Mineralium Deposita

https://doi.org/10.1007/s00126-017-0786-9

ARTICLE

Mineralized breccia clasts: a window into hidden porphyry-type

mineralization underlying the epithermal polymetallic deposit of Cerro
de Pasco (Peru)
Bertrand Rottier 1,2 & Kalin Kouzmanov 1 & Vincent Casanova 1 & Anne-Sophie Bouvier 3 & Lukas P. Baumgartner 3 &
Markus Wälle 4,5 & Lluís Fontboté 1

Received: 24 November 2016 / Accepted: 10 December 2017

# Springer-Verlag GmbH Germany, part of Springer Nature 2017

Abstract
Cerro de Pasco (Peru) is known for its large epithermal polymetallic (Zn-Pb-Ag-Cu-Bi) mineralization emplaced at shallow level,
a few hundred meters below the paleo-surface, at the border of a large diatreme–dome complex. Porphyry-style veins crosscut-
ting hornfels and magmatic rock clasts are found in the diatreme breccia and in quartz–monzonite porphyry dikes. Such
mineralized veins in clasts allow investigation of high-temperature porphyry-style mineralization developed in the deep portions
of magmatic–hydrothermal systems. Quartz in porphyry-style veins contains silicate melt inclusions as well as fluid and solid
mineral inclusions. Two types of high-temperature (> 600 °C) quartz–molybdenite–(chalcopyrite)–(pyrite) veins are found in the
clasts. Early, thin (1–2 mm), and sinuous HT1 veins are crosscut by slightly thicker (up to 2 cm) and more regular HT2 veins. The
HT1 vein quartz hosts CO2- and sulfur-rich high-density vapor inclusions. Two subtypes of the HT1 veins have been defined,
based on the nature of mineral inclusions hosted in quartz: (i) HT1bt veins with inclusions of K-feldspar, biotite, rutile, and minor
titanite and (ii) HT1px veins with inclusions of actinolite, augite, titanite, apatite, and minor rutile. Using an emplacement depth of
the veins of between 2 and 3 km (500 to 800 bar), derived from the diatreme breccia architecture and the supposed erosion
preceding the diatreme formation, multiple mineral thermobarometers are applied. The data indicate that HT1 veins were formed
at temperatures > 700 °C. HT2 veins host assemblages of polyphase brine inclusions, generally coexisting with low-density
vapor-rich inclusions, trapped at temperatures around 600 °C. Rhyolitic silicate melt inclusions found in both HT1 and HT2 veins
represent melt droplets transported by the ascending hydrothermal fluids. LA-ICP-MS analyses reveal a chemical evolution
coherent with the crystallization of an evolved rhyolitic melt. Quartz from both HT1 and HT2 veins also contains secondary, low-
temperature (~ 300 °C) brine and aqueous fluid inclusions that record the cooling of the system. Both vein types are locally
crosscut and/or reopened by a pre-diatreme polymetallic event consisting of pyrite, sphalerite with Bchalcopyrite disease,^
galena, chalcopyrite, tetrahedrite–tennantite, and minor quartz. LA-ICP-MS analyses of mineral and high-temperature fluid
inclusions hosted in HT1 and HT2 veins and in situ secondary-ion mass spectrometry oxygen isotope analyses of vein quartz
indicate a magmatic signature for the mineralizing fluids with no major meteoric water input and allow reconstruction of the
source and chemical evolution of fluids that formed these porphyry-style veins as snapshots of the early and deep mineralizations
at Cerro de Pasco. This detailed study of the porphyry-type mineralization hosted in clasts offers a unique opportunity to

Editorial handling: F. Melcher

Electronic supplementary material The online version of this article
(https://doi.org/10.1007/s00126-017-0786-9) contains supplementary
material, which is available to authorized users.

* Bertrand Rottier 3
Institute of Mineralogy and Geochemistry, University of Lausanne,
[email protected] 1015 Lausanne, Switzerland
4
Institute of Geochemistry and Petrology, ETH Zürich,
1
Department of Earth Sciences, University of Geneva, 8092 Zürich, Switzerland
1205 Geneva, Switzerland 5
Present address: CREAIT, CRC and CFI Services (CCCS), Bruneau
2
Present address: Bayerisches Geoinstitut, University of Bayreuth, Centre for Research and Innovation, Memorial University of
95440 Bayreuth, Germany Newfoundland, St. John’s, NL A1C 5S7, Canada
 Miner Deposita

reconstruct the late magmatic and early hydrothermal evolutions of porphyry mineralization underlying the world-class Cerro de
Pasco epithermal polymetallic (Zn-Pb-Ag-Cu-Bi) deposit.

Keywords Porphyry . Breccia clasts . Fluid inclusions . Silicate melt inclusions . Hydrothermal quartz . La-ICP-MS . Trace
elements . Oxygen isotopes . SIMS . Cerro de Pasco

Introduction mineralization in a magmatic–hydrothermal system with no

evidence of telescoping. This study is based on (i) detailed
Mineralized porphyry systems are the main global suppliers for petrography combined with scanning electron microscopy–
Cu and Mo and are important for Au, Re, Ag, Pd, Zn, Pb, and cathodoluminescence (SEM-CL) imaging of hydrothermal
many other economic elements (Sillitoe 2010). Porphyry sys- quartz veins, (ii) application of three independent mineral
tems consist of large volumes of hydrothermally altered, veined, thermobarometers (Ti in quartz, Huang and Audétat 2012; Zr
and mineralized rocks centered on porphyry stocks. A large va- in rutile, Tomkins et al. 2007; Zr in titanite, Hayden et al. 2008),
riety of mineralization styles including skarn, carbonate replace- (iii) fluid-inclusion microthermometry and LA-ICP-MS analy-
ment, sediment-hosted, and high- and intermediate-sulfidation sis, and (iv) in situ secondary-ion mass spectrometry (SIMS)
epithermal base and precious metal mineralization can be found oxygen isotope analysis of the hydrothermal vein quartz. The
in such systems (Sillitoe 2010). The comprehension of their tem- porphyry-style veins were formed at high temperatures (>
poral, spatial, and genetic relationships is a key for understanding 600 °C) and host CO2-rich high-density vapor inclusions and
the dynamics and complexity of such magmatic–hydrothermal polyphase brine inclusions. A prominent feature of these veins
systems. Close spatial and temporal relationship between is the presence of crystallized or glassy rhyolitic silicate melt
epithermal polymetallic mineralization and porphyry-style min- inclusions (SMIs) in the hydrothermal quartz. Occurrence of
eralization has been reported in several telescoped systems where SMIs in porphyry veins is rarely reported in the literature, and
polymetallic mineralization spatially overprints higher- their significance is still not fully understood (Pintea 1995,
temperature porphyry-style mineralization, such as Main stage 2014; Harris et al. 2003; Ivascanu et al. 2003; Stefanova et al.
veins at Butte, USA (Meyer et al. 1968; Lund et al. 2002; Rusk 2014; Rottier et al. 2016a). The chemical composition of the
et al. 2008a, b; Reed et al. 2013), El Rosario veins at Collahuasi, SMIs has been determined by LA-ICP-MS. We apply a multi-
Chile (Masterman et al. 2005), and the polymetallic veins and analytical approach to reconstruct the time–temperature–pres-
replacement bodies at Morococha, Peru (Kouzmanov et al. 2008; sure relationships and to study the origin and composition of the
Catchpole et al. 2011, 2015a, 2015b). fluids associated with the porphyry-style high-temperature hy-
Cerro de Pasco is a large epithermal polymetallic (Zn-Pb- drothermal veins.
Ag-Cu-Bi) deposit emplaced at shallow level, a few hundred
meters below paleo-surface, at the margin of a large diatreme–
dome complex (Einaudi 1977, Baumgartner et al. 2008, Geological setting
2009). Rottier et al. (2016a) and Rottier (2017) recently re-
ported direct evidence of three temporally distinct porphyry Cerro de Pasco is one of the largest epithermal polymetallic
events at Cerro de Pasco: (i) porphyry-style veins hosted by (BCordilleran^) deposits in the world. Historical production
hornfels and porphyritic igneous clasts in the diatreme breccia and remaining resources include more than 200 Mt (million
(PM1), (ii) shallow porphyry mineralization cropping out in tons) at around 7 wt% Zn, 2 wt% Pb, and 3 oz/t Ag in addition
the central part of the diatreme (PM2), and (iii) porphyry-style to at least 100 Mt at 1.3 wt% Cu and 1 g/t Au. Furthermore,
veins hosted by hornfels and porphyritic igneous clasts in Cerro de Pasco produced more than 1 billion oz of Ag during
east–west-trending quartz–monzonite porphyry dikes (PM3). colonial times (Baumgartner et al. 2008). The Cerro de Pasco
The present study focuses on the oldest identified porphyry district is located 300 km northeast of Lima on the western
event (PM1) recognized in hornfels and porphyritic igneous flank of the Eastern cordillera of Peru (Fig. 1a). It is part of the
clasts in the diatreme breccia that contain quartz–
molybdenite–(chalcopyrite)–(pyrite) veins. Mineralized breccia Fig. 1 Geological map and cross-section of the diatreme–dome complex„
clasts offer a unique opportunity to study the deep parts of and the different mineralization styles at Cerro de Pasco, compiled from
mineralizing systems and provide important geological infor- field observations and previous work of Rogers (1983), Baumgartner
et al. (2008), and the Volcan’s geological staff (modified from Rottier
mation (Spry et al. 1996; Müller et al. 2002; Vikre et al. 2014). et al., 2016a and b). Location of the drill holes where mineralized clasts
At Cerro de Pasco, the porphyry-style mineralization in clasts were found is indicated by the green and orange circles. Asterisk indicates
allows studying the genetic relationships between a world-class that few clasts correspond to one to three clasts every 10 m of intercepted
epithermal polymetallic deposit and deeper porphyry breccias, and numerous correspond to more than three clasts every 10 m
of intercepted breccia
Miner Deposita
 Miner Deposita

mid-Miocene metallogenic belt recognized in Central and The phreatomagmatic activity was followed by a period
Northern Peru (Noble and McKee 1999, Bissig et al. 2008; marked by volcanic and subvolcanic magmatisms that formed
Bissig and Tosdal 2009). (i) dacitic to rhyodacitic lava–dome complexes emplaced
The Cerro de Pasco district is characterized by a large along the western margin of the diatreme breccia, dated at
diatreme–dome complex 2.5 km in diameter emplaced along 15.40 ± 0.07 Ma (zircon U-Pb ID-TIMS; Baumgartner et al.
a major N 15° W-striking reverse fault, named Longitudinal 2009); (ii) a rhyodacitic porphyry stock cropping out in the
Fault (Fig. 1b). West of this fault, the diatreme–dome complex northern part of the diatreme–dome complex; (iii) several small
crosscuts metamorphosed Middle Paleozoic shales (Excelsior (each less than 10 × 10 m) porphyritic trachyte intrusions
Group) and conglomerates and sandstones of the Middle-Late cropping out in the central part of the diatreme breccia that
Triassic Mitu Group (Fig. 1b; Rosas et al. 2007; Spikings et al. are crosscut by low-density stockwork porphyry-style miner-
2016). East of the Longitudinal Fault, the diatreme–dome alization (Rottier et al. 2016a); and (iv) east–west-trending
complex is emplaced in an up to 1000-m-thick sequence of quartz–monzonite porphyritic dikes cutting the diatreme brec-
carbonate rocks belonging to the Late Triassic Chambará cia and the dacitic to rhyodacitic magmatic domes (Fig. 1b).
Formation, part of the Pucará Group (Angeles 1999; Rosas Two of these dikes have been dated at 15.35 ± 0.05 and 15.16
et al. 2007). The diatreme–dome complex was built up during ± 0.04 Ma (zircon U-Pb ID-TIMS; Baumgartner et al. 2009).
a succession of mid-Miocene phreatomagmatic and magmatic Numerous, 20-cm to 3 m-wide, E-W-trending, milled-matrix
events spanning 0.3 My (Fig. 1b; Einaudi 1968; Rogers 1983; fluidized breccia dikes were emplaced in various parts of the
Baumgartner et al. 2009). diatreme–dome complex crosscutting the diatreme breccia, the
An early phase of explosive activity produced a diatreme– dacitic and rhyodacitic domes, and the quartz–monzonite dikes.
breccia known locally as the Rumiallana Agglomerate, the most According to the morphology of the diatreme–dome complex
widespread lithology in the magmatic complex, dated by isotope and the occurrence and the type of the pyroclastic rocks (e.g.,
dilution thermal ionization mass spectrometry (ID-TIMS) zircon lapilli tuff), the total erosion since the formation of the diatreme–
U-Pb at 15.36 ± 0.03 Ma (Baumgartner et al. 2009). The geom- dome complex is estimated at < 1 km (Baumgartner 2007;
etry of the breccia body is not well-constrained; the deepest level Baumgartner et al. 2008; Rottier et al. 2016a).
reached by drilling is ~ 3300 m above sea level corresponding to Crosscutting relationships indicate that the economic
a vertical extent greater than 1150 m; the diameter of the breccia epithermal polymetallic mineralization occurred after the em-
body is close to 2000 m at the present-day surface (Fig. 1b). placement of the diatreme complex. It is developed mainly in
Three facies form the diatreme: (i) a lower facies consisting of carbonate rocks along the eastern margin of the diatreme–
non-stratified breccias, (ii) an intermediate facies marked by a dome complex (Fig. 1b; Einaudi 1977; Baumgartner et al.
stratified breccia, and (iii) an upper facies corresponding to a 2008). As reinterpreted by Rottier et al. (2016b), the following
fine-grained stratified tuff. The lower facies is only intercepted three main successive mineralizing stages are recognized: (A)
in drill holes and consists of massive non-stratified matrix-sup- pipe-like pyrrhotite-dominated bodies grading outwards into
ported polymictic breccias. The clasts are angular and generally Fe-rich sphalerite and galena rims, (B) a large N-S-trending
range from 2 to 10 cm in size but locally can reach a few meters funnel-shaped massive pyrite–quartz body, and (C) high-
and are dominantly composed of phyllites from the Excelsior sulfidation polymetallic mineralization consisting of a set of
Group, Mitu sandstone, carbonate rocks from the Pucará Group, E-W-trending Cu-Ag-(Au-Zn-Pb) enargite–pyrite veins
and subordinately altered dacite, porphyritic igneous rocks, and hosted by the diatreme breccia and large well-zoned Zn-
hornfels. The matrix makes up 20 to 50 vol% of the breccia and Pb-(Bi-Ag-Cu) carbonate replacement ore bodies in the east-
is composed of mud- to sand-sized fragments. Crosscutting re- ern part of the deposit (Fig. 1b). The enargite–pyrite veins
lationships are indicative of multiple breccia-forming events have been dated at 14.54 ± 0.08 and 14.41 ± 0.07 Ma (Ar-Ar
(Fig. 2a, b). The stratified breccia is the most common facies on alunite; Baumgartner et al. 2009).
cropping out at the surface (Fig. 2f). It is a matrix-supported
polymictic breccia with nature and size of the clasts similar to
those of the lower facies. The bed thickness is highly variable, Porphyry mineralization at Cerro de Pasco
from a few centimeters to several decimeters; some beds are
well-sorted while others are completely unsorted; no grading A link between the large epithermal polymetallic mineralization
was observed (Fig. 2e, f). The top facies of the diatreme breccia and a porphyry system at depth was already suggested by
corresponds to fine-grained well-sorted dacitic tuffs (Fig. 2c, d) Baumgartner et al. (2008). Large pervasively altered areas typical
and crops out only in the northwestern and the southwestern for a porphyry lithocap environment and minor occurrences of
parts of the diatreme–dome complex (Fig. 1b). It often shows porphyry-type mineralization have been identified at Cerro de
crossbedding, and locally, numerous accretionary lapilli occur Pasco (Rottier et al. 2016a; Rottier 2017). Pervasive alteration
(0.2–1 cm in diameter; Fig. 2d). This facies was interpreted as a affects the central and the northern parts of the diatreme–dome
base surge deposit by Baumgartner et al. (2008). complex as well as the sedimentary rocks north of it. It includes
Miner Deposita

