Journal of Volcanology and Geothermal Research 136 (2004) 1 – 30

www.elsevier.com/locate/jvolgeores

From andesitic volcanism to the formation of a porphyry Cu-Au

mineralizing magma chamber: the Farallón Negro Volcanic
Complex, northwestern Argentina
Werner E. Halter a,*, Nicolas Bain a, Katja Becker a, Christoph A. Heinrich a,
Marianne Landtwing a, Albrecht VonQuadt a, Alan H. Clark b, Anne M. Sasso b,1,
Thomas Bissig c, Richard M. Tosdal c
a
Isotope Geochemistry and Mineral Resources, Department of Earth Sciences, ETH Zentrum, 8092 Zurich, Switzerland
b
Department of Geological Sciences and Geological Engineering, Queen’s University, Kingston, ON, Canada K7L 3N6
c
Mineral Deposits Research Unit, University of British Columbia, 6339 Stores Road, Vancouver, BC, Canada V6T 1Z4

Received 14 July 2003; accepted 9 March 2004

Abstract

The geological relations and geochemical composition of extrusive and subvolcanic intrusive rocks of the Upper Miocene
Farallón Negro Volcanic Complex (NW Argentina) were studied to reconstruct the evolution of a deeply eroded andesitic
volcano over most its life history of some 3.5 my. This paper is part of an integrated study of the processes that generate
magmatic-hydrothermal fluids responsible for the formation of porphyry-Cu-Au deposits. Here we present new data on bulk
rock chemical and Nd, Sr and Pb isotopic compositions of unaltered volcanic and subvolcanic intrusive rocks, for which
geological relations and ages are constrained by field-based stratigraphy and extensive Ar – Ar geochronology.
The combined results indicate that voluminous volcanism started at ca. 9.7 Ma with the eruption of basaltic andesite to
andesite lava flows and pyroclastic deposits. This basic to intermediate volcanism persisted until 7.5 Ma and was followed at
7.35 Ma by the eruption of dacites as the last recorded major extrusive event. The oldest andesitic intrusions, emplaced at 9.0
Ma, were followed by hypabyssal intrusions of basaltic andesite, andesite and dacite composition, which intruded the volcanic
sequence between 8.3 and 6.1 Ma. Intrusions at 6.8 F 0.2 Ma led to the formation of over 800 Mt of porphyry-Cu-Au ore at
Bajo de la Alumbrera. It was formed by a large flux of magmatic-hydrothermal fluids channeled through several phases of
fractured dacitic porphyry, which together with immediately following barren andesitic intrusions make up a composite stock.
After this magmatic-hydrothermal event, only minor rhyodacitic to rhyolitic intrusions were emplaced until magmatic activity
ceased at 6.1 Ma.
Geochemical data for the magmatic rocks indicate that the evolution of the system was dominated by mixing of a mafic
magma with a felsic magma that contained a component of crustal material, as indicated by linear compositional correlations
between compatible and incompatible elements. Initial strontium isotopic ratios vary between 0.7047 and 0.7118 while
143
Nd/144Nd range from 0.5127 to 0.5124, and both isotopic variations correlate with each other and with SiO2 content of the
rocks. After 8.5 Ma, a continuous increase in the silica content of intrusive rocks suggests the gradual formation of a magma
chamber. A substantial magma reservoir (>7 km3) must have been established by the time of porphyry-Cu-Au mineralization, to

* Corresponding author. Tel.: +41-1-632-68-02; fax: +41-1-632-18-27.

E-mail address: [email protected] (W.E. Halter).
1
128 Charleswood, Hudson, Quebec, J0P 1H0.

0377-0273/$ - see front matter D 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jvolgeores.2004.03.007
2 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30

allow rapid release of a large volume of ore-forming magmatic brine as a result of wholesale crystallization and de-volatilization
of the magma chamber.
D 2004 Elsevier B.V. All rights reserved.

Keywords: magma chamber; andesite; porphyry; volcanism; magma evolution

1. Introduction deposits. The FNVC is a partly eroded Upper Mio-

cene volcano that hosts the world-class Bajo de la
Andesitic to dacitic volcanoes are the most wide- Alumbrera Cu-Au porphyry deposit (Guilbert, 1995;
spread and hazardous magmatic systems on land, but Sasso and Clark, 1998; Ulrich et al., 2002). Herein,
they also give rise to valuable concentrations of base we present new data derived from detailed field
and precious metals, including porphyry-style and mapping (Sasso, 1997; Bain, 2001; Becker, 2001),
epithermal Cu-Au-Mo deposits. Porphyry copper sys- whole-rock chemical and isotopic analyses and
40
tems are the most important primary source of Cu ore, Ar – 39Ar age determinations. This information
and magmatic-hydrothermal processes in these sys- allows the broad-scale reconstruction of the FNVC
tems have been investigated in great detail (e.g., volcanic edifice and its long-term evolution over
Gustafson and Hunt, 1975; Dilles, 1987; Cline and several million years, and provides an insight into
Bodnar, 1991; Sillitoe, 1993; Herzarkhani et al., 1999). the gradual establishment of a subjacent magma
In earlier studies (Ulrich et al., 1999, 2002; Ulrich and chamber that gave rise to magmatic-hydrothermal
Heinrich, 2002), we have demonstrated that metal ore formation as one of the last major events prior
ratios in porphyry-hosted ores are controlled primarily to the demise of the magmatic system.
by the composition of juvenile magmatic hydrothermal
fluids which, in turn, are dictated by the mineralogical
and chemical evolution of their source magmas. In the 2. Geological setting
mineralized porphyries themselves, the salient petro-
graphic and geochemical relationships of the original The Farallón Negro Volcanic Complex, formally
magmas are largely obliterated by intense hydrother- defined by Sasso (1997) and Sasso and Clark (1998),
mal alteration. More complete information is recorded is located in the Hualfı́n District of the Catamarca
in some epizonal plutonic complexes such as the Province in northwestern Argentina (Latitude 27j15
Yerington batholith, which have been exposed by later S), and covers some 700 km2. It lies at the northern
block-faulting, tilting and erosion (Dilles, 1987). An limit of the Sierras Pampeanas tectonic province,
alternative approach to understanding the fluid-gener- more than 200 km to the east of the present volcanic
ating magmas is to study the extrusive equivalents of arc, and close to the southern margin of the Puna
the porphyries, in volcanic systems that have been physiographic province (Fig. 1). The contact between
partly eroded (e.g., Keith et al., 1997). Although this these two provinces is the NE – SW-striking Tucu-
does not provide a direct three-dimensional view of the mán Transfer Zone (Urreiztieta et al., 1993). The
deeper subvolcanic structures beneath the deposits, it Sierras Pampeanas are dominated by major basement
has the potential of providing a more extensive time uplifts separated by flat plains (Jordan and Allmen-
record of magma evolution through the use of detailed dinger, 1986; Allmendinger, 1986), the result of
stratigraphic relationships. westward-migrating deformation initiated in the Late
This paper aims at a reconstruction of the evolution Miocene (Sasso, 1997; Coughlin et al., 1998). Re-
of an ore-generating andesitic igneous complex, the verse faults, generated during this event, divided the
Farallón Negro Volcanic Complex (FNVC; Sasso, FNVC into several structural blocks (González
1997) in northwestern Argentina, which had been Bonorino, 1950; Sasso, 1997), separated by base-
recognized by Sillitoe (1973) as particularly well ment rocks (Fig. 1).
suited to study the relationship between the evolution The FNVC is the easternmost manifestation of
of a stratovolcano and the formation of porphyry-type Neogene volcanic activity in this area, but magmatism
 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30 3

Fig. 1. Geologic setting of the Farallón Negro Volcanic Complex. Due to late movement on major reverse faults, the volcanic sequence was
variably uplifted and eroded. Relicts of the volcanic complex cover an area of approximately 700 km2. The area in the box is enlarged in Fig. 2.
The inset shows part of South America and the location of the complex in the context of the central Andes. Modified from Martinez et al. (1995)
and from Sasso (1997).

was coeval with extensive eruptions along the axis of may have been facilitated by the formation of local
the Cordillera Occidental (e.g., Sasso and Clark, extensional basins along the Tucumán Transfer Zone
1998), and thus it constituted part of an extremely (Sasso and Clark, 1998). The complex has long been
broad, albeit scattered, volcanic zone across the Andes considered to represent the remnants of a large strato-
(Clark et al., 1976). This areal dispersion of strictly arc- cone hosting numerous subvolcanic intrusions, which
type volcanism (Sasso and Clark, 1998) conforms was eroded to a sub-horizontal section through the base
neither to geodynamic models of slab flattening or slab of the system (Caelles et al., 1971; Llambias, 1972).
steepening (cf. Kay et al., 1999), and it may be Llambias (1970, 1972) first mapped in detail the
significant that the FNVC lies within a transect in distribution of volcanic and intrusive rocks in the
which the inclination of the present-day Wadati-Beni- northwestern part of the complex (Fig. 2). The no-
off Zone gradually decreases from north to south menclature defined by Llambı́as will be used through-
(Sasso, 1997; Sasso and Clark, 1998). Uprise of magma out this paper. Sasso (1997) provided a refinement of
4 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30

euaFrlt
ld
Ca
d
e re
I nf

Fig. 2. Geological map of the FNVC, modified from Llambias (1970) and Sasso (1997). Volcanic rocks are named by their mineralogical
composition; intrusion names are from Llambias (1970) with additions from this study. The NW and NE transsects, mapped in this study, are
outlined. Also shown are two cross sections though the complex perpendicular to the general stratigraphy of the volcanic rocks. These profiles
define a continuous section through the volcanic stratigraphy. Most new samples and newly determined ages are located on the profiles.

the eruptive and intrusive chronology proposed by porphyry (Stults, 1985; Dawson, 1994; Guilbert, 1995;
Llambı́as, employing 40Ar – 39Ar age dating and sat- Proffett, 1995, 1997; Sasso, 1997; Muller and Forres-
ellite images to delimit more precisely the volcanic tal, 1998; Clark and Sasso, 2000; Ulrich et al., 2002;
and hypabyssal units, and introducing numerous mod- Ulrich and Heinrich, 2002); mineralogical and chem-
ifications to the sequence of events. ical investigations of the Agua Tapada and the Bajo el
From the standpoint of economic geology, extensive Durazno intrusions (Allison, 1984; Suchomel, 1985;
attention has been focused on: the geology and chem- Sasso, 1997); the temporal and chemical evolution of
ical and fluid evolution of the Bajo de la Alumbrera intrusive and extrusive rocks in the Capillitas Valley
 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30 5

(Breitenmoser, 1999; Hug, 1999); and a refinement of by Caelles et al. (1971), and discussed in more detail
the intrusive and hydrothermal history of the Agua Rica by Llambias (1972). Significant changes were pro-
porphyry-epithermal center (Landtwing et al., 2002), posed by Sasso (1997) and Sasso and Clark (1998),
an easterly outlier of the FNVC. Sasso and Clark who provided 40Ar – 39Ar incremental-heating dates
(1998) summarized the salient features of the numerous for numerous igneous and hydrothermal minerals. On
magmatic-hydrothermal centers occurring within and this basis, the main period of volcanic activity lasted
in the vicinity of the FNVC. from approximately 9.2 to 6.8 Ma and fluctuated in
The eruptive chronology of the main FNVC was composition between basaltic andesite and dacite over
first outlined on the basis of conventional K-Ar dates time. In contrast, the hypabyssal intrusions that cut the

Fig. 3. Field and microscopic views of rocks from the FNVC. (A) View of Cerro Durazno from the North, showing the sedimentary rocks of the El
Morterito formation at the base and the consistingly E-dipping volcanics on top. (B) Typical andesitic deposits with alternating pyroclastic breccias
and homogeneous flows. (C) Close up of an andesitic flow breccia. (D) Microphotograph of a fresh pyroxene-hornblende andesite (NB 9(99))
typical of samples from which hornblende was used for Ar – Ar dating. (E) Coarse deposits of dacitic volcanoclastics in the Agua Tapada area. (F)
Agua Tapada dacite porphyry, mainly visible due to its strong bleaching alteration. The dike indicated on the picture can be followed up to the
intrusion and is possibly a feeder to the overlying dacitic volcanic rocks.
6 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30

