You are on page 1of 16

Chemical Geology 289 (2011) 124139

Contents lists available at ScienceDirect

Chemical Geology
j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / c h e m g e o

Iron and sulphur isotopes from the Carajs mining province (Par, Brazil):
Implications for the oxidation of the ocean and the atmosphere across the
ArchaeanProterozoic transition
Sbastien Fabre a,, Anne Ndlec a, Franck Poitrasson a, Harald Strauss b,
Christophe Thomazo b, c, Afonso Nogueira d
a

Gosciences Environnement Toulouse (GET), UMR 5563, Universit de Toulouse-CNRS-IRD, 14-16, avenue Edouard Belin, 31400 Toulouse, France
Institut fr Geologie und Palontologie, Westflische Wilhelms-Universitt Mnster, Corrensstrasse 24, 48149 Mnster, Germany
c
Universit Denis Diderot, Institut de Physique du Globe de Paris, 75251 Paris Cedex 05, France
d
Universidad federal do Par, Belem, Brazil
b

a r t i c l e

i n f o

Article history:
Received 22 July 2010
Received in revised form 19 July 2011
Accepted 24 July 2011
Available online 30 July 2011
Editor: J.D. Blum
Keywords:
BIF
Pyrite
Carajs
Iron and sulphur isotopes
Archaean
Palaeoproterozoic

a b s t r a c t
Mineralogical analyses and major and trace element geochemistry combined with iron and sulfur isotopes
were applied to banded iron formations (BIFs) and pyrite-bearing sandstones from the Carajs mineral
province, Brazil. The BIFs belong to the 2.7 Ga old Gro Par Group and show a rather simple mineralogy
(hematite and quartz or cryptocrystalline silica). Their REE + Y patterns suggest deposition under marine suboxic conditions. Fe isotopes show highly positive values (57FeIRMM-14 = 2.06 0.34 in average) which can
be explained by the partial oxidation of Fe 2+aq in the surface ocean, subsequent to the upwelling of deep
anoxic seawater. The resulting iron oxy-hydroxides likely formed in the shallow photic zone as a seasonal
process and acquired highly positive iron isotope signature values consistent with a Rayleigh distillation
model. The pyrite-bearing sandstones belong to the Aguas Claras Formation of (Palaeoproterozoic) younger
age and were cored at the base of the giant Azul manganese deposit. Diagenetic pyrite is characterized by
positive 57FeIRMM-14 = 1.02 0.2 and 34SVCTD (11.97 0.12) average values, that are different from
values of modern marine biogenic pyrite. Results likely indicate that dissolved sulfate was the limiting factor
controlling biogenic pyrite formation at that time. Near-zero 33S values (0.013 0.003) indicate minimal
mass-independent sulfur isotope fractionation (MIF-S), consistent with the 2.1 Ga age previously obtained
from the Aguas Claras pyrites. This result suggests deposition at a time when the oxygen partial pressure in
the atmosphere was already high enough (10 5 to 10 2 PAL- or Present Atmospheric Level-) to hinder MIF-S
by photochemical reactions. Incidentally, this also conrms that the giant Mn deposit of Carajs is in the same
age range (2.1 0.2 Ga) as other Mn ores in Gabon and South Africa, pointing to the importance of Mn
precipitation in Paleoproterozoic times.
2011 Elsevier B.V. All rights reserved.

1. Introduction
It is now broadly accepted that there were two major increases in
the atmospheric oxygen abundance; the rst rise occurred during the
so called Great Oxidation Event (GOE) between 2.4 and 2.2 Ga and the
second rise between 0.8 and 0.5 Ga (e.g., Kasting, 1993, 2001;
Caneld, 2005; Holland, 2006). Evidence supporting the existence of
the GOE includes (1) the termination of large mass independent
fractionation of sulfur isotopes in sedimentary suldes and sulfates
(Farquhar et al., 2000), (2) a change in the composition of palaeosols,
Rye and Holland (1998) particularly and others works like Kirk et al.,
2001 and Bekker et al., 2004, show that 2.44 Ga old palaeosols display

Corresponding author.
E-mail address: fabre@lmtg.obs-mip.fr (S. Fabre).
0009-2541/$ see front matter 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.chemgeo.2011.07.019

a loss of iron, nevermore observed thereafter (excepted in podzols)


and (3) the disappearance of detrital redox sensitive minerals: e.g.
uraninite (Ono et al., 2000) and pyrite (Frimmel, 2005). In fact, the
starting point of the GOE is still under debate because small amounts
of oceanic O2 were probably present in the environment more than
50 Ma before the disappearance of MIF-S as shown by Anbar et al.
(2007). Moreover, the Archaean ocean displays a redox variability
imposed by oxidative weathering of the continents which also puzzles
the Earth's oxygenation record (Reinhard et al., 2009).
BIFs and pyrites are likely to be among the best proxies to
reconstruct the pO2 evolution. Banded iron formations are chemical
sedimentary rocks, comprised of alternating layers of silica and ironoxy-hydroxides. Iron-formation including granular and banded
types temporal extension are mainly found between 3.2 and 1.8 Ga.
The peak of BIF formation was circa 2.5 Ga ago (Holland, 2005) but
there was many other depositional periods, likely reecting changes

S. Fabre et al. / Chemical Geology 289 (2011) 124139

125

Here, we present geochemical, Fe and S isotope data from banded


iron formation and pyrite-bearing sandstones from the Carajas mining
province, Par, Brazil. Together with their depositional conditions and
age, these new data highlight the evolution of iron and sulfur cycling in
near-surface environments in relation with the progressive oxidation of
the ocean and the atmosphere at the ArchaeanPalaeoproterozoic
transition.

of global magmatism and/or environmental changes (see the review


by Bekker et al., 2010). Traditionally, BIFs are viewed to have formed
due to Fe 2+aq oxidation derived from three possible mechanisms:
(1) oxygenic microbial photosynthesis in the upper layer of a
stratied oceanic water column with a deep anoxic layer (Cloud,
1968; Mel'Nik, 1982; Beukes and Klein, 1992; Beukes et al., 2003; Mc
Donald et al., 2005); (2) anaerobic photosynthetic Fe(II)-bacterial
oxidation (Widdel, 1993) or (3) UV photo-oxidation (Braterman et
al., 1983; Braterman and Cairnsmith, 1987). Whatever the case, BIFs
contain the ngerprints of the deep ocean anoxia.
Indeed, the level of O2 is affected by both geological and biological
processes. On one hand, O2 is produced by net photosynthesis, after the
burial of some organic matter and pyrite in sediments. On the other
hand, O2 is consumed by the oxidation of old buried sedimentary
organic matter and FeS2 after uplift and exposure to weathering on the
continents (Berner and Berner, 1987; Berner and Caneld, 1989;
Habicht and Caneld, 1996; Berner, 2001). Consequently, sedimentary
pyrite inuences the atmospheric oxygen level. The key process in
pyrite formation is the bacterial reduction of seawater sulfate (BSR) and
subsequent precipitation of iron sulde consuming reactive Fe of
continental origin (for a review, see Strauss, 1999, and Caneld, 2001).

2. Geological setting
The Carajs mineral province, Par state (Brazil) is the second largest
in the world in respect to iron, gold, manganese and copper deposits
(Grainger et al., 2008). It is situated in the southeastern part of the
Amazon Craton (Fig. 1). The study area is composed of an Archaean
crystalline basement, comprising ca 3.0 Ga old granites and gneisses of
the Xing, Pium and Plaqu complexes, and a volcano-sedimentary
succession, named the Itacainas Supergroup in the Carajs Basin,
whose depositional conditions and history are still being discussed. The
lithological units and their stratigraphic relationships are presented in
Fig. 1 after Pinheiro and Holdsworth (1997).

50OW

CINZE

NTO S

CARA

TRIKE

JS FA
ULTT

SYST

EM

CURURU

POJUCA
O
6S
6
S

-SLIP

6 OS

N-1
N-4

N-5

PARAUAPEBA S

AZUL

J
RA
CA

CAR

S FA
UL T

SYSTEM

SLIP

Carribbean
Sea
Northern
Amazon
Craton

Amazonas
Basin

50 W

BIF

s
de

An

Southern
Amazon Craton

Gro Par Group (2.75 Ga)


Igarap Pojuca Group

Granitoids (1.88 Ga)


guas Claras Formation
(debated age)

Estrela granitic Complex (2.5 Ga)

Phanerozoic

Archaean
cover

Igarap Salobo Group


Xingu Complex (2.86 Ga)

Gorotire Formation

MnF

Rancho Alegre
Granite

20 km

Carajs Mineral
Province (CMP)

Serra Grande Group

CURIONPOLIS

ESTRELA
GNEISS

AJ

CENTRAL
CARAJS
GRANITE

ST
RIK
E-

SERRA PELADA

CIGANO
GRANITE

BAHIA

500 km

SERRA DO
PA R E D O

A R AG UA I A B E LT

SALOBO

iv
er

Par
au
ap
R
eb
iv
as
er

51OW

Proterozoic

Pium Complex
Plaqu Suite

Fig. 1. Detailed geological map (a), geological overview (b), and stratigraphy (c) of the Carajs mining province.

Archaean
basement

126

S. Fabre et al. / Chemical Geology 289 (2011) 124139

The BIFs studied here form part of the late Archaean Gro Par Group,
that was dated at ca. 2.75 Ga through UPb zircon dating of associated
and crosscutting mac rocks (Trendall et al., 1998). The rocks underwent no more than low grade greenschist facies metamorphism and
only some low temperature brittle deformation.
The Aguas Claras Formation overlies the Gro Par Group. These
rocks are younger than the BIFs and probably overlie them unconformably although this contact has never been observed (Tolbert et al., 1971).
They represent a supracrustal sequence constituted by clastic sediments
(pelites, arenites, conglomerates and locally breccias) with a sericitic
matrix reecting anchimetamorphic conditions. Noguiera et al. (1994)
proposed a westward prograding sedimentary succession of pelites,
quartzites and conglomerates, corresponding to either an epicontinental shelf or a deltaic uvial domain. Deformation and metamorphism are
faint or absent. Age constraints for the Aguas Claras Formation are
somewhat controversial. A DOCEGEO (Rio Doce Geologia e Minerao
S.A.) report (1988) suggested that the Aguas Claras Formation was
much younger than the Gro Par Group. Moreover, the Aguas Claras
Formation contains giant manganese deposits (e.g. Azul mine, Fig. 1b),
regarded worldwide as typical Palaeoproterozoic ores. Mougeot et al.
(1996) performed PbPb dating on disseminated suldes from Aguas
Claras sandstones, yielding an age of 2.06 Ga. Later, Trendall et al. (1998)
acquired SHRIMP U/Pb dates for detrital zircons from these rocks. A
zircon age of 2681 5 Ma was assumed to represent syndepositional
volcanism, because of the nearly unweathered zircon appearance.
Consequently, the retained age for the Aguas Claras formation is
supposed to be Archaean in the recent review of Grainger et al. (2008).
Nevertheless, we prefer the 2.1 Ga age of Mougeot et al. (1996), as it was
obtained from syndiagenetic minerals and because we interpret the lack
of weathering in detrital zircons as the results of the short time lag
between protolith exhumation due to the Transamazonian orogeny
(2.22.0 Ga) and sediment deposition.
The Itacaiunas belt experienced oblique compressive deformation
due to roughly EW sinistral strikeslip shear zones at the end of
Palaeoproterozoic times (Arajo et al., 1988; Arajo and Maia, 1991;
Arajo et al., 1992). Arajo and Maia (1991) recognized two symmetrical
structural systems (Fig. 1): the Cinzento (northward) system and the
Carajs (southward) system. A-type granites dated at 1.88 Ga sealed this
deformation and are responsible for the development of some
hydrothermal AuCu mineralization (Grainger et al., 2002).
3. Material and methods
3.1. Sampling and mineral identication
The BIF samples were collected in the N4 iron mine (Fig. 1) at three
heights: 610, 625 and 640 m. The pyrite-bearing sandstones of the
Aguas Claras Formation are pieces from drill cores that were drilled at
the oor of the Azul manganese mine. Samples correspond to the
followings depths: 90.6, 92.2 and 92.25 m with respect to the
oor of the mine.
Petrographic observations and mineral identication were performed
using thin and polished sections examined under the optical microscope
and SEM (including EDS spectra). In addition, the ratio of silica vs iron
bands for the BIFs was determined by image analysis (Table 1).
Representative rock fragments from the BIFs were used for whole rock
analyses of major and trace elements. In addition, milligram-size samples

Table 1
Percentage of silica- and hematite-rich mesobands in BIF samples.
Samples

Silica-rich band (%)

Hematite-rich band (%)