Fig. 2 a, b Drill cores showing different generations of massive non- diatreme–dome complex). d Accretionary lapilli found in the fine-grained
stratified breccias (drill hole nos.: a DDH-S-CE-13-208. b DDH-S-CE- stratified tuff (northwestern part of the diatreme–dome complex). e, f
14-194). c Outcrop of fine-grained stratified tuff (northwestern part of the Outcrops of the stratified breccia (central part of diatreme–dome complex)
 Miner Deposita

mainly areas with pyrophyllite–quartz–pyrite mineral association to characterize fluid and silicate melt inclusions in quartz from
grading outward to illite–smectite–muscovite–pyrite and periph- the porphyry-type veins. SEM-CL was used in order to iden-
erally to chlorite–calcite–pyrite (Fig. 1b). These alteration zones tify different quartz generations in the veins, their relative time
are spatially disconnected from the main epithermal base metal relationships, and the paragenetic context of the studied fluid
mineralization. Our field observations and those of the geological and melt inclusions. SEM-CL images were acquired at the
staff of the Volcan Compañía Minera S.A (R. Bendezú, personal University of Lausanne using a CamScan MV2300 SEM
communication) show that they affect the quartz–monzonite dike and at the University of Geneva using a JEOL JSM7001F,
dated at 15.16 ± 0.04 Ma (Baumgartner et al. 2009) and are both equipped with a panchromatic CL detector.
overprinted by quartz–alunite alteration spatially associated to
the E-W-trending oxidized veins dated between 14.54 ± 0.06 Microthermometry
and 14.41 ± 0.07 Ma (Ar-Ar on alunite; Baumgartner et al.
2009; Fig. 1). The E-W oxidized veins are assumed to represent Around 300 fluid inclusions grouped in 33 fluid-inclusion as-
enargite–pyrite veins (Baumgartner et al. 2009). semblages (FIAs) were analyzed by microthermometry.
Three temporally distinct events of porphyry mineralization Microthermometric data were acquired using a THMSG 600
have been recognized at Cerro de Pasco; the main characteris- Linkam heating–freezing stage mounted on a DMLB Leica mi-
tics of each event are reported in Table 1. The first event (PM1), croscope calibrated using SynFlinc standards at − 56.6, 0.0 °C,
the one to which this contribution is devoted, consists of por- and 374.1 °C (Sterner and Bodnar 1984). For CO2-bearing fluid
phyry veins crosscutting hornfels and porphyritic igneous clasts inclusions, salinities were calculated using the CO2 clathrate
that are found in the southeast part of the diatreme–breccia (Fig. melting temperature and total homogenization of the CO2 phase
1b). Two types of quartz–molybdenite–(chalcopyrite)–(pyrite) using the equations of Diamond (1992). CO2 contents (mol%),
veins are found in the clasts. Early, thin (1–2 mm), and sinuous densities, and isochores were determined using Q2 and ISOC
HT1 veins are crosscut by slightly thicker (up to 2 cm) and software (Bakker and Brown 2003) and the equation of Duan
more regular HT2 veins. A molybdenite separate from one et al. (1992). Estimation of the liquid/vapor ratios is the largest
hornfels clast crosscut by numerous quartz– source of error using this approach, and therefore, it was quan-
molybdenite–(chalcopyrite)–(pyrite) HT1 veins (CP-16-BR- tified using the ImageJ software. For halite-saturated and liquid-
X4) has been dated by the Re-Os method at 15.6 ± 0.1 Ma rich fluid inclusions, salinities in weight percent NaCl equiv.,
(Rottier 2017; Table 1). densities, and isochores were determined using the SoWat soft-
The second event (PM2) is characterized by quartz–magne- ware (Driesner 2007; Driesner and Heinrich 2007).
tite–chalcopyrite–pyrite porphyry-style veinlets hosted by small
trachyte porphyry bodies cropping out in the central part of the Raman microspectroscopy
diatreme–dome complex (Fig. 1c; Rottier et al. 2016a). This
event took place after the emplacement of the trachyte intru- Raman microspectroscopy was used to identify minor mineral
sions dated at 15.29 ± 0.02 Ma and is older than a quartz–mon- phases, solid inclusions trapped in minerals or hosted by fluid
zonite dike dated at 15.16 ± 0.03 Ma (zircon U-Pb ID-TIMS; inclusions, gas composition of the vapor phase, and the presence
Baumgartner et al. 2009) which crosscuts the quartz–magne- of SO42− ion in the aqueous phase of fluid inclusions.
tite–chalcopyrite–pyrite porphyry-style veinlets (Rottier 2017). Measurements were performed using a confocal LABRAM in-
The last porphyry event (PM3) corresponds to quartz–mag- strument at the University of Geneva equipped with a 532.12-
netite–chalcopyrite–(pyrite) veins crosscutting porphyritic ig- nm Nd-YAG laser and an Olympus BX51 microscope. Gas
neous and hornfels clasts found in the quartz monzonite dikes. compositions were determined using a × 100 objective and the
One of the mineralized porphyritic igneous clasts has been following Raman settings: 500-μm confocal aperture, 100-μm
dated at 15.16 ± 0.04 Ma (zircon U-Pb ID-TIMS; Rottier slit, and an acquisition time of 2 × 120 s. The limit of detection
2017), an age overlapping within the error that was obtained for this setting is better than 0.13 mol% CO2 (Rosso and Bodnar
for the quartz–monzonite dike hosting it (15.16 ± 0.03 Ma; 1995). The presence of SO42− ion in the aqueous phase was
Baumgartner et al. 2009). determined using a × 50 objective, 500-μm confocal aperture,
100-μm slit, and acquisition time of 60 s. Measurements were
done from room temperature up to 450 °C; heating was per-
Analytical methods formed using the THMSG 600 Linkam heating–freezing stage.

Sample selection and SEM-CL petrography SIMS analyses

A total of eight clasts crosscut by HT1 or HT2 veins were Three different porphyry veins were selected for SIMS oxy-
selected for this study. Doubly polished thick sections gen isotope analyses: one HT1bt, one HT1px, and one HT2
(200 μm) were examined using a petrographic microscope vein. A total of 22 18O/16O isotopic ratio analyses were
Miner Deposita

Table 1 Summary of the main characteristics of the different porphyry events at Cerro de Pasco

Porphyry events PM1 PM2 PM3

Age 15.59 ± 0.12 Maa Between 15.286 ± 0.018a and 15.165 ± 0.043a (clast)
15.158 ± 0.035 Mab and 15.158 ± 0.035
Mab (host dike)
Occurrence Veins crosscutting hornfels and Stockwork of veins crosscutting Veins crosscutting hornfels and
porphyritic igneous clasts hosted porphyritic trachyte intrusions porphyritic igneous clasts hosted
in the diatreme breccia in a quartz monzonite dike
Recognized alteration Potassic and phyllic High-temperature sodic (albitization) Potassic
and low-temperature sodic(−calcic)
Type of veins Quartz–molybdenite–(chalcopyrite)– quartz–magnetite–chalcopyrite–pyrite Quartz–magnetite–chalcopyrite–
(pyrite) veins veins (pyrite) veins
Quartz precipitation > 600 °C First generation > 600 °C; second > 600 °C
temperature generation ≈ 350 °C
Sulfide precipitation > 600 °C ≈ 350 °C > 600 °C
temperatures
Pressure of formation 500–800 bar < 290 bar > 600 bar
References This study Rottier et al., (2016a) Rottier (2017)

a
Rottier (2017)
b
Baumgartner et al. (2009)

performed on quartz chips placed in indium mounts that in- from the estimated salinity was used as an internal standard; an
cluded two grains of the reference material (UNIL-Q1; Seitz empirical mass balance correction for major dissolved cations
et al. 2016). Measurements were performed at the SwissSIMS other than sodium, K, Fe, Mn, Zn, Cu, and Pb, was applied
ion probe national facility at the University of Lausanne, using (Heinrich et al. 2003). Quantification of the SMIs has been done
a Cameca IMS180-HR. The detailed procedure is provided by following the method of Halter et al. (2002). A fixed Al value of
Rottier et al. (2016a). 16 wt% Al2O3 was used as an internal standard for all analyzed
melt inclusions. This value corresponds to the calculated mean
LA-ICP-MS analyses Al2O3 content obtained from whole-rock XRF analyses of the
volcanic and subvolcanic rocks from the district, ranging in com-
LA-ICP-MS has been used to determine the composition of fluid position from dacite to rhyolite (mean = 15.97 ± 0.87 wt%
(n = 134) and silicate melt inclusions (n = 53), as well as the trace Al2O3, n = 12; Rottier et al. 2016a). The silica content of each
element content of quartz (n = 168), rutile (n = 23), and titanite SMI is calculated based on the equation SiO2 wt% = 100 −
(n = 5). Analyses were performed using a 193-nm ArF Excimer wt%(K 2 0) − wt%(Na 2 O) − wt%(Fe 2 O 3 ) − wt%(Al 2 O 3 ) −
laser with an energy-homogenized beam profile coupled with an wt%(MgO). For host correction (i.e., quartz) and quartz trace
ELAN 6100 DRC ICP quadrupole mass spectrometer (QMS) at element analyses, 99.5 wt% SiO2 was assumed as an internal
ETH Zurich (Günther et al. 1997, 1998; Heinrich et al. 2003). standard. Stoichiometric Ti (59.31 wt%) and Ca (19.25 wt%)
Summary of analytical conditions, data acquisition parameters, were, respectively, used as internal standards for analyses of rutile
and list of isotopes analyzed during each session are provided in and titanite.
the electronic supplementary material (ESM) Table A.1. To limit
the explosion of the analyzed quartz, an iris diaphragm was used Mineral thermobarometers
(Guillong and Heinrich 2007), allowing a progressive increase of
the laser beam diameter during ablation (Günther et al. 1998); the Ti in quartz
final ablation pits were between 20 and 50 μm in diameter. Rutile
and titanite were measured with a fixed beam diameter from 20 Titanium concentrations determined by LA-ICP-MS were
to 40 μm when they were cropping out at the surface and other- used as proxy for quartz precipitation temperatures, using
wise by ablation through quartz following similar procedures as the experimentally calibrated thermobarometer of Huang and
for SMIs and fluid inclusions (Pettke 2006). LA-ICP-MS tran- Audétat (2012). The activity of TiO2 (αTiO2) is assumed to be
sient signals were reduced using the SILLS software (Guillong equal to 1, considering the presence of rutile in the hydrother-
et al. 2008). The NIST SRM 610 glass standard was used as an mal veins and also in the host hornfels and igneous clasts. The
external standard and analyzed two times between a maximum main limit of this thermobarometer is the important effect of
of 20 unknowns to correct the instrument drift by linear regres- growth rate on the Ti concentration in quartz (Huang and
sions. For fluid inclusion analyses, Na concentration derived Audétat 2012), their thermobarometer being calibrated for a
 Miner Deposita