Table 1
Sample locations and descriptions
Sample Location Rock Type
Extrusive rocks
23-3(99) 3 431 810/6 985 897 Medium hornblende basaltic andesite
NB 4(99) 3 431 905/6 986 317 Large hornblende andesite
NB 5(99) 3 432 150/6 986 132 Medium pyroxene basaltic andesite
NB 9(99) 3 432 152/6 986 061 Large pyroxene basaltic andesite
NB 13(99) 3 431 730/6 986 704 Large hornblende andesite
NB 14(99) 3 431 623/6 986 817 Large hornblende andesite
NB 15(99) 3 431 830/6 986 611 Large hornblende andesite
NB 16(99) 3 431 840/6 986 588 Large hornblende andesite
NB 5(00) 3 439 726/6 982 690 Large pyroxene basaltic andesite
NB 6(00) 3 439 620/6 983 031 Small pyroxene basaltic andesite
NB 17(00) 3 439 623/6 982 680 Small pyroxene basaltic andesite
NB 22(00) 3 439 305/6 981 711 Small pyroxene basaltic andesite
NB 25(00) 3 439 491/6 981 362 Basalt
NB 31(00) 3 441 116/6 984 182 Medium hornblende andesite
NB 36(00) 3 431 494/6 987 727 Polymict dacitic breccia (groundmass)
NB 43(00) 3 432 636/6 983 713 Medium hornblende basaltic andesite
NB 49(00) 3 437 623/6 978 949 Small pyroxene basaltic andesite
NB 55(00) 3 432 028/6 985 129 Medium pyroxene basaltic andesite
NB 57(00) 3 448 387/6 980 023 Andesitic breccia (groundmass)
NB 63(00) 3 451 079/6 980 839 Medium pyroxene basaltic andesite
NB 65(00) 3 451 256/6 981 560 Andesite breccia (groundmass)
NB 66(00) 3 451 406/6 981 807 Medium pyroxene basaltic andesite
NB 67(00) 3 451 451/6 981 835 Andesitic breccia (groundmass)
NB 68(00) 3 451 548/6 981 756 Medium pyroxene basaltic andesite
NB 69(00) 3 451 838/6 981 937 Andesite breccia (groundmass)
NB 70(00) 3 453 036/6 984 026 Andesitic breccia (groundmass)
NB 76(00) 3 432 255/6 985 961 Basalt
DU 18-8(99) 3 443 875/6 992 497 Large hornblende andesite
DU 21-6(99) 3 443 689/6 992 327 Large hornblende andesite
DU 21-9(99) 3 442 700/6 989 139 Small pyroxene basaltic andesite

Intrusive rocks
9-9(99) 3 430 154/6 986 387 rhyodacite (Lomo Morado)
NB 8(99) 3 431 910/6 986 166 dacite porphyry (Agua Tapada)
NB 2(00) 3 430 962/6 986 293 Layered andesite
NB 13(00) 3 438 406/6 980 370 basaltic andesite dike
NB 18(00) 3 439 614/6 982 583 dacite porphyry
NB 32(00) 3 441 167/6 984 256 Layered andesite
NB 62(00) 3 433 195/6 983 055 basalt
NB 64(00) 3 451 144/6 981 022 large hornblende andesite
NB 73(00) 3 433 388/6 986 310 dacite porphyry (Agua Tapada)
NB 74(00) 3 432 842/6 984 970 dacite porphyry (Agua Tapada)
ML 5 3 441 803/6 982 705 andesite (Bajo el Durazno)
ML 7A 3 433 805/6 979 450 Monzonite (Alto de la Blenda)
ML 26 3 441 627/6 980 925 basaltic andesite dyke
ML 32 3 442 749/6 978 534 basalt (dyke)
ML 39 3 436 367/6 979 725 dacite porphyry (Las Pampitas)
ML 41 3 432 345/6 982 720 rhyodacite (Macho Muerto)
WH 8 3 144 725/6 985 015 andesite (Bajo el Durazno)
 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30 7

Table 1 (continued)
Sample Location Rock Type
Bajo de la Alumbrera (Ulrich, T. 1999)
BLA 67 N of Alumbrera andesite intrusion
BLA 48 S of Alumbrera decite intrusion
43-47.1 (22) drill whole Alumbrera dacite intrusion

Capillitas (Breitenmoser, T. 1999)

CAP 9 Capillitas Valley dacite lava flow unit 3
CAP 11 Capillitas Valley cpx lava flow
CAP 15 Capillitas Valley hb lava flow

Rock descriptions Matrix Phases Abund. Grain Size /Shape

Basalt Matrix plag and pyrox 40 – 60% <5 Am subhedral
phenocrysts plag 30% 0.2 – 1.0 cm euhedral
pyrox 10 – 20% 0.5 – 1.0 cm euhedral to subhedral
hb 10% <2.0 cm anhedral
Medium hornblende Matrix plag and hb 40 – 60% submicroscopic
basaltic andesite phenocrysts plag and hb 30 – 40% 0.3 – 1.5 cm subhedral to anhedral
hb 10 – 20% 0.10.2 cm to euhedral to subhedral
2.04.0 cm
pyrox 1 – 2% 0.10.1 cm to euhedral to subhedral
1.01.0 cm
Large pyroxene Matrix plag 60% <0.05 cm
basaltic andesite phenocrysts plag 22% 0.050.05 cm to subhedral
1.20.8 cm
pyrox 14% 0.10.1 cm to subhedral to euhedral
3.03.0 cm
hb 4% 0.2 – 0.5 cm, max subhedral to euhedral
1.0 cm
Medium pyroxene Matrix plag and pyrox 90% 0.3 cm euhedral to subhedral
basaltic andesite phenocrysts pyrox 5% 1.01.0 cm to subhedral
2.03.0 cm
plag <3% 1.0 – 3.0 cm subhedral
hb <2% 2.03.0 cm euhedral
Small pyroxene Matrix plag 60% submicroscopic to
basaltic andesite phenocrysts 0.05 cm
plag 24 – 36% 0.60.1 cm ehy euhedral
pyrox 4 – 16 0.5 – 0.8 subhedral
Medium hornblende Matrix plag 65% submicroscopic
andesite phenocrysts plag 21% 0.050.05 cm to subhedral
2.01.5 cm
hb 14% 0.050.05 cm to subhedral
0.52.0 cm
Large hornblende Matrix plag 30 – 60% submicroscopic
andesite phenocrysts plag 20 – 30% <1.0 – 4.0 cm euhedral to subhedral
hb 5 – 15% <0.1 – 1.5 cm euhedral to subhedral
bio 0 – 5% 0.3 cm euhedral
Andesite breccia Matrix plag 50 – 70% submicroscopic
phenocrysts plag 20 – 30% <1.0 – 4.0 cm euhedral to subhedral
hb 5 – 15% <0.1 – 1.0 cm euhedral to subhedral
bio 0 – 5% 0.3 cm euhedral
(continued on next page)
8 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30

Table 1 (continued)
Rock descriptions Matrix Phases Abund. Grain Size /Shape
Dacite Breccia Matrix plag and hb 50-60% submicroscopic
phenocrysts plag 10 – 30% 0.10.1 cm to euhedral to subhedral
0.20.5 cm
hb 10% 0.1 – 0.6 cm euhedral to subhedral
bio 10% 0.2 cm subhedral
qtz 2 – 5% 0.2 – 0.3 cm anhedral
Basaltic andesite Matrix plag and pyrox 50 – 60% microcrystalline
intrusion/dike phenocrysts plag 20 – 30% 0.1 – 4.0 cm euhedral to subhedral
pyrox 10% 0.5 – 1.0 cm euhedral to subhedral
hb 10% 1.0 – 4.0 cm subhedral
Andesite/monzonite Matrix plag 50 – 60% <0.25Am
intrusions phenocrysts plag 25% 0.1 – 1 cm anhedral to euhedral
hb 15 – 20% 0.1 – 5.0 cm euhedral to subhedral
Layered andesite Matrix plag 50 – 60% <0.25 Am
phenocrysts plag 25% 0.1 – 0.8 cm anhedral to euhedral
hb 10 – 15% 0.1 – 5.0 cm euhedral to subhedral
bio 5% <1.0 cm euhedral to subhedral
Rhyodacite Matrix plag 40 – 60% <3Am
phenocrysts plag 25% 1 – 5 cm euhedral to subhedral
hb <10% 0.3 cm subhedral
qtz 5 – 10% <4 cm euhedral
bio 5% 2 – 4 cm euhedral to subhedral
Dacite porphyry Matrix plag 60% submicroscopic
phenocrysts plag 22% 0.10.1 cm subhedral
to 0.41.0 cm
hb 10% 0.30.5 cm euhedral to subhedral
bio 8% 0.10.2 cm to subhedral
2.03.0 cm
qtz rare 0.30.5 to anhedral
0.80.8 cm
plag: plagioclase, hb: hornblende, bio: biotite, pyrox: pyroxene, qtz: quartz.

volcanic sequence appear to have followed a more by penetrative deformation and isoclinal folding of
systematic evolution from basaltic andesite to rhyolite deep water, flysch-type pelites and arenites of the
between 8.0 and 6.2 Ma. Suncho Formation (Mirre and Aceñolaza, 1972).
New field and petrographic interpretations pre- Also present are quartz + biotite + plagioclase + mus-
sented in this paper are based mainly on observations covite and quartz + microcline + andesine + hornblen-
in the northwestern part of the complex (hereafter de + biotite schists of high metamorphic grade,
referred to as the Main Zone; Fig. 2), but also include which form part of the Buey Muerto Schists (Gon-
information from the Cerro Durazno area, the Cap- zález Bonorino, 1950). Both units are intruded and
illitas area, and Agua Rica (Sasso, 1997; Breiten- metamorphosed by the granites of the Ordovician-
moser, 1999; Hug, 1999; Landtwing et al., 2002). Lower Silurian Capillitas batholith (Caelles et al.,
1971; McBride et al., 1976; A.H. Clark., unpubl.
40
Ar – 39Ar age data). This basement is overlain by
3. Petrography and field observations the approximately 200-m-thick Miocene El Morterito
Formation, a series of continental and marine red
3.1. Pre-volcanic rocks beds that includes conglomerate, sandstone and
siltstone (Turner, 1962, 1973; Martinez et al.,
The oldest basement rocks of this part of the 1995) and has been interpreted to represent sedi-
Sierras Pampeanas are schists and gneisses formed mentation into a pull-apart basin shortly prior to
 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30 9

