Isotopic fractionation of nitrogen and carbon in the synthesis of amino acids by microorganisms

$Page 1: Isotopic fractionation of nitrogen and carbon in the synthesis of amino acids by microorganisms$
Available online at www.sciencedirect.com

www.elsevier.com/locate/gca

Geochimica et Cosmochimica Acta 73 (2009) 3819–3848

Isotopic fractionation of nitrogen and carbon in Paleoarcheancherts from Pilbara craton, Western Australia: Origin

of 15N-depleted nitrogen

Daniele L. Pinti a,*, Ko Hashizume b, Akiyo Sugihara b, Marc Massault c,Pascal Philippot d

a GEOTOP and Departement des Sciences de la Terre et de l’Atmosphere, Universite du Quebec a Montreal, Montreal, QC., Canada H2X 3Y7b Department of Earth & Space Sciences, Graduate School of Science, Osaka University, Toyonaka, Osaka 560-0043, Japan

c UMR IDES-8148 (UPS-CNRS), Departement des Sciences de la Terre, Universite Paris XI SUD, Bat. 504, 91405 Orsay Cedex, Franced Equipe Geobiosphere Actuelle et Primitive, Institut de Physique du Globe de Paris, CNRS & Universite Denis-Diderot, Case 89, 4 place Jussieu,

75252 Paris cedex 05, France

Received 15 April 2008; accepted in revised form 17 March 2009; available online 26 March 2009

Abstract

Nitrogen and carbon isotopic compositions, together with mineralogy and trace element geochemistry, were studied in afew kerogen-rich Paleoarchean cherts, a barite and a dolomitic stromatolite belonging to the eastern (Dixon Island Forma-tion) and western (Dresser and Strelley Pool Chert Formations; North Pole Dome and Marble Bar) terranes of Pilbara Cra-ton, Western Australia. The aim of the study was to search for 15N-depleted isotopic signatures, often found in kerogens ofthis period, and explain the origin of these anomalies. Trace elements suggest silica precipitation by hydrothermal fluids as themain process of chert formation with a contamination from volcanoclastic detritus. This is supported by the occurrence ofhydrothermal-derived minerals in the studied samples indicating precipitation temperatures up to 350 �C. Only a dolomiticstromatolite from Strelley Pool shows a superchondritic Y/Ho ratio of 72 and a positive Eu/Eu� anomaly of 1.8, characteristicof chemical precipitates from the Archean seawater. The bulk d13C vs. d15N values measured in the cherts show a roughlypositive co-variation, except for one sample from the North Pole (PI-85-00). The progressive enrichment in 15N and 13C froma pristine source having d13C 6 �36& and d15N 6 �4& is correlated with a progressive depletion in N content and to vari-ations in Ba/La and Co/As ratios. These trends have been interpreted as a progressive hydrothermal alteration of the chertsby metamorphic fluids. Isotopic exchange at 350 �C between NH4

+(rock) and N2(fluid) may explain the isotopic and elemental

composition of N in the studied cherts. However, we need to assume isotopic exchange at 350 �C between carbonate C andgraphite to explain the large 13C enrichment recorded. Only sample PI-85-00 shows a large N loss (90%) with a positive d15Nvalue (+11&), while C (up to 120 ppm and d13C �38&) seems to be unaffected. This pattern has been interpreted as the resultof devolatilization and alteration (oxidation) of graphite by low-temperature fluids. The 15N-13C-depleted pristine source hasd 15N values from �7& to �4& and 40Ar/36Ar ratios from 30,000 to 60,000, compatible with an inorganic mantle N source,although the elemental abundance ratios N/C and 40Ar/C are not exactly the same with the mantle source. The componentalternatively could be explained by elemental fractionation from metabolic activity of chemolithoautotrophs and methano-gens at the proximity to the hydrothermal vents. However, ambiguities between mantle vs organic sources of N subsistand need further experimental work to be fully elucidated.� 2009 Elsevier Ltd. All rights reserved.

0016-7037/$ - see front matter � 2009 Elsevier Ltd. All rights reserved.

doi:10.1016/j.gca.2009.03.014

* Corresponding author. Fax +1 514 987 3635.E-mail address: [email protected] (D.L. Pinti).

mailto:[email protected]

3820 D.L. Pinti et al. / Geochimica et Cosmochimica Acta 73 (2009) 3819–3848

1. INTRODUCTION

Nitrogen compounds preserved in rocks, particularlyammonium (NH4

+), are the product of the mineralization

of organic matter (Honma and Itihara, 1981). The isotopiccomposition of sedimentary N (expressed with the notationd15N = (R/Rair � 1) � 1000 (R = 15N/14N; Rair = 0.00367)should thus reflect the primary isotope signal recorded dur-ing biosynthesis (e.g., Altabet and Francois, 1994).

Several studies have been focused on nitrogen isotopesin Archean rocks to determine if they can preserve the re-cord of ancient metabolic pathways (Sano and Pillinger,1990; Boyd and Philippot, 1998; Beaumont and Robert,1999; Pinti et al., 2001; Jia and Kerrich, 2004a,b; Uenoet al., 2004; Papineau et al., 2005; Van Zuilen et al., 2005;Nishizawa et al., 2007; Pinti et al., 2007).

Beaumont and Robert (1999) showed that nitrogen ex-tracted from Early Archean kerogens has a peculiar isoto-pic composition with d15N values from �6& to 0&,while Proterozoic and Phanerozoic kerogens have d15Nshifted towards 15N-enriched values (from +6& to+20&). The difference is supposed to represent a shift inthe metabolic pathways of nitrogen, from a biosphere dom-inated by N-fixation in an anoxic Archean ocean, to anoxygenated world where nitrifiers and denitrifiers flour-ished, cycling the nitrogen. Pinti et al. (2001) and then Uenoet al. (2004) confirmed the existence of nitrogen with d15Nvalues ranging from �7& to �4& in the Archean hydro-thermal silica veins of the 3.49 Ga Chert–Barite Unit atNorth Pole, Pilbara craton, Western Australia.

Pinti and Hashizume (2001) challenged these interpreta-tions by suggesting that the cherts studied by Beaumontand Robert (1999) are biased samples of Archean biota be-cause most of them are deposition of silica from hydrother-mal fluids. Pinti and Hashizume (2001) and Pinti et al.(2001) related the 15N-depleted values measured in Archeankerogens to inorganic sources of N (N2, NH3) metabolizedby chemosynthetic bacteria living at proximity to thehydrothermal vents. These communities are indeed ableto fix inorganic N, producing large negative isotopic shiftsin the biological residue (d15N from �4& to �10&; e.g.,Conway et al., 1994). Nishizawa et al. (2007) measuredthe isotopic signature of N contained in Archean fluidinclusions from silica veins at North Pole, Pilbara Craton.Isotopic variations were interpreted as a mixing between apaleoseawater component with d15NN2 from �0.7& to�0.2& and a hydrothermal component containing nitrogendepleted in15N (d15NN2 = �3.0 and d15NNH4+ = �12.6&).The biological assimilation of this hydrothermal NH4

+

might explain the 15N-depleted values found in Early Ar-chean kerogens.

Jia and Kerrich (2004a,b) further added to the contro-versy by their reports of d15N values up to +24& in ca.2.7 Ga old kerogens, heavier and more varied than thosemeasured in contemporary cherts of any other ages. Ker-rich et al. (2006) interpreted the high d15N values as a fossilresidual signature of a veneer atmosphere having an initialCI-chondritic composition with d15N values from +30& to+42&. The isotopic shift from +24& to 0& in the presentatmosphere was interpreted as a combination of three pro-

cesses: (1) degassing of 15N-depleted mantle N2 (d15N=�5 ± 2&); (2) progressive sequestration of chondrite-likeN2 atmosphere in sedimentary rocks by fixing organisms;and (3) return flux of 15N-depleted nitrogen to the atmo-sphere as a byproduct of some sort of metabolism. Re-cently, Pinti et al. (2007) observed d15N of +6& in a3.5 Ga chert from Marble Bar, Western Australia and inter-preted the N as biogenic. These varied N isotopic composi-tions suggest the possibility that the full suite of microbialreactions that constitute the modern nitrogen cycle werein operation since the Archean (Buick, 2007).

Question arose from these contrasted results as towhether these old rocks have preserved a pristine, biologi-cally fractionated N isotopic composition and how to dis-cern between biological and abiological sources ofnitrogen and carbon. With this in mind, we analyzed anew selected subset of kerogen-rich Paleoarchean cherts(with ages from 3.2 to 3.5 Ga) from different formationsof Pilbara Craton, Australia, where independent studiessuggested the occurrence of chemical and isotopic signa-tures of an ancient microbial activity. These are the DresserFormation at North Pole (Pinti et al., 2001; Ueno et al.,2004, 2006), the Apex Basalt at the ‘‘Schopf locality”, Chi-naman Creek (Schopf, 1993; Brasier et al., 2002) and theDixon Island Formation (Kiyokawa et al., 2006). We alsoanalyzed a barite, coeval of the chert veins of the DresserFormation at North Pole and a dolomitic stromatolite fromthe Strelley Pool Chert Formation, also from North Pole. Amulti-disciplinary approach including mineralogy, trace ele-ment geochemistry and stable isotope geochemistry wascarried out in order to decipher pristine biogenic N sourcesand fractionation processes.

2. GEOLOGICAL BACKGROUND AND SELECTED

SAMPLES

The Archean Pilbara Craton is composed of the 3.53–3.17 Ga East Pilbara Terrane representing the ancient nu-cleus of the craton; the 3.18 Ga Kurrana Terrane in thesoutheastern part of the craton; and the 3.27–3.11 Ga WestPilbara Terrane which is a collage of three distinct terranes(Van Kranendonk et al., 2007) (Fig. 1). Samples from thisstudy are from the East and West Pilbara terranes (Table1; Figs. 1 and 2). One sample from the West belongs tothe volcano-hydrothermal sequence of Dixon Island For-mation in the Cleaverville Group (Van Kranendonket al., 2007; Fig. 2a). Samples from the East belong to the3.51–3.42 Ga Warrawoona Group and the 3.43–3.32 GaKelly Group (Fig. 2b). Both groups consist dominantly ofvolcanic and minor sedimentary rocks. Famed sedimentaryunits in the Warrawoona group include the 3.49 Ga hydro-thermal chert–barite units of the Dresser Formation thatcontain putative stromatolites and microfossils (Walteret al., 1980; Groves et al., 1981) and the chert unit of theApex Basalt, which contains the controversial microfossilsdescribed by Schopf (1993). These rocks were weakly de-formed under low-grade metamorphic conditions andunconformably overlain by the Kelly Group (Buick et al.,1995; Van Kranendonk et al., 2007). The Kelly Group,from bottom to top, consists of the basal Strelley Pool

Ultrabasic rocks (2600-3500 Ma)

Metasedimentary and aciditic rocks (2600-3500 Ma)

Granitoid complexes (2600-3700 Ma)

BIF and Shale (2400-2700 Ma)

Basalt, dacite, sandstone (2400-2700 Ma)

Hamersley

500 100 km

Pilbara craton

2a

2b

Fig. 1. Simplified geological map of the Pilbara greenstone belt, Western Australia with the highlighted locations of the areas where the chertshave been sampled.

Origin of 15N-depleted nitrogen in Archean rocks 3821

Chert, Euro Basalt, Wyman Formation and Charteris Ba-salt. The Strelley Pool Chert contains putative stromatolitesdeposited in a shallow marine environment (Lowe, 1983;Hofmann et al., 1999; Van Kranendonk et al., 2003). Allthe samples selected for this study have been collected fromterrains that have been affected by low-grade metamorphicconditions (phrenite-pumpellyite, lower greenschist facies;200–350 �C).

2.1. Dixon Island Formation

Sample Pi-01-45 is a black kerogeneous chert belongingto the Black Chert Member of Dixon Island Formation.This member consists of, in ascending order, a �350 m-thick rhyolite tuff, and black and varicolored chert mem-bers (Kiyokawa et al., 2006; Fig. 2a). It has been suggestedthat Dixon Island Formation represents a pelagic hydro-thermal environment at �500–2000 m depth, and may havebeen on the slope of an immature island arc (Kiyokawaet al., 2006). The lower part of the Black Chert Membercontains carbonaceous peloids and spiral-, rod-, and den-drite-shaped bacteriomorphs with d13C values from�27& to �32&. The C isotopic signature suggests thatthe carbonaceous grains and bacteria-shaped material areof biogenic origin. They were possibly formed close to alow-temperature hydrothermal vent system promoting silic-ification and preservation of the organic material (Kiyoka-wa et al., 2006).

2.2. Dresser Formation at North Pole

The 3.49 Ga Dresser Formation at North Pole consistsof chert–barite and pillow basalt units. At the bottom, thebasalts are transected by thousands of silica and bariteveins varying from centimeters to kilometers long. Silicaveins show different colors (black, grey, white) reflectingdifferent abundances of Fe-carbonate, carbonaceous mate-rial, oxide and sulfides. Carbonaceous matter shows d13Corg

values from �30& to �35& (Ueno et al., 2004), while asso-ciated fluid inclusions have d13CCH4 as low as �56& (Uenoet al., 2006). The carbonaceous filaments have been inter-preted as microfossils of methanogens, whereas the associ-ated 13C-depleted inclusion fluids are thought to representmethanogenetic CH4 entrapped during the formation ofthe silica veins (Ueno et al., 2006). The chert–barite unitalso contains the oldest known occurrence of stromatolites(Walter et al., 1980) but debates continue over the biogenic-ity of these structures (Lowe, 1994). Samples Pi-47-00 andPi-85-00 are black kerogeneous cherts sampled in two dis-tinct silica veins, while sample Pi-II-08-00 is from a baritevein (Fig. 2b) of the chert–barite unit.

Three models have been proposed for the setting of thelowermost chert–barite unit of the Dresser Formation. Onemodel suggests bedded carbonate–gypsum evaporitesdeposited in a quiet, shallow marine lagoon separated fromthe open ocean by a sand bar, following analogy with SharkBay, Western Australia (Buick and Barnes, 1984). Silica

Tab

le1

Lo

cati

on

and

stra

tigr

aph

ical

info

rmat

ion

of

sam

ple

dch

erts

.

Sam

ple

Lo

cati

on

Lo

cali

tya

Gro

up

Fo

rmat

ion

Un

itA

ge(M

a)C

oo

rdin

ates

AM

Ggr

id50

Co

ord

inat

esge

ogr

aph

ical

Wes

t-P

ilbara

Gre

enst

one

Bel

t

Pi-

01-4

5-ca

rbo

nac

eou

sb

lack

cher

tD

ixo

nIs

lan

dC

lear

vill

eD

ixo

nIs

lan

dB

lack

Ch

ert

Mem

ber

3200

5070

80E

7719

900N

20�

37’

11.8

”L

at.

117�

4’4.

6”L

on

g.

East

-Pil

bara

Gre

enst

one

Bel

t

PA

NO

-D-1

36-0

bN

ort

hP

ole

Do

me

2.6;

Dre

sser

Min

eW

arra

wo

on

aD

ress

erch

ert–

bar

ite

3490

7543

00E

7662

100N

21�

7’28

.4’’

Lat

.11

9�26

’54

”L

on

g.

Pi-

47-0

0-ca

rbo

nac

eou

sb

lack

cher

tN

ort

hP

ole

Do

me

2.6;

Dre

sser

Min

eW

arra

wo

on

aD

ress

erch

ert–

bar

ite

3490

7530

00E

7659

300N

21�

9’0.

036”

Lat

.11

9�26

’10

.5L

on

g.

Pi-

85-0

0-b

lack

cher

tN

ort

hP

ole

Do

me

2.3;

Bar

ite

Ran

geW

arra

wo

on

aD

ress

erch

ert–

bar

ite

3490

7539

00E

7665

800N

21�

5’28

.4”

Lat

.11

9�26

’38

.2”

Lo

ng.

Pi-

II-0

8-00

-bar

ite

vein

No

rth

Po

leD

om

e2.

6;D

ress

erM

ine

War

raw

oo

na

Dre

sser

cher

t–b

arit

e34

9075

3400

E76

5950

0N21

�8’

53.3

”L

at.

119�

26’

24.2

”L

on

g.

Pi-

02-4

7-st

rom

ato

lite

No

rth

Po

leD

om

e2.

1;T

ren

dal

llo

cali

tyW

arra

wo

on

aP

ano

ram

aS

trel

ley

Po

ol

Ch

ert

3459

——

Pi-

02-0

7-b

recc

iate

dch

ert

Ch

inam

anC

reek

Sch

op

flo

cali

tyW

arra

wo

on

aA

pex

Bas

alt

Ap

exch

ert

3465

7801

47E

7656

011N

21�

10’

32.9

”L

at.

119�

51’

52.7

”L

on

g.

aN

um

ber

ing

of

loca

liti

esco

rres

po

nd

tofi

eld

sto

ps

des

crib

edin

Van

Kra

nen

do

nk

etal

.(2

001)

.R

efer

toth

isp

ub

lica

tio

nfo

ra

det

aile

dge

olo

gica

ld

escr

ipti

on

.b

PA

NO

-D-1

36-0

cher

th

asb

een

stu

die

db

yP

inti

etal

.(2

001)

.


and barite veins are interpreted as the product of fluid cir-culation during much later events. A second model suggeststhat the chert–barite unit was deposited on oceanic crustduring off-axis hydrothermal circulation and that the NorthPole represents a set of thrust-bound ophiolite slices (Kitaj-ima et al., 2001). A third model suggests that the protolithsof the chert–barite unit were deposited during hydrother-mal circulations in a volcanic caldera (Nijman et al.,1998; Van Kranendonk, 2006). In this model, the sil-ica ± barite veins are interpreted as fossil hydrothermalfluid pathways responsible for the deposition of the chert–barite unit.

