Classification of Archean Granitoids

Lithos 205 (2014) 208–235
Contents lists available at ScienceDirect
Lithos
journal homepage: www.elsevier.com/locate/lithos
Invited review article
The diversity and evolution of late-Archean granitoids: Evidence for the

onset of “modern-style” plate tectonics between 3.0 and 2.5 Ga
O. Laurent a,b,c,⁎, H. Martin a,b,c, J.F. Moyen a,b,c,d, R. Doucelance a,b,c
a
Clermont Université, Université Blaise Pascal, Laboratoire Magmas et Volcans, BP 10448, F-63000 Clermont-Ferrand, France
b
CNRS, UMR6524, LMV, F-63038 Clermont-Ferrand, France
c
IRD, R 163, LMV, F-63038 Clermont-Ferrand, France
d
Département de Géologie, Université Jean Monnet, 23 rue du Docteur Paul Michelon, 42023 Saint-Étienne, France
a r t i c l e i n f o a b s t r a c t
Article history: The end of the Archean aeon (3.0–2.5 Ga) was a period of fundamental change in many aspects of the geological
Received 11 December 2013 record. In Archean cratons, this timespan is marked by a considerable diversification in both the nature and
Accepted 15 June 2014 petrogenesis of granitoid rocks. In this article, we review the nature, petrogenesis and global evolution of late-
Available online 24 June 2014
Archean granitoids and discuss their geodynamic significance.
Late-Archean granitoids can be classified into four groups: (1) volumetrically-dominant and juvenile tonalites,
Keywords:
Archean–Proterozoic transition
trondhjemites and granodiorites (TTGs), whose geochemistry is consistent with an origin through partial melting
Continental crust of meta-igneous mafic rocks at various pressures; (2) Mg-, Fe- and K-rich, metaluminous (monzo)diorites and
Granitoids granodiorites, referred to as sanukitoids s.l., which derive primarily from hybridization between mantle perido-
Geochemistry tite and a component rich in incompatible elements; (3) peraluminous and K-rich biotite- and two-mica granites,
Geodynamics formed through melting of older crustal lithologies (TTGs and meta-sediments, respectively); and (4) hybrid
Plate tectonics high-K granites with mixed characteristics from the first three groups.
The chronology of granitoid emplacement in late-Archean times is different from one craton to another but, in
general, follows a very specific two-stage sequence: (1) a long period (0.2–0.5 Ga) of TTG emplacement; (2) a
shorter period (0.02–0.15 Ga) during which all other granitoid types were generated. We propose that this
sequence represents the first global subduction–collision cycle in the Earth's history. Although possibly present
in the geological record prior to 3.0 Ga, such mechanisms became progressively prevalent on a planetary scale
only between 3.0 and 2.5 Ga, indicating that the late-Archean geodynamic changes resulted from the global ini-
tiation of “modern-style” plate tectonics. The Archean–Proterozoic transition thus represents a major change in
the mechanisms of the Earth's heat loss: before 3.0–2.5 Ga, it took place by large-scale magmatic differentiation
characterized by generation of proto-continents that underwent crustal maturation locally, but without obvious
cyclic activity on a planetary scale. After this, heat loss became accommodated by plate tectonics and global
Wilson subduction–collision cycles. These changes were the consequence of the Earth's cooling, which in turn
controlled a number of different parameters locally (thickness, temperature, volume and rheology of the
crust). This explains why the changes took place over a short timespan (~0.5 Ga) relative to the Earth's history,
but at different times and with different characteristics from one craton to another.
© 2014 Elsevier B.V. All rights reserved.
Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
2. The diversity of late-Archean granitoids: a petrogenetic classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
2.1. TTGs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
2.1.1. Geology and petrography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
2.1.2. Key geochemical features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
⁎ Corresponding author at: Department of Mineralogy, Institute of Geosciences, Johann Wolfgang Goethe-University, Altenhöferallee 1, 60438 Frankfurt-am-Main, Germany. Tel.: +49
69 798 40133.
E-mail addresses: laurent@em.uni-frankfurt.de, oscarlaurent86@gmail.com (O. Laurent).
http://dx.doi.org/10.1016/j.lithos.2014.06.012
0024-4937/© 2014 Elsevier B.V. All rights reserved.
O. Laurent et al. / Lithos 205 (2014) 208–235 209
2.1.3. Petrogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211

2.2. Sanukitoids s.l. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
2.2.3. Petrogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
2.3. Biotite- and two-mica granites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
2.3.3. Petrogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
2.4. Hybrid granitoids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
2.4.3. Petrogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
2.5. A synthetic classification diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
3. Late-Archean crustal evolution: a two-stage sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
3.1. Evidence from geochronology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
3.2. Evidence from isotope data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
4. Geodynamic implications: late-Archean granitoid evolution reflecting subduction and continental collision . . . . . . . . . . . . . . . . . . . 223
4.1. First stage: genesis of TTGs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
4.1.1. Possible geodynamic environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
4.1.2. Late-Archean TTGs formed in a subduction setting: geometric and geochemical constraints . . . . . . . . . . . . . . . . . . 223
4.1.3. The meaning of “subduction” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
4.2. Second stage: genesis of sanukitoids s.l., biotite-, two-mica granites and hybrids . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
4.2.1. Possible geodynamic environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
4.2.2. Continent–continent collision: analogy with post-Archean collisional granitoids . . . . . . . . . . . . . . . . . . . . . . . 225
4.2.3. The meaning of “continent–continent collision” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
4.3. A global geodynamic model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
5. The onset of “modern-style” plate tectonics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
5.1. Comparison with Eo- to Mesoarchean granitoids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
5.2. Support from other geological evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
5.3. Possible origins of the changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
6. Summary and concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
Appendix A. Supplementary data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
1. Introduction Schematically, an Archean crustal segment consists of three

contrasting lithologies: (i) the so-called “grey gneisses”, a complex,
The temporal changes in the mechanisms of continental crust deformed and often migmatitic assemblage of meta-sedimentary and
extraction from the mantle are a consequence of the Earth's cooling meta-igneous rocks, whose major components (Moyen, 2011) are
and differentiation. The genesis of continental material started very low-K granitoid gneisses of the tonalite–trondhjemite–granodiorite
early in the Earth's history, as demonstrated by the Hadean ages (“TTG”; Jahn et al., 1981) series; (ii) supracrustal, meta-volcanic
(4.0–4.4 Ga) of the Acasta gneisses (north-western Canada; Bowring and meta-sedimentary rocks (generally recording greenschist- to
and Williams, 1999; Iizuka et al., 2006) and the Jack Hills zircons (West- amphibolite-facies metamorphism) forming the so-called green-
ern Australia; Wilde et al., 2001). Continental crust has been formed stone belts; and (iii) late and generally high-K granitoid plutons
continuously from that time, but both the growth rates and mecha- and batholiths, typically intruding the two former lithologies. These
nisms have changed over time. The most important changes took granitoids generally represent the last Archean geological event in
place between 3.0 and 2.5 Ga, at the very end of the Archean aeon. In- every craton worldwide (Condie, 1981; Heilimo et al., 2011; Martin,
deed, N 50% of the present-day continental volume formed prior to 1993; Moyen et al., 2003; Sylvester, 1994). Although this event is
2.5 Ga (Armstrong, 1981; Belousova et al., 2010; Dhuime et al., 2011; diachronous from one craton to another, it took place between 3.0
Guitreau et al., 2012; Kramers and Tolstikhin, 1997; Reymer and and 2.5 Ga ago on a planetary scale, and was followed by a ca.
Schubert, 1984; Taylor and McLennan, 1985, 1995) and the Archean 0.25 Ga-long global magmatic shutdown (Condie et al., 2009). There-
continental crust differs from the present day one in terms of petro- fore the late-Archean high-K granitic magmatism appears to be
graphic and chemical compositions (Condie, 1981, 1989; Keller and related to the final stabilization of the Archean cratonic lithosphere,
Schoene, 2012; Kemp and Hawkesworth, 2003; Martin, 1987; Taylor which testifies to the geodynamic changes of that period.
and McLennan, 1985, 1995), P–T conditions of metamorphic provinces Late-Archean granitoids cover a wide range of petrographic and
(Brown, 2006; Percival, 1994), geometry and lithology of volcano- mineralogical compositions (e.g. Champion and Sheraton, 1997;
sedimentary basins (Condie, 1981; Taylor, 1987; Veizer and McKenzie, Champion and Smithies, 1999; Dall'Agnol et al., 2006; Feng and Kerrich,
2003), and structures and rheology (Cagnard et al., 2011; Chardon 1992; Frost et al., 1998; Kampunzu et al., 2003; Käpyaho et al., 2006;
et al., 2009; Thébaud and Rey, 2013). Consequently, the end of the Mikkola et al., 2011b; Moyen et al., 2003; Sage et al., 1996; Schofield
Archean has long been regarded as a period of fundamental geodynamic et al., 2010; Sylvester, 1994; Yang et al., 2008). The most commonly
changes (Condie and O'Neill, 2010; Taylor and McLennan, 1995; described groups are (i) high-K calc-alkaline, slightly peraluminous
Windley, 1984), but the nature and origin of the changes are still (A/CNK = 1.0–1.1) biotite-bearing granites, and (ii) metaluminous
controversial. (A/CNK b 1.0), Mg- and K-rich (Mg# N 0.45; 1.5 b K2 O b 6.0 wt.%)
210 O. Laurent et al. / Lithos 205 (2014) 208–235
rocks of the sanukitoid suite (monzodiorites, monzogabbros and their Limpopo mobile belt). Late-Archean granitoids from this area, as
differentiated products). Both groups were found in almost all Archean well as from most Archean cratons worldwide, are characterized by a
terranes worldwide, but show some internal heterogeneity and com- very specific temporal evolution on which we base our attempt to deci-
plexity (e.g. Laurent et al., 2011; Martin et al., 2009; Meyer et al., pher both the nature and origin of the geodynamic changes that took
1994; Moyen et al., 2003; Sage et al., 1996; Whalen et al., 2004). place between 3.0 and 2.5 Ga.
Other, less common granitoid types, such as two-mica peraluminous
(A/CNK N 1.1) leucogranites and metaluminous to peralkaline syenites 2. The diversity of late-Archean granitoids: a petrogenetic
and “A-type” granites, were also found in some cratons. Recent work re- classification
ports other late-Archean granitoids which cannot be ascribed to any of
these groups, either because of mixed characteristics (Almeida et al., Compiling published data in order to establish a global “descriptive”
2010; Jayananda et al., 2006; Mikkola et al., 2011a,2011b; Prabhakar classification of late-Archean granitoids (i.e. based on petrography and/
et al., 2009) or local particularities (Laurent et al., 2014; Steenfelt or geochemistry) is a difficult task, severely hindered by (i) the internal
et al., 2005). On the other hand, TTGs are not restricted to early Archean variability of each group, owing to local conditions and to the potentially
times; they also represent a large fraction of the granitoids emplaced heterogeneous composition of available sources; (ii) local terminologies
between 3.0 and 2.5 Ga, adding even more complexity to the late- that make comparisons difficult; and (iii) the frequent compositional
Archean granitoid record. overlap between the different groups. Therefore, in this paper we pro-
Several studies have been conducted locally in order to identify, in a pose to build a classification for late-Archean granitoids based on both
given cratonic area, the composition of the source and the processes the nature of their sources and mechanisms of their petrogenesis. As
that led to the genesis of late-Archean granitoids, as well as their such, it should theoretically not be used prior to careful petrogenetic
temporal evolution (e.g. Almeida et al., 2013; Champion and Sheraton, modelling, although in most cases the granitoids belonging to a given
1997; Feng and Kerrich, 1992; Halla et al., 2009; Hill et al., 1992; Jahn group share several unique characteristics in terms of petrography
et al., 1988; Jayananda et al., 2000; Käpyaho et al., 2006; Mikkola and/or geochemistry, making them easy to identify.
et al., 2011b; Moyen et al., 2001; Oliveira et al., 2011; Percival et al., This “genetic” typology is based on a review of available data and
2006; Smithies and Champion, 2000; Whalen et al., 2004). However, models (see summary in Table 1). These discriminate late-Archean
despite efforts to formally define and classify late-Archean granitoids granitoids into four different groups:
(e.g. Heilimo et al., 2010; Martin et al., 2009; Moyen et al., 2003),
(1) sodic granitoids (TTGs) derived from melting of hydrous
there is a lack of global perspective about the significance of this par-
metabasalts at various depths;
ticular magmatism: the origin of granitoid diversity and the reason
(2) sanukitoids s.l., i.e. magmas resulting from interactions between
why they all were emplaced over a short timespan across all the cra-
the mantle and a component rich in incompatible elements, and
tons remain unclear. The present review aims to address these issues
their differentiated products;
by proposing a schematic classification of late-Archean granitoids
(3) biotite- and two-mica granites generated by partial melting of
based on their petrogenesis. This classification is mainly supported
pre-existing, felsic continental crust;
by examples from the Archean granitoid basement of South Africa,
(4) hybrid granitoids, generally granodiorites and granites, formed
where the transition from Archean to Proterozoic is well exposed be-
through interaction (e.g. metasomatism, mingling, mixing) be-
tween the Zimbabwe and Kaapvaal cratons (Pietersburg block and
tween magmas or sources of any of the former three groups.
Kaapvaal Craton Limpopo Belt In the following, we review the key geological and geochemical
Unexposed Unexposed traits of each group as well as their petrogenesis, based on examples
Granitoids / Greenstone belts Granulites (undifferentiated) from the northern Kaapvaal craton and the Limpopo Belt (South
Sedimentary basins Africa). The northernmost part of the Kaapvaal craton, known as the
~2.0 Ga-old Bushveld Pietersburg block (Fig. 1; de Wit et al., 1992; Eglington and Armstrong,
layered complex Zimbabwe
Craton 2004; Zeh et al., 2009), was tectonically and magmatically accreted
Crust-scale structure Fig. 2 between 3.1 and 2.7 Ga to the northern margin of a 3.6–3.2 Ga-old
nucleus of continental crust (Block et al., 2013; de Wit et al., 1992;
Central Zone e SZ
ipis Eglington and Armstrong, 2004; Laurent et al., 2013b; Poujol et al.,
Tsh
Pietersburg 2003; Zeh et al., 2009, 2013), now represented by the Swaziland and
block Witwatersrand blocks (Fig. 1). Subsequently, collision between the
N -Zoe
-
Palalatein SZ
tfon
inea
me
nt
Kaapvaal and Zimbabwe cratons resulted in the formation of the Limpo-
po mobile belt between 2.7 and 2.5 Ga, with reactivation at, or continu-
nL
Kimberley rc hiso ation until, ca. 2.0 Ga (Barton et al., 2006; de Wit et al., 1992; Eriksson
u
i-M
block azi mb et al., 2009; Kramers and Houri, 2011; Kramers et al., 2006; Roering
Thab
et al., 1992; van Reenen et al., 1987; Zeh et al., 2009 and references
t therein). Therefore, the Pietersburg block and Limpopo Belt consist
JOHANNESBURG en
of a portion of Archean crust formed over ~ 1 Ga across the Archean–
am
ine
Proterozoic boundary and intruded by large volumes of 3.1 to 2.6 Ga-

tonL
d old granitoid rocks (Fig. 2; Laurent et al., 2013b; Poujol et al., 2003;
er
an
rb
zil ck Zeh et al., 2009), which make it ideal for illustrating the present study.
ck nd
Ba
Lineament
AFR IC A a
Sw blo
blotersra
2.1. TTGs
twa
Colesburg
100 km TTGs are the dominant lithology of Archean crust, and consist of
Wi
silicic, plagioclase-rich plutonic rocks (tonalites, trondhjemites and

