Available online at www.sciencedirect.

com

Gondwana Research 12 (2007) 454 – 467

www.elsevier.com/locate/gr

U–Pb and Sm–Nd geochronology of amphibolites from the Curaçá Belt,

São Francisco Craton, Brazil: Tectonic implications
Luiz José Homem D'el-Rey Silva a,⁎, Elton Luiz Dantas b , João Batista Guimarães Teixeira c ,
Jorge Henrique Laux b , Maria da Glória da Silva c,d
a
Universidade de Brasília — Instituto de Geociências
b
Laboratório de Geocronologia, Universidade de Brasília, Campus Darcy Ribeiro, Asa Norte, 70.910-900, Brasília — DF, Brazil
c
Grupo de Metalogênese, Centro de Pesquisa em Geofísica e Geologia, Universidade Federal da Bahia, Campus Universitário de Ondina,
Sala 201-C, Salvador, 40170-290, Bahia, Brazil
d
Serviço Geológico do Brasil (CPRM) - Av. Pasteur 404, Urca, 22290-240 Rio de Janeiro — RJ, Brazil
Received 24 May 2006; received in revised form 29 July 2006; accepted 23 November 2006
Available online 16 January 2007

Abstract

The Curaçá terrane is part of the Itabuna–Salvador–Curaçá (I–S–C) Paleoproterozoic orogen in the São Francisco craton, northeastern Brazil,
and comprises supracrustal rocks, gneisses of their probable basement, amphibolites, and mafic-ultramafic Cu-bearing bodies (including the
Caraíba Cu-Mine), all affected by D1-D3 deformation events associated to M1-M3 metamorphism under high-T granulite and amphibolite facies,
and assisted by G1-G3 tonalitic-granodioritic-granitic intrusions. U–Pb and Sm–Nd Thermal Ionization Mass Spectrometry (TIMS) isotopic data
from amphibolite, tonalite, and granite, sampled in a well-known outcrop, indicate partial reset and heterogeneous modification of the original
isotopic systems, attributable to deformation and metamorphism. The ages obtained from these systems agree with each other, and also with other
previously published U–Pb data, and imply that 2.6 Ga is the crystallization age of the protolith of the amphibolite. Together with key structural
relationships, they also indicate a 2.08–2.05 Ga interval for M3 metamorphism, and make even a less precise age (2.2–2.3 Ga) acceptable, as it
suggests contamination in the amphibolite with material in a syn-D2 tonalite crystallized 2248 ± 36 Ma ago. The new data demonstrate the
existence of Neoarchean fragments of both oceanic and continental crusts and constrain the Archean-Paleoproterozoic development of the Curaçá
belt, the I–S–C orogen, and the São Francisco craton.
© 2006 International Association for Gondwana Research. Published by Elsevier B.V. All rights reserved.

Keywords: U–Pb and Sm–Nd isotopic data; Curaçá terrane; São Francisco craton; Archean to Paleoproterozoic tectonics; Caraíba Mine

1. Introduction belts relies on detailed structural studies of key outcrops

combined with geochronological methods (Gromet, 1991).
Proterozoic mobile belts or terranes of high metamorphic The Curaçá River Valley terrane is the N–S trending
grade are among the most characteristic features of continental northern segment of the Itabuna–Salvador–Curaçá (I–S–C)
shield areas (Windley, 1986; Condie, 1989). These terranes Paleoproterozoic orogenic belt that stretches between the
consist of polydeformed/metamorphosed rocks displaying a Mesoarchean Gavião, Serrinha, and Jequié cratonic blocks, in
generally older regional metamorphic banding that may hide the the central-eastern part of the São Francisco craton, Bahia,
evidence for polycyclic tectonic events recognisable only upon eastern Brazil (Fig. 1; Sabaté et al., 1990; Barbosa et al., 1996).
detailed mapping combined with lithogeochemistry (Passchier The existence of reworked Archean rocks in the Curaçá terrane
et al., 1990). Consequently accurate dating of protoliths (Inda and Barbosa, 1978; Lindenmayer, 1981), and across the
crystallization and/or deformation and metamorphism in such entire I–S–C orogen, has driven Leite (2002) and Barbosa and
Sabaté (2002, 2004) to postulate the participation of a fourth
⁎ Corresponding author. Tel.: +55 61 30324907; fax: +55 61 33474062. Archean block inside the Paleoproterozoic collision zone.
E-mail addresses: [email protected], [email protected] Geodynamic models for the Archean-Paleoproterozoic evolu-
(L.J.H. D'el-Rey Silva). tion of the São Francisco craton have been supported by recent
1342-937X/$ - see front matter © 2006 International Association for Gondwana Research. Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.gr.2006.11.008
 L.J.H. D'el-Rey Silva et al. / Gondwana Research 12 (2007) 454–467 455

Fig. 1. a — Summary display of the Gavião, Serrinha and Jequié blocks, and the Itabuna–Salvador–Curaçá (I–S–C) orogen, all within the geographical limits of the
Bahia State, and part of the São Francisco craton, Brazil. Numbers in brackets represent TDM model age data for each block (adapted from Barbosa and Sabaté,
2004). b — Simplified geological map of the part of northern Bahia to show the northern part of the I–S–C orogen (=Curaçá terrane), between the Gavião and the Serrinha
Archean blocks (combining Figs. 1 in D'el-Rey Silva, 1984, 1985; and in Oliveira et al., 2004). The Caraíba orebody and the area of Fig. 2 are highlighted. Details in text.

isotopic studies (Oliveira et al., 2004; Barbosa and Sabaté, crystallization ages of: 2580 ± 10 Ma for a sample of norite
2004), indicating that the I–S–C orogen and the three Archean collected at the Caraíba Mine (U–Pb Sensitive High Resolution
blocks preserve distinct Sm–Nd characteristics (Fig. 1a). Ion Microprobe – SHRIMP – data in Oliveira et al., 2004);
Early regional studies (Delgado and Souza, 1975) made the 2248 ± 36 Ma for syn-D2 tonalite in the Caraíba Airport outcrop;
Curaçá terrane well-known because of its high potential for Cu- and 2051 ± 16 Ma for syn-D3 granite intrusive in the Caraíba
sulphide mineralization associated with mafic–ultramafic (M– orebody (both U–Pb data, but respectively from zircons and
UM) bodies mapped within high metamorphic grade supra- monazite; D'el-Rey Silva et al., 1996), whereas the age of the
crustal rocks (metasediments and metavolcanics), gneisses, supracrustals was not directly determined.
migmatites, and metagranitoids. In fact, following the acciden- This paper presents the results of U–Pb and Sm–Nd
tal discovery of green copper oxide (malachite) nearly 140 years integrated studies carried out on samples of amphibolite,
ago in the site destined to become the Caraíba Mine (Fig. 1b), tonalite and granite of the Caraíba Airport outcrop, situated
six other smaller Cu-deposits have been found in the two last ∼5 km to the north of the Caraíba orebody (Fig. 2). This
decades, among over two hundred small bodies (D'el-Rey outcrop is one of the most visited places in the Curaçá belt
Silva, 1984, 1985). Structures record progressive (D1-D3) exactly because it displays clear evidence for a D1-D3
deformation (Fig. 2; D'el-Rey Silva, 1993) and conditions for progressive deformation under M1-M3 metamorphism, provid-
the main metamorphism have been estimated around 850°– ing a well-constrained structural-metamorphic framework for
720 °C (Jardim de Sá et al., 1982; Ackermand et al., 1987), or targeting geochronological studies.
up to 1000 °C, with pressures varying from 8–10 kbar (Leite,
2002; Leite et al., 2005), all on the basis of several 2. Regional geology and tectonic setting
geothermometers and sapphirine-bearing assemblages.
The first U–Pb isotopic data for rocks of the Curaçá belt, all The basement of the São Francisco Craton is mostly
from the surroundings of the Caraíba orebody (Fig. 2), indicated represented by the Gavião and Serrinha blocks, shown
456 L.J.H. D'el-Rey Silva et al. / Gondwana Research 12 (2007) 454–467