Miocene volcanism (Sasso, 1997; Sasso and Clark, breccia (Fig. 3E) with plagioclase, hornblende,
1998). biotite and quartz phenocrysts in both matrix and
clasts. It differs from underlying units in containing
3.2. Miocene volcanic rocks much larger phenocrysts and a higher abundance of
biotite and quartz. The matrix is holocrystalline.
Basaltic-andesite and andesite flows and pyroclas- Each rock type can have a range in chemical
tic breccias are the predominant volcanic rocks in the composition (see below) and field names are used
FNVC. Bedding generally dips 20j to 30j centripe- only for mapping purposes.
tally away from the Alto de la Blenda stock (Figs. 2 Volcanic units in the Cerro Durazno area are almost
and 3A), implying that most of the volcanic units were exclusively formed of brecciated, biotite-bearing an-
extruded from a broadly central vent in that area. The desite flows, with angular to rounded clasts. Very few,
volcanic topography has not been reconstructed com- thin unbrecciated flows were observed. This contrasts
pletely, but it is likely that the complex resembled a with units in the Main Zone where volcanism is
large polygenic stratovolcano with satellite cones dominated by hornblende andesites and pyroxene-
(Sasso, 1997). The thickness of individual extrusive hornblende andesites with a large proportion of mas-
units varies between a few and several hundred sive lava flows. In addition, the stratigraphic sequence
meters, and their lateral continuity is generally limit- at Cerro Durazno is more consistent: only biotite-
ed. As a result, strict stratigraphic correlation between bearing andesites and quartz-bearing dacites are pres-
even closely separated sections through the volcanic ent (Table 1) and the alternation of compositionally
sequence is difficult. However, broad clarification of distinct units is not exhibited.
the volcanic history is possible by combining contin-
uously mapped radial sections with high precision 3.3. Miocene intrusive rocks
geochronology (Sasso, 1997; this study).
On the western flank of Cerro Durazno, the vol- Numerous subvolcanic stocks and dikes intrude the
canic suite of the FNVC unconformably overlies the volcanic rocks (Fig. 3F). Individual stocks are typi-
El Morterito Formation (Fig. 3A). Much of the cally circular to elliptical in shape and are mineralog-
volcanic pile comprises lava flows and flow breccias ically distinct. Most intrusions are found in the Main
with decimeter-sized clasts that occur mainly at the Zone where they form a significant proportion of the
base of each unit (Fig. 3B and C). These breccias are outcrop area. A large number of stocks of 0.1 to 1 km
intercalated with thin lava flows. in diameter appear to be distributed around the large
In the following description, volcanic rocks are central Alto de la Blenda monzonite stock from which
named in the field after their phenocryst assem- numerous pyroxene-hornblende andesite and horn-
blage (Table 1). Hence, pyroxene-hornblende andes- blende andesite dikes emanate radially from this
ite is used to describe rocks containing plagioclase, center, albeit with a dominant NW – SE direction
hornblende and clinopyroxene as common phenoc- (Fig. 2).
rysts. Orthopyroxene is rare and magnetite is a Subvolcanic stocks throughout the complex com-
ubiquitous accessory phase (Fig. 3D). These rocks prise several individual magma pulses, emplaced in a
are devitrified, holocrystalline and have a fine- restricted time frame. Within a given stock, intrusions
grained plagioclase and clinopyroxene matrix exhib- can have a wide compositional range. For example,
iting flow textures. Rounded quartz xenocrysts are the Alumbrera porphyry stock is composed of intru-
rarely present. Hornblende andesite, which domi- sions ranging from dacite to andesite, with the most
nates the extrusive succession, is petrographically siliceous being intruded first and the most mafic last.
distinguished by the absence of phenocrystic py- The mineralogy of these intrusions is similar to that
roxene. Biotite-bearing andesite breccia occurs in of the volcanic rocks, but quartz-bearing dacite por-
the upper part of the stratigraphic column. Individ- phyries are much more abundant than in the
ual units are up to several meters thick and corresponding extrusive suites. The most detailed
commonly graded. The uppermost major unit of descriptions of compositionally diverse stocks from
the volcanic sequence is a quartz-bearing dacite the FNVC are given by Sasso (1997), Proffett (1995,
10 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30

1997), Ulrich et al. (2002) and Landtwing et al. emplaced towards the end of the magmatic activity
(2002). in the northwestern part of the FNVC. Rhyolitic
Both open pit exposure and core from deep drill intrusions cut all other intrusive units. They are
holes in the Alumbrera dacites show crenulated fine-grained and show millimeter-scale flow-banding.
contacts between mafic and felsic rocks, demonstrat- Extrusive equivalents of these rocks within the com-
ing that magma commingling occurred between an- plex are unknown.
desitic and dacitic magma (Fig. 4). Mafic enclaves in Basaltic andesite to dacite stocks emplaced be-
andesitic rocks described by Sasso (1997) are con- tween 8.5 and 6.8 Ma were hydrothermally altered
sistent with the hypothesis that magma mixing has and alteration patterns are centered on individual
occurred. intrusions (Sasso, 1997). Igneous contacts truncating
The oldest stocks (i.e., Tampa Tampa) are biotite- veins and alteration haloes in the Alumbrera stock
bearing andesite porphyries (Table 1), followed by indicate that each intrusion was hydrothermally al-
pyroxene-hornblende andesite intrusions with pheno- tered shortly after its emplacement and before the
crystic pyroxene, hornblende, plagioclase and mag- next intrusion was emplaced (Ulrich and Heinrich,
netite (Durazno basaltic andesite porphyry and Chilca 2002). Intensely altered stocks have a central quartz-
andesite porphyry). The groundmass of these intru- magnetite zone, surrounded by typical potassic alter-
sions is coarser-grained than that of volcanic rocks ation with hydrothermal biotite, K-feldspar and mag-
with the same mineralogy. Partially resorbed volcanic netite (Allison, 1984; Proffett, 1995, 1997; Sasso and
clasts are locally present. Younger intrusions are Clark, 1998; Ulrich and Heinrich, 2002). This alter-
texturally similar to the Durazno and Chilca intru- ation grades outwards into a propylitic alteration
sions, but contain no pyroxene phenocrysts. In the mineralogy (epidote + chlorite + calcite), which may
Agua Tapada area, an intrusion of hornblende andes- extend over more than 1 km into the surrounding
ite porphyry shows decimeter-thick flow banding. volcanic rocks. Feldspar-destructive sericitic alter-
Later emplaced quartz-bearing dacite porphyries ation overprints these alterations in the upper part
(with biotite instead of hornblende) have large phe- of hydrothermal systems. Advanced argillic alter-
nocrysts in a fine-grained matrix (Table 1). Associ- ation is associated with epithermal (vuggy-silica)
ated with quartz-bearing dacite intrusions are dikes of veins radiating around the Agua Tapada intrusion
one to a few meters in thickness (Fig. 3F) at Agua (Fig. 2). A low-sulfidation epithermal system of the
Tapada and between Agua Tapada and the Alumbrera same age (6.8 Ma) as the Alumbrera porphyry
stock. Rhyodacitic and rhyolitic intrusions were (Sasso, 1997) is responsible for Au-Ag ore formation
at the Farallón Negro vein deposit. The youngest
intrusive phases in the Alumbrera stock are relatively
mafic (andesite) that are not affected by alteration.
Alteration is also absent in late rhyodacites and
rhyolites.

40
4. Ar – 39Ar geochronology

Age relationships of volcanic, intrusive and hydro-

thermal activity have been clarified through laser-
induced, incremental-heating 40Ar – 39Ar dating in
the Queen’s University Geochronology Laboratory,
employing techniques initially developed by Sasso
Fig. 4. Sample of a drill core in deep parts of the Alumbrera pit (1997). New age determinations of unaltered horn-
showing evidence for mixing between andesitic and dacitic magma.
The irregular boundary with overlapping plagioclase phenocrysts blendes and biotites presented herein supplement
suggests that both magmas were partially liquid as they were earlier data reported by Sasso (1997) and summarized
brought into contact. by Sasso and Clark (1998).
 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30 11

Fig. 5. 39Ar release spectra from biotites and hornblendes showing the plateau ages and their 2r uncertainty. Comments on the release spectra
are given in Table 2.
12 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30

4.1. Analytical procedure 24.36 Ma (Sandeman et al., 1999). This age has
recently been reevaluated (Sandeman, personal com-
Samples (8 mg) were irradiated for 7.5 h at munication) and ages reported by Sasso (1997) and
McMaster nuclear reactor, Hamilton, Canada. The Sasso and Clark (1998) using the same standard were
monitor standard used is a biotite from a glassy adjusted accordingly. All samples were step-heated
rhyodacite tuff in Peru, with an assigned age of using a defocused argon laser beam and finally melted

Table 2
Ar – Ar age determinations
Sample Mineral Age data Comments
ML 41 Biotite Plateau: 6.26 F 0.15 Ma The age-spectrum is undisturbed
(95.8% of 39Ar) and reliable
Total gas: 6.31 F 0.17 Ma
ML 39(00) Hornblende Plateau: 7.22 F 0.28 Ma Well reproducible undisturbed
(98.2% of 39Ar) age-spectrum
Total gas: 7.22 F 0.33 Ma
14-4D(99) Hornblende Plateau: 7.27 F 0.35 Ma Well reproducible undisturbed
(94.6% of 39Ar) age-spectrum
Total gas: 6.99 F 0.44 Ma
NB 8(99) Biotite Plateau: 7.35 F 0.16 Ma The plateau age is reliable. The
(96.9% of 39Ar) low-temperature steps show minor
Total gas: 7.34 F 0.18 Ma evidence for excess-argon, but
the ages are within error of the
high-temperature ages and
therefore included in the plateau
NB 15(99) Hornblende Plateau: 7.78 F 0.28 Ma Well reproducible age-spectrum,
(94.2% of 39Ar) minor Ar-loss at lower laser-powers
Total gas: 7.44 F 0.33 Ma may have occurred, but the plateau
age is reliable
NB 70(00) Biotite Plateau: 8.0 F 0.2 Ma Reliable plateau age, but the
(75.6% of 39Ar) spectrum shows minor Ar-loss
Total gas: 7.5 F 0.2 Ma at low temperatures
NB 9(99) Hornblende Plateau: 8.05 F 0.37 Ma Well reproducible undisturbed
(96% of 39Ar) age-spectrum
Total gas: 7.99 F 0.44 Ma
DU 21-9(99) Hornblende Plateau: 8.89 F 0.71 Ma Well reproducible age-spectrum.
(81.2% of 39Ar) The large error is due to the low
Total gas: 8.39 F 0.89 Ma potassium content of the sample
ML 26 Hornblende Plateau: 8.93 F 0.41 Ma Well reproducible undisturbed
(87.7% of 39Ar) age-spectrum
Total gas: 8.75 F 0.46 Ma
DU 18-8(99) Hornblende Plateau: 8.95 F 0.56 Ma Well reproducible undisturbed
(92.3% of 39Ar) age-spectrum. The relatively large
Total gas: 8.57 F 0.71 Ma error is due to the low potassium
content of the sample
NB 31(00) Hornblende Plateau: 9.0 F 0.2 Ma Well reproducible undisturbed
(93.7% of 39Ar) age-spectrum
Total gas: 8.6 F 0.3 Ma
NB 32(00) Hornblende Plateau: 9.0 F 0.3 Ma The age-spectrum is not perfectly
(68.3% of 39Ar) reproducible (disturbed)
Total gas: 7.0 F 0.5 Ma
NB 6(00) Hornblende Plateau: 9.4 F 0.3 Ma Reproducible undisturbed
(94.6% of 39Ar) age-spectrum
Total gas: 9.7 F 0.4 Ma
Uncertainties are 2j.
 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30 13

with the focused beam. Hornblende samples yielded inferred central stratocone. The oldest quartz-bearing
only enough Ar gas to perform two to five heating flow is an 8.0 F 0.2 Ma dacite (NB70(00)) from Cerro
steps, and thus two aliquots of each sample were Durazno, though dacites represent the youngest vol-
analyzed. These yielded ages reproducible within canic rocks, dated at 7.27 F 0.35 Ma (14-4D(99)) at the
analytical uncertainty. Ages acquired in this study top of the NW transect (AB in Fig. 2). These dacites
are shown in Fig. 5 and summarized in Table 2 with have the same plateau age (within analytical uncer-
the corresponding percentages of the gas released, tainty) as the 7.35 F 0.16 Ma Agua Tapada dacitic
along with 2j uncertainties. Ages determined at intrusion (Sasso, 1997).
various levels of the volcanic pile in the Main Zone Biotite-bearing andesite intrusions between the
correlate very closely with those reported by Sasso Alto de la Blenda and the Bajo el Durazno stocks,
(1997), which are not repeated in Table 2 but are previously mapped as Chilca andesites (Llambias,
incorporated in Fig. 6. 1970), are dated at 9.0 F 0.2 Ma, significantly older
than the Chilca andesite porphyry dated at 7.9 Ma
4.2. Results (Sasso, 1997). These andesites constitute the oldest
dated intrusive rocks and are hereafter referred to as
The oldest hornblende 40Ar – 39Ar plateau age of Tampa Tampa andesites porphyry from the Sierra de
9.4 F 0.3 Ma was obtained from an andesite flow west Tampa Tampa in which they occur. This older age
of Bajo el Durazno; stratigraphically older units could agrees with crosscutting relationships reported by
not be dated because no fresh samples could be Llambias (1972), suggesting that these intrusions are
located. The lowermost unit (sample DU 18-8(99)) older than the 8.3 F 0.2 Ma Bajo el Durazno basaltic
overlying the El Morterito Formation at the base of andesite porphyry.
Cerro Durazno was dated at 8.95 F 0.56 Ma, suggest- Pyroxene-hornblende andesite from the Bajo el
ing that earlier volcanic activity was not represented Durazno porphyry was dated at 8.39 F 0.18 Ma (horn-
in this immediate area. Basaltic andesite and andesite blende: Sasso, 1997). The oldest hypabyssal felsic unit,
eruptions persisted at least until 8.05 F 0.37 Ma in the the Las Casitas Rhyolite, makes up a flow-banded plug,
interpreted by Llambı́as (op. cit.) as constituting the
terminal intrusive stage in the FNVC, but cross-cutting
hornblende andesite dikes yield an hornblende plateau
age of 8.0 F 0.12 Ma (Sasso, 1997). Contemporaneous
intrusion is represented by the Chilca andesite porphy-
ry (hornblende, 8.0 F 0.11 Ma: Sasso, 1997).
The central area of the eroded main stratocone is
dominated by the large 7.55 F 0.3 Ma Alto de la
Blenda stock (Sasso, 1997), the only unit in the
complex with, hypidiomorphic-granular textures.
Llambias (1970) described this body as monzonitic
based on petrographic studies and Sasso (1997) con-
firmed the presence of K-feldspar by cathodolumines-
cence imaging. This pluton occupies a central position
within the main cone, suggesting that it could be the
remnant of the main conduit. Extrusive equivalents
are missing.
Intrusion of dacite magma on a significant scale
is first manifested by the 7.35 F 0.16 Ma Agua
Tapada plug (‘‘quartz andesites of Bajo de Agua
Fig. 6. Age distribution of volcanic rocks and porphyries based on Tapada’’ of Llambias, 1970), emplaced along the
Ar – Ar dating, combining data form this study (Table 2) and Sasso northeastern margin of a NW-trending corridor. Pro-
(1997). gressively younger dacites intruded towards the SE
14 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30

