PUBLICATIONS

Geochemistry, Geophysics, Geosystems

RESEARCH ARTICLE Sulfur isotope evolution in sulfide ores from Western Alps:
10.1002/2014GC005459
Assessing the influence of subduction-related metamorphism
Key Points: Fabio Giacometti1, Katy A. Evans2, Gisella Rebay1, John Cliff3, Andrew G. Tomkins4,
 Fluid circulation and deformation
Piergiorgio Rossetti5, Gloria Vaggelli6, and David T. Adams7
favor the reequilibration of sulfur
isotopes 1
Dipartimento di Scienze della Terra e dell’Ambiente, Universita degli Studi di Pavia, Pavia, Italy, 2School of Applied Geol-
 Negligible fluid circulation and sulfur
release occurred during ogy, Curtin University, Perth, Western Australia, Australia, 3Centre for Microscopy, Characterization and Analysis, University
metamorphism of Western Australia, Perth, Western Australia, Australia, 4School of Geosciences, Monash University, Melbourne, Victoria,
 The mobilization of metals during Australia, 5Dipartimento di Scienze della Terra, Universita di Torino, Torino, Italy, 6CNR-Istituto di Geoscienze e Georisorse,
subduction is minimal in these study Torino, Italy, 7ARC Centre of Excellence for Core to Crust Fluid Systems and GEMOC National Key Centre, Department of
cases
Earth and Planetary Sciences, Macquarie University, Sydney, New South Wales, Australia

Supporting Information:
 Readme Abstract Sulfides entering subduction zones can play an important role in the release of sulfur and met-
 Supporting Information
als to the mantle wedge and contribute to the formation of volcanic arc-associated ores. Fractionation of
stable sulfur isotopes recorded by sulfides during metamorphism can provide evidence of fluid-rock interac-
Correspondence to:
F. Giacometti, tions during metamorphism and give insights on sulfur mobilization. A detailed microtextural and geo-
[email protected] chemical study was performed on mineralized samples from two ocean floor-related sulfide deposits
(Servette and Beth-Ghinivert) in high-pressure units of the Italian Western Alps, which underwent different
Citation: metamorphic evolutions. The combination of microtextural investigations with d34S values from in situ ion
Giacometti, F., K. A. Evans, G. Rebay, probe analyses within individual pyrite and chalcopyrite grains allowed evaluation of the effectiveness of
J. Cliff, A. G. Tomkins, P. Rossetti,
G. Vaggelli, and D. T. Adams (2014), metamorphism in modifying the isotopic record and mobilizing sulfur and metals and have insights on fluid
Sulfur isotope evolution in sulfide ores circulation within the slab. Textures and isotopic compositions inherited from the protolith are recorded at
from Western Alps: Assessing the Beth-Ghinivert, where limited metamorphic recrystallization is attributed to limited interaction with meta-
influence of subduction-related
metamorphism, Geochem. Geophys. morphic fluids. Isotopic modification by metamorphic processes occurred only at the submillimeter scale at
Geosyst., 15, 3808–3829, doi:10.1002/ Servette, where local interactions with infiltrating hydrothermal fluid are recorded by metamorphic grains.
2014GC005459. Notwithstanding the differences recorded by the two deposits, neither underwent intensive isotopic reequi-
libration or records evidence of intense fluid-rock interaction and S mobilization during metamorphism.
Received 17 JUN 2014 Therefore, subducted sulfide deposits dominated by pyrite and chalcopyrite are unlikely to release signifi-
Accepted 28 AUG 2014
cant quantities of sulfur to the mantle wedge and to arc magmatism sources at metamorphic grades below
Accepted article online 9 SEP 2014
Published online 18 OCT 2014 the lower eclogite facies.

1. Introduction
Subduction zones play an important role in global element geochemical cycles. Subduction, magmatism,
and hydrothermal circulation through ocean crust are the processes that control the exchange of major,
minor, and trace elements between the Earth’s interior and the atmosphere [e.g., Scambelluri et al., 1997;
Hoffman, 1998; Plank and Langmuir, 1998; Kerrick and Connolly, 2001; Manning, 2004; Bebout, 2007]. Subduc-
tion of hydrated ocean lithosphere drives metamorphic devolatilization reactions that produce H2O and
CO2-bearing fluids, which migrate to the slab-mantle interface [Kerrick and Connolly, 2001]. H2O-rich fluids
may then transfer leached chemical species from the slab into the mantle wedge. Water release from the
slab is considered necessary for hydration of the originally anhydrous mantle wedge and facilitates partial
melting [e.g., Schmidt and Poli, 1998; Manning, 2004; Scambelluri et al., 2004].
Although H2O and CO2 are the dominant end-members present in slab-derived fluids, S and other minor
components such as Cl and F play an important role in subduction zone processes. Each of these elements
can form aqueous complexes with various metals, mobilizing normally nonreactive elements that can be of
economic interest [e.g., Alt et al., 1993; Alirezaei and Cameron, 2001, and references therein; Cooke and Hol-
lings, 2005; Manning, 2011; Spandler and Pirard, 2013]. S-bearing fluids may leach and mobilize ore metals
from sulfide deposits undergoing subduction (i.e., volcanogenic massive sulfide deposits, VMS). Such slab-
derived fluids can introduce sulfur and ore metals into the overlying mantle wedge and contribute to the

GIACOMETTI ET AL. C 2014. American Geophysical Union. All Rights Reserved.

V 3808
 Geochemistry, Geophysics, Geosystems 10.1002/2014GC005459

formation of volcanic arc-associated

ores, such as porphyry copper deposits
[e.g., Sillitoe, 1972a, 1972b; Sawkins,
1990; Evans and Tomkins, 2011; Han-
nington, 2013; Evans et al., 2014, and
references therein]. Evidence that slab-
derived fluids contaminate the overly-
ing mantle wedge and arc magmatism
sources is provided by S isotopes. High
d34S values extracted from sulfide
inclusions in eclogite-facies diamonds
[Chaussidon et al., 1987] and from bulk
arc glasses and basalts [e.g., Alt et al.,
1993; de Hoog et al., 2001] have previ-
ously been interpreted as evidence
that slab-derived 34S-enriched fluids
were added to the mantle and to the
magmatic source of arc volcanism.
Moreover, slab-derived sulfur may con-
Figure 1. d34S variability for sulfides from different terrestrial reservoirs, meteor-
tribute to the redox budget of the
ites, and for modern seawater sulfate. Data represented with black solid lines
mantle wedge and possibly influence
were taken from Ohmoto and Rye [1979], Seal [2006], and Marini et al. [2011]. Gray
the redox state of arc magmas [Evans,
lines for ophiolites from the Apennines and Ligurian Alps [Garuti et al., 2009; Alt
et al., 2012] and from the Western Alps and New Caledonia [Evans et al., 2014].
2006, 2012; Kelley and Cottrell, 2009;
Black dotted lines refer to the present work.
Evans and Tomkins, 2011]. Sulfur can
occur in different oxidation states
(valence 22, 21, 0, 14, or 16) and may be released by processes associated with subduction, possibly after
a change in valence due to oxidation-reduction reactions. Since one mole of S61 has the potential to oxidize
eight moles of Fe21 to Fe31 when S61 is reduced to S22 [Kelley and Cottrell, 2009], S may be a significant
oxidant of subarc mantle.
Sulfide deposits are widespread in the Italian Western Alps and Northern Apennine ophiolites. These ophio-
lites are portions of the oceanic lithosphere of the Piedmont ocean of Jurassic to Late Cretaceous age [e.g.,
Lemoine et al., 1987, and references therein; Polino et al., 1990; Rubatto et al., 1998; Castellarin, 2001; Dal Piaz
et al., 2003; Carminati and Doglioni, 2012]. Ophiolites from the Western Alps are pervasively affected by
subduction-related metamorphism and deformation.
In this paper, we investigate two ocean floor-related sulfide deposits from the Western Alps: the
blueschists-facies Beth-Ghinivert (Chisone Valley) and the eclogite-facies Servette (St. Marcel Valley) depos-
its. These deposits underwent polyphase metamorphism from high pressure-low temperature (HP-LT) to
greenschist facies during the Alpine Orogeny.
A combination of textural observations, maps of trace element relative abundances, and in situ analyses of
d34S within individual pyrite and chalcopyrite grains (SIMS analyses) allows evaluation of the extent of
recrystallization and fluid-rock interaction during metamorphism. The comparison between the two depos-
its allows assessment of the relative roles of temperature, infiltration of external fluids, and deformation on
the metamorphic evolution of sulfide mineralization. Results are used to constrain the nature of metamor-
phic fluids in the subduction zones and the role of high-pressure metamorphism in the mobilization of sul-
fur and metals of economic interest.

2. Overview of Sulfur Isotopes

2.1. Ocean Floor-Related Sulfide Deposits
Several studies indicate that d34S values in sulfides found in unaltered mid-ocean ridge gabbros and basalts
(20.3 6 2.3& in MORB basalt) approach the composition of sulfides in meteorites (Figure 1) [Ohmoto and
Rye, 1979; Seal, 2006; Marini et al., 2011]. Sulfides from ocean-floor hydrothermal and sedimentary deposits,
on the other hand, show heterogeneous d34S values (ranging from 250 to 40&), at all scales from deposit

GIACOMETTI ET AL. C 2014. American Geophysical Union. All Rights Reserved.

V 3809
 Geochemistry, Geophysics, Geosystems 10.1002/2014GC005459

to single crystal, due to organic and inorganic processes of isotopic fractionation (Figure 1) [e.g., Sangster,
1971; Janecky and Seyfried, 1984; Alt and Chaussidon, 1989; Cook and Hoefs, 1997; Velasco et al., 1998; Garuti
et al., 2009].
Sulfur isotope fractionation is controlled by redox conditions, temperature, pH, and kinetic factors [Bachin-
ski, 1969; Ohmoto and Rye, 1979; Ohmoto and Lasaga, 1982; Ohmoto and Goldhaber, 1997; Seal, 2006]. There-
fore, variation of d34S at deposit and microscale can be due to a combination of processes, such as mixing
of different S sources (i.e., S-bearing fluids separated from magmas, sulfides leached from host magmatic or
sedimentary rock, organically and/or inorganically reduced seawater sulfate), cooling of hydrothermal fluids,
kinetic effects during precipitation of S-bearing minerals in open and closed systems, and variations of pH
and Eh conditions [Ohmoto and Rye, 1979; Puchelt and Hubberten, 1980; Alt and Chaussidon, 1989; Strauss
and Schieber, 1990; Calvert et al., 1996; Seal, 2006].