Fig. 2. Simplified geological map of the area surrounding the Caraíba orebody, highlighting the Curaçá antiform, the open pit, and the land track of the airport. Based
on a 1:10,000 scale geological mapping carried out by D'el-Rey Silva (1984, 1985) along 130 km of E–W tracks distributed in 18 lines 500 m apart one from each
other, some marked every 100 m, others marked every 50 m, a real frame (omitted for simplicity) opened in the field for geophysical survey. Mapping of the orebody
itself was at 1:1,000 scale (see Fig. 3a) whereas the part to the N of the open pit (belonging to DOCEGEO Company) was mapped at 1:5,000 scale (Lindenmayer et al.,
1984). See details in text.

respectively to the west and to the east of the Curaçá terrane, in bosa and Sabaté, 2004). Its westernmost part is covered by
their simplest map expression (Fig. 1b). The Gavião block is (meta)sediments deformed in the Neoproterozoic Brasiliano
mainly composed of gneiss-amphibolite associations, amphib- orogeny and by Quaternary limestones, whereas the eastern part
olite-facies tonalite-granodiorite orthogneisses dated at 2800– includes (to the west of the town of Bonfim, Fig. 1b) the Campo
2900 Ma as well as greenstone belts. The block also encloses a Formoso Cr-bearing mafic–ultramafic layered body, a major
3200–3400 Ma trondhjemite-tonalite-granodiorite suite (Bar- intrusion that is unconformably overlain by the Jacobina Group,
 L.J.H. D'el-Rey Silva et al. / Gondwana Research 12 (2007) 454–467 457

both of which are intruded by the Late Paleoproterozoic Campo Lindenmayer (1981), Lindenmayer et al. (1984), D'el-Rey Silva
Formoso granite (Silva, 1996; Leite, 2002). According to a (1984, 1985), and lately Mayer and Barnes (1996) and D'el-Rey
summary in Silva (1996) the Jacobina Group is part of a much Silva et al. (1988, 1994, 1996) to state that the Cu-bearing rocks
longer Paleoproterozoic rift-like volcanic-sedimentary basin are sill-like structures. It is noteworthy that the Caraíba orebody
extending to the boundary between the Gavião and Jequié is a Cu-only, chalcopyrite and bornite-rich deposit and
blocks. associated with deep ocean rocks such as banded iron formation
The Serrinha block consists of amphibolite-facies 2900– and evaporites (anhydrate) forsterite-marble. The third unit (the
3500 Ma ortho and paragneisses of granodiorite composition, as Caraíba complex) mostly comprises migmatitic gneisses and
well as migmatites and amphibolites (Barbosa and Sabaté, syntectonic intrusions (G1 and G2) generally consisting of grey
2004). The western margin is in contact with the Itiúba syenite, tonalites and granodiorites. Reddish-pink coloured, K-rich (G3)
whereas the eastern margin encloses Paleoproterozoic green- granites intruded all lithotypes and developed in association
stone sequences, and arc-like orthogneiss and granites that are with D3 (Fig. 2).
indicators of a back-arc basin system within the Serrinha block
(Silva, 1996). Farther to the east, this block underlies 3.2. Structures and metamorphism
Neoproterozoic (meta)sediments and, outside of Fig. 1b,
Mesozoic sediments of the Tucano basin. The Curaçá belt has been affected by a D1-D3 progressive
The I–S–C orogen (Fig. 1) is mainly composed of granulite- deformation assisted by large volumes of syntectonic G1-G3
facies tonalitic and charnockitic rocks with basic-ultrabasic granitoid intrusions. Associated metamorphisms M2 (granulite
enclaves, as well as supracrustal rocks, in both of its northern facies) and M3 (amphibolite facies) are quite evident every-
and southern segments (Barbosa and Sabaté, 2004). The where, but evidence for a M1 amphibolite facies metamorphism
2084 Ma old Itiúba syenite separates the 2580 Ma old Curaçá is solely preserved in paragneisses and amphibolites that occur
Valley Cu-district to the west, from the 2085 Ma old Jacurici as dm- to m-scale boudins or xenoliths entrained in G2
Valley Cr-district to the east (Fig. 1b; all U–Pb SHRIMP age of intrusions across the belt (Jardim de Sá et al., 1982; D'el-Rey
zircons; Oliveira et al., 2004). Silva, 1984, 1985).
Jardim de Sá et al. (1982) first reported M1 amphibolite-
3. Regional geology of the Curaçá high-grade terrane facies metamorphism on the basis of: 1 — Migmatites affected
by D2 and D3 events; 2 — Biotite inclusions (M1) found in
3.1. Lithostratigraphy crystals of orthopyroxene (M2) in paragneisses; and 3 — An
opx-rich border surrounding layers and boudins of mafic rocks
The lithotypes of the Curaçá terrane are divided into the (amphibolites) in paragneisses, all affected by F3 folds
Tanque Novo and Caraíba complexes, both including the Cu- associated with M3 metamorphism (their Fig. 4H). Whereas
bearing mafic-ultramafic bodies (Fig. 1b). The Tanque Novo these observations are respectively from outcrops to the south
Complex comprises Al-rich paragneisses, banded gneisses, and north of Caraíba (the two latter in the Tanque Novo
calcsilicate rocks, quartzites, iron formations and graphite Sequence, Surubim area; Fig. 1b), the existence of M1
gneisses, whereas the Caraíba complex comprises mostly metamorphism in the Caraíba area (Fig. 2) is demonstrated by
enderbitic and charnocktic orthogneisses, as well as migmatites. several inclusions of circular crystals of green hornblende in
Most of the regional geology and structural evolution of the hypersthene from the opx-rich border of mafic boudins in the
Curaçá belt (Lindenmayer, 1981; Figueiredo, 1981; Mandetta, Airport outcrop (photograph 45, page 153 of D'el-Rey Silva's
1982; Gáal, 1982; Hasuy et al., 1982; Jardim de Sá et al., 1982; MSc thesis, 1984). Hypersthene-free amphibolites have been
D'el-Rey Silva, 1984, 1985), published at the same time or soon largely documented across the Curaçá Valley (modal composi-
after the start-up of the Caraíba Mine operation, have shown the tions in page 99 of Lindenmayer, 1981). M 2 mineral
Curaçá terrane as formed of three main lithostratigraphic units assemblages include cordierite-sillimanite-garnet-biotite in
also recorded in the vicinities of the Caraíba orebody (Fig. 2), paragneisses, and andesine plus hypersthene (strongly pleo-
the area submitted to the most detailed mapping carried out in chroic and rich in Al2O3) in commonly granoblastic gabbroic
the Curaçá belt. rocks, even the Cu-bearing bodies (Lindenmayer, 1981). M3
The Tanque Novo Complex is a supracrustal sequence metamorphism resulted in the transformation of hypersthene
consisting mostly of quartzofeldspathic gneiss with thin into hornblende and biotite, in F3 hinges (photographs 50 and
intercalations of amphibolites, cordierite-sillimanite-garnet 51, D'el-Rey Silva's MSc thesis) and is characterized by biotite,
paragneiss, oxide-facies banded iron formation, calcsilicate hornblende, quartz and feldspar assemblages (Lindenmayer,
rocks, anhydrate-bearing marbles, forsterite-marbles and quart- 1981).
zites, as well as gneisses supposed to belong, at least in part, to Due to D1-D3 evolution, the older metamorphic banding (S1)
the crystalline basement. The second unit consists of Cu-poor or and intrafolial folds F1 are affected by two folding phases (F2
simply barren gabbros, gabbronorites, leucogabbros, and and F3), so that S1 is mostly a composite structure where
hyperites, together with Cu-rich hypersthenites, melanorites, evidences for M1 is commonly absent, but M2 and M3
and norites. A vast amount of geology and structural data assemblages dominate. S1 appears in most outcrops as a
collected regionally, and also on surface and in underground continuous banding traceable for 10 m, or more, and consisting
sites of the Caraíba Mine and other small bodies, drove of cm- to dm-thick bands of dark, more mafic (hornblende,
458 L.J.H. D'el-Rey Silva et al. / Gondwana Research 12 (2007) 454–467

