorite (DDH no. 1 to the Mariano vein, 61 m) Strongly chloritized biotite-granodiorite (DDH no. 1 to the Mariano vein, 61 m) Chloritite adjacent to the center of a chloritic vein (DDH no. 1 to the Mariano vein, 61 m)
237
18OV-SMOW () 11.5 11.9 12.3 10.8 12.0 11.7
Early, barren stage D5-70 Tourmaline D5-94 tou Tourmaline D5-94 qz Quartz D1-280 Tourmaline D3-69 Quartz D3-182 a Tourmaline
Main ore stage (tin and copper) D3-182 b Tourmaline D3-182 b R119 R119 qz1 R119 qz2 R119 qz3 R119 qz4 R119 qz5 R119 qz6 R119 cas1 R119 cas2 R119 cas3 R119 cas4 R119 cas5 R119 cas6 R119 chl R119 wol R113 R137 R137 qz1 R137 cas R137 qz2 D4 cas1 D4 cas2 R511 chl R511 qz D1-219 wol D1-219 qz R725 R429 R708 Cassiterite Quartz Quartz Quartz Quartz Quartz Quartz Cassiterite Cassiterite Cassiterite Cassiterite Cassiterite Cassiterite Chlorite Wolframite Wolframite Quartz Cassiterite Quartz Cassiterite Cassiterite Chlorite Quartz Wolframite Quartz Cassiterite Quartz Quartz
9.7 2.0 10.6 10.6 10.1 9.7 10.8 9.8 3.4 2.9 3.3 3.3 2.2 2.7 4.6 3.0 2.1 10.8 2.7 11.2 2.3 3.4 3.9 12.2 1.7 10.1 1.8 9.8 11.1 14.0 13.1 14.4 11.3 10.9 9.6 9.6 8.0 4.9
Late, barren stage R399 qz1 Quartz R399 qz2 Quartz R709 Quartz D5M-155 Quartz Whole-rock samples I20 Whole rock D1M-61d Whole rock D1M-61c Whole rock D1M-61b Whole rock D1M-61a Whole rock
Notes: Unless indicated otherwise, the samples are from the San Rafael vein and their location is shown in Figure 2
alteration minerals that formed indicate that they were acidic (Mlynarczyk et al., 2009, submitted). By contrast, the subsequent stages (i.e., tin-mineralized stage II, copper-mineralized stage III, and late, barren quartz stage IV) are all associated with a distinctive, strong chloritic alteration (and locally, silicification). The veins invariably have
0361-0128/98/000/000-00 $6.00
an open fracture-filling character, with ubiquitous vugs, likely indicative of hydrostatic conditions. Evidence for multiple vein-opening events suggests that the pressure must have fluctuated from slightly above lithostatic (700800 bars) to near hydrostatic (200300 bars) during ore deposition (assuming a roughly similar depth to that during stage I). The
237
238
WAGNER ET AL. TABLE 4. Hydrogen Isotope Data for Tourmaline and Chlorite from the San Rafael Sn-Cu Deposit (n = 6)
Sample no.
Mineral
Textural relationships Tourmaline vein, euhedral, prismatic tourmaline (DDH no. 5 to the San Rafael vein, 70 m) Tourmaline-quartz vein (DDH #1 to the San Rafael vein, 280 m) Tourmaline-cassiterite-chlorite-arsenopyrite vein (early, orange tourmaline) (DDH no. 3 to the Mariano vein, 182.5 m) Tourmaline-cassiterite-chlorite-arsenopyrite vein (late, green tourmaline, associated with cassiterite) - DDH no. 3 to the Mariano vein, 182.5 m Associated with quartz and needle-tin cassiterite, from a complexly banded quartz-chlorite -cassiterite-wolframite-chalcopyrite-pyrite vein (Fig. 5) Cassiterite-chlorite-quartz-arsenopyrite vein
DV-SMOW () 75 75 74
Early, barren stage D5-70 Tourmaline D1-280 Tourmaline D3-182 a Tourmaline Main ore stage D3-182 b Tourmaline R119 chl R511 chl Chlorite Chlorite
73 88 81
fluids circulating in the system were much cooler (260 380C, calculated from isochoric projection of fluid inclusion Th values) and had a moderate to low salinity (021 wt % NaCl equiv). During the main cassiterite stage, fluid temperatures were in the range of 370380C, whereas quartz was deposited at consistently lower temperatures of 290330C. Because the vein textures indicate multiple repetition of cassiterite and quartz deposition, the observed temperature difference between cassiterite- and quartz-hosted fluid inclusions likely reflects a sequence of distinct pulses of hotter magmatic and cooler meteoric fluids that mixed repeatedly in a hydrologically very dynamic system. The quartz-cassiterite oxygen isotope temperatures calculated for this stage (400420C) are in reasonable agreement with the temperature estimate obtained from pressurecorrected fluid inclusion data. However, as it was commonly necessary to base the isotopic temperatures on cassiterite and quartz from adjacent layers, which as noted above formed at different temperatures, the isotopic temperatures
6 4
Calcite Siderite
()
2 0 -2 -4 -6 -8 -10 5 10
18O
15
V-SMOW
20
25
()
FIG. 11. A plot of 13C versus 18O for different carbonates from the San Rafael deposit. The arrows link successive generations of siderite from the same sample. The two data points in the lower right corner of the diagram correspond to samples from the uppermost part of the deposit (elev 4820 m); the remaining samples originate from or below an elevation of 4330 m. 0361-0128/98/000/000-00 $6.00
are considered less reliable than the pressure-corrected fluid inclusion temperatures. The transition between the early, barren stage I and the subsequent tin and copper stages is recorded by rare tourmaline-cassiterite-chlorite-arsenopyrite veins, in which the early and main stage mineral assemblages overlap, indicating that stage II may have occurred very soon after stage I (Mlynarczyk and Williams-Jones, 2006). We emphasize that these veins are of the open fracture-filling type and are surrounded by envelopes of strong chloritic alteration, identical to the ore-stage veins. This observation is very important, as it implies that the onset of ore deposition (stage II) and the marked change in vein and alteration mineralogy were directly related to a change in structural style, i.e., a transition from a closed vein system (lithostatic conditions) to an open vein system (hydrostatic conditions). Although the tin stage (stage II) and the copper stage (stage III) were part of the same, protracted hydrothermal event, and were associated with the same chloritic alteration, the physico-chemical conditions of the ore fluid were markedly different, and, therefore, resulted in the formation of contrasting mineral assemblages. In order to constrain the depositional conditions for each stage, we have thermodynamically modeled them in fO2-pH and fO2-aS space, using phase diagrams combining aqueous and mineral equilibria in the CuFe-Sn-As-S-O-H system (Fig. 12). In addition, we have modeled the speciation of tin in the system Sn-Na-Cl-O-H, to constrain the total solubility of Sn as a function of pH, and the oxidation state of the hydrothermal fluid (Fig. 13). The details of the calculation methods and the source of thermodynamic data used are listed in the Appendix. As discussed above, the temperature and pressure during cassiterite precipitation (stage II) were approximately 370 to 380C and 200 to 300 bars, respectively, and the fluid had a salinity of about 20 wt percent NaCl equiv. The equilibrium mineral assemblage was very simple and consisted of cassiterite, Fe-rich chlorite (daphnite), and quartz. Subordinate arsenopyrite, which occurs locally, was the only sulfide present, and in one sample minor hematite was observed to grow coevally with cassiterite and chlorite (Mlynarczyk et al., 2009, submitted). Based on the observed mineral assemblages and the activity diagrams (Fig. 12), it is concluded that the conditions during precipitation of cassiterite evolved
13C V-PDB
238
STABLE ISOTOPE CONSTRAINTSSAN RAFAEL SN-CU DEPOSIT, SE PERU TABLE 5. Carbon and Oxygen Isotope Data for Vein Carbonates from the San Rafael Sn-Cu Deposit (n = 12) Sample no. Mineral Textural relationships Quartz-calcite-sphalerite-chalcopyrite infill of a reopened tourmaline vein (DDH no. 1 to the San Rafael vein, 118.5 m) Quartz-calcite infill of a reopened tourmaline vein (DDH no. 5 to the San Rafael vein, 137.9 m) Quartz-calcite infill of a reopened tourmaline vein (DDH no. 5 to the San Rafael vein, 165.8 m) Quartz-calcite-arsenopyrite infill of a reopened tourmaline vein (DDH no. 4 to the San Rafael vein, 230.5 m) Quartz-calcite infill of a reopened tourmaline vein (DDH no. 5 to the Mariano vein, 155 m) Early, anhedral, reddish, opaque siderite from a quartzsphalerite-pyrite-marcasite-chalcopyrite-galena-siderite vein Late, euhedral, greenish, translucent siderite from a quartzsphalerite-pyrite-marcasite-chalcopyrite-galena-siderite vein Late, euhedral, greenish, translucent siderite from a quartzsphalerite-pyrite-marcasite-chalcopyrite-galena-siderite vein Late, orange, translucent siderite from a quartz-cassiteritechlorite-siderite vein Greenish, translucent siderite from a quartz-fluorite-pyritemarcasite-siderite vein Early, anhedral, reddish, opaque siderite from a quartzchalcopyrite-fluorite-siderite vein Late, euhedral, greenish, translucent siderite from a quartzchalcopyrite-siderite veinlet in chloritized wall rock 13CV-PDB () 8.5 4.0 7.8 3.2 3.5 0.3 0.0 0.3 0.3 4.3 6.5 4.9
239
18OV-SMOW () 8.6 9.4 9.6 14.2 9.1 18.8 15.7 13.8 17.7 15.7 23.2 22.0
Late, barren stage D1-118 Calcite D5-137 D5-165 D4-230 D5M-155 R27 sid1 R27 sid2 R28 sid R57 sid R123 sid R708 sid2 R708 sid1 Calcite Calcite Calcite Calcite Siderite Siderite Siderite Siderite Siderite Siderite Siderite
from moderately reduced (log fO2 about 28 to 26, upper part of the arsenopyrite stability field) to more oxidizing (log fO2 above 25, magnetite-hematite boundary). The pH evolved from acidic to more alkaline, with the final pH at the end of massive cassiterite precipitation being slightly above about 4. At this pH value, the cassiterite solubility contours flatten out considerably (Fig. 13), i.e., the solubility decreases only very slightly upon further increase in pH. The sulfur concentration in the fluid is difficult to constrain, but a minimum value of aS of about 0.005 to 0.01 can be estimated from the stability field of arsenopyrite (Fig. 12b). During the sulfide stage (stage III), the temperature and pressure were approximately 300C (based on fluid inclusion trapping temperatures and stannite-sphalerite thermometry) and 200 to 300 bars (assuming hydrostatic conditions and a roughly similar depth to that during the previous mineralization stages), and the fluid salinity was around 5 wt percent NaCl equiv. The mineral assemblage consists mainly of chalcopyrite, pyrrhotite, Fe-rich chlorite (daphnite), quartz, and minor cassiterite, but locally other sulfides (e.g., arsenopyrite, sphalerite, galena, pyrite) are also present. Figure 12 shows the most relevant mineral stability relations, as well as the sulfur isotope contours that were calculated for chalcopyrite (see below). It can be deduced from the stability of chlorite (i.e., the daphnite end member as calculated from the measured chlorite composition) and pyrrhotite that during most of stage III, the oxidation state was lower (log fO2 below 35, upper daphnite stability limit in Fig. 12c) than during stage II, while the pH was likely similar (around 4). It should be noted, however, that in many locations in the deposit, pyrite is an abundant sulfide of stage III mineralization, and pyrrhotite displays an incipient or advanced replacement by pyrite, marcasite, or rare hematite. Although in some case this replacement of
0361-0128/98/000/000-00 $6.00
pyrrhotite was clearly late, the common presence of pyrite in stage III ores indicates that there must have been large fluctuations in fO2, and that the hydrothermal environment became progressively more oxidizing (i.e., moved from the pyrrhotite to the pyrite stability field in Fig. 12c) with time. Furthermore, the ubiquitous replacement of pyrrhotite by pyrite might indicate that the sulfidation state of the system increased with time (Fig. 12d), a feature that has been observed for many magmatic-hydrothermal systems (Einaudi et al., 2003). Additional information about the oxidation state and the fluid evolution during stage III mineralization comes from the sulfur isotope data. The calculated sulfur isotope contours show that an increase in oxidation state above a log fO2 of 31 to 29 would result in a considerable shift in the 34S of precipitated sulfides towards more negative values (Fig. 12). From the relatively narrow range in 34S values of the San Rafael sulfides it can be concluded that although the conditions during stage III became more oxidizing with time, they did not reach as high as the pyrite-hematite boundary (Fig. 12c). Source of sulfur The 34S values of sulfide minerals from the San Rafael deposit range between 2 and 6 per mil, and show relatively little variation with respect to the location in the deposit (see above). These rather uniform values point to a large-scale hydrothermal system with a homogeneous source of sulfur, which was likely of magmatic origin. As argued by Hattori and Keith (2001) for porphyry systems, the very narrow range in 34S values displayed by giant, granite-related deposits is best explained by a single magmatic source of sulfur, rather than by homogenization of the sulfur from a variety of sources in local country rocks having diverse S isotope compositions. It
239
240
Stability field of arsenopyrite
HSO4-
WAGNER ET AL.
T=380 C a P=300 bar Qtz saturation aH2O=0.873 Stability field of arsenopyrite T=380 C b P=300 bar Qtz saturation aH2O=0.873 pH=4
-18 -20 -22 -24
-18 -20 -22 -24
(20wt.% NaCl eq.) aS = 0.01
(20wt.% NaCl eq.)
hematite pyrite
SO42-
hematite
HSO4H2 S
Log fO2
Log fO2
-26
-26
pyrite magnetite
as asp
as as p
magnetite
-28 -30
-28
pyrrhotite
-32 -34 0
H2S
as loe
-30 -32
as loe
pyrrhotite
asp loe
2 4 6 8 10
HS-
as p loe
-3 -2 -1 0
12
-34 -4
pH
Stability field of chalcopyrite T=300 C c P=300 bar Qtz saturation aH2O=0.973
Log aS
Stability field of chalcopyrite T=300 C d P=300 bar Qtz saturation aH2O=0.973
-24
HSO4
-
prl
ms
ams=0.517
kfs
(5 wt.% NaCl eq.)
-24 -26
-26 -28
aK+= 0.005 aS = 0.01 34Stot=+3.0
(5 wt.% NaCl eq.)
pH=4 34Stot=+3.0
hematite
-28
HSO4-17.5 -15.0 -10.0 -5.0 0.0 +2.8
hematite
-30
Log fO2
Log fO2
bn+py cp pyrite
SO42-
-30 -32
-34
H2S
bn + cp py
pyrite
-32
-34
-17.5 -15.0 -10.0 -5.0 0.0 +2.8