2.2. Insights From Obducted and Subducted Ophiolites of the Piedmont Ocean
Deformation, recrystallization, dissolution-precipitation, fluid-rock interaction, diffusion, and isotope-
exchange reactions occurring during metamorphism are likely to cause progressive rehomogenization of
isotopic compositions within sulfides at different scales, even though primary isotope distribution patterns
have been in part preserved in many metamorphosed deposits [e.g., Sangster, 1971; Cook and Hoefs, 1997;
Alirezaei and Cameron, 2001; Wagner and Boyce, 2006; Evans et al., 2014].
Sulfide deposits from the Apennines, which did not undergo subduction, are likely to record isotopic char-
acteristics similar to those that may have existed in sulfide deposits in the Western Alps before they under-
went HP metamorphism. Bulk mineral analyses of S isotopes have been performed on a number of sulfide
deposits hosted by Northern Apennine ophiolites [Garuti et al., 2009]. This bulk mineral approach does not
allow an evaluation of intracrystalline heterogeneities. Nevertheless, the authors found that d34S was heter-
ogeneous at sample scale in all the different settings and sampling zones investigated: d34S values in sulfide
ores deposited at the ocean floor range from 20.6 to 11.3& (one outlier value 5 22.8&) for pyrite and
from 22.9 to 11.4& for chalcopyrite [Garuti et al., 2009].
Schwarzenbach et al. [2012], who performed bulk rock analyses of serpentinites and ophicalcites from the
Northern Apennines, found d34S ranges between 221.7 and 4.3&.
Previous bulk mineral studies of sulfides in the HP-serpentinites from the Voltri Massif in Ligurian Alps [Alt
et al., 2012] and from Monviso in Coattian Alps [Hattori and Guillot, 2007] produced d34S values that can be
ascribed to seafloor serpentinization. No evidence of changes in S isotopes can be associated with HP
recrystallization in these serpentinites.
Evans et al. [2014] performed in situ S isotope investigations and mapped trace elements of disseminated
sulfides in basic eclogites from the Zermatt-Saas Zone and in transitional eclogites from New Caledonia.
They found that d34S in pyrite ranges from 212& to 17& in the studied samples with variations up to 15&
within a single pyrite crystal and, commonly, polymodal distributions of isotopic ratios. They used a com-
bined textural and geochemical approach to interpret the observed isotopic heterogeneities, which were
consistent with limited metamorphic reequilibration of S isotopes in the two studied areas. The authors
conclude that many of the sulfides from the two localities preserve hydrothermal features, with localized
metamorphic recrystallization and, possibly, addition of slab-derived S during the retrogression.

3. Geological Settings
3.1. Beth-Ghinivert
The Beth-Ghinivert deposit lies in the Chisone Valley (Western Alps, Italy; Figure 2) and belongs to the
Schistes Lustres Unit of the Piedmont Zone. The Schistes Lustres Unit crops out between the neighboring
metasedimentary Brianconnais Unit (to the west) and the underlying Dora-Maira Unit to the east. The
Schistes Lustres is a main metasedimentary unit and consists of poly-deformed calcschists (calcareous shales
and pelites) that were deposited from late Jurassic onward on the Piedmont Ocean floor [e.g., Dal Piaz et al.,
1972; Lemoine et al., 1986; Lemoine and Tricart, 1986]. Meter to kilometer-sized blocks of mafic and/or ultra-
mafic rocks of ophiolitic affinity are scattered within these pelagic sediments, which are part of the internal
Penninic Domain [Dal Piaz et al., 1972; Lemoine et al., 1986; Lemoine and Tricart, 1986]. According to several

GIACOMETTI ET AL. C 2014. American Geophysical Union. All Rights Reserved.

V 3810
 Geochemistry, Geophysics, Geosystems 10.1002/2014GC005459

Figure 2. Simplified geological sketch of (left) Western Alps and the detail of the studied zones. (top right) Map of the Servette zone [after
Martin et al., 2008]. DB 5 Dent Blanche Unit; GP 5 Gran Paradiso nappe; MR 5 Monte Rosa nappe; SL 5 Sesia-Lanzo zone. (bottom right)
Map of the Beth-Ghinivert zone [after Agard et al., 2001]. BR5 Brianconnais Domain; DM 5 Dora Maira Massif; car 5 carpholite; cld 5 chlori-
toid; Epid 5 epidote; Law 5 lawsonite.

authors [e.g., Agard et al., 2001, and references therein], the units belonging to the Penninic Domain under-
went subduction during the closure of the Piedmont Ocean [e.g., Pognante, 1991; Agard et al., 2001; Tricart
and Schwartz, 2006]. Agard et al. [2001] performed a petrologic study of pelitic samples collected between
Sestriere and Fenestrelle and suggested that maximum P-T conditions increase significantly from west to
east and range from 1.2–1.3 GPa and 300–350 C in the west up to 2.0–2.1 GPa and 450–530 C in the east.
At Beth-Ghinivert, meter to decameter-sized blocks of metabasite occur as scattered irregular bodies within
the calcschists and record a polymetamorphic evolution at epidote-blueschist and greenschist-facies condi-
tions. Giacometti and Rebay [2013] proposed a metamorphic peak at 1 GPa and 492 6 35 C followed by
reequilibration at 0.82 6 0.12 GPa and 457 6 30 C. Relicts of the HP parageneses (blue amphibole, white
mica, chlorite, epidote, albite, quartz, minor titanite and, locally, sulfides) survived the retrogression, which
occurred at 0.5 6 0.1 GPa and temperatures below 450 C [e.g., Agard et al., 2001].
A 0.25–2.5 m thick zone of pyrite-rich stratiform mineralization is associated with the metabasites (metagab-
bros, glaucophanites, and greenschist-facies metabasites) and the calcschists [Novarese, 1900; Bouquet and
Forette, 1973; Giacometti and Rebay, 2013]. The deposit was exploited during the nineteenth century for
copper. Novarese [1900] asserted that the mineralized horizon has a lateral extension of at least 400 m (par-
allel to dip), a mean thickness of 0.5 m, and is continuous, parallel to the direction of strata, to a depth of at
least 80 m. Disseminated sulfides occur in the metabasites at the contact with the stratiform mineralization.
According to Novarese [1900], the metabasite is locally impregnated by sulfides. At least two different gen-
erations of pyrite have been recorded at Beth; common relicts preserving colloidal textures, which are typi-
cal low-temperature hydrothermal features, are overgrown by cubic metamorphic pyrite [Natale, 1966,
1969; Natale and Visetti, 1980]. These features are revealed by chemical etching. A sedimentary-exhalative
origin on the Piedmont Ocean floor is proposed for the Beth-Ghinivert deposit [e.g., Natale and Visetti,
1980].

3.2. Servette
The Servette deposit lies in the St. Marcel valley (Western Alps, Italy; Figure 2) in the Zermatt Saas Zone of
the Penninic Domain. The ophiolites of the Zermatt Saas Zone have been interpreted as oceanic lithosphere

GIACOMETTI ET AL. C 2014. American Geophysical Union. All Rights Reserved.

V 3811
 Geochemistry, Geophysics, Geosystems 10.1002/2014GC005459

belonging to the Piedmont Ocean, which was subducted below the Apulian margin between 50 and 40
Ma [e.g., Lemoine et al., 1986; Polino et al., 1990; Rubatto et al., 1998; Dal Piaz et al., 2003].
The Zermatt Saas Zone is considered to be composed of ophiolitic slices that recorded different P-T paths
during the Alpine orogeny [e.g., Martin et al., 2008; Rebay and Powell, 2012]. This interpretation is controver-
sial; Angiboust et al. [2009] proposed that the Zermatt Saas Zone is a continuous slice of oceanic lithosphere
that records relatively homogeneous pressure and temperatures. Exhumation to below 350 C occurred
around 44 Ma [Hunziker, 1974; Barnicoat and Cartwright, 1995].
Servette is the site of an economically exploited Fe-Cu sulfide deposit that has been interpreted as the
product of extensive hydrothermal alteration of the seafloor. The rocks underwent eclogite-facies metamor-
phism, but eclogites sensu stricto (omphacite 1 garnet rocks) are rare at Servette. Interlayered glaucophan-
ites, chlorite schists, and talc schists are the dominant lithologies [Castello, 1981; Cartwright and Barnicoat,
1999; Martin et al., 2008; Rebay and Powell, 2012]. These different rock types have been attributed to differ-
ent styles of hydrothermal alteration. The talc schists were considered as evidence of up-flow zones in black
smoker-type environments, whereas the glaucophanites and chlorite schists formed in zones of downward,
up-temperature, pervasive, fluid flow [Cartwright and Barnicoat, 1999; Martin et al., 2008]. Sulfides at Serv-
ette include pyrite, chalcopyrite with accessory sphalerite, bornite, pyrrhotite, neodigenite, marcasite, mack-
inawite, and rare native copper [e.g., Tartarotti et al., 1986; Krutow-Mozgawa, 1988; Martin et al., 2008].
The sulfide mineralization is concentrated in two major 3–4 m thick layers, and some minor layers (less than
1 m) intercalated between chlorite schists and glaucophanite. Sulfides are also disseminated in the host
rocks [Martin et al., 2004]. This deposit was exploited from the Middle Ages (or probably Roman times) up
to the 1950s with several interruptions of activity [Martin et al., 2004].
Martin et al. [2008] proposed a polymetamorphic evolution for Servette, starting with ocean-floor metamor-
phism followed by Alpine prograde blueschist facies toward peak eclogite-facies metamorphism at
550 6 60 C and 2.1 6 0.3 GPa. P-T-peak conditions of 2.3 GPa and 530 C were proposed by Angiboust et al.
[2009]. The extent of greenschist-facies retrogression is variable at Servette [Martin et al., 2008].

4. Methods
4.1. In Situ S Isotope Analyses
Eight samples of massive and disseminated mineralization (six from Servette and two from Beth-Ghinivert)
were selected after a detailed petrographic study was performed using transmitted and reflected light
microscopy on thin sections and reflected light microscopy on corresponding rock billets. The chosen sam-
ples were cut and mounted in EPOFIX epoxy with the in-house pyrite standard Sonora 3.
The mounts were polished with diamond paste down to 0.25 mm, cleaned, and a 30 nm gold coating was
applied on the mount surface. In situ analyses (a total of 491 on unknowns and 188 on standards) were per-
formed at the Centre for Microscopy, Characterization and Analysis (CMCA) at The University of Western
Australia in Perth (WA) using a CAMECA IMS 1280 ion probe.
Development of the in-house pyrite standard (Sonora 3) is described by Evans et al. [2014]. Briefly, Sonora 3
standard is subsamples taken from a 1 cm3 subsample of pyrite from a large 9 kg cube. The IMS-1280 was
used to test the isotopic heterogeneity of the Sonora 3 material, and the absolute d34S value obtained by
laser fluorination at McGill University is 1.61 6 0.08& relative to VCDT (Vienna Canyon Diablo Troilite).
A 10 keV, 1 nA focused Cs1 ion beam was used to presputter the analysis area for 10 s before automatic
secondary centering and 10 3 4 s analysis cycles were acquired. 32S2, 33S2, and 34S2 secondary ions were
collected using Faraday cup detectors and NMR regulation was used for all analyses. An electron gun was
used for charge compensation. Other conditions include: 133X magnification between the sample stage
and field aperture (FA), 4000 mm FA, 40 eV energy window with a 5 eV offset to the high-energy side, 70 mm
entrance slit, and 500 mm exit slits.
All the d34S values reported in this paper are referenced to VCDT (Vienna Canyon Diablo Troilite). Matrix cor-
rections for pyrite were made using the laser fluorination d34S for the Sonora 3 standard to calculate the
IMF on a daily basis. Matrix corrections for chalcopyrite were made using a matrix correction factor for chal-
copyrite relative to pyrite obtained from multiple analyses of chalcopyrite on the SIMS IMS 1280 at the

GIACOMETTI ET AL. C 2014. American Geophysical Union. All Rights Reserved.

V 3812
 Geochemistry, Geophysics, Geosystems 10.1002/2014GC005459

CMCA and laser fluorination measurements of chalcopyrite (n 5 5) standard subsamples at McGill Univer-
sity. Further details of the matrix correction procedure are provided by Evans et al. [2014].
This in situ microanalytical approach allows evaluation of microscale spatial variation of S isotope values
and reveals heterogeneities and/or chemical zoning that record fluid infiltration. S isotopes ratios are
reported as d34S, defined by:
h i h i21 
d34 S ð&Þ5 ð34 S=32 SÞsample 2ð34 S=32 SÞVCDT  ð34 S=32 SÞVCDT  1000:

Average external precision (1 SD) for d34S analyses of the standards was calculated (for each session day)
from the standard deviation of the drift-corrected d34S values measured in Sonora 3 and range between
0.19 and 0.32&. Overall, the 188 replicates on the Sonora 3 standard performed as part of this work return
a d34S value of 1.61 6 0.26&. Internal precision ranged from 0.14 to 0.30& (2r 5 0.28–0.60&), which is
close to the external precision calculated for the Sonora 3 standard.