EN 610
EN 625
EN 640

52.5
71.4
31.8

47.5
28.6
68.2

from selected meso- or microbands were taken using a tungstencarbide


bit (1 mm diameter) from fresh rock surface previously cleaned with
distilled water. Powdered BIF samples were subjected to X-ray diffraction
analyses and mineral quantication was achieved by using the PowderCell
software (Table 2).
3.2. Major and trace element geochemistry
Major and trace element contents of BIF whole-rock samples were
analyzed by SARM (CRPG, Nancy, France) (Table 3). Composition of the
microsamples from the three samples (610, 625 and 640) were
determined in Toulouse. Silica and total iron concentrations of these
microsamples were measured via ICP-OES (at ENSAT, Toulouse) after
alkaline fusion of the powdered samples with lithium tetraborate. The
analytical precision (b5%) was monitored with the geostandard IF-G.
Trace elements analyses were performed with an Agilent technologies
ICP-MS quadrupole 7500ce in the GET Laboratory Toulouse. Again,
analytical precision (also b5%) and accuracy of REE + Y abundances
were monitored using the IF-G geostandard. Correlation between
recommended and measured values provided an r 2 of 0.998,
except for Lanthanum, that was corrected accordingly, namely
[La] BIFs corrected = [La] BIFs measured [La] IF-G recommended concentrations /
[La] IF-G measured. Gd data are not reported due to the signicant effect
of polyatomic interferences on the measurements. Major and trace
element contents are presented in Table 3.
Sandstone whole-rock analyses were performed by the Antellis
society (Toulouse) after alkaline fusion and ICP-MS analysis. Mineral
composition of the diagenetic pyrite crystals were analyzed in situ on
polished sections of sandstones with the Electron Microprobe
CAMECA SX50 of the LMTG, Toulouse. Results are compiled in Tables 4
and 5.
3.3. Isotope geochemistry
3.3.1. Fe isotopes
Five milligram aliquots of sample powders from the BIFs and the
sandstone pyrites were rst digested in a 1.75 mL mixture of ultrapure
HCL-HF-HNO3 on a hot plate for 12 h, after which the solution was
evaporated dry. The remaining precipitate was taken up in 2.0 mL of 6N
HCl, and the sample was dried down. The remaining salt was redissolved
in 0.5 mL of 6N HCl, at which point it was ready for purication. All
samples were puried via anion exchange chromatography in a HCl
medium using 0.5 mL of Bio Rad AG1 X4, 200400 mesh resin loaded in
Teon columns. Samples were loaded and the matrix eluted in 4 mL of
6N HCl. Fe was eluted in 3 mL of 0.05N HCl. Isotopic compositions of Fe
were measured using a Thermo Electron Neptune MCICP-MS, which is
able to resolve Fe isotopes from the main interferences, such that Fe
peaks appear as at shoulders on the left side of interference peaks
(Weyer and Schwieters, 2003). A Ni spike of the same concentration as
Fe was added to each puried sample prior to analysis. Samples were
introduced to the instrument in a 0.05 HCl solution using tandem quartz
glass spray chamber arrangement (cyclone+ standard Scott double

Table 2
Mineral abundances (in %) in BIF microsamples. SM refers to silica-rich microsamples;
IM refers to iron-rich microsamples (see Fig. 2 for precise location).
Microsamples
610
610
610
625
625
625
640
640
640

SM
IM1
IM2
SM
IM1
IM2
SM
IM1
IM2

Quartz

Hematite

73.7
69.5
39.0
98.6
76.3
55.4

100
100
100
23.4
15.3
56.3
1.41
23.7
40.7

Magnetite

3.8

Goethite

Total

2.9
15.2
4.7

100
100
100
100
100
100
100
100
99.9

S. Fabre et al. / Chemical Geology 289 (2011) 124139

127

Table 3
Major and trace element contents of BIF whole rocks and microsamples; Isua BIFs (IF-G standard) for comparison. WR: whole rock, SM: silica-rich microsample IM: iron-rich
microsample (see Fig. 2 for precise location).
610

SiO2 (%)
TiO2
Al2O3
Fe2O3
MnO
MgO
CaO
Na2O
K2O
P2O5
SUM
Ba (ppm)
Sr
Th
Zr
Ti
Y
Cu
V
Cr
Mn
Co
Ni
La (ppm)
Ce
Pr
Nd
Sm
Eu
Tb
Dy
Ho
Er
Tm
Yb
Lu
Sum (13 REE)
Y/Ho
La/Lab
Eu/Eub
Ce/Ceb
Pr/Prb
Er/Erb

625

640

Isua BIF (IF-G)

WR

SM

IM1

IM2

WR

SM

IM1

IM2

WR

SM

IM1

IM2

WR

1
b DL
b DL
98.8
b DL
b DL
b DL
b DL
0.02
0.15
99
12
1.6
0.1
1.02
19.6
2.7
6.6
37.6
7.4
232
0.7
5.6
1.9
4.2
0.5
1.8
0.4
0.2
0.08
0.5
0.06
0.22
b DL
0.1
b DL
10
45
0.81
1.35
0.90
1.06

0.6
bDL
bDL
94.7
bDL
bDL
1.5
bDL
bDL
bDL
96.9
25
6
1
1
39
7
24

0.47
bDL
bDL
89.6
bDL
bDL
2.5
bDL
bDL
bDL
92.6
0.10
0.06
0.03
0.02
1
0.02
0.2

0.41
b DL
b DL
94.2
b DL
b DL
1.8
b DL
b DL
b DL
96.4
10
2
0
1
15
3
13

52
b DL
b DL
47.2
b DL
b DL
0.7
b DL
b DL
b DL
100
10
2
0.4
2
43
2
80

71
b DL
b DL
19.0
b DL
b DL
1.1
b DL
b DL
b DL
91.2
30
2
0.3
1
27
3
12

33
b DL
b DL
62.6
b DL
b DL
1.8
b DL
b DL
b DL
97.8
10
2
0.2
1
19
2
58

46
b DL
b DL
50.6
b DL
b DL
1.5
b DL
b DL
b DL
96.3
6
3
0.2
1
22
2
40

33
bDL
bDL
64.9
bDL
bDL
1.4
bDL
bDL
bDL
98.3
11
2
1
7
51
2
74

294

0.1

168

594

209

221

571

629

480

3
5
0.5
2
0.6
0.2
0.1
0.6
0.1
0.3
0.01
0.2
0.04
13
29
1.83
1.31
1.27
0.9
1.13

6
11
1
6
1
0.5
0.3
1
0.2
0.7
0.03
0.5
0.06
28
33
1.19
0.74
1.01
1
1.09

3
5
0.5
2
0.5
0.2
0.1
0.6
0.1
0.3
0.01
0.3
0.04
12
30
1.44
0.44
1.09
0.96
1.13

2
1
0.1
0.5
0.08
0.07
0.02
0.2
0.02
0.08
0.004
0.10
0.01
5
77
7.81
0.86
1.47
0.83
1.17

0.5
2
0.2
1
0.1
0.09
0.02
0.1
0.04
0.07
0.003
0.2
0.02
4
49
3.83
0.71
1.52
0.8
1.7

3
2
0.2
0.6
0.1
0.1
0.03
0.1
0.05
0.1
0.00
0.08
0.01
6
45
1.1
2.5
6.72
0.65
0.89

47
b DL
b DL
53.1
b DL
b DL
b DL
b DL
0.03
b DL
100
3
2
0.03
0.07
b DL
b DL
16
0.4
8
b DL
0.6
4
0.9
1
0.1
0.6
0.06
0.05
0.02
b DL
0.01
0.06
b DL
b DL
b DL
3

60
bDL
bDL
32.0
bDL
bDL
1.4
bDL
bDL
bDL
93.3
10
2
0.4
1
26
2
101

50
bDL
bDL
48.1
bDL
0.02
bDL
bDL
bDL
0.10
98
5
1
0.1
1
19
2
75
1
5
364
3
7
0.7
1
0.1
0.5
0.1
0.07
0.02
0.2
bDL
0.1
bDL
0.07
bDL
3

4
2
0.2
0.9
0.1
0.1
0.02
0.1
0.04
0.05
0.003
0.06
0.02
7
82
1.56
2.06
0.68
1.21
1.13

3
3
0.3
1
0.3
0.2
0.03
0.2
0.05
0.1
0.01
0.1
0.01
8
51
1.65
2.27
0.92
1.04
1.12

3
2
0.3
1
0.3
0.1
0.03
0.2
0.03
0.10
0.01
0.1
0.01
7
56
3.66
2.94
0.96
1.02
0.67

41
0.014
0.15
53.98
0.042
2
2
0.03
0.01
0.06
100
2
3
0.1
1
8
9
10
2
4
3
29
23
2.8
4
0.4
1.8
0.4
0.4
0.1
0.8
0.2
0.6
0.1
9
0.09
7
45
2.32
2.17
1.24
0.93
1.06

1.60
2.54
1.35
0.86

1.97
7.16
1.07
0.97

DL stands for detection limits. In the case of REE, Min value for Tm (1.14e-4 ppb) and Max value for Nd (8.11e-3 ppb).
b
Anomalies calculated following Bohlar et al. (2004, 2005).

Because of the very low concentrations and the great variability of the mesobands, the WR composition cannot always be expressed by a mixture of the composition of the
micro-samples, particularly in the case of the 640 samples.

pass) together with a PFA nebulizer and standard Ar gas. Typical signals
for 56Fe were ca 45 V, and a mass resolution (M/M) of 10,000 and
plateau width of 200 ppm were commonly achieved. Instrumentallyinduced mass fractionation was corrected using the mass bias measured

for Ni masses 60 and 61, combined with the standard-sample-standard


bracketing technique. A detailed report of the analytical techniques may
be found in Poitrasson and Freydier (2005). Iron and the internal
standard, Ni, were both set to a concentration of 3 ppm.

Table 4
Major element compositions of Aguas Claras sandstones.

Table 5
Chemical composition of the Aguas Claras pyrite (in wt.%) obtained by Electron Probe
Microanalyzer.

SiO2 (%)
TiO2
Al2O3
Fe2O3
MnO
MgO
CaO
Na2O
K2O
P2O5
LOI
Sum
Stot (%)

706P2-90,6

706P2-92,2

706P2-92,25

53.8
0.4
9.5
3.1
5.9
6.4
9.4
0.1
5.1
0.2
5.1
99
0.52

55.5
0.3
8.8
2.9
5.0
5.7
9.4
0.1
5.3
0.2
5.6
99
0.46

55.1
1.2
14.7
4.6
1.9
5.1
3.3
0.8
9.1
0.1
3
99
1.93

Label

Mn

Fe

Co

Ni

Cu

Zn

As

AZUL90-c1-2
AZUL90-1
AZUL90-c2-1
AZUL90-c2-2
AZUL90-c3-1
AZUL90-c4-1
AZUL90-c4-3
AZUL90-c5-2
AZUL90-c5-3
AZUL90-3
AZUL90-4

66.76
65.97
65.89
66.10
65.90
66.08
66.24
66.31
66.21
66.07
66.07

0.06
b DL
0.02
0.00
b DL
0.01
0.02
0.03
0.06
0.02
0.00

33.00
33.85
34.00
33.71
33.81
33.81
33.66
33.58
33.58
33.72
33.81

0.12
0.05
0.01
0.02
0.07
0.00
0.00
0.00
0.03
0.09
0.01

0.00
0.03
b DL
0.03
0.03
0.01
b DL
0.00
0.02
b DL
0.01

b DL
b DL
b DL
b DL
0.01
b DL
b DL
b DL
0.02
b DL
b DL

b DL
b DL
0.01
b DL
0.01
b DL
b DL
b DL
b DL
b DL
b DL

0.05
0.11
0.07
0.13
0.16
0.09
0.07
0.08
0.09
0.10
0.10

DL stands for detection limits.