biotite, pyroxene) minerals, and bands of light, more felsic mapped (Fig. 3a), and on the tectonic conditions prevailing in
minerals (feldspars and quartz, mostly). S1 may also appear as the I–S–C orogen.
mm- to cm-thick, generally continuous in ≤ 1 m-scale, or D3 flattening was so strong and melt-assisted that the real
somehow diffuse banding in metabasic rocks and amphibolites geometry of some of the 10 cm- to 1 m-scale F2 × F3 fold
such as the ones in the studied outcrop. S1 is also a banding interference patterns seen in the Curacá terrane is often obscured
parallel to S0 defined by the intercalation of 10 cm- to 50 cm- by disrupted hinges, melts injection, and/or extreme ductile
thick layers of acid gneiss (possibly meta-arkose and/or meta- thinning. However, the large amount of data from detailed
rhyolite), amphibolite, pelitic paragneiss, marble, banded iron mapping of the orebody and its surroundings (1978–1984; D'el-
formation, and calcsilicate rocks. Rey Silva, 1984, 1985) and from mine operation in the following
F1 folds are generally of 10 cm-scale, rootless and intrafolial 18 years (D'el-Rey Silva et al., 1988, 1994, 1996; D'el-Rey
relative to S1 (S0), whereas the F2 folds (and foliation S2) are Silva and Oliveira, 1999) allows a picture to be constructed of
mostly seen in outcrops of gneisses and orthogneisses, together the Caraíba Cu-deposit within the core of a 1000 m-high, tight,
with examples of 10–100 cm-sized patterns indicative of non-cylindrical F3 synform gently-moderately plunging to the
Ramsay's (1967) Type 2 of F2 × F3 fold interference pattern S (a parasite fold of the Curaçá antiform). It acquired the shape
(Jardim de Sá et al., 1982; D'el-Rey Silva, 1984) but could be of a N–S trending mushroom (Figs. 2 and 3a), due to the
better understood only after detailed mapping of the Caraíba interference of the F3 synform with several F2 folds, so that the
orebody and surrounding area (next section). F3 and S3 are also E–W trending and originally sub-horizontal F2 axis (B2) ap-
very common features in the Curaçá terrane. F3 folds are non- pears sub-vertical in the upper half of the orebody (Fig. 3b),
cylindrical, cm- to km-sized, generally tight, asymmetric, E- and the pre-D3 attitude of the layers also forces F3 fold axes to
verging, and their axial planes trend nearly N–S and dip 70°– plunge steeply to the N solely in the central and very inner part
75°W. The F3 axes (B3) plunge generally less than 30°, either in of the mushroom. Thus, the orebody consists, to the S of
a northerly or southerly direction. S3 is a typical mineral section 32 and to the N of section 37 (Fig. 3b), of a mineralized
foliation mostly defined by biotite, hornblende, and strongly layer affected by simple synform, whereas between sections 32
flattened quartz/feldspars, but the non-micaceous minerals to 37 (the central and richer part) it displays repetitions of the
display a prolate geometry and define a penetrative L3 mineral same layer due to several F2 hinges affected by the F3 synform
stretching lineation parallel to B3. Granites G3 are early- to late- (Fig. 3c and inset).
D3 (Fig. 2) and intruded sub-parallel to the axial trace of F3 The importance of the Airport outcrop and the Caraíba
folds. They most commonly display the S3 foliation, and orebody rests in the preservation of key evidence exactly where
represent typical lithotypes due to M3 metamorphism (Fig. 2). the I–S–C orogen was squeezed to its narrowest map
The Curaçá antiform (Delgado and Souza, 1975; Gáal, 1982) is expression (Fig. 1a-b). Similar D1-D3 structures and their field
a D3 structure, the hinge of which plunges to the N or to the S relationships are found further south, where the orogen is wider
(reference to the Surubim area) and is the site for most of the (in the Ipirá-Lajedinho area, Fig. 1a). In that area, largely
mafic–ultramafic bodies (Fig. 1b). Also in the Surubim area, protected by the northern margin of the Jequié block, abundant,
Jardim de Sá et al. (1982) and Hasuy et al. (1982) described m- to dm-scale, E–W trending F2 folds associated with
1 m- to 10 m-sized structures of Type 1 (domes and basins) mapable, ductile, D1-related lateral ramps, N–S trending D3
interference pattern between the regional F3 folds and gentle- transcurrent faults, as well as F2 × F3 interference patterns, some
open, very local F4 folds that have no influence in the geometry as large as the Caraíba mushroom, justifies interpreting the
and spatial distribution of the layers. orogen as the result of a long-lived, oblique collision-related
sinistral transpressional orogen (D'el-Rey Silva, 1993; Barbosa,
4. The Airport outcrop: key geological features preserved 1996), and the intensity of shortening peaks around Caraíba for
in an extremely deformed area two reasons: 1 — the area surrounding Caraíba lies 300 km
away from the Jequié block; and 2 — intrusion of the D2-related
The Airport outcrop is a flat-lying exposure of tonalitic syenite most likely added to make even stronger the E–W
orthogneiss with an amœboid surface area of nearly 1000 m2, compression responsible for the transposition of any pre-D3
and lies in the northern end of the land track of the Mine airport, structures that existed in the Caraíba region.
and is situated in the western limb and close to the hinge of the The tonalite in the Airport outcrop is yellowish-grey, well-
Curaçá antiform, in a similarly structural position as the banded, and composed of oligoclase-andesine, hornblende,
Caraíba orebody (Fig. 2). The orthogneiss also encloses tens of biotite, orthopyroxene, clinopyroxene, minor microcline and
boudins of amphibolite, the whole set intruded by sub-vertical, quartz, with garnet, magnetite, apatite, and zircon as acces-
sheet-like bodies of pink granites, and some quartz-feldspar sories. It also encloses cm- to dm-thick and N2 m-long bands of
veins. The outcrop itself displays important clues for acid composition, as well as b 1 m-sized lenses of non-
understanding the critical role of early E–W trending porphyritic granitoid, all aligned parallel to S2, this one a quite
structures, such as the F2 folds, in the evolution of the Curaçá homogeneous granulite-facies metamorphic banding consisting
terrane (next section), but a complete understanding of the of ≤ 1 cm-thick bands of flattened crystals of feldspar, quartz,
importance of the outcrop, and the importance of dating the minor biotite, intercalated with ≤ 1 cm-thick bands of pyroxene,
amphibolites (see Discussion) all demand an overview on the hornblende, plus biotite, and affected by numerous, 0.5–2.0 m-
geology of the Caraíba orebody, where such F2 folds were fully scale, very tight F3 folds that plunge ≤25° to the S and are
 L.J.H. D'el-Rey Silva et al. / Gondwana Research 12 (2007) 454–467
Fig. 3. a — Simplified geological map of the Caraíba orebody with indication of the S1 foliation, minor D3 structures, as well as the axial trace of the main F2 and F3 folds. The walls of each of the four upper benches of
the open pit were mapped (1978–1984) in a 1:1,000 scale, simultaneously with the opening of bench 395 m, the lower one at that time, so the orebody could be tracked in great detail as new exposures were available
every 10–20 m along strike, blast after blast. The geometry of the orebody match entirely the Cu-grade iso-contour map obtained on each bench by the plot of thousand data on the Cu-grade from drill holes of a 5 m-
square grid used for blasting. The eastern limb of the mushroom structure occurs only underground, from cross section 35 to the N, and has been projected on surface as a modification here introduced in the original map
by D'el-Rey Silva (1984, 1985); b — A 3-D schematic representation of the Caraíba mushroom (from D'el-Rey Silva et al., 1988); c — The overall pattern of F2 × F3 interference for cross section 36 (see also the inset) is
typical for sections 32–37 (the central part of the orebody). Diamond drill holes are indicated for different years. Adapted from D'el-Rey Silva et al. (1996) and D'el-Rey Silva and Oliveira (1999). See text for details.