4.2. EPMA Analyses

Trace element maps for As, Ni, Co, Mn, and Cu for three selected pyrite grains were made to investigate
compositional zoning. The maps were obtained using the JEOL 8530F Hyperprobe at the CMCA. Operating
conditions were 20 kV accelerating voltage, 100 nA beam current, 50 ms dwell time, and the pixel sizes of
1–2 lm depending on the size and scale of chemical heterogeneity in the area mapped.

4.3. Chemical Etching

After SIMS and EPMA analyses, chemical etching was performed using a 6M HNO3 solution to reveal textural
features. The solution was dropped on the mounts and left for 20 min and the mounts were then inspected.
This process was repeated until textures were revealed. Observed features were correlated with those
detected by in situ S analysis.

5. Petrographic Description
5.1. Samples From Beth-Ghinivert
The two samples from Beth-Ghinivert were collected from an outcrop at one of the mine entrances
(44 570 20.8700 N 6 590 42.4400 E). Sample BE51 comes from a stratiform 30 cm thick massive mineralization at
the base of a hectometer-sized metabasite block, which lies within the calcschist sequence. The sample
consists of pyrite (60 vol %), with tiny sphalerite and rare bornite inclusions, and quartz (40 vol %).
Chemical etching with 6 mol/L HNO3 allowed recognition of complex textures related to several growth
stages and the local occurrence of metamorphic recrystallization of irregular to polygonal aggregates (Fig-
ure 3a).
Sample BE66 was collected at the top of the stratiform 30 cm thick massive mineralization (where BE51
comes from), at the contact with the metabasite. This sample consists of two regions with different modal
composition and textural features. The first region consists of mostly euhedral to rounded pyrite (59 vol
%) that contains frequent bornite, chalcopyrite, sphalerite, and a few pyrrhotite inclusions, generally anhe-
dral chalcopyrite (40 vol %), minor quartz (1 vol %), and accessory magnetite (Figure 3b). The second
region is composed of chlorite (40 vol %), quartz (40 vol %), minor fine-grained sulfides (pyrite and chal-
copyrite; 20 vol %), and accessory white mica (Figure 3c). The mineralization is moderately deformed, but
folding and kink-banding occur in the silicate-rich portion at its margins (Figure 3c). We focused our atten-
tion on the sulfide-rich portion of the sample, where crystals were large enough to be studied in detail.

5.2. Samples From Servette

Six samples were analyzed from four lithologies (chlorite schist, glaucophanite, talc schist, and massive min-
eralization). Samples come from the talus located between N45 42.1110 , E7 27.3270 (alt. 1820 m) and
N45 42.0660 , E7 27.2800 (alt. 1792 m) as the mine is no longer accessible.
Samples belonging to the same rock type differ in their mineral modal compositions. Sulfides occur in all
the samples investigated in the matrix and as inclusions in garnet porphyroblasts. Both sulfide textural posi-
tions are associated with prograde to peak condition parageneses, which are locally overgrown by retro-
grade assemblages [Kawakami et al., 2006; Martin et al., 2008].

GIACOMETTI ET AL. C 2014. American Geophysical Union. All Rights Reserved.

V 3813
 Geochemistry, Geophysics, Geosystems 10.1002/2014GC005459

Figure 3. (a and b) Reflected and (c) transmitted plane-polarized light microscope images of the samples from Beth-Ghinivert. (Figure 3a) Sample BE51; pyrite aggregates in a quartz
matrix: chemical etching evidences relict growth structures (Figure 3b) in pyrite and local recrystallized irregular (see red dotted lines) to polygonal aggregates (see red solid lines). (b)
Sample BE66; cubic pyrite and anhedral chalcopyrite. Absence of polygonal aggregates. (Figure 3c) Sample BE66; chlorite-quartz-pyrite folded layers at the very contact between the
massive mineralization and the country metabasite. Pyrite crystals are deformed with locally resorbed margins. Irregular sulfide crystals fill the space among blastic silicates. Ccp 5 chal-
copyrite; Py 5 pyrite.

Samples SE13 and SE14 are from garnet-bearing chlorite schists. These samples consist mostly of chlorite
(up to 60 vol %) with up to centimeter-sized garnets (23 vol %), sulfides (10 vol %, pyrite and chalcopy-
rite with rare tiny sphalerite and galena inclusions), quartz (5 vol %), chloritoid porphyroblasts (2 vol %),
and accessory talc, albite, rutile (rimmed by ilmenite), blue amphibole (only in SE13), and late magnetite
(Figures 4a and 4b). The garnet crystals are zoned with respect to inclusion assemblages. Quartz, titanite,
pyrite, and chalcopyrite inclusions are aligned along a relict foliation in the core of garnets. Scattered quartz,
pyrite, talc, and rutile (rimmed by ilmenite) inclusions occur at the rim of crystals. Chloritoid and garnet rims
are partially replaced by late chlorite, which also occurs with magnetite in fractures within porphyroblasts.
In the matrix, chlorite, quartz, pyrite (which includes rutile), minor chalcopyrite, and rare blue amphibole in
SE13 are aligned parallel to the pervasive HP foliation. Quartz, chlorite, and talc (6blue amphibole and
albite in SE13) occur in the pressure shadows of garnet and chloritoid porphyroblasts.
Sample SE8 is a garnet-bearing glaucophanite. Garnet (30 vol %), chloritoid porphyroblasts (5 vol %),
and pseudomorphs after lawsonite porphyroblasts (consisting of epidote, phengite, paragonite, and albite;
20 vol %) occur within a foliated matrix made of glaucophane (33 vol %), clinozoisite (4 vol %), and
talc (5 vol %). Garnet and chloritoid are up to 9 mm in diameter and are partially replaced by late chlorite
at crystal rims and along fractures where rare tiny magnetite crystals were found. Garnets include rutile
(locally rimmed by ilmenite), quartz, glaucophane, rare talc, and pyrite. These phases locally define a spiral
internal foliation that testifies that garnet porphyroblasts grew simultaneously with the development of the
external foliation. Rutile in garnet and chloritoid both includes and is included by sulfides. Pyrite (2 vol %),
chalcopyrite (1 vol %), and rare bornite and sphalerite in the matrix are stable with garnet, glaucophane,
chloritoid, rutile, and clinozoisite (Figure 4c).
ATSV-001 is a garnet-sulfide-rich talc schist (Figure 4d), with 2–5 mm diameter garnet (40 vol %), pyrite
(25 vol %), chalcopyrite (5 vol %), felty-looking talc (20 vol %), and chlorite (10 vol %) and accessory
quartz. The garnet forms a framework of subequant grains. Pyrite occurs as submillimeter inclusions in gar-
net and with chlorite found in cracks in garnet. Polygranular patches of pyrite up to a few mm across occur
in the matrix. Chalcopyrite typically surrounds matrix pyrite (Figure 4d) and is occasionally intergrown with
silicates without pyrite. Talc, which is often iron stained, occurs in felty, deformed masses between pyrite
grains. Chlorite occurs on the rims of garnet, and as laths that crosscut talc. Rutile, rimmed by ilmenite, is
included in garnet and contains fine exsolution laminae and blebs of hematite.

GIACOMETTI ET AL. C 2014. American Geophysical Union. All Rights Reserved.

V 3814
 Geochemistry, Geophysics, Geosystems 10.1002/2014GC005459

Figure 4. Reflected plane-polarized light microscope images of some of the samples from Servette. (a) Sample SE14; garnet including pyrite, chalcopyrite, and rutile, some of which are
far from fractures occurring within the host mineral; pyrite and chalcopyrite also widespread in the chlorite matrix. (b) Sample SE13; polygonal pyrite (and subordinate chalcopyrite and
quartz) aggregate in the chlorite matrix. (c) Sample SE8; pyrite and subordinate chalcopyrite in the glaucophane matrix. Garnet porphyroblasts occur in the right side of the picture. (d)
Sample ATSV-001; pyrite and chalcopyrite occur both as small inclusions in garnet porphyroblasts and as polygonal aggregates in the matrix. (e) Sample SER18gl; pyrite and
chalcopyrite-rich glaucophanite: sulfides occur both included in garnet and in the matrix. (f) Sample SER18m; rounded pyrite crystals surrounded by anhedral chalcopyrite in a quartz
matrix. Ccp 5 chalcopyrite; Chl 5 chlorite; Gln 5 glaucophane; Grt 5 garnet; Py 5 pyrite; Qtz 5 quartz; Rt 5 rutile.

Sample SER18 contains the contact between a massive pyrite-chalcopyrite glaucophanite (SER18gl; Figure
4e) and a garnet and sulfide-bearing mineralization (SER18m; Figure 4f).
The glaucophanite (SER18gl) is mainly made of glaucophane (35 vol %), garnet (30 vol %), pseudo-
morphs after lawsonite (consisting of epidote, phengite, paragonite, and albite 17 vol %), sulfides (pyrite
with subordinate chalcopyrite and rare bornite; 16 vol %), minor quartz, and chlorite and accessory talc.
Chalcopyrite is present both in the matrix and as inclusions in garnets whereas pyrite is common in the
matrix, but only a few micrometer-sized inclusions occur in garnet (Figure 4e).
The massive mineralization (SER18m) is mainly made up of pyrite (45 vol %), garnet (25 vol %), chalcopy-
rite (15 vol %), and quartz (15 vol %). Rutile rimmed by ilmenite occurs as an inclusion in garnet in both
SER18gl and SER18m. The modal abundance of garnet increases toward the contact with the metabasite.
Pyrite is present both in the matrix and as inclusions in garnet whereas only interstitial chalcopyrite is

GIACOMETTI ET AL. C 2014. American Geophysical Union. All Rights Reserved.

V 3815
 Geochemistry, Geophysics, Geosystems 10.1002/2014GC005459

Table 1. Lowest, Highest, and Average d34S Values (&), Standard Deviations (&), and Number of Analyses for Pyrite (Py) and Chalcopy-
rite (Ccp) From the Eight Studied Samplesa
Lowest d34S (&) Highest d34S (&) Average d34S (&) r (&) Number of Analyses
SE 14 py in garnet 0.6 8.1 6.2 1.3 35
py matrix 22.0 7.6 5.1 1.9 28
ccp matrix 21.6 1.3 20.7 0.8 10
SE 13 py in garnet 3.3 8.4 5.2 1.5 35
py matrix 2.4 6.3 4.1 0.7 57
ccy in garnet 20.9 2.1 20.2 0.7 12
ccy matrix 20.7 20.3 20.5 0.2 4
ccy vein 22.1 20.6 21.4 0.6 4
SE 8 py in garnet 20.9 2.4 0.8 1.6 2
py matrix 8.1 13.9 9.5 1.5 39
ccp matrix 0.9 0.9 1
ATSV-001 py in garnet 7.2 9.6 8.7 0.8 7
py matrix 6.8 12.6 8.1 1.3 28
ccp matrix 1.6 2.7 2.2 0.4 6
ccp vein 2.2 3.0 2.5 0.2 12
SER18 gl py matrix 22.3 7.1 2.9 2.9 21
ccp in garnet 24.0 23.6 23.8 0.2 8
ccp matrix 23.8 23.0 23.3 0.3 7
SER18 min py in garnet 20.8 20.1 20.5 0.2 8
py matrix 0.1 1.2 0.8 0.2 27
ccp matrix 23.4 21.2 21.9 0.6 9
BE51 py 21.4 6.4 3.6 1.6 56
BE66 py 20.7 3.8 1.7 0.9 61
ccp 26.7 24.6 25.2 0.6 14
a
Data are reported for pyrite (py) and chalcopyrite (ccp) in each different microstructural domain (i.e., sulfide in the matrix or included
in garnet).

present in the matrix. The pyrite crystals (up to 2 mm) in the matrix have rounded margins suggesting
resorption (Figure 4f).