128

S. Fabre et al. / Chemical Geology 289 (2011) 124139

The hematite standard was run every 5 samples, and IRMM-14, the
European standard reference material for Fe isotopes, was analyzed
between each sample under the same analytical conditions. Each
sample was analyzed three times from which standard deviation was
calculated.
Fe isotope results are reported using the common delta notation
(in ) relative to the IRMM-14 iron isotope reference material:
57

Fe =



57


54
Fe= Fe

sample

57

54

Fe= Fe


IRMM14


1

1000:

Here, both 56Fe and 57Fe will be reported and are linked by the
following relation 56Fe = 2/3 57Fe. The IF-G value obtained here,
57Fe = 0.93 0.18 (corresponding to 56Fe = 0.64 0.13), is close to
that reported by Dauphas et al. (2004), namely 56Fe = 0.576 0.037.
As explained in Poitrasson et al. (2009), analytical reproducibility
was estimated on the basis of 135 analysis of our hematite standard
(ETH hematite standard) conducted over the 3.5 years of the
present study. Non-pooled analysis our case , yields
57Fe = 0.757 0.124 (2) and 56Fe = 0.510 0.079 (2).
3.3.2. S isotopes
Sedimentary sulde sulfur was extracted using an acidic chromous
chloride solution (Caneld et al., 1986). Liberated hydrogen sulde
was initially precipitated as zinc sulde in a zinc acetate solution and
subsequently converted to silver sulde. Multiple sulfur isotope
measurements ( 32S, 33S, 34S and 36S) were performed at the Institut
de Physique du Globe (IPGP) in Paris, France, using a ThermoFinnigan
MAT 253 mass spectrometer. Silver suldes samples were converted
to sulfur-hexauoride (SF6) via uorination in nickel tubes (Taylor,
2004). Puried SF6 was introduced into the mass spectrometer via a
dual inlet system and masses 127, 128, 129, and 131 (representing
32
S, 33S, 34S, and 36S, respectively) were recorded simultaneously (see
Handbook of stable isotope analytical techniques, chapter 20, for a
description of the analytical system). All replicate analyses of
standards were reproducible within 0.3. Subsequently, 33S and
36S values were calculated as (e.g., Hulston and Thode, 1965;
Farquhar et al., 2000):
33

33


0:515 
34
1 + S= 1000
1

36


1:90 
36
1 + S= 1000
1 :

S = S1000
and
36

S = S1000

The data are reported relative to V-CDT.


On the V-CDT scale, the standard materials IAEA-S1 show 33S and
36
S values in the same range as Ono et al. (2007) (see Table 7).
Reference gas is +17,001 vs S1, whereas S0 reference gas is
+16,701 vs CDT assuming that S1 = 0.3 vs CDT.
33S and 36S are calculated from the raw data and give values
similar within error to data reported from J. Farqhar's laboratory,
showing that the reference gasses lie on the same fractionation line.
No correction is made on 36S as our reference gas appears devoid of
the contaminant (likely C3F6) which is known to produce an isobare
interference on m/z = 131. Reproducibility for 33S (i.e. including
uorination) is better than N 0.008 (1).
4. Petrography and mineralogy
4.1. BIFs
Macroscopically, all BIF samples (640, 625 and 610) display the
meso- and microbanding usually described from these rocks (e.g. Pecoits

et al., 2009). Sample 610 was collected in the most iron-rich level
(so-called jaspilite ), likely corresponding to an iron-rich macroband
that is a few tens of meters thick and ca one kilometer in strike. Following
the classication of Trendall and Morris (1983), mesobands are
centimeter thick units, here either gray or reddish in color, that
respectively correspond either to an iron (hematite)-rich or a silicarich composition (Fig. 2 a, b and c). The silica mesobands, when colored
by admixed iron oxide dust, are referred to as jasper. The jasper varies in
color from deep brownish red (sample 640) to purple (sample 610). The
nature and intensity of the coloration seems to be controlled by the
mineralogy of the disseminated iron oxide dust. The silica and iron bands
are generally continuous showing little folding. The proportions of both
types of mesobands (silica-rich or S-bands, and iron-rich or Fe-bands)
were determined by image analysis (Table 1). In addition, each
mesoband often displays further varve-like alternation of microbands,
i.e. (sub) millimeter thick iron-rich and silica-rich laminae (Fig. 2).
Sample 610 differs from the other samples, as it contains fewer or no
silica mesobands. However, a microbanding with thicker Fe-microbands
(23 mm thick) and thinner S-microbands (1 mm thick) is still
recognizeable in this sample.
X-ray diffraction analyses of BIF samples (Table 2) reveal that the
major constituents are quartz and hematite with minor magnetite and
goethite (probably secondary). No carbonate or argillaceous minerals
have been detected. Quartz is below the X-ray detection limit (b5%) in
sample 610, that mainly contains cryptocrystalline silica. SEM observations show aggregates of euhedral platy hematite grains (Fig. 3a) of
variable sizes (110 m). Two oxide generations are sometimes
recognized, platy hematite and secondary acicular magnetite grown
on the hematite crystals (Fig. 3b).
4.2. Pyrite-bearing sandstones
Argillaceous sandstones of the Aguas Claras Formation are medium
gray to dark in color and display a well-dened millimeter-thick
stratication. The rocks were folded, but not metamorphized. In thin
section, the sandstone is composed of very ne-grained (b1 mm)
detrital quartz with subordinate sericite, biotite, chlorite, zircon and
opaque minerals. Finely disseminated pyrite grains are ubiquitous; a
few layers contain larger pyrite grains easily identied with the naked
eye (Fig. 2d and e). The pyrite-rich levels never crosscut the
stratication planes, therefore pointing to an early diagenetic nature
for the pyrite. Individual pyrite grains were selected, microdrilled and
analyzed; these grains are mainly euhedral and up to 3 mm in diameter
for the largest ones (Fig. 2f).
5. Major and trace element geochemistry
5.1. BIFs
Major and trace element contents of the analyzed BIF whole rocks
and microsamples (Table 3) are given in Table 3. Fe is the only element
enriched with respect to MORB (Fig. 4a, with normalization to MORB
following Hoffman, 1988, and Hess and Poldevaart, 1968), whereas Ba,
Th, Ce, Cu are close to the MORB values. The very low contents of all
other elements, especially Ti, Sr, V and Zr, are consistent with the lack of
a signicant contribution from continental crust for these chemical
sediments. At last, compared to present-day oceanic hydrothermal uid
composition (VonDamm, 1995; German and VonDamm, 2003), the
trace metal contents of the Carajs BIFs, enriched relative to seawater,
also seems to indicate mac-hosted hydrothermal sources (Table 3).
REE + Y distribution patterns normalized to Post-Archaean Australian Shales (PAAS: McLennan, 1989) are presented in Fig. 4(c). They
are consistent with the preliminary data of Klein and Ladeira (2002).
Sample 610 is characterized by smoother REE + Y patterns and higher
total REE concentrations ( = 2349 ppm) when compared to the
other samples ( = 621 ppm). Most samples display a strongly

S. Fabre et al. / Chemical Geology 289 (2011) 124139

129

640 IM1

Silica microband

Iron mesoband

Iron microband

640 SM1
Silica mesoband

640 IM2

1 cm
1 cm

625 IM1

Iron mesoband

625 SM

Silica mesoband

Pyrite
625 IM2

1 cm

610 IM1
610 SM1
610 IM2

Silica microband
Iron mesoband

1 cm

Fig. 2. BIF samples from N4 mine (sample numbers refer to present topographic height): (a). 640, (b) 625 and (c) 610 samples showing the microsampling points (SM for silica
mesoband and IM for iron mesoband); (d) Core fragment (706 P2 from depth 92.2 m) from the Aguas Claras Formation; (e) polished section of another Aguas Claras core sample
(stratication planes perpendicular to the section); (f) photomicrograph of diagenetic pyrites (black) in their arenaceous-argillaceous matrix.

superchondritic Y/Ho (see Table 3 and Fig. 5), ranging from 30 to 82,
with an average value of 50. This is considered typical for seawater
precipitates (Kamber et al., 2004). The samples also display positive
Eu anomalies, that are usually observed in Archaean BIFs and regarded
as witnessing a hydrothermal inuence (Bau and Dulski, 1996; Frei
et al., 2008; Planavsky et al., 2010). A positive La anomaly is observed
in most samples. Finally, the large negative Ce anomaly, typical of
modern oxidized seawater (Eldereld and Greaves, 1981; Alibo and
Nozaki, 1998), is either smaller, or absent, or even replaced by a small
positive anomaly, pointing to the less oxidized state of their Archaean
seawater source.

5.2. Pyrite-bearing sandstones


Table 4 provides the major element contents of the pyrite-bearing
sandstone samples from the Aguas Claras Formation. Actually,
elementary abundances conrm the rather shaly nature of the
so-called Aguas Claras sandstones. The rocks are arkoses (90.6 and
92.2) and wacke (92.25) following the classication scheme of Herron
(1988). Based on their Zr content and Y/Ho ratio, samples plot
between the average continental crust and modern seawater (Fig. 5),
which is consistent with both their terrigenous origin and their
marine deposition.

130

S. Fabre et al. / Chemical Geology 289 (2011) 124139

Intensity (cps)

8000

O
Fe

Platy hematite
(point 1)
EN 640 sample

6000
4000

Fe

2000

Fe

Si

10

15

20

25

Energy (keV)

Intensity (cps)

8000

6000

Fe

Magnetite (point 2)
EN 640 sample

4000
Fe

2000

Fe

Si

0
0

10

15

20

25

Energy (keV)
25000

Intensity (cps)

Si

20000

Quartz (point 3)
EN 640 sample

15000
10000

5000
0
0

10

15

20

25

Energy (keV)

Fig. 3. (a) Euhedral platy hematite by direct observation on the 625 sample (magnication 10,000; acc. V. 25 kV; Signal: SEI; WD: 11 mm. Spot size: 5 m); (b) Relationships
between iron minerals and quartz in the EN 640 sample (magnication 2500; Acc. V. 20 kV; Signal: BEC; WD: 11 mm. Spot size: 50 m); (c) EDS spectra corresponding to the 3
points 1, 2, and 3 depicted on the microphotograph.

Analyses of individual pyrite crystals (Table 5) are characterized


by rather high As contents (5301577 ppm) and variable Co contents
(131179 ppm). Mn contents are lower than 643 ppm and Ni
contents are lower than 299 ppm; Cu and Zn contents are even
lower (b163 ppm, b133 ppm). Arsenic in pyrite from marine
sediments is generally restricted to low contents (20300 ppm),
hence high As content in pyrites is sometimes regarded as indicative
of a hydrothermal origin. However, arsenian authigenic pyrites
(ca 3200 ppm as median value, with a range from 300 to
13,000 ppm) have been described from pyrite- and organic matterrich sediments (Lowers et al., 2007). Such high values are witness of a
highly reducing and/or suldic sedimentary environment and do not
challenge the syndiagenetic nature of the analyzed samples. This
interpretation is supported by the preserved sedimentary features,
that always point to the diaganetic nature of pyrites.

6.1.2. Pyrite-bearing sandstones


The Fe isotope values determined for the pyrite samples (Table 7)
are also relatively high and positive. The mean 57FeIRMM value is
1.03 0.4 ( 56FeIRMM-14 = 0.62 0.04), (n = 3; minimum : 0.88;
maximum = 1.25).
6.2. S isotopes
34S values for sedimentary pyrite are positive and cluster in a
tight range from 11.75 to 12.05 (Table 7), with an average value of
11.97 0.12. 33S values vary between 0.008 and 0.017, with
36S values ranging from 0.079 to 0.220.
7. Discussion
7.1. Origin and preservation of BIF samples

6. Isotope geochemistry
6.1. Fe isotopes
6.1.1. BIFs
Fe isotope results (Table 6) are remarkably constant, showing a
mean value of 57FeIRMM-14 = 2.06 0.34 ( 56FeIRMM-14 = 1.38
0.23). These are among the most positive values ever measured in
BIFs (Johnson et al., 2003; Rouxel et al., 2005; Johnson et al., 2008a,
2008b).

The lack of detrital minerals (such as clay) as well as the trace


element contents and distribution pattern argue for a marine
precipitation of the BIFs, specically in a distal position. The positive
Eu anomaly can be interpreted as the ngerprint of an oceanic
hydrothermal origin for the uid that provided both the REEs and Fe
(Bolhar et al., 2004; 2005). Finally, the lack of a large negative Ce
anomaly points to the anoxic nature of the deep oceanic waters as the
source of the precipitated iron (Kamber and Webb, 2001). Indeed,
anoxic waters are required to keep any iron that was leached from the
oceanic crust in solution.

S. Fabre et al. / Chemical Geology 289 (2011) 124139

640 WR
640 IM2
640 SM

IF-G

10

Table 6
Fe isotopic compositions of Carajs BIFs. SM: Silica-rich microsample, IM: iron-rich
microsample.

610 WR
610 IM2
610 SM

625 WR
625 IM2
625 SM

Samples

Sample/MORB

1
0.1
0.01
0.001
0.0001

PAAS normalised concentration

Ba Sr Th Ce Zr Sm Ti YC u
1

VC r

Mn Fe Co Ni

0.1

0.01
La Ce Pr Nd Sm Eu Tb Dy YH oE rT mY bL

Fig. 4. (a) Trace element concentrations of BIF whole rocks normalized to average
MORB. (b) PAAS-normalized REE + Y distribution patterns from BIFs microsamples.