Fig. 7. Synthetic stratigraphic column of the two profiles mapped in Fig. 2, together with the variation in SiO2, Cu, 87Sr/86Sr and eNd. Patterns
are the same as in Fig. 2. The composition of the extrusive rocks shows no systematic trend over time with respect to any of the elements and
isotopic ratios. Moreover, variations in mineralogy are not always reflected by variations in the chemical composition.
 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30 15

of this corridor, forming the Las Pampitas dacite ene is present in 30 – 40% of the units, and the
porphyry (7.22 F 0.28 Ma) complex. Following a frequency of its occurrence does not change from
0.2 –0.3-my quiescent period, the Bajo de la Alum- the bottom to top of the volcanic sequence. Only at
brera intrusions were emplaced between 6.98 F 0.08 the end of the volcanic activity (i.e., after 8.05 Ma) did
(biotite: Sasso, 1997) and 6.78 F 0.15 Ma (horn- the mineralogical composition of the extrusive rocks
blende: Sasso, 1997). The dacite intrusions were change, with the advent of biotite-bearing andesite.
followed by the emplacement of a suite of dikes The uppermost-preserved unit is a large body of a
and stocks in the NW quadrant of the complex: e.g., quartz-biotite-bearing dacite tuff.
the Macho Muerto rhyodacite porphyries, which When the reconstructed thickness is plotted
yield plateau ages of 6.26 F 0.15 Ma (sample ML against the ages of volcanic units (Fig. 8), it appears
41) and 6.04 F 0.07 Ma (Sasso, 1997), and the that at least in the mapped northwestern sector (Fig.
Loma Morada stock that was dated at 6.18 F 0.05 2), the volcanic edifice grew rapidly at the begin-
Ma (Sasso, 1997). The youngest hypabyssal rocks, ning of its development and that the rate of thick-
and the first of strictly rhyolitic composition, crop ening thereafter diminished markedly. Assuming
out in the Quebrada de los Leones. No rhyolites that the volcano had a conical shape and grew at
were dated isotopically, but these silicic bodies cut similar rates in all directions (integrated over time
all other units of the main stratocone. scales of a few hundred thousand years), we calcu-
The Chilca intrusion is cut by swarms of pyroxene- late that the volume of material ejected per 0.2 to
hornblende andesite and hornblende andesite dikes, 0.3 m.y. was approximately constant, approaching
which transect the core of the main volcanic complex. 310 km3/m.y. This linear relationship, we propose,
These dikes radiate around, and are cut by the Alto de la can be used to interpolate ages of other volcanic
Blenda stock, and are therefore not present therein. units within this column, and on this assumption,
However, one dike was dated at 6.36 F 0.11 Ma (whole we estimate that the stratocone construction began
rock: Sasso, 1997), indicating that mafic-to-intermedi- at about 9.6 Ma. At Cerro Durazno (Fig. 2), the
ate rocks were intruded through much of the evolution accumulation of volcanic material appears to have
of the FNVC, overlapping temporally with the em- been faster than in the Main Zone, because two
placement of dacitic as well as monzonitic magmas.

5. Stratigraphic and structural reconstruction

Cross sections from the mapped NW and NE

transects (AB and CD, respectively, in Fig. 2), from
which many of the geochronology samples were
taken, allow the reconstruction of a composite strati-
graphic column (Fig. 7) through the volcanic se-
quence of the Main Zone. The correlation between
the mapped transects is based on isotopic age deter-
minations and the measured average dips of volcanic
units.
Although this reconstruction may incorporate im-
portant gaps in the record of the volcanic activity, it
shows that the phenocryst assemblage of volcanic
rocks did not change systematically from bottom to Fig. 8. Cumulative measured thickness of volcanic units as a
function of time (full squares), together with the estimated volume
top of the volcanic sequence that records a minimum
of the volcano as a function of time. The volume was calculated
of 2.1 F 0.65 my. In addition to plagioclase, horn- assuming an overall conical shape. The linear growth through time
blende is present almost throughout the volcanic pile suggests that average magma production remained essentially
with randomly varying abundances and sizes. Pyrox- constant over 2.1 my of extrusive activity.
16 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30

horizons (at the base and the upper part of Cerro activity. Although the minimum age of the faulting
Durazno), separated by several hundreds of meters cannot be determined, we interpret this southwest-
of volcanic rocks (Fig. 3A), were dated at 8.95 F side-down movement and the 7.3-Ma-old quartz-bear-
0.56 and 8.89 F 0.71 Ma, respectively. Flank erup- ing dacitic magmatism (intrusive and extrusive) to
tions may have contributed to the growth of the reflect a partial caldera collapse. In a second stage,
volcanic pile in this area. movement along the fault was reversed, presumably
In the profile A – B of Fig. 2, the late quartz- as a result of the regional NW – SE contraction. This
bearing dacite breccia is separated from the strati- elevated the older andesitic volcanics with respect to
graphically lower parts of the sequence by a fault, yet the younger dacite breccia.
it is topographically lower than the stratigraphically
highest biotite-bearing andesite on the southeastern
side of the fault. The latter andesite dips southeast at 6. Whole-rock chemistry
the fault. These bedding attitudes are interpreted to
result from a two-step process (Fig. 9) in which the Extrusive and intrusive units were analyzed for
southeastern side of the fault was first lowered along a their whole-rock major and trace element composition
normal fault, creating drag-folds in the andesite brec- using X-ray fluorescence (Table 3). Other data from
cias (Bain, 2001). The age of this unit, and thus the the FNVC are available from Sasso (1997), Breiten-
maximum age of this deformation, is 7.8 Ma. Depo- moser (1999) and Ulrich (1999). A short description
sition of dacite breccia represents the last volcanic of the newly analyzed samples and their locations is
given in Table 1 and further information is presented
in Bain (2001). From Fig. 7 it is apparent that changes
in chemical composition do not strictly correlate with
changes in mineralogical composition. Quartz-bearing
samples (mapped as quartz-bearing dacites) may have
SiO2 contents as low as 59.9 wt.%, whereas pyroxene
bearing, biotite-free andesites have up to 62 wt.%
SiO2. We maintain mineralogically based rock names
where no information on the chemical composition is
available.
Most intrusive and extrusive rocks of the FNVC
form a high-K calc-alkaline series (Fig. 10; Caelles et
al., 1971; Sasso, 1997). The most primitive rocks are
lamprophyre dikes intruding basement granites in the
Capillitas area. They have SiO2 contents of about 42
wt.% (Breitenmoser, 1999). Samples from the Main
Zone have a minimum SiO2 content of 50.2 wt.%. The
most silica-rich rock is an intrusive rhyodacite with
66.4 wt.% SiO2. Rock compositions falling between
these two extremes define linear arrays for nearly all the
major and trace elements (Fig. 10). Only Cu shows a
Fig. 9. Cross section through the northwestern part of the mapped
area showing the relationship between the upper parts of the
non-systematic behavior, although a general tendency
andesite, the terminal dacite volcanic breccia and the major fault to decrease with increasing SiO2 is apparent from the
interpreted to be a reversly reactivated caldera fault. Also shown are data. Volcanic rocks extend to more mafic composi-
the ages for these units determined in this study. (A) is the tions than intrusive rocks and quartz is an abundant
interpreted situation around 7.3 Ma, (B) is the current observation. phenocryst phase only in intrusions, but both suites
These relationships are interpreted to result from normal faulting
(A) at approximately 7.3 Ma resulting in southeastwards plunging follow the same trend and cannot be distinguished on
dragged andesite bedding. Later reversal of the movement along the the basis of their composition. However, it appears that,
fault (B) is likely to have occurred during regional compression. for a given SiO2 content, the variation in the concen-
Table 3
Whole rock XRF anlyses of intrusive and extrusive rocks from the FNVC
Element sample
wt.% 23_3/99 NB9/99 NB5/99 NB4/99 NB15/99 NB16/99 NB13/99 NB14/99 DU21-9 DU21-6 DU18-8 NB6/00 NB13/00 NB36/00 NB55/00 NB69/00
SiO2 56.53 59.55 52.97 61.11 61.48 61.49 61.22 61.77 61.99 53.55 66.81 61.77 55.82 59.91 57.65 53.36
TiO2 0.96 0.79 1.22 0.79 0.79 0.8 0.79 0.8 0.87 0.97 0.71 0.69 1.08 0.79 1.15 1.29
Al2O3 15.14 16.54 17.36 16.49 16.19 16.78 16.85 16.66 17.08 17.04 14.25 17.61 17.12 13.93 17.32 17.29
Fe2O3 7.27 5.93 8.66 5.83 5.84 6.02 6 5.88 6.55 7.38 5.48 5.32 8.13 5.71 8.17 8.6

W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30
MnO 0.18 0.08 0.13 0.08 0.09 0.08 0.08 0.11 0.09 0.16 0.05 0.14 0.11 0.11 0.11 0.15
MgO 4.1 0.64 3.32 1.57 1.51 2.08 0.69 1.83 1.34 1.98 0.86 2.11 3.49 2.29 3.18 2.81
CaO 5.55 5.96 7.41 5.53 4.8 5.11 5.7 5.5 6.06 7.24 4.68 5.64 6.82 4.94 5.75 7.84
Na2O 2.45 2.94 3.14 3.01 2.87 3.02 2.9 2.99 2.77 2.89 2.42 3.7 2.26 2.45 2.55 3.25
K2O 3.36 3.46 3.64 3.69 3.46 3.45 3.47 3.46 2.82 2.44 2.57 1.96 3.51 2.21 3.34 2.7
P2O5 0.39 0.32 0.5 0.32 0.32 0.32 0.33 0.32 0.37 0.36 0.33 0.33 0.34 0.33 0.47 0.46
LOI 3.46 2.15 1.67 0.98 1.46 1.16 1.99 1.16 1.2 6 1.25 1.44 2.72 4.53 2.7 0.76

Sum 99.39 98.36 100.02 99.40 98.81 100.31 100.02 100.48 101.14 100.01 99.41 100.71 101.40 97.20 102.39 98.51