459
460 L.J.H. D'el-Rey Silva et al. / Gondwana Research 12 (2007) 454–467

associated with a sub-vertical, N–S striking, and penetrative Only after the detailed mapping (Fig. 4), can the boudins be
mineral foliation (S3) marked mostly by biotite, but also by seen to define nearly E–W trending trails affected by F3 folds,
hornblende plus quartz and feldspar. S3 can be seen cross- revealing a more complete story: during D2 the G2 tonalite
cutting at very low angle the S1 banding in the boudins. intruded continuous layers of E–W trending S1-foliated

Fig. 4. Simplified geological map of the Caraíba Airport outcrop to show the site of collection of the samples used in this study. B3 fold axis plunges ≤25° to the S, whereas
planar structures such as S1 in the amphibolite boudins, S2, and S3, all dip sub-vertically, unless in the very hinge of the F3 folds. G3 granite intrusions cut across some of the F3
hinges. Based on the original lithostructural mapping at 1:100 scale carried with the aid of a plane table and alidade by L.J.H. D'el-Rey Silva and J.A.C. de Morais (D'el-Rey
Silva, 1984, 1985). Figures indicative of F2 × F3 interference pattern (some observed, other highly suggested) are indicated in the areas surrounded by ellipses. Details in text.
 L.J.H. D'el-Rey Silva et al. / Gondwana Research 12 (2007) 454–467 461

amphibolite. The amphibolite layers underwent boudinage and also underwent D3 shortening. The boudins systematically
while the tonalite deformed by ductile flow and acquired the nested in the F3 hinges, with S1 and the long axis both parallel to
granulite-facies S2 banding. During D3 the boudins rotated S3, plus the abundant F2 × F3 folding interference patterns in the
nearly 90° around a vertical axis, so their longer horizontal axis, tonalite, or highly suggested in the amphibolites (Fig. 4),
originally oriented nearly E–W, became systematically oriented altogether strongly support interpreting that: 1 — S2 formed sub-
N–S, the boudin ellipses entrapped in the F3 hinges (Fig. 5a, b, e) parallel to S1; 2 — the G2 tonalite is a syntectonic intrusion; 3 —

Fig. 5. Features of the Airport outcrop, as seen in four view to the S photos (a, b, c, e) and in two photos from thin-sections (d, f ). Several boudins (a) define a E–W
trending layer affected by F3 folds, in the central part of the outcrop. The longer axis of the boudins parallels the hammer’s handle that points to the S. Technician R.
Belo is partially shown on the upper left corner, beside the plane table and alidade used for detailed mapping; b — A 10 cm-thick acid band and the grey tonalite
affected by an F3 fold around the northern margin of the largest boudin (dotted line) in the outcrop. The traces of S1 and AP3 are parallel one to each other; c —
Relationships between the F3 folded tonalite and the eastern side of the boudin source of sample 1A. Note the b50 cm-wide felsic pegmatoid with over 1 cm-big
crystals of orthopyroxene along the amphibolite-tonalite contact; d — Detail of a thin section from sample collected at the boudin-tonalite contact (in c). Large crystals
of hornblende (Hb) dominate in the left half of the picture, and occur together with plagioclase and some crystals of orthopyroxene (Opx), whereas the latter dominates
in the right half, together with larger crystals of plagioclase. Opx crystals grew at the expenses of the Hb crystals along the margin of the boudin, as also demonstrated
by circular crystals of Hb included in Opx (D'el-Rey Silva, 1984, 1985); e — Detail of a F3 fold affecting S2 around the head of a boudin. Note the traces of S1 (inside
the boudin) and S3 both parallel to the boudin's long axis; and f — Picture of part of a thin-section beside a F3 hinge. Note crystals of pyroxene (some marked by ⁎) and
plagioclase along S2, as well as ribbons of biotite and flattened crystals of plagioclase defining the S3 foliation.
462 L.J.H. D'el-Rey Silva et al. / Gondwana Research 12 (2007) 454–467

Table 1
Summary of the U–Pb data from thirteen fractions of zircons from 10 kg of the amphibolite
U Pb Pb206 Pb207⁎ (pct) Pb206⁎ (pct) Correl. Pb207⁎ (pct) Pb206⁎ Pb207⁎ Pb207⁎ (My)
(ppm) (ppm) Pb204 U235 U238 coeff. Pb206⁎ U238 U235 Pb206⁎
(obs) (rho) Age (My) Age (My) Age (My)
145 76 1596 7.56206 0.57 0.37703 0.571 0.99019 0.145466 0.08 2062 2180 2293 1.4
196 71 862 6.36684 0.83 0.360684 0.791 0.94863 0.128025 0.263 1985 2027 2071 4.6
49 17 448 5.38241 1.24 0.311302 1.200 0.96828 0.125399 0.309 1747 1882 2034 5.5
205 32 297 1.40039 1.84 0.142056 1.650 0.90839 0.071497 0.769 856 889 971 16
201 74 815 6.32118 0.66 0.360535 0.636 0.966734 0.127160 0.168 1984 2021 2058 3
174 65 857 6.42752 0.68 0.363910 0.663 0.975212 0.128100 0.150 2000 2036 2072 2.6
546 202 604 5.99644 0.45 0.344294 0.431 0.959408 0.126317 0.128 1907 1075 2047 2.3
330 108 817 7.42198 1.04 0.287785 1.040 0.997556 0.1870470 0.729 1630 2163 2716 1.2
90 13 916 2.0111 1.02 0.145027 1.020 0.925758 0.1005740 0.383 873 1119 1634 7.2
577 116 7745 3.55528 0.5 0.201059 0.496 0.991762 0.1282480 0.063 1181 1539 2074 1.1
313 132 1116 7.57359 0.35 0.362173 0.349 0.994050 0.1516650 0.004 1992 2181 2364 0.65
206 100 582 8.57972 0.7 0.392694 0.633 0.906718 0.1584590 0.294 2135 2294 2439 5
309 84 383 2.9353 1.02 0.199033 0.994 0.977454 0.1069610 0.215 1170 1391 1748 4

D1-D3 deformation was progressive; and 4 — lithological The Sm–Nd isotopic analyses followed the method
layers, as well as S1 and S2 foliation, may all have had an E–W described by Gioia and Pimentel (2000): whole rock powders
trend somewhere in the Curaçá terrane, before D3 deformation. (ca. 50 mg) were mixed with 149Sm–150Nd spike solution
Thus, it is possible that the amphibolite layers where affected by and dissolved in Savillex capsules. Sm and Nd extraction of
E–W trending, 10–100 m sized, isoclinal F2 folding, in the whole-rock samples followed conventional cation exchange
hinges of which the G2 tonalite intruded. The double row of techniques, using teflon columns containing LN-Spec resin
boudins in the central part of the outcrop (Fig. 4) suggests a (HDEHP — di-ethylhexil phosphoric acid supported on PTFE
10 m-sized F2 × F3 boomerang-like Type 2 interference pattern. powder). Sm and Nd samples were loaded on Re evaporation
The strength of interpreting and the validity of the U–Pb filaments of double filament assemblies and the isotopic
diffusion technique age data (next section) rely in the validity of measurements were carried out on a multi-collector Finnigan
these structural and metamorphism relationships demonstrated MAT 262 mass spectrometer in static mode. Uncertainties for
above, and also in the fact that such ages agree with others Sm/Nd and 143Nd/144Nd ratios are better than ± 0.4 % (1σ) and
obtained for the same rocks, using more precise techniques, and ± 0.005% (1σ) respectively, based on repeated analyses of
are subject of a detailed discussion at the end of this paper. international rock standards BHVO-1 and BCR-1. 143Nd/144Nd
ratios were normalized to 146Nd/144 Nd of 0.7219 and the decay
5. Geochronology constant (λ) used was 6.54 × 10− 12.