+2.9
magnetite
dap
adap=0.18
+3.0
-36 -38 -40 0
magnetite
-36 -38
pyrrhotite
bn cp +p o
-4 -3 -2 -1
pyrrhotite cp bn+po
HS-
H2S
10
12
-40 -5
pH
Log aS
FIG. 12. Log fO2-pH and log fO2-aS diagrams showing stability relationships in the system Fe-O-S, for stage II (oxide assemblage, diagrams a and b) and stage III (sulfide assemblage, diagrams c and d) of the paragenesis. The temperature and fluid salinity for which the diagrams were drawn were constrained from fluid inclusion microthermometry, whereas pressure was estimated. Diagrams a and b also show stability relationships for the Fe-As-S system (asp = arsenopyrite, as = native arsenic, loe = loellingite). Diagrams c and d show the stability relationships in the Cu-Fe-S system (cp = chalcopyrite, bn = bornite, py = pyrite) and of the principal alteration minerals (dap = daphnite, prl = pyrophyllite, ms = muscovite, kfs = K feldspar). The activities of alteration minerals were calculated from microprobe data (see Appendix for details), and the chemical reactions were balanced assuming conservation of Al and quartz saturation. Diagrams c and d (sulfide stage) also show superimposed S isotope contours, calculated for chalcopyrite, assuming aS = 0.01 and 34Stotal= 3.0 , inferred from the S isotope composition of the sulfides from this stage (see text for details). 0361-0128/98/000/000-00 $6.00
240
241
1.0
20 0
SMOW
10-1
Total Sn solubility (mol/kg)
- 31
DV-SMOW ()
log f O2 = - 32
-20 -40 -60 -80
Tertiary Andean meteoric water ?
Groundwater Magmatic water box
- 29
- 28 - 27
-100 -120 -20
10-4
- 22 - 23 - 24 - 25
- 26
Me te
10-3
Present-day Andean meteoric water
ori cw ate
r lin
- 30
10-2
Chlorite
Tourmaline
-10
18O
0
V-SMOW
10
20
()
wol cas tou
10-5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
pH
FIG. 13. Total solubility of tin (mol/kg) in the system Sn-Na-Cl-O-H calculated as a function of pH and fO2 at 380C and 300 bar (the inferred conditions for stage II). For details of the calculations, see the Appendix.
qz
-20
-10
18O
0 chl
Ore fluid ()
10
20
should be noted though that the San Rafael sulfides have slightly higher 34S values than some other tin deposits (Kontak, 1990), implying an initial relative enrichment in the 34S content of the sulfur source. This is not unexpected for S-type granitic magmas, which acquire most of their sulfur through assimilation of country rocks. Thus, for such plutons, the 34S values of ore sulfides are very similar to those of local igneous rocks and local country rocks, and reliable discrimination between magmatic and sedimentary sulfur is not possible (Ohmoto and Goldhaber, 1997). Source of ore fluids We have calculated the oxygen and hydrogen isotope composition of the ore fluids (Fig. 14) during different mineralization stages to constrain the most likely fluid sources for the San Rafael deposit. The isotopic composition of the early fluids, which were in equilibrium with stage I tourmaline, is very close to magmatic water values, which, considering the high temperature and salinity of these fluids, supports the idea that they were magmatic in origin. The water in equilibrium with late, Fe-rich tourmaline, which formed at the onset of cassiterite precipitation (single analysis in the upper right corner of the magmatic water box, Fig. 14) however, was enriched in deuterium and had isotopically lighter oxygen. Compared to the tourmaline data, the chlorite data demonstrate a fluid evolution characterized by a marked decrease in 18O and a slight increase in D with time. The water in equilibrium with ore-stage chlorite had 18O lower by about 7 per mil and D higher by 10 per mil than that of the fluid in equilibrium with early tourmaline (Fig. 14). It should be noted, however, that the D of water in equilibrium with chlorite cannot be precisely constrained due to the lack of accurate fractionation factors (Graham et al., 1987). The trend of decreasing fluid 18O with time is, nevertheless, clearly indicated by the
0361-0128/98/000/000-00 $6.00
FIG. 14. Plot of 18O versus D for tourmaline (open squares) and chlorite (open circles) from the San Rafael deposit and calculated water in isotopic equilibrium with these minerals (black squares and circles). The thick dashed trendline from tourmaline to chlorite represents the inferred evolution of fluid (water) during the formation of the deposit. The lower diagram (18O axis) shows the distribution of the calculated 18O values for water in equilibrium with the different ore and gangue minerals (cas = cassiterite, chl = chlorite, qz = quartz, tou = tourmaline, wol = wolframite). The calculations applied isotopic fractionation equations that are mostly based on well-constrained experiments in the appropriate temperature range (Matsuhisa et al., 1979; Graham et al., 1984, 1987; Carothers et al., 1988; Kotzer et al., 1993; Zheng, 1993; Zhang et al., 1994; Cole and Ripley, 1999).
oxygen isotope compositions of successive generations of quartz and cassiterite (see earlier section), and also by minerals which formed late in the paragenesis, such as calcite and siderite. The decrease in ore fluid 18O with time is paralleled by marked corresponding decreases in fluid salinity and temperature, as indicated by the fluid inclusion data (Fig. 7). These trends, which appear broadly aligned, could represent mixing of the early, hot, hypersaline (presumably magmatic) brine with a relatively warm fluid that had consistently lower 18O and D values. Considering the shallow level of emplacement of the San Rafael pluton and fluid inclusion evidence for very low salinity fluids (this could not be produced by simple boiling of a magmatic brine; Kontak and Clark, 2002) circulating in the system, a good candidate for this external fluid would be heated groundwater of meteoric origin that had partly exchanged its oxygen isotope composition with the host rocks. Unfortunately, it is currently not possible to properly constrain the isotopic composition of the late Oligocene, Andean meteoric waters, coeval with formation of the deposit, to further substantiate this hypothesis. This is because of the very
241
242
WAGNER ET AL.
large uncertainty concerning the rate of the uplift in this part of the Andes during the last 25 m.y. It is generally assumed that the elevation of the Central Andes in the late Oligocene was about a third of their current elevation (Gregory-Wodzicki, 2000; Anders et al., 2002), implying that ancient meteoric waters must have had considerably higher 18O and D values than their present-day counterparts, but the actual values can only be roughly estimated (Fig. 14). Applying both closed- and open-system scenarios (Taylor, 1977, 1997) for oxygen isotope exchange between water of meteoric origin and typical Andean granites (e.g., Longstaffe et al., 1983), the calculated 18O values of hot groundwater at temperatures of 200 to 250C (assuming moderate fluid/rock ratios between 0.1 and 1.0) would be in the range between 2 and 2 per mil. Ore deposition processes The rich tin ores of the San Rafael deposit testify to an unusually effective mechanism of ore deposition, which caused substantial supersaturation of tin and focused mineralization into several large fault-jogs, at depth in the lode (Mlynarczyk et al., 2003). Experimental studies and chemical modeling have shown that in natural hydrothermal systems the solubility of tin is highest in hot, reduced, saline, and acidic solutions (Fig. 13), and that the bulk of the tin is transported as stannous (Sn2+) chloride complexes (Eugster and Wilson, 1985; Pabalan, 1986; Taylor and Wall, 1993; Mller and Seward, 2001). The precipitation of cassiterite (SnO2), in which tin is in the tetravalent state, can, therefore, be induced by an increase in fO2 as well as by decreases in temperature and ligand (chloride) ion activity and an increase in pH. Possible mechanisms for tin ore formation could include reaction of the ore fluids with the host rocks, boiling, redox-coupled precipitation, mixing of the ore fluids with fluids of a markedly different composition, or a combination of some of the above (Eadington, 1985; Heinrich, 1990, 1995). The fluid-rock interaction, which accompanied ore deposition at San Rafael, produced a wide envelope of very strong chloritization but is not considered to have influenced ore formation as mass-balance calculations show that such alteration increases the acidity of the fluid (Mlynarczyk et al., 2009, submitted), leading to an increase in cassiterite solubility (Fig. 13). Redox-coupled precipitation (e.g., coupled precipitation of arsenopyrite and cassiterite) is also not considered to have played a significant role because it would have led to considerable heterogeneity in the compositions of the ore and alteration minerals, which was clearly not the case. On the other hand, structural evidence for a transition from a closed to an open vein system, and a concomitant change from lithostatic to hydrostatic conditions, implies that boiling may have taken place. Boiling would have oxidized the ore fluid, decreased its temperature and significantly increased its pH due to the loss of acidic volatiles. Thus, it would have promoted the precipitation of cassiterite. However, vapor-rich fluid inclusions are rare and are only found in stage 1 veins that formed prior to tin mineralization. Thus, if boiling occurred during the main ore-forming events, it was probably of limited extent. By contrast, the coincidence of the ore stage with the opening of the vein system and the widespread appearance of cooler, much more dilute fluids (down to almost zero salinity), as well as the ensuing periodic fluctuation