A point to address is whether the elemental abundances


determined here reect near-primary signatures or whether postdepositional alteration has resulted in elemental depletion or
enrichment. Two arguments should be considered in that respect.
First, Bau and Dulski (1996) pointed out that the impact of the postdepositional processes (diagenesis, metamorphism and surface
weathering) is negligible on the primary composition of BIFs.
Following Bolhar et al. (2004), diagenetic reduction of Eu(III) to
Eu(II) would have been accompanied by the mobility of Fe(II) which
would have destroyed the well-dened alternation of Fe-rich and Sirich microbands. Eu is consistently enriched relative to its neighbors
in all samples, regardless of absolute concentration so that either Eu or
Fe mobilization can be ruled out. In addition, important diagenetic

Zr (ppm)
BIF samples
from Carajas
610 samples
625 samples
640 samples
Aguas Claras
sandstones

Upper
continental
crust

100

10

Modern seawater

1
01

0.1

10

20
Chondritic
ratio

40

50

60

70

80

90

Y / Ho

Fig. 5. Zr vs Y/Ho diagram for Carajs BIF and Aguas Claras pyrite 921 samples (respective
data for modern seawater and continental crust are plotted for comparison).

Height Fe
(m)
(%)

610
610 SM (6)a
610 IM1(6)
610 IM2 (6)
610 mean valueb
625 SM (6)a
625
625 IM1 (6)
625 IM2 (6)
625 mean valueb
640 SM (6)a
640
640 IM1 (6)
640 IM2 (6)
640 mean valueb
I F- G s t a n d ar d
(Isua BIF) (6)
a

1000

131

Fe/Si

62.6
228.3
66.2
284.7
65.8
346.3
64.9
286.4
32.9
0.4
13.3
1.4
43.2
2.8
29.8
1.5
11.4
0.4
35.3
1.7
45.3
3.3
30.7
1.8
44.60
2.06

56FeIRMM-14 2
()

57FeIRMM-14 2
()

1.13
1.07
1.16
1.12
1.46
1.41
1.50
1.45
1.75
1.41
1.56
1.57
0.64

1.67
1.61
1.74
1.67
2.18
2.07
2.26
2.17
2.58
2.09
2.30
2.32
0.93

0.17
0.04
0.12
0.09
0.04
0.08
0.06
0.09
0.04
0.11
0.09
0.34
0.13

0.23
0.06
0.12
0.13
0.08
0.01
0.02
0.19
0.21
0.12
0.14
0.49
0.18

The number in brackets refers to the number of analysis for each sample.
Arithmetic values from data and not analytical uncertainties.

modications and continental contamination would have also


lowered the Y/Ho ratios, a modication that is not observed, as the
ratios are still noticeably superchondritic and seawater like. Second,
Dauphas et al. (2007) and Frost et al. (2007) examined the effect of
upper amphibolite/lower granulite facies metamorphism on the Fe
isotopic composition of BIFs and demonstrated that the mobility of Fe
is negligible even under higher grade metamorphic conditions.
Nevertheless, Valaas-Hyslop et al. (2008) demonstrated that 56Fe
values in high grade Biwabicks rocks were 0.3 lower than those in
their protolith. They concluded that metamorphism decreases the
56Fe values in rocks that also contain Fe-silicates, thus positive 56Fe
values cannot be explained by metamorphic reactions. Hence, as the
Carajs BIFs have experienced only lower greenschist facies conditions, we can be reasonably condent that the Fe isotopic compositions measured here are likely to be primary in nature, i.e. they have
been acquired during the oxidation of dissolved Fe originating from
the oceanic crust. Other recent works plea in favor of this scenario: for
instance, Steinhoefel et al. (2009) consider that the micrometer scale
differences in the Fe and Si isotopes in BIF are largely of primary
origin, so that post-depositional redistribution across layering and
recycling to the ocean could be considered as a minor phenomena.
Trace elements compositions of the BIFs argue for their seawater
origin. Following the traditional view, Fe originates from hydrothermal vents in the deep anoxic ocean and precipitates in the form of
oxy-hydroxide (FeOOH) in the relatively oxic upper layer of the ocean
or at a redoxcline between the lower and upper ocean, likely as a
consequence of seasonal upwelling. An alternative hypothesis
proposed by Steinhoefel et al. (2009) attributes the laminae to
changes in the upwelling dynamics of hydrothermal-rich water.
While the Archaean ocean is considered to have been anoxic in its
deeper part, oxygenic photosynthesis in the photic zone was the likely
cause of local oxygenation of the surface waters (Braterman and
Cairnsmith, 1987). Here, REE + Y adsorbed onto the surface of the
oxy-hydroxide particles (Bau, 1999) are responsible for a seawaterlike distribution pattern of the REE + Y for these chemical sediments.
Subsequently, the oxy-hydroxide precipitates transformed directly
into hematite. The expected Fe isotope fractionation will yield positive
values for the hematite (Johnson et al., 2005), as observed in this
study (Fig. 6). After deposition, only a minor amount, if any, of the
hematite has been reduced to magnetite, for instance by dissimilatory
iron reduction (DIR), a process that would have lowered the Fe
isotope values (Johnson et al., 2005; Severmann et al., 2006).
However, it is likely that this mechanism will have a little impact on
our bulk BIF isotope Fe composition because their magnetite content
is low (Table 2).

132

S. Fabre et al. / Chemical Geology 289 (2011) 124139

Table 7
Fe and S isotopic compositions of Aguas Claras diagenetic pyrites.
Samples

56FeIRMM-14

57FeIRMM-14

36S

33S

34SVCDT

92.2 (6)a
92.2(3)
92.2(3)
92.25 (6,3)
90.6 (6,3)
Mean valuesb
Standards
IAEA-S1 (3)

0.64

0.21

1.26

0.08

0.66
0.58
0.62

0.08
0.05
0.04

0.94
0.89
1.02

0.08
0.05
0.20

0.022
0.014
0.079
0.041
0.220
0.021
36S
0.907

0.63
0.63
0.63
0.63
0.63
0.12
2
0.075

0.009
0.015
0.015
0.017
0.008
0.013
33S
0.082

0.041
0.041
0.041
0.041
0.041
0.004
2
0.014

12.03
12.05
11.99
11.75
12.05
11.97
34S
0.320

0.12
0.12
0.12
0.12
0.12
0.13
2
0.107

a
b

In brackets is indicated the number of replicated analysis. The rst number refers to iron isotopes measurement, the second one to sulfur isotopes measurement.
Arithmetic values from data and not analytical uncertainties.

At last, a recent study by Halverson et al. (2011) on Neoproterozoic


Rapitan BIFs shows the same concordance between hematite
content and positive 57Fe signature. That seems to indicate similar
formation and depositional processes of the BIFs at two very different
times.
7.2. Oxidation state of the atmosphere during Aguas Claras sandstone
formation
The age of the Aguas Claras sandstones is a matter of debate (see
Section 2). The 2.06 Ga age proposed by Mougeot et al. (1996), that
we retain in this study, sets an older limit for the depositional age of
the giant Mn deposit found at the top of the Aguas Claras Formation.
Consistently, this Mn deposit can be regarded as being late
Paleoproterozoic in age, similar to the other giant Mn deposits in
the world (Roy, 1997; Gauthier-Lafaye and Weber, 2003).
Indirect evidence with respect to the age of deposition is gained
from multiple sulfur isotopes data in sedimentary suldes (pyrite) (cf.
Farquhar et al., 2000). It has been shown that sediments older than ca.
2.4 Ga display non-zero 33S values in most cases (see Ohmoto et al.,
2006 for controversial observations) and show a 33S/ 36S slope from
~0.9 to ~1.5 (Ono et al., 2003; Farquhar et al., 2007; Thomazo
et al., 2009a). Sediments that were formed thereafter always lack this
distinct isotopic signature (Farquhar et al., 2000; 2007; Guo et al.,
2009) and show a 33S/ 36S slope of ~7. Respective multiple sulfur
isotope signatures are viewed to result from photochemical dissociation of volcanogenic SO2 in an oxygen-poor atmosphere (Farquhar
et al., 2000). Pavlov and Kasting (2002) calculated that this reaction
can only occur if atmospheric oxygen levels were b10 5 the present
atmospheric level (PAL). Based on early accounts by Farquhar et al.
(2000) and the threshold value provided by Pavlov and Kasting
(2002), it can be concluded that, at atmospheric oxygen abundance
higher than 10 5 PAL, mass-independent sulfur isotope fractionation
during atmospheric reactions involving SO2 would have been
inhibited. Moreover, any transfer to the surface environments
would have inevitably resulted in a loss of these distinct signatures
due to oxidation.

Sea level
Shallow oxic ocean

Partial iron
oxidation
2+
Feaq

2+
[57Fe < 0]
Feaq

Fe3+

Fe(OH)3

Deep anoxic ocean


2+
[57Fe ~ 0]
Feaq

Fig. 6. Schematic view of Fe isotope exchange reactions between the different oceanic
reservoirs and BIF deposits.

The low 33S values obtained for sedimentary pyrite from the
Aguas Claras sandstone and a 33S/ 36S slope different from ~1
collectively argue that Aguas Claras pyrites experienced solely massdependent sulfur isotope fractionation. The lack of distinctly massindependent sulfur isotope fractionation is viewed as supporting a
depositional age that is younger than 2.4 Ga.
7.3. Iron cycling in the Archaean ocean
The modern ocean contains ca 0.002 ppm of dissolved iron
(Nordstrom et al., 1979), because of its globally oxidized state. 57Fe
values of particulate and dissolved Fe are close to 0 (0 for bulk
clastic sediments, 1,5 to 0 for hydrothermal vent uids, and 1 to
0 for FeMn crusts; Bennett et al., 2009; 0.02 to +0.46 in the
particulate fraction; Lacan et al., 2008; Radic et al., 2011). The
dissolved riverine ux of Fe has a variable iron isotope signature (i.e.
positive and negative values), as reported by Fantle and DePaolo
(2004), Bergquist and Boyle (2006), and Escoube et al. (2009). In the
present-day soils, that can be regarded as contributors to ocean iron
budget, Poitrasson et al. (2008) shown that iron isotope signature
( 57Fe) is close to the mean crustal value with a maximum range of
0.2. However, this Fe isotope reference frame cannot uniformly be
expanded to the Archaean ocean, because of the distinct changes in
the redox state of the oceanatmosphere-system which have
occurred since this period (Rye and Holland, 1998; Yamaguchi et al.,
2007).
As previously explained, the redox variations are the more likely
phenomenon able to induce isotopic variations. In fact, the isotopic
data require that an oxidant was present in the earliest period of
Earth's history, and the positive 57Fe values indicate that oxidation
did not run to completion, implying that the oxidation was limited.
7.3.1. Constraints on the isotopic signature of the BIF's hematite
We explore this idea by using a two-stage model in an idealized
supercial Archaean ocean equilibrated with atmospheric gasses on
one hand and with carbonate sediments on the other hand (Fig. 7a), in
order to constrain the pH, which is of utmost importance to constrain
the iron speciation. In our simplied approach, the only iron specie
considered thereafter is the aqueous iron species Fe(II). We suppose
that: (1) Fe(II) is oxidized in Fe(III) and (2) Fe(III) precipitates as iron
oxy-hydroxydes (FeOOH and/or Fe(OH)3) and then re-crystallizes
as hematite. The rst step is assumed to respond to a Rayleigh
distillation model, because the once formed Fe(III) doesn't interact
with aqueous Fe(II) anymore. The second step doesn't lead to any
fractionation, because all Fe(III) is assumed to convert in iron oxyhydroxydes. There is no real consensus about iron fractional
coefcient (see for example Saunier et al., submitted for publication),
so we use recent available values for aqueous iron complex (Schauble
et al., 2001, is taken as a representative value) in order to calculate an
iron isotope fractionation coefcient () between the different iron
aqueous species (see Table 8 for details).

S. Fabre et al. / Chemical Geology 289 (2011) 124139

133

Equilibrium with
atmospheric CO2

a
Microbial photosynthesis
dissolved O2

Fe(III)aq

Fe(II)aq

Fe oxy-hydroxydes

Oxydation
Isotopic fractionnation

Precipitation

Upper archean
ocean
(photic zone)

(Rayleigh distillation

Hematite
Redoxcline

Discontinuous Fe(II)aq
input e.g. seasonal
upwelling

Equilibrium with
carbonate minerals

b
57

Calculated Fe
in hematite ( )
9
8

T= 0C, log pO2= -2.69

T= 0C, log pO2= -4.69

T= 25C, log pO2= -2.69

T= 25C, log pO2= -2.69

2.06 , mean value

measured in Carajas BIFs

3
2
1

Proportion of Fe(II) oxydized

0
0

0.2

0.4

0.8

0.6

Corresponding values for


Fe(II) oxydized proportion

Fig. 7. (a) Details of the one-box model of the iron cycling in the Archaean supercial ocean. (b) Results of the model: 57Fe calculated for precipitated hematite in relation to the
consumed Fe(II)aq concentration values; inuence of temperature (wide symbols for T = 0 C and full symbols for T = 25 C symbols) and fugacity of O2 (triangles for
pO2 = 10 2.69 bars and squares for pO2 = 10 4.69 bars) are displayed.