5.1. Sample population and analytical procedures 5.2. U–Pb geochronology of zircons from amphibolite

The isotopic data derive from eleven samples of rocks The dated basic rock consists of orthopyroxene, hornblende,
collected in the Airport outcrop (Fig. 4), then treated and plagioclase, clinopyroxene, biotite, and quartz. Two populations
analyzed in the Laboratory of Geochronology — University of of zircons are distinguished, one comprising colourless,
Brasília. Zircon concentrates were extracted using conventional elongate, prismatic, and inclusion free crystals, the other
gravimetric (DENSITEST®) and magnetic (Frantz isodynamic consisting of clean, rounded, red to pink colour crystals.
separator) techniques applied to ca. 10 kg rock samples
collected in the largest boudin in the outcrop. Final purification
was achieved by handpicking using a binocular microscope. For
the conventional U–Pb analyses, fractions were dissolved in
concentrated HF and HNO3 (HF:HNO3 = 4:1) using micro-
capsules in Parr-type bombs. A mixed 205Pb–235U spike was
used. Chemical extraction followed standard anion exchange
technique, using Teflon micro columns, following procedures
modified from Krogh (1973). Pb and U were loaded together on
single Re filaments with H3PO4 and Si gel, and isotopic
analyses were carried out on a Finnigan MAT-262 multi-
collector mass spectrometer equipped with secondary electron
multiplier — ion counting. Procedure blanks for Pb, at the time
of the analyses, were better than 20 pg. PBDAT (Ludwig, 1993)
and ISOPLOT-Ex (Ludwig, 2001a) were used for data reduction Fig. 6. U–Pb ages defined by plotting the data in Table 1 and relative to zircons
and age calculation. Errors for isotopic ratios are 2σ. from amphibolite in the largest boudin in the Airport outcrop. Details in text.
 L.J.H. D'el-Rey Silva et al. / Gondwana Research 12 (2007) 454–467 463

Fig. 8. Plotting the data from eight samples of amphibolite in Table 2 in the Sm–
Nd diagram results in two isochronic ages. According to Frost and Frost (1995)
the causes for moving points out of true isochronic lines may be due to: 1 — Re-
Fig. 7. Detail concordia diagram for the three points in the upper part of the equilibrium at outcrop scale that would move the points along the vertical axis
diagram in Fig. 6, showing the accurate U–Pb age obtained for metamorphism (143Sm/144Nd ratio), then leading to a new isochronic straight line and affecting
M3 in the study outcrop. Details in text. the original age of the protolith; 2 — Variation in the igneous protolith at the
time of formation, a situation that would increase the 147Sm/143Nd ratio,
changing the inclination of the lines without re-equilibration of the system; and
Table 1 contains a summary of the U–Pb data from the 3 — Re-distribution of Sm relative to Nd during high-grade metamorphism,
most likely to be the case in the Airport outcrop, as the process would generate
analyses of 13 fractions of grains of zircons. The concordia parallel straight lines, but would not re-equilibrate the originally igneous data,
diagram for the analytical data shows considerable complexity therefore the ages will rather correspond to isochronic alignments.
in the U–Pb isotopic system, and the scattered plot of grains of
zircons indicates large discordance with the lower and upper
intercepts of the diagram (Fig. 6). 5.3. Sm–Nd geochronology in the Airport outcrop
We interpret the colourless zircons as representative of the
igneous protolith. Coincidently, their alignment provide the The systematic study of Nd isotopes was applied to ten
older U–Pb age of 2577 ± 110 Ma among other possible samples: eight of them collected in different boudins of
regressions. In despite the large error, this age is similar to the amphibolite (1A–H; Fig. 4), one sample of the G2 tonalite
most precise U–Pb SHRIMP age of 2580 ± 10 Ma obtained by (2A), and one sample of the G3 granite (3A), all listed in
Oliveira et al. (2004) for Caraíba norite, suggesting contempo- Table 2.
rary magmatic events. The population of red-pink zircons The isotopic data of G2 tonalite [sample 2A:TDM = 3101 Ma;
(interpreted as metamorphic crystals) plots along another εNd(t) = − 6.79] and G3 granite [sample 3A:TDM = 2719 Ma;
discordia and define an age of 2082 ± 4 Ma that we accept as εNd(t) = − 9.87] suggest derivation from partial melting of
a good result (Fig. 7) for the regional M3 metamorphism, Mesoarchean continental crust, likely to be the continental part
because it is very similar to crystallization ages of clearly of the basement of the Tanque Novo supracrustals. Plotted
syntectonic intrusions, such as G3 granites in the southern part together in a single 143Nd/144Nd versus 147Sm/144Nd iso-
of the Caraíba orebody (2051 ± 16 Ma age, U–Pb from chronic diagram (Fig. 8), the data from the eight samples of
monazite; D'el-Rey Silva et al., 1996), and the Itiúba Syenite amphibolites, appear to define two isochrones. The older age
(2084 ± 9 Ma U–Pb SHRIMP age; Oliveira et al., 2004). Points of 2603 ± 110 Ma, with εNd(t) values close to zero, is
probably related to Pb loss plot very near to the lower intercept interpreted as the age of crystallization of the igneous protolith
of the diagram, providing ages of 1.0 and 0.8 Ga that currently of the amphibolites. The younger age of 2022 ± 290 Ma and
we speculate to be geologically meaningless. the negative εNd(t) values are consistent with late isotopic re-

Table 2
Summary of the Sm–Nd data from the eight samples collected in the airport outcrop
Sample Rock Type Sm (ppm) Nd (ppm) 147
Sm/144Nd 143
Nd/144Nd ± 2σ ε(0) TDM (My) εNd(t)
1A Amphibolite 1.94 7.08 0.1658 0.512124 (12) − 10.02 2961 +0.14
1B Amphibolite 3.44 13.99 0.1487 0.511724 (09) − 17.83 3126 − 2.03
1C Amphibolite 2.82 8.76 0.1960 0.512619 (24) − 0.37 2884 − 0.14
1D Amphibolite 4.64 19.74 0.1393 0.511648 (17) − 19.31 2869 − 0.42
1E Amphibolite 5.10 20.19 0.1534 0.511816 (25) − 16.04 3104 − 1.79
1F Amphibolite 3.98 16.44 0.1466 0.511855 (32) − 15.28 2705 +1.21
1G Amphibolite 4.59 17.37 0.1606 0.511930 (22) − 13.81 2918 − 1.93
1H Amphibolite 2.06 7.73 0.1616 0.512015 (23) − 12.15 3048 − 0.60
2A Tonalite (G2) 7.24 35.39 0.1238 0.511241 (25) − 27.24 3101 − 6.79
3A Granite (G3) 1.79 19.84 0.0548 0.510202 (19) − 47.52 2719 − 9.87
464 L.J.H. D'el-Rey Silva et al. / Gondwana Research 12 (2007) 454–467

correspond to samples from the border of a boudin (1B); or from

small boudins (1E and 1G) with more fractioned Sm/Nd ratios
and the most negative εNd(t) values that reflect larger isotopic
interaction with the host rocks. Sample 1F (from the largest
boudin; Fig. 4) displays the best preserved original composition
and the highest positive value of εNd(t).