growth rate of 2–4 μm/day. If the growth rate is faster or Results

slower, the calculated temperature will be respectively
overestimated (up to 250 °C; Huang and Audétat 2012) or Mineralized clast petrography
underestimated. This effect cannot be quantified, and all tem-
peratures should be considered as approximate. Titanium dif- The clasts studied range from 2 to 10 cm in size; they are
fusion could also affect the original Ti concentration and mostly rounded and were found exclusively in the massive
therefore the calculated temperatures. At 600 °C, Ti diffuses non-stratified matrix-supported polymictic breccia, the basal
by ~ 15 μm over a million years (Cherniak et al. 2007a). unit of the diatreme. A total of 18 different holes drilled in the
According to the low Ti diffusion rate and the LA-ICP-MS eastern part of the diatreme–breccia were investigated (Fig.
spot size (20 to 50 μm) used, the potential effect of Ti diffu- 1b). Porphyry veins have been found in ten of them, all in
sion is considered as negligible; a detailed discussion on this hornfels clasts except one clast where the veinlets crosscut a
issue could be found in Mercer and Reed (2013). porphyritic igneous rock (Fig. 3c); more details (drill hole
number, location, depth, and relative abundance) are reported
in ESM Table A.2. Observations on the studied drill holes
Zr in rutile
suggest that mineralized clasts are more abundant in the south-
eastern part of the diatreme complex (Fig. 1b).
Hydrothermal rutile precipitation temperatures were calculated
Mineralized hornfels clasts have a fine-grained texture. Their
using the experimental thermobarometer of Tomkins et al.
composition was determined by combining classical petrogra-
(2007). Zircon and quartz need to be present to satisfy equilib-
phy, XRD, and Raman microspectroscopy. Based on their min-
rium controlling the Zr content in rutile (Tomkins et al. 2007).
eral compositions, three different types of hornfels clasts could
Analyzed rutile grains were located at the margin or inside the
be distinguished. The first type is characterized by a mineral
veins and were always intergrown with or included in quartz.
assemblage typical for potassic alteration, including quartz, K-
No zircon was found inside the veins, but in all studied clasts
feldspar, biotite, andalusite, albite, and pyrite (Fig. 3a). The
(hornfels or igneous), numerous small zircon grains (< 50 μm)
second is interpreted to result from overprinting lower-
are present at the border of the vein or in the immediate vicinity
temperature argillic alteration on early potassic alteration and
of the veins (< 100 μm from the border). We assumed that the
consists of quartz, K-feldspar, albite, kaolinite, dickite, and py-
analyzed rutile grains were precipitated from a Zr-saturated
rite (Fig. 3b, d). The third one results from phyllic alteration and
hydrothermal fluid. Therefore, calculated temperatures have
includes quartz, muscovite, illite, siderite, and pyrite (Fig. 3e).
to be considered as minimum temperatures. The error (2σ),
In addition, all the clasts contain grains of apatite, rutile, and
derived from the analytical and calibration uncertainties, is <
zircon. HT1 veins have been observed in the first two types of
30 °C (Tomkins et al. 2007). Zirconium diffusion in rutile is
clasts and HT2 veins in the last two types of clasts.
three orders of magnitude slower than Ti in quartz (Cherniak
Judging by their morphological and textural features as well
et al. 2007b), and therefore, the effect of Zr diffusion on cal-
as their mineral compositions, the thin, sinuous, HT1 and the
culated temperatures is assumed to be negligible.
slightly thicker, more regular HT2 quartz–
molybdenite–(chalcopyrite)–(pyrite) veins share common
Zr in titanite characteristics with A- and B-type veins defined by
Gustafson and Hunt (1975). Considering the fact that the veins
Hydrothermal titanite precipitation temperatures were calcu- are restricted to small clasts (< 10 cm) and therefore, the obser-
lated using the experimental thermobarometer of Hayden et al. vations are limited, we prefer to use the more general terminol-
(2008). This thermobarometer was calibrated for a titanite– ogies HT1 and HT2, with BHT^ referring to high temperature.
rutile–quartz–zircon assemblage. Analyzed titanite crystals
were always included in quartz forming the veins. Rutile in HT1 veins
the studied sample is a common mineral either in the quartz
veins or in the host lithology, and as already mentioned, small The 1–2 mm-thin, commonly sinuous HT1 veins are sulfide-
zircon grains were present in all clasts. Therefore, αTiO2 and poor (≪ 1 vol%); the main sulfides are molybdenite, pyrite, and
αSiO2 were fixed at 1. Calibration uncertainty associated with chalcopyrite. SEM-CL reveals a single highly luminescent
this thermobarometer is < ± 20 °C, and analytical uncertainty quartz generation contemporaneous with the sulfide precipita-
is close to 0 °C due to the high sensitivity of the tion event and characterized by subrounded oscillatory growth
thermobarometer (Hayden et al. 2008). Zirconium diffusion zoning (Fig. 4b, c). Molybdenite grains vary from 20 to 100 μm
in titanite is similar or slower than the diffusion in rutile in size, occur principally along the contact of the veins with the
(Cherniak et al. 2007b); therefore, as with rutile, the effect of host clasts (Fig. 5c), and subordinately are intergranular be-
Zr diffusion on calculated temperatures is considered as tween quartz crystals or as inclusions in the latter. Two distinct
negligible. subclasses of HT1 veins were defined based on the nature of
Miner Deposita

Fig. 3 Mineralized clasts in the massive non-stratified breccia from drill potassic alteration, and crosscut by an HT2 vein (sample BR-260). e
cores. a Hornfels clast affected by a high-temperature potassic alteration Hornfels clast strongly affected by phyllic alteration and crosscut by a
and crosscut by an HT1px vein (sample BR-259). b Hornfels clast affected HT2 vein (sample BR-267). f Hornfels clast affected by lower-
by a high-temperature potassic alteration and crosscut by an HT1bt vein temperature argillic alteration (kaolinite–dickite–(illite) assemblage),
(sample BR-257). c Porphyry clast crosscut by an HT1px vein (sample overprinting early potassic alteration, and crosscut by a HT2 vein
BR-300). d Hornfels clast affected by a lower-temperature argillic (sample BR-261)
alteration (kaolinite–dickite–(illite) assemblage), overprinting an early

Fig. 4 a SEM-CL image of a

HT2 vein crosscutting HT1bt
veins (Sample BR-312). b, c
SEM-CL images showing typical
quartz textures from HT1 veins
 Miner Deposita

Fig. 5 a Typical solid-inclusion assemblage found in quartz forming mineralization (sample BR-267); corresponding SEM-CL image is Fig.
HT1bt veins (sample BR-261). b Typical solid-inclusion assemblage 8c. Act actinolite, Aug augite, Bt biotite, Cpy chalcopyrite, Gn galena, K-
found in quartz from HT1px veins (sample BR-300). c Molybdenite Fsp K-feldspar, Mo molybdenite, Py pyrite, Rt rutile, SMI silicate melt
grain at the contact of HT1px quartz vein with hornfels (sample BR- inclusion, Sp sphalerite, Td-Tn tetrahedrite–tennantite
259). d Typical sulfide assemblage associated with the late polymetallic

mineral inclusions in the hydrothermal quartz: (i) HT1bt veins numerous fluid inclusions and some rare crystallized or glassy
commonly host numerous inclusions of K-feldspar, biotite, ru- SMIs. SEM-CL imaging revealed two distinct quartz genera-
tile, and minor titanite (Fig. 5a), and (ii) HT1px veins host nu- tions: (i) a main quartz generation characterized by euhedral
merous inclusions of actinolite, augite, titanite, apatite, and mi- crystals with oscillatory growth zoning and (ii) a second quartz
nor rutile (Fig. 5b). In both vein subclasses, mineral inclusions generation, minor in volume, overgrowing and showing subtle
are small, range in size from few to tens of micrometers, and are subrounded oscillatory growth zoning and, locally, dissolution
randomly distributed in quartz. Numerous fluid inclusions and textures over the first one (ESM Fig. A.1). Sulfides (molybde-
a few glassy silicate melt inclusions are commonly observed in nite and pyrite) are intergrown with both quartz generations.
quartz crystals from both HT1 vein subtypes.
Pre-diatreme polymetallic mineralization
HT2 veins
Commonly, HT1 and HT2 veins in diatreme clasts are cross-
The less sinuous morphology of the up to 2-cm-thick HT2 veins cut by late 100- to 200-μm-thick sulfide-rich veinlets that
(Fig. 3d–f) suggests that they formed under more brittle condi- mainly contain pyrite, sphalerite with Bchalcopyrite disease,^
tions compared to HT1 veins. In one hornfels clast, HT2 veins galena, chalcopyrite, and tetrahedrite–tennantite (Fig. 5d), as-
are observed to crosscut HT1 veins. This supports an early for- sociated occasionally with weakly luminescent quartz, as re-
mation of HT1 veins (Fig. 4a). HT2 veins are also sulfide-poor vealed by SEM-CL imaging. In part, these sulfide-rich vein-
(< 1 vol%), containing mainly pyrite and rare molybdenite lets are emplaced by the reopening of HT1 and HT2 veins.
grains. In contrast to HT1 veins, HT2 veins are composed of Based on the sulfide assemblages and the low CL lumines-
quartz almost free of mineral inclusions; only scarce rutile inclu- cence of the quartz, we conclude that the polymetallic veinlets
sions, from 10 to 50 μm in size, were found. HT2 veins host in the diatreme clasts probably formed at intermediate
Miner Deposita

temperatures < 350 °C. These intermediate-temperature sul- in HT1px veins varies from 19 to 64 ppm (Fig. 7a). Quartz from
fide-rich veinlets are restricted to the mineralized clasts and HT2 veins generally has lower Ti content (from 2 to 31 ppm)
therefore are pre-diatreme and earlier than the large epithermal than quartz from HT1 veins (from 81 to 114 pm), except sample
mineralizing event at Cerro de Pasco that crosscuts the BR-312 that has higher Ti content. No differences in Ti content
diatreme (Baumgartner et al. 2008; Rottier et al. 2016b). were observed between the two quartz generations that form the
HT2 veins (ESM Fig. A.1). Quartz from HT2 veins is also
characterized by significant enrichment in Al (up to
Trace element analysis of quartz, rutile, and titanite
2000 ppm) compared to that from HT1 veins (Fig. 6).
LA-ICP-MS profiles in quartz displaying oscillatory growth
LA-ICP-MS analyses were performed in areas devoid of solid
zoning in SEM-CL images reveal only a weak correlation be-
and fluid inclusions. For quartz and direct analyses of rutile
tween CL intensity and Ti content (ESM Fig. A.1). An elevated
grains, only analyses with flat LA-ICP-MS transient signals
Ti content has been considered to be the main reason for high
were considered. For rutile and titanite present as solid inclu-
CL intensity in high-temperature quartz (Rusk et al. 2008c;
sions in quartz, only signals clearly not disturbed by other
Müller et al. 2010; Mercer and Reed 2013; Mercer et al.
potential solid or fluid inclusions were considered. Potential
2015). The weak correlation observed in our analyses may be
contamination of the ablation signals by fluid, melt, and not
attributed to the different volume of material analyzed by SEM-
desired solid inclusions was carefully checked by tracking the
CL and by LA-ICP-MS. Only the first micrometers of the quartz
signals of Cs, Sr, Rb, Zr, W, and Fe.
are involved in generating the SEM-CL light intensity, whereas
LA-ICP-MS spot depth is typically in excess of 100 μm.
Quartz

Quartz crystals from eight porphyry-type veins (three HT1bt Rutile

veins, two HT1px veins, and three HT2 veins) were analyzed
by LA-ICP-MS. Analytical data are reported in ESM Rutile crystals from two HT1bt veins, two HT1px veins, and
Table A.3. Quartz of the different vein types displays Ti and two HT2 veins were analyzed by LA-ICP-MS. The analyzed
Al concentration ranges (Figs. 6 and 7a), comparable with grains were small (< 50 μm) and generally totally included in
ranges from quartz veins of other porphyry Cu-Mo deposits quartz. In most LA-ICP-MS analyses, the entire rutile grain
reported by Rusk et al. (2008c). was ablated, thus avoiding problems related to chemical zon-
In each vein type, the Ti concentration in quartz grains of the ing and diffusion inside the grains.
same generation is highly variable. In HT1bt and HT1px veins, Rutile contains up to several weight percent Fe and W and
SEM-CL images reveal just one quartz generation with small several thousand parts per million of Nb and Zr. Results are
differences of CL intensity (Fig. 4b, c), that do not correlate with reported in ESM Table A.3. Zirconium concentrations are
the Ti content in quartz. HT1bt veins show the most important quite similar in rutile from HT1 veins, with 836 to
variation in Ti, from 32 to 176 ppm, while the Ti concentration 1953 ppm for HT1bt veins and 699 to 3280 ppm for HT1px

Fig. 6 Concentrations of Ti and

Al in quartz from HT1bt, HT1px,
and HT2 veins. Inset shows the
large range of Al concentrations
in HT2 veins. Data from Rusk
et al. (2008c) of hydrothermal
quartz from porphyry veins of
various porphyry deposits are
plotted for comparison
 Miner Deposita

Fig. 7 a Box plots of Ti concentrations in quartz from the different boxes are the mean while the black dots represent the median value.
samples analyzed. b Box plots of Zr concentration in rutile from the BWhiskers^ of each box extend to the extreme values lying within 1.5
different samples analyzed. Number of analyses is indicated above each times the interquartile range beyond the edges of the box. The empty
box plot. Box’s height represents the interquartile range (i.e., representing circles and triangles represent outliers deviating more than ± 2.7σ. Full
the most frequent 50% of the distribution). The black lines dividing the circles represent isolated values

veins, whereas they are much lower in rutile from HT2 veins, diatreme polymetallic mineralization gave δ18O values of
with 153 to 277 ppm (Fig. 7b). In rutile from HT1 veins, a 17.4 and 16.8‰ (Fig. 8c).
weak positive correlation of Nb and Mo, with Zr, is recog-
nized, probably due to co-precipitation of these elements dur- Fluid inclusions
ing cooling of the fluid. In rutile from HT2 veins, no correla-
tion between any element and Zr has been noticed. Petrography, microthermometry, and Raman
microspectrometry
Titanite
Five fluid inclusion types were identified according to the
Five titanite grains included in quartz were analyzed: one from number, the nature, and the volume proportions of phases
an HT1bt vein and four from an HT1px vein. The analyzed present at room temperature. They are (i) CO2-rich high-
titanite grains were small (< 30 μm) and were entirely ablated density vapor inclusions (Fig. 9a, b), (ii) polyphase brine in-
during LA-ICP-MS analyses; compositions are reported in clusions (Fig. 9c, d), (iii) low-density vapor inclusions, (iv)
ESM Table A.3. Zirconium content in titanite is 803 ppm for brine inclusions (Fig. 9e), and (v) aqueous inclusions (Fig. 9f).
the HT1bt vein (sample BR-261) and varies between 402 and The CO2-rich high-density vapor inclusions were found
824 ppm for the HT1px vein (sample BR-300). Titanite con- exclusively in HT1bt and HT1px veins, whereas the polyphase
tains several thousand parts per million of Al, Fe, La, and Ce brine and the low-density vapor inclusions are restricted to
and several hundred parts per million of Mg, Y, and P. A HT2 veins. Brine and aqueous inclusions are low-temperature,
positive correlation between Nb, La, and Ce with Zr exists, secondary, and found in all vein types. Microthermometric
potentially indicating a preferential incorporation of these el- results are summarized in Table 2.
ements in high-temperature titanite. The CO2-rich high-density vapor inclusions are between 5
and 35 μm in size, and most of them contain one hematite flake
In situ SIMS oxygen isotope analysis of different and one anhydrite crystal (Fig. 9a, b). They are mostly organized
quartz generations as patches in the core of the quartz and also form secondary
trails. At room temperature, only liquid and vapor phases are
A total of 22 in situ SIMS oxygen isotope analyses of quartz visible; during cooling, liquid CO2 appears at around + 12 °C. In
from porphyry-type veins were performed on samples BR- the absence of liquid CO2, the inclusions contain 55 to 90%
257 (HT1bt vein), BR-259 (HT1px vein), and BR-267 (HT2 vapor and their L/V ratio is constant within the same assem-
vein). The results are nearly identical in the three samples and blage. The measured melting point of CO2 varies from − 56.0
show no correlation with the textures revealed by SEM-CL to − 56.5 °C; CO2 was the only gas detected by Raman
(Fig. 8): δ18O HT1bt vein = 12.0 ± 0.3‰ (n = 4), δ18O HT1px microspectroscopy. Clathrate melting occurs between − 1.0
vein = 12.1 ± 0.3‰ (n = 7), δ18O HT2 vein = 12.4 ± 0.2‰ and + 9.2 °C, and CO2 homogenizes to vapor between 10.2
(n = 7). Two analyses on quartz associated with the pre- and 14.0 °C. Total homogenization occurs by the disappearance
Miner Deposita