6. Isotope Characterization
Minimum, maximum, and average d34S values, standard deviations, and the number of analyses are sum-
marized in Table 1.

6.1. Samples From Beth-Ghinivert

6.1.1. BE51 (Massive Mineralization)
The d34S of pyrite ranges from 21.4 to 6.4& within a distance of 1 mm (Figures 5a, 5b, and 6a and Table 1).
There is no obvious correlation between isotopic variations and growth textures or with textures related to
metamorphic recrystallization (Figures 5a–5c).
Qualitative element mapping shows that the concentration of Mn shows a pattern consistent with growth
zoning (Figure 5d). However, d34S values in the same aggregate of pyrite crystals are scattered and do not
correlate with Mn distribution (Figure 5d). Metamorphic recrystallization textures are not related to Mn zon-
ing and growth textures or to sulfur isotopes variations.

6.1.2. BE66 (Mineralization Close to the Country Rock)

The d34S value ranges from 20.7 to 3.8& in pyrite (average value 5 1.7& 6 0.9&, n 5 61) and from 26.7
to 24.6& (average value 5 25.2 6 0.6&, n 5 14) in chalcopyrite (Figure 6b and Table 1). d34S values vary
by up to 3.4& within a single pyrite crystal. In one pyrite crystal (Figure 5e), isotopic zoning occurs, with
lower d34S values at the crystal core. In this sample, chalcopyrite inclusions in pyrite and chalcopyrite in the
matrix record statistically undistinguishable d34S values. Dav.py-ccp is 6.9 6 1.4&.
Qualitative trace element mapping of the pyrite shows that only Co varies to a detectable degree within
the crystal. The lowest d34S values observed in pyrite (20.7 6 0.4& to 0.8 6 0.2&) correspond to the Co-
rich core of a 2 mm sized crystal. This low-d34S and Co-rich core is overgrown by a high-d34S zone that is
characterized by cyclical variations of Co content (Figure 5f).

GIACOMETTI ET AL. C 2014. American Geophysical Union. All Rights Reserved.

V 3816
 Geochemistry, Geophysics, Geosystems 10.1002/2014GC005459

Figure 5. Textures and geochemical features in samples from Beth-Ghinivert. Numbers indicate the measured d34S values. (a) Sample
BE51; pyrite aggregate preserving growth textures. (b) The diagram shows the variation of d34S along three transects highlighting the iso-
topic heterogeneity of pyrite within a few tens of microns. (c) Sample BE51; well-preserved growth textures correspond to a zoned distri-
bution of Mn (shown in Figure 5d). Recrystallization is localized (see red dotted lines indicating boundaries). (d) Sample BE51; relative
distribution of Mn in pyrite, mimicking growth textures and evidencing lack of metamorphic rehomogenization of trace elements.
(e) Sample BE66; low d34S values occur at the core of this pyrite crystal, whereas higher d34S values occur at the rim. (f) Sample BE66; the
distribution of Co evidences relict growth textures. Dotted red lines evidence Co-rich cores (warm colors), which record low d34S values.
Oscillation of Co relative concentration is seen at crystal rims where higher d34S values occur.

6.2. Samples From Servette

6.2.1. SE13 and SE14 (Chlorite Schists)
d34S values were measured in pyrite and chalcopyrite from both samples (Figures 7a–7c and 8a–8e and
Table 1). d34S values for pyrite inclusions in garnet (Figures 8a and 8d) and pyrite in the matrix (Figures 8b

GIACOMETTI ET AL. C 2014. American Geophysical Union. All Rights Reserved.

V 3817
 Geochemistry, Geophysics, Geosystems 10.1002/2014GC005459

and 8e) span the same range of 2–8.5& (exclud-

ing two outliers in SE13 at 22.0 and 0.6&). Sam-
ples were heterogeneous with respect to sulfur
isotopes on the micron scale; d34S is homogene-
ous at the submillimeter scale only within totally
recrystallized pyrite aggregates in the matrix.
The d34S values of the two different textural posi-
tions in the two samples are statistically indistin-
guishable. However, pyrite included in garnet
records slightly heavier average isotopic composi-
tions than pyrite in the matrix. Average d34S val-
ues in SE13 are 5.2 6 1.5& (n 5 35) and
4.1 6 0.7& (n 5 57), respectively, for pyrite inclu-
sions in garnet and pyrite in the matrix. In sample
SE14, average d34S are 6.2 6 1.3& (n 5 35) and
5.1& 6 1.9& (n 5 28), respectively, for pyrite
inclusions in garnet and pyrite in the matrix (Fig-
ures 8a–8e and Table 1). d34S is homogeneous at
the submillimeter scale only within totally recrys-
tallized pyrite aggregates in the matrix.
d34S in chalcopyrite is more homogeneous than
in pyrite (Figure 8) and, as for pyrite, the average
Figure 6. Histograms showing the distribution of d34S values
observed in the samples from Beth-Ghinivert. (a) Sample BE51; wide values are statistically indistinguishable. In sam-
range of d34S values for pyrite. (b) Sample BE66; d34S values for pyrite ple SE13, the average d34S for chalcopyrite in
and chalcopyrite. garnet is 20.2 6 0.7& (n 5 12); d34S for chalco-
pyrite in the matrix is 20.5 6 0.2& (n 5 4), and
the average d34S for chalcopyrite in a vein crosscutting a garnet porphyroblast is 21.4 6 0.6& (n 5 4; Fig-
ures 8a–8c). In sample SE14, chalcopyrite inclusions were smaller than the ion probe beam and were not
analyzed. Average d34S for chalcopyrite from the matrix in SE14 is 20.7 6 0.8& (n 5 10; Figure 8e). The
Dav.py-ccp (5average d34Spyrite – average d34Schalcopyrite) value in garnet inclusions is 5.4 6 1.7& in sample
SE13; the Dav.py-ccp values in the matrix are 4.6 6 0.9& and 5.8 6 2.7& in SE13 and SE14, respectively.
Nonquantitative minor element maps of pyrite aggregates in the matrix of SE14 reveal that only Co, and to
some extent As, show heterogeneous distributions (Figure 7c). The lowest d34S values correspond to the

Figure 7. d34S values (&) in some of the studied microsites in chlorite schists for pyrite (py) and chalcopyrite (ccp). (a) Sample SE13; pyrite and chalcopyrite included in garnet. Heteroge-
neous d34S values occur in pyrite, in contrast to the relative homogeneity in chalcopyrite. (b) Sample SE14; pyrite aggregate in the matrix with wide d34S range. (c) Qualitative Co distribu-
tion in the aggregate of Figure 3b. Later pyrite is Co poor (cool colors) in comparison to pyrite deposited earlier (warm colors). Ccp 5 chalcopyrite; Py 5 pyrite.

GIACOMETTI ET AL. C 2014. American Geophysical Union. All Rights Reserved.

V 3818
 Geochemistry, Geophysics, Geosystems 10.1002/2014GC005459

lowest Co-As abundance, and texturally late Co-

rich pyrite is partially replaced by Co-poor
pyrite, but there is no consistent correlation
between the Co-As distribution and d34S.

6.2.2. SE8 (Glaucophanite)

d34S measured in pyrite from the matrix (average
9.5 6 1.5&, n 5 39; Figure 9a and Table 1) is sig-
nificantly heavier than d34S observed in the few
tiny pyrite inclusions in garnet (20.8 and 2.4&;
Figure 10a and Table 1). Values measured in the
matrix are mostly in the 8–10& range with four
analyses around 13–14&, and variations up to
5.8& within a 60 mm crystal (Figures 9a and 10d).
Only one chalcopyrite crystal in the matrix was large
enough to be analyzed. Its d34S is 0.9&, which gives
a Dpy-ccp of 8.6 6 1.5& if this value is compared
with the average d34S of pyrite from the matrix.

6.2.3. ATSV-001 (Talc Schist)

Pyrite d34S is relatively spatially homogeneous.
Most values that lie between 6.5 and 8.5&, with
the exception of two outliers from two analyses
within a few 10 s of microns of each other which
display heavier d34S values of around 12& (Fig-
ures 9b, 9c, and 10c–10e and Table 1). The aver-
age d34S for pyrite in garnet and in the matrix
(including the outliers) are, respectively,
8.7 6 0.8& (n 5 7) and 8.1 6 1.3& (n 5 28; Fig-
ures 10c and 10d and Table 1). Chalcopyrite, as in
other samples, is more homogeneous than pyrite
(Figures 9b and 10d–10e). The average d34S is
2.2 6 0.4& (n 5 6) in the matrix and 2.5 6 0.2&
(n 5 12) in a later vein (Figures 10d–10e). The
Dav.py-ccp value in the matrix is 5.9 6 1.7&.
There is no evidence of a systematic relationship
between pyrite d34S values and location within
the grain or with proximity to chalcopyrite (Fig-
ure 9b). Apart from two high d34S values
recorded in the matrix, which were rechecked
and found to be robust, the isotopic composition
in garnet inclusions and in the matrix are statisti-
cally undistinguishable (see Table 1).
Figure 8. Histograms showing the distribution of d34S values
observed in chlorite schists. Note the different scales of the histo- 6.2.4. SER18gl (Glaucophanite)
grams for appropriate comparison. Light gray is chalcopyrite (ccp) Pyrite is widespread in the matrix (Figure 11a),
and dark gray is pyrite (py). Pyrite is isotopically heterogeneous com- but only a few very small inclusions occur in gar-
pared to chalcopyrite at the sample scale. (a) Sample SE13; sulfide
inclusions in garnet. (b) Sample SE13; matrix sulfides. (c) Sample net and these were too small (<10 mm) to be
SE13; chalcopyrite in a vein crosscutting a garnet blast. (d) Sample analyzed. d34S in pyrite is partially heterogene-
SE14; pyrite inclusions in garnet. (e) Sample SE14; matrix sulfides. ous; values range from 22.3 to 17.1& within a
single matrix pyrite crystal: the average d34S
value is 2.9 6 2.9& (n 5 21; Figure 11a).
Matrix pyrite exhibits three modes of d34S values (Figures 11a and 12b). The first is represented by negative
d34S values and is only found in the core of crystals in the matrix. The other two groups, with values of

GIACOMETTI ET AL. C 2014. American Geophysical Union. All Rights Reserved.

V 3819
 Geochemistry, Geophysics, Geosystems 10.1002/2014GC005459

Figure 9. d34S values (&) in some of the studied samples. (a) Sample SE8 (glaucophanite); pyrite intergrown with rutile in the metamorphic peak foliation. (b) Sample ATSV-001 (talc
schist); pyrite rimmed by chalcopyrite in the matrix. d34S values of the pyrite and chalcopyrite are homogeneous except for two outlying high values in the pyrite. (c) Tiny pyrite inclu-
sions in garnet. ATSV-001. Ccp 5 chalcopyrite; Py 5 pyrite; Ru 5 rutile.