2.3. Apex chert Unit at Schopf locality

Chert Pi-02-07 belongs to the largely disputed Apexchert Unit, Marble Bar (Fig. 2b), which contains filamen-tous carbon believed to be cyanobacteria microfossils (seeSchopf, 2006 for the most recent review). Debate on thischert has been invigorated by the hypothesis of a simplehydrothermal, mineral origin for the bacterial pseud-omorphs observed by Schopf (1993), as suggested by Bra-sier et al. (2002). This is based: (a) on field observationswhich reveal that the Apex chert ‘microfossils’ occur withinmultiple generations of metalliferous hydrothermal veincherts, some 100 m to a 1.5 km long chert vein complexassociated with major synsedimentary growth faults; (b)controversial Raman spectroscopy carbon analyzes, whichreveal that C could be amorphous carbon reorganized inform of filamentous strains after devitrification processesof the chert veins (Brasier et al., 2002); c) d13C values from�27& to �34&, which can be both related to photosynthe-sis (d13C = �25 ± 10&; Schopf, 2006) or methanogenesis(Brasier et al., 2002). The Pi-02-07 sample is a brecciatedchert from the Apex Basalt Formation composed of mmto cm-scale lithoclasts surrounded with chalcedonic mi-cro-quartz, Al- and K-rich phyllosilicates from hydrother-mally altered feldspars, sulfides and carbonates (Brasieret al., 2002). The sample has been collected at the top ofthe hydrothermal chert breccia vein described by Brasieret al. (2002) as part of the ‘‘Schopf microfossil locality”.The sample contains native metals, barite, jarosite-aluniteand iron oxides (see Section 4), which suggest a direct pre-cipitation from hot (250–350 �C) hydrothermal fluids (Bra-sier et al., 2002). However, recent discovery of low-temperature hydrothermal halloysite [Al2Si2O5(OH)4] sug-gests that this sample may have undergone post-deposi-tional low-T hydrothermal alteration (Mineau et al., 2008).

2.4. Strelley Pool Chert at Trendall locality

Sample Pi-02-47 is a dolomitic stromatolite-like carbon-ate collected at the ‘‘Trendall locality” (Hofmann et al.,1999). The general stratigraphy of the formation consistsof a basal member of siliciclastic rocks (0.2–20 m thick), a10–20 m of dolomitic carbonates displaying stromatolite-like morphologies and extensive areas of crystal fans (nowreplaced by dolomite, but originally probably aragoniteor gypsum) that interrupted bedding, and a siliciclastic sed-imentary member that includes boulder-pebble conglomer-

Snapper Beach Formation

Lagoon Pillow Basalt

Dixon Island Formation

Cleaverville Group

Fortescue Group

Lizard Hills Formation

DixonIsland

1 km

Pi-45-01a

De Grey Group

Dresser Formation (ca. 3490 Ma)Duffer Formation at Marble Bar (ca. 3467 Ma)

Fortescue Group

Granitoid complexes

Fault

Sampled cherts

Pi-85-00

Pi-II-08-00Pi-47-00

Pi-02-07

Pano D-136-0North PoleDome Marble

Bar

10 km

b

Fig. 2. (a) Simplified geological map of the Dixon Island formation with the position of sample Pi-01-45 (modified from Kiyokawa et al.,2006). (b) Geological map of the North Pole Dome and Marble Bar area with location of sampled cherts redrawn and modified from theoriginal map from the Western Australia Geological Survey. ‘‘Trendall locality”, where stromatolite Pi-02-47 has been collected, is notreported. The Western Australia Geological Survey presently protects its position. The location of the sample PANO D 136-0 (Pinti et al.,2001), discussed here has been also reported.


ate, sandstone and tuffaceous siltstone (Hofmann et al.,1999; Van Kranendonk et al., 2003).

3. ANALYTICAL METHODS

Nitrogen and argon were analyzed at different times atOsaka University, by quadrupole mass spectrometry(QMG-420, Balzers�). The amounts of carbon and hydro-gen, reported as CO2 and H2O in Table 3, were quantifiedby a pressure gauge (crystal gauge M-320XG, ANELVA),followed by examination of the gas species by a gas chro-matography mass spectrometer (GC/MS: quadrupole massspectrometer, SUN200-FAM, attached with a GC, 6890-Agilent) connected to the nitrogen extraction line. TheCO2 was released by combusting the sample and trappingit at liquid nitrogen temperature in a cold finger duringthe extraction. Then, the CO2 was released from the cold

trap by increasing the temperature to about �72 �C, sepa-rating it from the residual gas species such as H2O, whichhave higher condensation temperatures. The compositionof the residual gas examined by the GC/MS was confirmedto be mainly in the form of CO2 and H2O. The detectionlimit of the CO2 and H2O pressures by the crystal gaugeis 0.1 Pa.

Nitrogen was extracted in gaseous molecular form usinga stepwise combustion method (Pinti et al., 2001), aimed todecipher the isotopic composition of volatiles trapped indifferent host phases and trapping forms, which often re-lease gases at temperatures unique to the phases/forms (Ta-ble 3). Only sample preparation is described here.Yamamoto et al. (1998) and Pinti et al. (2001, 2007) gavedetailed descriptions of the nitrogen isotopic analyzes andrecent technical improvements of the line at OsakaUniversity.

Table 2Major element, organic chemistry and trace elements of sampled cherts.

Element Pi-01-45 Pi-47-00 Pi-85-00 Pano D-136-0 Pi-02-47 Pi-02-07Rock type Black chert Black chert Black chert Black chert Stromatolite Apex chertLocality Dixon Island North Pole North Pole North Pole North Pole Schopf locality

Major elementsSiO2 95.8 97.3 97.97 98.77 15.98 98.86Al2O3 2.48 0.19 0.18 0.22 — —Fe2O3 — 0.21 0.87 0.14 1.8 0.27MnO — — — — 0.49 —MgO — — — — 16.98 —CaO — 0.21 — — 26.1 —Na2O 0.05 — — — — —K2O 0.63 — — 0.07 — —TiO2 0.14 — — — — —LOI 1.11 0.63 0.77 0.58 38.43 0.33Total 100.21 98.54 99.79 99.78 99.78 99.46Organic chemistryCO2 tot.% 1.04 0.57 0.52 0.64 38.92 —S tot.% 0.03 <0.01 0.51 0.15 <0.01 —N% 0.01 — 0.003 0.013 0.006 —C% 0.31 0.18 0.15 0.2 10.65 —C/Nmolar 36.2 — 58.3 17.9 2071 —d13C,& PDB �30.5 �32.8 �38.2 �36.9 — �18.7Trace elementsAs 7.2 14 9 16.8 10.4 47.2Ba 1109 113 46 472 160 588Bi 0.25 0.21 0.11 0.24 0.19 0.06Co 0.76 5.91 16.6 13.16 4.05 0.45Cr 182.6 11.3 6.9 10.3 5.2 9.2Cu 11.7 8.8 27.6 13.5 — —Ga 2.81 0.46 0.54 0.39 0.21 0.34Ge 0.47 1.07 0.57 1.04 0.15 1.73Hf 0.4 0.1 0.08 0.07 — 0.05Mo — 3.15 — 0.56 0.52 0.52Nb 0.95 0.14 0.18 0.12 — —Ni 8.6 19.3 63 40.7 82.6 6.2Pb — 3.49 — 9.11 — 16.4Rb 18.2 1.2 — 1.7 — —Sb 1.2 — — 1.4 1.1 15.5Sr 11.8 2.4 — 4.9 22.8 13Th 0.62 0.13 0.18 0.16 — —U 0.37 0.05 0.06 0.11 0.13 —V 31.6 7.4 4.6 3.6 1.9 3.3Zn — — 10.4 385.8 660.1 14.5Zr 14.2 4 3.8 2.6 — 1.7REE+YLa 1.859 0.305 0.404 0.247 0.209 0.692Ce 3.762 0.543 0.886 0.556 0.323 1.374Pr 0.457 0.076 0.103 0.07 — 0.183Nd 1.839 0.369 0.388 0.29 0.202 0.714Sm 0.363 0.15 0.088 0.089 0.047 0.099Eu 0.121 0.102 0.04 0.032 0.024 0.046Gd 0.241 0.13 0.082 0.058 0.066 0.086Tb 0.037 0.022 0.015 0.01 — 0.013Dy 0.225 0.113 0.082 0.065 0.051 0.059Y 1.46 0.775 0.443 0.324 1.007 0.434Ho 0.051 0.024 0.017 0.014 0.014 0.017Er 0.144 0.069 0.047 0.038 0.035 0.051Tm 0.024 0.009 0.008 — 0.006 0.008Yb 0.205 0.068 0.049 0.039 0.038 0.055Lu 0.034 0.011 0.009 0.006 0.005 0.01Eu/Eu� (0.67Sm+0.33Tb)MUQ

a 1.616 2.961 1.878 1.698 1.816b 2.097Ce/Ce� (0.5La+0.5Pr)MUQ 0.952 0.832 1.013 0.986 0.723c 0.901


Table 2 (continued)

Element Pi-01-45 Pi-47-00 Pi-85-00 Pano D-136-0 Pi-02-47 Pi-02-07Rock type Black chert Black chert Black chert Black chert Stromatolite Apex chertLocality Dixon Island North Pole North Pole North Pole North Pole Schopf locality

Pr/Pr* (0.5Ce+0.5Nd)MUQ 0.993 0.953 1.004 0.995 — 1.055Y/HOMUQ 28.6 32.3 26.1 23.1 71.9 25.5

a Trace element anomalies normalzed to the average mud from Queensland (MUQ) of Kamber et al. (2004).b The Eu/Eu* for stromatolite was calculated as Eu/[(Gd+Sm)^0.5]MUQ.c The Ce/Ce* anomaly was calculated as Ce/[0.5La+0.5Nd]MUQ.


We carried out the procedure for complete removal ofsuperficial contamination of the sample, which is often ig-nored and integrated in bulk N isotopic measurements.To quantify the amount of contaminating N and to deter-mine the temperature step at which this component is to-tally removed, we loaded and analyzed sample Pi-II-08-00without any preparation, except standard cleaning of thesample with ethanol and acetone in an ultrasonic bath for30 min (Pinti et al., 2001). Combustion was done with 25to 50 �C step resolution (Table 3), revealing a large N-,C-, H2O-rich component released between 300 and450 �C. The extracted N accounts for the 56% of the totalamount. The d15N(300–450�C) weighted value is�6.4 ± 1.0&, accompanied by some radiogenic 40Ar(40Ar/36Ar(300-450�C) of 2082 ± 36), carbon (16% of the totalextracted amount) and a substantial release of water (38%of the total released amount). Following combustion exper-iments on Pi-II-08-00, other samples were treated in differ-ent ways to eliminate the contamination. Sample Pi-85-00was baked at 200 �C for 3 h and the residual gas waspumped out. Sample Pi-47-00/1 and following samples werebaked at 200 �C overnight before the measurements, to re-move the contamination.

The isotopic composition of carbon was measured at theParis XI University and reported in Table 2. To eliminatesurface contamination, rock chunks were treated withHCl at 10% vol. in an ultrasonic bath and then rinsed withdistilled water. The rock was again washed in the ultrasonicbath with NaOH 0.1 N for a few minutes. The rock was re-duced to powder (10–20 lm) with an agate mortar and sub-mitted to an acid–alkaline–acid treatment (AAA). Toeliminate carbonates, the powder was washed for 30 mnat 80 �C with HCl 10% vol. and rinsed with distilled water.Extractable organic matter was eliminated by treating thepowder with NaOH 0.1 N at 80 �C for 30 min and rinsedagain with distilled water. Eventually carbonates precipi-tated and were eliminated by an additional cycle of treat-ment with HCl at 80 �C for 30 mn. After rinsing anddrying, 100–500 mg of powder was finally charged in aquartz tube with copper oxide. The quartz tube was thensealed under vacuum and combusted at 850 �C for30 min. The produced gas phase was then analyzed forthe 13C/12C ratio using a stable isotope mass spectrometerSIRA 10 of VG Instruments�. The obtained data are ex-pressed with the delta notation and normalized to the PeeDee Belemnite international standard. Total uncertaintieson the measurement are typically ±0.2&.

Together with N, Ar and C isotope analyzes, the majortrace elements and organic matter composition for selected

samples were analyzed at the SARM-CRPG of Nancy.Analytical details and uncertainties are reported in Cari-gnan et al. (2001). The samples were analyzed for their min-eral content at the LAMIC (Laboratoire deMicromanipulations, de Microanalyses et de Cryo-observa-tion) of GEOTOP-Montreal, using a HITACHI S-4300SE/N scanning electron microscope with field effect scanning atvariable pressures. Thin sections were slightly polished todetect fragile internal structures or minerals.

4. RESULTS

4.1. Mineralogy

Selected results are reported in Figs. 3a–l. Sample Pi-01-45 contains two K-bearing silicates within a microcrystal-line quartz matrix: abundant euhedral illite (Fig. 3a) withK = 6.5–9.8 wt% and minor subhedral Ba-(Fe)-feldspar(Fig. 3b), with K = 6–13 wt% (7.5 wt% in the Fe-richphases). Accessory minerals are Na-Ca plagioclases, pyrite,Ti-oxides and zircons. Sample Pi-47-00 contains a few euhe-dral to subhedral carbonates, Fe- and Fe-Mn oxides andpyrite within the microcrystalline quartz matrix. AK(3 wt%)-Na–Mn-rich aluminosilicate, possibly Gan-ophyillite ((K,Na)2(Mn,Al,Mg)8(Si,Al)12 O29(OH)7 � (8–9)H2O), has been found (Fig. 3c). Euhedral Ni-As-(Co) sul-fide, possibly Gersdorffite (NiAsS), is also observed(Fig. 3d). Gersdorffite is found in hydrothermal veinsformed at moderate temperatures, which is compatible withthe suggested depositional environment of the North Polesilica veins (Nijman et al., 1998). Within the microcrystal-line quartz matrix of black chert Pi-85-00, several euhedralcrystals of pyrite and chalcopyrite are preserved (Fig. 3e).The sulfides are zoned or surrounded by a Fe–Mg–Al–sili-cate anhedral phase not showing a clear stoichiometriccomposition but which could be clinochlore (Figs. 3e). Cli-nochlore could derive from hydrothermal alteration ofother Mg–Fe–silicates, such as pyroxene. Similar associa-tions were observed in black cherts recovered from theMarble Bar drill core #1 (Orberger et al., 2006a) and areassociated to late pervasive fluids that caused mica precip-itation. Barite vein Pi-II-08-00 is composed of anhedral tosubhedral Ca–Fe–Mg–Mn carbonates (ankerite) and barite(BaSO4) surrounded by a microcrystalline quartz matrix(Fig. 3f). Euhedral Ca–Mg carbonates and pyrite are foundassociated with these two prominent minerals (Fig. 3g).Subhedral illite, containing a few Fe (1.3 wt%) and a K con-tent of 8.4 wt% is also found within the quartz matrix(Fig. 3h). The stromatolite Pi-02-47 is composed exclusively

Table 3aElemental and isotopic composition of N, Ar and C in Pilbara cherts and stromatolites.

Combustionstep �C

Combustiontime (min)

Cppm

H2Oppm

Nppm

d15N(&)

± 36Ar10�12 cm3

±STP/g

40Ar/36Ar ± 40Ar�

10�8 cm3±STP/g

C/Nmolar

West-Pilbara Greenstone Belt-Dixon Island

Pi-01-45 (248.29 mg) - Bulk-rock

300 60 25 — 0.1 �4.7 1.6 579 32 345 13 2.9 0.7 211400 30 54 — 0.2 �5.2 1.9 1440 79 523 22 32.8 2.7 289500 30 100 — 0.4 �2.9 1.8 1591 105 1244 82 151 9 278600 30 142 — 6.0 6.1 2.0 5215 778 2358 373 1076 118 28700 30 150 — 17.7 �4.9 1.6 21637 2187 1886 200 3441 268 10800 30 137 — 65.7 �2.2 2.0 137045 13921 402 28 1463 353 2900 30 117 — 26.6 0.3 2.4 15903 3923 1080 243 1248 261 51000 30 90 — 20.5 1.6 1.6 17546 2218 800 85 885 108 51100 30 54 — 23.0 2.0 1.7 40292 3944 383 25 353 95 31200 30 54 — 6.2 1.5 2.1 14622 1534 327 22 47 31 101200 30 18 — 1.0 8.2 2.2 1027 170 501 58 21.1 5.1 201200 30 — — 0.6 6.2 1.9 687 87 456 41 11.0 2.5 —1200 30 — — 0.6 6.2 1.8 963 102 406 29 10.7 2.6 —R 1200 71 — 8.4 3.0 1.6 17300 1549 347 19 89 32 10Total 941 — 168.6 -0.5 0.9 258547 15410 634 29 8741 548 7R(500–1200 �C) 862 — 168.2 -0.5 0.9 256529 15410 635 29 8705 548 6

Pi-01-45 (38.06 mg) -Bulk-rock

300 30 34 — 0.3 -4.6 1.6 2695 112 308 8 3.2 2.2 146400 30 61 — 0.2 -7.2 1.5 2377 77 436 11 33.3 2.3 379500 30 151 — 0.3 -2.1 1.7 1144 85 1982 156 193 11 530550 30 213 — 0.4 -1.2 1.6 500 59 6812 897 326 26 554600 30 314 — 1.7 2.8 1.6 1163 174 7137 1218 796 83 218650 30 251 — 3.4 4.4 2.0 1035 267 10343 3051 1040 184 87700 30 163 — 6.1 -0.3 1.7 2152 443 6284 1474 1289 187 31750 30 117 — 10.6 0.9 1.5 6620 1057 2398 406 1392 163 13800 30 85 — 12.2 -0.3 1.7 3112 875 4078 1244 1177 226 8850 30 74 — 9.2 1.1 1.7 4016 763 2176 428 755 105 9900. 30 72 — 11.5 1.2 1.5 4206 849 2337 492 859 126 7950 30 56 — 6.8 2.9 1.8 2798 619 1713 380 397 68 101000 30 55 — 7.0 6.8 1.8 3074 649 1289 263 306 54 91100 30 74 — 12.0 6.9 1.7 3603 965 1187 290 321 69 71200 120 98 — 7.8 6.9 1.8 7381 1042 509 51 157 33 151200 24 — — 0.3 7.5 1.8 134 41 822 160 7.1 1.5 —1200 30 — — 0.3 9.1 1.6 79 36 1189 360 7.1 1.3 —1200 120 — — 0.7 7.8 1.6 495 138 572 83 13.7 2.9 —1200 120 — — 0.4 10.7 1.8 159 116 753 314 7.3 1.6 —1200 30 — — 0.1 8.3 1.8 44 28 487 122 0.8 0.3 —R 1200 98 — 9.8 7.3 1.4 8292 1060 529 46 193 33 12Total 1818 — 91.5 2.9 0.5 46788 2520 2236 158 9080 441 23R(500–1200�C) 1723 — 91.1 2.9 0.5 41716 2520 2463 177 9043 441 22