6. Discussion

The isotopic data suggest that the geochemical signature of

the amphibolites is compatible with basalts sourced from an
initially depleted mantle. These, combined with the strongly
similar ages of crystallization for Caraíba norites (Oliveira et al.,
2004) and airport amphibolite (this paper) leads to consider that
the protolith of the Caraíba orebody could have been either: a —
basaltic oceanic crust; b — basalts shed as flows within
sediments in the Tanque Novo sequence; or c — a gabbroic sill
intruded into the Tanque Novo sequence. All these possibilities
agree with the large inventory of data from the area surrounding
Caraíba, and from the Curaçá terrane as a whole. This inventory
Fig. 9. εNd(t) × Nd content Cartesian graph constructed with data from samples
1A–H in Table 2. requires that before D1 deformation the orebody's protolith was
a conformable body in the same sub-horizontal position of the
Tanque Novo volcanic-sedimentary pile, so that all lithotypes
equilibration of the Nd isotopic system and partial re- above came to experience the same D1-D3 structural/metamor-
equilibrium of Nd isotopic system during regional metamor- phic evolution.
phism. This process could have been theoretically related to The age of M1 metamorphism has not been determined.
contaminating fluids with low Sm/Nd ratio (McCulloch and However, once the field relationships in the Airport outcrop
Black, 1984), or loss of Sm relative to Nd during metamorphism. indicate a progressive evolution similar to the one observed in
The values obtained for the Nd isotopes of individual the Caraíba orebody, and the observation that the G2 intruded
samples of amphibolites indicate a heterogeneous isotope syntectonically, all support the interpretation that the structural
distribution reflected in varied TDM ages and variable values and metamorphic evolution of the Curaçá terrane may be
of εNd(t) (Fig. 9), and we found this related to the size of the bracketed between (2.35?)2.3–2.25 and 2.05 Ga (D'el-Rey
boudins. Samples with values of εNd(t) close to zero, or slightly Silva et al., 1996), with cooling lasting until 1900 Ma ago. The
positive (1A, 1C, 1D, and 1H) were all from large boudins new ages described in this paper suggest a more detailed
(Fig. 4) and are indicative of preservation of the igneous scenario: M2 granulite facies conditions peaked in the 2250–
protolith's isotopic composition with TDM model ages around 2200 Ma interval, whereas M3 amphibolite facies conditions
2.7–2.8 Ga, whereas TDM model ages around 2.9–3.1 Ga peaked at 2080–2050 Ma interval (Fig. 10).

Fig. 10. Summary diagrams for the tectonic evolution of the Curaçá belt. D1-D3 structures are observed since cm- to dm-scale up to the scale indicated below each
diagram. The leftmost diagram displays G1 conformable bodies in the Tanque Novo Complex, all affected by S1 and intrafolial F1 folds. M1 metamorphism probably
peaked in the 2.3–2.8 Gy interval. The 2.58 Gy old body (Oliveira et al., 2004) destined to be the Caraíba orebody is depicted in the center, with no implication on the
origin of the protolith. The central diagram displays S2 and asymmetric F2 folds affecting previous units and structures, as well as syntectonic G2 intrusions, all associated
to M2 metamorphism (T ≤ 1000 °C; Leite, 2002) peaked in the 2.25–2.20 Gy interval (this paper). The diagram to the right shows D1–D2 units and structures affected by
F3 folds and S3, and intruded by G3 bodies, all affected by M3 metamorphism peaked 2.08–2.05 Gy ago (our result combined with data in D'el-Rey Silva et al., 1996;
Oliveira et al., 2004). Final cooling lasted until ∼1.9 Gy ago. The Caraíba mushroom is to the left of the Curaçá antiform (Ca). Temperatures around 600 °C for M1 and
M3 fit data in Leite (2002) for amphibolite facies metamorphism in his PhD study area.
 L.J.H. D'el-Rey Silva et al. / Gondwana Research 12 (2007) 454–467 465

Three similarly important facts make the Curaçá belt a Combining our results with those of previous workers, we
classical area for tectonic analysis based on the combination of propose the following scenario for the tectonic evolution of the
accurate field geology, detailed structural observations, and region. The protoliths to the Airport amphibolites and the
careful isotope analysis of different systems. Tanque Novo sequence, both formed ∼2.6 Ga, when a wide
Firstly, the large errors associated with the ages in here ocean separated the Gavião and Serrinha blocks. The Caraíba
indicate open-system behaviour of the U–Pb and Sm–Nd Complex was evolving in this ocean as an island arc above an
isotopic systems in the amphibolite protolith. The difficulty of east-dipping subduction zone that plunged below the Serrinha
dating basic rocks in a high grade terrane resides in that such Block (Leite, 2002). M1 amphibolite facies metamorphism (our
rocks contain extremely small volumes of minerals suitable for results) peaked sometime between 2.35–2.80 Ga, as the rocks
high precision age determinations, and their isotopic systems entered in the subduction zone, then underwent layer-parallel
are commonly affected at varied intensities during severe shearing, and acquired D1 structures. D2 Structures developed
metamorphism, so the study of these systems will indicate as the rocks were buried further, and the E–W fold axes (F2)
varied levels of complexity, no matter if the tectonic evolution formed by differential flow affecting fold axes originally
was polyphase or polycyclic (Frost and Frost, 1995; Gruau trending N–S, a kind of embayment process pointing-down to
et al., 1996; Pidgeon and Wilde, 1998; Carson et al., 2002). the east, inside the N–S trending subduction zone and coeval
Secondly, the older cratonic blocks in the São Francisco with granulite facies metamorphism (M2). This metamorphism
Craton preserve different isotopic ratios (Fig. 1a) acquired condition peaked at 2.25–2.2 Ga, our result that on one hand fits
before the Paleoproterozoic. The I–S–C orogen may involve in the (2.2–2.1 Ga) interval suggested by Leite (2002) for
multiple collisions along the margins of the Gavião and Serrinha amalgamation of the island arc to the Serrinha Block, and on the
Mesoarchean blocks during the Neoarchean and Paleoproter- other hand, also suggests a back-arc scenario for part of the
ozoic collisions, and unravelling these presents a considerable Tanque Novo sequence as well as for the amphibolites and
challenge. the Cu-mineralized orebodies (see also Bello, 1986). Closure of
In the area south of the Itiúba syenite, Teixeira (1997) the ocean at ∼ 2.1 Ga resulted in the formation of a foreland
mapped the Caraíba and Tanque Novo (or Tanque Novo-Ipirá) basin to the west (all in Leite, 2002). The roots of the orogen
complexes folded/tectonically imbricate together with the São started to be uplifted as the subduction zone was choked when
José do Jacuípe Suite, a unit consisting of norites, gabbro- the Gavião Block began to enter into the subduction zone, and
norites, gabbros, peridotites, and pyroxenites, and suggested the ultimate Gavião – Serrinha continental collision at 2.08 Ga
that part of the Caraíba Complex defines a magmatic arc, (Leite, 2002) caused D3 deformation under M3 amphibolite
subsequently dated as Neoarchaean (quoted in Delgado et al., facies metamorphic conditions peaked at 2.08–2.5 Ga through-
2003), whereas part of the two other units matches the out the I–S–C orogen.
characteristics of oceanic crust. The interpretation here that
the Airport outcrop may preserve 2.6 Ga oceanic crust fits this 7. Conclusions
model well.
The third fact is that the long-lived and transpression- This paper has brought to light new U–Pb and Sm–Nd
related tectonics recorded in the I–S–C orogen has a great geochronology data obtained after detailed studies carried out in
potential to transform a single suture into a stack of sub- amphibolites (mainly), G2 tonalite, and G3 granite of the Airport
vertical suture segments, and the traces of these segments may outcrop, situated ∼ 5 km to the north of the Caraíba Cu-orebody,
spread over a wider field area than the original suture zone. In in the Curaçá terrane, northern São Francisco Craton, Brazil.
case of polycyclic evolution, this tectonic scenario could have The new data, coupled with well-known tectonic structures and
stacked segments of two collision sutures, creating a true other field data from the Airport outcrop, clearly imply D1-D3
geological puzzle. progressive deformational events under amphibolite-granulite-
We are aware that other regressions are possible with our U– amphibolite M1-M3 metamorphism, and G1-G3 magmatism.
Pb data. However, the U–Pb age of 2.6 Gy for the amphibolite The G2 tonalite and G3 granite in the Airport outcrop derive
protolith is sustained here because: a — it is the highest value (Sm–Nd data) from partial melting of a Mesoarchean
obtained; b — it derives from the population of clean zircons; continental crust existing in the Curaçá terrane. This Mesoarch-
c — it is within error of the Sm–Nd isochronic age we obtained ean crust is likely to be part of the basement of the Tanque Novo
for the same amphibolites; and d — both ages above are similar supracrustals. The igneous protolith of the amphibolites in the
to the SHRIMP age obtained for the Caraíba norite (Oliveira et outcrop is Neoarchean and crystallized ca. 2600 Ma ago (U–Pb
al., 2004). Metamorphic zircons in the amphibolites yield an age and Sm–Nd data). This may be also in the age-interval of
between 2.3 to 2.2 Ga. We suggest that the 2.3–2.2 Ga interval sedimentation in the northern part of the basin precursor of the
is the time in which the isotopic system in the amphibolites was Itabuna-Salvador-Curaçá orogen. Granulite facies metamor-
disturbed by incoming fluids derived from the G2 tonalite, and phism (M2) and amphibolite facies metamorphism (M3) in the
sustain the interpretation based on two facts: 1 — the G2 Curaçá terrane are respectively dated around 2250–2200 Ma,
tonalite is a syn-tectonic intrusion; and, 2 — pristine zircons and 2084–2050 Ma (U–Pb data). The progressive evolution
from the G2 tonalite in the same Airport outcrop provided an supports previous interpretation that M1 amphibolite facies
age of 2248 ± 36 Ma (D'el-Rey Silva et al., 1996), emphasizing metamorphism (still lacking a direct age determination) is as old
the significance of this early Paleoproterozoic D2 event. as 2300 Ma.
466 L.J.H. D'el-Rey Silva et al. / Gondwana Research 12 (2007) 454–467