Ni 27 <4 9 <4 <4 <4 <4 <4 <4 <4 <4 22 12 13 11 13

Pb 17 13 13 9 12 11 11 9 8 <4 7 36 30 30 42 42
V 118 218 226 166 187 230 236 175 123 232 82 90 238 207 151 201
Zn 81 61 92 57 63 73 64 71 64 88 51 84 82 83 100 90
Nb 38 34 43 28 31 31 32 30 29 26 27 37 32 43 46 44
Y 38 32 36 32 29 32 34 32 32 29 27 45 43 43 54 56
Zr 221 189 255 180 184 185 184 184 190 148 170 233 153 187 233 215
Th <7 <6 <7 <6 <6 <6 <6 <6 <7 <7 <6 <6 <6 <6 <6 <6
Cu 5 11 21 6 6 8 10 7 <2 10 <4 15 17 17 74 34
Cr 124 < 11 26 14 14 < 11 17 < 11 21 < 11 18 1 < 11 < 11 < 11 < 11
Ba 566 619 543 608 582 594 608 586 718 623 585 676 599 486 498 601
Rb 90 113 111 109 110 106 110 106 85 58 69 101 82 140 98 89
Sr 410 475 529 465 481 475 482 477 496 549 421 533 471 531 412 592
La 43 43 45 42 43 46 44 43 38 27 32 13 15 23 22 18
Ce 92 74 89 66 79 79 88 80 66 56 62 38 < 20 45 41 41
F < 270 < 270 < 270 < 270 < 270 274 < 270 278 < 270 < 270 < 270 < 270 199 512 385 67
U <6 <6 <7 <6 <6 <6 <6 <6 <6 <6 <6 9 <6 <6 15 28
Nd 37 32 32 28 22 31 30 34 27 31 28 22 < 10 23 29 22
Co 19 14 25 13 14 13 16 16 17 24 13 < 10 < 10 < 10 < 10 < 10
Ga 7 8 7 6 7 7 8 7 6 6 4 11 11 8 13 14
Sc 26 19 19 16 20 19 23 18 14 14 10 11 29 18 24 25
S 285 54 < 38 656 < 38 270 126 699 176 250 < 38 < 38 41 153 < 38 586
Ba 566 619 543 608 582 594 608 586 718 623 585 676 599 486 498 601
Pb 17 13 13 9 12 11 11 9 8 <4 7 36 30 30 42 42
(continued on next page)

17
 18
Table 3 (continued)
Element sample
NB32/00 NB5/00 NB76/00 NB25/00 NB43/00 NB17/00 NB49/00 NB22/00 NB62/00 NB63/00 NB65/00 NB66/00 NB67/00 NB68/00 NB70/00 NB8/99
SiO2 59.69 54.83 50.19 51.9 58.91 55.35 53.98 57.88 52.34 54.25 66.34 50.26 61.64 51.15 65.15 59.39
TiO2 0.7 1.14 1.4 1.15 1.01 1.09 1.3 0.79 1.22 1.22 0.93 1.45 0.52 1.34 0.55 0.65

W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30
Al2O3 17.24 17.78 17.78 18.04 17.46 17.68 17.43 17.14 15 16.49 14.16 16.9 19.04 16.36 15.23 15.61
Fe2O3 5.78 8.78 7.86 8.94 6.64 8.18 8.72 5.71 9.44 8.29 6.27 9.23 3.75 9.15 4.34 4.96
MnO 0.15 0.17 0.24 0.16 0.16 0.18 0.21 0.15 0.16 0.22 0.13 0.81 0.06 0.32 0.1 0.12
MgO 2.11 3.75 2.1 2.1 1.87 3.48 3.4 1.64 5.16 2.85 1.36 3.72 1.57 4.28 1.43 1.74
CaO 6.37 7.77 11.53 8.69 5.51 7.47 7.69 5.81 8.09 7 4.1 9.41 6.3 8.67 4.26 5.5
Na2O 2.99 2.69 2.95 2.95 3.29 2.68 2.72 3.49 2.24 3.17 2.04 3.32 4.07 3.03 3.34 2.22
K2 O 3.47 2.72 2.55 2.32 3.31 2.65 2.45 2.86 2.85 2.77 3.22 2.37 1.5 2.49 1.35 3.42
P2O5 0.34 0.43 0.48 0.59 0.39 0.43 0.46 0.31 0.33 0.35 0.3 0.52 0.18 0.51 0.25 0.28
LOI 1.49 0.33 3.89 5.42 2.45 0.56 1.83 2.75 3.6 1.63 1.52 2.13 1.86 1.37 5.4 6

Sum 100.33 100.39 100.97 102.26 101.00 99.75 100.19 98.53 100.43 98.24 100.37 100.12 100.49 98.67 101.40 99.89

Ni 11 12 28 11 11 11 12 11 124 11 11 41 11 65 11 <4
Pb 35 39 41 34 38 31 36 26 33 40 33 37 24 38 32 14
V 139 221 320 221 125 275 168 160 242 246 168 297 116 275 118 169
Zn 75 94 67 111 87 87 82 87 91 96 79 91 62 108 68 91
Nb 39 41 43 40 44 37 42 34 37 39 39 46 26 46 37 32
Y 49 52 52 50 52 48 48 42 45 49 49 52 36 55 37 27
Zr 179 190 176 219 240 180 211 193 165 170 180 203 157 215 150 158
Th <6 <6 <6 <6 <6 <6 <6 <6 <6 <6 <6 <6 <6 <6 <6 <6
Cu 21 18 67 28 15 17 23 14 17 22 19 32 28 26 12 5
Cr < 11 < 11 < 11 < 11 < 11 < 11 < 11 < 11 271 < 11 < 11 10 < 11 81 < 11 18
Ba 625 554 523 589 583 594 542 662 501 633 406 662 350 618 428 534
Rb 101 76 75 60 88 72 71 75 87 68 121 52 49 73 96 113
Sr 572 568 567 454 503 541 515 564 446 520 386 559 592 597 392 353
La 14 16 15 15 21 18 17 15 16 17 19 20 8 20 12 38
Ce 34 33 43 50 47 46 38 37 35 28 < 20 56 < 20 52 27 72
F 320 70 427 209 254 109 154 410 204 74 380 438 2974 149 275 325
U 13 17 22 12 10 <6 11 <6 <6 25 7 13 <6 24 <6 <6
Nd < 10 25 < 10 30 31 19 26 < 10 27 21 < 10 25 < 10 30 < 10 32
Co < 10 < 10 < 10 12 < 10 < 10 < 10 < 10 21 < 10 < 10 30 < 10 27 < 10 16
Ga 11 13 13 13 13 12 14 11 11 11 9 12 10 12 8 5
Sc 15 28 39 22 16 28 15 16 34 35 23 41 12 34 10 14
S 323 < 38 < 38 34 1142 < 38 173 < 38 < 38 < 38 < 38 687 1372 121 2057 < 38
Ba 625 554 523 589 583 594 542 662 501 633 406 662 350 618 428 534
Pb 35 39 41 34 38 31 36 26 33 40 33 37 24 38 32 14
 Element sample
9_9/99 NB73/00 NB64/00 NB2/00 WH8 ML7A ML5 ML39 ML32 ML26 ML41
SiO2 65.34 59.72 52.44 57.57 59.88 60.92 60.05 62.51 53.04 56.29 66.4
TiO2 0.39 0.64 0.97 0.75 0.72 0.86 0.66 0.59 1.07 0.94 0.34
Al2O3 15.83 16.47 16.38 15.85 16.14 16.63 16.8 15.66 17.91 16.3 15.81
Fe2O3 3.16 4.67 7.68 5.56 5.11 6.09 5.42 4.58 8.2 7.38 2.56
MnO 0.1 0.13 0.31 0.12 0.13 0.08 0.16 0.12 0.13 0.17 0.07
MgO 1.06 1.92 2.76 1.5 2.44 2.01 1.68 1.66 3.12 4.02 0.6

W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30
CaO 4.71 5.62 8.56 6.44 5.37 5.04 6.03 4.96 8.48 6.06 4.4
Na2O 3.79 2.91 3.03 2.88 3.13 3.14 3.25 2.67 3 2.66 2.91
K2 O 3.15 3.22 3.78 3.35 3.36 3.63 3.34 4.14 2.34 2.67 3.52
P2O5 0.18 0.28 0.39 0.3 0.26 0.32 0.32 0.26 0.43 0.35 0.15
LOI 1.74 4.23 3.75 3.34 2.62 1.37 2.57 2.74 1.51 2.01 2.73

Sum 99.45 99.81 100.05 97.66 99.16 100.09 100.28 99.89 99.23 98.85 99.49

Ni <4 11 12 13 16 <4 <4 <4 <4 18 <4

Pb 5 33 39 33 30 <4 <3 <3 9 16 <3
V 76 176 188 239 152 111 124 130 221 205 81
Zn 44 101 111 85 64 35 62 79 83 122 44
Nb 27 38 41 39 39 29 27 27 29 26 27
Y 21 41 52 42 38 30 27 23 34 33 16
Zr 142 168 196 151 159 203 166 130 174 180 139
Th <6 <6 <6 <6 <6 <6 <6 <6 <7 <7 <6
Cu <2 12 20 15 29 18 7 <4 7 14 <2
Cr 13 < 11 < 11 < 11 < 11 15 15 15 30 106 20
Ba 669 492 697 446 675 614 785 676 724 641 672
Rb 103 101 89 90 96 108 92 120 54 64 116
Sr 635 474 514 461 604 410 585 487 518 446 552
La 16 18 18 18 14 37 27 38 29 42 29
Ce 51 27 42 44 < 20 64 62 65 62 82 54
F 415 540 422 612 499 < 270 353 482 < 270 < 270 481
U <6 8 18 10 <6 <6 <6 <6 <6 <6 <5
Nd 19 < 10 20 19 < 10 27 18 18 27 35 30
Co 9 < 10 < 10 < 10 < 10 17 15 13 24 22 10
Ga 5 12 10 11 10 7 5 5 8 7 5
Sc 4 20 25 30 22 16 9 12 17 29 5
S < 38 28 < 38 < 38 < 38 44 1009 894 < 38 291 < 38
Ba 669 492 697 446 675 614 785 676 724 641 672
Pb 5 33 39 33 30 <4 <3 <3 9 16 <3
Major elements are in wt. %, traces in ppm <7: below limit.

19
20 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30

Fig. 10. Harker-type diagrams of whole rock chemical analyses of the FNVC igneous rocks. The suite of samples defines a trend in the high-K
calc-alkaline series, with some samples being shoshonitic. Most elements exhibit systematic variation with the SiO2 content (e.g., Na2O, Al2O3,
Fe2O3). Copper appears to decrease with increasing differentiation but the trend is not systematic.

trations of some elements (K, Al, Zr) is larger in variation holds for all the other elements, including
volcanic than in intrusive rocks (Table 4). Cu.
The concentrations of SiO2 in volcanic samples To investigate the variation of volcanic and intru-
from the Main Zone are plotted in Fig. 7 as a function sive rocks over time, we plotted the SiO2 content as a
of their stratigraphic position. Consistent with petro- function of the measured or inferred ages (Fig. 11).
graphic observations, the silica content of volcanic From this plot, it appears that, throughout the mag-
rocks does not vary systematically over time, and the matic history, melts with highly variable SiO2 con-
vague trend towards increasing SiO2 content is prob- tents were present at any given time. Prior to 8.5 Ma,
ably due to a sampling bias. The same non-systematic no systematic evolution of the silica content is appar-
 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30 21