East-Pilbara Greenstone Belt- North Pole, Dresser Formation

Pi-47-00/1 (190.20 mg)- Bulk-rock

300 30 7 0.23 0.01 2.7 2.0 30 9 3975 241 10.9 0.7 845350 30 2 0.17 0.01 1.7 2.3 55 10 2357 106 11.3 0.6 274400 30 2 0.46 0.01 4.5 1.8 54 12 10836 489 56.9 2.6 169450 30 2 2.0 0.03 5.4 1.3 89 19 18212 639 159.8 5.7 95500 30 3 4.2 0.03 6.9 1.3 80 17 19815 738 157.1 5.9 97550 30 6 0.79 0.03 7.4 1.5 41 15 20201 1033 82.4 4.3 262600 30 27 9.2 0.04 5.5 1.5 367 60 20867 572 755 21 724650 30 50 7.1 0.02 2.7 1.4 11 8 57570 5832 62.7 6.0 2612700 30 50 5.0 0.03 5.5 1.2 14 9 37802 3362 53.0 4.6 1749750 30 29 1.4 0.04 6.6 1.1 10 6 37776 4115 35.9 3.7 941800 30 20 8.4 0.03 5.5 1.2 21 10 17689 1290 36.0 2.6 685850 30 19 5.0 0.04 3.9 1.7 41 25 22856 1800 92.5 7.3 585900 30 17 2.3 0.03 3.4 1.3 20 10 20184 1478 40.1 2.9 567950 30 21 2.2 0.04 1.3 1.6 57 25 23512 1593 132 9 548


Table 3a (continued)

Combustionstep �C


Cppm

H2Oppm

Nppm

d15N(&)

± 36Ar10�12 cm3

±STP/g

40Ar/36Ar ± 40Ar�

10�8 cm3±STP/g

C/Nmolar

1000 30 17 0.05 0.05 �1.7 1.5 76 36 22738 1322 171 10 3631050 30 18 0.06 0.10 �0.1 1.2 122 27 26848 1281 324 15 2161100 30 19 0.06 0.14 �6.5 1.3 393 80 30931 1282 1203 50 1571150 30 23 0.04 0.23 �4.5 1.6 1221 260 31907 1170 3861 143 1181200 30 33 0.04 0.47 �3.0 1.3 1781 287 30683 937 5411 166 831200 408 46 — 0.67 0.2 1.4 7486 731 18375 452 13534 319 801200 30 2 — 0.01 4.5 1.8 — — — — — — 144R 1200 81 0.04 1.15 �1.0 1.0 9267 785 20740 1530 18945 360 82Total 411 49 2.07 �0.6 0.6 11970 836 22177 1349 26191 392 232R(500–1200�C) 398 46 2.01 �0.8 0.6 11742 836 22398 1374 25952 392 231

Pi-47-00/2 (135.61 mg) - Bulk-rock

400 60 — — 0.04 5.8 1.8 244 14 959 50 16.2 0.9 —500 30 — — 0.03 4.1 1.6 161 9 6868 451 105.5 4.1 —600 30 — — 0.06 2.1 1.4 401 19 8500 463 329 11 —700 30 — — 0.06 4.3 1.4 133 11 13608 1290 178 10 —800 30 — — 0.15 3.2 1.7 273 30 5172 628 133 10 —900 30 — — 0.17 2.4 1.6 449 37 2935 263 119 7 —1000 30 — — 0.21 1.3 1.4 239 28 7194 940 165 13 —1100 30 — — 0.34 �2.3 1.5 531 55 7980 913 408 28 —1200 120 — — 1.05 �5.8 1.4 1213 132 8670 986 1016 64 —1200 30 — — 0.11 �0.4 1.5 140 22 3952 575 51.4 3.9 —1200 120 — — 0.18 4.3 1.4 145 59 4300 1516 58.2 4.8 —1200 120 — — 0.09 3.1 1.6 73 55 3140 2048 20.6 1.7 —1200 — — 0.02 2.4 1.7 34 14 1395 439 3.7 0.3 —R 1200 — — 1.47 3.5 1.0 1610 157 7459 765 1150 64 —Total — — 2.54 �1.5 0.7 4035 180 6743 342 2602 73 —R(500–1200�C) 2.50 �1.6 0.7 3791 180 7116 364 2586 73 —


of Ca–Mg-carbonates with a few euhedral Fe-oxide crystals(Fig. 3i) and anhedral barite dispersed in the carbonates.Fe-oxide concretions surround the stromatolitic laminae(Fig. 3j). Apex chert Pi-02-07 is mostly composed of micro-crystalline quartz with abundant euhedral crystals of bariteand scarce Ca-plagioclase (Fig. 3l), very rare Ni–Cu–Cometal clothes (Fig. 3k) and jarosite-alunite solid solutions.A few iron oxides, with spherulitic habits or forming lepis-heres partially replaced euhedral barite crystals.

4.2. Major and trace elements

Bulk chemical analyzes (Table 2) reflect the dominanceof the microcrystalline quartz in cherts and of dolomite inthe Strelley Pool stromatolite Pi-02-47. Enrichments inAl2O3 (2.5 wt.%) and K2O (0.6 wt.%) in Pi-01-45 are relatedto the occurrence of abundant illite and some K–Ba-feld-spar (Figs. 3a–b). The slight enrichment of Fe2O3 (1.8wt%) in stromatolite Pi-02-47 compared to black cherts(Fe2O3 = 0.14 to 0.9 wt%) is caused by the iron oxide con-cretions (Figs. 3j), while in black cherts it is related tosparse pyrite (Fig. 3e).

The REE + Y are most useful to trace the origin of thestudied sedimentary rocks. The REE are insoluble and invery low concentrations in aqueous solution, thus theyare chiefly transported as particulate matter and reflectthe chemistry of their source (e.g., Rollinson, 1993; Hol-lings and Wyman, 2005). The effects of diagenesis andmetamorphism are minor and most Archean sedimentshave shown essentially undisturbed REE patterns except

for redox-sensitive elements such as Ce and Eu (e.g., Kam-ber et al., 2004; Bolhar et al. 2005; van den Boorn et al.,2007). Weathering/alteration effects are more difficult toevaluate. The REE are mobilized and fractionated due tothe breakdown of primary phases and the formation of sec-ondary phases, resulting in LREE enrichment in the weath-ered residue and HREE enrichment in the weatheringproduct (Nesbitt et al., 1996). These effects are importantfor understanding whether the measured N and C isotopicsignatures in these rocks are related to weathering processesincorporating exotic fluids rather than pristine biologicalsignatures.

Trace elements are reported in Table 2 together with Ce,Pr, Eu anomalies and Y/Ho, Pr/Sm and Nd/Yb ratios cal-culated after Bolhar et al. (2005) and normalized to theaverage mud from Queensland (MUQ; see Kamber et al.,2005 for details on this normalization). We reported thetrace element composition of PANO D-136-0 from Orber-ger et al. (2006b) with its companion, Pi-47-00. Indeed,these two cherts from the Dresser Formation silica veinsat North Pole share quasi-identical nitrogen release pat-terns (Fig. 8 this study and Fig. 2 in Pinti et al., 2001)and d15N and d13C values (Table 2 in Pinti et al., 2001;Tables 2 and 3; this study).

Fig. 4a shows the trace element normalized diagrams ofsampled cherts, compared to that of hot and cold hydro-thermal fluids (Hongo and Nozaki, 2001). Fig. 4b showsthe trace element normalized diagram of the Strelley Poolstromatolite (Pi-02-47) compared to average values forStrelley Pool stromatolites measured by Van Kranendonk

Fig. 3. Secondary electron (SE) or back-scattering (BSE) microphotographs of selected minerals founds in chert samples (A-L), together withX-ray mapping of Fe for stromatolite Pi-02-47 (J). See text for details.


et al. (2003), modern seawater proxies (Heron Reef micro-bialites; Webb and Kamber, 2000) and seawater (NPDW;Alibo and Nozaki, 1999). Cherts show relatively flat pat-terns, europium positive anomalies (Eu/Eu� >>1) and anabsence of clear La, Gd and Y positive anomalies(Fig. 4a). All these characteristics are typical of cherts pre-cipitated from hot or cold hydrothermal fluids with little orno influence of seawater (Bolhar et al., 2005; van den Boornet al., 2007). Dolomitic stromatolite Pi-02-47 shows a traceelement normalized pattern similar to that of seawater withLREE more depleted than MREE and HREE, a strong Ypositive spike and some La enrichment (Fig. 4b). However,there are some important differences between modern sea-

water or its sedimentary proxy and Archean stromatolites(Fig. 4b). There are no evident Gd anomalies and there isa mild positive Eu anomaly (calculated as Eu/Eu� = Eu/[p

(Gd�Sm)]MUQ = 1.82 because of missing Tb data; Table2). This value is slightly higher than the average value cal-culated for Strelley Pool stromatolites (Eu/Eu� =1.69 ± 0.06 from raw data of Van Kranendonk et al.,2003) but clearly different from that of modern seawater(Eu/Eu� = 1; Alibo and Nozaki, 1999) and seawater proxies(Eu/Eu� = 1.05 using raw data from Webb and Kamber,2000). Finally there is a less pronounced negative Ce anom-aly (calculated as Ce/Ce* = Ce/[(0.5La + 0.5Nd)]MUQ =0.72 because of missing Pr data; Table 2) than seawater

0.0001

0.001

0.01

0.1

La Ce Pr Nd Sm Eu Gd Tb Dy Y Ho Er Tm Yb Lu

Pi-02-47

Strelley Pool

NPDW x 10,000

(a)

(b)

0.001

0.01

0.1

1

10

La Ce Pr Nd Sm Eu Gd Tb Dy Y Ho Er Tm Yb Lu

Pi-01-45Pi-47-00Pi-85-00Pi-02-07Black Chimney White ChimneyPano D-136-0

Sam

ple/

MU

QS

ampl

e/M

UQ

Fig. 4. (a) REE+Y normalized diagrams of Pi-01-45, Pi-47-00, Pi-85-00 and Apex chert (Pi-02-07), normalized with average mud fromQueensland (MUQ; Kamber et al., 2005). The data are compared with REE patterns hot (black smokers) and cold (white smoker)hydrothermal fluids (Hongo and Nozaki, 2001). (b) REE+Y normalized diagrams measured in dolomite-iron bearing stromatolite Pi-02-47,normalized to MUQ, against average REE pattern measured in Strelley Pool stromatolites by Van Kranendonk et al. (2003) (average; n = 10).The REE+Y normalized diagram of Holocene microbialites from Webb and Kamber (2000) is reported as an analogue of modern seawaterproxies (dashed gray area). For sake of clarity, seawater and hydrothermal fluids REE contents are multiplied by a factor 10,000x and 1,000x,respectively.


(0.07; Alibo and Nozaki, 1999) but comparable with thatmeasured in Strelley Pool stromatolite of 0.68 (Van Kra-nendonk et al., 2003).

Figs. 5a and b illustrate the redox-sensitive Eu and Ceanomalies, respectively plotted against the Y/Ho ratios,whereas Fig. 5c shows a bivariate LREE/HREE ratio plot.For comparison, we reported the trace element values mea-sured in different type of cherts depending on their deposi-tional environment as discussed by van den Boorn et al.(2007). Three end-members of primary chert formationare recognized by van den Boorn et al. (2007) when combin-ing isotopic, petrographic and chemical data: (1) silicifica-

tion of volcanoclastic debris by silica-saturated seawaterand relatively weak or no hydrothermal input (S-cherts);and (2 and 3) in sediment-starved shallow or deep environ-ments, interaction between hot Si-rich fluid and seawater indifferent proportions promoting orthochemical depositionof silica (C-cherts).

When the Eu/Eu� ratios are plotted against the Y/Horatios, the data of van den Boorn et al. (2007) form threeclusters of points close to three possible end-members(Fig. 5a): (1) hydrothermal fluids (Eu/Eu� >> 5 and Y/Ho = 32; Mills and Elderfield, 1995); (2) upper crust rocks(Eu/Eu� = 0.95 and chondritic Y/Ho = 26; Taylor and


McLennan, 1981); and (3) Archean seawater represented byStrelley Pool stromatolites (Eu/Eu� = 2.05 and superchon-dritic Y/Ho = 70.3; Van Kranendonk et al., 2003). The Ar-chean seawater proxy end-member has an Eu-anomaly anda Y/Ho ratio slightly higher than those measured in modern

20

30

40

50

60

70

80

0 1 2 3 4 5 6

Y/H

o

Eu/Eu* (0.67Sm+0.33Tb)MUQ

High-T fluids

Archean SW proxy

Crust

C-chertsseawater-dominated

C-chertshydrothermal-dominated

S-chertsreplacement

Pi-02-47

Pi-47-00Pi-02-07

Pi-85-00

Pano D-136-0

Pi-01-45

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0 10 20 30 40 50 60 70 80

Ce/

Ce*

MU

Q

Y/Ho

Pi-02-47

Pi-47-00Pi-02-07

Pi-01-45

Pi-85-00Pano D136-0

Archean SW proxy

Crust

Hydrothermal jasper

Pi-47-00

Pi-02-07

Pi-85-00

Pano D-136-0

Pi-01-45

0.0

0.5

1.0

1.5

2.0

0.0 0.5 1.0 1.5 2.0

Nd/

Yb M

UQ

La/LuMUQ

Pi-02-47

C-chertshydrothermal-dominated

S-chertsreplacement

C-chertsseawater-dominated

(a)

(b)

(c)

seawater (Eu/Eu� = 1 and superchondritic Y/Ho = 54; Ali-bo and Nozaki, 1999) and this has been interpreted asreflecting a more anoxic ocean slightly buffered by hydro-thermal fluids (Van Kranendonk et al., 2003). Most ofour cherts have Y/Ho and Eu/Eu� ratios in-between thecrust and the hydrothermal fluids end-members suggestinga direct precipitation from hot fluids (C-cherts) possiblycontaminated with crustal material (Fig. 5a). Only dolo-mitic stromatolite Pi-02-47 clearly indicates a direct precip-itation from the Archean seawater (Fig. 5a), while DixonIsland chert Pi-01-45 could be considered as a S-chert.The Ce/Ce� vs. Y/Ho ratios (Fig. 5b) show essentially thesame pattern with most of the cherts plotting in the vari-ability field of hydrothermal-precipitated C-cherts. It isnoteworthy that the apparent negative Ce anomalies ob-served in this plot (Ce/Ce� from 0.7 to 0.95; Fig. 5b and Ta-ble 2) are not real (oxidation of Ce3+ to Ce4+ andscavenging in an oxygenated ocean) but they are deter-mined by the overabundance of La and the deficiency inPr, the two elements preceding and following Ce. Bolharet al. (2005) exhaustively discussed of this problem. RealCe negative anomalies are those which are accompaniedby Pr/Pr� higher than 1.1, as in the case of modern seawaterproxies (Hernon Reef; Ce/Ce� = 0.7–0.9 and Pr/Pr� = 1.1–1.2; Webb and Kamber, 2000). Our cherts show Pr/Pr�

ratios of 0.95–1.06, which are within the values normallyobserved in Archean hydrothermal precipitates (Bolharet al., 2005).

Large concentrations of Th (160–620 ppb), Zr (2.6–14 ppm) and Hf (54–400 ppb) (Table 2) and co-variationof Hf vs. Zr (not plotted here) are additional indicationsof the contamination of our cherts by felsic and mafic vol-canoclastic debris (Kamber et al., 2004). The less contami-nated are Pi-47-00, Pi-02-07, Pi-85-00 and PANO D-136-0.The trace element compositions of these cherts plot insidethe variability field of hydrothermal-precipitated C-cherts(Fig. 5a and b). This is supported by independent observa-tion of hydrothermal-derived minerals and native metals(Fig. 3) and by the fact that they have been sampled fromhydrothermal depositional settings (see Section 2 and refer-ences therein). The largest amount of crustal contamination(5x the other cherts) seems to be Dixon Island Pi-01-45(Hf = 398 ppb; Zr = 14 ppm; Table 2), which could be con-sidered a S-Chert (Fig. 5a). This seems to be clearly indi-

Fig. 5. (a) Eu/Eu� ratio calculated as Eu/[(0.67Sm+0.33Tb)]MUQ

against the raw Y/Ho ratio. The three shaded areas labelled ‘‘S-chert” and ‘‘C-chert” have been obtained by plotting REE datafrom van den Boorn et al. (2007). See text for details. Eu/Eu� ratiofor Strelley Pool stromatolite Pi-02-47 was calculated as Eu/[Sm�Gd]0.5 because of Tb was not measured (Table 2). The threeend-members named ‘‘crust”, ‘‘Archean SW proxy” and ‘‘high-Tfluids” have been calculated using data from Rollinson (1993), VanKranendonk et al. (2003) and Mills and Elderfield (1995),respectively. (b) The Ce/Ce� ratio calculated as Ce/[(0.5La+0.5Pr)]-

MUQ against the raw Y/Ho ratio. The shaded area labeled‘‘hydrothermal jasper” has been obtained using data from Bolharet al. (2005). (c) Bivariate plot of two LHREE/HREE ratios (La/Lu and Nd/Yb) normalized to the Queensland mud (MUQ). Thethree shaded areas labelled ‘‘S-chert” and C-chert” have beenobtained by plotting REE data from van den Boorn et al. (2007).