Fig. 8 Summary of δO18 SIMS analyses of quartz from HT1bt (a), HT1px (b), and HT2 veins (blue values) and from late polymetallic mineralization
(orange values, c; similar to Fig. 5d)

Fig. 9 a CO2-rich high-density vapor inclusions hosted in HT1bt vein. b CO2-rich high-density vapor inclusion in an HT1px vein. c, d Polyphase brine
inclusions hosted in HT2 vein. e Low-temperature brine in an HT1bt vein. f Low-temperature aqueous fluid inclusions in an HT1bt vein
Table 2 Microthermometry data for fluid inclusion assemblages

FIA Veins n Type L% TmCO2 σ Tmchl σ ThCO2 σ TmIce σ TmSyl? σ TmHal σ Thtotal σ Mode Salinity eq σ XCO2 σ d (g/cm3) Pmin (bar)

300L_8 HT1px 11 LlCO2 30 − 56.2 0.1 3.2 1.6 10.2 1.9 421 8.0 V 11.5 2.2 0.14 0.008 ≈ 0.38 360
300I_2 HT1px 13 LlCO2 20 − 56.5 0.1 2.9 2.0 11.5 1.9 382 13.6 V 11.8 2.8 0.22 0.004 ≈ 0.31 290
300E_9 HT1px 14 LlCO2 45 − 56.2 0.3 − 0.3 1.1 10.5 1.2 550 12.0 V 15.8 1.1 0.07 0.003 ≈ 0.6 1470
300L_6 HT1px 9 LlCO2 25 − 56.4 0.1 − 0.8 1.0 11.0 2.1 412 15.0 V 16.3 1.0 0.17 0.003 ≈ 0.35 230
259C4_1 HT1px 7 LlCO2 15 − 56.1 0.1 0.6 3.4 14.0 1.1 391 11.0 V 14.4 3.7 0.28 0.004 ≈ 0.27 205
259C4_2 HT1px 7 LlCO2 10 − 56.0 0.0 − 1.0 0.5 13.5 399 9.1 V 16.6 0.4 0.37 0.001 ≈ 0.23 217
259C2_2 HT1px 9 LlCO2 15 − 56.3 0.1 2.3 1.8 12.5 386 14.4 V 12.6 2.2 0.27 0.006 ≈ 0.27 298
257C4_1 HT1bt 7 LlCO2 10 − 56.0 0.0 2.6 1.1 12.5 Nd Nd V 12.4 1.4 0.38 0.000 ≈ 0.23 Nd
311C_1 HT1bt 7 LlCO2 45 − 56.4 0.0 9.1 0.5 12.0 394 25.4 V 1.8 1.0 0.10 0.000 ≈ 0.53 590
311E_2 HT1bt 8 LlCO2 15 − 56.5 0.0 7.4 0.8 12.5 384 20.9 V 5.0 1.5 0.29 0.000 ≈ 0.28 370
311D_5 HT1bt 6 LlCO2 45 − 56.5 0.1 7.6 0.8 12.0 395 10.8 V 4.7 1.4 0.10 0.001 ≈ 0.53 540
311B_1 HT1bt 5 LlCO2 10 − 56.3 0.0 6.2 0.9 11.5 Nd Nd V 7.0 1.5 0.38 0.001 ≈ 0.23 Nd
267C2 HT2 15 P. brine 10–20 141 3.8 377 11.0 L? 44.7 1.1 0.9a 748a
267C3 HT2 14 P. brine 10–20 132 6.8 408 23.6 L? 48.0 2.6 0.93a 709a
267C6_1 HT2 8 P. brine 10–20 137 2.8 364 5.0 L? 43.3 0.5 0.88a 763a
267C6_3 HT2 14 P. brine 10–20 140 4.4 378 9.8 L? 44.7 1.0 0.90a 757a
267C6_2 HT2 7 P. brine 10–20 145 6.7 369 2.5 L? 43.8 0.2 0.88a 747a
267C7_2 HT2 14 P. brine 10–20 136 4.7 318 11.5 L? 39.1 1.0 0.83a 810a
313A2 HT2 8 P. brine 10–20 153 11.4 273 10.9 L? 35.8 0.7 0.80a 790a
313A2b HT2 6 P. brine 10–20 145 12.2 306 12.9 L? 38.2 1.0 0.82a 785a
313A1 HT2 8 P. brine 10–20 165 12.7 393 9.6 L? 42.3 1.0 0.91a 728a
312_2 HT2 14 P. brine 10–20 Nd Nd 344 19.7 L? 41.4 1.8 0.86a 843a
312_1 HT2 7 P. brine 10–20 118 0.8 338 8.1 L? 40.9 0.7 0.85a 819a
261_2_2C HT1bt 7 Aqueous 80 − 16.0 0.9 264 13.4 L 19.4 0.7 0.96 44
261_2_3 HT1bt 9 Aqueous 80 − 23.4 2.5 235 6.3 L 24.5 1.7 1.03 25
257_C4_2 HT1bt 11 Aqueous 80 − 15.2 0.7 231 9.6 L 18.8 0.6 0.99 25
257_C2_1 HT1bt 9 Aqueous 80 − 3.6 0.2 226 12.5 L 5.8 0.2 0.88 25
257_C3_1 HT1bt 7 Aqueous 80 − 1.1 0.1 245 5.4 L 1.9 0.2 0.82 36
257_C4_3 HT1bt 4 Brine 75–80 222 21.9 239 9.2 Lb 30.8 4.0 1.1 24
257_C4_4 HT1bt 3 Brine 75–80 251 18.5 277 22.6 Lb 34.8 1.2 1.1 45
257_C4_5 HT1bt 2 Brine 75–80 191 9.9 226 11.3 Lb 31.4 0.5 1.1 19
257_C3_2 HT1bt 8 Brine 75–80 274 19.0 275 19.0 Hb 36.3 1.3 1.2 1756
257_C3_3 HT1bt 6 Brine 75–80 253 14.2 256 10.2 La 34.9 0.9 1.1 31

Data are reported as an average for each fluid inclusion assemblage with 1σ standard deviation. Host mineral of all studied inclusions is quartz
L% mean volume percent of the liquid phase in the inclusions, TmCO2 temperature of melting of solid CO2. Tmchl temperature of melting of CO2 clathrate, ThCO2 temperature of the total homogenization
of the CO2 phase, TmIce ice melting temperature, TmSyl? temperature of melting of solid phase identified as sylvite, TmHal halite melting temperature, Thtotal temperature of total homogenization, V
vapor, L liquid, H halite, Salinity eq calculated salinity in weight percent NaCl equiv., XCO2 measured CO2 concentration, d calculated density, Pmin calculated minimal pressure of entrapment
a
Densities and pressures calculated assuming a homogenization temperature of 600 °C
b
Principal homogenization behavior of the FIAs
Miner Deposita
Miner Deposita

of liquid between 382 and 421 °C (Fig. 10; purple and red salinities are considered just as indicative as, for the recorded
diamonds), except for one FIA (no. 300E9; Table 2) which melting temperatures, no satisfactory equation exists to resolve
homogenizes at around 550 °C. Calculated salinities are differ- the compositions dominated by FeCl, KCl, and NaCl found in
ent in FIAs from HT1bt veins, with values between 1.8 and the polyphase brine inclusions (see Fig. 11 and ESM Fig. A.3).
12.4 wt% NaCl equiv., and in FIAs from HT1px veins, with The polyphase brine inclusions were heated up to 580 °C with-
values between 11.5 and 16.6 wt% NaCl equiv. (Fig. 10). out reaching total homogenization; a very small vapor bubble
Raman analyses of the liquid phase in CO2-rich inclusions re- remained. Judging from the remaining bubble size, total homog-
veal the presence of SO42− ion in solution (ESM Fig. A.2). enization of such inclusions most likely occurs at temperatures
Analyses were performed from room temperature to 450 °C, around or slightly above 600 °C. Raman analysis of the vapor
and an increase of the area of SO42− peak at 980 cm−1 was phase of polyphase brine inclusions has detected the presence of
observed with increasing temperature, probably due to the pro- CO2 (ESM Fig. A.2). No microthermometric measurements
gressive dissolution of the anhydrite crystals present in the fluid were done on the low-density vapor inclusions.
inclusions (Fig. 9b and ESM Fig. A.2). HSO42− ions and H2S Low-temperature brine inclusions occur in secondary trails.
have not been detected in the liquid and the vapor phases, re- They are generally small, around 5 μm, with a few larger in-
spectively. The presence of detectable SO42− indicates that the clusions up to 25 μm in size (Fig. 9e). The vapor phase typically
fluid was sulfur-rich, and in combination with the presence of occupies ~ 20% of the inclusion volume. Around 50% of the
hematite, this is indicative of the oxidizing nature of the fluid measured brine inclusions homogenized by vapor disappear-
trapped in the inclusions. Such oxidizing conditions may not ance and 50% by halite disappearance and both types of ho-
represent the initial composition of the trapped fluid and could mogenization behavior were observed in some FIAs. Total va-
also be related to post-entrapment H2 diffusion out of the inclu- por disappearance occurs from 183 to 277 °C, and halite disso-
sions (Mavrogenes and Bodnar 1994). lution temperatures vary between 191 and 275 °C, correspond-
The polyphase brine inclusions and low-density vapor inclu- ing to salinities from 30.8 to 36.3 wt% NaCl equiv. (Fig. 10;
sions form secondary trails through quartz crystals. A few clear green squares). Total homogenization by halite dissolution is
boiling assemblages with coexisting polyphase brine and low- common for brine inclusions in porphyry systems (e.g., Rusk
density vapor in the same trail are observed. The polyphase et al. 2008a; Lecumberri-Sanchez et al. 2015) and is commonly
brine inclusions range from 10 to 25 μm in size and always interpreted as due to high trapping pressures (Becker et al.
contain two transparent crystals and rarely an opaque crystal 2008; Lecumberri-Sanchez et al. 2012) or as indicative of het-
and/or a hematite flake (Fig. 9c, d). The vapor phase composes erogeneous entrapment of fluid saturated in halite (Lecumberri-
30 to 50 vol% of the inclusion volume. One transparent crystal is Sanchez et al. 2015). Raman analysis of these inclusions did not
rounded, pale green in color, and melts between 118 and 165 °C; detect gas (e.g., CO2, CH4…) in the vapor phase.
it has been interpreted as sylvite. The other transparent mineral is Aqueous inclusions form secondary trails. Inclusions range
halite and melts between 273 and 408 °C. Salinities were calcu- in size from 5 to 40 μm. Vapor occupies ~ 20 vol% of the
lated based on halite melting temperatures and vary between inclusions. No solids were observed (Fig. 9f). The ice melting
35.8 and 48.0 wt% NaCl equiv. (Fig. 10; blue circles). These temperatures range from − 1.1 to − 23.4 °C, corresponding to

Fig. 10 Microthermometric
results of salinity (wt% NaCl
equiv.) and homogenization
temperature (°C). Each FIA is
represented by its mean value,
and the error bars represent ± 1σ.
Purple and red diamonds are
CO2-rich high-density vapor
FIAs, respectively, hosted in
HT1px and HT1bt veins. Blue
circles represent polyphase brine
FIAs hosted in HT2 veins, and
homogenization temperatures
were fixed at 600 °C. Green and
yellow squares are brine and
aqueous FIAs in HT1bt veins,
respectively
 Miner Deposita

Fig. 11 Element concentration

ratios relative to Na, K, and Fe (X/
Na + K + Fe) of individual fluid-
inclusion assemblages obtained
from LA-ICP-MS analyses. Fluid
inclusions are sorted by fluid type
and by salinity decrease
Miner Deposita