0361-0128/98/000/000-00 $6.00
of fluid salinity and temperature (above 360C and about 21 wt % NaCl equiv during formation of massive layers of cassiterite, but below 300C and 216 wt % NaCl equiv during formation of intervening quartz layers) is readily explained by incursions of heated groundwaters of meteoric origin which mixed with magmatic fluids (Mlynarczyk et al., 2003). A model invoking the mixing of magmatic fluids with waters having lower 18O and D could also explain the observed systematic decrease of cassiterite and quartz 18O values with time. Assuming that hot groundwaters of meteoric origin would not have had sufficient time to fully equilibrate with the host rocks, they likely would be cooler and much more oxidizing than the magmatic fluids. Numerical simulation of complex fluid-granite equilibria (Dolejs and Wagner, 2008) has shown that rock-buffered low-salinity aqueous fluids will become much more oxidizing as temperature decreases from 400 to 200C. Therefore, mixing of hot groundwaters with tin-bearing magmatic fluids would have oxidized the fluid system, increased the overall pH, and decreased temperature and ligand activity, all of which would have destabilized tin chloride complexes and triggered cassiterite precipitation. A further evaluation of boiling and fluid mixing scenarios and their relative importance is possible from the evolution of the oxygen isotope composition of the ore fluid. Boiling increases the 18O of the remaining liquid because of preferential partitioning of the light isotope into the vapor, and this is reflected in the paragenesis by a gradual increase in 18O with time. By contrast, mixing of magmatic waters with groundwaters of meteoric origin (isotopically light) should lead to progressive decreases in the 18O values of the ore fluid. It should be noted, however, that the temperature decrease associated with mixing will increase the isotopic fractionation between the ore fluid and the precipitating minerals, which, in the case of the oxygen isotope composition of cassiterite and wolframite, will partly offset the effect of addition of low18O groundwaters. Because of such complexities, quantitative modeling of boiling and fluid mixing mechanisms of tin ore formation are required to discriminate between them. Modeling of fluid evolution scenarios To test the hypothesis of fluid mixing and compare it to one of boiling of the ore fluid, the oxygen isotope composition of silicate and oxide minerals precipitating by each mechanism was calculated as a function of temperature. Because the available experimental studies of isotopic liquid-vapor partitioning in the H2O-NaCl system are restricted to conditions below the critical point of pure water (Horita et al., 1993, 1995), it was not possible to adequately model boiling for the temperature-composition space at San Rafael. Based on boiling models for low-salinity fluids (e.g., Truesdell and Nathenson, 1977; Wagner et al., 2005), it can be predicted that boiling would produce a progressive increase in mineral 18O values. This would be the consequence of preferential partitioning of the lighter 16O isotope into the vapor phase (Horita and Wesolowski, 1994), yielding an increasingly 18O-rich residual brine. Furthermore, the effect would be enhanced by the cooling of the residual brine as a consequence of enthalpy transfer from the liquid to the vapor that is required to maintain the enthalpy balance. Hence, the first-order effect of boiling on the 18O of the precipitated minerals is broadly
242
243
similar to that of fluid cooling and can therefore be simulated by the latter. Based on the fluid inclusion data and the measured oxygen isotope data for stage I silicates, the parameters of the starting magmatic fluid were assumed to be a temperature of 500C, a salinity of 45 wt percent NaCl equiv, and a 18O value of 11 per mil. One model considered that this fluid cooled due to boiling down to a temperature of 200C, whereas the other considered its mixing with a cooler, dilute fluid (230C, zero salinity), assuming three possible 18O compositions of 2.0, 0.0, 2.0 percent representing the composition of hot groundwaters of meteoric origin that had exchanged their oxygen with their host rocks. A stepwise decrease of ore fluid temperature took place in each model and in the case of mixing was accompanied by a decrease in salinity and fluid 18O, proportional to the aliquot of diluting fluid added. For simplicity, a linear relationship between the temperature and composition of the fluid mixture was assumed, although it is actually the enthalpy, not the temperature, that varies linearly with composition during mixing (e.g., Reed and Spycher, 1984; Spycher and Reed, 1989). The influence of fluid salinity on the concentration of oxygen was taken into account (Schwinn et al., 2006) because the salinity of the magmatic end member was quite high but, in the absence of pertinent experimental data, the mineral-water fractionation factors were assumed to be salinity independent. The latter were calculated for every temperature step (using the equations of Matsuhisa et al., 1979; Zhang et al., 1994; and Cole and Ripley, 1999), enabling the 18O composition of quartz, cassiterite, chlorite, and wolframite in equilibrium with the fluid to be determined. Selected modeling results are presented in Table 6 and the calculated changes in the 18O values of quartz and cassiterite
with time are shown in Figure 15, together with the measured ranges of 18O values and fluid inclusion trapping temperatures for ore stage quartz and cassiterite. From the calculations, it is evident that boiling and fluid mixing have a contrasting influence on the oxygen isotope composition of the precipitating minerals. Whereas mixing with hot, low-18O fluid produces a systematic decrease in mineral 18O values (followed by a minor increase below 250C), boiling results in a systematic increase of the 18O values of the oxides and silicates with time, a scenario clearly not supported by the paragenetic information and the isotopic data for San Rafael. In addition to the good agreement between the predicted and observed trend of decreasing mineral 18O values, the mixing model also reproduces closely the relatively narrow range of mineral 18O compositions observed in the deposit (Fig. 15). It is, however, noteworthy that this is only the case when the diluting fluid is assumed to be relatively hot (230C), as mixing of the magmatic brine with cold (25100C), dilute water would drive the mineral 18O compositions toward very high values. The clear lack of a trend of increasing mineral 18O values at San Rafael (apart from that observed in the lower-temperature stage IV, which is consistent with the mixing model), therefore, precludes a significant role for boiling in cassiterite deposition. By contrast, the mixing of a magmatic brine with heated, dilute groundwaters of meteoric origin reproduces remarkably well the observed 18O composition of the ore and gangue minerals. In our model, fluid mixing was simulated as a continuous process to capture the first-order effects, whereas the complex banding of the veins with multiple generations of the main ore and gangue minerals suggests that fluid mixing occurred repeatedly in a hydrologically very dynamic system.
TABLE 6a. Calculated Oxygen Isotope Composition of Quartz, Cassiterite, Chlorite and Wolframite, in Equilibrium with a Magmatic Brine (T = 500C, 18O = 11.0), which is Mixing with Meteoric Water (T = 230C, 18O = 2.0) Tmix (C) F = mass Fraction brine 1.00 0.95 0.90 0.85 0.80 0.75 0.70 0.65 0.60 0.55 0.50 0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 500 487 473 460 446 433 419 406 392 379 365 352 338 325 311 298 284 271 257 244 230 11.0 10.4 9.9 9.4 8.9 8.4 7.9 7.4 6.9 6.5 6.0 5.6 5.1 4.7 4.3 3.9 3.5 3.1 2.7 2.4 2.0 18O (fluid) Quartz 13.3 13.0 12.6 12.3 12.0 11.8 11.6 11.4 11.2 11.0 10.9 10.9 10.8 10.8 10.8 10.9 11.0 11.2 11.4 11.6 12.0 18O (mineral) Cassiterite 6.7 6.2 5.7 5.2 4.7 4.3 3.9 3.5 3.1 2.8 2.6 2.3 2.1 2.0 1.9 1.9 1.9 2.0 2.1 2.4 2.7 Chlorite 10.1 9.5 9.0 8.5 8.0 7.5 7.0 6.6 6.2 5.8 5.4 5.1 4.8 4.5 4.3 4.1 4.0 3.9 3.9 3.9 4.0 Wolframite 7.8 7.3 6.8 6.3 5.8 5.4 5.0 4.6 4.2 3.8 3.5 3.2 2.9 2.6 2.4 2.1 1.9 1.8 1.6 1.5 1.5
0361-0128/98/000/000-00 $6.00
243
244
WAGNER ET AL. TABLE 6b. Calculated Oxygen Isotope Composition of Quartz, Cassiterite, Chlorite and Wolframite, in Equilibrium with a Magmatic Brine (T = 500C, 18O = 11.0), Which is Progressively Cooling 18O (fluid) T (C) 500 480 460 440 420 400 380 360 340 320 300 280 260 240 230 11.0 11.0 11.0 11.0 11.0 11.0 11.0 11.0 11.0 11.0 11.0 11.0 11.0 11.0 11.0 Quartz 13.3 13.6 13.9 14.3 14.7 15.1 15.6 16.1 16.7 17.3 18.0 18.7 19.6 20.5 21.0 18O (mineral) Cassiterite 6.7 6.8 6.8 6.9 7.0 7.1 7.4 7.6 8.0 8.4 8.9 9.5 10.3 11.2 11.8 Chlorite 10.1 10.1 10.1 10.1 10.2 10.2 10.3 10.5 10.7 10.9 11.2 11.6 12.0 12.7 13.0 Wolframite 7.8 7.8 7.9 8.0 8.1 8.2 8.4 8.5 8.7 8.9 9.2 9.5 9.8 10.3 10.5
Concluding Remarks The hydrothermal history of the San Rafael vein system began with hot, acidic, saline and reducing fluids, which had an isotopic composition very close to that of typical magmatic fluids. These fluids likely exsolved from a late granitic melt related to the San Rafael igneous center and produced extensive sericitic and tourmaline alteration of the wall rocks (stage
22 20 18 16 14 12 10 8 6 4 2 0
a
Boiling (cooling)
Stage IV quartz Stage II-III quartz +2.0
0.0 -2.0
Mixing
Low-T fluid: 230 C 0 wt.% NaCl eq.
O
High-T fluid: 500 C 45 wt.% NaCl eq. 18O=+11