3


cated by the ratios between LREE (Nd and La) and theHREE (Yb and Lu) (Fig. 5c). In Fig. 5c, Dixon IslandPi-01-45 and Apex chert Pi-02-07 are plotted inside the S-chert field. However, in other graphs, Apex chert Pi-02-07is plotted closer to the hydrothermal C-cherts (Fig. 5aand b). The only explanation might be that Apex cherthas been partially affected by post-depositional alteration,exhibiting higher LREE/HREE ratios (Fig. 4a) as those ex-pected for residual weathered rocks (Nesbitt et al., 1996).

4.3. Organic chemistry and d13C

Table 2 reports some data from organic matter extractedfrom the samples and analyzed at CRPG Nancy. The mostimportant result is the C/N ratios for black chert Pi-01-45(36.2), Pi-85-00 (58.3) and PANO D-136-0 (17.9), whichare close to minimum C/N values measured in Archean ker-ogens (Beaumont and Robert, 1999). The d13C values mea-sured at Orsay University range from �30.5& to �38.2&,typical of Early Archean kerogens from Pilbara (Beaumontand Robert, 1999; Ueno et al., 2004; Ueno et al., 2006;Kiyokawa et al., 2006). Apex chert shows a d13C value of�18.7&. This value is clearly less 13C-depleted than typicald13C values reported in literature for the bulk carbon inApex Chert (d13C = �22.5 to �30&; Strauss and Moore,1992).

4.4. Nitrogen and argon isotopes

The N and Ar isotopic composition of cherts is reportedin Tables 3a and 3b, together with the H (calculated asH2O) and C contents and the C/N atomic ratio. The uncer-tainties reported in the tables are 1r. Black cherts Pi-01-45from Dixon Island (Fig. 6) and Pi-47-00 from North Pole(Fig. 8) were analyzed in duplicate in order to check forthe ubiquity of their volatile contents. The C combustionrelease patterns in chert Pi-01-45/1 and Pi-01-45/2 show amain C component released at 600–700 �C followed by avery slight C enrichment at 1100–1200 �C (Fig. 6). Thelow-temperature C is not related to the thermal dissociationof CaCO3, which is produced at 600 �C during stepwisepyrolysis (Boyd et al., 1997) or to decrepitation of fluidinclusions in quartz (575 �C; Sano and Pillinger, 1990). Inboth cases the release of C is produced in a very narrowrange of temperatures (100 �C; Boyd et al., 1997; Kendricket al., 2006). Our C component is released at a broad tem-perature range between 500 �C and 1000 �C (Fig. 6). Fur-thermore, the measured C/N(R500–1000�C) molar ratio of6.3 is incompatible with those observed during thermaldecomposition of carbonates (105; Sano et al., 1994), butcompatible with an organic source (Pinti et al., 2007).

Nitrogen shows two different release patterns for Pi-01-45/1 and Pi-01-45/2 at 1000–1200 �C, while they are essen-tially the same at 700–900 �C (Fig. 6). Between 700 and800 �C, extended to 900 �C for Pi-01-45/2, a large amountof nitrogen is released with d15N(R700–900�C) values as lowas +0.6 ± 0.7& (Pi-45-02/2) or �2.8 ± 1.6& (Pi-45-02/1).At higher temperatures, chert Pi-01-45/1 shows a decreasein the N amount extracted at each step with a contempora-neous increase in the d15N value (Fig. 6), which is character-

istic of a diffusion-controlled fractionation (Boyd et al.,1993; Pinti et al., 2007). Chert Pi-01-45/2 shows a distinctsecond N release at 1000–1200 �C. This component has ad15NR1000–1200�C value heavier (+7.0 ± 1.0&; Table 3) thanthat measured at low temperature (2.0 ± 1.3&). In bothsamples (Fig. 6), the low-temperature N component isaccompanied by a large release of 40Ar�. Radiogenic argonrelease at high-temperature is 5–10 � lower than that re-leased at 700–900 �C (Table 3; Fig. 6). Radiogenic 40Arhas been calculated as the amount in excess of the atmo-spheric value. The non-radiogenic 40Ar/36Ar ratio for theArchean atmosphere should lower than the present valueof 295.5 (Cadogan, 1977), yet the difference will not sub-stantially change our results.

Fig. 7 illustrates co-variations existing between the re-leased amounts of N and 40Ar* and the d15N value in samplePi-01-45/2. Between 800 and 1200 �C, two mineral phasesrelease N and 40Ar* in different amounts and with differentisotopic signatures. The first mineral phase (40Ar*/N�0.0006; d15N approx.�2&) is likely the abundant illite ob-served in the sample (Fig. 3a). Boyd et al. (1993) showed thatammonium (NH4

+) in phyllosilicates is released at 900 �Cby pyrolysis. In the case of combustion, the release temper-ature is probably lower (Robert and Halbout, 1990). Indeed,N combustion pattern of mica from the Kittys Gap Chert atPilbara Craton shows a mid-temperature release patternpractically identical to that of Pi-01-45 (Hashizume et al.,2006). The second mineral phase (40Ar*/N <0.0001; d15N� +7&), could be K-feldspars, observed in minor amountsin Pi-01-45 (Fig. 3) and which should release N at tempera-tures between 1100 and 1200 �C (Boyd et al., 1993). Hashiz-ume et al. (2006) observed the same mixing relationship inmicas of 3.5 Ga Kittys Gap chert, Panorama Formation,Pilbara Craton except that the measured 40Ar*/N molar ra-tios were one order of magnitude higher than in Pi-01-45(40Ar*/N from 0.003 to 0.025; d15N from 0& to +7 &).Hashizume et al. (2006) interpreted the 40Ar*/N co-varia-tions as resulting from a different capacity of retention ofthe electrically-charged NH4

+ and neutrally-charged noblegas 40Ar. Ammonium will be strongly adsorbed in-betweennegatively charged layers of phyllosilicates, while 40Ar* willbe preferentially lost since it is no longer a cation. Differ-ences in the bounding capacity of NH4

+ and 40Ar in differentmineral phases (mica and feldspars) could explain the co-variations illustrated in Fig. 7 (Honma and Itihara, 1981).However it is still unclear if ammonium released at lowand high temperatures, showing two distinct d15N signa-tures, comes from two different sources or if this is the resultof a ‘‘mineralogical” control on the isotopic fractionation ofnitrogen (Boyd et al., 1993).

Fig. 8 shows the release pattern for N, Ar, C and H ofblack chert Pi-47-00 from North Pole silica veins in two re-peated analyzes carried out at different periods. It is worthto note that the N release patterns of these samples arestrikingly similar to that of PANO D-136-0, another blackchert analyzed by Pinti et al. (2001) and collected from oneof the numerous North Pole silica veins cross-cutting theDresser Formation (Fig. 2b). The release patterns of Nand Ar for the two samples are similar: (i) N is mainly re-leased at very high temperature (1200 �C) with a decrease

Table 3bElemental and isotopic composition of N, Ar and C in Pilbara cherts and stromatolites.

Combustionstep �C


Cppm

H2Oppm

Nppm

d15N(&)

± 36Ar10�12 cm3

±STP/g

40Ar/36Ar ± 40Ar�

10�8 cm3±

STP/gC/Nmolar

East-Pilbara Greenstone Belt- North Pole, Dresser Formation

Pi-85-00 (327.80 mg)

300 60 0.038 4.4 1.6 55 1 699 25 2.2 0.1 9400 30 3 8.5 0.182 2.7 1.9 72 1 1759 63 10.6 0.5 17450 30 9 0.026 12.5 1.9 27 3 3218 329 7.9 0.5 400500 30 75 7.0 0.050 11.2 1.6 27 3 3971 410 9.8 0.6 1760550 30 35 48.5 0.049 11.2 1.6 22 4 5645 1137 11.6 0.9 825600 30 43 12.1 0.018 13.0 2.2 70 2 3091 178 19.7 1.1 2884650 30 107 6.6 — — — — — — — — — —700 30 222 4.1 0.018 14.2 2.3 12 2 4649 976 5.2 0.6 14036750 30 265 10.9 0.030 7.4 1.8 10 2 3847 850 3.6 0.3 10337800 30 174 0.27 0.021 14.9 2.1 7 2 4718 1385 3.3 0.3 9870850 30 76 3.2 0.031 6.2 1.6 8 2 4919 1283 3.8 0.4 2846900 30 72 2.0 0.029 10.1 1.7 11 2 5145 1022 5.3 0.4 2899950 30 60 1.0 0.006 7.6 1.8 9 2 7921 2009 6.5 0.6 116591000 30 40 1.6 0.017 11.1 1.2 11 1 6744 662 7.2 0.5 27161100 30 41 1.6 0.019 14.6 1.2 25 2 6070 548 14.6 0.7 24871200 30 30 1.1 0.028 19.1 1.4 46 5 6405 741 28.2 1.0 12441200 180 47 2.0 0.030 10.6 1.9 106 32 8664 2527 88.7 1.9 18141200 30 17 0.61 0.002 8.1 3.9 10 5 10427 5600 9.8 0.6 79841200 290 21 0.61 0.014 2.7 4.5 28 51 — — 62.9 1.5 1681R 1200 115 4.3 0.075 12.2 1.3 189 61 10302 6329 189.6 2.7 1777Total 1337 111.7 0.609 8.3 0.7 557 62 5705 2157 300.9 3.5 2560R(500–1200�C)) 1325 103.3 0.363 11.2 0.5 402 62 7277 2990 280.2 3.4 4252

Pi-II-08-00 (309.70 mg)

300 30 18 7.4 0.359 �6.3 1.7 60.3 0.1 2296 92 12.1 0.6 57350 30 18 8.7 0.212 �9.5 1.4 75.4 0.1 2272 81 14.9 0.6 97400 30 7 7.1 0.085 �4.1 1.4 98.0 0.1 2066 65 17.4 0.6 102450 30 8 5.8 0.083 �1.9 2.2 96.0 0.1 1815 58 14.6 0.6 108500 30 9 0.9 0.060 3.1 1.6 49.1 0.1 1454 64 5.7 0.3 185525 30 6 1.9 0.028 3.2 1.6 7.2 0.1 1012 116 0.5 0.1 232550 30 5 1.6 0.026 4.3 1.7 5.7 0.1 929 120 0.4 0.1 220575 30 5 1.9 0.024 4.3 1.8 4.0 0.1 1345 206 0.4 0.1 229600 30 5 1.2 0.025 6.3 1.7 6.9 0.1 1513 178 0.8 0.1 255625 30 4 1.6 0.014 7.2 1.4 6.3 0.1 1411 173 0.7 0.1 311650 30 2 0.84 0.006 8.6 2.1 5.5 0.1 1988 261 0.9 0.1 360675 30 2 0.10 0.005 9.3 2.3 5.6 0.1 2344 305 1.1 0.2 332700 30 2 0.24 0.003 9.4 2.9 4.1 0.1 3822 578 1.5 0.2 865750 30 6 0.39 0.009 9.3 1.6 8.6 0.1 3483 367 2.7 0.3 769800 30 2 0.07 0.020 19.4 1.2 12.6 0.1 4107 358 4.8 0.5 129850 30 1 0.08 0.023 21.2 1.6 15.8 0.1 3205 218 4.6 0.3 62900 30 2 0.08 0.014 11.5 1.4 21.9 0.2 1277 74 2.2 0.2 136950 30 3 0.10 0.005 25.3 2.3 23.1 0.4 983 56 1.6 0.1 6781000 30 2 0.02 0.015 9.7 1.3 23.1 0.7 1379 82 2.5 0.2 1401050 30 3 0.09 0.017 11.3 1.3 24.2 1.2 1962 132 4.0 0.3 1751100 30 3 0.10 0.018 14.4 1.2 21.5 2.1 1494 141 2.6 0.2 1961150 30 5 0.39 0.021 16.9 1.2 28.7 3.5 1299 136 2.9 0.2 3081150 180 24 1.0 0.029 9.4 1.7 33.7 21.0 2208 1191 6.4 0.3 9941200 300 80 4.6 0.206 �0.8 1.5 1151.1 56.3 1352 53 121.6 1.2 4561200 30 40 1.5 0.005 13.9 2.7 22.9 5.6 447 43 0.3 0.1 9739R 1150 30 1.4 0.049 12.5 1.1 62 21 1790 646 9.3 0.3 705R 1200 120 6.2 0.211 �0.5 1.5 1174 57 1335 52 122.0 1.2 664Total 261 48 1.312 �1.2 0.6 1811 60 1550 41 227.3 2.0 232R(500–1200�C) 211 19 0.572 5.7 0.6 1482 60 1432 50 168.4 1.6 430

East-Pilbara Greenstone Belt- North Pole, Panorama Formation-Strelley Pool Chert

Pi-02-47 (228.70 mg)

300 30 13 122 0.223 1.5 1.5 4550 0.4 300 1 2.2 0.6 70400 30 19 104 0.243 1.1 1.4 4721 0.6 298 1 1.3 0.6 91


Table 3b (continued)

Combustionstep �C


Cppm

H2Oppm

Nppm

d15N(&)

± 36Ar10�12 cm3

±STP/g

40Ar/36Ar ± 40Ar�

10�8 cm3±STP/g

C/Nmolar

450 30 21 24 0.092 4.0 1.1 1740 0.4 301 2 1.0 0.4 264500 30 31 19 0.067 2.4 1.2 1512 0.3 301 2 0.8 0.4 539500 60 21 0.19 0.013 6.1 1.4 240 0.6 293 6 1802550 30 105 2.3 0.061 �0.2 1.2 1497 0.3 300 2 0.6 0.4 1992600 30 83 8.2 0.110 �0.4 1.4 2496 0.3 302 2 1.6 0.5 883650 30 444 3.8 0.149 �0.4 1.2 3095 0.3 300 2 1.3 0.5 3483700 30 9899 38 0.636 0.5 1.4 7517 1 301 1 3.9 0.8 18163700 60 18881 106 0.995 �1.2 1.6 7360 1 303 1 5.5 0.8 22136750 30 14815 13 0.073 2.3 1.9 350 3 314 5 0.6 0.2 235908800 30 33288 3.1 0.014 4.0 1.5 2 4 385 193 0.02 0.03 2785105850 30 1017 14 0.002 7.8 6.3 — — — — — — 714349900 30 130 0.09 0.003 9.6 4.6 5 2 274 31 — — 546991000 30 55 2.0 0.005 8.2 2.4 17 1 332 25 0.06 0.04 137761200 30 6 0.01 0.009 9.4 2.5 28 3 557 40 0.7 0.1 8151200 30 1 0.01 0.000 8624R500 52 20 0.081 3.0 1.0 1751 1 300 2 0.8 0.4 750R700 28780 144 1.631 �0.5 1.1 14877 1 302 1 9.4 1.2 20587R1200 7 0.02 0.009 7.6 2.9 28 4 557 40 0.7 0.1 953Total 78830 460 2.694 0.2 0.7 35129 8 301 1 19.7 1.8 34133R(500–1200�C) 78777 210 2.137 �0.2 0.9 24118 8 302 1 15.1 1.5 43005

East-Pilbara Greenstone Belt-Chinaman Creek, Apex Basalt

Pi-02-07 (188.06 mg)

300 90 39 73 0.591 0.5 1.0 514 1 341 5 2.4 0.3 77400 90 32 66 0.234 8.9 0.8 260 1 348 7 1.4 0.2 157450 60 5 47 0.070 8.6 1.2 90 1 458 16 1.5 0.1 85500 60 4 16 0.086 8.1 0.9 90 1 565 19 2.4 0.2 59550 45 4 5.5 0.105 6.0 0.9 69 1 678 41 2.6 0.3 49600 30 9 2.2 0.182 2.5 1.0 58 1 979 43 4.0 0.2 59650 30 11 0.40 0.084 1.1 0.9 37 1 1238 68 3.5 0.2 156700 30 10 0.55 0.040 2.0 1.2 34 1 1637 96 4.5 0.3 297800 30 9 2.0 0.047 4.7 1.1 37 1 2769 155 9.2 0.5 233900 30 10 0.98 0.040 9.5 1.1 37 1 4063 231 13.8 0.8 2811000 30 36 0.51 0.028 10.7 1.3 42 1 4530 261 17.6 0.9 15381100 30 66 0.40 0.024 4.4 0.9 36 4 5471 590 18.7 1.0 31961200 30 20 — 0.043 0.9 0.8 50 10 5442 1018 25.7 1.1 5381200 60 4 — 0.028 �0.7 1.2 19 19 10251 9915 19.3 0.9 1501200 60 4 — 0.021 �0.2 1.0 54 19 2632 831 12.7 0.5 1991200 120 3 — 0.019 1.3 1.1 — — — — 11.3 0.7 1921200 30 1 — 0.005 1.4 1.4 11 10 2619 2003 2.6 0.3 174R1200 31 — 0.116 0.4 0.5 91 49 7039 2244 71.6 1.7 309Total 267 215 1.647 3.6 0.4 1394 49 1322 148 153.1 2.4 189R(500–1200�C) 191 29 0.752 4.0 0.3 575 2 2892 389 147.9 2.4 92


of the d15N values, from +5.2 ± 0.4& at 300–900 �C to�2.0 ± 0.7& at 1000–1200 �C (Fig. 8); (ii) radiogenic40Ar� is released at 600 �C and then at 1200 �C, the latterassociated with N. Some differences can be detected. Thefirst is a flattening of the N and Ar isotopic patterns, fromsample Pi-47-00/1 to sample Pi-47-00/2 (Fig. 8). Thisbehavior can be easily explained by a loss of resolution.N and Ar have been extracted with a resolution of 50 �C/step from Pi-47-00/1 and 100 �C/step from Pi-47-00/2.However, both samples show a progressive decrease inthe d15N with the temperature, from � +5& to �5&

(Fig. 8). However, for sample Pi-47-00/1, the released40Ar� amount at 550 �C is 20 times lower than that releasedat 1200 �C. For Pi-47-00/2, the first 40Ar� release amount isonly 3 times lower than that at 1200 �C. This difference

could be caused by sample heterogeneity. Carbon showsalso two release peaks (Fig. 8); one centered at 700 �Cand the other at 1200 �C. The relatively high C/N ratio of2500 measured for this C component, its temperature of re-lease (700–750 �C) and the apparent decoupling with radio-genic argon (Fig. 8) are typical of thermal decomposition ofcarbonates (Sano et al., 1994; Boyd et al., 1997). However,residual C released at the higher temperatures of 900–1200�C from Pi-47-00/1 is well correlated with N (not plottedhere) and shows a C/N ratio of 53, compatible with an or-ganic source (Pinti et al., 2007). The combustion pattern ofwater measured in Pi-47-00/1 shows three distinct releasepeaks at 500�, 600–650 �C and 800 �C, which are com-pletely decoupled from those of N, Ar and C (Fig. 8). Itmight be possible that H2O released at 500 �C is related

Chert Pi-01-45/01Dixon Island Fm. (3.2 Ga)

Chert Pi-01-45/02Dixon Island Fm. (3.2 Ga)

-10

-5

0

5

10

δ 15N, ‰

0

250

500

750

1000

N, n

mol

e/g

0

5000

10000

15000

40Ar/ 36Ar

0

200

400

600

800

40A

r*, p

mol

e/g

05

1015202530

C, μ

mol

e/g

300 400 500 600 700 800 900 1000 1100 1200Temperature, ºC

-10

-5

0

5

10

δ 15N, ‰

0

1000

2000

3000

4000

5000

N, n

mol

e/g

0

500

1000

1500

2000

2500

3000

40Ar/ 36Ar

0

500

1000

1500

2000

40A

r*, p

mol

e/g

2468

10121416

C, μ

mol

e/g

300 400 500 600 700 800 900 1000 1100 1200Temperature, ºC

Fig. 6. Combustion patterns measured for C, N and Ar in two repeated experiments on samples from Dixon Island black kerogeneous chertPi-01-45. Units of concentration have been converted to nmole/g (N), pmole/g (40Ar�) and lmole/g (C and H2O) from data of Table 3 for sakeof clearness. Data for temperature step 1200 �C are reported as R(1200 �C) as in Tables 3(a and b).


to decrepitation of some aqueous fluid inclusions (Kendricket al., 2006), while that released at 600–650 �C may be re-lated to the thermal decomposition of carbonate minerals.