granodiorites) with a sodic composition (K2O/Na2O b 1) and resulting
Fig. 1. Schematic geological map of the Kaapvaal Craton (yellow) and the Limpopo Belt
from differentiation of tholeiitic and hydrous mafic rocks (for recent
(red) of southern Africa, after Eglington and Armstrong (2004) and Poujol et al. (2003). and detailed review about TTGs see Halla et al., 2009; Moyen, 2011;
The grey rectangle represents the area covered by Fig. 2. Moyen and Martin, 2012; Martin et al., 2014). TTGs are ubiquitous and
abundant in all Archean cratons, even in b 3.0 Ga-old terranes where 2.1.3. Petrogenesis
they are still the volumetrically dominant rock type (Table 1). For exam- It is widely accepted that TTGs were generated by differentiation of a
ple more than half the surface of the Pietersburg block, in South Africa, basaltic parent, either by partial melting (e.g. Atherton and Petford,
consists of grey gneisses, themselves largely made up of TTGs (Fig. 2). 1993; Barker and Arth, 1976; Ellam and Hawkesworth, 1988; Foley
et al., 2002; Martin, 1986, 1987, 1994; Martin and Moyen, 2002;
Martin et al., 2014; Rapp et al., 1991, 2003) or fractional crystallization
2.1.1. Geology and petrography (e.g. Kamber et al., 2002; Kleinhanns et al., 2003). Moyen and Martin
In the field TTGs consist either of several banded, deformed and/or (2012) stressed out that a number of observations are at odds with frac-
migmatized magmatic phases intruding each other (Fig. 3a), or of a sin- tional crystallization (notably the lack of any mafic and intermediate
gle, homogeneous orthogneiss (Fig. 3b). In some places TTGs can form phases in TTG suites), pointing to partial melting of a meta-basalt, leav-
well-defined plutons where the rocks are weakly- to un-deformed, for ing a garnet amphibolitic or eclogitic residue, as a more likely origin.
instance around the Murchison greenstone belt in the Pietersburg This hypothesis is consistent with the major element geochemistry of
block (Fig. 2; Baderoukwe and Maranda intrusions; Poujol et al., 1996; the Pietersburg block TTG samples (Fig. 6).
Poujol, 2001). The internal diversity within Archean TTGs could have resulted from
TTGs are typically characterized by high modal contents of sodic pla- subsequent subordinate fractional crystallization (Martin, 1987), but
gioclase and quartz with respect to alkali feldspar. For example, TTGs of recent studies have shown that it depends primarily on the pressure
the Pietersburg block consist of fine- to medium-grained (1–5 mm) and temperature conditions of melting (Almeida et al., 2011; Halla
trondhjemites and granodiorites, made up of an assemblage of sodic et al., 2009; Moyen, 2011; Moyen and Martin, 2012; Moyen and
(An15–30) plagioclase (~50%), quartz (~30%) and biotite (5–15%), with Stevens, 2006). Many late-Archean TTGs display high Sr and low Y–
subordinate K-feldspar (b 5%). Accessory minerals are magnetite, apa- HREE contents (Table 2; Fig. 5), thus belonging to the “high-” and
tite, epidote (pistacite and/or allanite) and zircon. No hornblende was “medium-pressure” groups of Moyen (2011). Such magmas derived
indentified in TTGs from the Pietersburg block, although in other places from melting of metabasaltic sources at ≥10 kbar, which roughly corre-
it can be present in the most mafic rocks (Moyen and Martin, 2012). sponds to depths N35 km. Such conditions allow for the presence of gar-
net in the residue, accounting for low HREE contents, whilst being
beyond the stability field of plagioclase, explaining the TTGs' richness
2.1.2. Key geochemical features in Al2O3, CaO and Sr and the lack of negative Eu anomaly. As noted
The major-element geochemistry of TTGs is well represented by sam- above, a small fraction of late-Archean TTGs show low Sr, Al2O3 and
ples from the Pietersburg block in South Africa (Table 2; Fig. 4). TTGs are CaO contents and less fractionated REE patterns, locally with a negative
calcic to calc-alkaline, slightly peraluminous (1.0 ≤ A/CNK ≤ 1.1) silica- Eu anomaly (e.g. Almeida et al., 2011; Althoff et al., 2000; Halla et al.,
rich magmatic rocks (65 ≤ SiO2 ≤ 75 wt.%) with low contents of ferro- 2009), which pinpoints melting at shallow levels in the plagioclase sta-
magnesian oxides (FeOt + MgO + MnO + TiO2 ≤ 5 wt.%; Fig. 4a–c). bility field and a garnet-free or garnet-poor residue.
Typically, they are K-poor and Na-rich, leading to low K2O/Na2O ratios
(≤ 0.5; Fig. 4b). CaO contents range from 1 to 5 wt.%, resulting in 2.2. Sanukitoids s.l.
moderate CaO/(Na2O + K2O) ratios (Fig. 4d).
Contrastingly, trace element contents in TTGs show significant This group encompasses a wide diversity of rocks, which share a
variations depending on the pressure at which their meta-basaltic common petrogenesis: they derive partially or totally from interactions
source melted. This issue has been discussed in detail by Almeida between a mantle peridotite and a component rich in incompatible
et al. (2011), Halla et al. (2009), Moyen (2011) and Moyen and elements. This includes the sanukitoids s.s., first described in the
Martin (2012). They broadly range between a “high-HREE” or “low- Superior Province (Bédard, 1996; Shirey and Hanson, 1984; Stern and
pressure” end-member (high HREE–Y and low Sr contents) and a Hanson, 1991; Stevenson et al., 1999; Sutcliffe et al., 1990) and recently
“low-HREE” or “high-pressure” one (low HREE–Y and high Sr contents). redefined on the basis of geochemical criteria by Heilimo et al. (2010)
It has been proposed that late-Archean TTGs (b3.0 Ga-old) are richer in and Martin et al. (2005, 2009). Those were further described in
Sr and poorer in Y–HREE than the older TTGs (Champion and Smithies, most late-Archean terranes (see Table 1), including the Slave Craton
2007; Martin and Moyen, 2002), thus belonging to the “high-pressure” (Davis and Hegner, 1992), the Wyoming Province (Frost et al., 1998),
group. This is well illustrated by the composition of the Pietersburg the Baltic shield (Bibikova et al., 2005; Halla, 2005; Halla et al., 2009;
TTGs (Table 2; Fig. 5). These are moderately rich in incompatible Heilimo et al., 2010; Käpyaho et al., 2006; Kovalenko et al., 2005;
elements (1 to 10 times the primitive mantle values), with negative Lobach-Zhuchenko et al., 2005, 2008; Samsonov et al., 2005), the
Nb and Ta anomalies, a positive Sr anomaly and a typically fractionated Dharwar Craton of South India (Jayananda et al., 1995; Krogstad et al.,
REE pattern (Fig. 5a), due to low HREE concentrations (Yb ≤ 1.5 ppm; 1995; Moyen et al., 2001, 2003; Sarvothaman, 2001), the North China
LaN/YbN N 30). Y contents are also low (≤ 10 ppm), leading to high Craton (Jahn et al., 1988; Wang et al., 2009; Yang et al., 2008), the
Sr/Y (20–200; Fig. 5b). TTGs also show moderate contents in HFSE Pilbara Craton in Western Australia (Smithies and Champion, 1999a,
(1 ≤ Nb ≤ 7 ppm; 45 ≤ Zr ≤ 245 ppm) and REE, and are poor in tran- 2000), the Amazonian Craton (Althoff, 1996; Leite et al., 2004;
sition elements (V b 20 ppm; Fig. 5c–d). Their REE patterns generally Medeiros and Dall'Agnol, 1988; Oliveira et al., 2009) and southern
lack any significant Eu anomaly (Fig. 5d). Africa (Kampunzu et al., 2003; Laurent et al., 2011; Zhai et al., 2006).
On the other hand, some studies also reported “low-” and “medium- This group also includes other late-Archean rocks that differ from
pressure” TTGs of late-Archean age, for example in the Amazonian cra- sanukitoids s.s. in compositional detail, but whose petrogenesis is very
ton (Almeida et al., 2011; Althoff et al., 2000), the Baltic shield (Halla similar (see Section 2.2.2). Such rocks were described as quartz diorites
et al., 2009; Mikkola et al., 2011b) and the Superior Province (Feng in the North Atlantic Craton (Steenfelt et al., 2005) and the Baltic shield
and Kerrich, 1992; Sage et al., 1996; Whalen et al., 2002), with higher (Mikkola et al., 2011b), as well as “syenite–quartz monzonite–granite
Y contents, flatter HREE profiles, as well as lower Sr and Al2O3 contents suite” in the Superior Province (Feng and Kerrich, 1992; Sage et al.,
than “high-pressure” TTGs. This shows that the late-Archean period 1996) and includes the Matok pluton in the northern Kaapvaal craton
represents a diversification of TTG magmatism rather than a strict evo- (Laurent et al., 2014). All these granitoids are collectively grouped
lution towards the “high-pressure” group. However, it must be noted here under the term of “sanukitoids sensu lato (s.l.)”.
that in all the cases mentioned above, the “low-pressure” TTGs proper Sanukitoids s.l. generally represent a minor, albeit ubiquitous, com-
are clearly subordinate in volume with respect to “medium-” and ponent of late-Archean terranes (roughly b15% in surface). For instance,
“high-pressure” TTGs. in the northern Kaapvaal craton and the Limpopo Belt, this group is only
212
Table 1
O. Laurent et al. / Lithos 205 (2014) 208–235

A list of examples of the four late-Archean granitoid types discussed in the present review, from most cratonic areas worldwide. This table is non-exhaustive and only reports examples for which a comparison could be made on the basis of sufficient
criteria (mostly based on petrography, whole-rock geochemistry and petrogenesis). The comparison is somewhat hypothetical for some terranes, because of the lack of geochemical data (particularly for the Slave and Zimbabwe Cratons).
TTG Sanukitoids s.l. Biotite- and two-mica granites “Hybrid” granitoids

a
Name Age (Ga) Name Age (Ga) Name Age (Ga) Name Age (Ga) Ref.b
Kaapvaal • Goudplaats–Hout River unit [3.40] • Matok pluton ~2.69 • Turfloop and Lekkersmaak 2.84–2.77 • Mashashane, Mashishimale, ~2.69 [1] to
(±Limpopo) • Groot-Letaba Duiwelskloof unit 3.30–2.85 (• Bulai pluton) (~2.60) batholith Matlala, Moletsi plutons [8]
• Baderoukwe, Malati Pump, • Various intrusions embedded in
Maranda intrusions TTG gneisses
Amazonian • Arco Verde, Caracol, Mariazinha 3.00–2.84 • Rio Maria suite ~2.87 • Mata Surrão, Xinguara granites 2.87–2.86 • Guarantã suite 2.87–2.86 [9] to
craton tonalites [14]
(Carajás • Agua Fria, Mogno trondhjemites
Province)
Baltic Shield • Various trondhjemites, tonalites [3.10] • Amindomaoja, Arola, Bergaul, Bolshozero, Chalka, 2.74–2.68 • Katajavaara, Konivaara, Pieni 2.70–2.68 • Leucogranite–granodiorite 2.72–2.68 [15]
and granodiorites 2.95–2.74 Elmus, Hautavaara, Hizhjärvi, Ilomantsinjärvi, Tuomaanjärvi intrusions suite (Kuhmo district) to
(many local names) Jalonvaara, Kaapinsalmi, Kaartojärvet, Koitere, (Kuhmo district) [24]
Kuittila, Kurgelampi, Kuusamo, Lieksa, Loso, Nilsiä, • Vaamankallio
Njuk, Panozero, Peuravaara, Purnu, Sharavalampi, leucogranite + leucosomes
Siikalahti, Sjargozero, Sysmanjärvi, Tulos, Voloma (Suomussalmi district)
plutons
• Lentiira and Riihivaara “quartz diorites”
Dharwar • Peninsular gneisses 3.00–2.55 • Dod gneisses ~2.55 • Bangalore leucogranite 2.54–2.51 • Felsic phases of Closepet 2.62–2.52 [25]
• Krishnagiri tonalites • Bisanattam granitoid • Kadiri, Krishnagiri, Lepakshi, batholith to
• Gneisses in the Hutti–Gurgunta • Mafic phases of Closepet batholith Madanapalle, Nandi hills granites • Arsikere–Banavara and [30]
area • Kolar granodiorite Chitradurga–
• Guramkonda, Vendodu granites Jampalnaikankote–Hosdurga
granites
• “Transitional TTG” in the
Hutti–Gurgunta area
East European • Granitoids of the West Azov group [3.25] • Osipenkovskii massif 2.92–2.86 [31]
(Ukrainian • Parts of the Obitochnen 3.12–2.91 • Parts of the Obitochnen Complex
Shield) Complex
Madagascar • Nosy Boraha suite [3.35] • Masoala suite (very heterogeneous, different granitoid 2.54–2.51 [32]
3.15–2.75 phases with characteristics of all three groups)
North Atlantic • Disko Bugt, Fiskefjord, Ikkattoq 3.20–2.80 • Disko Bugt diorites 3.00–2.80 • Qôrqut granite ~2.55 [33],
gneisses, Âta tonalite • Qeqertaussaq diorite • Rodebay granodiorite? [34]
North China • Taishan granitoid gneisses 2.75–2.55 • Taishan diorites 2.55–2.52 • Monzogranites of Qinhuangdao 2.55–2.44 • Granodiorites of Anziling, 2.55–2.52 [35]
• Eastern Hebei gneisses • Zhong Tian Men granodiorites pluton Jiekouling, Qinhuangdao to
• Fuping Complex • Hornblendites and diorites of Anziling, Jiekouling, • Some granites of the Wutai plutons [38]
Qinhuangdao plutons Complex • Some granites of the Wutai
Complex
Pilbara • Unnamed tonalitic, trondhjemitic 3.26–2.99 • Peewah suite: Peewah and Wallareenya ~2.95 • Unnamed monzo- and 2.94–2.93 • Portree Complex ~2.95 [39]
and granodioritic gneisses granodiorites, Geemas, Jallogoonina, Jones Wells, syenogranites
Malindra, Nr. 1 and Toweranna stocks
Slave • Banded gneisses [3.5] • Part of the “Type I” granitoids 2.63–2.60 • “Type II” granitoids 2.60–2.58 • Part of the “Type I” granitoids 2.63–2.60 [40],
3.10–2.65 • Concession and Defeat suites • Morose and Prosperous suites [41]
Superior Province • Lac Flavrian, Bourlamaque, Round 2.89–2.70 • Abitibi, Mattachewan batholiths, Lebel and Otto 2.72–2.69 • GMG granite suite (Pontiac Subp.) 2.69–2.65 • Watabeag batholith, Garrison 2.69–2.68 [42]
Lake complexes (Abitibi Subp.) stocks (Abitibi Subp.) • Ghost Lake batholith stock (Abitibi Subp.) to
• Tonalites, trondhjemites, • Stormer Lake and Okanse Lake plutons (Sioux Lookout terrane) • Granodiorites of the KMGS [49]
granodiorites (Wabigoon Subp.) (Berens River Subp.) • Granites and granodiorites suite (Wabigoon Subp.)
• Gutcher Lake, Hawk Lake, • Lac Fréchette and Lac Rémigny complexes (Wabigoon Subprovince) • Hermann Lake stock
Jubilee Lake suites (Wawa Subp.) (Pontiac Subp.) (Wawa Subp.)
• Barnum, Blalock, Burchell Lake, Lappé, Penassen
Lake suites (Quetico Subp.)
• Mafic phases of the KMGS suite, Jackfish Lake, Lac
des Îles, plutons, Eye-Dashwa, Norway Lake, Roaring
River, van Nostrand suites (Wabigoon Subp.)
• Dickenson Lake, Kabenung Lake, Maskinonge Lake,
Troupe Lake stocks (Abitibi Subp.)
O. Laurent et al. / Lithos 205 (2014) 208–235