Table 4
Isotopic compositions of intrusive and volcanic rocks
87
Samples Rb (ppm) Sr (ppm) Sr/86Sr Sm (ppm) Nd (ppm)
143
Nd/144Nd
Samples from this study
23-3(99) 117.9 699.0 0.70599 4.4 23.2 0.512506
9-9(99) 30.3 238.4 0.70863 5.0 23.9 0.512434
NB 4(99) 109.0 465.0 5.5 27.0 0.512443
NB 5(99) 88.0 528.8 0.70472 5.9 26.9 0.512669
NB 8(99) 99.9 329.3 0.70678 4.5 22.3 0.512430
NB 9(99) 99.4 463.0 0.70734 5.5 27.3 0.512448
NB 13(99) 117.5 469.8 0.70720 4.9 23.0 0.512500
NB 14(99) 106.7 399.2 0.70806 5.3 26.3 0.512419
NB 15(99) 117.1 451.4 0.70669 5.1 25.6 0.512430
NB 16(99) 134.7 422.1 0.70681 5.7 28.5 0.512436
NB 2(00) 106.0 477.0 0.70799 5.3 22.6
NB 5(00) 76.0 568.0 0.70532 3.4 17.6 0.512486
NB 6(00) 88.4 410.5 0.70734 3.6 21.8 0.512394
NB 13(00) 82.0 471.0 4.5 20.0 0.512492
NB 18(00) 5.7 39.4 0.70714 4.3 20.1 0.512242
NB 25(00) 4.7 288.1 0.70639 5.6 27.6 0.512488
NB 31(00) 14.5 49.7 0.70546 3.5 17.1 0.512551
NB 32(00) 37.0 338.2 0.70584 4.3 21.5 0.512541
NB 43(00) 33.9 233.0 0.70785 5.3 26.1 0.512483
NB 49(00) 154.0 356.7 0.70625 5.8 22.8
NB 55(00) 276.8 412.0 0.70781 7.2 35.2 0.512420
NB 57(00) 135.4 487.3 0.70718
NB 62(00) 64.1 278.6 0.70732 4.7 21.7
NB 69(00) 80.0 557.6 0.70742 5.1 24.3 0.512444
NB 70(00) 96.0 392.0 8.1 45.6 0.512484
NB 73(00) 7.9 293.9 0.71182 3.7 18.9 0.512489
NB 74(00) 27.1 294.6 0.70721 4.7 23.1
NB 76(00) 5.8 567.0 0.70860 5.0 23.0 0.512565
DU 18-8(99) 56.0 317.9 4.5 22.6 0.512492
DU 21-6(99) 46.5 460.5 0.70641 4.4 20.1 0.512473
CAP 9 103.7 443.0 0.70789 4.2 24.4
CAP 11 62.9 612.0 0.70604 5.9 23.7 0.512521
BLA 67 68.0 737.5 0.70678 5.1 26.7 0.512500
BLA 48 48.5 717.5 0.70650 4.5 22.9
43-47.1 (22) 3.9 17.9 0.512499
2j unertainty F 0.000015 F 0.000015

87
Samples from Sasso (1997) Rb (ppm) Sr (ppm) Sr/86Sr Pb (ppm) 206
Pb/204Pb 207
Pb/204Pb 208
Pb/204Pb

201 108 529 0.70567 13 18.716 15.649 38.746

204 78 401 0.70737 10 18.731 15.644 38.705
206 152 692 0.70809 7 18.705 15.644 38.812
216 91 549 0.70612 6 18.884 15.662 38.866
229 76 623 0.70624 9 18.657 15.609 38.687
238 95 601 0.70527 6 18.724 15.634 38.752
244 108 479 0.70581 7 18.708 15.653 38.833
269 183 169 0.70754 6 19.059 15.684 38.834
271 38 511 0.70542 4 18.622 15.633 38.567
284 96 559 0.70672 4 18.670 15.654 38.839
288 86 658 0.70573 9 18.712 15.639 38.808
296 93 712 0.71026 15 18.613 15.644 38.775
311 98 600 0.70629 15 18.651 15.638 38.642
312 44 606 0.70711 10 18.681 15.664 38.813
2j unertainty F 0.005 F 0.005 F 0.012
Rb, Sr, Sm and Nd from this study were determined using TIMS.
Sr, Rb and Pb analyses of the samples form Sasso (1997) were obtained by XRF.
2 j uncertainties on the 87Sr/86Sr and the 143Nd/144Nd ratio: F 0.000015.
22 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30

Fig. 11. Variation of the SiO2 content of intrusive and extrusive rocks as a function of time (dated samples and inferred ages from Fig. 7). Early
variations are non-systematic, but after 8.5 Ma, the most silica-rich intrusive rocks follow a linear increase in the degree of differentiation with
time.

ent. After ca. 8.5 Ma, the most silica-rich intrusions to concentrate the REE, Sr and Pb; Sm-Nd was
follow a systematic evolution towards more evolved separated on a second column containing teflon pow-
compositions up to the end of the magmatic activity. der coated with di-2-ethyl-hexyl-orthophosphoric ac-
This trend defines an upper boundary to the SiO2 id. Sr was cleaned with a second small cation
content of all magmas; volcanic rocks have silica exchange column. Sm and Nd were loaded using
contents that are generally less than or equal to the HCl on double Re filaments and measured as metal.
intrusive rocks. After the formation of the Alumbrera Nd isotopic ratios were normalized to 146Nd/144Nd =
stock around 6.8 Ma, evidence for any further mafic 0.7219. Analytical reproducibility was estimated by
to intermediate magmatism is missing. periodical measurements of the La Jolla standard (Nd)
as well as the NBS 987 (Sr). The mean of 12 runs
during this work was 143Nd/ 144 Nd = 0.511841 F
7. Isotopic compositions 0.000007. Sr was loaded with HNO3 on Ta filaments;
Sr ratios were normalized to 87Sr/86Sr = 8.37521. Ten
To help identify the origin of contributing magma runs of the NBS 987 standard show an 87Sr/86Sr ratio
components, Sr, Nd and Pb isotopic compositions were of 0.710235 F 0.000006. Pb was loaded with silica
determined for selected samples from the complex. gel and H3PO4 on single Re filaments and measured
Sample locations and a brief description are given in with Faraday cups in static mode. Blanks for the entire
Table 1; detailed information is documented in Sasso procedure were < 1 ng Pb. Standardization was done
(1997), Breitenmoser (1999) and Becker (2001). with the NBS SRM 982 lead standard. The total
uncertainty is less than 0.1%.
7.1. Analytical procedure
7.2. Results
Isotopic compositions were measured at the ETH
Zurich (Sr and Nd) and the U.S.G.S., Menlo Park (for Fig. 12A compares variations in the initial Sr and
Pb) using thermal ionisation mass spectrometry on Nd isotopic compositions of whole-rock samples from
whole rock samples for simultaneous collection of the FNVC. Data from other volcanic centers in the
isotope masses. Powders were dissolved using HF and central and southern Andes, compiled by Kay et al.
HCl. and were treated in teflon bombs using HF- (1999) are shown for comparison. Variations in iso-
HNO3-HClO4. A cation-exchange column was used topic compositions are interpreted as recording con-
 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30 23

Fig. 12. (A) Initial 87Sr/86Sr vs. 143Nd/144Nd plot for samples from the FNVC. The most primitive composition was found in an extrusive rock
from the Agua Tapada area and plots on the mantle array (uncontaminated). Other samples follow a mixing trend between this primitive melt
and a crustal component. Also shown is the field of samples from other andesitic complexes in the Andes, compiled by Kay et al. (1999). (B – F)
Plots of initial 87Sr/86Sr, 143Nd/144Nd, 206Pb/204Pb, 207Pb/204Pb and 208Pb/204Pb against SiO2. Each of these plots shows a correlation between
the isotopic composition and the SiO2 content in volcanic rocks (arrows), but a more blurred distribution for the intrusive rocks. 2r uncertainties
are smaller than symbol sizes.

tamination of a mantle melt with basement rocks (e.g., lution from the bottom to the top of the volcanic pile
Francis et al., 1980; James, 1982; Barreiro and Clark, (Fig. 7). However, despite the small total variation, the
87
1984; Colucci et al., 1991; McDowell et al., 1999) and Sr/86Sr ratio measured in volcanic rocks shows an
the correlated variations suggest that contamination in increase with increasing SiO2 content (Fig. 12). This
both isotopic systems occurred simultaneously. In the trend is not apparent in intrusive rocks for which
FNVC, samples with the smallest radiogenic compo- strontium isotopic composition appears to be indepen-
nent (Cap 11 and 271) are labeled in Fig. 12. Exten- dent of the chemical composition. A similar conclu-
sion of this trend towards the most mafic rock of the sion can be drawn for Nd and Pb isotopic compositions,
complex (42 wt.% SiO2) yields a 87Sr/86Sr ratio of i.e., volcanic rocks appear to follow a systematic
0.705 and a 143Nd/144Nd ratio of 0.51260. evolution as a function of the SiO2 content, but intru-
The Sr and Nd isotopic compositions are consistent sive rocks do not. As most volcanic rocks analyzed
with the randomly variable major and trace element were formed before 8.0 Ma and most analyzed intru-
concentrations and do not show any systematic evo- sions were emplaced after 8.0 Ma, the correlation
24 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30

Fig. 13. Plot of the Sr isotopic ratio vs. the SiO2/TiO2 ratio in intrusive and extrusive rocks. No correlation is apparent between the degree of
crustal contamination (indicated by the Sr isotopic ratio) and the ratio between the immobile Ti and Si, suggesting that contamination did not
occur through a partial melting of the crust, nor through a hydrothermal fluid.

between the isotopic composition and the SiO2 content Alto de la Blenda intrusion. The first intrusive rocks
could also be related to an evolution of the system over are the Tampa Tampa andesites, which were
time. emplaced at 9.0 F 0.3 Ma. These rocks contain
The nature of the contaminant was investigated by biotite and have a similar composition as biotite-
comparing the behavior of an immobile element (Ti) bearing breccias of the same age on the western
with that of Si as a function of the Sr isotopic ratio. flank of Cerro Durazno. It is thus likely that this
There is no correlation between the 87Sr/86Sr ratio and and similar intrusions represent conduits for the
the TiO2/SiO2 ratio in either the intrusive or the extrusive rocks on the flanks of the main stratocone.
extrusive rocks (Fig. 13). Similar relationships hold This relationship is consistent with higher than
if Zr is considered instead of Ti or if Si is replaced by average growth rates of the volcanic pile in this
K or Rb. area compared to other parts of the complex. Later
intrusions include the Bajo el Durazno basaltic
andesite and the Chilca andesite porphyries. The
8. Discussion: geological evolution of the FNVC intrusion of the Alto de la Blenda monzonite stock
occurred at the end of the volcanic activity in the
Fig. 14 shows and idealized cross section through central part of the system. Pyroclastic dacite brec-
the FNVC, based on our field and geochronological cias at the northwestern end of the complex are the
observations. This reconstruction was established us- youngest volcanic units. These have a composition
ing the profiles of Fig. 2, age determinations, estimat- nearly identical to that of the Agua Tapada stock,
ed thicknesses and a minimum depth of 3 km which probably represents the remnant of a satellite
estimated for the Alumbrera porphyry, based on conduit on the flank of the main vent. Evidence for
lithostatic pressures of 110 MPa from fluid inclusion a relatively shallow depth of emplacement of the
microthermometry (Ulrich and Heinrich, 2002). Agua Tapada intrusions is given by the vuggy silica
Based on this reconstruction, the following history alteration. This alteration forms in the upper part of
of magmatism and hydrothermal activity can be magmatic-hydrothermal systems and thus implies a
interpreted. thin overlying volcanic pile and emplacement on the
Magmatism was initiated at ca. 9.6 Ma by local intrusion at the periphery of the stratocone. Con-
extension in a regionally compressive regime (Sasso temporaneous with this event was the sector col-
and Clark, 1998). The early evolution of the system lapse of a caldera at the northwestern extremity of
is characterized by a rapid growth of the stratocone, the system (Fig. 9). This marked the end of the
centered approximately on the location of the later volcanic activity, after which no further magma was
 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30 25

Fig. 14. Schematic but correctly scaled reconstruction of the Farallón stratovolcano. The thickness of volcanic units is based on the cross
sections of Fig. 2 and schematically extrapolated across the volcano. The intrusions present in the mapped areas are also shown. The correlation
between ages and thickness of volcanic units is based on Figs. 6 and 7 and radiometric ages. Additional constraints are given by the depth of the
Alumbrera intrusions and the relative depth of the alteration pattern in Alumbrera, Agua Tapada and Las Pampitas.