A Rayleigh distillation equation is used to calculate the 57Fe of the


Fe(III):
h
i
1
57
57
1000
FeIII = Feo + 1000 ::f

where 57Feo is the initial Fe isotope composition, namely the iron


isotope signature of the hydrothermal uids coming out of the oceanic
crust, that is assumed to be 57Feo = 0, f is the oxidized fraction of
Fe(II) and is the calculated Fe 2+Fe 3+ fractionation factor. We
assume that all the Fe(III) is converted into iron oxy-hydroxyde as the
very high calculated saturation indices (see the kinetic paragraph
below) seems to indicate. It appears that ca 44% of the initial iron mass
must precipitate in order to generate 57Fe at ca 2 as observed
(Fig. 7b). This value is used in order to calculate the time required to
oxidize the aqueous iron.
7.3.2. Kinetics of iron oxidation in the Archean ocean
The Archaean seawater composition is not known, so we use the
classical composition of present-day seawater (Nordstrom et al., 1979)
with several assumptions: (1) pH is adjusted to a p(CO2) = 0.1 bars
(Kastings, 1993) and the ocean is equilibrated with carbonate minerals,

Table 8
Polynomial coefcients for calculating iron- factors aqueous species and corresponding isotopic fractionation coefcient ().
Coefcient of the polynomial expansion: 103ln = Aixi =
106/T2
Species/
mineral
1000.ln(56-54)
1000.ln(57-54)
1000.ln(56-54)
1000.ln(57-54)
1000.ln(57-54)

[FeII(H2O)6]2+
III

[Fe (H2O)6]

3+

Pyrite

Fe(III)/Fe(II)
Pyrite/Fe(II)

Source

A1

A2

A3

Schauble
et al.
(2001)

0.555
0.8325
1.149
1.7238
1.2437

0.0016
0.0024
0.011
0.01665
4.82E-03

0.0002
0.0003
0.00006
0.00009
2.68E-05

Blanchard
et al.
(2009)
T = 0 C
T = 25 C
T = 0 C
T = 25 C

1.014
1.012
1.004
0.990

[FeII(H2O)6]2+ is referred in the text as Fe(II)aq, in the same way, [FeIII(H2O)6]3+


refers to Fe(III) aq.

134

S. Fabre et al. / Chemical Geology 289 (2011) 124139

which lead to a pH= 6.35, (2) the iron concentrations are comprised
between 0.1 and 1 mmol/kgw as suggested by Walker and
Brimblecombe (1985) and Bjerrum and Caneld (2002) for the oceans
prior to 1.8 Ga, (3) and the redox potential (pe) depends only of the
atmospheric p(O2). We use the PHREEQC program (Appelo and
Parkhurst, 1999) developed by the U.S. Geological Survey for
modeling waterrock interactions, coupled with the Lawrence
Livermore National Laboratory database. The program determines
the chemical speciation of aqueous solutes, the fugacities of gasses,
and attributes a stable aqueous species to each component of the
solution. It also calculates the activity coefcients of water and all
aqueous species, as well as the saturation index of the solution with
respect to all the solid phases of the database. The speciation
calculation points that the major species is Fe(II)aq compared to all
other complexes (e.g. FeCl +, FeSO4) (see Appendix 1 for molalities
and activities).
The rate of oxidation of dissolved Fe 2+ by O2 is given by Singer and
Stumm (1970):
dmFe2
dt



9
12
2
= 2:91e + 1:33e aOH pO2 mFe2 +

2+
is
where t is time in seconds, aOH is the activity of the hydroxyl ion, mFe
the total molality of ferrous iron in solution and pO2 (atm) is the oxygen
partial pressure. Let us now examine the chemical constraints of the
model.
First, Pavlov and Kasting (2002) propose that pO2 levels between
2.45 and 2.0 Ga were 10,000 times less than present atmospheric level
(PAL). On the opposite, other authors (Beukes et al., 2002; Mojzsis et
al., 2003), according to eld evidences, show that pO2 at the GEO
should have been as high as 10 2 PAL. Clearly, the partial pressure of
oxygen was oscillating before the GEO and so we choose to test the
limit of the admitted values for this parameter. In our modeling, the
O2 quantity is explicitly considered un-limitating for the duration of
the iron oxy-hydroxide precipitation. We assumed, as in present-day
ecosystems, that planctonic bloom is seasonal in response to nutrient
supply by upwellings.
Second, we test two limit cases: (1) various iron concentrations,
respectively 5.58 ppm (0.1 mmol/kgw) and 55.8 ppm (1 mmol/kgw)
and (2) various p(O2), respectively 10 2 and 10 4 PAL. The kinetic
equation included in the model enables to calculate the time needed
to oxidize nearly 40% of the dissolved iron in the surface ocean, in
contact with the atmosphere. The results are comprised between four
years (logpO2 = 4.69 bars and [Fe] = 55.8 ppm) and a week
(logpO2 = 2.69 bars and [Fe] = 5.58 ppm). This is very fast compared to residence time values (10 6 years) proposed by Johnson et al.
(2003) for a global anoxic ocean. It is noticeable that, in our
hypothesis, the presence of a chemocline will reduce the iron
residence time in the deep anoxic ocean, but this residence time is
not constrained. Nevertheless, Archaean iron cycling seems to have
been organized in two decoupled stages (1) a short supercial cycle
and (2) a long deeper one, both of them likely connected by seasonal
upwellings. Moreover, the lack of variation observed in the isotopic
values suggests that, at our limited observation scale, the iron
mass is probably equilibrated.
Previous studies (Johnson et al., 2003; Rouxel et al., 2005) showed
that 56Fe for Fe-oxide minerals range from 1 to +1, but have an
average value of +0.04 (n = 73). In a 56Fe vs time diagram
(Fig. 8a), the Fe isotope values measured for the Carajs BIF plot
among the most positive ones ever measured for that type of rocks
except for older ones (e.g. Akilia/Isua; Whitehouse and Fedo, 2007;
Thomazo et al., 2009b; Dauphas et al., 2004). The highly positive Fe
isotopic signatures recorded for the Carajs BIF requires the existence
of a complementary isotopically light Fe isotope reservoir, such as the
surface seawater that contains the dissolved Fe(II). Negative 56Fe
values (in the range 2.1 to 0.5) measured in Archaean

stromatolites or in Archaean marine diagenetic pyrites (Fig. 8b)


possibly provide the evidence for the existence of such a reservoir
(Rouxel et al., 2005; von Blanckenburg et al., 2008).

7.4. Iron and sulfur cycling in the Paleoproterozoic ocean


7.4.1. Comparison with modern times
Diagenetic pyrite formation is the result of the following events:
reactive Fe 2+ combines with aqueous suldes in the sediment pore
water (Fig. 9) to form an aqueous FeS species, which subsequently
turns to FeS (mackinawite), the precursor of pyrite (Rickard and
Luther, 1997; Benning et al., 2000). The positive Fe isotope ratios
measured in the diagenetic pyrites of the Aguas Claras formation
( 57 Fe IRMM-14 = 1.03 0.40; 56 Fe IRMM-14 = 0.63 0.08) are
puzzling when compared to modern values that are systematically
lower. Severmann et al. (2006) and Butler et al. (2005) clearly
showed that modern pyrite is far from isotopic equilibrium with
Fe 2+aq relative to the theoretically predicted fractionation factor. It
displays a variety of Fe isotopic signatures, including negative ones.
Besides, from a theoretical point of view, Polyakov et al. (2007)
suggested that equilibrium precipitation of pyrite could fractionate
the heavy Fe isotope compared to the light one. Polyakov and
Soultanov (2011) propose that Fe isotope composition of pyrite is
controlled by equilibrium between aqueous Fe and a FeS precursor
or that kinetic isotope effects may have been important. Moreover,
as explained by Rouxel et al. (2008) the isotopic fractionation
between Fe(II) and pyrite is dependant of the pathways of pyrite
formation.
The Aguas Claras pyrite Fe isotopic composition displays the
expected positive values which appear to be characteristic for the
time immediately after rise of atmospheric oxygen (see Fig. 8b after
Rouxel et al., 2005). In order to test this model, we use the values of
Schauble et al. (2001) and Blanchard et al. (2009) to calculate the
fractionation factor between Fe(II) and pyrite. Again, let us mention that
there is no complete agreement on these values and recent experimental work (Saunier et al., submitted for publication) will probably make it
clear. Values are reported in Table 8 for the T interval of 020 C, and we
note no major differences in the calculated fractionation factor in this
interval. Considering a Rayleigh distillation model, only a very little
fraction (10 4) of the initial reactive iron mass must have been
consumed to reach the 57Fe (ca 1) value observed in the natural
samples of Aguas Claras. The modern pyrite formation frequently
utilizes all the available reactive Fe in the sediment (Berner, 1983).
Consequently, no isotope fractionation should occur and the pyrite
isotopic signature would reect the isotopic signature of dissolved Fe in
the pore water. Recent experimental work by Guilbaud et al. (2011) is
along the lines of that hypothesis.
In addition, modern day seawater contains appreciable amounts of
dissolved sulfate (28 mM) which is likely not exhausted in the
sediment. Hence, strongly negative 34S values are frequently
observed for biogenic pyrite resulting from the metabolism of sulfate
reducing bacteria (SRB) in modern marine sediments (e.g., Strauss,
1997). During Paleoproterozoic times, an opposite situation is
proposed (Fig. 9). Seawater sulfate is likely a limiting factor even
after the initiation of oxidative continental weathering concomitant to
the Great Oxidation Event at around 2.4 Gyr ago (see Guo et al., 2009
for respective multiple isotope evidence). Hence, near-complete
reduction of sulfate can be expected in the sediment and would
result in a 34S signature for the diagenetic pyrites that is nearly
identical to the signature of seawater sulfate. Sulfur isotope values of
ca. 12 have been measured here for sedimentary pyrites from the
Aguas Claras Formation. This value is close to a range in 34S values of
1214 (Schrder et al., 2008) or 1217 (Guo et al., 2009) that was
suggested as representative for Palaeoproterozoic seawater sulfate
around 2.1 Ga ago.

S. Fabre et al. / Chemical Geology 289 (2011) 124139

2.0

Stage2

Stage3
a

Rise of atm O2

135

Stage1

1.0

0.0
-1.0
Fe oxide minerals (N4 Mine)

-2.0

Fe oxide mineral separates from BIFs


(Johnson et al., 2003, 2008)
Fe oxide mineral separates from BIFs
(other ref. in Rouxel et al., 2005)
Archaean microbialites (von Blanckenburg et al., 2008)

-3.0

Fe-rich carbonates (after Bekker et al., 2010)

-4.0
0.0 0.5 1.0 1.5

1.8

2.2

2.0

2.4

2.8

Rise of BSR

Age (Ga)

2.0

2.6

1.0
0.0
-1.0
-2.0
-3.0

Diagenetic pyrite (Aguas Claras)


Pyrite in black shales (Anbar et Rouxel, 2007)
Diagenetic pyrite (Rouxel et al., 2005)
Zinbabwe pyrite (Archer et Vance, 2006)
Modern sedimentary pyrite (Severmann et al., 2006)

-4.0
0.0 0.5 1.0 1.5

1.8

2.0

2.2

2.4

2.6

2.8

Age (Ga)
Fig. 8. Temporal evolution of 56Fe values (after Rouxel et al., 2005): (a) BIFs (oxides); black square: mean value for Carajs BIFs; (b) Suldes; black diamond: mean value for Aguas
Claras pyrite. The gray area corresponds to 56Fe values of Fe derived from igneous rocks (at 0.1) and modern marine sedimentary pyrite (down to 1.0). The mean mac Earth
is 56Fe = 0.068 0.021 (Poitrasson et al., 2005).