Black chert Pi-85-00 has been also collected from one ofthe North Pole silica veins. However, the N and Ar isotopicand elemental combustion release patterns are completelydifferent from that found in Pi-47-00 (Figs. 8 and 9).Nitrogen is released at 500–550 �C (d15N(R500–550�C) =+11.2 ± 1.1 &) together with little C and H (Fig. 9; Table3); at 850–900 �C (d15N(R850–900�C) = +8.1 ± 1.2&); andthen at 1200 �C (d15N(R1200�C) = +12.2 ± 1.3&). A smallamount of N might have been released at 750 �C (Fig. 9) witha d15N value of +7.4 ± 1.8&. Essentially, all the N compo-nents released from this chert are 15N-enriched. Radiogenic40Ar is slightly released at 600 �C, and then most of it is re-leased at 1200 �C (Fig. 9). Water is primarily released at550 �C and a little at 750 �C. Carbon is essentially releasedat 700–800 �C. The measured C/N ratios up to 14,000strongly suggest a carbonate-rich source (Sano et al., 1994).

The barite vein Pi-II-08-00 shows a singular release pat-tern for all elements, between 300 and 500 �C. Largeamounts of H and 15N-depleted nitrogen, though with little40Ar� and C, are released at low temperatures. This compo-nent likely originates from surface contamination (Pinti

et al., 2001), because this sample was not baked overnight(see details in Section 3). Nitrogen is then released at800–850 �C with an average d15N(R800-850�C) value of+20.4 ± 1.0& associated with a small release peak of radio-genic 40Ar�. The C/N ratio of this component is 94. Theoccurrence of 40Ar� and the temperature of 800–850 �Ccould indicate the thermal decomposition of some mica(Boyd et al., 1993), although it is not accompanied by therelease of a hydrogenous phase as water (Fig. 9). The mostimportant N component for Pi-II-08-00 is released duringthe first step at 1200 �C, together with large amounts of40Ar�, H and C (Fig. 9). The d15N value of this componentis �0.8 ± 1.5& (Table 3). Between 300 and 650 �C and1050–1200 �C, C and H are released from two distinct res-ervoirs showing different C/H (expressed as H2O) molar ra-tios (3.2 and 27, respectively; Fig. 10). The C/N molarratios of these two components are also distinctly different.The low temperature component has a C/N molar ratio of�60 (Table 3), while the C/N molar ratio at 1200 �C is�450 (Table 3). The release of N and C at 1200 �C frombarite vein Pi-II-08-00 was accompanied by a small releaseof sulfur, which deposited on the wall of the glass furnaceused for the sample combustion. The simplest explanationis the thermal decomposition of barite started, possibly

Pi-01-45/2 (800-1200 C)

Increasing extraction temperature

-2

0

2

4

6

8

10

12

0 1 10-4 2 10-4 3 10-4 4 10-4 5 10-4 6 10-4 7 10-4 8 10-4

δ15N

, ‰

40Ar*/N (mol/mol)

Fig. 7. Covariations of the 40Ar�/N molar ratio measured for mid-and high-temperature steps (800-1200 �C) in chert Pi-01-45/2against the d15N value. These variations represent the mixingbetween two N-Ar components located within two K-bearingminerals (illite and K-Ba-feldspars). See text for details.


accompanied by the release of H2O and C from the numer-ous H2O-CO2-H2S-CH4 inclusions that have been observedin these barite formations (Sommer II and Gibson Jr.,1986). Indeed, experimental heating of BaSO4 in vacuumshows that the thermal decomposition of barite is producedat 1400 K, i.e. 1127 �C (L’Vov and Novichikhin, 1997).

Fig. 11 depicts the release pattern of volatiles in thedolomitic stromatolite Pi-02-47. A smaller N amount is re-leased at lower temperatures of 300–500 �C together with40Ar�, H but no carbon. This N could be related to super-ficial contamination not completely removed during over-night baking. Nitrogen is then released at around 700 �C,together with large amounts of atmospheric-like argon (Ta-ble 3), C and H. The d15N value and the 40Ar/36Ar ratio ofthis low-temperature release component are close to that ofthe modern atmosphere (d15N(R700�C) = �0.5 ± 1.1&;40Ar/36Ar = 301 ± 1), but the N2/40Ar ratio is 3 � higherthan the present atmospheric ratio (N2/40Ar(R700�C) =8.8 � 104). Carbon exhibits two major release peaks at700 and 800 �C, while N, Ar and H show only one releasepeak at 700 �C. This could indicate the occurrence of twocarbonate phases in the sample with the first one enrichedin aqueous inclusions and nitrogen (Fig. 11). Sample Pi-02-47 contains also Fe-oxides that could contain N andAr (Pinti et al., 2007). However, experiments carried outon Fe-oxyhydroxides and Fe-rich layers from Banded IronFormations showed that N is mainly released between 400and 600 �C and H2O is released mainly at 400 �C (Pintiet al., 2007; Hashizume et al., 2008). A small amount ofradiogenic 40Ar* is released at 1200 �C (40Ar/36Ar =557 ± 40) but this is not accompanied by N nor H (Fig. 11).

Finally, Apex chert Pi-02-07 (Fig. 11) shows a nitrogenrelease pattern which is mostly decoupled from that ofradiogenic 40Ar* and H. Nitrogen is released at 300 �C

(d15N = +0.5 ± 1.0&); at 600 �C (d15N = +2.5 ± 1.0&

and C/Nmolar = 59) and then at 1200 �C (d15N(R1200 �C) =+0.4 ± 0.5& and C/Nmolar = 309). The release amountsof radiogenic 40Ar* increase with increasing combustiontemperature (Fig. 11), which are accompanied by an in-crease of the 40Ar/36Ar from ratios close to the atmosphere(40Ar/36Ar = 341 ± 5) up to 5400 (Table 3). Carbon is re-leased at 300–500 �C, a little at 650–700 �C and then at1000–1200 �C.

5. DISCUSSION

There are exhaustive literatures on the N contaminationof rock samples such as cherts, diamonds, basaltic glasses,or meteoritic samples (e.g., Hashizume and Sugiura, 1990;Pineau and Javoy, 1994; Hashizume and Sugiura, 1995,1997; Boyd et al., 1997; Pinti et al., 2001; Brauer andHahne, 2005; Busigny et al., 2005; Ader et al., 2006; Pintiet al., 2007) and all point out that stepped combustion isprobably the best method for discriminating between or-ganic contaminant N and pristine N trapped in the minerallattice. These works clearly showed experimentally thatcontaminant-N is essentially extracted between ambienttemperature and 450 �C. Although several precautions weretaken during laboratory handling and analysis, and a pre-heating treatment of the sample was carried out overnight(see Section 3), a few samples show abundant N, H, andC released at temperatures between 200 and 450 �C (Figs.9 and 11). This component is related to superficial contam-ination and its occurrence is well known. In the followingdiscussion, the only components considered as pristine arethose released by combustion between 500 and 1200 �C(Table 3), as clearly identified by Pinti et al. (2001).

At high release temperatures, N can be affected by isoto-pic fractionation following a diffusion-controlled process(Boyd et al., 1993; Pinti et al., 2007). This behavior is sim-ilar to that of N during prograde metamorphism. Ammo-nium is lost from mica by devolatilization, whose d15Nvalue increases with decreasing concentration (Haendelet al., 1986). In our samples, only nitrogen released between900 and 1200 �C in chert Pi-01-45 (Fig. 6) and that releasedbetween 750 and 900 �C in stromatolite Pi-02-47 (Fig. 11)could be partially affected by this kinetic fractionation.However, Arrhenius plots for N (the log of the N fractionextracted at each combustion temperature versus the in-verse of the extraction temperature) for these two samplesdid not show a clear correlation (not plotted here) perhapsindicating that mixing with N components released at hightemperature disturbed the kinetic release of N in these twosamples. In the following discussion, these values will notbe considered further. We define and discuss here isotopicsignatures of trapped components as those measured atthe fraction corresponding to the highest N release amounts(Figs. 6, 8, 9 and 11).

5.1. C and N isotopic variability in Paleoarchean cherts

The nitrogen components extracted from the sampledPaleoarchean cherts have isotopic signatures spanning fromca. �5& up to +12& (Figs. 6, 8, 9 and 11). This range of

Chert Pi-47-00/1Dresser Fm. (3.5 Ga)

Chert Pi-47-00/2Dresser Fm. (3.5 Ga)

-10

-5

0

5

10

δ 15N, ‰

0

10

20

30

40

50

60

70

80

90

N, n

mol

e/g

0

20000

40000

60000

80000

40Ar/ 36Ar

0

2500

5000

7500

10000

40A

r*, p

mol

e/g

0

2

4

6

8

C,μ

mol

e/g

00.10.20.30.40.50.6

H2O

, μm

ole/

g

300 400 500 600 700 800 900 1000 1100 1200Temperature, ºC

-5

0

5

10

δ 15N, ‰

0

25

50

75

100

125

N, n

mol

e/g

0

5000

10000

15000

40Ar/ 36Ar

0

100

200

300

400

500

600

40A

r*, p

mol

e/g

300 400 500 600 700 800 900 1000 1100 1200

Temperature, ºC

Fig. 8. Combustion patterns measured for H (here calculated as H2O), C, N and Ar in repeated experiments on samples from North Polesilica vein black chert Pi-47-00. Units are the same than in Fig. 6.


values is basically the same as those observed by Beaumontand Robert (1999) (d15N values from �6.2& to +13&; op.cit.) and by Pinti et al. (2001) (d15N values from �7& to+12&; op. cit.), when samples with the same period andof the same metamorphic grade are compared. The mea-sured d15N values for the Paleoarchean cherts are lowerthan that measured in cherts from later periods, particularlythose from the Neoarchean (d15N values from +6& to+35&; Beaumont and Robert, 1999; Jia and Kerrich,2004a,b). Yet, the persistency of light and heavy N in thosecherts need to be clarified.

Pinti et al. (2001) suggested N loss from a 15N-depletedsource (d15N 6 �7&) by metamorphic-induced devolatil-ization (Bebout and Fogel, 1992), to explain the inverse cor-relation between the d15N values and the N concentrationsobserved among cherts affected by different metamorphicgrades, from greenschist to amphibolite. Nitrogen loss bypost-depositional devolatilization (which can be modelledby a Rayleigh distillation process) will favor the loss of

lighter 14N compared to heavier 15N, increasing the d15Nvalue of the residue, i.e. the metamorphosed rock.

Here, we plotted the d15N(R500–1200�C) values (represent-ing those from the bulk, uncontaminated rock) againstthe bulk d13C values of the sampled cherts (Fig. 12a). Thed15N and d13C values measured in chert PANO D-136-0(Pinti et al., 2001; this study) are included in Fig. 12a forcomparison. For cherts Pi-47-00 and Pi-01-45, the averagedvalues ± standard deviation of the two repeated analyzeswere used. Except for sample Pi-85-00, all cherts show arough co-variation between the d15N and the d13C(Fig. 12a).

This co-variation could be explained by a mixing be-tween two different N and C sources or by the progressiveC and N loss from a unique source induced by devolatiliza-tion, with the consequent progressive 13C and 15N enrich-ment in the rock. In the case of a mixing, one of thesources must be a 15N-13C-depleted one, with d15N andd13C values lower than �4& and �36&, respectively.

Chert Pi-85-00Dresser Fm. (3.5 Ga)

Barite Pi-II-08-00Dresser Fm. (3.5 Ga)

0

5

10

15

20

δ 15N, ‰

0

5

10

15

N, n

mol

e/g

0

5000

10000

15000

20000

40Ar/ 36Ar

0

25

50

75

100

40A

r*, p

mol

e/g

05

10152025

C,μ

mol

e/g

00.5

11.5

22.5

3

H2O

,μm

ole/

g

300 400 500 600 700 800 900 1000 1100 1200Temperature, ºC

-20

-10

0

10

20

30

δ 15N, ‰

0

5

10

15

20

25

30

N, n

mol

e/g

0

1000

2000

3000

4000

5000

40Ar/ 36Ar

0

20

40

60

40A

r*, p

mol

e/g

0.02.55.07.5

10.012.5

C,μ

mol

e/g

00.10.20.30.40.5

H2O

,μm

ole/

g

300 400 500 600 700 800 900 1000 1100 1200Temperature, ºC

Fig. 9. Combustion patterns measured for H (here calculated as H2O), C, N and Ar in black chert from the North Pole silica veins (Pi-85-00)and in barite vein (Pi-II-08-00). Units are the same than in Fig. 6. For sample Pi-II-08-00, data at 1150 �C is calculated as sum of two repeatedsteps (see Table 3b).


The C isotopic ratio of this hypothetical source is in therange of those normally observed in Archean kerogens(Beaumont and Robert, 1999; Jia and Kerrich, 2004a,b;Ueno et al., 2004). The N isotopic ratio is within the valuesmeasured in the North Pole kerogeneous black cherts (Pintiet al., 2001; Ueno et al., 2004). If biological, this source canbe assimilated to the activity of chemolithoautotrophs andmethanogens bacteria in hydrothermal environments (Pintiet al., 2001; Ueno et al., 2004, 2006).

The second source shows d15N and d13C values greaterthan +4& and �19&, respectively. If similar N isotopic ra-tios have been measured in Paleoarchean cherts (Beaumontand Robert, 1999; Pinti et al., 2001, 2007), the C isotopicratios are quite unusual. Similar d13C values have beenmeasured in 3.8 Ga old graphite from Isua (West Green-land) extracted from black slates (d13C = �18.7&; VanZuilen et al., 2005); in graphite from Paleoarchean carbona-

ceous chert from Barberton Greenstone Belt (South Africa)(d13C = �20.7 ± 1.7&; Van Zuilen et al., 2007); and in Fe-Mn oxides of a Paleoarchean chert from Marble Bar, Pil-bara (d13C = �19.9 ± 0.1&; Pinti et al., 2007). Thesed13C values are 10–20 & lower than those normally ob-served in the Archean (e.g., Beaumont and Robert, 1999;Jia and Kerrich, 2004b). However, both C and N isotopicratios are within the values observed in modern marine sed-imentary organic matter (e.g., Peters et al., 1978). The mea-sured d13C in Apex chert is indeed characteristic ofphotosynthetic fixation of CO2 by cyanobacteria, via ribu-lose-1,5-bishosphate carboxylase (RuBisCo), while thed15N ratio is similar to that observed in N undergone to afixation-assimilation-nitrification-denitrification metabolicpathway (Shen et al., 2006).

If we assume an isotopic mixing between these two bio-logical sources, then we have to admit that during the

0

1 10-7

2 10-7

3 10-7

4 10-7

5 10-7

0 1 10-6 2 10-6 3 10-6 4 10-6 5 10-6 6 10-6 7 10-6

300-675 ˚C1050-1200 ˚C

y = -8.02e-9 + 0.311x R 2= 0.66

y = -8.26e-9 + 0.037x R 2= 0.97

H2O

, mol

/g

C, mol/g

Incr

ea

sin

g t

em

pe

ratu

re

Increasing temperature

Fig. 10. The ratio between C and H (expressed here as H2O) released at low-temperature steps of 300 and 675 �C and at high-temperature of1050 to 1200 �C in barite Pi-II-08-00. The low temperature released volatiles are related to modern contamination. C and H2O released at hightemperature are from fluid inclusions in barite.