O
250
300
350
400
450
500
TOre fluid (C)

18 16 14 12 10 8 6 4 2 0 -2 -4
I). The modeling of tin solubility suggests that these fluids could have transported high concentrations of tin, but they did not deposit any cassiterite. The onset of cassiterite deposition came with a major change in the plumbing of the hydrothermal system, inferred to correspond to a change from lithostatic to hydrostatic pressure conditions, upon a major reopening of the San Rafael lode. The change in structural style, indicated by the ubiquitous open fracture-filling character of stage II-IV veins, was associated with a drop in temperature and major changes in fluid chemistry. This is evident from the strong chloritic alteration and the massive deposition of cassiterite and, subsequently, sulfide minerals. The ore fluids were considerably less saline and somewhat cooler than the early brines, were more oxidizing, and had lower 18O and D values. The precipitation of cassiterite was most likely caused by a marked increase in the oxidation state of the ore fluid (supported by evidence of rare hematite), an increase in pH, and to a lesser extent, decreases in temperature and chloride activity. The two depositional mechanisms that could explain the geologic evidence are boiling of the magmatic brine or its mixing with cooler, oxidizing meteoric waters. Quantitative modeling of mineral-water oxygen fractionation predicts the observed decrease of fluid 18O values with time for the case of fluid mixing, whereas this modeling suggests that boiling will produce the opposite trend in 18O values. Also, the general
()
18O Quartz
()
Boiling (cooling)
Stage II-III cassiterite +2.0 0.0 -2.0
Mixing
250
300
350
400
450
500
TOre fluid (C)