7.4.2. Comparison with Archaean times


Fe and S isotope data from pyrites of Archaean sediments display
very different values from either present-day or Palaeoproterozoic
pyrites. Actually, positive isotope Fe values have not been registered
before 2.6 Ga (Fig. 8). Most Archaean pyrites have negative iron isotope
signatures. We suggest that this can be easily explained, because it is
well know that the BSR did not appear before 2.7 Ga, as evidenced by the
appearance of very negative 34S in pyrites (Grassineau et al., 2001).
This new metabolic pathway is more efcient than the older
dissimilatory iron reduction (DIR) and, therefore, it will override the
DIR (Johnson et al., 2008a,b). Moreover, it will become the main process
for the production of diagenetic pyrite in sediments. This will have
noticeable isotopic consequences. The DIR yields dissolved ferrous iron
with negative 56Fe values (Percak-Dennett et al., 2011), that can even
becomes lighter and lighter by successive cycling. In Archean times,
pyrite could precipitate using available dissolved ferrous iron (of
biogenic DIR or abiogenic hydrothermal origin) and abiogenically
produced sulfur or suldic species (mainly of volcanogenic or
hydrothermal origin), none of them being limited. In these conditions,
iron isotope signatures in pyrites possibly become sometimes slightly
heavier, but were unlikely to reach positive values. In addition, their S
isotope signatures remained rather close to zero (Caneld and Raiswell,

1999). Global trends in S and Fe isotopes at the end of the Archaean are a
consequence of the evolution of the main microbial ecosystems on
Earth.

8. Conclusions
Iron and sulfur isotopes from BIFs and pyrite-bearing sandstones
from the Carajs mineral province provide a valuable tool to explore
the redox changes of Earth's ocean and atmosphere between 2.7 and
2.1 Ga ago. The 2.7 Ga old Carajs BIFs display a marine hydrothermal
signature witnessing anoxic deep water uxes to the locally oxic
shallow photic zone. Partial (ca 40%) oxidation of aqueous Fe(II) into
Fe(III) and the following transformation of iron hydroxide into iron
oxide (hematite) is responsible for the highly positive Fe isotope
values measured in these rocks, namely 57FeIRMM-14 = 2.06 0.34.
The terrigenous sandstones from the Aguas Claras Formation
contain early diagenetic pyrite. The lack of a distinct MIF-S signature
and a typical post-GOE 33S/ 36S array provide some independent
support for the previously proposed depositional age of 2.1 Ga. This is
also considered as indirect evidence for the Palaeoproterozoic age of
the overlying giant Carajas Mn deposits.

136

S. Fabre et al. / Chemical Geology 289 (2011) 124139

Fe & S
(Aguas Claras, -2.06 Ga)

Continental iron (oxides)


Imputs by rivers to the oceans
3+
Fe [57Fe ~ 0]

Sea level

SO4

3+

Fe
2-

2-

2+

Fe

FeS

FeS2 (pyrite)

a
Fe & S
(Present-day ocean)

57Fe > 0
34S > 0

Continental iron (oxides)


Imputs by rivers to the oceans
3+
Fe [57Fe ~ 0]

Sea level
23+

Fe
2-

SO4
B.S.R.
2-

2+

Fe

FeS

FeS2 (pyrite)

1. Fe species considered in the simulations (Data are from thermo.com.


V8.R6.230 prepared by Jim Johnson at Lawrence Livermore National
Laboratory)

2-

SO4
B.S.R.

SO4

Appendix 1. Data used for the numerical modeling

57Fe ~ 0
34S << 0

Fe2+ species

Fe3+ species

Fe+ 2 = Fe+ 2
log k = 0.0
Fe+ 2 + H2O = FeOH+ + H+
log k = 9.5
Fe+ 2 + Cl = FeCl+
log k = 0.14
Fe+ 2 + CO3 2 = FeCO3
log k = 4.38
Fe+ 2 + HCO3 = FeHCO3+
log k = 2.0
Fe+ 2 + SO42 = FeSO4
log k = 2.25
Fe+ 2 + HSO4 = FeHSO4+
log k = 1.08
Fe+ 2 + 2HS = Fe(HS)2
log k = 8.95
Fe+ 2 + 3HS = Fe(HS)3
log k = 10.987
Fe+ 2 + HPO42 = FeHPO4
log k = 3.6
Fe+ 2 + H2PO4 = FeH2PO4+
log k = 2.7
Fe+ 2 + F = FeF+
log k = 1.0

Fe+ 3 = Fe+ 3
log k = 0.0
Fe+ 3 + H2O = FeOH2+ + H+
log k = 2.19
Fe+ 3 + 2 H2O = Fe(OH)2+ + 2 H+
log k = 5.67
Fe+ 3 + 3 H2O = Fe(OH)3 + 3 H+
log k = 12.56
Fe+ 3 + 4 H2O = Fe(OH)4 + 4 H+
log k = 21.6
2 Fe+ 3 + 2 H2O = Fe2(OH)2+ 4 + 2 H+
log k = 2.95
3 Fe+ 3 + 4 H2O = Fe3(OH)4+ 5 + 4 H+
log k = 6.
Fe+ 3 + Cl = FeCl+ 2
log k = 1.48
Fe+ 3 + 2 Cl = FeCl2+
log k = 2.13
Fe+ 3 + 3 Cl = FeCl3
log k = 1.13
Fe+ 3 + SO4 2 = FeSO4+
log k = 4.04
Fe+ 3 + HSO4 = FeHSO4+ 2
log k = 2.48
Fe+ 3 + 2 SO4 2 = Fe(SO4)2
log k = 5.38
Fe+ 3 + HPO4 2 = FeHPO4+
log k = 5.43
Fe+ 3 + H2PO4 = FeH2PO4+ 2
log k = 5.43
Fe+ 3 + F = FeF+ 2
log k = 6.2
Fe+ 3 + 2 F = FeF2+
log k = 10.8
Fe+ 3 + 3 F = FeF3
log k = 14.0

Fig. 9. Schematic model for iron and sulfur cycling in the ocean during late Archaean to
Palaeoproterozoic times (a) compared to present day situation (b); arrow sizes are
proportional to the uxes.

Aguas Claras pyrites are characterized by positive 57FeIRMM-14


(1.02 0.2) and positive 34S (11.97 0.12). These sulfur isotope
values differ from the strongly negative ones frequently observed for
modern marine sediments, where pyrite is formed via bacterial sulfate
reduction. We consider the positive 34S values of sedimentary pyrite
in the Aguas Claras Formation as a consequence of BSR without any S
isotope fractionation, because of the limited amount of sulfate in these
marine sediments deposited in the immediate aftermath of the Great
Oxidation Event. In contrast, only a small part of the reactive Fe
reservoir was used, leading to the expected positive Fe isotope ratios
in the pyrites. This is distinctly different from the case of modern
diagenetic pyrites, where the size of the reactive Fe pool (and not the
sulfate supply) is the likely limiting factor for the growth of
sedimentary pyrite. This is also different from the case of Archaean
diagenetic pyrites, that were likely produced by different processes.
Acknowledgments
Field work and iron isotope analyses were supported by the French
ECLIPSE-INSU-CNRS program (Proj. No. 43, P.I.: A. Ndlec) and the
BrazilianFrench CAPES-COFECUB program (Proj. No. 442/04/06) of R.
Trindade and A. Ndlec. The heads and staff of CVRD N4 and Azul
mines are warmly thanked for having given access to the quarries and
core samples. Sulfur isotopes analysis was performed at Laboratoire
de Gochimie des Isotopes Stables, Institut de physique du globe de
Paris Sorbonne. Paris Cit, Universit Paris Diderot, UMR CNRS 7154,
1 rue Jussieu, 75238 Paris cedex 05, France.
Fruitful discussions and technical assistance by T. Aigouy, C.
Boucayrand, C. Cavar-Hester, G. Berger, A. Fugmann, E. Henry, F. Hubert,
J.F. Mena, J. M. Montel, F. and P. de Parseval are also acknowledged. A
constructive review of a former version by O. Rouxel and two anonymous
reviews of the latter version greatly improved the present manuscript.

2. Archaean anoxic seawater, data after Nordstrom et al. (1979)


equilibrated with pCO2 = 0.1 bars and calcite (which stands for
carbonates sediments)
Description of the seawater solution
pH = 6.55
pe = 4.0
Activity of water = 0.981
Ionic strength = 0.65
Total alkalinity (eq/kg) = 1.780e 3
Total CO2 (mol/kg) = 2.407e 3
Temperature comprised between 0 and 25 C
Solution composition

Elements (oxidation
number)

Molality (mol/
kgw)

Elements (oxidation
number)

Molality (mol/
kgw)

C(4)
Cl

2.407e 3
5.658e 1

K
Mn
N(5)
S(6)
U

1.058e 2
3.773e 9
4.849e 6
2.927e 2
1.437e 8

Ca
Fe total (2 before
oxidation)
Mg
N( 3)
Na
Si

1.066e 2
1.03e 3 and
1.05e 5
5.509e 2
1.724e 6
4.854e 1
7.384e 5

S. Fabre et al. / Chemical Geology 289 (2011) 124139

Iron species in seawater before oxidation

Case [Fe] = 5.58 ppm or 0.1 mmol/kgw


Fe species

Molality

Fe(2) sum
Fe+ 2
FeCl+
FeSO4
FeHCO3+
FeCO3
FeOH+
FeHSO4+

1.035e 4
7.888e 5
1.598e 5
6.269e 6
2.386e 6
3.238e 8
4.863e 9
1.529e 11

Case [Fe] = 55.8 ppm or 1 mmol/kgw

Activity

Fe species

Molality

Activity

2.533e 5
1.203e 5
7.285e 6
1.796e 6
3.763e 8
3.661e 9
1.151e 11

Fe(2) sum
Fe+ 2
FeCl+
FeSO4
FeHCO3+
FeCO3
FeOH+
FeHSO4+

1.036e 3
7.891e 4
1.599e 4
6.259e 5
2.367e 5
3.212e 7
4.867e 8
1.527e 10

2.535e 4
1.204e 4
7.275e 5
1.7826e 5
3.734e 7
3.665e 8
1.150e 10

References
Alibo, D.S., Nozaki, Y., 1998. Rare earth elements in seawater: particle association, shale
normalisation, and Ce oxidation. Geochim. Cosmochim. Acta 63, 363372.
Anbar, A., Duan, Yun, Lyons, T.W., Arnold, G.L., Kendall, B., Creaser, R.A., Kaufman, A.J.,
Gordon, G.W., Scott, C., Garvin, J., Buick, R., 2007. A Whiff of oxygen before the Great
Oxidation Event? Science 317, 19031906.
Appelo, C.A.J., Parkhurst, D.L., 1999. User's guide to PHREEQC (version 2) A computer
program for speciation, batch-reaction, one-dimensional transport, and inverse
calculations. Water-Ressources Investigations Rep. 994259, U.S. Geological
Survey, Denver, Colorado.
Arajo, O.J.B., Maia, R.G.N., 1991. Programma de levantamentos geolgicos bsicos do
Brasil. Programa grande Carajs. Serra dos Carajs folha SB. 22-Z-A. Estado do Par.
Texto. DNPM/CPRM, Brasilia, p. 164.
Arajo, O.J.B., Maia, R.G.N., Jorge Joo, X.S., Costa, J.B.S., 1988. A mega estruturao
arqueana da Folha Serra dos Carajs. VII Congresso Latino-Americano de Geologia,
pp. 324333.
Arajo, O.J.B., Costa, J.B.S., Pinheiro, R.V.L., Maia, R.G.N., Macambira, E.M.B., Vale, A.G.,
Sequiera, J.B., Jorge Joo, X.S., 1992. Arcabouo estructural do complexe de bacias
transcorrente arqueanas da regio da Serra dos Carajs. 37e 592 Cong. Bras. Geol.
So Paulo, pp. 547549. 25/39.
Bau, M., 1999. Scavenging of dissolved Yttrium and Rare Earths by precipitating iron
oxyhydroxyde: experimental evidence for Ce oxidation, Y-Ho fractionation, and
lanthanide tetrad effect. Geochim. Cosmochim. Acta 63, 6777.
Bau, M., Dulski, P., 1996. Distribution of yttrium and rare-earth elements in the Penge
and Kuruman iron-formations, Transvaal Supergroup, South Africa. Precambrian
Res. 79, 3755.
Bekker, A., Holland, H.D., Wang, P.L., RumbleIII, D., Stein, H.J., Hannah, J.L., Coetzee, L.L.,
Beukes, N.J., 2004. Dating the rise of atmospheric oxygen. Nature 427, 117120.
Bekker, A., Slack, J.F., Planavsky, N., Krapez, B., Hofmann, A., Konhauser, K.O., Rouxel, O.J.,
Wing, B.A., 2010. Iron formation: the sedimentary product of a complex interplay
among mantle, tectonic, and biospheric processes. Econ. Geol. 105, 467508.
Bennett, S.A., Rouxel, O.J., Schmidt, K., Garbe-Schnberg, D., Statham, P.J., German, C.R.,
2009. Iron isotope fractionation in a buoyant hydrothermal plume from the MidAtlantic Ridge at 5S. Geochim. Cosmochim. Acta 73, 56195634.
Benning, L.G., Winkin, R.T., Barnes, H.L., 2000. Reaction pathways in the FeS system
below 100 C. Chem. Geol. 167, 2551.
Bergquist, B.A., Boyle, E.A., 2006. Iron isotopes in the Amazon River system: weathering
and transport signatures. Earth Planet. Sci. Lett. 248, 5468.
Berner, R.A., 1983. Sedimentary pyrite formation: an update. Geochem. Cosmochim.
Acta 48, 605615.
Berner, R.A., 2001. Modeling atmospheric O2 over Phanerozoic time. Geochim.
Cosmochim. Acta 65, 685694.
Berner, E.K., Berner, R.A., 1987. The Global Water Cycle: Geochemistry of the
Environment. Prentice Hall, Englewoods Cliffs. 397 pp.
Berner, R.A., Caneld, D., 1989. A model for atmospheric oxygen over Phanerozoic time.
Am. J. Sci. 289, 333361.
Beukes, N.J., Klein, C., 1992. Models for iron formation deposition. In: Schopf, J.W., Klein,
C. (Eds.), The Proterozoic Biosphere. Cambridge University Press, Cambridge,
pp. 147151.
Beukes, N.J., Dorland, H., Gutzmer, J., Nedachi, M., Ohmoto, H., 2002. Tropical laterities,
life on land, and the history of atmospheric oxygen. Geology 30, 491494.
Beukes, N.J., Burger, A.M., Gutzmer, J., 2003. The geology and genesis of high-grade
hematite ore deposits. Appl. Earth Sci. 112, 1825.
Bjerrum, C.J., Caneld, D.E., 2002. Ocean productivity before 619 about 1.9 Gyr ago
limited by phosphorus adsorption onto iron oxides. Nature 417, 159162.
Blanchard, M., Poitrasson, F., Mheut, M., Lazzeri, M., Mauri, F., Balan, E., 2009. Iron
isotope fractionation between pyrite (FeS2), hematite (Fe2O3) and siderite
(FeCO3): a rst-principle density functional theory study. Geochim. Cosmochim.
Acta 73, 65656578.
Bolhar, R., Kamber, B.S., Moorbath, S., Fedo, C.M., Whitehouse, M.J., 2004. Characterisation of early Archean chemical sediments by trace elements signatures. Earth
Planet. Sci. Lett. 222, 4360.
Bolhar, R., van Kranendonk, M.J., Kamber, B.S., 2005. A trace element study of siderite
jasper banded iron formation in the 3.45 Warrawoona Group, Pilbara CratonFormation from hydrothermal uids and shallow water. Precambrian Res. 137,
93114.