Paleoarchean, the biosphere was already diversified withoxygenic (or anoxygenic) photosynthetic mats, stromato-lites, and cyanobacterial plankton at the surface of theocean, and anoxygenic photosynthesizers, methanogens atdepth, and hyperthermophiles (cf. and likely chemosynthe-sizers) at MOR, as suggested by Nisbet and Sleep (2001).This hypothesis is appealing for the dramatic consequencesthat they could have on the long-standing debate on theoxygenation of the Earth (Ohmoto, 1997); or on the typeof carbonaceous matter preserved in Apex chert (e.g.,Schopf 1993; Brasier et al., 2002). However, the mixinghypothesis cannot explain the strong positive N isotopicshift of chert Pi-85-00 (d15N = 11&; Fig. 12a), which isnot accompanied by an equivalent enrichment of 13C(d13C = �38&).

An alternative explanation to the N and C isotopic co-variations illustrated in Fig. 12a is the occurrence of apost-depositional alteration process of a pristine sourcedepleted in 15N and 13C. This process could be relatedto metasomatic overprinting, as observed in the NorthPole silica veins (Ueno et al., 2004); metamorphic devola-tilization (Pinti et al., 2001); or finally hydrothermal alter-ation (Busigny et al., 2005). There are sparse evidencesthat our samples have been partially or largely overprintedby one of these processes: chert Pi-85-00 shows clinochloreovergrowth around sulfides (Fig. 3d), possibly indicatinglater pervasive hydrothermal alteration; Apex chert Pi-02-07 shows an anomalous increase in the LREE/HREEratio (Figs. 4a and 5c) which could indicate someweathering.

Trace elements and REE could be useful in decipheringamong these processes. There is no evident relation betweenthe C and N isotopic composition and LREE/HREE ratios.This excludes surface weathering as a prominent cause ofthe N and C isotopic shift. However, the d15N and the

d13C values show a meaningful correlation with the Ba/La ratio (Fig. 12c and d) except for Pi-85-00 (Fig. 12c)and to a less extent with the Co/As ratio (not plotted here).Indeed, Ba/La shows an inverse correlation with the Co/Asratio (Fig. 12b). These rough co-variations could be ex-plained by hydrothermal alteration of the cherts. The blackcherts from the North Pole veins (Pi-47-00, Pi-85-00 andPANO D-136-0) plot on the left side of the Ba/La vs. Co/As diagram, close to the Co/As values which are measuredin hydrothermal hot and cold fluids (Metz and Trefry, 2000;Hannington et al., 2005; Von Damm, 1995). For the Ba/Laratio, we assumed that its value in hydrothermal fluidsshould be close to that of the mantle (�10; e.g., Rollinson,1993). During weathering, the Ba, which is highly mobileshould be removed while La, a LREE, is immobile thusbeing concentrated in the weathered residue. However, dur-ing hydrothermal alteration, Ba is often enriched, as ob-served in basalts (Humphris and Thompson, 1978). TheBa enrichment should thus produce a net increase of theBa/La ratio, as observed in Fig. 12b. Although Co andAs are both relatively mobile elements in hydrothermal sys-tems (Hollings and Wyman, 2005), enrichment only of Ashas been observed in ore deposits associated with hydro-thermal alteration and formed by metamorphic fluid flow(Craw, 2002). Thus we could speculate that the inverse cor-relation observed in Fig. 12b indicates a progressive hydro-thermal alteration of our cherts by hydrothermal/metamorphic fluids. Chert Pi-85-00 does not follow thesame trend between d15N and the Ba/La ratio, indicatinga different process of alteration.

N and eventually C isotopic shifts can be produced byisotopic exchange between the chert and the metasomaticfluids, either of hydrothermal or metamorphic origin (Beb-out and Fogel, 1992). Busigny et al. (2005) suggested a pro-cess of isotopic exchange between (N2)fluid and (NH4

+)rock

Dolomitic stromatolite Pi-02-47Strelley Pool Chert unit (3.4 Ga)

Apex Chert Pi-02-07Apex Basalt Fm. (3.5 Ga)

-5

0

5

10

15

δ 15N, ‰

0

10

20

30

40

50

N, n

mol

e/g

0

2500

5000

7500

10000

40Ar/ 36Ar

0

10

20

30

40

40A

r*, p

mol

e/g

0123456

C, μ

mol

e/g

012345

H2O

,μm

ole/

g

300 400 500 600 700 800 900 1000 1100 1200Temperature, ºC

-2

0

2

4

6

8

10

12

14

16

δ 15N, ‰

0

20

40

60

80

100

120

N, n

mol

e/g

100

200

300

400

500

600

40Ar/ 36Ar

0.0

1.0

2.0

3.0

4.0

5.0

40A

r*, p

mol

e/g

0

1000

2000

3000

C,μ

mol

e/g

0

4

8

H2O

, μm

ole/

g

300 400 500 600 700 800 900 1000 1100 1200Temperature, ºC

Fig. 11. Combustion patterns measured for H (here calculated as H2O), C, N and Ar in stromatolite from Trendall locality (Pi-02-47) and inApex Chert from Chinnaman Creek (Pi-02-07). Units are the same than in Fig. 6. For sample Pi-02-47, data at 500 and 700 �C are calculatedas sum of two repeated steps (see Table 3b).


to explain N isotopic shift associated with basalt alteration,following the reaction:

15N 14N ðfluidÞ þ 14NHþ4 ðrockÞ �15NHþ4 ðrockÞ þ

14N 2ðfluidÞ ð1Þ

Isotopic exchange in cherts could occur also throughNH4

+-NH3 and NH3-N2 reactions (Haendel et al., 1986;Pinti et al., 2001), yet the one represented in eq. (1) seemsto be the most common (Busigny et al., 2005; Svensenet al., 2008). Nitrogen in silicate rocks is often preservedin the ammonium form (NH4

+) (Honma and Itihara,1981), whereas hydrothermal fluids percolating throughthe North Pole silica veins seems to contain N mostly asN2 (Nishizawa et al., 2007). Hanschamn (1981) has calcu-lated the fractionation factors (a) for the NH4

+-N2 ex-change reaction, which could have taken place in a rangeof temperature between 100 and 350 �C. The lower limit

is fixed by the occurrence of halloysite in Pi-02-07 (Mineauet al., 2008), a low-T alteration product. The upper limit isthe maximum hydrothermal temperature reached by Pi-02-07, based on the presence of native metals (this study; Bra-sier et al., 2002, 2005). Temperatures of 350 �C are alsocompatible with hydrothermal/metamorphic events in thearea (Kitajima et al., 2001). For C, it is more complicatedto evaluate what kind of isotopic exchange occurred. Uenoet al. (2004) assumed post-depositional metasomatism (oxi-dation) for the North Pole graphite, with a fractionationfactor a = 1.0015.

The effect of N (and C) loss on its isotopic compositioncan be modeled as (a) batch volatilization (one step pro-cess), where fluids that are released equilibrate with the rockin a closed system; as (b) Rayleigh distillation, where eachsmall aliquot of volatile that is released is immediately re-

0

0.5

1

1.5

2

0 200 400 600 800 1000

Co/

As

Ba/La

Pi-85-00

Pano D-136-0

Pi-47-00

Pi-01-45Pi-02-07

hydrothermal fluids

hydrothermal alteration

-40

-35

-30

-25

-20

-15

0 200 400 600 800 1000

δ13C

, ‰

Ba/La

Pi-85-00

Pano D-136-0Pi-47-00

Pi-01-45

Pi-02-07

-6

-4

-2

0

2

4

6

8

10

12

14

0 200 400 600 800 1000

δ15N

, ‰

Ba/La

Pi-85-00

Pano D-136-0

Pi-47-00

Pi-01-45

Pi-02-07

-40

-35

-30

-25

-20

-15

-6 -4 -2 0 2 4 6 8 10 12 14

δ13C

, ‰

δ15N, ‰

Pi-85-00Pano D-136-0

Pi-47-00

Pi-01-45

Pi-02-07

(b)(a)

(d)(c)

Fig. 12. (a) The bulk C and N isotopic composition of studied cherts, including PANO D-136-0 (C data from this study; N data from Pintiet al., 2001). (b) The Ba/La ratio versus the Co/As ratio for the studied cherts. The shaded area labeled ‘‘hydrothermal fluids” representsvariations of these two ratios measured in hot and cold hydrothermal fluids. Data are from Von Damm (2005), Hannington et al. (2005) andMetz and Trefy (2000). (c, d) The N and C isotopic composition of studied cherts is plotted against the Ba/La ratio.


moved from the rock in an open system. The equations thatgovern these two processes are, respectively:

dR ¼ dR0 þ ð1� F Þ1000lna ð2Þ

dR ¼ dR0 � 1000½F ða�1Þ � 1� ð3Þ

where dR0 and dR are respectively the N and C isotopiccomposition before and after the metamorphic/hydrother-mal alteration and a is the fractionation factor. F is thefraction of nitrogen and carbon that remains in the rockafter devolatilization reactions.

Fig. 13a shows results from the calculation of the isoto-pic fractionation of a theoretical pristine N and C source

(here taken as PANO D-136-0) affected by Rayleigh distil-lation-type loss. Solid lines represent isotopic exchange be-tween ammonium and N2 at temperatures between 100 and350 �C, followed by oxidation of organic matter. Fraction-ation factors for N have been extrapolated from Hans-chamn (1981) and those of C from Ueno et al. (2004).These processes could be able to produce the desired isoto-pic shift observed in chert Pi-85-00 (Fig. 13a), assuming aninitial isotopic N and C composition of � �6& and �42&.However, to explain the C isotopic shifts observed in theother cherts we should consider fractionation factors of1.0035–1.010 for reactions produced at 350 �C (Fig. 13a;dashed lines), assuming that C and N share the same F va-

-4

-2

0

2

4

6

8

10

12

0 0.2 0.4 0.6 0.8 1

δ15N

, ‰

Fraction N remaining

Pi-02-07

Pi-85-00

Pi-47-00

PANO D-136-0

NH4

+chert

- N2 fluid

350˚C

100˚C

350˚C

100˚C

-40

-35

-30

-25

-20

-15

-5 0 5 10 15

δ13C

, ‰

δ15N, ‰

α = 1.0035

α = 1.0100

α = 1.0015

Pi-02-07

Pi-85-00

Pi-47-00

PANO D-136-0

NH4

+chert

- N2 fluid

350˚C100˚C

Pi-01-45

b

a

Fig. 13. (a) The bulk C and N isotopic composition of studied cherts, including PANO D-136-0 (C data from this study; N data from Pintiet al., 2001). Straight lines represent the result of Rayleigh distilation of a chert having a pristine N and C isotopic composition equal to that ofPANO D-136-0. The two straight lines having a gentle slope have been calculated for an isotopic exchange between NH4

+(rock) and N2(fluid)

and for metasomatic alteration of C at temperatures of 100 and 350 �C. Fractionation factor ‘‘a” = 1.0015 has been calculated by Ueno et al.(2004) for graphite contained in the North Pole silica veins. The two steeper straight lines have been calculated for an isotopic exchangebetween NH4

+(rock) and N2(fluid) at 350 �C and using two hypothetical fractionation factors for C. The large extent of the fractionation

expected for C could be explained only by exchange of heavier C from carbonates to graphite C (see text for details). The fractions (F values)for the remaining C and N are assumed to be the same time-to-time. (b) The fraction of remaining N in the cherts (F) from North Pole andChinaman Creek against the N isotopic composition. Straight lines represent results of a batch volatilization and isotopic exchange betweenNH4

+(rock) and N2(fluid) at 100 and 350 �C. Curves represent results of a Rayleigh volatilization and isotopic exchange between NH4

+(rock) and

N2(fluid) at 100 and 350 �C. Fractionation factors calculated from data of Hanschmann (1981).


lue time-to-time. The only plausible mechanism to makekerogen enriched in 13C to such a large extent is by isotopicexchange with carbonate in presence of metamorphic CO2

fluids (e.g., Valley and O’Neil, 1981; Ueno et al., 2002).

This enrichment is favored by the relatively high diffusivi-ties of a 13C-enriched carbonate phase and a 13C-depletedgraphitic phase with very slow diffusion rate (Hoefs,2004). On the basis of experimental work from Scheele


and Hoefs (1992), Ueno et al. (2004) suggested that isotopicexchange between carbonate carbon and organic carbon atthe temperature of 350 �C is minimal and thus this processshould be negligible at North Pole. However, Morikiyo(1984) showed that isotopic exchange between carbonateand graphite might occur down to 270 �C. Using data fromMorikiyo (1984), we obtain a fractionation factor acc–

gc = 1.0159 for a temperature of 350 �C. Thus, isotopic ex-change between N2 and NH4

+ and between carbonate Cand graphite C promoted by pervasive CO2-N2 hydrother-mal or metamorphic fluids at 350 �C could be a plausibleexplanation of the C and N isotopic variability measuredin our cherts (Fig. 13a).

If one of these processes of N and C loss took place, weshould observe an inverse correlation between the fractionof remaining N or C in the rock (F) and the observed15N and 13C isotopic shift. Cherts PANO D-136-0, Pi-47-00 and Pi-85-00 belong basically to the same silica veinsplaced around the monzogranite batholith of North Pole(Fig. 2) and Apex chert Pi-02-07 is also collected in a hydro-thermal silica vein in the upper Apex Basalt Chert unit.These cherts contain N dispersed within the silica matrix,possibly as kerogeneous compound or in tiny K-bearingphase (this study). With this in mind, we compared whetherthere is a relationship between F and the d15N value(Fig. 13b). F is calculated here by assuming PANO D-136-0 the undisturbed protolith (Pinti et al., 2001) and nor-malizing N abundances of respective cherts by that ofPANO D-136-0. Indeed, there is a rough inverse correlationbetween the fraction F and the measured d15N values(Fig. 13b). Chert Pi-47-00, the companion of PANO D-136-0 (see N release pattern for both these two cherts; this

10000

20000

30000

40000

50000

60000

70000

80000

-10 -5

40Ar

/36Ar

δ1

Mantle

1200 ˚C

1200 ˚C

Fig. 14. The d15N values against the 40Ar/36Ar ratios measured in release(white dots; data from Pinti et al., 2001) and chert Pi-47-00 (black dots; Tvalues in the Archean, as constrained by Cartigny et al. (2001a,b) and Maare likely lower than the values reported here but model-dependent (Samember label ‘‘sedimentary” has been calculated using data from Sano e

study; Pinti et al., 2001), has N concentrations and bulkd15N values in agreement with the selective loss of 35% ofits initial N content for both a batch volatilization (straightline) and a Rayleigh distillation (curved lines), in the rangeof 100–350 �C. The N concentration and isotopic composi-tion of Apex chert Pi-02-07 are compatible with the loss of87% of its initial N content, by a Rayleigh distillation at350 �C, while those of chert Pi-85-00 are compatible withthe loss of 91% of its initial N content, by a Rayleigh distil-lation at 100 �C (Fig. 13b). Chert Pi-01-45 from Dixon Is-land has not been reported in Fig. 13b, despite the factthat it seems to be affected by the same process of the NorthPole and Chinaman Creek cherts (Fig. 13a). Indeed, the Ncontent of Pi-45-01 is one or two order of magnitude largerthan that measured in the other cherts (Table 2) because ofthe large amount of ammonium-bearing mica, and thus itcannot be directly compared.

If the model is validated for N (Fig. 13b), problems arisefor C. Carbon is from different sources (organic C and car-bonate) in cherts (Figs. 6–9) and the amount of C has notbeen measured during isotopic analyzes. If we assume thatthe C released at 1200 �C is the pristine organic C compo-nent (Pinti et al., 2001), then there is a rough relation be-tween the C amount and the d13C (not reported here).The C-richest chert and least devolatilized chert being Pi-85-00 (CR1200�C = 115 ppm and d13C = �38.2&) and Apexchert Pi-02-47 having the lowest CR1200�C content of 31 ppmand the highest d13C of �18.7&.

We would like to spend a few words about the remotepossibility that the observed C and N isotopic variationsare related to abiologic C and N compounds producedthrough Fischer-Tropsch-Type (FTT) synthesis. FTT syn-

0 5 105N, ‰

650 ˚C

700-750 ˚C

Sedimentary

temperature steps from 500 and 1200 �C for chert PANO D-136-0able 3a). The shaded area represents the possible upper mantle d15Nrty and Zimmermann (1999). The 40Ar/36Ar ratios for the Archeanrda et al., 1985) and thus difficult to precisely quantify. The end-t al. (2001) and it represents crustal N.


thesis requires reactions involving CO and H2 in the pres-ence of a catalyst such as Ni, Co or Fe in a temperatureand pressure range of about 200–400 �C and 1-100 atm,respectively. Typical products include C1–C4 hydrocarbonsand higher polymerized carbon compounds (Lancet andAnders, 1970). FTT hydrocarbons are consistently depletedin 13C owing to a kinetic isotopic effect (McCollom and See-wald, 2006). Thus, highly 13C-depleted (d13C from -20 to -50&) graphitic carbon present in Archean cherts is claimedby several authors to be the product of FTT synthesis underhydrothermal conditions (e.g., Horita, 2005 for a review onthe debate). Isotopic fractionation by abiotic production ofN-based polymers and amino acids is much less docu-mented. The few existing experimental data, if not for theFTT synthesis, gave contradictory results. Miller–Ureyreactions of CH4–NH3–H2 mixtures have produced N-con-taining nonvolatile soluble organics (amino acids, organicacids) and polymers, whose d15N values are 8–11& greaterthan the starting NH3 (Kung et al., 1979). Plasma-dischargeof CO–N2–H2 produced carbonaceous materials, whichhave d15N values of �3& to �17& relative to the reactantN2 (Kerridge, 1999). With so few experimental evidences onthe isotopic fractionation of N, it is difficult to evaluate thepossibility whether the N and C co-variations illustrated inFig. 12 and 13 may be explained by abiotic production.