Wyoming • Banded gneisses of the Washakie [3.35] • Louis Lake batholith 2.67–2.63 • Bears Ears, Middle Mountain, New ~2.55 • Felsic phases of the Bridger ~2.67 [50],
Province block, Native Lake 2.95–2.80 • Mafic phases of the Bridger batholith Fork Lake, South Pass granites batholith [51]
gneisses
Yilgarn • Banded gneiss and migmatite [3.20–2.75] • “Mafic” granitoids (Eastern Goldfields): Lawlers, ~2.66 • “Low-Ca granites” ~2.66 • “High-Ca granites” 2.69–2.66 [52]
Liberty, Maori Queen, Mt. Lucky, Porphyry plutons (=Transitional TTG) to
(Eastern Goldfields) [54]
• “High-HFSE” and “syenitic”
granitoids (Eastern Goldfields)
Zimbabwe • “TTGs” of the Tati-Selkirk suite [3.20] • “Sanukitoids” of the Tati-Selkirk suite 2.70–2.65 • Chillimanzi suite? 2.67–2.60 [55]
• Mashaba tonalite, Umwindsi 3.00–2.70 • “Potassic granites” of the to
gneiss, Shamvaian clasts Tati-Selkirk suite [58]
a
Ages between brackets are ages of oldest crust in the area of late-Archaean granitoid intrusions, either from inherited zircon grains in younger granitoids or Nd–Hf model ages.
b
Reference list: [1] Henderson et al., 2000; [2] Jaguin et al., 2012; [3] Kröner et al., 2000; [4] Laurent et al., 2011; [5] Laurent et al., 2013a; [6] Laurent et al., 2013b; [7] Laurent et al., 2014; [8] Zeh et al., 2009; [9] Almeida et al., 2010; [10] Almeida et al.,
2011; [11] Almeida et al., 2013; [12] Althoff et al., 2000; [13] Leite et al., 2004; [14] Oliveira et al., 2009; [15] Bibikova et al., 2005; [16] Halla, 2005; [17] Halla et al., 2009; [18] Heilimo et al., 2010; [19] Heilimo et al., 2011; [20] Käpyaho et al., 2006; [21]
Kovalenko et al., 2005; [22] Lobach-Zhuchenko et al., 2005; [23] Lobach-Zhuchenko et al., 2008; [24] Mikkola et al., 2011b; [25] Jayananda et al., 1995; [26] Jayananda et al., 2006; [27] Moyen et al., 2001; [28] Moyen et al., 2003; [29] Prabhakar et al.,
2009; [30] Vijaya Kumar et al., 2011; [31] Bibikova et al., 2008; [32] Schofield et al., 2010; [33] Garde, 1997; [34] Steenfelt et al., 2005; [35] Jahn et al., 1988; [36] Liu et al., 2004; [37] Wang et al., 2009; [38] Yang et al., 2008; [39] Smithies and Champion,
2000; [40] Davis and Bleeker, 1999; [41] Davis and Hegner, 1992; [42] Breaks and Moore, 1992; [43] Feng and Kerrich, 1992; [44] Sage et al., 1996; [45] Stern et al., 1989; [46] Stern and Hanson, 1991; [47] Stevenson et al., 1999; [48] Sutcliffe et al.,
1990; [49] Whalen et al., 2004; [50] Frost et al., 1998; [51] Frost et al., 2006; [52] Cassidy et al., 1991; [53] Champion and Sheraton, 1997; [54] Champion and Smithies, 2003; [55] Bagai et al., 2002; [56] Jelsma et al., 1996; [57] Kampunzu et al., 2003;
[58] Zhai et al., 2006.
213
29°E 30°E 31°E
Botswana Zimbabwe
100 km
Musina
Mozambique
Bu
tral
Zone N
Cen
g
b e r
n s
S o u t p a
23°S
Louis Trichardt
n e
Zo
Sout
hern GGB
Marginal
Ho
ut
Mk
Riv
er GHR
ne
o
Sh
Z
Mt Mo ea
r GLD
RGB
Tzaneen Lk
Polokwane
Phalaborwa
MGB
Ms
24°S
ent
PGB Ma Lin
e am
n
Tf iso
Thabazimbi- M u rch
Sanukitoids and related rocks (2.7-2.6 Ga) Proterozoic intrusive complexes (~2.0 Ga)
Sedimentary cover (<2.6 Ga)
’’Hybrid’’ granites (~2.69 Ga)
Greenstone belts (3.3-2.8 Ga)
Biotite-, two-mica granites (~2.78 Ga) Limpopo Belt : granulites
TTG gneisses and associated small Undifferentiated granitoids of the
Bt-(Ms-)granite bodies (3.3-2.8 Ga) Witwatersrand block (>3.1 Ga)
Major shear zone Town
Fig. 2. Spatial distribution of the different late-Archean granitoid groups in the Pietersburg block and part of the Central Zone of the Limpopo Belt in South Africa (after Kramers et al., 2006;
Laurent et al., 2013b and Robb et al., 2006). The abbreviations refer to the granitoid plutons or units described in this study: Bu = Bulai; GHR = Goudplaats–Hout River gneiss suite;
GLD = Groot Letaba–Duiwelskloof gneiss unit; Lk = Lekkersmaak; Ma = Mashishimale; Mk = Matok; Ms = Mashashane; Mt = Matlala; Mo = Moletsi; Tf = Turfloop. Also abbrevi-
ated are the names of the Giyani (GGB), Murchison (MGB), Pietersburg (PGB) and Rhenosterkoppies (RGB) greenstone belts.
Fig. 3. Field photographs showing examples of late-Archean granitoids from the Pietersburg block and the Central Zone of the Limpopo Belt: TTGs, with (a) typically polyphased outcrop of
highly deformed and partly migmatitic orthogneisses of TTG composition (Goudplaats–Hout River gneiss unit); and (b) example of homogeneous, foliated and unmigmatized granodio-
ritic gneiss (Groot Letaba–Duiwelskloof gneiss unit); Sanukitoids s.l., with (c) K-feldspar porphyritic, relatively dark granodiorite, containing comagmatic mafic enclaves of monzodioritic
composition (Bulai pluton); and (d) dark, fine-grained (monzo)diorite containing phenocrysts of plagioclase (Plag) and clinopyroxene (Cpx) (Matok pluton); (e) Biotite- and two-mica
granites: homogeneous, equigranular biotite-bearing monzogranite (Turfloop batholith); (f) Hybrid granites: heterogeneous, coarse-grained and porphyritic monzogranite (Mashashane
pluton).
represented by two intrusions, the Bulai and Matok plutons (Laurent 2011; Lobach-Zhuchenko et al., 2005; Mikkola et al., 2011a). Granites
et al., 2011, 2014; Fig. 2). s.s. are invariably scarce.
The granodiorites often contain large (1–6 cm) K-feldspar pheno-
2.2.1. Geology and petrography crysts within a medium-grained (0.1–1 cm) matrix made up of pla-
Sanukitoids are syn- to post-tectonic intrusions, ranging from well- gioclase (An20–35), quartz, and very characteristic mafic aggregates of
foliated to undeformed. They most often consist of composite magmatic biotite and amphibole. Ortho- and clinopyroxene are relatively com-
complexes: for instance, the dominant phase of the Bulai and Matok mon, especially in the mafic clots where they occur as relict cores within
plutons is a porphyritic granodiorite (Fig. 3c) associated with abundant biotite and amphibole. Magnetite, ilmenite, apatite, zircon, allanite and,
mafic bodies of quartz-(monzo)diorite (Fig. 3d), whose size ranges from locally, titanite are abundant accessory phases.
small microgranular magmatic enclaves (MME) to bodies several
kilometres in size. In some areas, the latter locally form individual and 2.2.2. Key geochemical features
monogenic stocks, for instance in the Superior Province (Sage et al., A number of definitions based on strict geochemical criteria have
1996; Stevenson et al., 1999), the North China craton (Wang et al., already been proposed for sanukitoids s.s. The original one only consid-
2009), the Pilbara craton (Smithies and Champion, 2000) or the Baltic ered LILE- and LREE-rich diorites with high Mg# (N 0.6; e.g. Stern et al.,
shield (Heilimo et al., 2010). Some sanukitoid intrusions are also associ- 1989), but it has since been widened to include their differentiated
ated with lamprophyre dykes and/or ultrapotassic rocks (Laurent et al., products (granodiorites and granites of the “sanukitoid suite”; Heilimo
Table 2
Average major- and trace-element compositions of the four types of late-Archean granitoids in the Pietersburg block and the Central Zone of the Limpopo Belt (South Africa).
TTG Sanukitoids s.l. Biotite- and two-mica Hybrid granites

granites
N = 23 N = 77 N = 34 N = 104
Average St. dev. Average St. dev. Average St. dev. Average St. dev.
Major elements (wt.%)

SiO2 71.20 2.12 63.57 6.43 72.54 1.59 70.65 3.82
TiO2 0.26 0.11 1.22 0.55 0.26 0.14 0.61 0.33
Al2O3 15.42 0.74 14.48 1.26 14.42 0.48 13.36 0.60
FeOt 1.63 0.75 5.52 2.24 1.30 0.49 2.98 1.57
MnO 0.03 0.01 0.08 0.04 0.02 0.01 0.05 0.03
MgO 0.66 0.36 1.89 1.03 0.44 0.23 0.89 0.80
CaO 2.37 0.75 3.93 1.75 1.34 0.37 1.82 1.10
Na2O 5.37 0.61 3.60 0.31 4.36 0.37 3.86 0.33
K2O 1.77 0.66 3.32 1.05 4.07 0.72 4.26 0.88
P2O5 0.11 0.05 0.54 0.31 0.10 0.06 0.22 0.15
A/NK 1.44 0.14 1.54 0.22 1.25 0.07 1.23 0.14
A/CNK 1.03 0.04 0.88 0.09 1.03 0.03 0.94 0.06
K2O/Na2O 0.34 0.15 0.95 0.37 0.95 0.22 1.12 0.27
Trace elements (ppm)

Cs 4.1 2.8 1.1 1.0 5.1 4.1 7.3 3.8
Rb 71 37 90 39 155 53 186 57
Ba 532 243 1292 542 751 237 795 255
Nb 5.1 1.8 20.5 7.8 7.3 3.9 19.6 5.9
Sr 638 241 527 202 254 128 251 122
Zr 151 53 475 190 155 92 332 150
Y 8 4 44 18 15 11 44 14
La 24.2 11.1 86.8 36.0 40.8 25.2 70.9 24.0
Ce 47.6 21.9 186.3 74.0 74.2 45.7 145.1 45.8
Pr 4.58 1.72 22.79 9.28 7.64 4.75 17.05 5.69
Nd 18.6 7.9 88.6 37.9 27.7 16.8 62.1 20.9
Sm 3.01 1.23 14.43 6.18 4.56 2.33 10.64 3.63
Eu 0.88 0.31 2.81 1.03 0.69 0.25 1.66 0.73
Gd 2.27 0.91 11.26 4.66 3.19 1.51 8.96 3.19
Tb 0.30 0.13 1.64 0.66 0.49 0.24 1.38 0.47
Dy 1.31 0.56 8.50 3.35 2.74 1.57 7.78 2.66
Ho 0.24 0.10 1.57 0.65 0.53 0.34 1.55 0.54
Er 0.62 0.27 4.23 1.81 1.52 1.06 4.58 1.57
Tm 0.08 0.03 0.63 0.28 0.22 0.16 0.67 0.22
Yb 0.65 0.43 3.73 1.65 1.33 0.96 4.41 1.47
Lu 0.09 0.06 0.54 0.24 0.20 0.14 0.64 0.21
Hf 4.0 1.2 11.7 4.3 4.8 2.3 9.0 3.6
Ta 0.27 0.12 0.97 0.41 0.52 0.42 1.24 0.40
Th 3.7 1.5 8.9 9.6 10.9 4.6 10.3 4.5
U 1.0 0.4 1.0 0.7 2.1 1.3 2.5 1.2
Pb 13.4 4.5 n.d. n.d. 25.8 5.6 21.4 5.6
Sc 3 1 12 6 3 1 8 5
V 16 2 98 48 8 7 53 33
Cr 13 6 79 49 11 6 29 38
Co 3 1 19 15 2 1 7 6
Ni 5 1 32 25 5 5 15 16
Ga 19 2 20 3 18 2 21 3
Ba/Rb 8.8 5.6 17.9 11.7 5.3 2.4 5.1 3.3
Sr/Y 94.9 56.0 15.3 13.4 28.5 25.4 6.0 3.3
Nb/Ta 19.5 7.7 22.0 4.8 19.7 10.8 17.0 5.0
Lu/Hf 0.021 0.010 0.050 0.038 0.057 0.072 0.076 0.026
(La/Yb)N 31.5 13.4 22.1 17.7 32.6 21.6 11.8 5.9
EuN/Eu* 1.10 0.32 0.73 0.27 0.62 0.23 0.51 0.13
et al., 2010; Martin et al., 2005) as well as the “high-Ti” sanukitoids variable K2O contents (1.5 ≤ K2O ≤ 5.0 wt.%) leading to a wide range
(Martin et al., 2009), also referred to as “Closepet-type” granitoids of K2O/Na2O ratios, albeit generally ≥0.5 (Table 2; Fig. 4b). Such rocks
(Moyen et al., 2003), which are distinctively rich in FeOt, TiO2 and are typically characterized by high contents of ferromagnesian oxides
other HFSE. Our group of sanukitoids s.l. is even looser, as it also com- (5 ≤ FeOt + MgO + MnO + TiO2 ≤ 25 wt.%) and CaO (Fig. 4c–d).
prises granitoids that do not match all the criteria proposed by these Primitive mantle-normalized trace element patterns of sanukitoids
definitions. However, they are all demonstrably formed by similar s.l. are relatively similar to those of TTGs, notably characterized by
processes as sanukitoids s.s., i.e. through interaction at mantle level negative anomalies in Nb–Ta, but shifted to higher concentrations in
between peridotite and an incompatible element-rich component. all incompatible elements (Fig. 5a). These rocks are typically rich in Ba
All sanukitoids s.l. share several key geochemical characteristics. They (generally N1000 ppm) and Sr (generally N 400 ppm), which results in
are calc-alkaline to alkali-calcic, metaluminous (0.7 ≤ A/CNK ≤ 1.0) high Ba/Rb and Sr/Y ratios (Fig. 5b), and also have high contents in tran-
granitoids characterized by a large range of SiO2 contents, extending sition elements such as V (N50 ppm; Fig. 5c), Ni (15–200 ppm) and Cr
down to mafic compositions (45 ≤ SiO2 ≤ 70 wt.%), in contrast with (20–500 ppm). This dual geochemical character, i.e. richness in both
other late-Archean granitoids (Fig. 4a–b). They are potassic, with “crustal” and “mantle” elements, is a unique and very typical feature
12 1.2
(a) (b)
Al2O3 / (CaO + Na2O + K2O) (mol.)