salinities of 1.9 to 24.5 wt% NaCl equiv. (Fig. 10; yellow Melt inclusions
squares). One FIA shows ice melting temperatures below −
21.2 °C, in metastable absence of hydrohalite. Total homoge- Petrography and description
nization of aqueous inclusions occurs by vapor disappearance
at temperatures between 226 and 264 °C (Fig. 10). Raman SMIs, generally 5–10 μm in diameter but locally up to 25 μm
analyses of these inclusions indicate the presence of small (Fig. 12), were identified in HT1 and HT2 veins. They mainly
amounts of CO2. occur as isolated inclusions, rarely grouped in melt-inclusion
assemblages. Petrography combined with SEM-CL imaging
Fluid composition reveals that the silicate melt inclusions are randomly distrib-
uted and are not spatially associated to fluid inclusions or to
Results of LA-ICP-MS analyses of fluid inclusions are plotted particular quartz growth bands. SEM-CL images do not show
as selected element concentration ratios [X/(Na + K + Fe)]. magmatic xenocrystic quartz grains and suggest that all quartz
They are classified by fluid inclusion and vein types and did grow locally (Figs. 4 and 8 and ESM Fig. A.1).
sorted by salinity decrease (Fig. 11). Only element concentra- In HT1 veins, the SMIs are glassy and contain a shrinkage
tion ratios [X/(Na + K + Fe)] are plotted and discussed to avoid bubble filling ~ 10 vol% of the inclusion and rarely a crystal
errors induced by the determination of sodium content from occupying < 5 vol% of the SMI (Fig. 12a–d). In HT2 veins,
microthermometry measurements. the SMIs are either glassy like in HT1 veins or totally crystal-
The fluid composition of CO2-rich high-density vapor in- lized, forming a dark mass of fine-grained crystals with no
clusions is similar in HT1bt and HT1px veins for most ele- discernible vapor bubble (Fig. 12e, f).
ments except for Au, Cu, and Zn that show higher concentra-
tion ratios in fluid inclusions in HT1bt veins (Fig. 11). Fluid Melt inclusion composition
compositions of the polyphase brine inclusions from HT2
veins show lower concentration ratios for Zn, Sr, Mo, Cu, A total of 53 melt inclusions from HT1 and HT2 veins were
Au, Li, and Cs than those in HT1 veins (Fig. 11). analyzed by LA-ICP-MS, 39 from HT1bt veins, 4 from HT1px
Concentration ratios in the late fluids (the low-temperature veins, and 10 from HT2 veins; results are reported in ESM
brine and aqueous fluid inclusions, Fig. 11) are generally Table A.4. Only inclusions fully enclosed by quartz and not
scattered with a large standard deviation for each FIA. intersected by any crack or fluid inclusion trail were selected
Compared to the CO2-rich high-density vapor inclusions and for analysis. The analyzed SMIs have been compared to
polyphase brine inclusions from HT1 and HT2 veins, concen- whole-rock data of dacite and rhyodacite domes and quartz
tration ratios of B, Li, Ca, Sr, Cs, and Ba are significantly monzonite dikes from the Cerro de Pasco district (Rottier et al.
higher whereas those of Pb, Zn, and Fe are lower (Fig. 11). 2016a) and also to published data of SMI compositions hosted
Mo/(Na + K + Fe) ratios show important variations from in quartz from miarolitic cavities from various barren granitic
6.3 × 10−3 to 4.9 × 10−5 and tend to decrease with the relative systems (Zajacz et al. 2008) and to SMIs found in porphyry-
timing from HT1bt veins to HT1px veins and to HT2 veins type quartz veins from the Elatsite porphyry Cu-Au deposit
(Fig. 11). The obtained values are in the range of published (Bulgaria; Stefanova et al. 2014); only SMI compositions
data from Cu-Mo and Mo porphyry systems (Ulrich et al. classified by Zajacz et al. (2008) and Stefanova et al. (2014)
2002; Landtwing et al. 2005, 2010; Klemm et al. 2007, as not affected by the hydrothermal fluid were selected for
2008; Kouzmanov and Pokrovski 2012; Lerchbaumer and comparison (Fig. 13).
Audétat 2013). All analyzed SMIs are rhyolitic (calculated SiO2 content
Similar to Mo, concentration ratios for Cu show important ranges from 71.2 to 76.5 wt%), K2O is strongly enriched and
variations and range from 0.05 to 2.6 × 10−4, except for one varies from 4.2 to 7.9 wt%, Fe2O3 varies from 0.2 to 1.6 wt%,
FIA from an HT1px vein showing a higher concentration ratio and CaO, MgO, TiO2, and MnO are always below 0.5 wt%.
of 0.24 (Fig. 11); no systematic variations according to salinity The Na2O content of the SMIs varies from 2.8 to 4.1 wt%
or homogenization temperature are observed. A general de- except for SMIs hosted in HT2 veins where Na2O is strongly
crease of Cu concentration ratios in FIAs is recognized from depleted by < 1.5 wt% (Fig. 13). Such Na2O depletion is due
HT1 to HT2 veins. Obtained Cu concentration ratios in CO2- probably to post-entrapment loss of Na (Zajacz et al. 2009),
rich high-density vapor inclusions from HT1 veins are lower possibly related to a slower cooling rate of the SMIs hosted in
by one order of magnitude or more, compared to published HT2 veins than the one hosted in HT1 veins, which is also
data for intermediate-density fluid inclusion compositions suggested by the fact that the SMIs tend to be coarsely crys-
from other Cu-Mo and Mo porphyry systems; in contrast, tallized (Bodnar and Student 2006).
Zn concentration ratios are higher by up to one order of mag- The strong incompatible behavior of Cs was used as a
nitude (Ulrich et al. 2002; Landtwing et al. 2005, 2010; proxy for the degree of crystallization and fractionation of
Klemm et al. 2007, 2008; Stefanova et al. 2014). the residual melt (Audétat et al. 2008). Silicate melt inclusions
 Miner Deposita

Fig. 12 a–d Glassy silicate melt inclusions hosted in quartz from HT1bt or HT1px veins. e, f Crystallized silicate melt inclusions hosted in quartz from
HT2 veins

show an increase in incompatible element concentrations like attributed to post-entrapment diffusion, whereas Zn and Pb
Rb, W, Mo, and Sb and a decrease in compatible element show no correlation with Cs and their concentrations are
concentrations like Sr with the increase in the fractionation roughly constant (Fig. 13). Tungsten, Sb, and Mo are enriched
degree of the trapped melt (marked by the increase of Cs compared to the whole-rock data and show a weak positive
content). Similar evolution trends were reported in SMIs from correlation with Cs (Fig. 13); W varies from 0.10 to 290 ppm,
miarolitic cavities (Zajacz et al. 2008) and from porphyry Sb from 0.26 to 195 ppm, and Mo from 1 to 35 ppm. As for
veins in the Elatsite porphyry Cu-Au-PGE deposit Cu, Zn, and Pb, these values are relatively similar to the ones
(Stefanova et al. 2014). Base metal concentrations in SMIs from miarolitic cavities from granitic systems (Zajacz et al.
range from 20 to 200 ppm (with three values above 2008). Zinc, Pb, Cu, W, and Sb are enriched compared to
300 ppm) for Zn, from 0.7 to 84 ppm (with two values above
200 ppm) for Pb, and from 1.3 to 194 ppm for Cu. Few SMIs Fig. 13 Plot of selected elements or element ratios against Cs„
from HT1bt veins carry higher Cu values (up to 800 ppm Cu; concentration obtained from LA-ICP-MS analyses of individual SMI
Fig. 13); these values are likely due to post-entrapment Cu and whole-rock LA-ICP-MS data. Black squares represent whole-rock
diffusion through quartz (Zajacz et al. 2009; Rottier et al. analyses (LA-ICP-MS) of dacite and rhyodacite domes and quartz–
monzonite dikes from Cerro de Pasco (data from Rottier et al., 2016a).
2017). Base metal concentrations of the SMIs are slightly Full red, purple, and blue circles represent SMIs hosted in HT1bt, HT1px,
enriched compared to the whole-rock analyses of dacite and and HT2 veins, respectively. Gray diamonds are SMIs hosted in quartz
rhyodacite domes and quartz monzonite dikes but relatively from miarolitic cavities from various barren granitic systems from Zajacz
similar to base metal concentrations of SMIs hosted in et al. (2008). Gray triangles and empty gray reversed triangles are,
respectively, SMIs homogenized in an autoclave and SMIs not
miarolitic cavities from barren granitic systems (Zajacz et al. homogenized prior to LA-ICP-MS analyses, hosted in quartz from
2008; Fig. 13). Copper content of the analyzed SMIs presents porphyry-type veins from the Elatsite Cu-Au-PGE porphyry deposit
a weak positive correlation with Cs discarding the high values (Bulgaria); data are from Stefanova et al. (2014)
Miner Deposita
 Miner Deposita

SMIs hosted in quartz veins from Elatsite (Stefanova et al. ranging from 640 to 680 °C (Fig. 14); (ii) the Zr in rutile
2014; Fig. 13). Elevated Ag concentrations from 0.2 to thermobarometer gives temperatures of formation between
11.7 ppm were detected in 15 SMIs. 700 and 780 °C for HT1bt veins, between 680 and 840 °C for
HT1px, and between 560 and 605 °C for HT2 veins (Fig. 14);
and (iii) the Zr in titanite thermobarometer gives temperatures
Discussion of formation of 750 °C for HT1bt veins and between 710 and
750 °C for HT1px (Fig. 14). We obtained similar temperatures
Comparison of the mineral thermobarometers using the Zr in rutile and the Zr in titanite thermobarometers
(Fig. 14), whereas the temperatures obtained with Ti in quartz
The effect of pressure on the Zr in rutile and Zr in titanite thermobarometers are up to 150 °C lower (and up to 100 °C
thermobarometers is small (~ 5 °C for 1 kbar and ~ 10 °C higher for sample BR-312; Fig. 14). Such differences could
for 1 kbar, respectively; ESM Fig. A.4; Tomkins et al. 2007; indicate that (i) quartz grows in disequilibrium with the fluid
Hayden et al. 2008). This effect is more important for the Ti in or (ii) quartz growth rate during vein formation was lower (or
quartz thermobarometer (~ 70 °C for 1 kbar; ESM Fig. A.4; higher for sample BR-312) than that used for the calibration of
Huang and Audétat 2012). the thermobarometer (2–4 μm/day; Huang and Audétat 2012).
The Zr in rutile and the Zr in titanite thermobarometers are
considered to be robust due to the weak effect of pressure Pressure–temperature reconstruction
estimates, the very low diffusion rate of Zr in rutile and titanite,
and their independence relative to mineral growth rate HT1 veins
(Tomkins et al. 2007; Hayden et al. 2008), when compared
to that of Ti in quartz. The Zr in rutile thermobarometer was Isochores for the CO2-rich high-density vapor inclusions hosted
successfully used to reconstruct vein formation temperatures in in HT1bt and HT1px veins are similar (Fig. 15). Two to three
the Butte porphyry Cu-Mo deposit (Mercer and Reed 2013). isochores were constructed per FIA to illustrate the existing
To estimate formation temperatures, pressure of formation of range inside one FIA. For both HT1bt and HT1px veins, two
the HT1bt, HT1px, and HT2 veins was fixed at 800 bar, the distinct families of isochores could be discerned: (i) Bhigh-
estimated maximum emplacement pressure according to the angle^ isochores and (ii) Blow-angle^ isochores (Fig. 15). The
diatreme architecture and the interpreted erosion preceding the slope of the isochores is controlled by the density of the fluid
formation of the diatreme breccia. At this pressure, (i) the Ti in inclusions. The high-angle isochores are constructed from fluid
quartz thermobarometer yields temperatures of formation be- inclusions with ~ 45% liquid (FIA nos. 311C_1 and 311D_5 for
tween 550 and 730 °C for HT1bt veins, between 505 and HT1bt veins; FIA no. 300E_9 for HT1px veins; Figs. 9a, b and
630 °C for HT1px veins, and between 370 and 545 °C for 15) whereas the low-angle isochores are from fluid inclusions
HT2 veins except sample BR-312 that returns higher values with 10 to 30% of liquid (Table 2). The existence in both vein

Fig. 14 Box plots of isobaric temperature, obtained at pressure fixed at samples analyzed. Box height represents the interquartile range (i.e.,
0.9 kbar for HT1bt veins, at 1.3 kbar for HT1px, and 0.85 kbar for HT2 representing the most frequent 50% of the distribution). The black lines
veins, using the Ti in quartz thermobarometer of Huang and Audétat dividing the boxes are the mean while the black dots represent the median
(2012) indicated by Qz at the bottom of each column, the Zr in rutile value. BWhiskers^ of each box extend to the extreme values lying within
thermobarometer of Tomkins et al. (2007) indicated by Rt at the bottom of 1.5 times the interquartile range beyond the edges of the box. The empty
each column, and the Zr in titanite thermobarometer of Hayden et al. circles and triangles represent outliers deviating more than ± 2.7σ. Full
(2008) indicated by Tnt at the bottom of each column, for the different circles represent isolated values
Miner Deposita

water loss, resulting in density decrease of the inclusions and

changes in the slope of the isochores. The fact that salinity is
relatively constant inside a single FIA discards the hypothesis of
heterogeneous entrapment. Moreover, no evidence of reopening
has been observed on the SEM-CL images; thus, we favor the
post-entrapment modification hypothesis.
The high-angle isochores never cross the liquid–vapor immis-
cibility field in the H2O-NaCl system (Driesner 2007). Formation
pressures were estimated using the intersection between Zr in
rutile isopleths and the isochores. The obtained pressures are
between 1800 and 2300 bar for HT1bt veins and between 1500
and 2750 bar for HT1px veins. These pressures are in a similar
range than the ones obtained for the Butte porphyry Cu-Mo
deposit by Rusk et al. (2008a) on similar types of fluid inclusions.
The high pressure would imply an emplacement depth between
5.5 and 10 km, which is significantly higher than typical em-
placement depth (2 to 5 km) reported for porphyry deposits
(Lecumberri-Sanchez et al., 2015) and, at Cerro de Pasco, than
the estimated emplacement depth of the porphyry veins based on
field evidence. The formation of the porphyry mineralization
(15.56 ± 0.1 Ma, Re-Os; Rottier 2017; Table 1) only slightly
predates the diatreme breccia emplacement (15.36 ± 0.03 Ma;
Baumgartner et al. 2009). The diatreme breccia hosting the min-
eralized clasts has been recognized to a depth of 1150 m below
present surface. Post-diatreme erosion is assumed to be < 1 km
(Baumgartner 2007; Baumgartner et al. 2008). This vertical ex-
Fig. 15 Pressure–temperature diagram. The red and burgundy lines
tent is in line with the generally assumed vertical extent of
represent isochores calculated for CO 2 -rich high-density vapor
inclusions in HT1 bt veins (a) and HT1px veins (b); two or three diatreme breccia bodies formed during large phreatomagmatic
isochores were constructed per FIAs. Red lines are the Bhigh-angle^ eruptions that is considered to not exceed 2 km (White and
and burgundy lines the Blow-angle^ isochore families. Colored thick Ross, 2011; Valentine and White, 2012). If this vertical extent
lines and shaded fields are Zr isopleths of rutile, Zr isopleths of titanite,
of the diatreme breccia is realistic, it means that at the time of the
and Ti isopleths of quartz from HT1bt veins (a) and HT1px veins (b). Solid
lines represent the mean value, and shaded fields represent the mean ± 1σ breccia formation, the porphyry veins were at a maximum depth
range of values of Zr concentration of rutile and titanite and Ti close to 2 km. Time gap between the mineralization and the
concentration of quartz from HT1bt veins (a) and HT1px veins (b). formation of the diatreme breccia is 0.23 ± 0.15 My, and during
Dashed field is the liquid–vapor immiscibility field in the H2O-NaCl
this time, erosion of material could be assumed to be also less
system for fluids having salinities of 2, 5, 10, and 15 wt%
than 1 km. Such eroded thickness would require an erosion rate
of ≈ 3 mm/year. Therefore, emplacement depth of the porphyry
types of two isochore families yielding distinctly different pres- veins should be between 2 and 3 km (between 500 and 800 bar,
sure estimations is difficult to interpret. To better constrain trap- assuming a lithostatic regime). The emplacement depth between
ping conditions of the CO2-rich high-density vapor inclusions, 5.5 and 10 km suggested by the high-angle isochores would
we have included, in Fig. 15, the liquid–vapor immiscibility require an unrealistic erosion of > 3.5 km of material between
field of the H2O-NaCl system for fluids having salinities of 2, the formation of the porphyry mineralization and the
5, 10, and 15 wt% (Fig. 15; Driesner 2007). Addition of CO2 to phreatomagmatic eruption. Such a high erosion rate is not sup-
the H2O-NaCl system would expand the solvus to higher pres- ported by the existence of related Late Miocene detrital units in
sures, and thus, the immiscibility fields can be considered as the area or by other geological evidence.
minimum pressure estimates. The low-angle isochores con- The incoherent pressure estimations using the high-angle
structed for HT1bt and HT1px veins cross this solvus at temper- isochores could be due to (i) the H2O-NaCl-CO2 system not
atures close to or below 450 °C, which is in contradiction with approximating the fluid trapping conditions, (ii) post-
the temperature obtained with the three independent mineral entrapment modification of the fluid inclusions leading to a
thermobarometers and also with the sinuous shape of the veins. density overestimation, and (iii) local overpressure. Local flu-
This discrepancy could be due to (i) heterogeneous trapping of id overpressure can occur in magmatic–hydrothermal systems
vapors and brines, (ii) late trapping of these fluid inclusions, and where sealed fracture networks reach the magmatic chamber
(iii) post-entrapment modification of the inclusions, such as underneath, (P ≈ magmatic pressure; Carrigan et al. 1992).
 Miner Deposita