3–4& and 5–7.1&, respectively, occur in the rim of the same pyrite crystal as that in which the lowest val-
ues are observed.
In chalcopyrite, d34S is relatively spatially homogeneous ranging from 24 to 23.6& in garnet inclusions
(average d34S 5 23.8 6 0.2&, n 5 8; Figure 12a) and from 23.4 to 23& in the matrix (average
d34S 5 23.3 6 0.3&, n 5 7; Figure 12b and Table 1). The Dav.py-ccp value in the matrix is 6.2 6 3.1&.
6.2.5. SER18m (Massive Mineralization)
The d34S of pyrite included in garnet shows little spatial variability and ranges from 20.8 to 20.1& with an
average of 20.5 6 0.2& (n 5 8; Figure 12c). The d34S of pyrite in the matrix ranges from 0.1 to 1.2&
(n 5 27; Figures 11b and 12d): the average is 0.8& 6 0.2&. The d34S of chalcopyrite from the matrix ranges
from 22 to 21.2& with a single low outlier of 3.4& and an average of 21.9& 6 0.6& (n 5 9). The
Dav.py-ccp value in the matrix is 2.7 6 0.8&.
The isotopic compositions of pyrite in garnet, pyrite in the matrix, and chalcopyrite in the matrix are signifi-
cantly different and rather homogeneous within each category (1r  0.6&; Table 1). This is in contrast with
observation from the host rock SER18gl, where pyrite in the matrix is extremely heterogeneous.

7. Discussion
7.1. Length Scales of Sulfur Isotopic Disequilibrium
The combined textural and isotopic investigation shows that isotopic equilibrium at Servette and Beth-
Ghinivert occurs on very limited length scales, if at all:
1. In both deposits, the d34S composition of pyrite is heterogeneous (with variations up to 9.4& at Servette
and up to 7.8& at Beth-Ghinivert) at the submillimeter scale, indicating limited length scales of equilibrium
(Figures 5a, 5b, and 11). Mean d34S values in a given sample have 1r standard deviations up to 2.9&, which
are greater than the internal precision of the analyses (2r 5 0.28–0.62&; see section 4.1 and Table 1).
2. Sulfur isotope in chalcopyrite may have reequilibrated on longer length scales than in pyrite. Chalcopyrite is rela-
tively homogeneous on the mm scale in any sample with 1r standard deviations in the 0.2–0.8& range (Table 1).
3. Chalcopyrite and pyrite are not in isotopic equilibrium. Even taking into account the propagation of errors in
calculating Dav.py-ccp and considering the lowest possible values for each sample, Dav.py-ccp is generally greater

GIACOMETTI ET AL. C 2014. American Geophysical Union. All Rights Reserved.

V 3820
 Geochemistry, Geophysics, Geosystems 10.1002/2014GC005459

than 3.1& at both deposits, with values up to

10.1&. These values are higher than those
expected at either the peak temperature pro-
posed for Servette (i.e., T 5 550 6 60 C) [Martin
et al., 2008] and Beth-Ghinivert (492 6 35 C) [Gia-
cometti and Rebay, 2013] and at greenschist-facies
retrogression temperatures around 300 C (Figure
13). At these temperatures, Dav.py-ccp is expected
to be 0.8&, 0.9&, and 1.4&, respectively, based
on fractionation factors reported by Seal [2006]
(see references therein for details). Dav.py-ccp > 6&
requires temperatures below 0 C which are not
likely, even for processes occurring at the ocean
floor.

7.2. Preservation of Protolith Isotopic

Composition
The limited length scale of isotopic equilibrium
observed at Servette and Beth-Ghinivert may have
been inherited from the protolith or have devel-
oped during metamorphism. At Beth-Ghinivert, pri-
mary pyrite growth textures are widespread and
unaffected by the limited metamorphic recrystalli-
zation. Analogous textures were described by sev-
eral authors in sulfides at Beth-Ghinivert [Natale,
1966, 1969; Natale and Zucchetti, 1966; Natale and
Visetti, 1980] and were interpreted as ocean-floor
relicts.
Metamorphic recrystallization appears not to
have caused homogenization of Mn and Co
growth zoning so d34S values are likely to reflect
those inherited from the ocean floor (Figure 5).
Deformation and metamorphic mobilization of
sulfides are extremely localized (millimeter scale)
and are always coupled (Figure 3c). Their mutual
occurrence supports an important role for defor-
mation in mobilization of S, as has been sug-
gested for other metamorphosed VMS deposits
[Marshall and Gillian, 1993; Tomkins, 2007].
We therefore propose that most textural and iso-
topic features at Beth-Ghinivert were produced
during ocean-floor metasomatism and that the
limited extent of deformation-induced sulfur
Figure 10. Histograms showing the distribution of d34S values remobilization is responsible for the preservation
observed in (a and b) SE8 and (c–e) ATSV-001. Note the different
of these primary features.
scales of the histograms for appropriate comparison. Light gray is
chalcopyrite (ccp) and dark gray is pyrite (py). (Figure 10a) Pyrite
7.3. S Isotopes Records of Metamorphic
inclusions in garnet with low d34S values. (Figure 10b) Matrix pyrite
Processes
mainly ranges between 8 and 10&. (Figure 10c) Pyrite inclusions in
garnet and chalcopyrite in a vein (Figure 10e) are isotopically similar
7.3.1. Heterogeneous Pyrite
to the corresponding minerals in the matrix (Figure 10d).
At Servette, metamorphic textures are wide-
spread in sulfides (Figure 4b). In samples SE8,
SE14, and SER18gl, pyrite in garnets and some matrix pyrite cores record different compositions compared
to pyrite rims in the matrix (Figures 7b, 9a, and 11 and Table 1). These texturally distinct d34S populations in
pyrite cannot be explained without metamorphic redistribution of d34S.

GIACOMETTI ET AL. C 2014. American Geophysical Union. All Rights Reserved.

V 3821
 Geochemistry, Geophysics, Geosystems 10.1002/2014GC005459

Figure 11. Comparison between isotopically zoned pyrite from the host rock and unzoned pyrite from the massive mineralization. (a) Het-
erogeneous d34S values in core-rim zoned pyrite from the matrix of SER18gl. (b) Homogeneous d34S values from a rounded pyrite in the
matrix of sample SER18m. (c) d34S values measured along transects are reported in a diagram. Ccp 5 chalcopyrite; Gln 5 glaucophane;
Grt 5 garnet; Py 5 pyrite.

Previous workers have established that effective sulfur diffusion through garnet is unlikely at the
temperature of interest, so most of the sulfide inclusions in garnet should have behaved as closed sys-
tems after inclusion in garnet, as long as the grains are not close to later veins [e.g., Crowe, 1994;
Kawakami et al., 2006]. Therefore, sulfides in garnet should preserve the isotopic composition acquired
before garnet crystallization (during pre-Alpine oceanic or during the early stages of prograde
metamorphism). Sulfides in the matrix coexist with HP minerals and are likely to record isotopic compo-
sitions reflecting peak metamorphism, and/or, possibly, local interaction with retrogression-related
fluids.
The limited length-scale isotopic equilibrium recorded by pyrite and disequilibrium recorded by pyrite-
chalcopyrite pairs at Servette can be related to partial reequilibration during subduction-related metamor-
phism. The isotopic heterogeneities in pyrite recognized in every sample (Figures 8, 10, 12a, and 12b), and
replacement textures revealed by the irregular distribution of Co in pyrite (Figure 7c) suggest interaction
between sulfides and fluids during deformation and recrystallization associated with subduction-related
metamorphism. The preservation of this sulfur isotope heterogeneity at the microscale indicates that, at

GIACOMETTI ET AL. C 2014. American Geophysical Union. All Rights Reserved.

V 3822
 Geochemistry, Geophysics, Geosystems 10.1002/2014GC005459

larger scales, diffusive homogenization of S iso-

tope variability in pyrite is minimal, or does not
occur at all, at the P-T conditions experienced
by these rocks.
7.3.2. Homogeneous Pyrite d34S in SER18m
Only sample SER18m records homogeneous
d34S in pyrite. A possible cause for this homoge-
neity is that the protolith was homogeneous
with respect to d34S. However, this hypothesis is
considered unlikely because ocean-floor sulfide
massive deposits are typically heterogeneous at
the microscale, as observed at Beth-Ghinivert
[e.g., Sangster, 1971; Janecky and Seyfried, 1984;
Alt and Chaussidon, 1989; Cook and Hoefs, 1997;
Velasco et al., 1998; Garuti et al., 2009; Alt et al.,
2012]. It is therefore more likely that isotopic
reequilibration occurred during metamorphism.
Isotopic reequilibration may have been facili-
tated by the high modal abundance of sulfides
and by the absence of a silicate matrix [cf.
Crowe, 1994]. Alternatively, or additionally,
minor dissolution and recrystallization in the
pyrite matrix may have provided an effective
means to focus metamorphic fluids through the
massive mineralization, although the extent to
which this may have occurred is not known.
7.3.3. Homogeneous Chalcopyrite
In contrast to the heterogeneous d34S observed
in pyrite, d34S in chalcopyrite is homogeneous,
within error, at the mm to cm scale. This obser-
vation can be explained by the dramatically dif-
ferent responses of pyrite and chalcopyrite to
deformation during metamorphism. Pyrite is
brittle during deformation in many natural set-
tings to temperatures well beyond 400 C [Barrie
et al., 2011], whereas chalcopyrite dynamically
recrystallizes at temperatures as low as 300 C
Figure 12. Histograms showing the statistic distribution of d34S values [Cox, 1987], depending on the differential stress
observed in the host rock (glaucophanite) and in the massive mineral- regime and strain rate. Pyrite is consequently
ization. Note the different scales of the histograms for appropriate sluggish to reequilibrate and tends to preserve
comparison. Light gray is chalcopyrite (ccp) and dark gray is pyrite
(py). (a) Sample SER18gl; chalcopyrite in garnet is isotopically similar original textural and isotopic features more
to that in the matrix (compare with Figure 12b). (b) Sample SER18gl; faithfully than chalcopyrite [Bachinski, 1977].
isotopically heterogeneous pyrite in the matrix. (c) Sample SER18m;
Under certain circumstances, this difference in
pyrite in garnet. (d) Sample SER18m; pyrite and chalcopyrite in the
matrix. Note the isotopic homogeneity of pyrite in sample SER18m,
behavior may have contributed to the observed
compared to other samples.
chalcopyrite d34S values that are too light to
have been in equilibrium with coexisting pyrite.
If chalcopyrite modes decreased during recrystallization with a loss of heavy sulfur, if chalcopyrite recrystal-
lized in equilibrium with a light sulfur source, or if chalcopyrite modes increased by incorporation of light
sulfur then d34S of chalcopyrite would decrease, without a corresponding decrease in the pyrite d34S. The
occurrence of high Dav.py-ccp values at Beth, as well as Servette, suggests that the process that caused the
decoupling of the pyrite and chalcopyrite isotopes was not part of the subduction process. High Dav.py-ccp
values could reflect disequilibrium inherited from the protolith, although they could have been produced
during retrogression associated with exhumation.

GIACOMETTI ET AL. C 2014. American Geophysical Union. All Rights Reserved.

V 3823
 Geochemistry, Geophysics, Geosystems 10.1002/2014GC005459

Figure 13. (a) The 1000lnai-H2S expresses the fractionation factor between S species ‘‘i’’ and H2S coexisting at isotopic equilibrium as a func-
tion of temperature [after Seal, 2006]. (b) Dpy-ccp expresses the theoretical difference between d34S in pyrite and chalcopyrite at equilibrium
as a function of temperature. The black thick lines correspond to the temperature conditions estimated at Servette and Beth-Ghinivert.
The dotted lines indicate the theoretical Dpy-ccp corresponding values, which are significantly lower than those observed in the studied
samples.