Biotite-granites
Alkalic
Na2O + K2O – CaO (wt.%)
8
1.0
Sanukitoids
4
c
lci
i-ca
k al 0.8
Al TTG
0
Calcic
lilcic
kaa
al lk
c--a
al lc
CCa
–4 0.6
45 50 55 60 65 70 75 80 0.0 0.5 1.0 1.5 2.0
SiO2 (wt.%) K2O / Na2O
50
(c) (d)
FeOt + MgO + MnO + TiO2 (wt.%)
20
20
Al2O3 / (FeOt + MgO)

10
10
5
5
2
2
1
1
0.5
0.5
45 50 55 60 65 70 75 80 0 0.3 0.6 0.9 1.2 1.5
SiO2 (wt.%) CaO / (Na2O + K2O)
Sanukitoids s.l. Biotite- and two-mica granites Hybrid granites

TTG
Granitic phases Small bodies in TTG gneisses Dark enclaves
Fig. 4. Major-element geochemistry of the four groups of late-Archaean granitoids, illustrated with examples from the Pietersburg block and the Limpopo Belt (symbols) as well as liter-
ature data (fields; compilations from Martin et al., 2009; Moyen, 2011; Moyen and Martin, 2012). (a) MALI index ([Na2O + K2O]–CaO) vs. SiO2 diagram of Frost et al. (2001); (b) A/CNK
(molar Al2O3/[CaO + Na2O + K2O] ratio) vs. K2O/Na2O ratio; (c) sum of “mafic” oxides (FeOt + MgO + MnO + TiO2) vs. SiO2; and (d) Al2O3/(FeOt + MgO) ratio vs. CaO/(Na2O + K2O)
ratio.
of all sanukitoids s.l. Some of them, such as the Bulai and Matok granit- of the metasomatic agent that could be represented, in addition to TTG
oids, are very rich not only in Ba and Sr, but also in other incompatible melts, by H2O- or CO2-rich brines (Lobach-Zhuchenko et al., 2008;
elements such as REE and HFSE (ΣREE ≥ 250 ppm; 20 ≤ Y ≤ 70 ppm; Stern et al., 1989), fluids or melts derived from sediments (Halla,
300 ≤ Zr ≤ 1000 ppm; Table 2; Fig. 5c,d), a typical feature of the 2005; King et al., 1998; Laurent et al., 2011; Mikkola et al., 2011b;
“high-Ti” sanukitoids (Martin et al., 2009; Moyen et al., 2003). Thereby, Wang et al., 2009), or even alkaline melts (Heilimo et al., 2010) and
some sanukitoids s.l. display a Sr negative anomaly in multi-element carbonatites (Steenfelt et al., 2005). In the case of sanukitoids s.l. show-
patterns (Fig. 5a), owing to a greater enrichment in REEs than Sr with ing notably high FeOt/MgO ratios, low Al2O3 and flat HREE profiles, such
respect to primitive mantle compositions. Sanukitoids s.l. show a wide as the Matok pluton from the Pietersburg block (Figs. 2, 4 and 5) and
range of LaN/YbN ratios (10–75, but most often N25), and a variable Eu other granitoids reported from different cratons (Jahn et al., 1988;
anomaly (0.5 ≤ EuN/Eu* ≤ 1.0; Fig. 5d). Mikkola et al., 2011b; Steenfelt et al., 2005), asthenospheric mantle
could also have played a role as a source component (Laurent et al.,
2.2.3. Petrogenesis 2014).
Experimental and geochemical constraints show that sanukitoids s.s. On the other hand, the widespread felsic rocks of sanukitoid
result from interaction between mantle peridotite and a component suites (i.e. granodiorites) derive from differentiation of the most
rich in incompatible elements, most often a TTG melt (Almeida et al., mafic magmas, regardless of the nature and source of the latter, by
2013; Dey et al., 2012; Halla et al., 2009; Heilimo et al., 2010; Martin fractionation of amphibole- and plagioclase-rich solids with little or
et al., 2009; Moyen et al., 2001, 2003; Oliveira et al., 2011; Rapp et al., no involvement of pre-existing felsic crust (Bédard, 1996; Laurent
1999, 2010; Smithies and Champion, 2000; Stern and Hanson, 1991). et al., 2013a; Lobach-Zhuchenko et al., 2005, 2008; Oliveira et al.,
Geochemical variability of sanukitoids s.l. would result either from (i) 2010; Stern and Hanson, 1991). Therefore, most of the chemical vari-
different petrogenetic processes at mantle levels, especially regarding ability in sanukitoids s.l. (including the felsic rocks) results from mantle
the physical conditions of the metasomatic interactions (P, T and relative processes, as described above, rather than reflecting contamination
volume of peridotite and contaminant; Martin et al., 2009; Mikkola et al., at shallower levels of emplacement. This legitimates the identity of
2011b; Rapp et al., 1999, 2010), and/or (ii) the nature and composition sanukitoids s.l. which, although compositionally diverse, are all related
1000 500
(a) Average sanukitoids (b)
Average Bt-granites 200
Sample / Primitive mantle Average TTG
100
100 50
Sr / Y
20
10
10 5
Er Yb 1
Rb Th Ta Ce Nd Hf Eu Tb Y
1 0.5
Ba Nb La Sr Zr Sm Gd Dy Ho Tm Lu 1 2 5 10 20 50 100
Ba / Rb
1200 2.0
(c) (d)
1000
1.5
800
EuN / Eu* Positive
Zr (ppm)
600 1.0 No Eu anomaly
Negative
400
0.5
200
0 0
0 50 100 150 200 250 0 200 400 600 800 1000
V (ppm) ΣLREE (ppm)
Sanukitoids s.l. Biotite- and two-mica granites Hybrid granites

TTG
Granitic phases Small bodies in TTG gneisses Dark enclaves
Fig. 5. Trace-element geochemistry of the four groups of late-Archaean granitoids, illustrated with examples from the Pietersburg block and the Limpopo Belt (symbols) as well as
literature data (fields; compilations from Martin et al., 2009; Moyen, 2011; Moyen and Martin, 2012). (a) Multi-element patterns of average composition for each group, normalized to
primitive mantle concentrations (“Pyrolite Silicate Earth” of McDonough and Sun, 1995); (b) Sr/Y vs. Ba/Rb ratios; (c) HFSE (represented by Zr) vs. transition element (represented by
V) concentrations; and (d) amplitude of the Eu anomaly (EuN = Eu concentrations normalized to C1 chondrites; McDonough and Sun, 1995; Eu* = √SmN·GdN) vs. sum of light rare
earth elements (ΣLREE = La + Ce + Nd).
somehow to interactions between peridotite and incompatible 2012) and the Kaapvaal Craton (Henderson et al., 2000; Meyer et al.,
element-rich components at mantle levels. 1994; this study).
After TTGs, biotite- and two-mica granites are the second most
widespread lithology in late-Archean terranes (e.g. Moyen et al.,
2.3. Biotite- and two-mica granites 2003; Sylvester, 1994). For instance, they are largely exposed in the
Pietersburg block as large plutons or batholiths (Turfloop and
This group basically consists of “purely crustal-derived” granites Lekkersmaak granites; Fig. 2) or occur as diffuse sheets, small bodies
(i.e. generated by melting of grey gneisses: TTGs ± metasediments). and dykes intrusive in, and intimately associated with, the TTGs of the
They are widespread in every Archean craton (Sylvester, 1994; see Goudplaats–Hout River and Groot Letaba–Duiwelskloof gneiss units
Table 1), including the Superior Province (Breaks and Moore, 1992; (Laurent et al., 2013b; Robb et al., 2006).
Feng and Kerrich, 1992), the Slave Craton (Davis and Hegner, 1992),
the Wyoming Province (Frost et al., 1998, 2006), the Baltic Shield 2.3.1. Geology and petrography
(Käpyaho et al., 2006; Martin and Querré, 1984; Mikkola et al., Texturally, the biotite- and two-mica granites usually consist of ho-
2011b), the Dharwar Craton of South India (Dey et al., 2012, 2014; mogeneous, grey to pink, equigranular granites and leucogranites
Jayananda et al., 2006; Moyen et al., 2001, 2003; Vijaya Kumar et al., whose grain size ranges from 0.1 to 0.5 cm (Fig. 3e). Some porphyritic
2011), the North China Craton (Jahn et al., 1988; Liu et al., 2004; Yang varieties do occur, but are generally largely subordinate (e.g. in the
et al., 2008), the Pilbara and Yilgarn Cratons in Western Australia Turfloop batholith of the Pietersburg block; Henderson et al., 2000;
(Champion and Sheraton, 1997; Champion and Smithies, 1999, 2007), Kröner et al., 2000). Like sanukitoids, they are generally late- to post-
the Amazonian Craton (Almeida et al., 2010, 2013; Althoff et al., tectonic and show variable levels of deformation from one area to
2000; Leite et al., 2004), the Tanzania Craton (Mshiu and Maboko, another. Their modal composition is fairly constant, with ~ 30% K-
3 CaO granites are described as partial melts of older felsic crust, especially
TTGs (e.g. Almeida et al., 2013; Althoff et al., 2000; Dey et al., 2012;
Feng and Kerrich, 1992; Frost et al., 2006; Henderson et al., 2000;
Moyen et al., 2001, 2003; Mshiu and Maboko, 2012; Sylvester,
Low-K mafic rocks High-K mafic rocks
1994; Vijaya Kumar et al., 2011; Whalen et al., 2004). Major element
compositions of such granites are consistent with this model, matching
the composition of experimental melts derived from tonalites
(Fig. 6). Furthermore, these rocks are equivalent to the potassic gran-
ite subgroup of Archean grey gneisses defined by Moyen (2011),
Tonalites Metasediments which he interpreted as partial melts of sodic TTGs at crustal pres-
sure (b 10 kbar). A plagioclase-rich and garnet-free residue accounts
for the low Ba/Rb and Sr/Y ratios and negative Eu anomalies of these
granites.
As pointed out by Watkins et al. (2007), most TTGs are too sodic to
Al2O3 K2O
5 produce the large volumes of late-Archean potassic granites. Therefore,
(FeOt + MgO) Na2O
these authors proposed an alternative hypothesis, assuming that these
Fig. 6. Ternary diagram Al2O3/(FeOt + MgO); 3 ∗ CaO; 5 ∗ (K2O/Na2O) on which the
granites could derive ultimately from a highly metasomatized, K-rich
composition of late-Archean granitoids from the Pietersburg block and Limpopo Belt is mantle source. Nevertheless, their very silicic nature, the absence of in-
plotted (symbols as in Figs. 4 and 5). The multiplication factors were only added for scaling termediate to mafic phases and the lack of geochemical affinities with
purposes. The fields represent the composition of melts derived from a range of potential sanukitoids s.l. (Figs. 4 and 5), which definitely derive from such an
sources (tonalites, metasediments, low- and high-K mafic rocks), determined using the
enriched mantle source, are at odds with this model. Moreover, the
major-element compositions of experimental studies: Patiño-Douce and Beard (1995),
Singh and Johannes (1996), Skjerlie and Johnston (1996), Watkins et al. (2007) for K2O contents of melts produced experimentally by partial melting of
tonalites; Patiño-Douce and Beard (1996), Patiño-Douce and Johnston (1991), Vielzeuf low-K TTGs range from 2 to 6 wt.% at 70–75 wt.% SiO2 (Patiño-Douce
and Holloway (1988) for metasediments; Beard and Lofgren (1991), Rapp et al. (1999), and Beard, 1995; Singh and Johannes, 1996; Skjerlie and Johnston,
Rapp and Watson (1995), Wolf and Wyllie (1994) for low-K mafic rocks and Sisson 1996; Watkins et al., 2007), with an average K2O content of 3.8 wt.%,
et al. (2005) for high-K mafic rocks.
which is comparable to that of biotite- and two-mica granites in the
Pietersburg block (4.1 wt.%; Table 2). On the other hand, melting of
metagreywackes and metapelites would provide an additional source
of K2O: the two-mica-bearing late-Archean granites, such as the
feldspar, ~ 30% sodic plagioclase (An5–15), and ~30% quartz; the other Lekkersmaak pluton in South Africa, show A/CNK values N1.1 similar
major mineral phases are biotite, and in places primary muscovite. to that of post-Archean S-type granites, pointing to the contribution of
For instance, in South Africa, most of the Lekkersmaak pluton such a metasedimentary source (e.g. Breaks and Moore, 1992; Jaguin
(Jaguin et al., 2012) as well as some phases of the Turfloop batholith et al., 2012; Jahn et al., 1988). Therefore, we consider that late-
(Henderson et al., 2000; Kröner et al., 2000) are two-mica granites. Archean biotite- and two-mica granites derive from variable contribu-
Accessory phases are ilmenite, magnetite, apatite, zircon, allanite tions of two distinct crustal sources, i.e. sodic orthogneisses (TTGs)
and, locally, pistacite. Exceptionally, the two-mica phases also bear and paragneisses (metasediments).
Al-rich minerals such as cordierite, garnet and tourmaline, and are
associated with rare metal-bearing (e.g. Sn and Li) pegmatite fields
(e.g. Breaks and Moore, 1992; Feng and Kerrich, 1992; Frost et al., 2.4. Hybrid granitoids
1998; Kinny, 2000).
Late-Archean granitoids with hybrid characteristics were identi-
2.3.2. Key geochemical features fied from most cratons. These formed through interaction (e.g. meta-
As shown by examples from South Africa (Table 2; Fig. 4a–c), typ- somatism, mingling, mixing) between magmas or sources of any of
ical chemical characteristics of biotite- and two-mica granites are the previously defined groups. They form an extremely heteroge-
their high silica contents (68 ≤ SiO2 ≤ 75 wt.%), peraluminous affinity neous family that cannot be defined on geochemical criteria alone,
(A/CNK ≥ 1.0), low amounts of ferromagnesian oxides (1 ≤ FeOt + because sources and petrogenetic mechanisms differ from one place
MgO + MnO + TiO2 ≤ 4 wt.%), and a clear potassic signature, with to another. They most commonly result from interactions between
K2O of ~ 4 wt.% resulting in K2O/Na2O ratios N 0.5. Low CaO concen- sanukitoids s.l. and biotite- or two-mica granites, for example the
trations (b2 wt.%) and moderate Al2 O3 contents lead to low CaO/ Mashashane, Mashishimale, Matlala and Moletsi plutons of the
(Na2O + K2O) and high Al2O3/(FeOt + MgO) ratios (Fig. 4d). Pietersburg block (Fig. 2), as well as several granitoids in the Superior
The trace element geochemistry of biotite- and two-mica granites Province (Stevenson et al., 1999; Whalen et al., 2004) and the Dharwar
from the Pietersburg block in South Africa is fairly representative of craton of southern India (Jayananda et al., 2006; Moyen et al., 2001).
the whole group. Their primitive mantle normalized patterns are similar Some relatively common granodiorites and monzogranites, which are
to those of TTGs, except for higher contents in highly incompatible clearly too mafic to represent pristine crustal melts, also fall into this
elements (see Table 2) such as Rb (average = 155 ppm) and Th (aver- category (e.g. “high-Ca” granites of the Yilgarn Craton; Champion and
age = 11 ppm), systematic negative Ba and Sr anomalies and higher Sheraton, 1997). In addition, this family includes all granitoids recog-
concentrations in Y–HREE (Fig. 5a). Thus, such granites have lower nized as being hybrids between TTGs and sanukitoids or biotite-
Ba/Rb and Sr/Y ratios than TTGs (Fig. 5b). Both HFSE and transition bearing granites, i.e. the so-called “Transitional TTGs” from the Yilgarn
element contents are low (Zr b 265 ppm; V b 20 ppm; Table 2; Fig. 5d), Craton (Champion and Smithies, 2003) and the Dharwar Craton (Dey
and their REE contents are moderate, with variously fractionated et al., 2012; Jayananda et al., 2006; Prabhakar et al., 2009) as well as
REE patterns (15 ≤ LaN/YbN ≤ 65 for most samples) but significant monzonite–granodiorite–granite suites of the Amazonian Craton
negative Eu anomalies (EuN/Eu* ~ 0.5; Fig. 5d). (Almeida et al., 2010, 2013) and Superior Province (Feng and Kerrich,
1992). The rare Archean alkaline granites and syenites (Champion and
2.3.3. Petrogenesis Sheraton, 1997; Corfu et al., 1989; Feio et al., 2012; Mikkola et al.,
In all Archean cratons worldwide, including the Kaapvaal Craton, 2011a; Sage et al., 1996; Smithies and Champion, 1999b) may also be
slightly peraluminous, potassic and calc-alkaline biotite-bearing akin to this group, as they presumably derived from a lower crustal,
metasomatized basaltic source (Smithies and Champion, 1999b, mica granites by the absence of primary muscovite, their greater
2000). richness in biotite and the presence of magmatic epidote (pistacite),
The global distribution of hybrid granitoids is very heteroge- which is ubiquitous in hybrid granitoids from the Pietersburg block in
neous. They are not reported in every late-Archean craton, and where South Africa as well as in the Amazonian and Dharwar Cratons
they do occur, they can represent either a very small volume (e.g. in (Almeida et al., 2010, 2013; Jayananda et al., 2006). Hornblende is
the Kuhmo district of the Baltic shield; Mikkola et al., 2011b) or a signif- generally rare or only present as relics: some monzogranites from
icant crustal component, like in the Pietersburg block in South Africa the Pietersburg block contain cm-sized mafic clots, interpreted as
where they form several large granitic plutons (Fig. 2). pseudomorphs of early amphibole crystals into biotite + epidote as-
semblages (Laurent, 2012), a characteristic also reported from hy-
brid granitoids of the Dharwar Craton (Jayananda et al., 2006).
2.4.1. Geology and petrography Accessory minerals are magnetite, ilmenite, apatite, titanite and
In general, late-Archean hybrid granitoids are well-defined plutons, allanite.
like in the Pietersburg block (Fig. 2). They generally consist of com-
posite magmatic complexes where, in contrast to the sanukitoids,
the monzogranitic phases largely prevail over granodiorites and diorites. 2.4.2. Key geochemical features
For example, in the northern Kaapvaal craton, the latter are restricted There is no systematic geochemistry of hybrid granitoids, because
to scarce, small (b 1 m) enclaves enclosed within the dominant coarse- the different components that are involved in their genesis differ from
grained (0.5–2 cm), equigranular to porphyritic monzogranites one craton to another. Nevertheless, as reported by all authors, they
(Fig. 3f). In other areas, granodioritic phases can be more abundant are typically intermediate in composition between the granitoids
(e.g. Almeida et al., 2010; Champion and Smithies, 2003) but there, from the other three groups (Almeida et al., 2010, 2013; Champion
too, mafic rocks are scarce or even absent, and clearly distinguish and Sheraton, 1997; Champion and Smithies, 2003; Jayananda et al.,
these intrusions from sanukitoids s.l. 2006; Moyen et al., 2001). For example, the Mashashane, Mashishimale,
Most often, the dominant monzogranites consist of sodic plagioclase Matlala and Moletsi monzogranites from the Pietersburg block
(An5–20) (~30%), K-feldspar (~30%), quartz (~25%), with biotite (b15%) resemble biotite- and two-mica granites in being calc-alkaline to
as the dominant mafic mineral. They differ from classic biotite- and two- alkali-calcic, silica-rich rocks (SiO2 N 68 wt.%), with high K2O/Na2O
Na2O / K2O Na2O / K2O