0361-0128/98/000/000-00 $6.00
FIG. 15. Diagrams showing the effect of boiling (cooling) and mixing of magmatic and meteoric water derived fluids on the oxygen isotope composition of quartz (a) and cassiterite (b). Because no experimental data are available for isotopic liquid-vapor fractionation in the H2O-NaCl system at conditions above the critical point of water, the boiling model has only considered the cooling effect (resulting from enthalpy transfer from the liquid to the vapor). Therefore, the boiling curves are labeled as boiling (cooling). The starting conditions for the magmatic brine were a temperature of 500C, a salinity of 45 wt percent NaCl equiv, and a 18O value of 11 per mil. Mixing lines were calculated assuming that groundwater had a temperature of 230C, zero salinity and 18O values between 2.0 and 2.0 per mil (labels on the mixing lines).
18O Cassiterite
244
245
evolution of the hydrothermal system toward domination by strongly dilute, cooler fluids is consistent with an incursion of hot groundwaters of meteoric origin. As the bulk of the tin ore at San Rafael is restricted to large fault-jogs at depth in the lode, we conclude that these jogs provided sites favorable for repeated episodic mixing between magmatic and meteoric fluids, which resulted in unusually efficient, structurally focused cassiterite deposition. Acknowledgments The authors wish to thank Ing. Fausto Zavaleta Cruzado, General Manager of MINSUR S.A., Ing. Luis Alva Florian and Ing. Otto Velarde Junes, successive managers of the San Rafael mine, Ing. Julver Alvarez Romero, Ing. Pastor Luque Malag, Ing. Ladislao Guilln Cardenas, and Ing. Luis Santalla Medrano, mine geologists, as well as other staff of the San Rafael mine, for logistic support and helpful collaboration during the field work. The assistance of Harold E. Waller and Nestor Roldan from EMINASA and Mario Arenas Figueroa, consulting geologist, were much appreciated. The authors would also like to thank the reviewers, B. Lehmann and A. Campbell, as well as associate editor J. Muntean and Editor L. Meinert for their constructive comments, which helped to improve this paper. The research was funded by NSERC and FQRNT grants to AEW-J and NERC support of the Isotope Community Support Facility at SUERC. November 6, 2008; February 18, 2009
REFERENCES
Aja, S.U., 2002, The stability of Fe-Mg chlorites in hydrothermal solutions. II. Thermodynamic properties: Clays and Clay Minerals, v. 50, p. 591600. Aja, S.U., and Dyar, D.M., 2002, The stability of Fe-Mg chlorites in hydrothermal solutions. I. Results of experimental investigations: Applied Geochemistry, v. 17, p. 12191239. Alderton, D.H.M., 1989, Oxygen isotope fractionation between cassiterite and water: Mineralogical Magazine, v. 53, p. 373376. Alderton, D.H.M., and Harmon, R.S., 1991, Fluid inclusion and stable isotope evidence for the origin of mineralizing fluids in southwest England: Mineralogical Magazine, v. 55, p. 605611. Anders, M.H., Gregory-Wodzicki, K.M., and Spiegelman, M., 2002, A critical evaluation of late Tertiary accelerated uplift rates for the Eastern Cordillera, Central Andes of Bolivia: Journal of Geology, v. 110, p. 89100. Arenas, M.J., 1980, Mapa geologico superficial del Distrito Minero San Rafael, Puno: Unpublished Report MINSUR Archives. Bakker, R.J., 2003, Package FLUIDS. 1. Computer programs for analysis of fluid inclusion data and for modelling bulk fluid properties: Chemical Geology, v. 194, p. 323. Barton, P.B., 1969, Thermochemical study of the system Fe-As-S: Geochimica et Cosmochimica Acta, v. 33, p. 841857. Bodnar, R.J., 1993, Revised equation and table for determining the freezing point depression of H2O-NaCl solutions: Geochimica et Cosmochimica Acta, v. 57, p. 683684. Bortnikov, N.S., Zaozerina, O.N., Genkin, A.D., and Muravitskaya, G.N., 1990, Stannite-sphalerite intergrowthspossible indicators of conditions of ore deposition: International Geology Review, v. 32, p. 11321144. Cathelineau, M., 1988, Cation site occupancy in chlorites and illites as a function of temperature: Clay Minerals, v. 23, p. 471485. Carothers, W.W., Adami, L.H., and Rosenbauer, R.J., 1988, Experimental oxygen isotope fractionation between siderite-water and phosphoric acid liberated CO2-siderite: Geochimica et Cosmochimica Acta, v. 52, p. 24452450. Clark, A.H., Palma, V.V., Archibald, D.A., Farrar, E., Arenas, M.J., and Robertson, R.C.R., 1983, Occurrence and age of tin mineralization in the Cordillera Oriental, Southern Peru: ECONOMIC GEOLOGY, v. 78, p. 514520. Clark, A.H., Farrar, E., Kontak, D.J., Langridge, R.J., Arenas Figueroa, M.J., France, L.J., McBride, S.L., Woodman, P.L., Wasteneys, H.A., Sandeman, H.A., and Archibald, D.A., 1990, Geologic and geochronologic constraints 0361-0128/98/000/000-00 $6.00
on the metallogenic evolution of the Andes of southeastern Peru: ECONOMIC GEOLOGY, v. 85, p. 15201583. Clayton, R.N., and Mayeda, T.K., 1963, The use of bromine pentafluoride in the extraction of oxygen from oxides and silicates for isotopic analysis: Geochimica et Cosmochimica Acta, v. 27, p. 4352. Cole, D.R., and Ripley, E.M., 1999, Oxygen isotope fractionation between chlorite and water from 170 to 350C: A preliminary assessment based on partial exchange and fluid/rock experiments: Geochimica et Cosmochimica Acta, v. 63, p. 449457. Collins, P.L.F., 1981, The geology and genesis of the Cleveland tin deposit, Western Tasmania: Fluid inclusion and stable isotope studies: ECONOMIC GEOLOGY, v. 76, p. 365-392. Dolejs, D., and Wagner, T., 2008, Thermodynamic modeling of non-ideal mineral-fluid equilibria in the system Si-Al-Fe-Mg-Ca-Na-K-H-O-Cl at elevated temperatures and pressures: Implications for hydrothermal mass transfer in granitic rocks. Geochimica et Cosmochimica Acta, v. 72, p. 526554. Dolejs, D., Mlynarczyk, M.S.J., and Williams-Jones, A.E., 2009, S-type peraluminous granitic rocks associated with a world-class tin deposit: a petrogenetic study of the San Rafael stock, southeastern Peru. J. Petrol. (submitted) Eadington, P.J., 1985, The solubility of cassiterite in hydrothermal solutions in relation to some lithological and mineral associations of tin ores, in Taylor, R.P., and Strong, D.F., eds., Recent advances in the geology of graniterelated mineral deposits: Canadian Institution of Mining and Metallurgy, Special Volume 39, p. 2532. Einaudi, M.T., Hedenquist, J.W., and Inan, E.E., 2003, Sulfidation state of fluids in active and extinct hydrothermal systems: Transitions from porphyry to epithermal environments: SEG Special Publications, v. 10, p. 285313. Eugster, H.P., and Wilson, G.A., 1985, Transport and deposition of ore-forming elements in hydrothermal systems associated with granites, in Halls, C., ed., High heat production (HHP) granites, hydrothermal circulation and ore genesis: Institution of Mining and Metallurgy Conference, London, p. 8798. Goldstein, R.H., and Reynolds, T.J., 1994, Systematics of fluid inclusions in diagenetic minerals: Society of Sedimentary Geology Short Course, v. 31, 199 p. Graham, C.M., Atkinson, J., and Harmon, R.S., 1984, Hydrogen isotope fractionation in the system chlorite-water [abs.]: Natural Environment Research Council (NERC) Publication Series D, no. 25, p. 139. Graham, C.M., Viglino, J.A., and Harmon, R.S., 1987, Experimental study of hydrogen isotope exchange between aluminous chlorite and water and of hydrogen diffusion in chlorite: American Mineralogist, v. 72, p. 566579. Gregory-Wodzicki, K.M., 2000, Andean paleoelevation estimates: A review and critique: Geological Society of America Bulletin, v. 112, p. 10911105. Grnvold, F., and Stlen, S., 1992, Thermodynamics of iron sulfides. II. Heat capacity and thermodynamic properties of FeS and of Fe0.875S at temperatures from 298.15 K to 1000 K, of Fe0.98S from 298.15 K to 800 K, and of Fe0.89S from 298.15 K to about 650 K. Thermodynamics of formation: Journal of Chemical Thermodynamics, v. 24, p. 913936. Harrison, T.M., Duncan, I., McDougall, I., 1985, Diffusion of 40Ar in biotite: Temperature, pressure and compositional effects: Geochimica et Cosmochimica Acta, v. 49, p. 24612468. Haynes, F.M., 1985, Determination of fluid inclusion compositions by sequential freezing: ECONOMIC GEOLOGY, v. 80, p. 14361439. Hattori, K.H., and Keith, J.D., 2001, Contribution of mafic melt to porphyry copper mineralization: Evidence from Mount Pinatubo, Philippines, and Bingham Canyon, Utah, USA: Mineralium Deposita, v. 36, p. 799806. Heinrich, C.A., 1990, The chemistry of hydrothermal tin (-tungsten) ore deposition: ECONOMIC GEOLOGY, v. 85, p. 457481. 1995, Geochemical evolution and hydrothermal mineral deposition in Sn (-W-base metal) and other granite-related ore systems: some conclusions from Australian examples, in Thompson, J.F.H., ed., Magmas, fluids and ore deposits: Mineralogical Association of Canada (MAC) Short Course, v. 23, p. 203220. Heinrich, C.A., and Eadington, P.J., 1986, Thermodynamic predictions of the hydrothermal chemistry of arsenic, and their significance for the paragenetic sequence of some cassiterite-arsenopyrite-base metal sulfide deposits: ECONOMIC GEOLOGY, v. 81, p. 511529. Holland, T.J.B., and Powell, R., 1998, An internally consistent thermodynamic data set for phases of petrological interest: Journal of Metamorphic Geology, v. 16, p. 309343.
245
246