7.4. Effects of Pressure, Temperature, Redox, and pH on d34S

7.4.1. Temperature
Isotopic reequilibration, diffusion rates, and fractionation factors during metamorphism are sensitive to tem-
perature [e.g., Bachinski, 1969; Ohmoto and Rye, 1979; Ohmoto and Lasaga, 1982; Seal, 2006]. If sulfur in flu-
ids was dominated by reduced species such as H2S and HS2 (see below) with a constant isotopic
composition during Alpine metamorphism, then evolution of temperatures between 300 and 600 C during
prograde and retrograde metamorphism [cf. Hunziker, 1974; Barnicoat and Cartwright, 1995; Martin et al.,
2008] would cause less than 2& variation in the d34S of pyrite precipitated from a fluid of fixed composition
(Figure 13a). Calculations are based on fractionation factors reported by Seal [2006]. Thus, temperature vari-
ation alone cannot explain the 9.4& core-to-rim increase in d34S observed in pyrite in sample SER18gl (Fig-
ure 11a), or the 14& garnet-to-matrix increase in sample SE8 (Figures 10a and 10b).
7.4.2. Pressure
Little is known about the influence of pressure on isotope fractionation, but Reed and Palandri [2006] and
Wagner and Boyce [2006] observed that the solubility of pyrite is minimally affected by pressure compared
to temperature. Effects of pressure on S speciation and on isotope fractionation were proposed during
degassing of magmas [e.g., de Moor et al., 2011] but these processes can be discounted at Servette, where
magmas were not formed. We therefore consider that variations in pressure recorded by the samples stud-
ied here did not influence S isotope fractionation directly.
7.4.3. pH
There is no general consensus with respect to the effects of pH on S isotope-fractionation during sulfide-
fluid interaction. However, Wagner and Boyce [2006] indicated that fractionation between different S-
bearing aqueous species is essentially controlled by their oxidation state, with limited influence from pH, so
pH effects are discounted here.
7.4.4. Redox and Sulfur Speciation
Calculations reported for analogous metallic parageneses suggest that, at peak metamorphic conditions,
fO2 was <10220 bar and fS2 was <1024 bar [e.g., Condie, 1967; Itaya et al., 1985; Shi, 1992; Kontny et al., 1997;

GIACOMETTI ET AL. C 2014. American Geophysical Union. All Rights Reserved.

V 3824
 Geochemistry, Geophysics, Geosystems 10.1002/2014GC005459

Wood, 1998; Pokrovski and Dubrovinsky, 2011, and references therein]. These values are an estimate, at least
for Servette, because the original calculations were made at pressures 1.1 GPa. Extrapolation is necessary
because there are no data available for pressures higher than 1.1 GPa.
Pokrovski and Dubrovinsky [2011] proposed that the dominant sulfur species at pressure and temperature
above 0.5 GPa and 250 C is the trisulfide ion S32, but, due to the limited information currently available in
the literature about this ion, estimates of its solubility and of isotopic fractionation between this and other
sulfur species cannot be performed.
If S32 is excluded from consideration then the dominant sulfur species in solution at the conditions of inter-
est are expected to be H2S and HS2, as previously proposed by Reed and Palandri [2006], Wagner and Boyce
[2006], and Evans et al. [2014]. Sulfur isotope fractionation between these reduced species is minimal at the
temperatures and pressures of interest [Evans et al., 2014], so sulfides crystallized from a fluid would be
expected to record the sulfur isotope characteristics of that fluid. Similarly, dissolution of sulfides would not
be expected to significantly affect the sulfur isotope composition of the residue.
If sulfate or SO2-bearing fluids were present then open system processes could have a much stronger effect
on sulfur isotope values. 34S sequesters preferentially into more oxidized sulfur species, so a heavy d34S
value in pyrite could result from formation from, or equilibration with, sulfate-bearing fluids, for example, or
a light d34S value could result from loss of sulfate derived from pyrite dissolution. However, there is little evi-
dence that conditions were sufficiently oxidizing for sulfate stability during prograde and peak
metamorphism.
The occurrence of magnetite only in late greenschist-facies veins may record a decrease of S2 activity and/
or an increase of O2 activity in retrograde fluids, or it may record the release of ferric iron from prograde
minerals such as garnet. If more oxidized fluids were present, then oxidized sulfur species may have coex-
isted with sulfide phases, with consequences for the sulfur isotope composition of these phases. In this
work, we have focused on sulfides Further work is necessary to explore this possibility.

7.5. Conceptual Model

The observed limited metamorphic modification of preexisting S isotope heterogeneities is to be expected
if the solubility of S in the metamorphic fluids is low, or if fluids are only sparingly present. In either case, S
would be mobilized on very limited length scales, and d34S values would be controlled by the isotopic com-
position of the protoliths on these length scales and by fluid:crystal fractionation factors. Thus, the hetero-
geneities inherited from the protolith may not be strongly modified by the recrystallization process if sulfur
isotope equilibration is limited to length scales of 10–100 s of microns. It therefore seems likely that fluid
communication was limited at the sample and, as a consequence, at the deposit scale. Pervasive circulation
of S-buffered fluids through hydrothermal ore deposits during shallow-level subduction is therefore not
supported by the evidence.
Fluid flow at Beth-Ghinivert is likely to have been more restricted than at Servette, given the preservation of
primary isotopic features and the limited evidence for fluid release in neighboring areas within the Schistes
Lustres Unit [Agard et al., 2000]. This is consistent with the lower metamorphic P-T conditions at this location
and with the estimated structural positions of the deposit in the subducting slab during peak metamor-
phism (Figure 14); at the depths of 30 km a large amount of water is retained by sediments and by the
oceanic crust, as previously observed in other structurally and genetically comparable areas of the Western
Alps [Phillipot and Scambelluri, 1996; Agard et al., 2000]. Therefore it is likely that very limited fluid circulation
occurred during peak metamorphism Beth-Ghinivert. More intense circulation of metamorphic fluid
occurred at Servette, where higher temperatures were attained during subduction, which is consistent with
its inferred structural position in the subducting slab (Figure 14).

7.6. Comparison With Previous Works

A number of other studies have produced results consistent with those presented here and attributed the
preservation of d34S values inherited from the protoliths to limited water release during metamorphic
recrystallization [Alirezaei and Cameron, 2001; Hattori and Guillot, 2007; Alt et al., 2012]. However, there is
some evidence for more extensive fluid-induced modification of sulfur isotopes during subduction and
exhumation. Alt et al. [2013] attributed the modification of the isotopic composition of sulfides in HP ser-
pentinites from the Betic Cordillera to the release of a significant amount of water. Evans et al. [2014] used

GIACOMETTI ET AL. C 2014. American Geophysical Union. All Rights Reserved.

V 3825
 Geochemistry, Geophysics, Geosystems 10.1002/2014GC005459

Figure 14. Model for a subduction zones and the formation of a volcanic front [after Schmidt and Poli, 1998]. The model highlights the
continuous water release up to 200 km (see green arrows). Dotted red lines indicate pressures and depths estimated for the metamor-
phic peaks at Beth-Ghinivert [Giacometti and Rebay, 2013] and Servette [Martin et al., 2008] and help distinguish the position of the two
deposits in the subducting slab during the eo-Alpine event.

the results of in situ sulfur isotope analysis to infer that contributions from blueschist-retrogression fluids
are recorded by metabasites from the Zermatt Saas, but isotopic heterogeneities preserved at the milli-
meter scale suggest that fluid/rock ratios were not large enough to be compatible with a fluid-buffered sys-
tem with respect to sulfur.
It is therefore clear that the extent of sulfur isotope modification varies. Likely causes for the variation
include the extent to which fluids were focused through the sampled rocks during metamorphism or retro-
gression, the extent of deformation, and the pressure and temperature conditions. In situ sampling
approaches are more likely to recognize limited length-scale equilibration, whereas bulk sampling
approaches provide a useful dm-scale average compositions.

7.7. Implications for Cycling of Sulfur and Chalcophile Elements in Subduction Zones
Despite the evidence of local interaction of sulfides with S-bearing fluids during the different stages of the
metamorphic evolution at Servette, the absence of pyrrhotite and magnetite replacing pyrite (through reac-
tions like FeS2 5 FeSx 1 (2 2 x)S) suggests limited S release from the deposit itself. This conclusion is consistent
with observations by Dale et al. [2009] who studied samples from the Zermatt-Saas Zone and asserted that no
significant decrease of S content accompanied recrystallization. Under these circumstances, little or no S would
be released from sulfides into the mantle wedge at temperatures lower than, or equal to, those studied.
If breakdown of pyrite and chalcopyrite, which host siderophile and chalcophile elements, in the studied
deposits, is limited, then it is unlikely that these elements are mobilized from subducted sulfide ore deposits
at pressures and temperatures below that of the blueschist-ecologite transition [cf. Tomkins, 2010].

8. Conclusions
Interpretation of S isotopes records in two sulfide deposits from the Italian Western Alps, combined with
mineralogical, textural, and geochemical study has provided valuable insights into the mobilization of sulfur
and metal-sulfur complexes during the Alpine Orogeny. Evidence of fluid-rock interaction during metamor-
phism is found at Servette, but not at the lower grade Beth Ghinivert deposit. The low fluid/rock ratio during
metamorphism at Beth-Ghinivert is consistent with the depth reached during subduction. Sulfur isotopes
were not homogenized at the sample scale within pyrite at either deposit, whereas dynamic recrystalliza-
tion of chalcopyrite during deformation may have promoted homogenization of S isotopes at short length
scales for this mineral.
Neither of the deposits record evidence of intense fluid circulation or sulfide release during metamorphism.
Therefore, the release of reduced sulfur to the mantle wedge from subducted sulfide deposits is expected
to be minimal at pressure and temperatures less than those of the eclogite facies.

GIACOMETTI ET AL. C 2014. American Geophysical Union. All Rights Reserved.