However, the progressive evolution recorded in the Curaçá D'el-Rey Silva, L.J.H., Oliveira, J.G., 1999. Geology of the Caraíba Copper
terrane as a whole does not rule out the possibility that other parts Mine and its surroundings in the Paleoproterozoic Curaçá Belt- Curaçá River
Valley, Bahia, Brazil. In: da Silva, M. da G., Misi, A. (Coordenadores) Base
of the Curaçá terrane, or elsewhere in the I–S–C orogen, still metal Deposits of Brazil, MME/CPRM/DNPM, 25-32.
hide evidence for a polycyclic evolution. The results in here D'el-Rey Silva, L.J.H., Cavalcante, P.R.B., Mota, E.R., Rocha da, A.M.R.,
emphasize the importance of combining geochronology, 1988. Controle estrutural da mina de cobre Caraiba: implicações na lavra e
lithostructural and metamorphic detailed studies of key outcrops na tectônica das faixas móveis do Proterozóico Inferior. XXXV Congresso
Brasileiro de Geologia, Belém, Brazil, Anais SBG, vol. 1, pp. 16–29.
for unravelling the evolution of high-grade terrains, and
D'el-Rey Silva, L.J.H., Oliveira, J.G., Lima e Silva, F.J., 1994. The Mushroom-
constitute a contribution for the geochronology approach of shaped Caraíba Cu-deposit, Vale do Curaçá-Ba: Understanding the Structural
high-grade terranes in general. Evolution of the Paleoproterozoic, Granulitic Ipirá-Curaçá Belt within the São
Francisco Craton. XXXVIII Congresso Brasileiro de Geologia, Camboriú,
Acknowledgements Brazil, SBG, Boletim de Resumos Expandidos, vol. 2, pp. 175–177.
D'el-Rey Silva, L.J.H., Oliveira, J.G., Gaál, E.G., 1996. Implication of the
Caraíba Deposit's structural controls on the emplacement of the Cu-bearing
Dr. Sérgio L. Junges is thanked for support during isotopic hypersthenites of the Curaçá Valley, Bahia-Brazil. Revista Brasileira de
analyses at the Geochronology Lab — UnB. Elton L. Dantas Geociências 26 (3), 181–196.
thanks CNPq for grant 471144/03-7. The first author dedicates Delgado, I. de M., Souza, J.D., 1975. Projeto cobre Curaçá. Geologia
his part in this work to his wife and sons, and to the memory of Econômica do Distrito Cuprífero do Rio Curaçá — Bahia, Brasil. CPRM/
geologist José Genário de Oliveira (deceased in 2000), an DNPM, internal report, 30 volumes.
Delgado, I. de M., de Souza, J.D., da Silva, L.C., Silveira Filho, N.C., dos
enthusiastic of the Caraíba Project, who largely facilitated the Santos, R.A., Pedreira, A.J., Guimarães, J.T., Angelim, L.A. de A.,
visit to the mine and collection of the samples in the Airport Vasconcelos, A.M., Gomes, I.P., de Lacerda Filho, J.V., Valente, C.R.,
outcrop. Two anonymous reviewers are thanked for excellent Perrotta, M.M., Heineck, C.A., 2003. Geotectonics of the Atlantic Shield.
comments that helped to make better this piece of work. Dr. Geology, Tectonics and Mineral Resources of Brazil: Text, Maps, and GIS,
Alan Collins (The University of Adelaide, Australia) is deeply CPRM, pp. 227–334. 692 p., Chapter V.
Figueiredo, M.C.H., 1981. Geoquímica da rochas metamórficas de alto grau do
acknowledged for his superb help on final editing. This paper is nordeste da Bahia-Brazil. Geologia e Recursos Naturais do Estado da Bahia,
a contribution to IGCP project 509, Palaeoproterozoic Super- SME, Série Textos Básicos 4, 1–71.
continents and Global Evolution. Frost, C.D, Frost, B.R., 1995. Open-system dehydration of amphibolites,
Morton Pass, Wyoming: elemental and Nd and Sr isotopic effects. The
References Journal of Geology 103, 269–284.
Gáal, E.G., 1982. Evaluation of geological, geophysical, and geochemical data
Ackermand, D., Herd, R.K., Reinhardt, M.C., Windley, B.F., 1987. Sapphirine of the Curaçá Valley, and the Mundo Novo-Uauá-Rio Capim area.
parageneses from the Caraíba Complex, Bahia, Brazil: the influence of Fe2+– Consulting report for the Mineração Caraíba Ltda, MID-066/82.
Fe3+ distribution on the stability of Sapphirine in natural assemblages. Gioia, S.M.L., Pimentel, M.M., 2000. The Sm–Nd method in the geochronol-
Journal of Metamorphic Geology 5, 323–339. ogy laboratory of the University of Brasilia. Anais da Academia Brasileira
Barbosa, J.S.F., 1996. O embasamento Arqueano e Proterozóico Inferior de de Ciencias 72, 219–245.
Estado da Bahia. In: Barbosa, J.S.F., Dominguez, J.M.L. (Eds.), Geologia da Gromet, L.D., 1991. Direct dating of deformational fabrics. In: Heaman, L.,
Bahia, Texto Explicativo SME-SGM, Capítulo III, pp. 63–83. Ludden, J.N. (Eds.), Short Course Handbook on Applications of Radiogenic
Barbosa, J.S.F., Sabaté, P., Dominguez, J.M.L., 1996. O estado da Bahia na Isotope Systems to Problems in Geology. Mineralogists Association of
Plataforma Sul-Americana, suas sub-divisões, critérios de empilhamento Canada, pp. 167–192. Chapter 5.
estratigráfico das coberturas plataformais e ciclos geotectônicos. In: Gruau, G., Rosing, M., Bridgwater, D., Gill, R.C.O., 1996. Resetting of Sm–Nd
Barbosa, J.S.F., Dominguez, J.M.L. (Eds.), Geologia da Bahia, Texto systematics during metamorphism of N3.7 Gy rocks: implications for isotopic
Explicativo SME-SGM, Capítulo II, pp. 41–61. models of early Earth differentiation. Chemical Geology 133, 225–240.
Barbosa, J.S.F., Sabaté, P., 2002. Geological features and the Paleoproterozoic Hasuy, Y., D'el-Rey Silva, L.J.H., Lima e Silva, F.J., Mandetta, P., de Moraes, J.A.,
collision of four Archean crustal segments of the São Francisco Craton, Oliveira, J.G., Miola, W., 1982. Geology and copper mineralization of Curaçá
Bahia, Brazil. Anais da Academia Brasileira de Ciencias 74 (2), 343–359. River Valley, Bahia. Revista Brasileira de Geociências 12 (1-3), 463–474.
Barbosa, J.S.F., Sabaté, P., 2004. Archean and Paleoproterozoic crust of the São Inda, H.A.V., Barbosa, J.S.F., 1978. Mapa geológico do Estado da Bahia, escala
Francisco Craton, Bahia, Brazil: Geodynamic Features. Precambrian 1:1.000.000. Texto explicativo. SME/CPM.
Research 133, 1–27. Jardim de Sá, E.F., Arcanjo, C.J., Legrand, J.-M., 1982. Structural and
Bello, R.M.S., 1986. Jazida de Cobre de Surubim, Vale do Curacá, Bahia: metamorphic history of part of the high-grade terrane in the Curaçá Valley,
Mineralogia, Petrografia e Petrogênese. Ph.D. thesis, University of São Bahia, Brazil. Revista Brasileira de Geociências 12, 251–262.
Paulo (USP). Krogh, T.E., 1973. A low-contamination method for hydrothermal decompo-
Carson, J.C., Ague, J.J.A, Grove, M., Coath, D.C., Harrison, M.T., 2002. U–Pb sition of zircon and extraction of U and Pb for isotopic age determinations.
isotopic behaviour of zircon during upper-amphibolite facies fluid Geochimica et Cosmochimica Acta 37, 485–494.
infiltration in the Napier Complex, east Antarctica. Earth and Planetary Leite, C.M.M., 2002. A Evolução Geodinâmica da Orogênese Paleoproter-
Science Letters 199, 287–310. ozóica nas Regiões de Capim Grosso-Jacobina e Pintadas-Mundo Novo
Condie, K.C., 1989. Plate Tectonics and Crustal Evolution, Third ed. Pergamon (Bahia-Brasil): Metamorfismo, Anatexia Crustal e Tectônica. Ph.D. thesis,
Press, Oxford-UK. Federal University of Bahia (UFBA).
D'el-Rey Silva, L.J.H., 1984. Geologia e controle estrutural do depósito Leite, C.M.M., Barbosa, J.S.F, Nicollet, C., Kienast, J.R., Fuck, R.A., 2005.
cuprífero Caraíba, Vale do Curaçá, Bahia, Brasil. M.Sc. thesis, Federal Ultrahigh-temperature metamorphism of sapphirine-bearing granulite from
University of Bahia (UFBa). the 2.0 Gy Itabuna-Salvador-Curaçá orogen, Bahia, Brazil. III Sympósio
D'el-Rey Silva, L.J.H., 1985. Geologia e controle estrutural do depósito sobre o Cráton do São Francisco, Salvador, Brasil, AnaisISSN :1808-6047,
cuprífero Caraíba, Vale do Curaçá, Bahia. Geologia e Recursos Naturais do pp. 83–86.
Estado da Bahia, SME, Série Textos Básicos 6, 51–123. Lindenmayer, Z.G., 1981. Evolução geológica do Vale do Curaçá e dos corpos
D'el-Rey Silva, L.J.H., 1993. Geologia Estrutural e Tectônica na área do Projeto máfico-ultramáficos mineralizados a cobre. M.Sc. thesis, Federal University
Lajedinho-Ipirá. Consulting report for CBPM, Salvador, 23 pp., 23 figures. of Bahia (UFBa).
 L.J.H. D'el-Rey Silva et al. / Gondwana Research 12 (2007) 454–467 467