HT2 veins rich MVPs. The CO2-rich high-density vapor inclusions in

HT1bt and HT1px veins probably result from fluid exsolution
Homogenization of the polyphase brine inclusions hosted in of CO2- and sulfur-rich mafic magmas injected into the un-
HT2 veins was not reached at 580 °C, and homogenization derlying magmatic chamber as recorded in other porphyry
temperatures are estimated to be slightly above 600 °C. Some deposits (Hattori and Keith 2001; Maughan et al., 2002;
polyphase brine FIAs are spatially associated with low- Halter et al. 2005, Buret et al. 2016). Mixing of the CO2-
density vapor inclusions and interpreted as boiling assem- and sulfur-rich fluids with more Bclassical^ porphyry fluids
blages. Thus, temperatures, densities, and pressures obtained might occur, considering that introduction of CO2-rich fluids
reflect the true trapping conditions. The Zr in rutile thermom- decreased the activity of H2O in the felsic magma and thus
eter applied to HT2 veins yields temperatures between 560 likely promoted its crystallization by elevating solidus and
and 605 °C, in good agreement with homogenization temper- liquidus temperatures (Lowenstern 2001). Such mixing could
atures of polyphase brine FIAs. Pressures of entrapment for explain the measured high Cs and Rb concentration ratios in
the polyphase brine inclusions estimated from phase relations the CO2-rich high-density vapor inclusions that are not coher-
in the H2O-NaCl system (Driesner and Heinrich 2007) are ent with fluids exsolved from mafic magma and which, taking
between 710 and 850 bar, assuming a homogenization tem- into account the fluid\melt partition coefficients for these ele-
perature of 600 °C (Table 2). This estimation should be taken ments (Zajacz et al. 2008), require an evolved parental melt
with caution considering that polyphase brine inclusions are (dacitic to rhyolitic in composition). The relatively wide salin-
dominated by Fe and K instead of Na (Fig. 11 and ESM Fig. ity ranges of CO2-rich high-density vapor inclusions (1.8 to
A.3) and that the minor presence of CO2 has been also recog- 12.4 wt% NaCl equiv. in HT1bt veins and from 11.5 to
nized (ESM Fig. A.2). On the other hand, the estimation of 16.6 wt% NaCl equiv. in HT1px veins) could be an indication
710 and 850 bar is in the same range as the pressure estimated of such a mixing or could be due to minor phase separation
from the architecture of the diatreme breccia. during fluid ascent.
Concentration ratios of X/(Na + K + Fe), X being Na, K, Fe,
Low-temperature fluids Zn, and Cs, are similar in coexisting vapor and brine fluid
inclusions from various porphyry copper deposits (Ulrich
Calculated pressures, using the SoWat software (Driesner et al. 2002; Landtwing et al. 2005, 2010; Klemm et al. 2007;
2007; Driesner and Heinrich 2007), for the low-temperature Stefanova et al. 2014), suggesting that phase separation has no
brines in secondary trails homogenized by halite disappear- impact on the ratios, i.e., that elemental ratios of the parental
ance are higher than 1700 bar. Such high pressures are not fluid are preserved. This observation is coherent with the fact
compatible with the pressure estimations developed above. that vapor–liquid partition coefficients are almost similar for
Such inclusions might probably result from a heterogeneous Na, K, Fe, and Zn (Kouzmanov and Pokrovski 2012 and
entrapment of halite-saturated fluid, as proposed by references therein; Zajacz et al., 2017); no experimental data
Lecumberri-Sanchez et al. (2015). The reconstructed are available for the other elements. Therefore, the fact that at
isochores, for aqueous and brine inclusions homogenized by Cerro de Pasco, the concentration ratios X/(Na + K + Fe) for
vapor disappearance, are steep, and assuming a similar entrap- Na, K, Fe, Zn, and Cs in CO2-rich high-density vapor inclu-
ment pressure as we used for HT1 and HT2 veins (~ 800 bar), sions from the HT1 veins differ significantly from those in
the obtained entrapment temperatures are between 270 and polyphase brine inclusions in the lower-temperature (~
320 °C. 600 °C) HT2 veins (Fig. 11 and ESM Fig. A.3) may indicate
that they represent different fluids. The lower Cs/(Na + K +
Fluid evolution Fe) ratios of the the polyphase brine inclusions suggest that
fluid trapped in HT2 veins is exsolved from a less differenti-
Occurrence of CO2-rich high-density vapor inclusions as ated magma, which is more mafic than the one that exsolved
those observed in HT1bt and HT1px veins is not common in the fluid forming the HT1 veins (Audétat et al. 2008). Such an
porphyry Cu-(Mo)-(Au) deposits (Kouzmanov and Pokrovski increase of the proportion of fluids exsolved from mafic
2012). These CO2-rich inclusions are also strongly enriched in magmas into the mineralizing fluids is also coherent with the
sulfur as indicated by the presence of anhydrite daughter crys- increase of Fe/(Na + K + Fe) ratios and decrease of Zn/(Na +
tals and the detectable amount of SO42− ions in the fluid phase K + Fe) ratios in polyphase brine inclusions (Fig. 11).
of the inclusions (Fig. 9b and ESM Fig. A.2). CO2 and sulfur The scattered salinities and concentration ratios and the
strongly partition into the magmatic volatile phase (MVP) strong enrichment in Sr, Ba, and Ca of the low-temperature
during magmatic fluid exsolution; if magmas reach fluid sat- brine and aqueous secondary inclusions are typical for fluid
uration deep in the crust, CO2-rich MVP exsolve and the re- linked to feldspar destabilization during phyllic alteration. We
sidual magma is depleted in CO2 (Lowenstern, 2001). Thus, cannot totally discard the possibility that these secondary low-
evolved, dacitic magmas unlikely exsolve CO2- and sulfur- temperature fluid inclusions are trapped in quartz after the
Miner Deposita

emplacement of the diatreme–breccia, but no similar inclu- the Pasto Bueno tungsten–polymetallic deposit (up to 10‰:
sions have been observed in quartz phenocrysts in the Landis and Rye 1974), located 200 km north of Cerro de
diatreme–breccia surrounding the mineralized clasts, suggest- Pasco. Such heavy signatures for magmatic fluids and there-
ing that the low-temperature brine and aqueous secondary fore for the magma itself could reflect significant assimilation
fluid inclusions were trapped before the breccia formation. of crustal rocks enriched in δ18O (Bindeman 2008; Auer et al.
They would, thus, record the cooling of the hydrothermal 2009). Strong contamination by old continental crust of the
system at depth, as also suggested by the occurrence of pre- magmas at Cerro de Pasco has been evidenced by the Hf
diatreme polymetallic mineralization in reopened HT1 and isotope signatures of zircons with εHf ranging from − 1 to −
HT2 veins. 4 (Baumgartner et al. 2009), a typical feature of Miocene
magmatism in central Peru (Bissig and Tosdal 2009).
Origin of fluids: constraints from oxygen isotope The homogeneous δ18O values of the fluid in equilibrium
composition of hydrothermal quartz with quartz from HT1bt, HT1px, and HT2 veins and also with
quartz associated to the pre-diatreme polymetallic mineraliza-
The δ18O values of the fluid in equilibrium with the different tion points towards a similar δ18O composition of the magmas
quartz generations were calculated using high- and low- from which the fluids are exsolved. The high δ18O values of
temperature fractionation curves of Matsuhisa et al. (1979) as- the mineralizing fluids exclude significant input of meteoric
suming a temperature of 700 °C for HT1bt and HT1px veins and water in the system, even when the latter cooled down to ~
600 °C for HT2 vein, based on the mineral thermobarometers 300 °C.
and the fluid inclusion microthermometry (Fig. 16). The tem-
perature of the pre-diatreme polymetallic mineralization has Significance and origin of silicate melt inclusions
been fixed at 300 °C based on the sulfide assemblage (Fig. 16). in hydrothermal veins
The δ18OFluid values are 11.0‰ (n = 4, σ = ± 0.3) for the
HT1bt vein, 11.1‰ (n = 7, σ = ± 0.3) for the HT1px vein, No spatial relationship has been observed between the SMIs
10.9‰ (n = 7, σ = ± 0.20) for the HT2 vein, and 10.2‰ (n = and the CO2-rich high-density vapor inclusions from the HT1
2, σ = ± 0.4) for the pre-diatreme polymetallic mineralization veins or with the polyphase brine inclusions from the HT2
(Fig. 16). The δ18O signature of fluids in equilibrium with the veins. In the absence of a genetic link, our data cannot be used
different quartz generations (around 11‰) is slightly shifted to determine fluid/melt partition coefficients. The SMIs are
towards heavier compositions compared to the typical signa- hosted in hydrothermal quartz, and therefore, the silicate melt
ture for primary magmatic fluids (6–9‰; Taylor 1997). droplets forming the inclusions were transported by the high-
Similar enriched compositions were also calculated for mag- temperature hydrothermal fluids. The strong variability in
matic fluids forming porphyry Cu-Mo mineralization in the composition of SMIs hosted in the same vein and occasionally
Morococha district (up to 12‰; Catchpole et al. 2015a) locat- in the same quartz crystal indicates that SMIs do not represent
ed 100 km south of Cerro de Pasco and also for fluids forming the parental magma from which the fluids have been exsolved.

Fig. 16 δO18 SIMS analyses of

quartz from HT1bt, HT1px, and
HT2 veins and from the late
polymetallic mineralization and
the calculated δO18 values of the
fluid in equilibrium with quartz
using the equation of Matsuhisa
et al. (1979) at temperatures of
700 °C for HT1bt and HT1px
veins, 600 °C for HT2 veins
based on microthermometry and
results obtained from the
thermobarometers, and 300 °C for
pre-diatreme polymetallic
mineralization based on the
sulfide assemblage
 Miner Deposita

It also indicates that SMIs did not re-equilibrate with the fluid could be due to minor crystallization of Sr-rich phases, such as
during their transport. An alternative hypothesis would be that apatite, not taken into account in the model.
the fluids transporting the silicate melt droplets also strongly No systematic chemical evolution between SMIs hosted in
vary in composition, a fact that is not supported by the FIAs. HT1bt, in HT1px, and in HT2 veins was observed; moreover,
Moreover, the chemical evolution drawn by the SMIs is the entire range of Cs concentrations (from 10 to 100 ppm) is
marked by an enrichment of elements incompatible in igneous observed in almost all veins. It has been even observed that the
minerals, such as Rb, and a depletion of compatible elements, SMIs within one individual quartz crystal yield Cs contents dif-
such as Sr, suggesting that the chemical variations observed in fering up to 30 ppm. These results suggest that there is no corre-
the SMIs are controlled by igneous processes. lation between the relative timing of SMI entrapment and the
SMIs trapped in vein quartz show a large range of Cs con- initial silicate melt composition. Thus, we interpret the investi-
tents, from 3 to 102 ppm. This could be interpreted as entrap- gated SMIs as melt droplets sampled at various isolated, not
ment of melt droplets at different degrees of crystallization and solidified parts of the underlying magma chamber (like miarolitic
differentiation (Audétat et al. 2008; Audétat and Lowenstern cavities) by the ascending high-temperature magmatic hydrother-
2014). Cesium, Sr, and Rb concentrations in SMIs have been mal fluids. This hypothesis could explain the absence of system-
reconstructed using a simple crystallization model of an atic patterns in the chemical composition of the SMIs. No direct
evolved felsic melt in a closed system (Fig. 17). Crystal pro- relationship could be established between the SMIs and the mag-
portions chosen for this model are 50% plagioclase, 35% K- ma from which the mineralizing fluids were exsolved, and there-
feldspar, and 15% quartz, in agreement with products crystal- fore, they cannot be used to evaluate the efficiency of ore-
lizing from a granitic melt (Whitney 1988); the starting com- forming processes. Nonetheless, these SMIs are direct indication
position was fixed at 3.7 ppm Cs, 90 ppm Rb, and 400 ppm Sr, of the occurrence of heterogeneously crystallized silicic plutonic
based on composition of Andean rhyolites (http://georoc. bodies at depth, at the time of magmatic fluid exsolution.
mpch-mainz.gwdg.de/georoc/).
The concentration of an element X in the residual melt was Cerro de Pasco mineralizing system
calculated assuming an equilibrium fractional crystallization
using the equation CL = Ci/[D − F(1 − D)], where CL is the This study, combined with other results on the porphyry-type
concentration of the element of interest in the residual melt, and polymetallic epithermal mineralization at Cerro de Pasco
Ci is the initial concentration of the element of interest, D is the (Baumgartner et al. 2009, Rottier et al. 2016a, 2016b, Rottier
bulk partition coefficient of the element between minerals and 2017), confirms that the large epithermal polymetallic deposit
melt, and F is the amount of remaining magmatic liquid. Bulk of Cerro de Pasco formed in the shallow part of a polycyclic
partition coefficient was fixed at 0.026 for Cs, 0.214 for Rb, porphyry system, as observed in other districts such as
and 8.12 for Sr, based on partition coefficient data compiled in Morococha (Catchpole et al. 2015a, 2015b). It also reveals a
the GERM database (http://earthref.org/GERM/). The model complex history of the Cerro de Pasco magmatic–hydrother-
is able to reproduce the compositional evolution recorded in mal system with at least three temporally distinct porphyry-
the SMIs. Strontium concentrations of the SMIs are more type and one polymetallic mineralization events before the
depleted than those estimated by the model; this discrepancy formation of the diatreme breccia and the large epithermal