Wyoming
Greenland
(a) TTG China (b) Sanukitoids
Superior s.l. Baltic
Baltic Pilbara
Dharwar China
Amazonian
Superior
2A
2A
B
B
MS
MS
/CN
/CN
Wyoming Dharwar
2F
2F
K
Na2O / K2O Na2O / K2O
(c) Biotite-, (d) Hybrid

two-mica granites
granites Wyoming
Baltic
Amazonian Superior
Baltic
Dharwar
Dharwar
Wyoming
Superior
2A
2A
B
B
MS
MS
/CN
/CN
China Tanzania
2F
2F
K
Fig. 7. Proposed ternary classification diagram for late-Archean granitoids: 2 ∗ A/CNK (molar Al2O3/[CaO + Na2O + K2O] ratio); Na2O/K2O ratio; 2 ∗ (FeOt + MgO)wt.% ∗ (Sr + Ba)wt.%
(=FMSB). The multiplication factors were only added for scaling purposes. Examples of each group of late-Archean granitoids were plotted on the diagram, including samples from
the Pietersburg block and the Limpopo Belt in South Africa (symbols) as well as 7 other cratons worldwide (for references, see Table 2). This diagram can be used as a petrogenetic
indicator, each pole representing an end-member source of late-Archean granitoids: (1) pre-existing crustal rocks (pole A/CNK); (2) low-K mafic rocks (pole Na2O/K2O); and (3) a
LILE-rich, metasomatized mantle (pole FMSB). The hybrid granitoids plot in the convergence area of the other three groups. See text for details.
ratios (0.5–1.5) (Table 2; Fig. 4a–b). However, they are poorer in 2.5. A synthetic classification diagram
Al (12 ≤ Al 2 O 3 ≤ 14 wt.%) and richer in ferromagnesian oxides
(1 ≤ FeO t + MgO + MnO + TiO 2 ≤ 7 wt.%; Fig. 4c) leading to The review presented above pinpoints the involvement of three end-
metaluminous affinities (0.85 ≤ A/CNK ≤ 1.05; Fig. 4b) and lower member petrogenetic processes in the origin of late-Archean granites. We
Al2O3/(FeOt + MgO) ratios (Fig. 4d), which is more typical of sanukitoids therefore propose to use a synthetic ternary diagram for purposes of clas-
s.l. sification, in which each pole represents a key geochemical characteristic
The same observation can be made for trace elements. Trace element of granitoids produced through one of these processes (Fig. 7):
patterns of hybrid granites are similar to those of biotite- and two-mica
(1) Na2O/K2O for melting of meta-igneous mafic rocks with low to
granites in their high Rb, Th contents (Table 2) and negative Ba, Sr
moderate K2O contents;
anomalies (Fig. 5a), as well as their Ba/Rb ratios (Fig. 5b). However,
(2) FMSB ([FeOt + MgO]wt.% ∗ [Sr + Ba]wt.%) for the interactions be-
like sanukitoids, they are also very rich in REE and HFSE (Fig. 5c–d),
tween peridotite and components rich in incompatible elements
which results in lower Sr/Y ratios than in any other granitoid type
(in particular LILE); and
(Fig. 5b). Zr and V contents are also intermediate between those of
(3) A/CNK for melting of Al-rich, felsic crustal lithologies (TTGs,
sanukitoids s.l. and high-Al granites (Table 2; Fig. 5c). Due to high
metasediments).
HREE contents (3.0 ≤ Yb ≤ 10.4 ppm), REE patterns are moderately
fractionated (5.0 ≤ LaN/YbN ≤ 17.6) and have a marked negative Eu We plotted the composition of late-Archean granitoids from South
anomaly (0.2 ≤ EuN/Eu* ≤ 0.7; Fig. 5d). Africa together with those from other Archean cratons worldwide in
this ternary diagram (Fig. 7). Three important observations can be
2.4.3. Petrogenesis made: (i) all members of a given group show relatively similar compo-
Fig. 6 shows that the major-element composition of hybrid gran- sitions regardless of their geographic origin, highlighting the relevance
itoids does not unambiguously point to an origin through melting of of the classification; (ii) the diagram is fairly effective at discriminating
a single crustal lithology. Most authors that have studied such granites late-Archean granitoids, as the different groups are clearly distinct from
worldwide therefore propose a mixed origin, which basically involves each other: TTGs trend towards the Na2O/K2O pole, whilst biotite- and
any of the magma or sources of the previously defined groups: mixing two-mica granites plot close to the A/CNK pole and sanukitoids s.l. are
between a differentiated sanukitoid magma and a TTG liquid (Almeida concentrated around the FMSB end-member; (iii) there is however a
et al., 2010), contamination of melts originating from the lower mafic significant overlap between the different granitoid groups, especially
crust by an older crustal component (Jayananda et al., 2006; Smithies in the central part of the ternary diagram where they all converge to-
and Champion, 2000), hybridization of TTG magmas by enriched man- wards the field of hybrid granitoids, a problem obviously inherent in
tle (Feng and Kerrich, 1992; Prabhakar et al., 2009) or a complex inter- the petrogenesis of the latter. This issue further emphasizes that there
play of these different mechanisms (Champion and Sheraton, 1997). In is a compositional continuum between the four granitoid groups de-
the Pietersburg block, the hybrid granites are akin to both sanukitoid fined here and that, therefore, many samples actually represent
suites (high FeOt + MgO + MnO + TiO2 as well as high REE and HFSE magmas derived from relative contributions of the three end-
contents; Figs. 4c, 5), and biotite- and two-mica granites (low CaO/ members. Likewise, this illustrates that this diagram must be used in ad-
[Na2O + K2O], similar Rb, Ba, Th contents and K2O/Na2O; Figs. 4, 5). dition to other, classical geochemical projections, and cannot substitute
Therefore, in that case, they likely originate from interactions between careful petrogenetic modelling.
felsic magmas derived from the differentiation of sanukitoids and
crustal melts (similar in composition to biotite- and two-mica gran- 3. Late-Archean crustal evolution: a two-stage sequence
ites). This model is supported by structural and textural evidence of
magma interaction (magma mingling, rapakivi textured feldspars and 3.1. Evidence from geochronology
quartz ocelli; Laurent, 2012; heterogeneity in zircon populations and
compositions; Laurent et al., 2013b), as well as variable Nd–Hf isotope It has long been demonstrated that during the Neoarchean,
compositions in hybrid granitoids (εNd = 0 to −5; Laurent, 2012) com- emplacement of dominantly granitic and potassic magmas followed
pared with sanukitoids s.l. (εNd = −3 to −4; Laurent et al., 2014). the massive formation of sodic TTGs in most Archean cratons
Emplacement ages (1 box = 1 sample) Ages of detrital zircons

Sanuk. Hybrid Bt-(Ms-) from the PGB (1 box = 1
s.l. granites granites TTG analytical spot)
plutons
Lekkersmaak plutons
3.3 IV
TTG
inTTG
Turfloop&&Lekkersmaak
III
3.2
bodies in
II
Small bodies
Stage 1 (~0.5 Ga)

3.1
Age (Ga)
Ib ? Emplacement of TTG
Small
3.0
Turfloop
Ia
2.9
2.8 Stage 2 (~0.15 Ga)

Emplacement of biotite- and two-mica
2.7 granites, hybrid granites, sanukitoids s.l.
Fig. 8. Compilation of (1) U–Pb and Pb–Pb emplacement ages obtained for the four granitoid groups in the Pietersburg block (data from Brandl and Kröner, 1993; Brandl et al., 1996; de Wit
et al., 1993; Henderson et al., 2000; Kröner et al., 2000; Laurent et al., 2013b; Poujol, 1997, 2001; Poujol and Robb, 1999; Zeh et al., 2009, 2013) — the different groups are colour-coded as in
Figs. 4 to 7; and (2) all age data from detrital zircons in the Pietersburg greenstone belt (PGB; grey boxes) (Zeh and Gerdes, 2012). See text for discussion.
Superior Amazonian Karelian Kaapvaal Dharwar N.China Pilbara

Province craton craton craton craton craton craton
2.40
2.45
2.50
2.55
2.60
2.65
2.70
2.75
Age (Ga)
Magmatic
phase 2
2.80
2.85 Magmatic
phase 1
2.90
2.95
3.00
3.05
3.10
3.15
3.20
TTG Sanukitoids "Hybrid" Biotite- and two-
magmatism s.l. granitoids -mica granites
Fig. 9. Top: map of exposed Archean crust (after Bleeker, 2003) in grey, and selected cratons of interest (in red) where the late-Archean granitoid evolution is well documented. Bottom: for
each one of these areas, simplified age distribution for the four groups of late-Archean granitoids defined in this study (modified and completed after Heilimo et al., 2011). The same
sequence is systematically repeated in every terrane, with a preliminary, long-lasting period of TTG emplacement (Phase 1) and a subsequent, much shorter time during which all the
other three groups are emplaced together or in rapid succession (Phase 2). For references and interpretation, see text and Table 1.
(e.g. Almeida et al., 2013; Champion and Smithies, 2003; Condie, 1981; during that period, but Zeh and Gerdes (2012) identified dis-
Heilimo et al., 2011; Jahn et al., 1988; Martin and Querré, 1984; Moyen crete episodes of more intense magmatic activity at ~ 3.30,
et al., 2003; Sylvester, 1994). Such an evolution is well illustrated by a ~ 3.25, ~ 3.20, ~ 3.10 and ~ 2.95 Ga (Fig. 7).
compilation of emplacement ages obtained from all granitoid rocks (2) Between 2.84 and 2.69 Ga, a briefer period during which all
of the Pietersburg block (Fig. 8). According to these data, the tempo- other types of granitoids were generated. Firstly, small masses
ral evolution of late-Archean granitoid magmatism in this area took of biotite- and two-mica granites intruded TTGs at 2.84 Ga;
place in two main stages: this event was followed, 2.75 to 2.79 Ga-ago, by the for-
mation of the compositionally similar, but larger Turfloop
(1) Between 3.30 and 2.85 Ga, a long period of almost exclusively and Lekkersmaak batholiths. All the hybrid granites were
TTG magmatism was recorded by emplacement ages from the then emplaced within a very short time span at around 2.69 Ga,
gneiss units. The ages of detrital zircons from sediments in the together with the sanukitoids s.l. of the Matok pluton and coeval
Pietersburg greenstone belt span this whole range (Fig. 8), in- with another major episode of crustal anatexis (e.g. Kreissig et al.,
dicating that TTG emplacement was more or less continuous 2000; Taylor et al., 2014).
Fig. 9 shows that there is a striking similarity between this emplace- tholeiitic mafic rocks; and (ii) the subsequent melting of the latter, giv-
ment sequence in the northern Kaapvaal Craton and granitoid evolution ing rise to the TTGs. The homogeneously high Hf–Nd isotope composi-
in other Archean terranes worldwide. Initially there is a long period (up tions and the wide spread of model ages over this long timespan
to 0.5 Ga, but generally in the range of 0.2–0.3 Ga) of almost exclusive TTG indicate that mafic rocks were being continuously generated from the
magmatism, building up large volumes of felsic continental crust. Imme- mantle and subsequently re-melted to produce TTGs over several
diately after TTG emplacement, and within a shorter timespan (~0.05 Ga hundreds of Ma, and that the length of time between the two steps
on average), the whole range of high-K granitoids (sanukitoids s.l., remained relatively short throughout this period. In sharp contrast,
biotite-and two mica-bearing granites, hybrid granites) is generated the isotopic signature of the second stage points to the coeval involve-
and emplaced. ment of mantle sources (many sanukitoids have positive εNd(t);
The planetary scale of this event is highlighted by the fact that, in Fig. 11) and recycling of the continental crust generated during the
most cratons, it took place at the end of Archean aeon, between 3.0 first stage. This second stage has a greater diversity of sources and
and 2.5 Ga (Fig. 9). However, a comparative scrutiny of the evolution petrogenetic mechanisms, which can be separated into three catego-
shows two main temporal differences from one craton to another: ries: (i) melting of an “enriched” mantle, to generate mafic magmas
which subsequently differentiated to give rise to sanukitoid (s.l.)
(1) The absolute lengths of the two magmatic stages vary widely suites; (ii) melting of pre-existing crustal rocks (mostly TTGs and
from one region to another. Extreme examples are the Amazo- minor metasediments) to produce biotite- and two-mica granites;
nian Craton, where the episode of TTG emplacement lasted (iii) interaction between magma types from (i) and (ii) (or their
~ 0.1 Ga, and the second magmatic stage less than 30 Ma sources) to generate the hybrid granitoids. Crustal recycling may
(Fig. 9), and the Kaapvaal Craton, in which these two events therefore have resulted either from direct intracrustal melting (pro-
spanned over ~ 0.50 and ~ 0.15 Ga, respectively (Figs. 8 and 9). ducing biotite- and two-mica granites) or introduction of crustal-
However, the relative duration of the two stages was identical derived material into the mantle, forming the metasomatized mantle
in both cases. source of the sanukitoids s.l. and explaining why some of them have
(2) The transition from one magmatic stage to the other took relatively evolved isotopic signatures (Figs. 10 and 11).
place at very different times globally (as early as ~ 2.95 Ga in
the Pilbara Craton, but as late as ~ 2.55 Ga in the Dharwar and 4. Geodynamic implications: late-Archean granitoid evolution
North China cratons; Fig. 9), highlighting the diachronic character reflecting subduction and continental collision
of late-Archean magmatism, as already discussed by Almeida
et al. (2013), Heilimo et al. (2011) and Champion and Smithies 4.1. First stage: genesis of TTGs
(2003).
4.1.1. Possible geodynamic environments
3.2. Evidence from isotope data The geochemical diversity of TTGs, which is the main component
of Archean crust, indicates that these were generated by melting of
The temporal evolution of late-Archean granitoid magmatism is metabasalt at various depths, which can be achieved in several
recorded not only by the geochronology, but also by Hf-Nd isotope sys- different geodynamic settings (see Moyen, 2011; Moyen and
tematics. Fig. 10 shows the evolution of initial isotope compositions of Martin, 2012 for review). These include (i) subduction-like settings
Hf (zircon) and Nd (whole-rock) in granitoids in the northern Kaapvaal in which subducted meta-basalts (oceanic crust or plateaus) melted
Craton. The first magmatic stage (3.33–2.85 Ga) is characterized by pos- due to higher Archean geothermal gradients (e.g. Defant and
itive εHf values in zircons from TTGs (εHf(t) = 0 to +5; Fig. 10a), as well Drummond, 1990; Martin, 1986, 1987, 1999; Martin and Moyen,
as a close similarity between magmatic ages of TTGs (2.85–3.33 Ga; 2002; Martin et al., 2014; Rapp et al., 2003); (ii) tectonic environ-
Fig. 8) and their Nd model ages (2.95–3.30 Ga; Fig. 10b). Both lines ments unrelated to plate boundaries, such as the progressive matura-
of evidence unequivocally indicate that this was a period of juvenile tion of an oceanic plateau above a long-lived mantle plume (e.g.
magmatism and therefore of continental crust extraction from the Bédard, 2006, 2013; Bédard et al., 2003; Smithies, 2000; Willbold
mantle. In contrast, granitoids emplaced during the second stage (bi- et al., 2009; Zegers and van Keken, 2001) and (iii) intermediate scenar-
otite- and two-mica granites, sanukitoids s.l. and hybrid granites) ios, such as delamination at the base of a magmatically or tectonically
clearly show more scattered and more evolved, “enriched” signa- over-thickened (N30 km) mafic crust (e.g. Bédard et al., 2003;
tures (εNd(t) = 0 to − 5; εHf(t) = + 2 to − 5), pointing to significant Johnson et al., 2013; Moyen, 2011; van Thienen et al., 2004). Here, we
reworking and recycling processes. In other words, the geological propose that subduction is arguably the most plausible geodynamic en-
history of the Pietersburg block is characterized by long-lasting, in- vironment to explain the genesis of TTGs in the first stage of the late-
tense juvenile crust formation and ended up with the recycling of Archean granitoid evolution.
this preexisting crust.
A review of available whole-rock Nd isotope data from late- 4.1.2. Late-Archean TTGs formed in a subduction setting: geometric and
Archean granitoids in four reasonably well-documented cratons geochemical constraints
(Baltic Shield, North China Craton, East Dharwar Craton and Superior As discussed earlier, most of late-Archean (b3.0 Ga-old) TTGs are
Province; after Dey, 2013) shows that this evolution is systematic Sr-rich and Y–HREE-poor, thus corresponding to the “low-HREE”
(Fig. 11). The first stage of the magmatic sequence defined in group of Halla et al. (2009) and the “medium-” to “high-pressure”
Section 3.1 is characterized either by (i) emplacement of juvenile groups of Moyen (2011), which these authors consider as having been
TTGs, having dominantly positive εNd(t), over a long period; or (ii) generated at pressures N1.0 GPa (i.e. at depths N35 km). Such melting
major crust-forming events depicted by Nd model ages of TTGs is unlikely to have taken place in the crust, especially in the Archean
ranging over that time span (Fig. 11). In comparison, the much context where a higher heat flux from the mantle resulted in a weaker
shorter second magmatic stage is always associated with greater crust, unable to sustain thicknesses N30 km (England and Bickle,
variability of Nd isotope compositions (up to 10 epsilon units; Fig. 11) 1984; Johnson et al., 2013; Rey and Coltice, 2008). Thus, the genesis
and obvious involvement of older crustal components in granitoid of these magmas requires that (hydrous) mafic rocks were somehow
sources (highly negative εNd(t)). introduced deep into the mantle.
According to the petrogenetic model of TTGs described in This requirement is easily achieved in the context of a subduction
Section 2.1.3, new crust was formed during the first magmatic setting. It may also be fulfilled by other scenarios, though, especially if
stage in two successive steps: (i) melting of the mantle to produce the base of the mafic crust becomes gravitationally unstable owing to
(a) +8 (b) +6
Detrital zircons from the Pietersburg greenstone belt
(Blich
ert-To
DM Stage 2 Stage 1
ft and
Putch
el, 20
10)
DM
+4 +3 (dePaolo,
1981) TTG Model Ages
εNd (t) Whole rock