137

Braterman, P.S., Cairnsmith, A.G., 1987. Photoprecipitation and the banded iron
formations: some quantitative aspects. Orig. Life Evol. Biosph. 17, 221228.
Braterman, P.S., Cairnsmith, A.G., Sloper, R.W., 1983. Photo-oxydation of hydrated Fe2+
signicance for banded iron formations. Nature 303, 163164.
Butler, I.B., Archer, C., Vance, D., Oldroyd, A., Rickard, D., 2005. Fe isotope fractionation
on FeS formation in ambient aqueous solution. Earth Planet. Sci. Lett. 236,
430442.
Caneld, D.E., 2001. Biogeochemistry of sulphur isotopes. In: Valley, J.W., Cole, D.R.
(Eds.), Stable Isotope Geochemistry, pp. 607633.
Caneld, D.E., 2005. The early history of atmospheric oxygen. Annu. Rev. Earth Planet.
Sci. 33, 136.
Caneld, D., Raiswell, R., 1999. The evolution of the sulfur cycle. Am. J. Sci. 299, 697723.
Caneld, D., Raiswell, R., Westrich, J., Reaves, C., Berner, R., 1986. The use of chromium
reduction in the analysis of inorganic sulphur in sediments and shales. Chem. Geol.
54, 149155.
Cloud, P.E., 1968. Atmospheric and hydrospheric evolution on primitive Earth. Science
160, 729736.
Dauphas, N., van Zuilen, M., Wahwa, M., Davis, A.M., Marty, B., Janney, P.E., 2004. Clues
from Fe isotope variations on the origin of early Archean BIFs from Greenland.
Science 306, 20772080.
Dauphas, N., Cates, N.L., Mojzsis, S.J., Busigny, V., 2007. Identication of chemical
sedimentary protoliths using iron isotopes in the N3750 Ma Nuvvuagittuq
supracrustal belt, Canada. Earth Planet. Sci. Lett. 254, 272287.
DOCEGEO (Rio Doce Geologia e Minerao S.A.) report, 1988. Provncia Mineral de
Carajs. Litostratigraa e principais depsitos minerais. 35 Congresso Brasileiro de
Geologia, vol. 165.
Eldereld, H., Greaves, M.J., 1981. Rare-earth element geochemistry of oceanic
ferromanganese nodules and associated sediments. Geochim. Cosmochim. Acta
45, 513528.
Escoube, R., Rouxel, O., Sholkovitz, E., Donard, O.F.X., 2009. Iron isotope systematics in
estuaries: the case of North River, Massachusetts (USA). Geochim. Cosmochim.
Acta 73, 40454059.
Fantle, M.S., DePaolo, D.J., 2004. Iron isotopic fractionation during continental
weathering. Earth Planet. Sci. Lett. 228, 547562.
Farquhar, J., Bao, H.M., Thiemens, M., 2000. Atmospheric inuence of Earth's earliest
sulphur cycle. Science 289, 756758.
Farquhar, J., Peters, M., Johnston, D.T., Strauss, H., Masterson, A., Wiechert, U., Kaufman,
A.J., 2007. Isotopic evidence for Mesoarchean anoxia and changing atmospheric
sulphur chemistry. Nature 449, 706709.
Frei, R., Dahl, P.S., Duke, E.F., Frei, K.M., Hansen, T.R., Frandsson, M.M., Jensen, L.A., 2008.
Trace element and isotopic characterization of Neoarchean and Paleoproterozoic
iron formations in the Black Hills (South Dakota, USA): assessment of chemical
change during 2.91.9 Ga deposition bracketing the 2.42.2 Ga rst rise of
atmospheric oxygen. Precambrian Res. 162, 441474.
Frimmel, H.E., 2005. Archaean atmospheric evolution: evidence from the Witwatersrand gold elds, South Africa. Earth Sci. Rev. 70, 146.
Frost, C.D., von Blanckenburg, F., Schoenberg, R., Frost, B.R., Swapp, S.M., 2007.
Preservation of Fe isotope heterogeneities during diagenesis and metamorphism
on banded iron formation. Contrib. Mineral. Petrol. 153, 211235.
Gauthier-Lafaye, F., Weber, F., 2003. Natural nuclear ssion reactors: time constraints
for occurrence, and their relation to uranium and manganese deposits and to the
evolution of the atmosphere. Precambrian Res. 120, 81100.
German, C.R., VonDamm, K.L., 2003. Hydrothermal processes. In: Eldereld, H. (Ed.),
Treatise on Geochemistry (V). Elsevier, pp. 181222.
Grainger, C.J., Groves, D.I., Costa, C.H.C., 2002. The epigenetic sediment-hosted Serra
Pelada Au-PGE deposits and its potential genetic association with Fe-oxide CuAu
mineralisation within the Carajas Mineral Province, Amazon Craton, Brazil. Soc.
Econ. Geol. Spec. Publ. 127, 1941.
Grainger, C.J., Groves, D.I., Tallarico, F.H.B., Fletcher, I.R., 2008. Metallogenesis of the
Carajs Mineral Province, Southern Amazon Craton, Brazil: varying styles of
Archaean through Paleoproterozoic to Neoproterozoic base- and precious-metal
mineralisation. Ore Geol. Rev. 33, 451489.
Grassineau, N., Nisbet, E.G., Biskle, M.J., Fowler, C.M., Lowry, D., Mattey, D.P., Abell, P.,
Martin, A., 2001. Antiquity of the biological sulphur cycle: evidence from sulphur
and carbon isotopes in 2700 million-year-old rocks of the Belingwe Belt, Zinbabwe.
Proc. R. Soc. Lond. B 268, 113119.
Guilbaud, R., Butler, I.B., Ellam, R.M., 2011. Abiotic pyrite formation produces a large Fe
isotope fractionation. Science 24, 15481551.
Guo, Q., Strauss, H., Kaufman, A.J., Schrder, S., Gutzmer, J., Wing, B., Baker, M.A., Bekker,
A., Kim, S.-T., Jin, Q., Farquhar, J., 2009. Reconstructing Earth's surface oxidation
across the ArcheanProterozoic transition. Geology 37, 399402.
Habicht, K.S., Caneld, D.E., 1996. Sulphur isotope fractionation in modern microbial
mats and the evolution of the sulphur cycle. Nature 382, 342343.
Halverson, G.P., Poitrasson, F., Hoffman, P.F., Ndlec, A., Montel, J.M., Kirby, J., 2011. Fe
isotope and trace element geochemistry of the Neoproterozoic syn-glacial Rapitan
iron formation. Earth Planet. Sci. Lett. 309, 100112.
Herron, M.M., 1988. Geochemical classication of terrigenous sands and shales from
core or log data. J. Sed. Petrol. 58, 820829.
Hess, H.H., Poldevaart, A., 1968. Basalts: The Poldevaart Treatise on Rocks of Basaltic
Composition. Interscience Publisher, New York.
Hoffman, A.W., 1988. Chemical differentiation of the Earth: the relationships between
mantle, continental crust and oceanic crust. Earth Planet. Sci. Lett. 90, 297314.
Holland, H.D., 2005. Sedimentary mineral deposits and the evolution of Earth Near
surface environments. Econ. Geol. 100, 14891509.
Holland, H.D., 2006. The oxygenation of the atmosphere and oceans. Philos. Trans. R.
Soc. B 361, 903915.