Fig. 17 a Rb vs Cs and b Sr vs Cs diagrams of the SMIs analyzed by LA-ICP-MS. Empty ellipses represent the results of the crystallization model in
closed system of an evolved felsic melt
Miner Deposita

polymetallic deposit. Fluid inclusion and oxygen stable iso- similar Pb/Zn ratios found at Cerro de Pasco do not imply a
tope studies on the large epithermal polymetallic deposit point direct genetic link but may indicate that fluids forming the
to involvement of high-salinity magmatic fluids (Baumgartner epithermal polymetallic mineralization and those recorded in
et al. 2008; Rottier 2017). the different porphyry-type mineralization events share com-
Kouzmanov and Pokrovski (2012) showed that Pb/Zn ratios mon features, potentially a common source. It is also interesting
measured in fluid inclusions are constant inside a single por- to note that the Pb/Zn ratios from the whole-rock analyses of
phyry system, whereas they vary from 1 to 6 over different dacite and rhyolite domes, and quartz–monzonite dikes are also
systems (Fig. 11 in Kouzmanov and Pokrovski 2012). This similar (Fig. 18). Zinc and Pb have similar partition coefficients
ratio is not affected by phase separation processes and thus between melt and hydrothermal fluids (Zajacz et al. 2008).
reflects the original ratio of the silicate melt at the timing of Therefore, taking into account the similar Pb/Zn ratios found
fluid exsolution (Kouzmanov and Pokrovski 2012). At Cerro in the magmatic rocks, in the polyphase brines from the differ-
de Pasco, the Pb/Zn ratio of the bulk ore from the epithermal ent porphyry-type mineralization events, and in the bulk ore of
polymetallic mineralization is similar to the Pb/Zn ratios of the the shallow large epithermal mineralization, a common source
polyphase brine inclusions found in (i) HT2 veins (studied from a similar deep magmatic reservoir can be assumed.
here), (ii) quartz–magnetite–chalcopyrite–pyrite porphyry-
style veinlets (PM2; Rottier et al. 2016a), and (iii) the quartz–
magnetite–chalcopyrite veins crosscutting clasts found in the Summary and conclusions
quartz monzonite dikes (PM3; Rottier 2017; Fig. 18). The
Our detailed study of porphyry-type quartz veins in hornfels and
porphyritic igneous clasts found in the phreatomagmatic
diatreme breccia allows reconstructing the different hydrother-
mal events at depth, occurring prior to and below the large
epithermal polymetallic mineralization of Cerro de Pasco.
Oxygen isotope data indicate a magmatic origin of the fluids
forming the quartz in the porphyry-style HT1bt, HT1px, and
HT2 veins and also that in the pre-diatreme polymetallic miner-
alization. Application of multiple mineral thermobarometers and
estimation based on the diatreme architecture indicate that HT1bt
and HT1px quartz–molybdenite–(chalcopyrite)–(pyrite) veins
were formed at temperatures above 700 °C at pressures from
500 to 800 bar. Fluids, forming the HT1 veins, are CO2-rich
(XCO2 between 0.1 and 0.4) and S-rich and are generated dur-
ing fluid exsolution from a mafic magma injected in the felsic
magma chambers. These fluids have probably partly mixed with
fluids exsolved from the felsic magma as well. HT2 quartz–
molybdenite–(chalcopyrite)–(pyrite) veins formed at lower tem-
perature (~ 600 °C) and at similar pressures. Cesium concentra-
tions of the polyphase brine inclusions indicate that fluids
forming the HT2 veins are less evolved than the ones forming
HT1 veins, suggesting a more important contribution of fluids
exsolved from the mafic magma. Fluid inclusions trapped in
both HT1 and HT2 veins show lower Cu but similar Mo and
Fig. 18 Cu-Zn-Pb ternary diagram. Red and purple full circles are mean
higher Zn concentration ratios ([X]/[Na + K + Fe]) compared to
values of CO2-rich high-density vapor FIAs, respectively, from HT1bt various porphyry Cu-Mo deposits (Ulrich et al. 2002;
veins and HT1px veins. Blue circles represent mean values of polyphase Landtwing et al. 2005, 2010; Klemm et al. 2007, 2008;
brine FIAs from HT2 veins. Empty red, purple, and blue circle represent Stefanova et al. 2014). The SMIs found in hydrothermal quartz
SMI compositions in HT1bt, HT1px, and HT2 veins, respectively. Light
and dark gray full circles represent mean values polyphase brine FIAs,
from HT1bt, HT1px, and HT2 veins represent droplets of melts
respectively, from the quartz–magnetite–chalcopyrite–(pyrite) porphyry- from various isolated and not solidified parts of the deeply
style veinlets (PM2; Rottier et al., 2016a) and from the quartz–magnetite– seated magma chamber sampled by the high-temperature mag-
chalcopyrite. Black squares represent LA-ICP-MS whole-rock analyses matic fluids during their ascent. Cooling of the system down to
of dacitic and rhyodacite domes and quartz–monzonite dikes from Cerro
de Pasco (Rottier et al., 2016a). Blue star represents the Cu-Pb-Zn ratio
300 °C prior to the formation of the diatreme breccia is indicated
calculated for the bulk ore of the epithermal polymetallic deposit using by the occurrence of polymetallic mineralization in veins re-
values published by Baumgartner et al. (2008) stricted to the clasts and also by the P-T-X characteristics of
 Miner Deposita

the low-temperature brines and aqueous fluid inclusions. The disappearance from porphyry copper and other hydrothermal ore
deposits. Econ Geol 103:539–554
latter exhibits signatures typical of fluids associated to
Bindeman I (2008) Oxygen isotopes in mantle and crustal magmas as re-
feldspar-destructive phyllic alteration. vealed by single crystal analysis. Rev Mineral Geochem 69:445–478
The results demonstrate the existence of porphyry minerali- Bissig T, Ullrich TD, Tosdal RM, Friedman R, Ebert S (2008) The time-
zation and a low-temperature polymetallic stage at a depth that space distribution of Eocene to Miocene magmatism in the central
Peruvian polymetallic province and its metallogenetic implications.
predates the diatreme and the formation of the large shallow
J S Am Earth Sci 26:16–35
epithermal polymetallic Cerro de Pasco deposit. The Pb/Zn Bissig T, Tosdal RM (2009) Petrogenetic and metallogenetic relationships in
ratios of the studied hypersaline fluid inclusions from the dif- the Eastern Cordillera Occidental of Central Peru. J Geol 117:499–518
ferent porphyry-type events are similar and overlap the bulk Pb/ Bodnar RJ, Student JJ (2006) Melt inclusions in plutonic rocks: petrog-
raphy and microthermometry. In: Webster JD (ed) Melt Inclusions in
Zn ratio of the large shallow polymetallic mineralization and
Plutonic Rocks Mineral. Assoc. Canada, Short Course 36: 1–26
those from the different magmatic rocks. This similarity sug- Buret Y, Von Quadt A, Heinrich C, Selby D, Wälle M, Peytcheva I (2016)
gests a common deep magmatic reservoir of the fluids respon- From a long-lived upper-crustal magma chamber to rapid porphyry
sible for the different mineralization styles at Cerro de Pasco. copper emplacement: reading the geochemistry of zircon crystals at
Bajo de la Alumbrera (NW Argentina). Earth Planet Sci Lett 450:
120–131
Acknowledgements The present investigation was carried out with the
Carrigan CR, Schubert G, Eichelberger J (1992) Thermal and dynamical
support of the Swiss National Science Foundation (grant 200020_
regimes of single-and two-phase magmatic flow in dikes. J Geophys
134872). Funding for the SwissSIMS facility was provided by the KIP
Res 97:377–392
6 BCASA^ to LPB. We thank Volcan Companía Minera S.A. for provid-
Catchpole H, Kouzmanov K, Fontboté L, Guillong M, Heinrich CA
ing logistic support and geological data. We gratefully acknowledge Jean-
(2011) Fluid evolution in zoned Cordilleran polymetallic veins—
Marie Boccard and Fabio Capponi for their help with sample preparation
insights from microthermometry and LA-ICP-MS of fluid inclu-
and XRF analyses. We are grateful to Celestine Mercer, Zoltan Zajacz,
sions. Chem Geol 281:293–304
associate editor Frank Melcher, and an anonymous reviewer for their
Catchpole H, Kouzmanov K, Putlitz B, Seo J, Fontboté L (2015a) Zoned
constructive comments which significantly improved the manuscript.
base metal mineralization in a porphyry system: origin and evolution
of mineralizing fluids in the Morococha District, Peru. Econ Geol
110:39–71
References Catchpole H, Kouzmanov K, Bendezú A, Ovtcharova M, Spikings R,
Stein H, Fontboté L (2015b) Timing of porphyry (Cu-Mo) and base
metal mineralisation (Zn-Pb-Ag-Cu) in a magmatic-hydrothermal
Angeles C (1999) Los sedimentos Cenozoicos de Cerrode Pasco: system—Morococha district, Peru. Mineral Deposita 50:895–922
Estratigrafia, sedimentación y tectónica. In: Machare J, Benavides- Cherniak DJ, Watson EB, Wark DA (2007a) Ti diffusion in quartz. Chem
Caceres V, Rosas S (eds) Sociedad Geológica del Perú, vol. 5, pp Geol 236:65–74
103–118 Cherniak DJ, Manchester J, Watson EB (2007b) Zr and Hf diffusion in
Audétat A, Pettke T, Heinrich CA, Bodnar RJ (2008) The composition of rutile. Earth Planet Sci Lett 261:267–279
magmatic–hydrothermal fluids in barren and mineralized intrusions. Diamond LW (1992) Stability of CO2 clathrate hydrate + CO2 liquid +
Econ Geol 103:877–908 CO2 vapour + aqueous KCl-NaCl solutions: experimental determi-
Audétat A, Lowenstern JB (2014) Melt inclusions. In: Scott SD (ed) nation and application to salinity estimates of fluid inclusions.
Geochemistry of mineral deposits. Treatise on Geochemistry, vol Geochim Cosmochim Acta 56:273–280
13, 2nd edn. Elsevier, pp 143–173. https://doi.org/10.1016/B978- Driesner T (2007) The system H2O-NaCl. Part II: correlations for molar
0-08-095975-7.01106-2 volume, enthalpy, and isobaric heat capacity from 0 to 1000°C, 1 to
Auer S, Bindeman I, Wallace P, Ponomareva V, Portnyagin M (2009) The 5000bar, and 0 to 1 X-NaCl. Geochim Cosmochim Acta 71:4902–
origin of hydrous, high-δ18O voluminous volcanism: diverse oxy- 4919
gen isotope values and high magmatic water contents within the Driesner T, Heinrich CA (2007) The system H2O-NaCl. I. Correlations
volcanic record of Klyuchevskoy volcano, Kamchatka, Russia. for molar volume, enthalpy, and isobaric heat capacity from 0 to
Contrib Mineral Petr 157:209–230 1000 degrees C, 1 to 5000 bar, and 0 to 1 X-NaCl. Geochim
Bakker RJ, Brown PE (2003) Computer modelling in fluid inclusion Cosmochim Acta 71:4880–4901
research. In Samson I, Anderson A, Marshall D (eds) Fluid inclu- Duan Z, Møller N, Weare JH (1992) An equation of state for the CH4-
sions: analysis and interpretation. Mineralogical Association of CO2-H2O system: II. Mixtures from 50 to 1000°C and 0 to 1000 bar.
Canada, Short Course Series, 32: 175–204 Geochim Cosmochim Acta 56:2619–2631
Baumgartner R (2007) Sources and evolution in space and time of hy- Einaudi MT (1968) Pyrrhotite-pyrite-sphalerite relations at Cerro de
drothermal fluids at the Cerro de Pasco Cordilleran base metal de- Pasco, Peru. Unpublished Ph.D. thesis. Harvard University,
posit, Central Peru Terre & Environnement 66. https://doi.org/10. Cambridge, MA, p 381
13097/archiveouverte/unige:98205 Einaudi MT (1977) Environment of ore deposition at Cerro de Pasco,
Baumgartner R, Fontboté L, Vennemann T (2008) Mineral zoning and Peru. Econ Geol 72:893–924
geochemistry of epithermal polymetallic Zn-Pb-Ag-Cu-Bi mineral- Guillong M, Heinrich CA (2007) LA– ICP–MS analysis of inclusions:
ization at Cerro de Pasco, Peru. Econ Geol 103:493–537 Improved ablation and detection. Biennial Conference ECROFI
Baumgartner R, Fontboté L, Spikings R, Ovtcharova M, Schneider J, Pace XIX, Bern, Switzerland, 17–20 July, p 82
L, Gutjahr M (2009) Bracketing the age of magmatic-hydrothermal Guillong M, Meier DL, Allan MM, Heinrich CA, Yardley BWD (2008)
activity at the Cerro de Pasco epithermal polymetallic deposit, central SILLS: a Matlab-based program for the reduction of laser ablation
Peru: a U-Pb and 40Ar/39Ar study. Econ Geol 104:479–504 ICP-MS data of homogeneous materials and inclusions. In:
Becker SP, Fall A, Bodnar RJ (2008) Synthetic fluid inclusions. XVII. Sylvester P (ed) Laser-ablation-ICPMS in the earth sciences: current
PVTX properties of high salinity H2O–NaCl solutions (>30 wt% practices and outstanding issues. Mineralogical Association of
NaCl): application to fluid inclusions that homogenize by halite Canada, Vancouver, B.C., pp 328–333
Miner Deposita