εHf (t) Zircon
0 CHUR 0 CHUR
–4 –3
Isotopic evolution of Pietersburg
TTGs samples
Stage 2 Stage 1
–8 –6
2.50 2.70 2.90 3.10 3.30 3.50 2.50 2.70 2.90 3.10 3.30 3.50
Age (Ga) Age (Ga)
Fig. 10. Hf–Nd isotopic evolution of late-Archean granitoids in the Pietersburg block (symbols as in Figs. 4 to 7): (a) εHf(t) of zircon vs. age diagram for samples investigated by Zeh et al.
(2009, 2013) (one symbol = average of several LA-MC-ICPMS zircon analyses, typically ~15) and individual analyses of detrital zircon grains from sediments of the Pietersburg greenstone
belts (grey circles; Zeh and Gerdes, 2012). DM model is after Blichert-Toft and Putchel (2010); (b) εNd(t) of whole rock vs. age diagram (data from Henderson et al., 2000; Laurent, 2012
and Laurent et al., 2014); the dashed grey lines represent the isotopic evolution of TTG samples for which no absolute age is available (data from Kreissig et al., 2000). The two stages of the
magmatic evolution defined on geochronological grounds in Section 3.1 and Fig. 8 are also shown (yellow and grey areas). Note that Stage 1 corresponds to (i) a long period of
suprachondritic Hf isotope composition and (ii) a major crust-forming event as witnessed by TTG model ages, whereas Hf–Nd signatures are more scattered and evolve towards negative
values during Stage 2.
+4 +8
+3 +6
DM TTG Model Ages
+2 DM
+4
εNd (t) Whole rock
εNd (t) Whole rock
+1
+2
0 CHUR
0 CHUR
–1
–2
–2
Baltic shield –4 North China
–3
–4 Finland –6 craton
–5 –8 Mafic rock
One sample at –10
–6 –10
2.60 2.70 2.80 2.90 3.00 3.10 3.20 3.30 3.40 2.30 2.40 2.50 2.60 2.70 2.80 2.90 3.00 3.10
Age (Ga) Age (Ga)
+6 +6
+4 DM +4 DM TTG Model Ages
+2 +2
εNd (t) Whole rock
εNd (t) Whole rock
0 CHUR 0 CHUR
–2 –2
Early reworking ?
–4 –4
Superior Province East Dharwar
–6 –6
Canada craton
–8 –8 India
–10 –10
2.50 2.60 2.70 2.80 2.90 3.00 3.10 2.40 2.50 2.60 2.70 2.80 2.90 3.00 3.10 3.20
Age (Ga) Age (Ga)
Fig. 11. Compilation of εNd(t) of whole rocks vs. age diagram for late-Archean granitoids in four different regions. Each sample is assigned to one of the four groups of our classification:
TTGs (squares), sanukitoids s.l. (triangles), biotite- and two-mica granites (circles) and hybrid granites (diamonds). The grey dashed lines represent the isotopic evolution of TTG samples
for which no absolute age is available. The two stages of the magmatic evolution are as in Fig. 9. The isotopic evolution is comparable to that of late-Archean granitoids in the Pietersburg
block (Fig. 10), as the first stage (yellowish fields) is characterized either by emplacement of juvenile TTGs (and coeval mafic rocks) or represents a protracted period of crust formation (Nd
model ages), whereas the second stage (greyish fields) shows much more scatter over a shorter timespan, and overall a salient trend towards more negative εNd(t). Modified and
completed after Dey (2013); for references, see Table 1.
the formation of eclogites at depth. Formation of an eclogitic root can 4.1.3. The meaning of “subduction”
be due to partial melting of mafic material at the base of an oceanic The conclusion that late-Archean TTGs formed in a “subduction”
crust or plateau, producing garnet- and pyroxene-bearing residues environment does not imply necessarily that those “subduction” zones
(e.g. Bédard, 2006; Zegers and van Keken, 2001). However, such were similar to present-day situations (i.e. a long-lived, continuous
delaminated residues would consist of anhydrous and refractory ma- slab regularly plunging along an active margin several thousands of
terial, which is unlikely to melt any further and thus yield the great kilometres long), neither that they were the driving force of plate tec-
volumes of TTG observed in late Archean cratons. Alternatively, de- tonics at that time. Moyen and van Hunen (2012) proposed that, unlike
lamination can take place “spontaneously” if the mafic crust is the present-day situation, an Archean subducted slab could not have
thick enough for eclogitic assemblages to be stable at its base, but been stable, but would have been subjected to frequent break-off, lead-
this can take place only if the crust has a komatiitic composition ing to the episodic sinking of basaltic fragments into the mantle. Such a
(N 21 wt.% MgO; Johnson et al., 2013), whilst the source of TTGs is basal- process is a possible scenario for the late-Archean period and would
tic rather than komatiitic. The only way to generate “medium-” and typically result in several pulses of TTG magmatism, as seen in the
“high-pressure” TTGs by melting delaminated mafic rocks is therefore Pietersburg block (Zeh and Gerdes, 2012; see Section 3.1 and Fig. 8)
to consider a tectonically over-thickened (i.e. gravitationally unstable) and several other terranes of that age (Figs. 9 and 11). Likewise, if one
basaltic crust (Moyen, 2011). considers melting of subducted oceanic plateaus, the episodic produc-
Although plausible to explain the origin of TTGs as such, this tion of the latter would account for successive pulses of TTG magmatism
possibility is challenged by the whole evolution of late-Archean (Martin et al., 2014).
granitoids, and in particular the subsequent emplacement of On the other hand, if late-Archean subduction was actually an unsta-
sanukitoids (s.l.) during the second stage of the magmatic se- ble, transient process, then it was probably unable to drive plate tecton-
quence. Sanukitoids were generated by interactions between man- ics efficiently (Moyen and van Hunen, 2012; O'Neill et al., 2007b; Sizova
tle peridotite and LILE-, LREE-rich material (see Sections 2.2.3 and et al., 2010; van Hunen and van den Berg, 2008). Recently, Bédard et al.
3.2). Such material had to be water-rich in order to account for (2013) proposed a solution to this paradox proposing that nuclei of
the generally high H 2 O contents of the “primary” sanukitoid continental crust were actually the active tectonic agent, drifting in
magmas (Laurent et al., 2013a; Lobach-Zhuchenko et al., 2008; response to the entrainment of their stiff mantle keel in mantle convec-
Oliveira et al., 2009, 2010). Delaminated basalts would not be rich tion currents. At their leading edges, such drifting proto-cratons would
enough in incompatible elements and water to represent such material. override oceanic, mafic crust or plateaus, which in turn would be buried
Alternatively, melting of delaminated basaltic rocks might have pro- deep in the mantle and eventually melt to produce “medium-” to “high-
duced LILE-, LREE-rich and hydrous TTGs, which in turn were responsi- pressure” TTGs; a scenario equivalent to subduction from a geometric
ble for metasomatism in the surrounding mantle. In that case, (and geochemical) point of view.
sanukitoids would be expected to have been produced at any time
together with TTGs, in response to the successive events of mantle re- 4.2. Second stage: genesis of sanukitoids s.l., biotite-, two-mica granites and
enrichment by delamination of lower crustal material. Instead, the evo- hybrids
lution of late-Archean granitoids shows that sanukitoids s.l. are rather
associated with biotite-, two-mica and hybrid granites and largely re- 4.2.1. Possible geodynamic environments
stricted to the second stage of the magmatic evolution (see Section 3 In all late-Archean terranes the granitoids emplaced during the
and Figs. 8 to 11). second magmatic stage (sanukitoids s.l., biotite- and two-mica granites,
Moreover, many authors proposed that the mantle source of some hybrid granitoids) result from different degrees of interaction between
sanukitoids s.l. interacted with sediments, or sediment-derived two main source end-members: (i) the local continental crust and (ii)
fluids/melts (Halla, 2005; King et al., 1998; Laurent et al., 2011, mantle peridotite metasomatically enriched in incompatible elements.
2014; Lobach-Zhuchenko et al., 2008; Mikkola et al., 2011b), which This implies that both geochemical reservoirs were associated geomet-
are most easily introduced to mantle depth by subduction. Delami- rically, and requires a thermal anomaly to account for their coeval melt-
nation of supracrustal material could be an alternative to subduction ing. There are two different geodynamic environments which fit these
(e.g. Bédard, 2013), but such buoyant material would be hard to constraints: (i) an event of terrane accretion (i.e. continental collision)
transfer to lower crustal depths even considering a classical convective where crustal thickening followed by slab breakoff, retreat or litho-
overturn. And if so, it is unlikely that it would survive its transfer spheric delamination provides the required heat source (Dey et al.,
through a hot, convective and certainly partially molten crust. Finally, 2014; Feng and Kerrich, 1992; Halla et al., 2009; Heilimo et al., 2010;
the radiogenic isotope signature of many sanukitoids (Fig. 11) indicates Laurent et al., 2013a, 2014; Mikkola et al., 2011b; Smithies et al., 2007;
that the interactions between mantle and enriched components took Whalen et al., 2004), or (ii) an intraplate setting, where the
place shortly (b0.3 Ga and most often b0.1 Ga) before emplacement lithosphere is affected by a thermal event such as plume upwelling
(Halla, 2005; Käpyaho et al., 2006; Kovalenko et al., 2005; Laurent (Jayananda et al., 2000) or rifting and break-up of a stable continent
et al., 2011, 2014; Shirey and Hanson, 1984). This supports the (Smithies and Champion, 2000). Following most authors, we prefer
hypothesis that sanukitoid genesis took place soon after a subduc- the model of terrane amalgamation and continental collision, as
tion event. discussed below.
Importantly, the presence of “low-pressure” (Sr- and Al-poor, Y- and
HREE-rich) TTGs in the late-Archean geological record (Almeida et al., 4.2.2. Continent–continent collision: analogy with post-Archean collisional
2011; Feng and Kerrich, 1992; Halla et al., 2009; Whalen et al., 2002; granitoids
see Section 2.1.3) does not preclude a convergent plate setting. As pro- Clearly, the range of granitoids formed during the second stage of
posed by several authors, subduction can take place in an “intra-ocean- the late-Archean evolution are more akin to those formed in post-
ic” setting, e.g. beneath a thick oceanic plateau (Almeida et al., 2011, Archean collisional settings than intraplate environments. Indeed,
2013; Halla et al., 2009; Heilimo et al., 2010). This particular setting calc-alkaline, K-rich granitoids generated in syn- to late-collisional con-
would adequately account for the coexistence, and even the co- texts of b 2.5 Ga-old orogens also derive from the respective contri-
magmatic character in some cases (Almeida et al., 2011), of “low- bution of local crust and metasomatized lithospheric mantle, whilst
pressure” and “high-pressure” TTGs. The latter thus represent “slab granitoids formed in purely intraplate settings are invariably alka-
melts”, whilst the former originated through partial melting at the line and less diverse (Bonin, 2004; Clemens et al., 2009; Debon and
base of the thick, overlying mafic pile heated up by thermal conduc- Lemmet, 1999; Downes et al., 1997; Ferré et al., 1998; Frost et al.,
tion and advection. 2001; Harris et al., 1986; Liégeois et al., 1998; Turner et al., 1992). For
10000
instance, Cenozoic magmatism in the Himalayas is characterized by the Melting of enriched
coexistence of high-K granitoids generated by melting of local crust (ei- mantle
5000
ther metasediments for the High Himalayan Leucogranites or meta- Melting of
igneous lithologies for calc-alkaline granitoids; e.g. Aikman et al., crustal lithologies
2012; Chung et al., 2003; Guo and Wilson, 2012; Harrison et al., 1997, 2000
1999; Le Fort et al., 1987; Reichardt et al., 2010; Zeng et al., 2011) and
Sr + Ba (ppm)
high-K calc-alkaline, shoshonitic to ultrapotassic, mafic and intermedi- 1000 UC
Mantle enrichment
ate volcanics, derived from metasomatized phlogopite- and/or BC
amphibole-bearing mantle peridotite (e.g. Gao et al., 2010; Mahéo 500

et al., 2009; Miller et al., 1999; Williams et al., 2004; Xu et al., 2001;
OIB
"Mantle
Zhao et al., 2009).
200
In order to support this analogy, we present here a comparison of the
granitoids emplaced during the second stage of the late-Archean
MORB
array"
100
magmatic evolution with Paleozoic granitoids from the French Massif
Central (FMC). There, erosion has exposed the mid-crustal levels
1000
of the Variscan orogenic belt and in particular, huge amounts of Carbon-
iferous syn- to late-collisional granitoids. These granitoids can be classi- 500
fied into three groups (Fig. 12):
(1) Low FeO t + MgO (b10 wt.%) peraluminous muscovite- or UC