V 3826
 Geochemistry, Geophysics, Geosystems 10.1002/2014GC005459

Acknowledgments References
The authors acknowledge the facilities
Agard, P., B. Goff e, J. L. R. Touret, and O. Vidal (2000), Retrograde mineral and fluid evolution in high-pressure metapelites (Schistes Lustr es
and the scientific and technical
unit, Western Alps), Contrib. Mineral. Petrol., 140, 269–315, doi:10.1007/s004100000190.
assistance of the Australian
Agard, P., L. Jolivet, and B. Goff e (2001), Tectonometamorphic evolution of the Schistes Lustres complex: Implications for the exhumation
Microscopy and Microanalysis
of HP and UHP rocks in the Western Alps, Bull. Soc. Geol. Fr., 172(5), 617–636, doi:10.2113/172.5.617.
Research Facility at the Centre for
Alirezaei, S., and E. M. Cameron (2001), Variations of sulfur isotopes in metamorphic rocks from Bamble Sector, southern Norway: A laser
Microscopy, Characterization and
probe study, Chem. Geol., 181, 23–45, doi:10.1016/S0009-2541(01)00266-2.
Analysis, the University of Western
Alt, J. C., and M. Chaussidon (1989), Ion microprobe analyses of the sulfur isotopic composition of sulfides in hydrothermally altered rocks,
Australia, a facility funded by the
DSDP/ODP Hole 504B, Proc. Ocean Drill. Program Sci. Results, 111, 41–45, doi:10.2973/odp.proc.sr.111.151.1989.
University, State and Commonwealth
Alt, J. C., W. C. Shaks III, and M. C. Jackson (1993), Cycling of sulfur in subduction zones: The geochemistry of sulfur in the Mariana Island
Governments. K.E. thanks the
Arc and back-arc trough, Earth Planet. Sci. Lett., 119, 477–494, doi:10.1016/0012-821X(93)90057-G.
Australian Research Council for grant
Alt, J. C., W. C. Shanks III, L. Crispin, L. Gaggero, E. M. Schwarzenbach, L. Gretchen, G. L. Fr€ uh-Green, and S. M. Bernasconi (2012), Uptake of
FT120100579 that contributed toward
carbon and sulfur during seafloor serpentinization and the effects of subduction metamorphism in Ligurian peridotites, Chem. Geol.,
this research. This is a TiGeR paper and
322–323, 268–277, doi:10.1016/j.chemgeo.2012.07.009.
a contribution from the ARC Centre of
Alt, J. C., E. M. Schwarzenbach, G. L. Fr€ uh-Green, W. C. Shanks III, S. M. Bernasconi, C. J. Garrido, L. Crispini, L. Gaggero, J. A. Padron-Navarta,
Excellence for Core to Crust Fluid
and C. Marchesi (2013), The role of serpentinites in cycling of carbon and sulfur: Seafloor serpentinization and subduction metamor-
Systems (http://www.ccfs.mq.edu.au).
phism, Lithos, 178, 40–54, doi:10.1016/j.lithos.2012.12.006.
F.G. acknowledges funding from the
Angiboust, S., P. Agard, L. Jolivet, and O. Beyssac (2009), The Zermatt-Saas ophiolite: The largest (60 km wide) and deepest (c. 70–80 km)
Italian Ministry of Education, University
continuous slice of oceanic lithosphere detached from subduction zone?, Terra Nova, 21, 171–180, doi:10.1111/j.1365-
and Research for his PhD scholarship.
3121.2009.00870.x.
J. C. Alt and J. H. Dilles are thanked for
Bachinski, D. J. (1977), Sulfur isotopic composition of ophiolitic cupriferous iron sulfide deposits, Notre Dame Bay, Newfoundland, Econ.
carefully reviewing this manuscript.
Geol., 72, 243–257, doi:10.2113/gsecongeo.72.2.243.
Details of the analyses of S isotopes
Barnicoat, A. C., and I. Cartwright (1995), Focused fluid flow during subduction: Oxygen isotope data from high-pressure ophiolites of the
(d34S and r values) which support this
western Alps, Earth Planet. Sci. Lett., 132, 53–61, doi:10.1016/0012-821X(95)00040-J.
work are available as supporting
Barrie, C. D., M. A. Pearce, and A. P. Boyle (2011), Reconstructing the pyrite deformation mechanism map, Ore Geol. Rev., 39, 265–276, doi:
information.
10.1016/j.oregeorev.2011.03.006.
Bebout, G. E. (2007), Metamorphic chemical geodynamics of subduction zones, Earth Planet. Sci. Lett., 260, 373–393, doi:10.1016/
j.epsl.2007.05.050.
Bouquet, J., and M. C. Forette (1973), Sur la deerite de l’ensemble des calchschistes pi emontais a Troncea (Italie), Bull. Soc. Fr. Mineral. Cris-
tallogr., 96, 314–316.
Calvert, S. E., H. G. Thode, D. Yeung, and R. E. Karlin (1996), A stable isotope study of pyrite formation in the Late Pleistocene and Holocene
sediments of the Black Sea, Geochim. Cosmochim. Acta, 60(7), 1261–1270, doi:10.1016/0016-7037(96)00020-8.
Carminati, E., and C. Doglioni (2012), Alps vs. Apennines: The paradigm of a tectonically asymmetric Earth, Earth Sci. Rev., 112, 67–98, doi:
10.1016/j.earscirev.2012.02.004.
Cartwright, I., and A. C. Barnicoat (1999), Stable isotope geochemistry of Alpine ophiolites: A window to ocean-floor hydrothermal altera-
tion and constraints on fluid-rock interaction during high-pressure metamorphism, Int. J. Earth Sci., 88, 219–235, doi:10.1007/
s005310050261.
Castellarin, A. (2001), Alps–Apennines and Po Plain-frontal Apennines relations, in Anatomy of an Orogen: The Apennines and the Adjacent
Mediterranean Basins, edited by G. B. Vai and J. P. Martini, pp. 177–195, Kluwer Acad., Norwell, Mass.
Castello, P. (1981), Inventario delle mineralizzazioni a magnetite, ferro-rame e manganese del complesso piemontese dei calcescisti con
pietre verdi in valle d’Aosta, Ofioliti, 6, 5–46.
Chaussidon, M., F. Albarède, and S. M. F. Sheppard (1987), Sulphur isotope heterogeneity in the mantle from ion probe measurements of
sulphide inclusions in diamonds, Nature, 330, 242–244.
Condie, K. C. (1967), Oxygen, carbon dioxide and sulfur fugacities during diagenesis and low-grade metamorphism of late precambrian
subgraywackes from Northern Utah, Am. Mineral., 52, 1153–1159.
Cook, N. J., and J. Hoefs (1997), Sulphur isotope characteristics of metamorphosed Cu-(Zn) volcanogenic massive sulphide deposits in the
Norwegian Caledonides, Chem. Geol., 135, 307–324, doi:10.1016/S0009-2541(96)00119-2.
Cooke, D. R. C., and P. H. Hollings (2005), Giant porphyry deposits: Characteristics, distribution, and tectonic controls, Econ. Geol., 100, 801–
818, doi:10.2113/gsecongeo.100.5.801.
Cox, S. F. (1987), Flow mechanisms in sulphide minerals, Ore Geol. Rev., 2, 133–171, doi:10.1016/0169-1368(87)90026-6.
Crowe, D. E. (1994), Preservation of original hydrothermal d34S values in greenschist to upper amphibolite volcanogenic massive sulfide
deposits, Geology, 22, 873–876, doi:10.1130/0091-7613(1994)022<0873:POOHSV>2.3.CO;2.
Dal Piaz, G. V., J. C. Hunziker and G. Martinotti (1972), La zona Sesia-Lanzo e l’evoluzione tettono-metamorfica delle Alpi nord occidentali
interne, Mem. Soc. Geol. Ital., 11, 433–466.
Dal Piaz, G. V., A. Bistacchi, and M. Massironi (2003), Geological outline of the Alps, Episodes, 26(3), 175–180.
Dale, C. W., K. W. Burton, D. G. Pearson, A. Gannoun, O. Alard, T. W. Argles, and I. J. Parkinson (2009), Highly siderophile element behaviour
accompanying subduction of oceanic crust: Whole rock and mineral-scale insights from a high-pressure terrain, Geochim. Cosmochim.
Acta, 73, 1394–1416, doi:10.1016/j.gca.2008.11.036.
de Hoog, J. C. M., B. E. Taylor, and M. J. Bergen (2001), Sulfur isotope systematics of basaltic lavas from Indonesia: Implications for the sulfur
cycle in subduction zones, Earth Planet. Sci. Lett., 189, 237–252.
de Moor, J. M., P. L. King, Z. D. Sharp, and T. P. Fischer (2011), A model for sulfur speciation and sulfur isotope fractionation during mag-
matic degassing on Earth and Mars, paper presented at 42nd Lunar and Planetary Science Conference, Lunar and Planetary Institute,
Houston, Texas.
Evans, K. A. (2006), Redox decoupling and redox budgets: Conceptual tools for the study of earth systems, Geology, 34, 489–492, doi:
10.1130/G22390.1.
Evans, K. A. (2012), The redox budget of subduction zones, Earth Sci. Rev., 113, 11–32, doi:10.1016/j.earscirev.2012.03.003.
Evans, K. A., and A. G. Tomkins (2011), The relationship between subduction zone redox budget and arc magma fertility, Earth Planet. Sci.
Lett., 308, 401–409, doi:10.1016/j.epsl.2011.06.009.
Evans, K. A., A. G. Tomkins, J. Cliff, and M. L. Fiorentini (2014), Insights into subduction zone sulfur recycling from isotopic analysis of
eclogite-hosted sulfides, Chem. Geol., 365, 1–19, doi:10.1016/j.chemgeo.2013.11.026.
Garuti, G., A. Pura, J. A. Proenza, and F. Zaccarini (2009), Sulfur-isotope variations in sulfide minerals from massive sulfide deposits of the
Northern Apennine Ophiolites: Inorganic and biogenic constrains, Ofioliti, 34(1), 43–62, doi:10.4454/ofioliti.v34i1.377.

GIACOMETTI ET AL. C 2014. American Geophysical Union. All Rights Reserved.