Lindenmayer, Z.G. and six co-authors, 1984. Projeto Cu Bahia, Alvará 872, significance for orogen-parallel escape tectonics. Precambrian Research
DNPM 812.998/73. Relatório Final de Pesquisa, DOCEGEO-CVRD, 128, 143–165.
3 volumes. Passchier, C.W., Myers, J.S., Kröner, A., 1990. Field Geology of High-Grade
Ludwig, K.R., 1993. PBDAT. A computer program for processing Pb–U–Th Gneiss Terranes. Springer-Verlag, Berlin.
isotope data. USGS Open File Report, vol. 88-542. 34 pp. Pidgeon, R.T., Wilde, S.A., 1998. The interpretation o complex zircon U–Pb
Ludwig, K.R., 2001. Users Manual for Isoplot/Ex version 2.47. A geochrono- systems in Archean granitoids and gneisses from the Jack Hills, Narryer
logical toolkit for Microsoft Excel. Berkeley Geochronology Centre Special Gneiss Terrane, Western Australia. Precambrian Research 91, 309–322.
Publication, vol. 1a. 55 pp. Ramsay, J.G., 1967. Folding and Fracturing of Rocks. McGraw Hill, New York.
Mandetta, P., 1982. Aspectos geológicos e petrogenéticos das associações Sabaté, P., Marinho, M.M., Vidal, P., Caen-Vachette, M., 1990. The 2-Ga
máfico-ultramáficas da região de Caraíba, Vale do Curacá/Ba. MsC thesis, peraluminous magmatism of the Jacobina-Contendas Mirante belts (Bahia,
Federal University of Bahia (UFBa). Brazil): geologica and isotopic constraints on the sources. Chemical
Mayer, W.D., Barnes, S.-J., 1996. Unusually high concentrations of magnetite at Geology 83, 325–338.
Caraíba and other Cu-sulphide deposits in the Curaçá Valley, Bahia, Brazil. da Silva, M. da G., 1996. Sequências metasedimentares, vulcanosedimentares e
The Canadian Mineralogist 34, 717–731. greenstone belts do Arqueano e Proterozóico Inferior. In: Barbosa, J.S.F.,
McCulloch, M.T., Black, L.P., 1984. Sm-Nd isotopic systematics of Enderby Dominguez, J.M.L. (Eds.), Geologia da Bahia, Texto Explicativo SME-
Land granulites and evidence for the redistribution of Sm and Nd during SGM, Capítulo IV, pp. 86–102.
metamorphism. Earth and Planetary Science Letters 71, 46–58. Teixeira, L., 1997. O complexo Caraíba e a suíte São José do Jacuípe no cinturão
Oliveira, E.P., Windley, B.F., McNaughton, N.J., Pimentel, M.M., Fletcher, I.R., Salvador-Curaçá (Bahia, Brasil): petrologia, geoquímica e potencial
2004. Contrasting copper and chromium metallogenic evolution of terranes metalogenático. Ph.D. Thesis, Federal University of Bahia (UFBa).
in the Paleoproterozoic Itabuna-Salvador-Curaçá orogen, São Francisco Windley, B.F., 1986. The Evolving Continents, Second ed. John Wiley and Sons
craton, Brazil: new zircon (SHRIMP) and Sm–Nd (model) ages and their Ltd, UK.