100 BC
cordierite-bearing granites, with moderate incompatible
element contents (e.g. Couturié, 1977; Ledru et al., 2001; 50
Mantle enrichment
LaN
OIB
Turpin et al., 1990). All were generated by partial melting of
a crustal (generally meta-sedimentary) source (Barbey et al.,
"Mantle
1999; Williamson et al., 1992, 1997). 10
(2) High FeO t + MgO, metaluminous, intermediate to mafic
array"
(SiO2 = 45–60 wt.%) monzo- and syenodiorites, rich in K 5
MORB
and incompatible elements (4 ≤ K2O ≤ 9 wt.%), locally referred
to as “vaugnerites” (Aït-Malek, 1997; Couzinié et al., 2014;
Galán et al., 1997; Gardien et al., 2011; Ledru et al., 2001; 1000
Michon, 1987; Montel and Weisbrod, 1986; Solgadi et al.,
2007). Their genesis requires an enriched mantle source
500
(e.g. Agranier, 2001; Couzinié et al., 2014; Debon and Lemmet,
1999; von Raumer et al., 2013).
(3) “Sub-alkaline” granites, which are K-rich, generally amphibole-
Nb + Zr + Y (ppm)
UC
200
bearing granitoids containing abundant microgranular mafic BC
enclaves (e.g. Didier et al., 1989; Solgadi et al., 2007). These
Mantle enrichment
granitoids are believed to result from mixing or mingling of 100
OIB
the previous two groups (Downes et al., 1997; Solgadi et al.,
2007) which gave rise to a great variety of hybrid and hetero- 50 "Mantle
geneous phases.
array"
MORB
As shown in Fig. 12, the composition of these three groups of granit- 20
oids from the FMC overlaps those of granitoids emplaced during the sec-
ond stage of the late-Archean granitoid evolution. The “vaugnerites” are 10
the richest in both ferromagnesian oxides and incompatible elements, 0.2 0.5 1 2 5 10 20 50
thus displaying the same features as sanukitoids s.l. On the other FeOt + MgO (wt.%)
hand, FMC peraluminous granites are comparable to late-Archean
biotite- and two-mica granites, in the sense that they derive from Peraluminous
"Vaugnerites"
granites
similar petrogenetic processes, i.e. melting of the dominant crustal
"Subalkaline" granites
lithologies (which are mostly TTGs in Archean terrains and and their enclaves
BC UC Bulk / Upper Crust
metasediments in post-Archean orogens, thus accounting for the
limited occurrence of two-mica granites before 2.5 Ga). In addition, a Fig. 12. Plots of incompatible element contents (Sr + Ba; La concentrations normalized to
variety of granitoids range between these two end-members, i.e. C1 chondrite value of McDonough and Sun, 1995; and sum of HFSE = Nb + Zr + Y)
the late-Archean hybrid granitoids and the late Carboniferous “sub-al- plotted as a function of FeOt + MgO (wt.%) concentrations. These plots illustrate the sim-
kaline” ones. Therefore, it can be reasonably concluded that ilarity between post-Archean, late-collisional granitoid associations (represented by the
three groups of late-Variscan granitoids of the French Massif Central (FMC): “vaugnerites”,
the granitoids which characterize the second magmatic stage of the
peraluminous granites, “subalkaline” granites — coloured fields; database available on
late-Archean evolution were generated in a syn- to late-collisional request), and the late-Archean granitoids emplaced during the second stage of the mag-
geodynamic environment. matic evolution discussed in Section 3: sanukitoids s.l., biotite-bearing granites and hybrid
This hypothesis is also supported by isotope data from late-Archean granitoids (represented by samples from the Pietersburg block; symbols as in Figs. 4 to 7).
granitoids: the second stage of the magmatic evolution is indeed The composition of the average bulk and upper continental crust (data from Rudnick and
Gao, 2003) as well as the composition of mantle-derived magmas (MORB and OIB) are
marked by a greater diversity of initial Nd–Hf isotope signatures than
also presented for comparison. The dashed line separating compositions of magmas
the first one, and notably of a spread towards less radiogenic composi- derived from crustal- and enriched mantle sources is qualitatively represented by the
tions (see Figs. 9 and 11 and Section 3.2). In the post-Archean record, “intracrustal” differentiation trend, i.e. the line joining the upper- and bulk crustal
such distinct isotopic evolutions are generally observed when a conver- compositions.
gent (i.e. subduction) environment turns into continental collision, for
example in the Lachlan fold belt in SE Australia (e.g. Hawkesworth (which can be limited, for example in the case of a “hot” or retreating
et al., 2010; Kemp et al., 2009) or the Kohistan area in Pakistan orogen). Nevertheless, we emphasize here that the common factor in
(Bouilhol et al., 2013). The drift towards less radiogenic Nd–Hf iso- all these scenarios is the involvement of both metasomatized mantle
tope compositions is also very consistent with an increase in and crustal lithologies in the genesis of granitoid magmas, irrespective
reworking and recycling processes, respectively associated with of the geometry of the system and the heat source at the origin of
thickening of the crust during collision (genesis of biotite- and two- crust and lithospheric mantle melting.
mica granites) and melting of “enriched” mantle sources with low
Sm/Nd and Lu/Hf ratios during the post-collisional stage.
Finally, sanukitoids s.l., biotite-, two-mica granites and hybrid 4.3. A global geodynamic model
phases are either coeval or emplaced within a short time span immedi-
ately after the waning of TTG magmatism. This very specific evolution The above evidence points to the evolution of late-Archean granit-
took place on a planetary scale between 3.0 and 2.5 Ga, but at slightly oids recording an episode of subduction (first stage) and subsequent
different times from one craton to another (see Section 3.1). If the continent–continent collision (second stage). This conclusion is consis-
activity of mantle plumes was at the origin of these granitoids, then tent with those of previous studies focusing on the diversity of late-
plume upwelling would have taken place at different times from Archean granitoids in a given cratonic area (Almeida et al., 2011,
place to place, but always immediately after a period of TTG magmatism, 2013; Dey et al., 2014; Feng and Kerrich, 1992; Halla et al., 2009;
which is implausible. Moreover, there is a strong structural control on Käpyaho et al., 2006; Moyen et al., 2003; Smithies and Champion,
granitoid emplacement during the second stage: magmas specifically 2000; Whalen et al., 2004) and with reviews of geodynamic evolution
intrude along crustal-scale structures (Champion and Sheraton, 1997; in individual Archean terranes (Percival et al., 2006; Smithies et al.,
Jayananda et al., 1995; Moyen et al., 2001), at the boundaries between 2007; van Kranendonk et al., 2007). Hence, a tentative, generic
distinct crustal domains (Bibikova et al., 2005; Heilimo et al., 2010; geodynamic evolution for late-Archean terranes is proposed in Fig. 13.
Lobach-Zhuchenko et al., 2005; van Kranendonk et al., 2007) and/or in During the first stage, subduction takes place with possible episodic
association with regional tectono-metamorphic overprints (Laurent break-off and foundering of “rafts” of basaltic oceanic crust and/or oce-
et al., 2011, 2014). In addition, in a given cratonic area, late-Archean anic plateaus (Fig. 13a). This leads to (i) several pulses of juvenile TTG
granulites are generally coeval with the granitoids emplaced during magmatism over a long period of time and (ii) metasomatism of the
the second stage of the magmatic evolution (Percival, 1994). All lines overlying “mantle wedge” by fluids or melts derived from metabasalts
of evidence thus support the granitoids being spatially and temporally (i.e. TTGs) and/or crust-derived, terrigenous sediments. Subduction
associated with terrane amalgamation rather than a plume or rifting can take place either along a pre-existing, Paleoarchean continental
event. nucleus (e.g. in the northern Kaapvaal Craton) or in an “intra-oceanic”
setting, for instance beneath a thick and recent oceanic plateau
(Almeida et al., 2011, 2013; Halla et al., 2009) (Fig. 13a; see also
4.2.3. The meaning of “continent–continent collision” Section 4.1). Ultimately, this first stage leads to the closure of an
The terminology of “continent–continent collision” that we use here oceanic basin and welding of two (proto)continental blocks (conti-
has a very general meaning and refers primarily to the amalgamation of nental collision). At that stage, sanukitoids s.l. can be generated by
two continental blocks. Continental collision can be accommodated by significant interaction between the last TTG melts and peridotite,
different orogenic styles, including retreating or advancing plate bound- whilst thickening of the crust triggers its partial melting, giving rise
aries followed by collision proper, which can itself take place and evolve to biotite- and two-mica granites (Fig. 13b). Sanukitoid s.l. can also
according to different scenarios (such as slab retreat, break-off, or ongo- be emplaced during the late- to post-collisional period owing to
ing convergence and formation of an orogenic plateau) depending on thermal relaxation and extension, which induce melting of the previ-
the thermo-mechanical conditions of the system (Cawood et al., 2009; ously enriched lithospheric mantle (Fig. 13c). Slab break-off, slab
Duretz and Gerya, 2013; Vanderhaeghe, 2010; and references therein). retreat or lithosphere delamination, which is likely to take place in
These different possibilities vary primarily in terms of geothermal gradi- the late- to post-collisional stages (e.g. Bonin, 2004; Duretz and
ents, kinematics, deformation patterns and extent of crustal thickening Gerya, 2013; Liégeois et al., 1998; van Hunen and Allen, 2011;
(a) ~20 km (b) Tectonically and magmatically (c) Thermal relaxation

Collapse
thickened crust
Formation of (continent-continent ‘‘orogen’’)
juvenile TTG crust Deposition of large shelf sequences
Older crust
Mafic crust
Crustal melting Crustal melting

Melting of
metasomatized
Metasomatized Hybridization mantle
‘‘Subduction’’ mantle
of oceanic crust Mantle TTG/mantle
metasomatism
Asthenosphere upwelling?
Magmatic stage 1 Magmatic stage 2

Subduction - TTG Collision - Sanukitoids s.l., crust-derived- and ‘hybrid’ granitoids
~ 0.2 to 0.5 Ga ~ 0.02 to 0.15 Ga
TTG magmatism Sanukitoids s.l. Biotite- and two-mica ‘‘Hybrid’’ granitoids

granites
Fig. 13. Synthetic, conceptual cartoon illustrating the proposed geodynamic environment corresponding to the evolution of late-Archean granitoids, i.e. a subduction–collision sequence.
See text for discussion.
Vanderhaeghe, 2010), has been consistently proposed as tectonic

triggers for the generation of sanukitoids s.l. (Almeida et al., 2013; TTG
Halla et al., 2009; Laurent et al., 2014; Lobach-Zhuchenko et al., Na2O / K2O
H.
2008; Mikkola et al., 2011b; Whalen et al., 2004). The resulting ther-
Bt. γ Snk. 3.6
mal anomaly can also promote further crustal anatexis. During both
syn- (Fig. 13b) and late-collisional (Fig. 13c) periods, interaction be- 3.5
Age (Ga)
tween sanukitoids s.l. and crust-derived melts potentially generates
a wide spectrum of hybrid granitoids. 3.4
ra
ba
Depending on local conditions, such as the composition of the
3.3
Pil
sources involved in magma genesis, the nature of the resulting granit-
st
oids can vary from one craton to another. This reliably explains (i) 3.2
Ea
why some granitoid types are more common in certain cratons than
( a)
others, and (ii) the internal differences within each group from distinct
terranes. On the other hand, as shown in Fig. 13b–c, the granitoids
related to the second magmatic stage can be generated at different
periods during continent-continent collision, adequately explaining
2A
B
MS
the differences in the relative and absolute chronology of granitoid
/ CN
emplacement from one craton to another. Furthermore, the geometry
2F
Na2O / K2O
K
of the system and orogenic style are also likely to play a role. For
instance, the Pietersburg block is believed to have formed as a 3.9
“Cordilleran” convergent margin, with successive accretion of micro-
continents at the northern edge of the Kaapvaal craton (e.g. Zeh et al., 3.8
Age (Ga)
2013). In this context, a protracted period of crustal reworking and
3.7
nd
mantle enrichment would be expected, explaining the long duration
nla
of the second stage (~ 0.15 Ga) relative to other cratons (see Figs. 8 3.6
ee
and 9).
3.5
Gr
5. The onset of “modern-style” plate tectonics
(b)
to 3.0 Ga
5.1. Comparison with Eo- to Mesoarchean granitoids
A two-stage evolution of granitoid magmatism, from dominant

2A
B
MS
TTG magmas to the emplacement of high-K granites, has also been
/CN
2F
reported in some early Archean terranes: (i) the area of the Barber-
K
ton greenstone belt (southeastern Kaapvaal Craton; see Fig. 1) Na2O / K2O
3.6
between ~ 3.55 and ~ 3.09 Ga (Clemens et al., 2010; Kamo and Davis,
1994; Sanchez-Garrido et al., 2011; Yearron, 2003; Zeh et al., 2009); 3.5
(ii) the North Atlantic Craton (Greenland) between ~ 3.85 and
Age (Ga)
3.4
~ 3.50 Ga (Friend and Nutman, 2005; Nutman et al., 2007, 2013);
and (iii) the East Pilbara Craton (NW Australia) between ~ 3.50 and
n
3.3
rto
~ 3.25 Ga (Barley and Pickard, 1999; Bickle et al., 1989; van

rbe
Kranendonk et al., 2007). 3.2

Ba
For each region, when these granitoids are plotted in the ternary 3.1
synthetic classification diagram (see Fig. 7 and Section 2.5), the oldest
(c)
to 3.0 Ga
granitoids match the “TTG field”, whilst the youngest one show more
diversity (Fig. 14). In the case of the East Pilbara (Fig. 14a) and SW
Greenland (Fig. 14b), these “second-stage” granitoids spread between
the fields of the biotite- and two-mica granites, on one hand, and that GMS suite
2A
B
MS
of the TTGs, on the other hand. However, both terranes strikingly lack
/CN
2F
a sanukitoid s.l. component (Fig. 14a–b), and are therefore not equiva-
K
lent to late-Archean domains. The situation is rather different in the