5.2. 15N-depleted nitrogen in the North Pole silica veins

In the devolatilization model illustrated in Fig. 13a,b, the15N-13C-depleted isotopic signature of chert PANO -136-0and in some extent that of chert Pi-47-00 would best representthe pristine isotopic composition of N and C. This hypothesisis supported by the release of the 15N-depleted nitrogen com-ponent observed at high temperatures (from 900 to 1200 �C)in Pi-47-00 (Fig. 8; this study) and PANO D-136-0 (Fig. 3 in

0.000

0.005

0.010

0.015

0.020

0.025

0.030

0 5 10-4 1

y = 0.002

y = 0.001

N/C

(mol

/mol

)

40Ar/C (

Pano D-136-0 (800-12

Pi-47-00/1

Barite fluids

Fig. 15. The N/C vs 40Ar/C molar ratios measured in high temperature ret al., 2001) and PI-47-00/1 (black circles; Table 3a). The molar compositichert dykes of North Pole is also reported for comparison (white star; d

Pinti et al., 2001). These two cherts show a very similar releasepattern for N, indicating a common N source firmly embed-ded in the microcrystalline silica. We might speak here of an‘‘armored effect” for nitrogen as that existing for graphiteand produced by quartz recrystallization (Wada and Suzuki,1982). This 15N-13C-depleted source might indicate organicmatter metabolized by organisms living close to hydrother-mal sites. Chemosynthetic bacteria and methanogensmetabolize inorganic source of C (likely CO2) and N (mainlyNH4

+ and possibly N2) to obtain their energy resources,through chemosynthesis and methanogenesis (Conwayet al., 1994). The hydrothermal origin, as chemical precipi-tates, of most of the studied Paleoarchean cherts gives anindependent support to this hypothesis (this study; Pintiet al., 2001; Orberger et al., 2006a; Nijman et al., 1998; Uenoet al., 2004; Kiyokawa et al., 2006; Ueno et al., 2006).

Jia and Kerrich (2004a,b) and Jia (2006) suggested thatthe observed 15N-depleted values in Paleoarchean chertscould be produced by an overprinting of the original com-ponent by the inorganic upper mantle N, which is known tohave a d15N =�5 ± 2& (e.g., Marty and Zimmermann,1999; Cartigny et al., 2001b). Fig. 14 illustrates the ‘‘in-tra-chert” co-variations between the d15N values and the40Ar/36Ar ratios measured in steps from 500 up to 1200�C for chert PANO-D-136-0 and Pi-47-00. Two commonN components reside in these cherts. The first, released be-tween 500 and 800 �C is 15N-enriched and it has an isotopicsignature of N and Ar compatible with a crustal source(d15N P +2& and 40Ar/36Ar 6 10,000; Sano et al., 2001).The second source is 15N-depleted and it shows N and Arisotopic signatures (d15N 6 -7 & and 40Ar/36Ar P50,000; Fig. 14) indistinguishable from that of the presentupper mantle d15N = �5 ± 2& and 40Ar/36Ar =35,000 ± 10,000; Marty and Zimmermann, 1999; Moreiraet al., 1998; Raquin et al., 2008). The upper mantle acquired

10-3 1.5 10-3 2 10-3

4464 + 7.2923x R2= 0.874

4621 + 178.87x R2= 0.996

mol/mol)

00 ˚C)

(800-1200 ˚C)

elease steps of cherts PANO D-136-0 (white circles; data from Pintion of fluids from barite (sample PI-II-08-00) syngenetic to the blackata calculated as sum of the 1200 �C release steps; Table 3b).


the negative d15N values as far back as the Archean as indi-cated by measured values on 3.2 Ga old diamonds (Carti-gny et al., 2001a). However, there is no isotopic evidencethat the Archean mantle had the same Ar isotopic signatureas today. The Archean mantle should have a 40Ar/36Ar ra-tio lower than the present mantle ratio of 35,000 ± 10,000(Sarda et al., 1985), but how much lower is model-depen-dent. Nonetheless, the isotopic similitude of N and Ar be-tween the North Pole silica veins and the Archean uppermantle subsists (Fig. 14).

To ascertain the origin of this 15N-depleted source, weshould look in detail at the intra-chert elemental and isoto-pic variation of these two cherts. Fig. 15 illustrates themeaningful correlation between 40Ar/C and N/C atomicratios for the component released at high temperature(800–1200 �C) from chert PANO D-136-0 and Pi-47-00.The relationship indicates a binary mixing between a C-poor component released at high temperature (1200 �C)and having C/N ratios of ca. 50 (Table 3, this study; Table2; Pinti et al., 2001) and a C-rich component released atlower temperatures and having C/N ratios higher than400. The C-poor component is associated with the releaseof the 15N-depleted component (Fig. 14). The low C/Nratios are incompatible with the occurrence of a mantlecomponent, although not necessarily with the hydrothermalcomponent, i.e., the volatile fraction degassed from thesource mantle, which is inferred to have a C/N ratio muchlower than the source composition (C/N � 450; Marty,1995; Cartigny et al., 2001a). Cartigny et al. (2001a) pro-vided evidence from diamonds that the C/N ratio in themantle did not significantly change since the Archean.Alternatively, we may interpret that the measuredd15N 6 �7& and the low C/N ratios if we assume thatthe source of C and N is organic and represents the meta-bolic products of chemosynthetic bacteria around hydro-thermal vents (Pinti and Hashizume, 2001; Shen et al.,2006). The slopes of the two correlations give the N/40Aratomic ratio for the two cherts. Chert PANO D-136-0 hasN/40Ar atomic ratio of 179, while chert Pi-47-00 showsN/40Ar ratio of 7. The close relationship between N and40Ar suggests that N in those cherts is trapped in form ofammonium in a K-bearing phase. In this case, the differencein the N/40Ar ratios calculated for the two cherts could beexplained by differences in the time-integrated NH4

+/K ra-tio in each rock and both elements have been acquiredin situ and not derived from an external exotic source. Onthe other hand, the N/40Ar ratio for PANO D-136-0 iswithin the ratios observed in MORBs (N/40Ar = 250± 80; Marty and Zimmermann, 1999; Cartigny et al.,2001b) suggesting the addition of a pure mantle N andAr component to C in this chert. In fact, this possibilityis supported by the correlation observed among the hightemperature fractions of PANO D-136-0 between N/Cand 36Ar/C (not shown here), as well as with 40Ar/C.

It is noteworthy that the isotopically heavier N compo-nent in chert PANO D-136-0 and Pi-47-00 has N/C and40Ar/C ratios (Fig. 15) and d15N values (Table 3) compati-ble with those observed in fluids trapped in barite from thesame North Pole silica veins. Low-temperature released Nis likely trapped in fluid inclusions (Sano and Pillinger,

1990; Pinti et al., 2001), thus the occurrence of intraforma-tional fluids in North Pole silica veins is not surprising. Thehigher d15N values of 0 to +3 & observed in Fig. 14 arecompatible with those measured in fluid inclusions trappedin quartz veins precipitated contemporaneously with theblack cherts (Nishizawa et al., 2007), possibly indicatingthat different generations of fluids interacted or participatedto the precipitation of these cherts.

6. CONCLUSIONS

This study has established important details and ‘‘modusoperandi” for discriminating among pristine sources anddetermining the degree of preservation of those signals.Among others, the main conclusions of this work are:

- A 15N-depleted nitrogen source (d15N 6 �6&) occurredin most samples but it is partially (chert Pi-47-00) or lar-gely (Apex chert) overprinted by 15N and 13C enrichmentcaused by isotopic exchange between the rock and thehydrothermal/metamorphic fluids at temperatures from100 to 350 �C (Fig. 12). This isotopic enrichment isrelated to a progressive loss of N in the samples thatcan be modeled by a Rayleigh distillation-type devolatil-ization process (Figs. 13a and b). This implies that wecannot use indiscriminately the d13C value for determin-ing the biogenicity of Archean rocks.

- The large C isotopic fractionation cannot be explainedby metasomatic reactions (as for chert Pi-85-00; Uenoet al., 2004) but requires isotopic exchange between theisotopically heavier C from carbonates and graphitic Cat temperatures of 350 �C. Ueno et al. (2004) did notconsider this process as viable to explain the C isotopicfractionation observed in the North Pole silica veins,because of the low temperature involved. However, stud-ies carried out in the metamorphic domain indicate thatthey are possible down to 270 �C (Morikiyo, 1984).

- Following the devolatilization model illustrated inFig. 13a, the samples containing pristine N and C isoto-pic signatures are those of the North Pole silica veins,particularly PANO D-136-0 that was studied by Pintiet al. (2001). When we study in detail the intra-chert Nand Ar isotopic variability (Fig. 14), both PANO D-136-0 and its companion Pi-47-00 (Fig. 14; this study)show mixing between two sources, the most depleted in15N having d15N values and 40Ar/36Ar ratios compatiblewith those measured in the mantle.

- The hypothesis of an inorganic N mantle overprinting isnot fully compatible with the elemental abundance ratiosbetween C, N and Ar (Fig. 15), although it cannot becompletely ruled out. The N, C and Ar isotopic variabil-ity in North Pole silica veins is best explained by theoccurrence of N and Ar transported by intra-forma-tional fluids having d15N 6 0& and 40Ar/36Ar 6 10,000,as those observed in fluid inclusions in intra-formationalbarite (this study) and associated quartz veins (Nishiza-wa et al., 2007). This source is intimately mixed with asource having d15N 6 �7&, 40Ar/36Ar P 60,000 andpossibly d13C 6 �36&, which is interpreted as organicin origin and derived from the metabolic activity of


chemolithoautotrophs and methanogens in a hydrother-mal environment (Pinti et al., 2001; Ueno et al., 2004,2006). The N/40Ar ratios in black cherts from NorthPole are variable, indicating possibly different NH4

+/Ktime-integrated ratios, but chert PANO D-136-0 has aN/40Ar ratio which encompasses those from present-day MORB. This ambiguity between an organic vs man-tle source for the d15N-depleted nitrogen in Archeancherts needs further experimental work.

- The N and C isotopic composition measured in Apexchert revealed that it was strongly modified by metamor-phic reactions, yet we cannot exclude an organic originfor these two elements. It is likely that the organic15N-13C-depleted source found in North Pole silica veinsoccurred also in Apex chert, then modified by hydrother-mal or metamorphic-induced reactions. Biomass derivedfrom a metabolic activity of hyperthermophiles would bemuch more compatible with the hydrothermal origin ofApex chert (Brasier et al., 2002) rather than occurrenceof mesothermophiles such as cyanobacteria.

- Stromatolite from Strelley Pool (Pi-02-47) and fluid inclu-sions in barite from the Dresser formation, North Pole(Pi-II-08-00) shows d15N from -0.8 to -0.5 &, atmosphericto slightly radiogenic 40Ar/36Ar ratios (301–1350) andN/36Ar ratios 2–4 times higher than the present atmo-sphere. These values are close to those of the Archeanpaleoseawater (d15N � �0.5 ± 1& and N/36Ar � 4 �104), as proposed by Pinti et al. (2001) and Nishizawaet al. (2007). These latter isotopic evidences need to bestudied further in order to determine whether the N andAr isotopic composition of the atmosphere-hydrospherealready reached equilibrium since the Archean.

ACKNOWLEDGMENTS

Criticisms from one anonymous reviewer, R. Kerrich and thehandling editor (J. Horita) greatly improved the paper. Fruitful dis-cussions with M. Jebrak and B. Ghaleb on REE geochemistry weregreatly appreciated. This work was funded by the JSPS Short-TermFellowship (Grant No. L02540) and the NSERC (Grant No.314496-05) to D.L.P. We wish to thank Y. Sano, M. Nishizawaand Y. Ueno for fruitful discussions during DLP visit at TokyoUniversity in 2007 (JSPS Short-Term Fellowship grant no. S-07050). R. Mineau assisted D.L.P. during SEM imagery andEDEX analyzes of chert at the LAMIC laboratory, GEOTOP.M. Laithier kindly drew the geological maps of Pilbara and NorthPole. Grant-in-Aid for Science Research from MEXT (Grant No.14702016) and JSPS (Grant No. 17340168) supported KH. PPacknowledges support from the CNRS-INSU under the Pro-gramme National de Planetologie and GEOMEX. This is GEO-TOP contribution 2009-0006.

REFERENCES

Ader M., Cartigny P., Boudou J. P., Oh J. H., Petit E. and JavoyM. (2006) Nitrogen isotopic evolution of carbonaceous matterduring metamorphism: methodology and preliminary results.Chem. Geol. 232, 152–169.

Alibo D. S. and Nozaki Y. (1999) Rare earth elements in seawater:particle association, shale-normalization, and Ce oxidation.Geochim. Cosmochim. Acta 63, 363–372.

Altabet M. A. and Francois R. (1994) Sedimentary nitrogenisotopic ratio as a recorder for surface ocean nitrate utilization.Global Biogeochem. Cy. 8, 103–116.

Beaumont V. and Robert F. (1999) Nitrogen isotope ratios ofkerogens in Precambrian cherts: a record of the evolution ofatmosphere chemistry? Precambrian Res. 96, 63–82.

Bebout G. E. and Fogel M. L. (1992) Nitrogen isotope composi-tions of metasedimentary rocks in the Catalina Schist, Califor-nia: implications for metamorphic devolatilization history.Geochim. Cosmochim. Acta 56, 2839–2849.

Bolhar R., Van Kranendonk M. J. and Kamber B. S. (2005) Atrace element study of siderite-jasper banded iron formation inthe 3.45 Ga Warrawoona Group, Pilbara Craton - Formationfrom hydrothermal fluids and shallow seawater. Precambrian

Res 137, 93–114.

Boyd S. R., Hall A. and Pillinger C. T. (1993) The measurement ofd15N in crustal rocks by static vacuum mass-spectrometry -Application to the origin of the ammonium in the CornubianBatholith, Southwest England. Geochim. Cosmochim. Acta 57,

1339–1347.

Boyd S. R., Wright I. P. and Pillinger C. T. (1997) Stepped-heatingof carbonates and carbon-bearing quartz grains. Chem. Geol.

134, 303–310.

Boyd S. R. and Philippot P. (1998) Precambrian ammoniumbiogeochemistry: a study of the Moine metasediments. Scot-

land. Chem. Geol. 144, 257–268.

Brasier M. D., Green O. R., Jephcoat A. P., Kleppe A. K., VanKranendonk M. J., Lindsay J. F., Steele A. and Grassineau N.V. (2002) Questioning the evidence for Earth’s oldest fossils.Nature 416, 76–81.

Brasier M. D., Green O. R., Lindsay J. F., McLoughlin N., SteeleA. and Stoakes C. (2005) Critical testing of Earth’s oldestputative fossil assemblage from the 3.5 Ga Apex chert, China-man Creek, Western Australia. Precambrian Res 140, 55–102.

Brauer K. and Hahne K. (2005) Methodical aspects of the 15N-analysis of Precambrian and Palaeozoic sediments rich inorganic matter. Chem. Geol. 218, 361–368.

Buick R. (2007) The earliest records of life on Earth. In In Planets

and Life. The emerging science of Astrobiology (eds. T. S.Woodruff III and J. A. Baross). Cambridge University Press,

Cambridge, pp. 237–264.

Buick R. and Barnes K. R. (1984) Cherts in the WarrawoonaGroup: early Archaean silicified sediments deposited inshallow-water environments. In Archean and Proterozoic basins

of the Pilbara, Western Australia (eds. J. R. Muhling, D. I.Groves and T. S. Blake). Geology Dept. (Key Centre) and

University Extension, The University of Western Australia, pp.

37–53.

Buick R., Thornett J. R., McNaughton N. J., Smith J. B., BarleyM. E. and Savage M. (1995) Record of emergent continentalcrust 3.5 billion years ago in the Pilbara craton in Australia.Nature 375, 574–577.

Busigny V., Ader M. and Cartigny P. (2005) Quantification andisotopic analysis of nitrogen in rocks at the ppm level usingsealed tube combustion technique: a prelude to the study ofaltered oceanic crust. Chem. Geol. 223, 249–258.

Cadogan P. H. (1977) Paleoatmospheric argon in Rhynie chert.Nature 268, 38–40.

Carignan J., Hild P., Mevelle G., Morel J. and Yeghicheyan D.(2001) Routine analyzes of trace element in geological samplesusing flow injection and low pressure on-line liquid chroma-tography coupled to ICP-MS: a study of geochemical referencematerials BR, DR-N, UB-N, AN-G and GH. Geostand.

Newsletter 25, 187–198.

Cartigny P., Harris J. W. and Javoy M. (2001a) Diamond genesis,mantle fractionations and mantle nitrogen content: a study of


d13C-N concentrations in diamonds. Earth Planet. Sci. Lett.

185, 85–98.

Cartigny P., Jendrzejewski N., Pineau F. and Petit E. (2001b)Volatile (C, N, Ar) variability in MORB and the respective rolesof mantle source heterogeneity and degassing: the case of theSouthwest Indian Ridge. Earth Planet. Sci. Lett. 194, 241–257.

Conway N. M., Kennicutt M. C. and Van Dover C. L. (1994)Stable isotopes in the study of marine chemosynthetic-basedecosystems. In Stable isotopes in ecology, environmental science

(eds. K. Lajtha and R. H. Michener). Blackwell Sci. Publ.,

Oxford, pp. 158–186.

Craw D. (2002) Geochemistry of late metamorphic hydrothermalalteration and graphitisation of host rock, Macraes gold mine,Otago Schist, New Zealand. Chem. Geol. 191, 257–275.

Groves D. I., Dunlop J. S. R. and Buick R. (1981) An early habitatof life. Sci. Am. 245, 56–65.

Haendel D., Muhle K., Nitzsche H.-M., Stiehl G. and Wand U.(1986) Isotopic variations of the fixed nitrogen in metamorphicrocks. Geochim. Cosmochim. Acta 50, 749–758.

Hannington M. D., Jonasson I. R., Herzig P. M. and Sven Petersen(2005) Physical and chemical processes of seafloor mineraliza-tion at Mid-Ocean ridges. In Seafloor Hydrothermal Systems(eds. S. E. Humphris, R. A. Zirenberg, L. S. Mullineaux, R. E.Thomson), American Geophysical Union, Geophys. Monogr.91, 115–157.