extruded and only andesitic to rhyolitic intrusions compositions (< 45% SiO2) to rhyodacitic composi-
formed. tions (>70% SiO2). This is difficult to explain by
The first of these andesite to rhyolite intrusions is fractional crystallization of the observed phenocryst
the small Las Pampitas stock at about 7.22 Ma. This phases (spinel-magnetite, pyroxene, hornblende, bio-
is followed by the polyphase Alumbrera stock with tite, plagioclase and quartz) from any of the observed
its porphyry copper system between 7.0 and 6.75 bulk rock compositions. Rather, these linear trends
Ma. The region was again under a compressive suggest that intermediate rock compositions result
regime after the cessation of volcanic activity (Fig. from the mixing of mafic and felsic magma. Magma
1) as shown by structural evidence. It is possible that mixing has previously been suggested as a mechanism
the return from local extension to compression to generate magmas of andesitic compositions, mainly
towards the end of the magmatic activity contributed based on petrographic evidence from mafic enclaves
to the evolution from mainly volcanic to mainly (Mandeville et al., 1996; Wolf and Eichelberger, 1997;
intrusive activity and, eventually, to cessation of Coombs et al., 2000; Cole et al., 2001) and has been
magmatism. proposed for the FNVC by Sasso (1997). Indications
Throughout this evolution, bulk rock compositions for such mixing processes are mafic enclaves in felsic
define a near-linear array for nearly all major and trace rocks and crenulated contacts between the two magma
elements (Fig. 10), extending from extremely basic types (Fig. 4). Isotopic data suggest that the potential
26 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30

sources for the mixing components are a mafic, mantle- By contrast, the presence of a significant magma
derived magma and a silica-rich magma with a signif- chamber is required at the end of the lifetime of the
icant crustal component. volcanic complex to allow the formation of the
Isotopic data indicate that almost all samples have porphyry copper deposit (Ulrich and Heinrich,
incorporated variable amounts of a component more 2002). After 8.5 Ma, the systematic evolution in the
radiogenic than the primitive mantle source. The SiO2 content (Fig. 11) probably reflects the progres-
isotopic variations correlate with each other and with sive establishment of a continuously evolving sub-
the concentrations of all major elements, including volcanic magma chamber. This magma chamber is
SiO2, suggesting that the silica-rich magma was the likely to be heterogeneous, given the variability in
one carrying the strongest crustal signature (Fig. 12). melt composition at any one time, and the systematic
This correlation is mainly observed at an early stage SiO2 increase is only visible in the most differentiat-
of the evolution; it is absent in late intrusive rocks ed part of the systems. The presence of mafic
even though isotopic ratios indicate similar degrees of enclaves in some dacitic intrusions shows that con-
contamination for both rock types. A possible expla- trasting magmas are mixing upon emplacement.
nation for this observation is that additional mixing Thus, we suggest that the increase in the SiO2 content
and small degrees of fractionation within the magma of the most silica-rich intrusions in each stock reflects
chamber erased the original correlation. Consistent an increasing proportion of felsic melt at the roof of
with this interpretation is the long-term evolution of the magma chamber over time and, thus, an increas-
intrusive rocks indicated by the continuous increase in ing proportion of felsic magma in younger stocks.
the SiO2 content of the most felsic intrusions over This also suggests that the magma chamber had to
time (Fig. 11). evolve over some 2.4 m.y. (between 8.5 and 6.1 Ma)
Contamination of the silica rich melt through partial to account for this continuous evolution in the melt
melting of the crust would tend to enrich incompatible composition.
elements preferentially over compatible elements in the
melt, but Fig. 13 suggests that in volcanic rocks Si and
Ti were added together to the melt. We consider this 9. Formation of the porphyry deposit
plot and the correlated variations in Sr and Nd isotopic
compositions as an indication that no selective con- Early intrusions emplaced during volcanic activity
tamination (i.e., by partial melting or through a fluid are only slightly altered and mineralized, regardless of
phase) has occurred. No additional contamination their level of exposure. Intensely altered intrusions are
specific to intrusive rocks is apparent. found in the Agua Tapada, the Las Pampitas and the
Based on the high variability and the non-system- Alumbrera stocks, all emplaced after the main volcanic
atic evolution in the bulk rock chemistry investiga- activity had ceased. Thus, it appears that the extensive
tions of the Tatara –San Pedro volcanic complex in degassing associated with the volcanic activity pre-
the Southern Volcanic Zone, Chile, Dungan et al. vented the release of large amounts of fluid into
(2001) suggest that andesitic stratovolcanoes might individual stock to allow the formation of a major
not be underlain by large magma chambers. During deposit during volcanic activity. After significant vol-
the early stages of the magmatic evolution, the non- canic activity had stopped fluid release from the
systematic variation of the SiO2 content of intrusive magma chamber could be focused into single intrusive
and extrusive rocks as a function of time (Fig. 11) events and cause large-scale alteration. The three
and the variation in radiogenic isotopic compositions highly altered intrusions are notably aligned along a
are compatible with this concept. Indeed, we interpret NW – SE trend with progressively younger ages to-
the early high variability in magma chemistry (Fig. wards the SE. This contrast with the otherwise random
11) to represent a stage corresponding to the gradual distribution of intrusions over time and presumably
initiation of a magma chamber in the upper crust indicates a change in the local tectonic constraints
through multiple pulses of mantle-derived melts, from extensional to compressional. This could explain
forming a network of dikes and sills at a level of the cessation of the volcanic activity and of associated
neutral buoyancy. extensive degassing.
 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30 27

The release of enough Cu and Au to the ore- fluid from the system, as later intrusions are essen-
forming fluid that generated the Alumbrera deposit tially unaltered.
requires the presence of a large upper crustal magma
chamber of at least 7 km3 (Halter et al., in press).
This magma chamber was only established after the 10. Summary and conclusions
volcanic system had been active for a certain period
of time and, thus, precursor volcanic activity appears The extrusive activity of andesitic volcanoes is
as a pre-requisite to the formation of a large por- extremely complex and cannot be precisely recon-
phyry system. structed by mapping selected cross-sections, because
From the three, mainly dacitic stocks emplaced locations of flows and breccia deposits vary at the
after volcanism had ceased, only Alumbrera shows scale of years to thousands of years in response to the
economic ore concentrations. The Agua Tapada and actively evolving topography of the volcano (e.g.,
Las Pampitas stocks are highly altered, but explo- Dungan et al., 2001). On the longer time scale of
ration drilling showed no significant mineralization. the entire life of an andesite volcano, however, the
This difference might simply relate to the level of study of an extinct and deeply eroded system such as
erosion. Agua Tapada shows intense pervasive ser- the Farallón Negro Volcanic Complex offers the
icitic alteration and radially distributed veins with advantage that evidence from the integrated extrusive
vuggy silica alteration. Both these alteration types and intrusive magmatic activity can be combined and
are characteristic for shallow hydrothermal activity, correlated. This less precise but broader view can
consistent with the emplacement of this stock at the yield critical information about the long-term evolu-
periphery of the system. The center of the Las tion of subvolcanic magma chambers and their role in
Pampitas stock is formed of a massive, barren the formation of hydrothermal ore deposits.
quartz-magnetite altered stock at the erosion level. Published and new field data, chemical and isoto-
This type of alteration is typical for the deeper part pic analyses as well as 40Ar – 39Ar geochronology
of a porphyry system, consistent with the emplace- provide a comprehensive data set for reconstructing
ment of this stock under a thick volcanic pile in the the evolution of the Farallón Negro Volcanic Com-
middle of the complex. Thus, it is possible that plex. During the first f 1.2 Ma of dominantly
these stocks had the same potential to form eco- eruptive activity, there is no evidence for the existence
nomic ore deposits, but in the case of Agua Tapada, of a major magma chamber, even though a magma
the ore would be buried under several hundred reservoir may already have been gradually forming by
meters of sericitic alteration and in the case of continued andesite eruption through a central conduit.
Las Pampitas, the mineralization would already be Non-systematic variation of basaltic andesite to an-
eroded. desite composition of the extrusive rocks, and tight
Assuming that successive intrusions in a given linear correlations of the concentrations of all major
stock extract magma from progressively deeper parts and trace elements in all igneous rocks suggest that
of the magma chamber, the progressive evolution the compositional variability was dominated by mix-
from more silica-rich intrusions towards more mafic ing between mafic and felsic magmas. From about 8.5
intrusions, such as observed at Alumbrera, suggests Ma onward, an upper crustal magma chamber was
that the magma chamber was layered. Pulses of probably established, from which subvolcanic stocks
magmatic-hydrothermal magnetite and K-feldspar + and equivalent volcanic rocks in the northeastern part
biotite + magnetite alteration alternated with multiple of the volcano were extracted. The main evidence for
intrusive pulses within the Alumbrera stock. Metals the existence of a magma reservoir, probably extend-
necessary to generate the ore fluid were released ing under much of the Main Zone (> 15 km horizontal
from sulfide melts, present as relict inclusions in diameter), is a gradual evolution in the composition of
amphibole from andesites, emplaced after ore forma- increasingly felsic subvolcanic intrusions between 8.5
tion (Halter et al., 2002). The alteration events at and 6.1 Ma. As a major event terminating the volcanic
Agua Tapada, Las Pampitas and eventually Alum- activity, deposition of copious quartz-bearing dacite
brera are inferred to have exhausted much of the breccias accompanied a partial caldera collapse.
28 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30

Starting some 0.4 Ma after termination of the main References

extrusive activity and, hence, following a period of
quiescence, the intrusion of multiple pulses of dacitic Allison, A.E., 1984. The geology of the Bajo el Durazno porphyry
to andesitic magma into a composite stock generated copper-gold prospect, Catamarca Province, Argentina. M.Sc.
thesis, University of Arizona, Tuscon., 167 pp.
the porphyry Cu-Au deposit of Bajo de la Alumbrera.
Allmendinger, R.W., 1986. Tectonic development, southeastern
Ore formation probably occurred as a result of pro- border of the Puna plateau, Northwestern Argentine Andes.
gressive solidification of the magma chamber and the Geological Society of America Bulletin 97 (9), 1070 – 1082.
exsolution of a large volume of magmatic-hydrother- Bain, N., 2001. Petrological and geochemical contribution to the
mal brine. Mass-balance requirements imposed by the Farallón Negro Volcanic Complex, NW-Argentina. Diploma
thesis, Swiss Federal Institute of Technology, Zurich, 92 pp.
metal budget of the deposit and analyzed ore fluid
Barreiro, B.A., Clark, A.H., 1984. Lead isotopic evidence for
compositions (Ulrich and Heinrich, 2002) necessitate evolutionary changes in magma – crust interaction, central
that a minimum of 0.6 km3 of magmatic-hydrothermal andes, Southern Peru. Earth and Planetary Science Letters 69
brine was extracted from a magma reservoir of at least (1), 30 – 42.
7 km3 within a short period of time ( < 0.2 Ma), and Becker, K., 2001. Sm-Nd and Rb-Sr isotope geochemistry on the
magmatic source of the igneous rocks in the Farallón Negro
focused upward through the internally fractured por-
Volcanic Complex, NW Argentina. Diploma thesis, Swiss Fede-
phyry stock. Our evidence suggests that a long-term ral Institute of Technology, Zurich, 103 pp.
volcanic evolution over several million years may Breitenmoser, T., 1999. Geology and geochemistry of the calc-al-
have been essential to prepare the magmatic system kaline Farallón Negro Volcanic Complex at Capillitas, NW-
for the final event of magma chamber crystallization Argentina. Diploma thesis, Swiss Federal institute of Techno-
logy, Zurich, 105 pp.
and hydrothermal ore fluid exsolution.
Caelles, J.C., Clark, A.H., Farrar, E., McBride, S.L., Quirt, S.,
1971. Potassium-argon ages of porphyry copper deposits and
associated rocks in the Farallon Negro-Capillitas District, Cata-
Acknowledgements marca, Argentina. Economic Geology and the Bulletin of the
Society of Economic Geologists 66 (6), 961 – 964.
Clark, A.H., Sasso, A.M., 2000. ‘‘The shoshonite porphyry Cu-Au
This study was partly supported by ETH research
association at Bajo de la Alumbrera, Catamarca Province,
funds and by the Swiss National Science Foundation Argentina’’, by D. Muller and P. Forrestal: a discussion. Minera-
(grant 20-59544-99 to C.H.). Extensive logistic logy and Petrology 68 (4), 297 – 304.
support was granted by Minera Alumbrera Ltd. and Clark, A.H., et al., 1976. Longitudinal variations in the metalloge-
Mount Isa Mines and we would like to thank Dave netic evolution of the central Andes; a progress report, Metal-
logeny and plate tectonics. Geological Association of Canada,
Keough and Steve Brown in particular for their help.
Toronto, Canada, pp. 23 – 58.
Mario Alderete, from IMAD (Tucumán, Argentina), Cline, J.S., Bodnar, R.J., 1991. Can economic porphyry copper
provided authorization to access the Farallón Negro mineralization be generated by a typical calc-alkaline melt. Jour-
exploration prospect. The authors would like to nal of Geophysical Research-Solid Earth and Planets 96 (B5),
acknowledge the help of Anthony Harris (University 8113 – 8126.
Cole, J.W., Gamble, J.A., Burt, R.M., Carroll, L.D., Shelley, D.,
of Queensland, Australia) in the interpretation and
2001. Mixing and mingling in the evolution of andesite-dacite
drawing of the map, Eduardo Llambı́as (University of magmas; evidence from co-magmatic plutonic enclaves, Taupo
La Plata, Argentina) and Daniel Rubiolo (Servicio Volcanic Zone, New Zealand. Lithos 59 (1 – 2), 25 – 46.
Geológico Minero Argentino, SEGEMAR) for con- Colucci, M.T., Dungan, M.A., Ferguson, K.M., Lipman, P.W.,
structive discussions and insight into the regional Moorbath, S., 1991. Precaldera lavas of the southeast San-Juan
volcanic field-parent magmas and crustal interactions. Journal
geology. Analytical and field data from the earlier
of Geophysical Research-Solid Earth and Planets 96 (B8),
research of A.M.S. incorporated herein was funded by 13413 – 13434.
grants from the Natural Sciences and Engineering Coombs, M.L., Eichelberger, J.C., Rutherford, M.J., 2000. Mag-
Council of Canada to A.H.C. and Edward Farrar, ma storage and mixing conditions for the 1953 – 1974 erup-
constituting a contribution to the Queen’s University tions of Southwest Trident Volcano, Katmai National Park,
Alaska. Contributions to Mineralogy and Petrology 140 (1),
Central Andean Metallogenetic Project. The manu-
99 – 118.
script has benefited from constructive comments by Coughlin, T.J., O’Sullivan, P.B., Kohn, B.P., Holcombe, R.J.,
Jake Lowenstern, John Dilles, Shan De Silva, David 1998. Apatite fission-track thermochronology of the Sierras
John and two anonymous reviewers. Pampeanas, central western Argentina; implications for the
 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30 29