V 3827
 Geochemistry, Geophysics, Geosystems 10.1002/2014GC005459

Giacometti, F., and G. Rebay (2013), Structural and petrological evolution of the Beth-Ghinivert zone (Schistes Lustres—Italian Western
Alps), Rend. Online Soc. Geol. Ital., 29, 70–73.
Hannington, M. D. (2013), The role of black smokers in the Cu mass balance of the oceanic crust, Earth Planet. Sci. Lett., 374, 215–226, doi:
10.1016/j.espl.2013.06.004.
Hattori, K. H., and S. Guillot (2007), Geochemical character of serpentinites associated with high to ultrahigh-pressure metamorphic rocks
in the Alps, Cuba, and the Himalayas: Recycling of elements in subduction zones, Geochem. Geophys. Geosyst., 8, Q09010, doi:10.1029/
2007GC001594.
Hoffman, A. W. (1998), Chemical differentiation of the Earth: The relationship between mantle, continental crust, and oceanic crust, Earth
Planet. Sci. Lett., 90, 297–314, doi:10.1016/0012-821X(88)90132-X.
Hunziker, J. C. (1974), Rb-Sr and K-Ar Age Determination and the Alpin Tectonic History of The Western Alps, Mem. degli Ist. di Geol. Mineral.
dell’Univ. di Padova., vol. 31, edited by Padova, pp. 1–55.
Itaya, T., R. N. Brothers, and P. M. Black (1985), Sulfides, oxides and sphene in high-pressure schists from new Caledonia, Contrib. Mineral.
Petrol., 91, 151–162, doi:10.1007/BF00377762.
Janecky, D. R., and W. R. Seyfried Jr. (1984), Formation of massive sulfide deposits on ocean ridge crests: Models for mixing between hydro-
thermal solutions and seawater, Geochim. Cosmochim. Acta, 48, 2723–2738, doi:10.1016/0016-7037(84)90319-3.
Kawakami, T., D. J. Ellis, and A. G. Christy (2006), Sulfide evolution in high-temperature to ultrahigh-temperature metamorphic rocks from
L€
utzow–Holm Complex, East Antarctica, Lithos, 92, 431–446, doi:10.1016/j.lithos.2006.03.057.
Kelley, K. A., and E. Cottrell (2009), Water and the oxidation state of subduction zone magmas, Science, 325, 605–607, doi:10.1126/
science.1174156.
Kerrick, D. M., and J. A. D. Connolly (2001), Metamorphic devolatilization of subducted oceanic metabasalts: Implications for seismicity, arc
magmatism and volatile recycling, Earth Planet. Sci. Lett., 189, 19–29, doi:10.1016/S0012-821X(01)00347-8.
Kontny, A., G. Friedrich, H. J. Behr, H. de Wall, E. E. Horn, P. Moller, and G. Zulauf (1997), Formation of ore minerals in metamorphic rocks of
the German continental deep drilling site (KTB), J. Phys. Res., 102(8), 18323–18336, doi:10.1029/96JB03395.
Krutow-Mozgawa, A. (1988), M etamorphisme dans les s ediments riches en fer ou magnesium de la couverture des ophiolites pi emontaises
(mine de Servette, Val d’Aoste), PhD thesis, 166 pp., Univ. Pierre Marie Curie, Paris.
Lemoine, M., and P. Tricart (1986), Les Schistes lustr es piemontais des Alpes occidentales: Approche stratigraphique, structurale et
s
edimentologique, Eclogae Geol. Helv., 79, 271–294.
Lemoine, M., et al. (1986), The continental margin of the Mesozoic Tethys in the Western Alps, Mar. Pet. Geol., 3, 179–199, doi:10.1016/
0264-8172(86)90044-9.
Lemoine, M., P. Tricart, and G. Boillot (1987), Ultramafic and gabbroic ocean floor of the Ligurian Tethys (Alps, Corsica, Apennines): In
search of a genetic model, Geology, 15(7), 622–625, doi:10.1130/0091-7613(1987)15<622:UAGOFO>2.0.CO;2.
Manning, C. E. (2004), The chemistry of subduction-zone fluids, Earth Planet. Sci. Lett., 223, 1–16, doi:10.1016/j.epsl.2004.04.030.
Manning, C. E. (2011), Sulfur surprises in deep geological fluids, Science, 331, 1018–1019, doi:10.1126/science.1202468.
Marini, L., R. Moretti, and M. Accornero (2011), Sulfur isotopes in magmatic-hydrothermal systems, melts and magmas, Rev. Mineral. Geo-
chem., 73, 423–492, doi:10.2138/rmg.2011.73.14.
Marshall, B., and L. B. Gillian (1993), Remobilization, syn-tectonic processes and massive sulfide deposits, Ore Geol. Rev., 8, 39–64, doi:
10.1016/0169-1368(93)90027-V.
Martin, S., G. Godard, and G. Rebay (2004), Walking on a Palaeo Ocean floor. The subducted tethys in the Western Alps—An excursion
guide, J. Virtual Explorer, 16(2), 1–46.
Martin, S., G. Rebay, J. R. Kienast, and C. M evel (2008), An eclogitised oceanic paleo-hydrothermal field from the St. Marcel Valley (Italian
Western Alps), Ofioliti, 33(1), 49–63, doi:10.4454/ofioliti.v33i1.359.
Natale, P. (1966), Sulla pirite di alcuni giacimenti piritoso-cupriferi stratiformi delle Alpi Occidentali, Bollettino della Associazione Mineraria
Subalpina, 3–4, 356–363.
Natale, P. (1969), Re-crystallization and remobilization in some stratiform pyrite deposits of the Western Alps, paper presented at Con-
vegno Sulla Rimobilizzazione dei Minerali Metallici e Non Metallici, Universita degli Studi di Cagliari, Cagliari, Italy.
Natale, P., and A. Visetti (1980), Contributo alla conoscenza minerogenica delle piriti di origine esalativo-sedimentaria, Bollettino della Asso-
ciazione Mineraria Subalpina, 1, 180–210.
Natale, P., and S. Zucchetti (1966), Studi sui giacimenti piritoso-cupriferi stratiformi della Alpi Occidentali—Nota I. Compendio selle cono-
scenze attuali delle piriti stratiformi, L’Industria Mineraria, 17, 443–451.
Novarese, V. (1900), La miniera del Beth e Ghinivert. Rassegna mineraria della industria chimica e delle industrie mineralurgiche e metallur-
giche, 12, 7 (97–99), 8 (113–115), and 9 (131–133).
Ohmoto, H., and M. B. Goldhaber (1997), Sulfur and carbon isotopes, in Geochemistry of Hydrothermal Ore Deposits, edited by H. L. Barnes,
pp. 517–611, John Wiley, N. Y.
Ohmoto, H., and A. C. Lasaga (1982), Kinetics of reactions between aqueous sulfates and sulfides in hydrothermal systems, Geochim. Cos-
mochim. Acta, 46, 1727–1745, doi:10.1016/0016-7037(82)90113-2.
Ohmoto, H., and R. O. Rye (1979), Isotopes of sulfur and carbon, in Geochemistry of Hydrothermal Ore Deposits, edited by H. L. Barnes, pp.
509–567, John Wiley, N. Y.
Phillipot, P., and M. Scambelluri (1996), The composition and behaviour of fluids in high-pressure rocks from the Alps: A review, in Studies
on Metamorphic Rocks and Minerals of the Western Alps. A Volume in Memory of Ugo Pognante, Boll. Museo Reg. Sci. Nat., vol. 13, edited
by B. Lombardo, pp. 75–102, Torino, Italy.
Plank, T., and C. H. Langmuir (1998), The chemical composition of subducting sediment and its consequences for the crust and mantle,
Chem. Geol., 145, 325–394, doi:10.1016/S0009-2541(97)00150-2.
Pognante, U. (1991), Petrological constrains on the eclogite- and blueschist-facies metamorphism and P-T-t path in western Alps, J. Meta-
morph. Geol., 9, 5–17, doi:10.1111/j.1525-1314.1991.tb00501.x.
Polino, R., G. V. Dal Piaz, and G. Gosso (1990), Tectonic erosion at the Adria Margin and accretionary processes for the Cretaceous orogeny
of the Alps, Mem. Soc. Geol. Fr., 156, 345–367.
Pokrovski, G. S., and L. S. Dubrovinsky (2011), The S3- ion is stable in geological fluids at elevated temperatures and pressures, Science, 331,
1052–1054, doi:10.1126/science.1199911.
Puchelt, H., and H. W. Hubberten (1980), Preliminary results of sulfur isotope investigations on deep sea drilling project cores from Legs 52
and 53, in Initial Reports of the Deep Sea Drilling Project, edited by T. Donnelly et al., vol. 51–53, pp. 1145–1148, U.S. Gov. Print. Off.,
Washington, D. C.
Rebay, G., and R. Powell (2012), Eclogite-facies sea-floor hydrothermally-altered rocks: Calculated phase equilibria for an example from the
Western Alps at Servette, Ofioliti, 37(1), 55–63, doi:10.4454/ofioliti.v37i1.405.

GIACOMETTI ET AL. C 2014. American Geophysical Union. All Rights Reserved.

V 3828
 Geochemistry, Geophysics, Geosystems 10.1002/2014GC005459

Rubatto, D., D. Gebauer, and M. Fanning (1998), Jurassic formation and eocene subduction of the Zermatt-Saas-Fee ophiolites: Implications
for the geodynamic evolution of the Central and Western Alps, Contrib. Mineral. Petrol., 132, 269–287, doi:10.1007/s004100050421.
Reed, M. H., and J. Palandri (2006), Sulfide mineral precipitation from hydrothermal fluids, Rev. Mineral. Geochem., 61, 609–631, doi:
10.2138/rmg.2006.61.11.
Sangster, D. F. (1971), Sulphur isotopes, stratabound sulphide deposits, and ancient seas, Soc. Mining Geol. Japan, Special Issue Proc. IMA-
IAGOD Meeting 70-3, 295–299.
Sawkins, F. J. (1990), Integrated tectonic-genetic model for volcanic-hosted massive sulfide deposits, Geology, 18, 1061–1064, doi:10.1130/
0091-7613(1990)018<1061:ITGMFV>2.3.CO;2.
Scambelluri, M., G. B. Piccardo, P. Philippot, A. Robbiano, and L. Negretti (1997), High salinity fluid inclusions formed from recycled sea-
water in deeply subducted alpine serpentinite, Earth Planet. Sci. Lett., 148, 485–499, doi:10.1016/S0012-821X(97)00043-5.
Scambelluri, M., O. M€ untener, L. Ottolini, T. T. Pettke, and R. Vannucci (2004), The fate of B, Cl and Li in the subducted oceanic mantle and
in the antigorite breakdown fluids, Earth Planet. Sci. Lett., 222, 217–234, doi:10.1016/j.epsl.2004.02.012.
Schmidt, M. W., and S. Poli (1998), Experimentally based water budgets for dehydrating slabs and consequences for arc magma genera-
tion, Earth Planet. Sci. Lett., 163, 361–379, doi:10.1016/S0012-821X(98)00142-3.
Schwarzenbach, E. M., G. L. Fr€ uh-Green, S. M. Bernasconi, J. C. Alt, W. C. Shanks III, L. Gaggero, and L. Crispini (2012), Sulfur geochemistry of
peridotite-hosted hydrothermal systems: Comparing the Ligurian ophiolites with oceanic serpentinites, Geochim. Cosmochim. Acta, 91,
283–305, doi:10.1016/j.gca.2012.05.021.
Seal, R. R. (2006), Sulfur isotope geochemistry of sulfide minerals, Rev. Mineral. Geochem., 61, 633–677, doi:10.2138/rmg.2006.61.12.
Shi, P. (1992), Fluid fugacities equilibria in the Fe-Si-O-H-S system, Am. Mineral., 77, 1050–1066.
Sillitoe, H. R. (1972a), A plate tectonic model for the origin of porphyry copper deposits, Econ. Geol., 67, 184–197, doi:10.2113/
gsecongeo.67.2.184.
Sillitoe, H. R. (1972b), Geology of the Los Pelambres porphyry copper deposit, Chile, Econ. Geol., 68, 1–10, doi:10.2113/gsecongeo.68.1.1.
Spandler, C., and C. Pirard (2013), Element recycling from subducting slabs to arc crust: A review, Lithos, 170–171, 208–223, doi:10.1016/
j.lithos.2013.02.016.
Strauss, H., and J. Schieber (1990), A sulfur isotope study of pyrite genesis: The mid-Proterozoic Newland Formation, Belt Supergroup, Mon-
tana, Geochim. Cosmochim. Acta, 54(1), 197–204, doi:10.1016/0016-7037(90)90207-2.
Tartarotti, P., S. Martin, and R. Polino (1986), Geological data about the ophiolitic sequences in the St. Marcel Valley (Aosta Valley), Ofioliti,
11(3), 343–346.
Tomkins, A. G. (2007), Three mechanisms of ore remobilisation at the amphibolite facies Montauban Zn-Pb-Au-Ag deposit, Miner. Deposita,
42, 627–637, doi:10.1007/s00126-007-0131-9.
Tomkins, A. G. (2010), Windows of metamorphic sulfur liberation in the crust: Implications for gold deposit genesis, Geochim. Cosmochim.
Acta, 74, 3246–3259, doi:10.1016/j.gca.2010.03.003.
Tricart, P., and S. Schwartz (2006), A north-south section across the Queyras Schistes lustr es (Piedmont zone, Western Alps): Syn-collision
refolding of a subduction wedge, Eclogae Geol. Helv., 99, 429–442, doi:10.1007/s00015-006-1197-6.
Velasco, F., J. Sanchez-Espa~ na, A. J. Boyce, A. E. Fallick, R. Saez, and G. R. Almod
ovar (1998), A new sulphur isotopic study of some Iberian
Pyrite Belt deposits: Evidence of a textural control on sulphur isotope composition, Miner. Deposita, 24, 4–18, doi:10.1007/
s001260050182.
Wagner, T., and A. J. Boyce (2006), Pyrite metamorphism in the Devonian Hunsr€ uck slate of Germany: Insights from laser microprobe sulfur
isotope analysis and thermodynamic modeling, Am. J. Sci., 306, 525–552, doi:10.2475/07.2006.02.
Wood, S. A. (1998), Calculation of activity-activity and log fO2—pH diagrams, in Techniques in Hydrothermal Ore Deposits Geology, Rev. Econ.
Geol., vol. 10, edited by J. P. Richards and P. B. Larson, Soc. of Econ. Geol. Inc., Littleton, Colo.

GIACOMETTI ET AL. C 2014. American Geophysical Union. All Rights Reserved.

V 3829