138

S. Fabre et al. / Chemical Geology 289 (2011) 124139

Hulston, J., Thode, H., 1965. Variations in the 33S, 34S and 36710 S contents of
meteorites and their relation to chemical and nuclear effects. J. Geophys. Res. 70,
34753484.
Johnson, C.M., Beard, B.L., Beukes, N.J., Klein, O'Leary, J.M., 2003. Ancient geochemical
cycling in the earth as inferred from Fe isotope studies of banded iron formation
from the Transvaal craton. Contrib. Mineral. Petrol. 144, 523547.
Johnson, C.M., Roden, E.E., Welch, S.A., Beard, B.L., 2005. Experimental constraints on Fe
isotope from fractionation during magnetite and Fe carbonate formation coupled to
dissimilatory hydrous ferric oxide reduction. Geochim. Cosmochim. Acta 69,
963993.
Johnson, C.M., Beard, B.L., Klein, C., Beukes, N.J., Roden, E.E., 2008a. Iron isotopes
constrain biologic and abiologic processes in banded iron formation genesis.
Geochim. Cosmochim. Acta 72, 151169.
Johnson, C.M., Beard, B.L., Roden, E.E., 2008b. The iron isotope ngerprints of redox and
biogeochemical cycling in modern and ancient Earth. Annu. Rev. Earth Planet. Sci.
36, 457493.
Kamber, B.S., Webb, G.E., 2001. The geochemistry of late Archaean microbial carbonate:
implications for ocean chemistry and continental erosion. Geochim. Cosmochim.
Acta 65, 25092525.
Kamber, B.S., Bolhar, R., Webb, G.E., 2004. Geochemistry of late stromatolites from
Zimbabwe: evidence for microbial life in restricted epicontinental seas. Precambrian Res. 132, 379399.
Kasting, J.F., 1993. Earth's early atmosphere. Science 259, 920926.
Kasting, J.F., 2001. Earth historythe rise of atmospheric oxygen. Science 293, 819820.
Kirk, J., Ruiz, J., Chesley, J., Titley, S., Walshe, J., 2001. A detrital model for the origin of
gold and suldes in the Wiwatersrand basin based on ReOs isotopes. Geochim.
Cosmochim. Acta 13, 21492159.
Klein, C., Ladeira, E.A., 2002. Petrography and geochemistry of the least altered banded
iron formation of the Archaean Carajs Formation, Northern Brazil. Econ. Geol. 97,
643651.
Lacan, F., Radic, A., Jeandel, C., Poitrasson, F., Sarthou, G., Pradoux, C., Freydier, R., 2008.
Measurement of the isotopic composition of dissolved iron in the open ocean.
Geophys. Res. Lett. 35, L24610. doi:10.1029/2008GL035841.
Lowers, H.A., Breit, G.N., Foster, A.L., Whitney, J., Yount, J., Uddin, M.N., Muneem, A.A.,
2007. Arsenic incorporation into authigenic pyrite, Bengal Basin sediment,
Bangladesh. Geochim. Cosmochim. Acta 71, 26992717.
Mc Donald, I., Boyce, A.J., Butler, I.B., Herrigton, R.J., Polya, D.A., 2005. Mineral Deposit
and Earth Evolution. Geological Society of London, London.
McLennan, S.M., 1989. Rare earth elements in sedimentary rocks: inuence of
provenance and sedimentary processes. In: Lipin, B.R., McKay, G.A. (Eds.),
Geochemistry and Mineralogy of Rare Earth Elements: Rev. Mineral., 21,
pp. 169200.
Mel'Nik, Y.P., 1982. Precambrian banded iron-formations, physicochemical conditions of formations. Developments in Precambrian Geology, 5. Elsevier,
Amsterdam.
Mojzsis, S.J., Coath, C.D., Greenwood, J.P., Mc Keegan, K.D., Harrisson, T.M., 2003. Massindependent isotopes effects in Archean (2.5 to 3.8 Ga) sedimentary sulphides
determined by ion microprobe analysis. Geochim. Cosmochim. Acta 67,
16351658.
Mougeot, R., Respaut, J.P., Briqueu, L., Ledru, P., Milesi, J.P., Lerouge, C., Marcoux, E.,
Huhn, S.B., Macambira, M.J.B., 1996. Isotope geochemistry constraints for Cu, Au
mineralizations and evolution of the Carajs 763 province (Para, Brazil). 39e
Congresso Brasileiro de Geologia, Salvador, Anais, SBG, vol 7, pp. 321324.
Noguiera, A.C.R., Truckenbrodt, W., Costa, J.B.S., Pinheiro, R.V.L., 1994. Analise
faciologica e estructural da formao Aguas Claras, Pre-cambriano da Serra dos
Carajs. IV Symp. Geol. Amaznia. Belm, p. 363.
Nordstrom, D.K., Plummer, L.N., Wigley, T.M.L., Wolery, T.J., Ball, J.W., Jenne, E.A., Basset,
R.L., Crerar, D.A., Florence, T.M., Fritz, B., Hoffman, M., Holdren Jr., G.R., Lafon, G.M.,
Mattigod, S.V., McDuff, R.E., Morel, F., Reddy, M.M., Sposito, G., Thraikill, J., 1979. A
comparison of computerized chemical models for equilibrium calculations in
aqueous species. Chemical Modeling in aqueous systems, speciation, sorption,
solubility and kinetics. American Chemical Society, 857892 pp.
Ohmoto, H., Watanabe, Y., Ikemi, H., 2006. Sulphur isotope evidence for an oxic Archean
atmosphere. Nature 442, 908911.
Ono, S., Fayek, M., Ohmoto, H., 2000. Origin of uraninite in the Elliot lake district, Canada
and the atmospheric oxygen level of 2.3 Ga Earth. Evolution of the Atmosphere,
Ocean, Crust and Biosphere, pp. 6774.
Ono, S., Eigenbrode, J.L., Pavlov, A.A., Kharecha, P., Rumble, D., Kasting, J.F., Freeman,
K.H., 2003. New insights into Archaean sulfur cycle from massi-independent
sulfur records from the Hamersley basin, Australia. Earth Planet. Sci. Lett. 213,
1530.
Ono, S., Shanks, W.C., Rouxel, O., Rumble, D., 2007. S-33 constraints on the seawater
sulfate contribution in modern seaoor hydrothermal vent suldes. Geochim.
Cosmochim. Acta 71, 11701182.
Pavlov, A.A., Kasting, J.F., 2002. Mass-independent fractionation of sulphur isotopes in
Archean sediments: strong evidence for an anoxic Archean atmosphere. Astrobiology 2, 2741.
Pecoits, E., Gingras, M.K., Barley, M.E., Kappler, A., Posth, N.R., Konhauser, K.O., 2009.
Petrography and geochemistry of the Dales Gorge banded iron formation:
paragenetic sequence and implications for palaeo-ocean chemistry. Precambrian
Res. 172, 163187.
Percak-Dennet, E.M., Beard, B.L., Xu, H., Konishi, H., Johnson, C.M., Roden, E.E., 2011.
Iron issotope fractionation during microbial dissimilatory iron oxide reduction in
simulated Archaean seawater. Geobiology 9, 205220.
Pinheiro, R.V.L., Holdsworth, R.E., 1997. Reactivation of Archaean strikeslip fault
systems, Amazon region, Brazil. J. Geol. Soc. 154, 99103.

Planavsky, N., Bekker, A., Rouxel, O.J., Kamber, B., Hofmann, A., Knudsen, A.H., Lyons,
T.W., 2010. Rare Earth element and Yttrium compositions of Archean and
Paleoproterozoic Fe formations revisited: new perspectives on the signicance
and mechanisms of iron formation deposition. Geochim. Cosmochim. Acta 74,
63876405.
Poitrasson, F., Freydier, R., 2005. Heavy iron isotope composition of granites determined
by high resolution MCICP-MS. Chem. Geol. 222, 132147.
Poitrasson, F., Viers, J., Martin, F., Braun, J.J., 2008. Limited iron isotope variations in
recent lateritic soils from Nsimi, Cameroon: implications for the global Fe
geochemical cycle. Chem. Geol. 253, 5463.
Poitrasson, F., Roskosz, M., Corgne, A., 2009. No isotope fractionation between molten
alloys and silicate melt to 2000 C and 7.7 GPa: experimental evidence and
implications for planetary differentiation and accretion. Earth Planet. Sci. Lett. 278,
376385.
Polyakov, V.B., Soultanov, D.M., 2011. New data on equilibrium iron isotope
fractionation among suldes: Constraints on mechanisms of sulde formation in
hydrothermal and igneous systems. Geochim. Cosmochim. Acta 75, 19571974.
Polyakov, V.B., Clayton, R.N., Horita, J., Mineev, S.D., 2007. Equilibrium iron isotope
fractionation factors of minerals: reevaluation from the data of nuclear inelastic
resonant X-ray scattering and Mossbauer spectroscopy. Geochim. Cosmochim. Acta
71, 38333846.
Radic, A., Lacan, F., Murray, J.W., 2011. Iron isotopes in the seawater of the equatorial
Pacic Ocean: new constraints for the oceanic iron cycle. Earth Planet. Sci. Lett. 306,
110.
Reinhard, C., Raiswell, R., Scott, C., Anbar, A.D., Lyons, T.W., 2009. A late archean suldic
sea stimulated by early oxidative weathering of the continents. Science 326,
713716.
Rickard, D., Luther, G.W., 1997. Kinetics of pyrite formation by the oxidation of iron (II)
monosulphide in aqueous solutions between 25 and 125 C: the mechanism.
Geochim. Cosmochim. Acta 61, 135147.
Rouxel, O.J., Bekker, A., Edwards, K.J., 2005. Iron isotope constraints on the Archaean
and Paleoproterozoic Ocean Redox State. Science 307, 10881090.
Rouxel, O., Shanks III, P.F., Bach, W., Edwards, K.J., 2008. Integrated Fe- and S-isotope
study of seaoor hydrothermal vents at East Pacic Rise 910N. Chem. Geol. 252,
214227.
Roy, S., 1997. Genetic diversity of manganese deposition 812 in the terrestrial
geological record. Geol. Soc. Spec. Publ. 119, 527.
Rye, R., Holland, H.D., 1998. Paleosols and the evolution of atmospheric oxygen; a
critical review. Am. J. Sci. 298, 621672.
Saunier, G., Pokrovsky, G.S., Poitrasson, F., submitted for publication. First experimental
detremination of iron isotope fractionation between hematite and aqueous
solution at hydrothermal conditions. Geochim. Cosmochim. Acta.
Schauble, E.A., Rossman, G.R., Taylor, B.E., 2001. Theoretical estimates of equilibrium Fe
isotope fractionations from vibrational spectroscopy. Geochim. Cosmochim. Acta
65, 24872497.
Schrder, S., Bekker, A., Beukes, N.J., Strauss, H., van Niekerk, H.S., 2008. Sulfate
evaporites in the 2.22.1 Gyr shallow-marine Lucknow Formation, South Africa:
evidence for the rise in seawater sulfate content during the Paleoproterozoic
positive carbon isotope excursion. Terra Nova 20, 108117.
Severmann, S., Johnson, C.M., Beard, B.L., McManus, J., 2006. The effect of early
diagenesis on the Fe isotope compositions of porewaters and authigenic minerals
in continental margin sediments. Geochim. Cosmochim. Acta 70, 20062027.
Singer, P.C., Stumm, W., 1970. Acid mine drainagethe rate limiting step. Science 167,
11211123.
Steinhoefel, G., Horn, I., von Blanckenburg, F., 2009. Micro-scale tracing of Fe and Si
signatures in banded-iron formations using femtosecond laser ablation. Geochim.
Cosmochim. Acta 73, 53435360.
Strauss, H., 1997. The isotopic composition of sedimentary sulfur through time.
Palaeogeogr. Palaeoclimatol. Palaeoecol. 132, 97118.
Strauss, H., 1999. Geological evolution from isotope proxy signalssulphur. Chem. Geol.
161, 89101.
Taylor, B.E., 2004. Biogenic and thermogenic sulfate 835 reduction in the Sullivan Pb
ZnAg deposit, British Columbia (Canada): evidence from micro-isotopic analysis
of carbonates and sulphide in bedded ores. Chem. Geol. 204, 215236.
Thomazo, C., Pinti, D., Busigny, V., Ader, M., Hashizume, K., Philippot, P., 2009a.
Biological activity and the Earth's surface evolution: insights from carbon, sulfur,
nitrogen and iron stable isotopes in the rock record. C. R. Palevol. 8, 665678.
Thomazo, C., Ader, M., Farquhar, J., Philippot, P., 2009b. Methanotrophs regulated
atmospheric sulfur isotope anomalies during the Late Archean (Tumbiana
Formation, Western Australia). Earth Planet. Sci. Lett. 279, 6575.
Tolbert, G.E., Tremaine, J.W., Melcher, G.C., Gomes, C.B., 1971. The recently discovered
Serra dos Carajs Iron Deposits, Northern Brazil. Econ. Geol. 66, 985994.
Iron formations: facts and problems. In: Trendall, A.F., Morris, R.C. (Eds.), Developments
in Precambrian geology, 6. Elsevier, Amsterdam.
Trendall, A.F., Basei, M.A.S., Laeter, J.R., Nelson, D.R., 1998. SHRIMP zircon UPb
constraints on the age of the Carajs formation, Gro Par Group, Amazon Craton.
J. S. Am. Earth Sci. 11, 265277.
Valaas-Hyslop, E., Valley, J.W., Johnson, C.M., Beard, B.L., 2008. The effects of
metamorphism on O and Fe isotope compositions in the Biwabiks Iron Formation,
northern Minnesota. Contrib. Mineral. Petrol. 155, 313328.
von Blanckenburg, F., Mamberti, M., Schoenberg, R., Kamber, B.S., Webb, G.E., 2008. The
iron isotope composition of microbial carbonate. Chem. Geol. 249, 113128.
VonDamm, K.L., 1995. Controls on the chemistry and temporal variability of seaoor
hydrothermal uids. In: Humpris, S.E., Zierenberg, S.A., Mullineaux, L.S., Thomson,
R.E. (Eds.), Seaoor Hydrothermal Systems: Physical, Chemical, Biological and
Geological Interactions. American Geophysical Union, pp. 222247.

S. Fabre et al. / Chemical Geology 289 (2011) 124139


Walker, J.C.G., Brimblecombe, P., 1985. Iron and sulfur 860 in the pre-biologic ocean.
Precambrian Res. 28, 205222.
Weyer, S., Schwieters, J.B., 2003. High precision Fe isotopes measurements with high
mass resolution MCICPMS. Int. J. Mass Spectrom. 226, 355368.
Whitehouse, M.J., Fedo, C.M., 2007. Microscale heterogeneity of Fe isotopes in N 3.71 Ga
banded iron formation from the Isua Greenstone Belt. Southwest Greenland.
Geology 35, 719722.

139

Yamaguchi, K.E., Johnson, C.M., Beard, B.L., Beukes, N.J., Gutzmer, J., Ohmoto, H., 2007.
Isotopic evidence for iron mobilization during Paleoproterozoic lateritization of the
Hekpoort paleosol prole from Gaborone, Botswana. Earth Planet. Sci. Lett. 256,
577587.