mechanism of plateau uplift in the Andes. Geology (Boulder) Landtwing, M.R., Dillenbeck, E.D., Leake, M.H., Heinrich, C.A.,
26 (11), 999 – 1002. 2002. Evolution of the breccia-hosted porphyry Cu-Mo-Au de-
Dawson, S.E., 1994. The occurrence of gold at the Bajo de la posit at Agua Rica, Argentina: progressive unroofing of a mag-
Alumbrera porphyry copper-gold deposit, Northwestern Argen- matic hydrothermal system. Economic Geology and the Bulletin
tina. Masters thesis, University of Arizona, Tuscon, 135 pp. of the Society of Economic Geologists 97 (6), 1273 – 1292.
Dilles, J.H., 1987. Petrology of the Yerington Batholith, Nevada— Llambias, E.J., 1970. Geologia de los Yacimientos Mineros Agua
evidence for evolution of porphyry copper ore fluids. Economic de Dionisio, Prov. de Catamarca, Rep. Argentina. Revista de la
Geology 82 (7), 1750 – 1789. Asociacion Argentina de Mineralogia, Petrologia y Sedimento-
Dungan, M.A., Wulff, A., Thompson, R., 2001. Eruptive stratigra- logia 1, 2 – 32.
phy of the Tatara – San Pedro complex, 36 degrees S, southern Llambias, E.J., 1972. Estructura del grupo volcanico Farallon Ne-
volcanic zone, Chilean Andes: Reconstruction method and gro, Catamarca, Republica Argentina. Revista de la Asociacion
implications for magma evolution at long-lived arc volcanic Geologica Argentina 27 (2), 161 – 169.
centers. Journal of Petrology 42 (3), 555 – 626. Mandeville, C.W., Carey, S., Sigurdsson, H., 1996. Magma mixing,
Francis, P.W., Thorpe, R.S., Moorbath, S., Kretzschmar, G.A., fractional crystallization and volatile degassing during the 1883
Hammill, M., 1980. Strontium isotope evidence for crustal con- eruption of Krakatau volcano, Indonesia. Journal of Volcano-
tamination of calc-alkaline volcanic rocks from Cerro Galan, logy and Geothermal Research 74 (3 – 4), 243 – 274.
Northwest Argentina. Earth and Planetary Science Letters 48 Martinez, L., Meilán, D., Maza, A.E., 1995. Mapa geológico de la
(2), 257 – 267. provincia de Catamarca, República Argentina. Secretaria de
Guilbert, J.M., 1995. Geology, alteration, mineralization, and gene- Minerı́a, Dirección Nacional del Secvicio Geológico.
sis of the Bajo de la Alumbrera porphyry copper-gold deposit, McBride, S.L., Caelles, J.C., Clark, A.H., Farrar, E., 1976. Paleo-
Catamarca Province, Argentina. In: Pierce, F.W., Bolm, J.G. zoic radiometric age provinces in the Andean basement, Lati-
(Eds.), Porphyry copper deposits of the American Cordillera. tudes 25j – 30j. Earth and Planetary Science Letters 29 (2),
Arizona Geological Society Digest. Arizona Geological Society, 373 – 383.
Tucson, pp. 646 – 656. McDowell, F.W., Housh, T.B., Wark, D.A., 1999. Nature of the crust
Gustafson, L.B., Hunt, J.P., 1975. The porphyry copper deposit at beneath west-central Chihuahua, Mexico, based upon Sr, Nd, and
El Salvador, Chile. Economic Geology and the Bulletin of the Pb isotopic compositions at the Tomochic volcanic center. Geo-
Society of Economic Geologists 70 (5), 857 – 912. logical Society of America Bulletin 111 (6), 823 – 830.
Halter, W.E., Pettke, T., Heinrich, C.A., 2002. The origin of Cu/ Mirre, J.C., Aceñolaza, F.G., 1972. El hallazgo de Oldhamia sp.
Au ratios in porphyry-type ore deposits. Science 296 (5574), (traza fosil) y su valor como evidencia de edad Cambrica para el
1844 – 1846. supuesto Precambrico del borde occidental del Aconquija, Prov.
Halter, W.E., Heinrich, C.A., Pettke, T., 2004. Laser-ablation ICP- de Catamarca. Ameghiniana 9, 72 – 78.
MS analysis of silicate and sulfide melt inclusions in an ande- Muller, D., Forrestal, P., 1998. The shoshonite porphyry Cu-Au
sitic complex II: magmas genesis and implications for ore-for- association at Bajo de la Alumbrera, Catamarca Province,
mation. Contributions to Mineralogy and Petrology (in press). Argentina. Mineralogy and Petrology 64 (1 – 4), 47 – 64.
Herzarkhani, A., Williams-Jones, A.E., Gammons, C.H., 1999. Fac- Proffett, J.M., 1995. Geology of the Bajo de la Alumbrera porphyry
tors controlling copper solubility and chalcopyrite deposition in copper-gold deposit, Catamarca province, Argentina. Minera
the Sungun porphyry copper deposit, Iran. Mineralium Deposita Alumbrera Internal Report, vol. 2.
34 (8), 770 – 783. Proffett, J.M., 1997. Geology of the Bajo de la Alumbrera porphyry
Hug, A., 1999. Petrography and genesis of the Capillitas Diat- copper-gold deposit, Catamarca province, Argentina. Minera
reme, Farallón Negro Volcanic Complex, NW-Argentina. Di- Alumbrera Internal Report.
ploma thesis, Swiss Federal Institute of Technology, Zurich, Sandeman, H.A., Archibald, D.A., Grant, J.W., Villeneuve, M.E.,
57 pp. Ford, F.D., 1999. Characterization of the chemical composition
James, D.E., 1982. A combined O, Sr, Nd, and Pb isotopic and trace and 40Ar – 39Ar systematics of intralaboratory standard MAC-83
element study of crustal contamination in central Andean lavas: biotite, Radiogenic age and isotopic studies; report. Current
I. Local geochemical variations. Earth and Planetary Science Research-Geological Survey of Canada, vol. 12. Geological
Letters 57 (1), 47 – 62. Survey of Canada, Ottawa, ON, Canada, pp. 13 – 26.
Jordan, T.E., Allmendinger, R.W., 1986. The Sierra Pampeanas of Sasso, A.M., 1997. Geological evolution and metallogenetic rela-
Argentina: a moderne analogue of Rocky Mountain foreland tionships of the Farallon Negro volcanic complex, NW Argen-
deformation. American Journal of Science 260, 737 – 764. tina. PhD thesis, Queen’s University, Kingston, Ontario, 843 pp.
Kay, S., Mpodozis, C., Coira, B., 1999. Neogene magmatism, tec- Sasso, A.M., Clark, A.H., 1998. The Farallón Negro Group,
tonism, and mineral deposits of the Central Andes (22j to 33j S northwest Argentina: magmatic, hydrothermal and tectonic
Latitude). In: Skinner, B.J. (Ed.), Geology and Ore Deposits of evolution and implications for Cu-Au metallogeny in the
the Central Andes. Special Publication, Society of Economic Andean back-arc. Society of Economic Geologists Newsletter
Geology, Littleton, CO, pp. 27 – 59. 34, 8 – 17.
Keith, J.D., et al., 1997. The role of magmatic sulfides and mafic Sillitoe, R.H., 1973. The tops and bottoms of porphyry copper
alkaline magmas in the Bingham and Tintic mining districts, deposits. Economic Geology and the Bulletin of the Society
Utah. Journal of Petrology 38 (12), 1679 – 1690. of Economic Geologists 68 (6), 799 – 815.
30 W.E. Halter et al. / Journal of Volcanology and Geothermal Research 136 (2004) 1–30

Sillitoe, R.H., 1993. Gold-rich porphyry copper deposits; geological brera, Argentina (vol 96, pg 1719, 2001). Economic Geology
model and exploration implications. In: Kirkham, R.V., Sinclair and the Bulletin of the Society of Economic Geologists 97
W.D. Thorpe, R.I., Duke, J.M. (Eds.), Mineral deposit mode- (8), 1863 – 1888.
lingSpecial Paper-Geological Association of Canada Geological Ulrich, T., Gunther, D., Heinrich, C.A., 1999. Gold concentrations
Association of Canada, Toronto, ON, Canada, pp. 465 – 478. of magmatic brines and the metal budget of porphyry copper
Stults, A., 1985. Geology of the Bajo de la Alumbrera porphyry deposits. Nature 399 (6737), 676 – 679.
copper and gold prospect, Catamarca province, Argentina. Mas- Ulrich, T., Gunther, D., Heinrich, C.A., 2002. Evolution of a por-
ters thesis, University of Arizona, Tuscon, 75 pp. phyry Cu-Au deposit, based on LA-ICP-MS analysis of fluid
Suchomel, B.J., 1985. Geology of the Bajo de Agua Tapada base inclusions: Bajo de la Alumbrera, Argentina (vol 96, pg 1743,
and precious metal prospect, Catamarca Province, Argentina. 2001). Economic Geology and the Bulletin of the Society of
M.Sc thesis, University of Arizona, Tuscon, 75 pp. Economic Geologists 97 (8), 1888 – 1920.
Turner, J.C., 1962. Estratigrafia de la region al naciente de Laguna Urreiztieta, M., Rosello, E.A., Gapais, D., LeGorve, C., Cobbold,
Blanca (Catamarca). Revista Asociation Geologica Argentina P.R., 1993. Neogene dextural transpression at the southern
17, 11 – 46. edge of the Altiplano-Puna (NW Argentina). II International
Turner, J.C., 1973. Descripcı́on geológica de la Hoja 13b, Cha- Symposium on Andean Tectonics, Oxford. Editions de l’Office
schuil. Dirección General de Industria Minera, 78. de la Recherche Scientifique et Technique d’Outre-mer, Paris,
Ulrich, T., 1999. Genesis of the Bajo de la Alumbrera porphyry Cu- pp. 267 – 269.
Au deposit, Argentina: Geological, fluid geochemical, and iso- Wolf, K.J., Eichelberger, J.C., 1997. Syneruptive mixing, degas-
topic implications. Doctoral thesis, Swiss Federal Institute of sing, and crystallization at Redoubt Volcano, eruption of De-
Technology, Zurich, 207 pp. cember, 1989 to May 1990. Journal of Volcanology and
Ulrich, T., Heinrich, C.A., 2002. Geology and alteration geo- Geothermal Research 75 (1 – 2), 19 – 37.
chemistry of the porphyry Cu-Au deposit at Bajo de la Alum-