WAGNER ET AL. Longstaffe, F.J., Clark, A.H., McNutt, R.H., and Zentilli, M., 1983, Oxygen isotopic compositions of Central Andean plutonic and volcanic rocks, latitudes 26-29 south: Earth and Planetary Science Letters, v. 64, p. 918. Matsuhisa, Y., Goldsmith, J.R., and Clayton, R.N., 1979, Oxygen isotopic fractionation in the system quartz-albite-anorthite-water: Geochimica et Cosmochimica Acta, v. 43, p. 11311140. Mattey, D.P., and Macpherson, C.G., 1993, High-precision oxygen isotope microanalysis of ferromagnesian minerals by laser fluorination: Chemical Geology, v. 105, p. 305318. McBride, S.L., Robertson, R.C.R., Clark, A.H., Farrar, E., 1983, Magmatic and metallogenic episodes in the northern tin belt, Cordillera Real, Bolivia: International Journal of Earth Sciences, v. 72, p. 685713. McCrea, J.M., 1950, On the isotopic chemistry of carbonates and a palaeotemperature scale: Journal of Chemical Physics, v. 18, p. 849857. Menzie, W.D., Reed, B.L., and Singer, D.A., 1988, Models of grades and tonnages of some lode tin deposits, in Hutchison, C.S., ed., Geology of tin deposits in Asia and the Pacific; mineral concentrations and hydrocarbon accumulations in the ESCAP region: Selected papers from the International Symposium on the Geology of Tin Deposits, Nanning, China, October 1984, p. 7388. Migdisov, A.A., Williams-Jones, A.E., Lakshtanov, L.Z., and Alekhin, Y.V., 2002, Estimates of the second dissociation constant of H2S from the surface sulfidation of crystalline sulfur: Geochimica et Cosmochimica Acta, v. 66, p. 17131725. Mlynarczyk, M.S.J., and Williams-Jones, A.E., 2005, The role of collisional tectonics in the metallogeny of the Central Andean tin belt: Earth and Planetary Science Letters, v. 240, p. 656667. 2006, Zoned tourmaline associated with cassiterite: implications for fluid evolution and tin mineralization in the San Rafael Sn-Cu deposit, SE Peru: Canadian Mineralogist, v. 44, p. 347365. Mlynarczyk, M.S.J., Sherlock, R.L., and Williams-Jones, A.E., 2003, San Rafael, Peru: geology and structure of the worlds richest tin lode: Mineralium Deposita, v. 38, p. 555567. Mlynarczyk, M.S.J., Wagner, T., Williams-Jones, A.E., and Dolejs, D., 2009, Geology and geochemistry of alteration at the San Rafael Sn-Cu deposit, SE Peru: ECONOMIC GEOLOGY (in review) Mller, B., and Seward, T.M., 2001, Spectrophotometric determination of the stability of tin (II) chloride complexes in aqueous solution up to 300C: Geochimica et Cosmochimica Acta, v. 65, p. 41874199. Nakamura, Y., and Shima, H., 1982, Fe and Zn partitioning between sphalerite and stannite [abs.]: Joint Meeting of the Society of Mining Geology of Japan, Association of Mineralogy, Petrology and Economic Geology, and Mineralogical Society of Japan, p. A8. Nekrasov, I.Y., Sorokin, V.I., and Osadchii, E.G., 1976, Partition of iron and zinc between sphalerite and stannite at T = 300 to 500C and P = 1kb: Doklady Earth Sciences, v. 226, p. 136138. 1979, Fe and Zn partitioning between stannite and sphalerite and its application in geothermometry: Physics and Chemistry of the Earth, v. 11, p. 739742. Oelkers, E.H., and Helgeson, H.C., 1990, Triple-ion anions and polynuclear compexing in supercritical electrolyte solutions: Geochimica et Cosmochimica Acta, v. 54, p. 727738. Ohmoto, H., 1972, Systematics of sulfur and carbon isotopes in hydrothermal ore deposits: ECONOMIC GEOLOGY, v. 67, p. 551578. Ohmoto, H., and Goldhaber, M.B., 1997, Sulfur and carbon isotopes, in Barnes, H.L., ed., Geochemistry of hydrothermal ore deposits, 3rd ed: New York, Wiley, p. 517611. Pabalan, R.T., 1986, Solubility of cassiterite (SnO2) in NaCl solutions from 200C350C, with geologic applications: Unpublished Ph.D. thesis, Pennsylvania State University, 141 p. Palma, V.V., 1981, The San Rafael tin-copper deposit, Puno, SE Peru: Unpublished M.Sc. thesis, Kingston, Canada, Queens University, 235 p. Pashinkin, A.C., Muratova, V.A., Moiseyev, N.V., and Bazhenov, J.V., 1991, Heat capacity and thermodynamic functions of iron diarsenide in the T range 5 K to 300 K: Journal of Chemical Thermodynamics, v. 23, p. 827830. Primmer, T.J., 1985, Discussion on the possible contribution of metamorphic water to the mineralising fluid of south-west England: Preliminary stable isotope evidence: Proceedings of the Ussher Society, v. 6, p. 224228. Reed, M.H., and Spycher, N.F., 1984, Calculation of pH and mineral equilibria in hydrothermal waters with application to geothermometry and studies of boiling and dilution: Geochimica et Cosmochimica Acta, v. 48, v. 14791492.
Holland, T., Baker, J., and Powell, R., 1998, Mixing properties and activitycomposition relationships of chlorites in the system MgO-FeO-Al2O3-SiO2H2O: European Journal of Mineralogy, v. 10, p. 395406. Horita, J., and Wesolowski, D.J., 1994, Liquid-vapor fractionation of oxygen and hydrogen isotopes of water from the freezing point to the critical temperature: Geochimica et Cosmochimica Acta, v. 58, p. 34253447. Horita, J., Wesolowski, D.J., and Cole, D.R., 1993, The activity-composition relationship of oxygen and hydrogen isotopes in aqueous salt solutions. I. Vapor-liquid water equilibration of single salt solutions from 50 to 100C: Geochimica et Cosmochimica Acta, v. 57, p. 27972817. Horita, J., Cole, D.R., and Wesolowski, D.J., 1995, The activity-composition relationship of oxygen and hydrogen isotopes in aqueous salt solutions. III. Vapor-liquid equilibration of NaCl solutions to 350 C: Geochimica et Cosmochimica Acta, v. 59, p. 11391151. Jackson, P., Changkakoti, A., Krouse, H.R., and Gray, J., 2000, The origin of greisen fluids of the Foleys zone, Cleveland tin deposit, Tasmania, Australia: ECONOMIC GEOLOGY, v. 95, p. 227236. Johnson, J.W., Oelkers, E.H., and Helgeson, H.C., 1992, SUPCRT92: A software package for calculating the standard molal thermodynamic properties of minerals, gases, aqueous species, and reactions from 1 to 5000 bar and 0 to 1000 C: Computers and Geosciences, v. 18, p. 899947. Jowett, E.C., 1991, Fitting iron and magnesium into the hydrothermal chlorite geothermometer: GAC-MAC-SEG Joint Meeting, Program with Abstract, p. A62. Kelley, S.P., and Fallick, A.E., 1990, High precision spatially resolved analysis of 34S in sulphides using as laser extraction technique: Geochimica et Cosmochimica Acta, v. 54, p. 883888. Kelley, S.P., Fallick, A.E., McConville, P., and Boyce, A.J., 1992, High precision, high spatial resolution analysis of sulfur isotopes by laser combustion of natural sulfide minerals: Scanning Microscopy, v. 6, p. 129138. Kelly, W.C., and Turneaure, F.S., 1970, Mineralogy, paragenesis and geothermometry of tin and tungsten deposits of the eastern Andes, Bolivia: ECONOMIC GEOLOGY, v. 65, p. 609680. Kojima, S., and Sugaki, A., 1984, Phase relations in the Cu-Fe-Zn-S system between 500 and 300C under hydrothermal conditions: ECONOMIC GEOLOGY, v. 80, p. 158171. Kontak, D.J., 1990, A sulfur isotope study of main-stage tin and base metal mineralization at East Kemptville tin deposit, Yarmouth County, Nova Scotia, Canada: evidence for magmatic origin of metals and sulfur: Economic Geology, v. 85, p. 399-407. Kontak, D.J., and Clark, A.H., 1988, Exploration criteria for tin and tungsten mineralization in the Cordillera Oriental of southeastern Peru, in Taylor, R.P., and Strong, D.F., eds., Recent advances in the geology of granite-related mineral deposits: Canadian Institution of Mining and Metallurgy, Special Volume 39, p. 157169. 2002, Genesis of the giant, bonanza San Rafael lode tin deposit, Peru: origin and significance of pervasive alteration: Economic Geology, v. 97, p. 1741-1777. Kontak, D.J., Clark, A.H., Farrar, E., Pearce, T.H., Strong, D.F., and Baadsgaard, H., 1986, Petrogenesis of a Neogene shoshonite suite, Cerro Moromoroni, Puno, southeastern Peru: Canadian Mineralogist, v. 24, p. 117135. Kontak, D.J., Clark, A.H., Farrare, E., Archibald, D.A., and Baadsgaard, H., 1987, Geochronological data for tertiary granites of the southeast Peru segment of the Central Andean tin belt: ECONOMIC GEOLOGY, v. 82, p. 16111618. Kontak, D.J., Cumming, G.L., Krstic, D., Clark, A.H., and Farrar, E., 1990, Isotopic composition of lead in ore deposits of the Cordillera Oriental, southeastern Peru: ECONOMIC GEOLOGY, v. 85, p. 15841603. Kotzer, T.G., Kyser, T.K., King, R.W., and Kerrich, R., 1993, An empirical oxygen- and hydrogen-isotope geothermometer for quartz-tourmaline and tourmaline-water: Geochimica et Cosmochimica Acta, v. 57, p. 34213426. Laubacher, G., 1978, Estudio geologico de la region norte del Lago Titicaca: Peru, Instituto Geologia y Mineria, v. 5, 120 p. LeBoutillier, N.G., Camm, G.S., Shail, R.K., Bromley, A.V., Jewson, C., and Hoppe, N., 2002, Tourmaline-quartz-cassiterite mineralization of the Lands End granite at Nanjizal, west Cornwall: Proceedings of the Ussher Society, v. 10, p. 312318. Lehmann, B., 1990, Metallogeny of tin: Lecture Notes in Earth Sciences, v. 32, Berlin,Springer-Verlag, 211 p. Linnen, R.L., and Williams-Jones, A.E., 1995, Genesis of a magmatic metamorphic hydrothermal system: The Sn-W polymetallic deposits at Pilok, Thailand: ECONOMIC GEOLOGY, v. 90, p. 11481166. 0361-0128/98/000/000-00 $6.00
246