Hanschamn G. (1981) Berechnung von Isotopieeffekten auf quan-tenchemischer Grundlage am Beispiel stick-stoffhaltiger Mole-kule. ZfI-Mitt. 41, 19–39.

Hashizume K., Sugihara A., Pinti D. L., Orberger B. and WestallF. (2006) Search for primordial biogenic isotopic signatures ofnitrogen in Archean sedimentary rocks. Geochim. Cosmochim.

Acta(Suppl. 1), A235.Hashizume K. and Sugiura N. (1990) Precise measurement of

nitrogen isotopic composition using a quadrupole mass spec-trometer. J. Mass Spectrom. Soc. Jpn. 38, 269–286.

Hashizume K. and Sugiura N. (1995) Nitrogen isotopes in bulkordinary chondrites. Geochim. Cosmochim. Acta. 59, 4057–

4069.

Hashizume K. and Sugiura N. (1997) Isotopically anomalousnitrogen in H-chondrite metal. Geochim. Cosmochim. Acta. 61,

859–872.

Hashizume K., Soyama H., Cloquet C., Pinti D. L., Rouchon V.,Orberger B. and Jayananda M. (2008) Covariation of nitrogenand iron isotopic ratios in a Banded Iron Formation. Geochim.

Cosmochim. Acta(Suppl. 1), A356.Hoefs J. (2004) Stable Isotope Geochemistry. Springer-Verlag,

Berlin, pp. 244.Hofmann H. J., Grey K., Hickman A. H. and Thorpe R. I. (1999)

Origin of 3.45 Ga coniform stromatolites in WarrawoonaGroup, Western Australia. Geol. Soc. Am. Bull 111, 1256–1262.

Hollings P. and Wyman D. (2005) The geochemistry of traceelements in igneous systems: principles and examples frombasaltic systems. In Rare-element geochemistry and mineral

deposits (eds. R. Linnen and I. Samson). Geological Association

of Canada, Short Course 17, 1–17.

Hongo Y. and Nozaki Y. (2001) Rare earth element geochemistryof hydrothermal deposits and Calyptogena shell from the IheyaRidge vent field, Okinawa Trough. Geochem. J. 35, 347–354.

Honma H. and Itihara Y. (1981) Distribution of ammonium inminerals of metamorphic and granitic rocks. Geochim. Cosmo-

chim. Acta 45, 983–988.

Horita J. (2005) Some perspectives on isotope biosignatures forearly life. Chem. Geol. 218, 171–186.

Humphris S. E. and Thompson G. (1978) Trace element mobilityduring hydrothermal alteration of oceanic basalts. Geochim.

Cosmochim. Acta 42, 127–136.

Jia Y. (2006) Nitrogen isotope fractionations during progressivemetamorphism: a case study from the Paleozoic Coomametasedimentary complex, southeastern Australia. Geochim.


Jia Y. and Kerrich R. (2004a) A Reinterpretation of the crustal N-isotope record: evidence for a 15N-enriched Archean atmo-sphere? Terra Nova 16, 102–108.

Jia Y. and Kerrich R. (2004b) Nitrogen 15–enriched Precambriankerogen and hydrothermal systems. Geochem. Geophys. Geo-

syst. 5, Q07005, doi:10.1029/2004GC000716.Kamber B. S., Bolhar R. and Webb G. E. (2004) Geochemistry of

late Archaean stromatolites from Zimbabwe: evidence formicrobial life in restricted epicontinental seas. Precambrian

Res. 132, 379–399.

Kamber B. S., Greig A. and Collerson K. D. (2005) A new estimatefor the composition of weathered young upper continental crustfrom alluvial sediments, Queensland, Australia. Geochim.


Kendrick M. A., Phillips D. and Miller J. M. (2006) Part I.Decrepitation, degassing behavior of quartz up to 1560 �C:analysis of noble gases and halogens in complex fluid inclusionassemblages. Geochim. Cosmochim. Acta 70, 2540–2561.

Kerrich R., Jia Y., Manikyamba C. and Naqvi M. (2006) Secularvariations of N-isotopes in terrestrial reservoirs and oredeposits. In Evolution of Early Earth’s Atmosphere, Hydro-

sphere, and Biosphere—Constraints from Ore Deposits, Memoir

(eds. S.E. Kesler and H. Ohmoto), Geological Society of

America, Boulder 198, 81–104.

Kerridge J. F. (1999) Formation and processing of organics in theearly solar system. Space Sci. Rev. 90, 275–288.

Kitajima K., Maruyama S., Utsunomiya S. and Liou J. G. (2001)Seafloor hydrothermal alteration at an Archaean mid-oceanridge. J. Metamorph. Geol. 19, 581–597.

Kiyokawa S., Ito T., Ikehara M. and Kitajima F. (2006) MiddleArchean volcano-hydrothermal sequence. Bacterial microfossil-bearing 3.2 Ga Dixon Island Formation, coastal Pilbaraterrane, Australia. Bull. Geol. Soc. Am 118, 3–22.

Kung C. C., Hayatsu R., Studier M. H. and Clayton R. N. (1979)Nitrogen isotope fractionations in the Fischer-Tropsch andUrey-Miller reaction. Earth Planet. Sci. Lett. 46, 141–146.

L’Vov B. V. and Novichikhin A. V. (1997) Quantitative interpre-tation of the evaporation coefficients for the decomposition orsublimation of some substances in vacuo. Thermochimica Acta

290, 239–251.

Lancet M. S. and Anders E. (1970) Carbon Isotope Fractionationin the Fischer-Tropsch Synthesis and in Meteorites. Science

170, 980–982.

Lowe D. R. (1983) Restricted shallow-water sedimentation of EarlyArchean stromatolitic and evaporitic Strata of the Strelley PoolChert, Pilbara Block, Western Australia. Precambrian Res. 19,

239–283.

Lowe D. R. (1994) Abiological origin of described stromatolitesolder than 3.2 Ga. Geology 22, 387–390.

Marty B. (1995) Nitrogen content of the mantle inferred from N2-Ar correlation in oceanic basalts. Nature 377, 326–329.

Marty B. and Zimmermann L. (1999) Volatiles (He, C, N, Ar) inmid-ocean ridge basalts: assessment of shallow-level fraction-ation and characterization of source composition. Geochim.


McCollom T. M. and Seewald J. S. (2006) Carbon isotopecomposition of organic compounds produced by abiotic syn-thesis under hydrothermal conditions. Earth Planet. Sci. Lett.

243, 74–84.

Metz S. and Trefry J. H. (2000) Chemical and mineralogicalinfluences on concentrations of trace metals in hydrothermalfluids. Geochim. Cosmochim. Acta 64, 2267–2279.

http://dx.doi.org/10.1029/2004GC000716


Mills R. A. and Elderfield H. (1995) Rare earth element geochem-istry of hydrothermal deposits from the active TAG Mound,26�N Mid-Atlantic Ridge. Geochim. Cosmochim. Acta 59, 3511–

3524.

Mineau R., Pinti D. L. and Valentin C. (2008) Do we really knowapex chert? newly identified micro- and dubiofossils. Geochim.

Cosmochim. Acta 72 (Suppl. 1), A633.

Moreira M., Kunz J. and Allegre C. (1998) Rare gas systematics inpopping rocks: isotopic and elemental compositions in theupper mantle. Science 279, 1178–1180.

Morikiyo T. (1984) Carbon isotopic study on coexisting calcite andgraphite in the Ryoke metamorphic rocks, northern Kisodistrict, central Japan. Contrib. Mineral. Petrol. 87, 251–

259.

Nesbitt H. W., Young G. M., McLennan S. M. and Keays R. R.(1996) Effects of chemical weathering and sorting on thepetrogenesis of siliciclastic sediments, with implications forprovenance studies. J. Geol. 104, 525–542.

Nijman W., De Bruijne K. C. H. and Valkering M. E. (1998)Growth fault control of Early Archaean cherts, barite moundsand chert–barite veins, North Pole Dome, eastern Pilbara,Western Australia. Precambrian Res. 88, 25–52.

Nisbet E. G. and Sleep N. H. (2001) The habitat and nature ofearly life. Nature 409, 1083–1091.

Nishizawa M., Sano Y., Ueno Y. and Maruyama S. (2007)Speciation and isotope ratios in fluid inclusions from seafloorhydrothermal deposits at ca. 3.5 Ga. Earth Planet. Sci. Lett 254,

332–344.

Ohmoto H. (1997) When did the Earth’s atmosphere become oxic?The Geochemical News 93, 26–27.

Orberger B., Rividi N. and Hofmann A. (2006a) Origin of blackcherts from the Marble Bar Drill Core #1 (Pilbara, W-Australia). Geochim. Cosmochim. Acta(Suppl. 1), A461.

Orberger B., Rouchon V., Westall F., devries S. T., Pinti D. L.,Wagner C., Wirth R. and Hashizume K. (2006b) Microfaciesand origin of some Archaean cherts (Pilbara, Australia). InProcesses on the Early Earth (eds. W. U. Reimold and R. L.Gibson). Geological Society of America, Boulder, Spec. Publ,

pp. 133–156, 405.

Papineau D., Mojzsis S. J., Karhu J. A. and Marty B. (2005)Nitrogen isotopic composition of ammoniated phyllosilicates:case studies from Precambrian metamorphosed sedimentaryrocks. Chem. Geol. 216, 37–58.

Peters K. E., Sweeney R. E. and Kaplan I. R. (1978) Correlation ofcarbon and nitrogen stable isotope ratios in sedimentaryorganic matter. Limnol. Oceanogr. 23, 598–604.

Pineau F. and Javoy M. (1994) Strong degassing at ridge crests: thebehavior of dissolved carbon and water in basalt glasses at14�N, Mid-Atlantic Ridge. Earth Planet. Sci. Lett. 123, 179–

198.

Pinti D. L. and Hashizume K. (2001) 15N-depleted nitrogen inEarly Archean kerogens: clues on ancient marine chemosyn-thetic-based ecosystems? A comment to Beaumont, V., Robert,F., 1999. Precambrian Res. 96, 62–82. Precambrian Res 105,

85–88.

Pinti D. L., Hashizume K. and Matsuda J. (2001) Nitrogen andargon signatures in 3.8–2.8 Ga metasediments: clues on thechemical state of the Archean ocean and the deep biosphere.Geochim. Cosmochim. Acta 65, 2301–2315.

Pinti D. L., Hashizume K., Orberger B., Gallien J. -P., Cloquet C.and Massault M. (2007) Biogenic nitrogen and carbon in Fe-Mn-oxyhydroxides from an Archean chert, Marble Bar,Western Australia. Geochem. Geophys. Geosyst. 8,doi:10.1029/2006GC001394.

Raquin A., Moreira M. A. and Guillon F. (2008) He, Ne and Arsystematics in single vesicles: mantle isotopic ratios and origin

of the air component in basaltic glasses. Earth Planet. Sci. Lett.

274, 142–150.

Robert F. and Halbout J. (1990) A numerical model for isotopicpatterns from thermal-extraction experiments. Meteoritics 25,

291–299.

Rollinson H. (1993) Using geochemical data. Pearson-Prentice Hall,London, 352 pp.

Sano Y. and Pillinger C. T. (1990) Nitrogen isotopes and N2/Arratios in cherts: an attempt to measure time evolution ofatmospheric d15N value. Geochem. J. 24, 315–325.

Sano Y., Nagao K. and Pillinger C. T. (1994) Carbon and noblegases in Archean chert. Chem. Geol. 112, 327–342.

Sano Y., Takahata N., Nishio Y., Fischer T. P. and Williams S. N.(2001) Volcanic flux of nitrogen from the Earth. Chem. Geol.

171, 263–271.

Sarda P., Staudacher T. and Allegre C. J. (1985) 40Ar/36Ar inMORB glasses: constraints on atmosphere and mantle evolu-tion. Earth Planet. Sci. Lett. 72, 357–375.

Scheele N. and Hoefs J. (1992) Carbon isotope fractionationbetween calcite, graphite and CO2: an experimental study.Contrib. Mineral. Petrol. 112, 35–45.

Schopf J. W. (1993) Microfossils of the Early Archean Apex chert -New evidence of the antiquity of Life. Science 260, 640–646.

Schopf J. W. (2006) Fossil evidence of Archaean life. Phil. Trans.

R. Soc. B 361, 869–885.

Shen Y., Pinti D. L. and Hashizume K. (2006) Biogeochemicalcycles of sulfur and nitrogen in the Archean ocean andatmosphere. In Archean Geodynamics and Environments,(eds. K. Benn, J. C. Mareschal and K. C. Condie). AmericanGeophysical Union, Washington D.D. Geophys. Monogr 164,

305–320.

Sommer II M. A. and Gibson Jr. E. K. (1986) Volatile determi-nations of individual fluid inclusions within the 3.4 B.Y. NorthPole Barites from the Warrawoona Group, NorthwesternAustralia. Lunar Planet. Sci. XVII, Lunar. Planet. Inst.,Houston. pp. 815-816 (abstr.).

Strauss H. and Moore T. B. (1992) Abundances and isotopiccompositions of carbon and sulfur species in whole rock andkerogen samples. In Biosphere: A Multidisciplinary Study (eds.J. W. Schopf and C. Klein). Cambridge University Press,

Cambridge, pp. 709–798.

Svensen H., Bebout G., Kronz A., Li L., Planke S., Chevallier L.and Jamtveit B. (2008) Nitrogen geochemistry as a tracer offluid flow in a hydrothermal vent complex in the Karoo Basin,South Africa. Geochim. Cosmochim. Acta 72, 4929–4947.

Taylor S. R. and McLennan S. M. (1981) The composition andevolution of the continental crust: rare earth element evidencefrom sedimentary rocks. Phil. Trans. R. Soc. Lond. A 301, 381–

399.

Ueno Y., Yurimoto H., Yoshioka H., Komiya T. and MaruyamaS. (2002) Ion microprobe analysis of graphite from ca. 3.8 Gametasediments, Isua supracrustal belt, West Greenland: rela-tionship between metamorphism and carbon isotopic compo-sition. Geochim. Cosmochim. Acta 66, 1257–1268.

Ueno Y., Yoshioka H., Maruyama S. and Isozaki Y. (2004)Carbon isotopes and petrography of kerogens in similar to3.5 Ga hydrothermal silica dikes in the North Pole area,Western Australia. Geochim. Cosmochim. Acta 68, 573–589.

Ueno Y., Yamada K., Yoshida N., Maruyama S. and Isozaki Y.(2006) Evidence from fluid inclusions for microbial methano-genesis in the early Archaean era. Nature 440, 516–519.

Valley J. W. and O’Neil J. R. (1981) 13C/12C exchange betweencalcite and graphite: a possible thermometer in Grenvillemarbles. Geochim. Cosmochim. Acta 45, 411–419.

Van den Boorn S., Van Bergen M., Nijman W. and Vroon P.(2007) Dual role of seawater and hydrothermal fluids in Early

http://dx.doi.org/10.1029/2006GC001394


Archean chert formation: evidence from silicon isotopes.Geology 35, 939–942.

Van Kranendonk M. J. (2006) Volcanic degassing, hydrothermalcirculation and the flourishing of early life on Earth: A reviewof the evidence from c. 3490–3240 Ma rocks of the PilbaraSupergroup, Pilbara Craton, Western Australia. Earth Sci. Rev.

74, 197–240.

Van Kranendonk M. J., Hickman A. H., Williams I. R., andNijman W. (2001). Archaean geology of the East Pilbara granite-

greenstone terrane, Western Australia–A field guide Record

2001/9. Geological Survey of Western Australia, Perth.Van Kranendonk M. J., Webb G. E. and Kamber B. S. (2003)

Geological and trace element evidence for a marine sedimentaryenvironment of deposition and biogenicity of 3.45 Ga stromat-olitic carbonates in the Pilbara Craton, and support for areducing Archaean Ocean. Geobiology 1, 91–108.

Van Kranendonk M. J., Hugh Smithies R., Hickman A. H. andChampion D. C. (2007) Review: secular tectonic evolution ofArchean continental crust: interplay between horizontal andvertical processes in the formation of the Pilbara Craton,Australia. Terra Nova 19, 1–38.

Van Zuilen M. A., Mathew K., Wopenka B., Lepland A., Marti K.and Arrhenius G. (2005) Nitrogen and argon isotopic signa-tures in graphite from the 3.8-Ga-old Isua Supracrustal Belt,Southern West Greenland. Geochim. Cosmochim. Acta 69,

1241–1252.

van Zuilen M. A., Chaussidon M., Rollion-Bard C. and Marty B.(2007) Carbonaceous cherts of the Barberton Greenstone Belt,South Africa: isotopic, chemical and structural characteristicsof individual microstructures. Geochim. Cosmochim. Acta 71,

655–669.

Von Damm K. L. (1995) Controls on the chemistry and temporalvariability of seafloor hydrothermal fluids. In Seafloor Hydro-thermal Systems (eds. S. E. Humphris, R. A. Zirenberg, L. S.Mullineaux, R. E. Thomson), American Geophysical Union,Geophys. Monogr. 91, 222–247.

Wada H. and Suzuki K. (1982) Carbon isotopic thermometrycalibrated by dolomite-calcite solvus temperatures. Geochim.


Walter M. R., Buick R. and Dunlop J. S. R. (1980) Stromatolites 3,400–3,500 Myr old from the North Pole area, WesternAustralia. Nature 284, 443–445.

Webb G. E. and Kamber B. S. (2000) Rare earth elements inHolocene reefal microbialites: a new shallow seawater proxy.Geochim. Cosmochim. Acta 64, 1557–1565.

Yamamoto T., Hashizume K., Matsuda J. and Kase T. (1998)Multiple nitrogen isotopic components co-existing in ureilites.Meteoritics Planet. Sci. 33, 857–870.

Associate editor: J. Horita

Isotopic fractionation of nitrogen and carbon in the synthesis of amino acids by microorganisms

Documents