Günther D, Frischknecht R, Heinrich CA, Kahlert HJ (1997) Capabilities Lecumberri-Sanchez P, Steele-MacInnis M, Weis P, Driesner T, Bodnar
of an argon fluoride 193 nm excimer laser for laser ablation induc- RJ (2015) Salt precipitation in magmatic-hydrothermal systems as-
tively coupled plasma mass spectrometry microanalysis of geologi- sociated with upper crustal plutons. Geology 43:1063–1066
cal materials. J Anal At Spectrom 12:939–944 Lerchbaumer L, Audétat A (2013) The metal content of silicate melts and
Günther D, Audétat A, Frischknecht R, Heinrich CA (1998) Quantitative aqueous fluids in subeconomically Mo mineralized granites: impli-
analysis of major, minor and trace elements in fluid inclusions using cations for porphyry Mo genesis. Econ Geol 108:987–1013
laser ablation–inductively coupled plasma mass spectrometry. J Lowenstern JB (2001) Carbon dioxide in magmas and implications for
Anal At Spectrom 13:263–270 hydrothermal systems. Mineral Deposita 36:490–502
Gustafson LB, Hunt JP (1975) The porphyry copper deposit at El Lund K, Aleinikoff JN, Kunk MJ, Unruh DM, Zeihen GD, Hodges WC,
Salvador, Chile. Econ Geol 70:857–912 du Bray EA, O’Neill JM (2002) SHRIMP U-Pb and 40Ar/39Ar age
Halter WE, Pettke T, Heinrich CA, Roten-Rutishauser B (2002) Major to constraints for relating plutonism and mineralization in the Boulder
trace element analysis of melt inclusions by laser-ablation ICP-MS: batholith region, Montana. Econ Geol 97:241–267
methods of quantification. Chem Geol 183:63–86 Masterman GJ, Cooke DR, Berry RF, Walshe JL, Lee AW, Clark AH
Halter WE, Heinrich CA, Pettke T (2005) Magma evolution and the (2005) Fluid chemistry, structural setting, and emplacement history
formation of porphyry Cu–Au ore fluids: evidence from silicate of the Rosario Cu-Mo porphyry and Cu-Ag-Au epithermal veins,
and sulfide melt inclusions. Min Dep 39:845–863 Collahuasi district, northern Chile. Econ Geol 100:835–862
Harris AC, Kamenetsky VS, White NC, van Achterbergh E, Ryan CG Matsuhisa Y, Goldsmith JR, Clayton RN (1979) Oxygen isotopic frac-
(2003) Melt inclusions in veins: linking magmas and porphyry Cu tionation in the system quartz-albite-anothite-water. Geochim
deposits. Science 302:2109–2111 Cosmochim Acta 43:1131–1140
Hattori KH, Keith J (2001) Contribution of mafic melt to porphyry copper Maughan DT, Keith JD, Christiansen EH, Pulsipher T, Hattori K, Evans
mineralization: Evidence from Mount Pinatubo, Philippines, and NJ (2002) Contributions from mafic alkaline magmas to the
Bingham Canyon, Utah, USA. Mineral Deposita 36:799–806 Bingham porphyry Cu–Au–Mo deposit, Utah, USA. Mineral
Hayden LA, Watson EB, Wark DA (2008) A thermobarometer for sphene Deposita 37:14–37
(titanite). Contrib Mineral Petr 155:529–540 Mavrogenes JA, Bodnar RJ (1994) Hydrogen movement into and out of
Heinrich CA, Pettke T, Halter WE, Aigner-Torres M, Audétat A, Günther fluid inclusions in quartz: experimental evidence and geologic im-
D, Hattendorf B, Bleiner D, Guillong M, Horn I (2003) Quantitative plications. Geochim Cosmochim Acta 58:141–148
multi-element analysis of minerals, fluid and melt inclusions by
Mercer CN, Reed MH (2013) Porphyry Cu-Mo stockwork formation by
laser-ablation inductively-coupled-plasma mass-spectrometry.
dynamic, transient hydrothermal pulses: Mineralogic insights from
Geochim Cosmochim Acta 67:3473–3496
the deposit at Butte, Montana. Econ Geol 108:1347–1377
Huang R, Audétat A (2012) The titanium-in-quartz (TitaniQ)
Mercer CN, Reed MH, Mercer CM (2015) Time scales of porphyry cu
thermobarometer: a critical examination and re-calibration.
deposit formation: insights from titanium diffusion in quartz. Econ
Geochim Cosmochim Acta 84:75–89
Geol 110:587–602
Ivascanu PM, Pettke T, Kouzmanov K, Heinrich CA, Pintea I, Rosu E,
Meyer C, Shea E, Goddard C (1968) Ore deposits at Butte, Montana. In:
Udubasa GH (2003) The magmatic to hydrothermal transition:
Ridge JD (ed) Ore deposits of the United States 1933–1967, 2.
Miocene Deva porphyry copper-gold deposit, South Apuseni Mts,
American Institute of Minerals, Metals, and Petroleum Engineers,
Romania. Society for Geology Applied to Mineral Deposits, SGA
New York, pp 1373–1416.
Biennial Meeting, 7th, Athens, Greece, 2003, Proceedings, pp 283–286
Klemm LM, Pettke T, Heinrich CH (2007) Hydrothermal evolution of the Müller D, Kaminski K, Uhlig S, Graupner T, Herzig PM, Hunt S (2002)
El Teniente deposit, Chile: porphyry Cu-Mo ore deposition from The transition from porphyry-to epithermal-style gold mineraliza-
low-salinity magmatic fluids. Econ Geol 102:1021–1045 tion at Ladolam, Lihir Island, Papua New Guinea: a reconnaissance
Klemm LM, Pettke T, Heinrich CA (2008) Fluid and source magma study. Mineral Deposita 37:61–74
evolution of the Questa porphyry Mo deposit, New Mexico, USA. Noble D, McKee EH (1999) The Miocene metallogenic belt of central
Mineral Deposita 43:533–552 and northern Peru. In: Skinner BJ (ed) Geology and ore deposits of
Kouzmanov K, Bendezú A, Catchpole H, Ageneau M, Pérez J, Fontboté L the central Andes, vol. 7. Society of Economic Geology Special
(2008) The Miocene Morococha district, central Peru-Large-scale Publication, pp. 155–193. https://doi.org/10.5382/SP.07.05
epithermal polymetallic overprint on multiple intrusion-centered por- Pettke T (2006) In-situ laser-ablation-ICP-MS chemical analysis of melt
phyry systems. In: PACRIM Congress 2008, The Pacific rim: mineral inclusion and prospects for constraining subduction zone magmas.
endowment, discoveries & exploration frontiers—extended abstracts; In: Webster JD (ed) Melt inclusions in plutonic rocks. Mineralogical
Gold Coast Queensland, November 24-26, 2008, pp 117–121 Association of Canada, short course series 36: 51–80
Kouzmanov K, Pokrovski GS (2012) Hydrothermal controls on metal Pintea I (1995) Fluid inclusions evidence for liquid magmatic immisci-
distribution in porphyry Cu (-Mo-Au) systems. Soc Econ Geol bility between hydrous salt melt and silicate melt as primary source
Special Publ 16:573–618 of ore metals in porphyry copper systems from Apuseni Mountains
Landis GP, Rye RO (1974) Geologic, fluid inclusion, and stable isotope (Romania). Bol Soc Esp Mineral 18:184–185
studies of the Pasto Bueno tungsten-base metal deposit, northern Pintea I (2014) The magmatic immiscibility between silicate-, brine-, and
Peru. Econ Geol 69:1025–1059 Fe-S-O melts from the porphyry (Cu-Au-Mo) deposits in the
Landtwing MR, Pettke T, Halter WH, Heinrich CA, Redmond PB, Carpathians (Romania): a review. Rom J Earth Sci 87:1–32
Einaudi MT, Kunze K (2005) Copper deposition during quartz dis- Reed M, Rusk B, Palandi J (2013) The Butte magmatic-hydrothermal
solution by cooling magmatic-hydrothermal fluids: the Bingham system: one fluid yields all alteration and veins. Econ Geol 108:
porphyry. Earth Planet Sci Lett 235:229–243 1379–1396
Landtwing MR, Furrer C, Redmond PB, Pettke T, Guillong M, Heinrich Rogers R (1983) Structural and geochemical evolution of a mineralized
CA (2010) The Bingham Canyon porphyry Cu-Mo-Au deposit. III. volcanic vent at Cerro de Pasco, Peru. Unpublished PhD thesis.
Zoned copper-gold ore deposition by magmatic vapor expansion. Tucson, University of Arizona, pp 116
Econ Geol 105:91–118 Rosso K, Bodnar R (1995) Microthermometric and Raman spectroscopic
Lecumberri-Sanchez P, Steele-MacInnis M, Bodnar RJ (2012) A numerical detection limits of CO2 in fluid inclusions and the Raman spectro-
model to estimate trapping conditions of fluid inclusions that homog- scopic characterization of CO2. Geochim Cosmochim Acta 59:
enize by halite disappearance. Geochim Cosmochim Acta 92:14–22 3961–3975
 Miner Deposita

Rosas S, Fontboté L, Tankard A (2007) Tectonic evolution and paleoge- rifting in Peru and implications for the early disassembly of western
ography of the Mesozoic Pucará Basin, central Peru. J S Am Earth Pangaea. Gondwana Res 35:124–143
Sci 26:16–35 Spry PG, Paredes MM, Foster F, Truckle JS, Chadwick TH (1996)
Rottier B (2017) Magmatic and hydrothermal fluid processes at the origin Evidence for a genetic link between gold-silver telluride and por-
of the giant porphyry-related epithermal polymetallic deposit of phyry molybdenum mineralization at the Golden Sunlight Deposit,
Cerro de Pasco (Central Peru), Phd thesis, University of Geneva, p Whitehall, Montana; fluid inclusion and stable isotope studies. Econ
434. https://doi.org/10.13097/archive-ouverte/unige:95569 Geol 91:507–526
Rottier B, Rezeau H, Casanova V, Kouzmanov K, Moritz R, Schlöglova Taylor HP (1997) Oxygen and hydrogen isotope relationships in hydro-
K, Wälle M, Fontboté L (2017) Trace element diffusion and incor- thermal mineral deposits. In: Barnes HL (ed) Geochemistry of hy-
poration in quartz during heating experiments. Contrib Mineral drothermal ore deposits, vol 3. John Wiley and Sons, Inc., New
Petrol 172:23 York, pp 229–302
Rottier B, Kouzmanov K, Bouvier AS, Baumgartner L, Wälle M, Rezeau Tomkins HS, Powell R, Ellis J (2007) The pressure dependence of the
H, Bendezú R, Fontboté L (2016a) Heterogeneous melt and hyper- zirconium-in-rutile thermometer. J Metamorph Geol 25:703–713
saline liquid inclusions in shallow porphyry type mineralization as Ulrich T, Günther D, Heinrich CA (2002) The evolution of a porphyry
markers of the magmatic-hydrothermal transition (Cerro de Pasco Cu-Au deposit, based on LA-ICP-MS analysis of fluid inclusions:
district, Peru). Chem Geol 447:93–116 Bajo de la Alumbrera, Argentina. Econ Geol 97:1888–1920
Rottier B, Kouzmanov K, Wälle M, Bendezú R, Fontboté L (2016b)
Valentine GA, White JDL (2012) Revised conceptual model for maar-
Sulfide replacement processes revealed by textural an LA-ICP-MS
diatremes: subsurface processes, energetics, and eruptive products.
trace element analyses: example from the early mineralization stages
Geology 40:1111–1114
at Cerro de Pasco, Peru. Econ Geol 111:1347–1367
Rusk BG, Reed MH, Dilles JH (2008a) Fluid inclusion evidence for Vikre PG, Graybeal FT, Koutz FR (2014) Concealed basalt-matrix
magmatic-hydrothermal fluid evolution in the porphyry copper- diatremes with Cu-Au-Ag-(Mo)-mineralized xenoliths, Santa Cruz
molybdenum deposit at Butte, Montana. Econ Geol 103:307–334 porphyry Cu-(Mo) system, Pinal County, Arizona. Econ Geol 109:
Rusk BG, Miller BJ, Reed MH (2008b) Fluid inclusion evidence for the 1271–1289
formation of Main stage polymetallic base-metal veins, Butte, White JDL, Ross P-S (2011) Maar-diatreme volcanoes: A review. J
Montana, USA. Ariz Geol Soc Dig 22:573–581 Volcanol Geotherm Res 201:1–29
Rusk BG, Lowers HA, Reed MH (2008c) Trace elements in hydrother- Whitney JA (1988) The origin of granite: the role and source of water in
mal quartz: relationships to cathodoluminescent textures and in- the evolution of granitic magmas. Geol Soc Am Bull 100:1886–
sights into vein formation. Geology 36:547–550 1897
Seitz S, Baumgartner LP, Bouvier A-S, Putlitz B, Vennemann T (2016) Zajacz Z, Halter WE, Pettke T, Guillong M (2008) Determination of fluid
Quartz standards for oxygen isotope and trace element analysis by ⁄ melt partition coefficients by LA-ICPMS analysis of coexisting
SIMS. Geost Geoanal Res 41:69–75 fluid and silicate melt inclusions: controls on element partitioning.
Sillitoe RH (2010) Porphyry copper systems. Econ Geol 105:3–41 Geochim Cosmochim Acta 72:2169–2197
Stefanova E, Driesner T, Zajacz Z, Heinrich CA, Petrov P, Vasilev Z Zajacz Z, Hanley JJ, Heinrich CA, Halter WE, Guillong M (2009)
(2014) Melt and fluid inclusions in hydrothermal veins: the mag- Diffusive reequilibration of quartz-hosted silicate melt and fluid in-
matic to hydrothermal evolution of the Elatsite porphyry Cu-Au clusions: are all metal concentrations unmodified? Geochim
deposit, Bulgaria. Econ Geol 109:1359–1381 Cosmochim Acta 73:3013–3027
Sterner SM, Bodnar RJ (1984) Synthetic fluid inclusions in natural quartz Zajacz Z, Candela PA, Piccoli PM (2017) The partitioning of Cu, Au and
I. Compositional types synthesized and applications to experimental Mo between liquid and vapor at magmatic temperatures and its
geochemistry. Geochim Cosmochim Acta 48:2659–2668 implications for the genesis of magmatic-hydrothermal ore deposits.
Spikings R, Reitsma MJ, Boekhout F, Mišković A, Ulianov A, Chiaradia Geochim Cosmochim Acta 207:81–101
M, Gerdes A, Schaltegger U (2016) Characterization of Triassic