STABLE ISOTOPE CONSTRAINTSSAN RAFAEL SN-CU DEPOSIT, SE PERU Robie, R.A., and Hemingway, B.S., 1995, Thermodynamic properties of minerals and related substances at 298.15 K and 1 Bar (105 Pascals) pressure and at higher temperatures: U.S. Geological Survey Monograph, v. 2131, 461 p. Robie, R.A., Seal, R.R. II, and Hemingway, B.S., 1994, Heat capacity and entropy of bornite (Cu5FeS4) between 6 and 760 K and the thermodynamic properties of phases in the system Cu-Fe-S: Canadian Mineralogist, v. 32, p. 945956. Roedder, E., 1984, Fluid inclusions: Reviews in Mineralogy, v. 12, 646 p. Rosenbaum, J., and Sheppard, S.M.F., 1986, An isotopic study of siderites, dolomites and ankerites at high temperatures: Geochimica et Cosmochimica Acta, v. 50, p. 11471150. Ryzhenko, B.N., Shvarov, Y.V., and Kovalenko, N.I., 1997, The Sn-Cl-F-C-SH-O-Na system: Thermodynamic properties of components within the conditions of the Earths crust: Geochemistry International, v. 35, p. 10161020. Sandeman, H.A., Clark, A.H., and Farrar, E., 1995, An integrated tectonomagmatic model for the evolution of the Southern Peruvian Andes (1320S) since 55 Ma: International Geology Review, v. 37, p. 10391073. Sandeman, H.A., Clark, A.H., Farrar, E., and Arroyo-Pauca, G., 1996, A critical appraisal of the Cayconi Formation, Crucero Basin, southeastern Peru: Journal of South American Earth Sciences, v. 9, p. 381392. Schwinn, G., Wagner, T., Markl, G., and Baatartsogt, B., 2006, Quantification of mixing processes in ore-forming hydrothermal systems by combination of stable isotope and fluid inclusion analyses: Geochimica et Cosmochimica Acta, v. 70, p. 965982. Sharp, Z.D., 1990, A laser-based microanalytical method for the in situ determination of oxygen isotope ratios of silicates and oxides: Geochimica et Cosmochimica Acta, v. 54, p. 13531357. Sharp, Z.D., Atudorei, V., and Durakiewicz, T., 2001, A rapid method for determination of hydrogen and oxygen isotope ratios from water and hydrous minerals: Chemical Geology, v. 178, p. 197210. Sheppard, S.M.F., 1994, Stable isotope and fluid inclusion evidence for the origin and evolution of hercynian mineralizing fluids, in Seltmann, R., Kmpf, H., and Mller, P., eds., Metallogeny of collisional orogens: Czech Geological Survey, Prague, p. 4960. Shimitzu, M., and Shikazono, N., 1985, Iron and zinc partitioning between coexisting stannite and sphalerite: A possible indicator of temperature and sulfur fugacity: Mineralium Deposita, v. 20, p. 314320. Shock, E.L., Oelkers, E.H., Johnson, J.W., Sverjensky, D.A., and Helgeson, H.C., 1992, Calculation of the thermodynamic properties of aqueous species at high pressures and temperatures: Journal of the Chemical Society Faraday Transactions, v. 88, p. 803826. Shock, E.L., Sassani, D.C., Willis, M., and Sverjensky, D.A., 1997, Inorganic species in geological fluids: Correlations among standard molal thermodynamic properties of aqueous ions and hydroxide complexes: Geochimica et Cosmochimica Acta, v. 61, p. 907950. Shvarov, Y.V., 1978, Minimization of the thermodynamic potential of an open chemical system: Geochemistry International, v. 15, p. 200203. 1981, A general equilibrium criterion for an isobaric-isothermal model of a chemical system: Geochemistry International, v. 18, p. 3845. Shvarov, Y.V., and Bastrakov, E., 1999, HCh: A software package for geochemical equilibrium modeling. Users guide: Australian Geological Survey Organization, 61 p. Smith, M., Banks, D.A., Yardley, B.W.D., and Boyce, A., 1996, Fluid inclusion and stable isotope constraints on the genesis of the Cligga Head Sn-W deposit, S.W. England: European Journal of Mineralogy, v. 8, p. 961974. Spycher, N., and Reed, M.H., 1989, Evolution of a Broadlands-type epithermal ore fluid along alternative P-T paths: Implications for the transport and deposition of base, precious, and volatile metals: ECONOMIC GEOLOGY, v. 84, p. 328359. Sterner, S.M., Hall, D.L., and Bodnar, R.J., 1988, Synthetic fluid inclusions. V. Solubility relations in the system NaCl-KCl-H2O under vapor-saturated conditions: Geochimica et Cosmochimica Acta, v. 52, p. 9891005. Sugaki, A., Kitakaze, A., and Kojima, S., 1990, Sphalerite stars in chalcopyrite; are they always the results of an unmixing process; discussion: Mineralium Deposita, v. 25, p. 8283.
247
Sun, S., and Eadington, P.J., 1987, Oxygen isotope evidence for the mixing of magmatic and meteoric waters during tin mineralization in the Mole granite, New South Wales, Australia: ECONOMIC GEOLOGY, v. 82, p. 4352. Sverjensky, D.A., Shock, E.L., and Helgeson, H.C., 1997, Prediction of the thermodynamic properties of aqueous metal complexes to 1000C and 5 kb: Geochimica et Cosmochimica Acta, v. 61, p. 13591412. Taylor, H.P., 1977, Water/rock interactions and the origin of H2O in granitic batholiths: Journal of the Geological Society London, v. 133, p. 509558. 1997, Oxygen and hydrogen isotope relationships in hydrothermal mineral deposits, in Barnes, H.L., ed., Geochemistry of hydrothermal ore deposits, 3rd ed.: New York, Wiley, p. 229302. Taylor, J.R., and Wall, V.J., 1993, Cassiterite solubility, tin speciation and transport in a magmatic aqueous phase: ECONOMIC GEOLOGY, v. 88, p. 437460. Taylor, R.G., 1979, Geology of tin deposits: Developments in Economic Geology, v. 11, 543 p. Truesdell, A.H., and Nathenson, M., 1977, The effects of boiling and dilution on the isotopic compositions of Yellowstone thermal waters: Journal of Geophysical Research, v. 82, p. 36943704. Turneaure, F.S., 1960a, A comprarative study of major ore deposits in Bolivia: ECONOMIC GEOLOGY, v. 55, p. 217254. 1960b, A comparative study of major ore deposits in Bolivia. Part II: ECONOMIC GEOLOGY, v. 55, p. 574606. Vidal, O., Parra, T., and Trottet, F., 2001, A thermodynamic model for Fe-Mg aluminous chlorite using data from phase equilibrium experiments and natural pelitic assemblages in the 100-600C, 1-25 kb range: American Journal of Science, v. 301, p. 557592. Wagner, T., Boyce, A.J., and Fallick, A.E., 2002, Laser combustion analysis of 34S of sulfosalt minerals: Determination of the fractionation systematics and some crystal-chemical considerations: Geochimica et Cosmochimica Acta, v. 66, p. 28552863. Wagner, T., Boyce, A.J., Jonsson, E., and Fallick, A.E., 2004, Laser microprobe sulphur isotope analysis of arsenopyrite: experimental calibration and application to the Boliden Au-Cu-As massive sulphide deposit: Ore Geology Reviews, v. 25, p. 311325. Wagner, T., Williams-Jones, A.E., and Boyce, A.J., 2005, Stable isotope-based modeling of the origin and genesis of an unusual Au-Ag-Sn-W epithermal system at Cirotan, Indonesia: Chemical Geology, v. 219, p. 237260. Walshe, J.L., 1986, A six-component chlorite solid-solution model and the conditions of chlorite formation in hydrothermal and geothermal systems: ECONOMIC GEOLOGY, v. 81, p. 681703. Walshe, J.L., Halley, S.W., Anderson, J.A., and Harrold, B.P., 1996, The interplay of groundwater and magmatic fluids in the formation of the cassiterite-sulfide deposits of western Tasmania: Ore Geology Reviews, v. 10, p. 367387. Wiggins, L.B., and Craig, J.R., 1980, Reconnaissance of the Cu-Fe-Zn-S system: sphalerite phase relationship: ECONOMIC GEOLOGY, v. 75, p. 742751. Williamson, B.J., Spratt, J., Adams, J.T., Tindle, A.G., and Stanley, C.J., 2000, Geochemical constraints from zoned hydrothermal tourmalines on fluid evolution and tin mineralization: an example from fault breccias at Roche, SW England: Journal of Petrology, v. 41, p. 14391453. Wilkinson, J.J., Jenkin, G.R.T., Fallick, A.E., and Foster, R.P., 1995, Oxygen and hydrogen isotopic evolution of Variscan crustal fluids, south Cornwall, U.K: Chemical Geology, v. 123, p. 239254. Zhang, L., Liu, J., Chen, Z., and Zhou, H., 1994, Experimental investigations of oxygen isotope fractionation in cassiterite and wolframite: ECONOMIC GEOLOGY, v. 89, p. 150157. Zhang, X., and Spry, P.G., 1994, FO2PH: a quickBASIC program to calculate mineral stabilities and sulphur isotope contours in log fO2-pH space: Mineralogy and Petrology, v. 50, p. 287291. Zheng, Y.F., 1993, Calculation of oxygen isotope fractionation in hydroxylbearing silicates: Earth and Planetary Science Letters, v. 120, p. 247263.
0361-0128/98/000/000-00 $6.00
247
248
WAGNER ET AL.
APPENDIX Thermodynamic Calculations We modeled the depositional conditions prevalent during the main cassiterite stage and the late sulfide stage at San Rafael by constructing phase diagrams showing aqueous and mineral equilibria in the Cu-Fe-Sn-As-S-O-H system and contours of sulfur isotope fractionation. In addition, we carried out a series of speciation calculations in the model system Sn-Na-Cl-O-H in order to obtain the total solubility of Sn as a function of pH and the oxidation state of the hydrothermal fluid. All the calculations were carried out with the HCh software package (Shvarov and Bastrakov, 1999), which models heterogeneous equilibria and reaction progress by minimization of the Gibbs free energy of the total system (Shvarov, 1978, 1981). Thermodynamic data for most aqueous species were taken from the SUPCRT92 database and subsequent updates (Johnson et al., 1992; Shock et al., 1997; Sverjensky et al., 1997). Data for a number of aqueous Sn species came from Ryzhenko et al. (1997). Thermodynamic data for rockforming silicate and oxide minerals were taken from the internally consistent dataset of Holland and Powell (1998). The data for cassiterite, pyrite, chalcopyrite, bornite, and pyrrhotite (not contained in this dataset) were compiled from Robie and Hemingway (1995). As heat capacity functions for pyrrhotite and bornite are not given in Robie and Hemingway (1995), the experimentally determined heat capacity data from the original sources (pyrrhotite: Grnvold and Stlen, 1992; bornite: Robie et al., 1994) were fitted with a four-term polynominal of the form used by Holland and Powell (1998). It should be noted that the entropy values for bornite and pyrrhotite given in Robie and Hemingway (1995) also originate from Grnvold and Stlen (1992) and Robie et al. (1994), thereby ensuring internal consistency of these datasets. Data for calculating the stability fields of arsenic phases (arsenopyrite, loellingite, native arsenic) were compiled from Robie and Hemingway (1995) and Barton (1969). The heat capacity function for loellingite was fitted to the original data given in Pashinkin et al. (1991), which were the source of the entropy data tabulated in Robie and Hemingway (1995). All calculations of individual activity coefficients of aqueous species applied an extended Debye-Hckel model using the b-gamma equation for NaCl as the background electrolyte (Oelkers and Helgeson, 1990; Shock et al., 1992). The set of equations of Zhang and Spry (1994) and the most recent set of isotopic fractionation factors given in Ohmoto and Goldhaber (1997) were used to calculate the sulfur isotope contours. The model of Zhang and Spry (1994), which excludes the aqueous S2 species, is preferred over the original formalism of Ohmoto (1972), because the second dissociation constant of H2S is too small for S2 to be a significant species at geologically realistic values of pH (Migdisov et al., 2002).
0361-0128/98/000/000-00 $6.00
248

223

Transféré par

Informations du document

Description originale:

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

223

Transféré par

Droits d'auteur :

Formats disponibles

2009 Society of Economic Geologists, Inc. Economic Geology, v. 104, pp.

author: e-mail: thomas.wagner@erdw.ethz.ch

Granitoids (Late Oligocene):

STABLE ISOTOPE CONSTRAINTSSAN RAFAEL SN-CU DEPOSIT, SE PERU

Copper stopes Tin orebodies

STABLE ISOTOPE CONSTRAINTSSAN RAFAEL SN-CU DEPOSIT, SE PERU

chl qz cp tou cas cas chl sp

2.7 9.7 qz 10.1 3.0 3.3 cas 2.2

cas4 4.6 qz6 qz5

qz3 qz2 cas2

chl 10.6 3.4 R 119

STABLE ISOTOPE CONSTRAINTSSAN RAFAEL SN-CU DEPOSIT, SE PERU

cas chl qz cas cp qz

sph po py cas stn chl cas apy

STABLE ISOTOPE CONSTRAINTSSAN RAFAEL SN-CU DEPOSIT, SE PERU

Homogenization temperature (C)

Total salinity (wt.% NaCl eq.)

Stage III quartz / fluorite

STABLE ISOTOPE CONSTRAINTSSAN RAFAEL SN-CU DEPOSIT, SE PERU

5.4 qz D3-27 R 404

3.8 4.2 R 119

4.0 R 511a J 410 R 428

STABLE ISOTOPE CONSTRAINTSSAN RAFAEL SN-CU DEPOSIT, SE PERU

18OV-SMOW () 11.5 11.9 12.3 10.8 12.0 11.7

-18 -20 -22 -24

-18 -20 -22 -24

(20wt.% NaCl eq.) aS = 0.01

(20wt.% NaCl eq.)

(5 wt.% NaCl eq.)

aK+= 0.005 aS = 0.01 34Stot=+3.0

(5 wt.% NaCl eq.)

-17.5 -15.0 -10.0 -5.0 0.0 +2.8

-36 -38 -40 0

STABLE ISOTOPE CONSTRAINTSSAN RAFAEL SN-CU DEPOSIT, SE PERU

Total Sn solubility (mol/kg)

-20 -40 -60 -80

Tertiary Andean meteoric water ?

Groundwater Magmatic water box

-100 -120 -20

Present-day Andean meteoric water

STABLE ISOTOPE CONSTRAINTSSAN RAFAEL SN-CU DEPOSIT, SE PERU

High-T fluid: 500 C 45 wt.% NaCl eq. 18O=+11

TOre fluid (C)

TOre fluid (C)

STABLE ISOTOPE CONSTRAINTSSAN RAFAEL SN-CU DEPOSIT, SE PERU

Vous aimerez peut-être aussi