Barberton area where some samples of the ~ 3.1 Ga-old Granite– Fig. 14. Comparison of early Archean (N3.0 Ga-old) granitoids with the fields of
Monzogranite–Syenite (GMS) suite plot within the field of sanukitoids late-Archean (3.0–2.5 Ga) granitoids (inset in the top left corner; Snk. = sanukitoids s.l.,
s.l. (Fig. 14c). Although these samples are subordinate in volume, they Bt. γ = biotite- and two-mica granites, H. = hybrids) in the synthetic classification
possibly depict the involvement of enriched mantle sources, which is diagram of Fig. 7 (see also Section 2.5). Early Archean data are from three different
Paleo- to Mesoarchean terranes in which a secular evolution of granitoid compositions
consistent with the proposed occurrence of a ≥3.23 Ga-old subduction
has been reported: (a) the East Pilbara Craton, NW Australia; (b) SW Greenland; (c) the
event in this area (Moyen et al., 2006). Therefore, the evolution of gran- area of the Barberton Greenstone Belt, South Africa. In a given area, granitoids are classi-
itoids in the Barberton region possibly results from successive subduc- fied into two groups: (i) “early” granitoids (open dark green symbols) and (ii) “late” gran-
tion and collision, thus representing an early example of the late- itoids (filled light green symbols) on the basis of published models for crustal evolution in
Archean magmatic sequence described in this review. these terranes. The adjacent timelines indicate the approximate age range for each group.
In (c), the dashed field encloses samples of the Granite–Monzogranite–Syenite (GMS)
However, to our knowledge, the diversity of granitoids reported in suite of the Barberton region. Data are from published or personal databases (Moyen,
the Barberton area has no older equivalent. This means that the typical 2011; Moyen and Martin, 2012; J.F. Moyen).
late Archean granitoid evolution as described here is either lacking, or
very unusual, in terranes older than 3.0 Ga. In contrast, and despite
some local differences, it took place in almost every Archean craton
between 3.0 and 2.5 Ga. Since this magmatic evolution is very similar
to that of modern subduction–collision settings (see Section 4), it can be Ga (Fig. 15a), thus “modern” behaviour could not be adopted until the
concluded that such geodynamic mechanisms have only been operating very end of the Archean. This threshold would actually have been
globally since 3.0 Ga. Thus, whilst subduction and continental collision reached somewhat earlier (~ 3.0 Ga) considering that komatiites asso-
could have taken place before 3.0 Ga as discrete events in both space ciated with basalts in the oceanic crust would have significantly decreased
and time, it is only from the late-Archean that global plate tectonics its buoyancy (Barbey and Martin, 1987; Johnson et al., 2013; Martin et al.,
were characterized by a cyclic pattern of successive assembly and 2014). Alternatively, Hargraves (1986) proposed that a higher heat pro-
break-up of continental masses, with the late-Archean evolution being duction in the early Earth could have been accommodated by increasing
the very first of these cycles. the ridge length rather than the thickness of oceanic crust. In this case,
the size of oceanic plates would have been smaller than today (de Wit
5.2. Support from other geological evidence and Hart, 1993; Lagabrielle et al., 1997), and the geometry of subduction
systems different. If this was the case, then the establishment of “modern-
Our interpretation of the late-Archean granitoid evolution is sup- style”, stable subduction zones by the end of the Archean reflects a pro-
ported independently: (i) the chemistry of continental basalts changed gressive increase in oceanic plate size.
across the Archean–Proterozoic boundary, reflecting an evolution The emergence of continental collision and the increasing impor-
towards LILE-richer mantle sources (Condie, 1989; Keller and Schoene, tance of reworking processes at the end of the Archean are due to an
2012; Smithies et al., 2004, 2007), which is consistent with the initiation evolution of both the volume and thermo-mechanical characteristics
of crustal recycling into the mantle owing to either rising volumes of of continental crust. Firstly, a trivial but fundamental parameter is
continental crust or increasingly efficient subduction processes; (ii) Hf that, before being able to recycle continental crust, a significant volume
and O isotopic compositions of detrital and magmatic zircons indicate of this crust must already exist. Hence, the late-Archean period follows a
changes in crustal growth rates and granitoid sources at 3.0–2.5 Ga, long period of predominantly juvenile magmatism (Dhuime et al., 2011;
interpreted as the onset of subduction- and collision-driven continental Guitreau, 2012; Kramers and Tolstikhin, 1997; Taylor and McLennan,
recycling (Dhuime et al., 2011; Valley et al., 2005); (iii) Shirey and 1985, 1995; Fig. 15b) and thus marks the moment when the volume
Richardson (2011) noted that lithospheric inclusions in diamond crys- of the continental blocks became sufficient for them to be welded to
tals changed from peridotitic to peridotitic and eclogitic at ~ 3.0 Ga, each other to form collisional belts, and undergo significant internal dif-
interpreted as reflecting the initiation of subduction and collision pro- ferentiation (producing large amounts of anatectic granite, i.e. biotite-
cesses; (iv) the appearance of shelf sequences and mature sediments and two-mica granites). Secondly, these mechanisms depend not only
in supracrustal lithologies, typical of Wilson cycles, at the end of the on the volume of continental crust, but also on its ability to thicken
Archean (Veizer and McKenzie, 2003) and (v) the style of crustal defor- and to maintain thickening during a sufficiently long time for the
mation changed from the late Archean to the middle Proterozoic, with recycling mechanisms to start and develop. In other words, conti-
an increasing number of orogenic belts characterized by predominantly nental crust must be stiff enough for deformation to be focused
horizontal deformation and focused high-strain zones (Cagnard et al., along sub-linear orogenic belts. England and Bickle (1984) and Rey
2011; Chardon et al., 2009; Smithies et al., 2007). and Coltice (2008) proposed that due to higher Archean heat flux, the
In addition, Wilson cycles are characterized by a global balance continental crust was weaker and unable to sustain thicknesses N30
between crustal formation and crustal recycling into the mantle, such km, such that collision between crustal blocks did not result in the for-
that the net continental growth is presently close to zero (Cawood mation of collisional belts at plate boundaries. Instead, deformation of
et al., 2009; Clift and Vannucchi, 2004; Clift et al., 2009; Hawkesworth Archean continental crust was dominated by volume forces, coaxial
et al., 2010; Scholl and von Huene, 2009). Many studies pointed out shortening and lateral flow rather than sustainable vertical thickening
that the crustal growth rate significantly decreased between 3.0 and (Chardon et al., 2009; Choukroune et al., 1995; England and Bickle,
2.0 Ga ago, i.e. by the end of the Archean (Dhuime et al., 2011; 1984; Rey and Coltice, 2008; Thébaud and Rey, 2013). Specifically, Rey
Guitreau, 2012; Kramers and Tolstikhin, 1997; Taylor and McLennan, and Coltice (2008) proposed that the continental crust was too hot
1985, 1995), confirming that modern Wilson cycles were not operating and weak to sustain high topographic anomalies (N 2 km) until 2.5 Ga
efficiently prior to that. This observation is also supported by the ago (Fig. 15c). Therefore, collisional orogens would have become stable
present study (see Section 3.2), which shows that late-Archean gran- only from 3.0 to 2.5 Ga (Fig. 15c).
itoids derived largely from crustal recycling processes (intracrustal Importantly, the strength of the crust, and thus its capacity to sustain
melting; introduction of crustal-derived material and formation of high topography, has important implications for hypsometry. It has
metasomatized mantle; see also Figs. 10 and 11). Hence, the geodynamic been proposed that continental emersion was not significant until
changes at the Archean–Proterozoic transition also correspond to a the late-Archean (Flament et al., 2008; Rey and Coltice, 2008), resulting
major shift in the balance between growth and recycling rates of the in a major obstacle to the establishment of effective sedimentary
continental crust. cycles before ~ 3.0 Ga. The absence of thick sedimentary deposits
on the oceanic crust (coupled to less efficient subduction processes)
5.3. Possible origins of the changes implies that the recycling of sediments into the mantle remained
negligible until that time. The emergence of sanukitoids s.l.,
Because of the higher heat flux from the mantle, Archean oceanic some of which required a sediment-metasomatized mantle
crust was thicker, more buoyant and less coupled to the underlying con- source (Halla, 2005; King et al., 1998; Laurent et al., 2011, 2014;
vective mantle (e.g. Abbott et al., 1994; Hargraves, 1986; O'Neill et al., Lobach-Zhuchenko et al., 2008; Mikkola et al., 2011b), is therefore
2007a,b; van Thienen et al., 2004). This must have resulted in smaller- a good chronological marker of the onset of crustal recycling into
scale, less coherent and “weaker” Archean subduction, unable to drive the mantle.
plate tectonic processes and Wilson cycles efficiently as it presently To summarize, the establishment of “modern-style” subduction–col-
does (Moyen and Martin, 2012; O'Neill et al., 2007b; Sizova et al., lision mechanisms depended on several different parameters, namely
2010; Moyen and van Hunen, 2012; van Hunen and van den Berg, the thickness, temperature, composition and rheology of oceanic and
2008; see also Section 4.1.3). Abbott et al. (1994) proposed that, besides continental crusts as well as the net volume of continental material.
its composition, the thickness and temperature of oceanic crust exert a However, as noted by Moyen and Martin (2012), the single mechanism
first-order control on its capacity to subduct. They concluded that mod- that controls all these parameters is the evolution of the Earth's thermal
ern, steep subduction geometry could only be achieved with a relatively regime. This global control over a number of local parameters adequately
thin (b 10 km), cold crust (Fig. 15a). They calculated that most oceanic accounts for the ambiguous nature of the late-Archean geodynamic
crust reached the 10 km thickness threshold between ~ 2.5 and ~ 2.0 changes: they follow a global scheme (similar magmatic evolution in
(a) 50 6. Summary and concluding remarks

(Percentages are
cumulative)
This review shows that 3.0- to 2.5 Ga-old granitoids are very diverse,
Thickness of oceanic crust (km)
99% in relation to the TTG-dominated magmatic record of N3.0 Ga. These

40
late-Archean granitoids can be separated into four groups:
No subduction or
‘flat’ subduction (1) TTGs, i.e. volumetrically dominant low-K tonalites, trondhjemites
30
and granodiorites, which were generated from melting of hy-
drous, meta-igneous mafic rocks. Melting took place principally
95% at great depth (N35 km);
20
(2) Sanukitoids s.l., i.e. (Fe + Mg)- and K-rich magmas
50% (monzodiorites, monzogabbros and their differentiation
1% products) derived from interaction between mantle peridotite
10
and an incompatible element-rich component;
‘Classical’ (3) Biotite- and two-mica-bearing, peraluminous and high-K
subduction
granites derived from melting of pre-existing crustal lithologies,
100
(b) i.e. older TTGs and metasediments;
(4) hybrid granitoids with a wide range of petrographic and geo-
Guitreau (2012) chemical characteristics, resulting from interaction between
Volume of continental crust
(% relative to present-day)
75 magmas or source rocks from the first three groups.

Dhuime et al.
(2011) The emplacement of these different granitoid groups follows a very
Kramers and typical, two-stage sequence in all Archean cratons worldwide: (i) a
Tolstikhin (1997) long period (0.2 to 0.5 Ga) of exclusively TTG magmatism, followed by
50
(ii) a shorter stage (0.02–0.15 Ga) during which all the other granitoid
groups were emplaced. This peculiar evolution, coupled to available
Nd–Hf isotope data on late-Archean granitoids, shows that juvenile
25 crust generation through TTG magmatism gave way to crustal recycling
Taylor and McLennan at two distinct levels, i.e. by classic intracrustal differentiation, and
(1985, 1995)
introduction of crustal-derived material into the mantle as a source
component for sanukitoids s.l.
0 10 Both the nature of the granitoids related to the two stages and dura-
(c)
Max. elevation of sustainable relief (km)
tion of these stages are consistent with a subduction–collision evolu-

Strength in compression of the
tion. In particular, granitoids emplaced in modern collisional settings

continental crust (1012 N.m–1)
50
8 closely resemble those of the second stage, being characterized by
coeval involvement of metasomatized mantle sources, and intracrustal
40 melting. This type of evolution is scarce in the N3.0 Ga-old Archean
6 record, and takes place only locally, indicating that the global extent of
APT
30 late-Archean geodynamic changes likely heralds the initiation of sub-
duction–collision (Wilson) cycles on a planetary scale and thus of “mod-
4 ern-style” plate tectonic processes.
20
This late-Archean change in global geodynamics likely resulted from
the evolution of several parameters, such as the thickness, temperature
2 and rheology of oceanic and continental crusts, as well as an increase in
10
the global volume of continents. In summary, the onset of “modern-
style” plate tectonics resulted from the stabilization, by the end of the
0
Archean, of lithospheric plates that were large, cold and stiff enough
1.00 1.50 2.00 2.50 3.00 3.50 4.00
to undergo subduction and collision. This complex interplay of several
Age (Ga)
factors induced local differences from one craton to another, but the
overall evolution depended on a single, unifying mechanism: the pro-
Fig. 15. Evolution of some key properties of the oceanic and continental crusts between 4.0
and 1.0 Ga, being possible causes for the onset of “modern-style” plate tectonics during the
gressive cooling of the Earth. As a result of this global control, the late-
Archean–Proterozoic Transition (APT): (a) progressive thinning of the oceanic crust Archean geodynamic changes took place within a relatively short period
through time, according to thermal models (modified after Abbott et al., 1994). Between (~0.5 Ga) and affected the whole planetary system.
2.0 and 3.0 Ga, ~50% of oceanic crust became thinner than 10 km, the “threshold” below
which “classic” subduction is stable; (b) several models of crustal growth, all predicting
Acknowledgements
that at the end of the Archean, ~70% of the present-day volume of continental crust was
already formed. This highlights the switch from dominantly juvenile to dominantly
recycling processes at that time; (c) stiffening of the continental crust through time, as a This work was principally funded by two scientific co-operations be-
result of its cooling. The compressive strength of the crust increased significantly (by a fac- tween France and South Africa, the PICS n°4828 as well as the !Khure
tor of 2) during the last 0.5 Ga of the Archean, so that at the end of this period it was able to programme, and was also supported by the European Research Council
sustain positive topographic anomalies of N2 km, i.e., collisional orogens (modified after
(ERC StG 279828). This paper was originally requested on the basis of an
Rey and Coltice, 2008). See text for discussion.
abstract presented at the EGU General Assembly 2012, and O.L.'s partic-
ipation to this conference was possible thanks to funding by Région
Auvergne. We are grateful to N. Eby and T. Horscroft for their invitation,
their kind editorial assistance and their patience throughout the prepa-
every craton, within a relatively short time span on a planetary scale) ration of the manuscript. We also appreciated the detailed and con-
but with local characteristics (different ages from one craton to another; structive reviews by R. Dall'Agnol and an anonymous referee that
diversity in magma sources and relative chronology). contributed to significantly improve the original manuscript. We
would like to thank many colleagues for fruitful discussion regarding Bédard, J.H., Harris, L.B., Thurston, P., 2013. The hunting of the snArc. Precambrian
Research 229, 20–48.
the ideas presented in this paper, in particular J.H. Bédard, S. Block, B. Belousova, E.A., Kostitsyn, Y.A., Griffin, W.L., Begg, G.C., O'Reilly, S.Y., Pearson, N.J., 2010.
Bonin, L.S. Doucet, J. Halla, E. Heilimo, G. Stevens, A. Villaros and A. The growth of the continental crust: constraints from zircon Hf-isotope data. Lithos
Zeh. We also thank F. van Wyk de Vries for her corrections of the English 119 (3–4), 457–466.
Bibikova, E.V., Petrova, A., Claesson, S., 2005. The temporal evolution of sanukitoids in the
language. This research was financed by the French Government Karelian Craton, Baltic Shield: an ion microprobe U–Th–Pb isotopic study of zircons.
Laboratory of Excellence initiative n°ANR-10-LABX-0006 and the Lithos 79, 129–145.
European Regional Development Fund. This is Laboratory of Excellence Bibikova, E.V., Lobach-Zhuchenko, S.B., Artemenko, G.B., Claesson, S., Kovalenko, A.V., Krylov,
I.N., 2008. Late Archean magmatic complexes of the Azov Terrane, Ukrainian
ClerVolc contribution number 108.
Shield: geological setting, isotopic age, and sources of material. Petrology 16
(3), 211–231.
Appendix A. Supplementary data Bickle, M.J., Bettenay, L.F., Chapman, H.J., Groves, D.I., McNaughton, N.J., Campbell, I.H., de
Laeter, J.R., 1989. The age and origin of younger granitic plutons of the Shaw Batholith
in the Archaean Pilbara Block, Western Australia. Contributions to Mineralogy and
Supplementary data to this article can be found online at http://dx. Petrology 101, 361–376.
doi.org/10.1016/j.lithos.2014.06.012. Bleeker, W., 2003. The late Archean record: a puzzle in ca. 35 pieces. Lithos 71, 99–134.
Blichert-Toft, J., Puchtel, I.S., 2010. Depleted mantle sources through time: evidence from
Lu–Hf and Sm–Nd isotope systematics of Archean komatiites. Earth and Planetary
References Science Letters 297, 598–606.
Block, S., Moyen, J.-F., Zeh, A., Poujol, M., Jaguin, J., Paquette, J.-L., 2013. The Murchison
Abbott, D., Drury, R., Smith, H.F.S., 1994. Flat to steep transition in subduction style. Greenstone Belt. South Africa: accreted slivers with contrasting metamorphic condi-
Geology 22, 937–940. tions. Precambrian Research 227, 77–98.
Agranier, A., 2001. Un exemple de magmatisme ultrapotassique syn-orogénique: les Bonin, B., 2004. Do coeval mafic and felsic magmas in post-collisional to within-plate
lamprophyres du Massif Central Français. Approche géochimique et géochronologique. regimes necessarily imply two contrasting, mantle and crustal, sources? A review.
Unpublished M.Sc. thesis, Université Blaise Pascal, Clermont-Ferrand II, 51 pp. Lithos 78, 1–24.
Aikman, A.B., Harrison, T.M., Hermann, J., 2012. The origin of Eo- and Neo-himalayan Bouilhol, P., Jagoutz, O., Hachar, J.M., Dudas, F.O., 2013. Dating the India-Eurasia collision
granitoids, Eastern Tibet. Journal of Asian Earth Sciences 58, 143–157. through arc magmatic records. Earth and Planetary Science Letters 366, 163–175.
Aït-Malek, H., 1997. Pétrologie, géochimie et géochronologie U/Pb d'associations acide- Bowring, S.A., Williams, I.S., 1999. Priscoan (4.00–4.03 Ga) orthogneisses from northwest-
basiques: exemples du SE du Velay (Massif central français) et de l'anti-Atlas occi- ern Canada. Contributions to Mineralogy and Petrology 134, 3–16.
dental (Maroc). Unpublished Ph.D. thesis, Université de Nancy, 297 pp. Brandl, G., Kröner, A., 1993. Preliminary results of single zircon studies from various Ar-
Almeida, J.A.C., Dall'Agnol, R., Dias, S.B., Althoff, F.J., 2010. Origin of the Archean chaean rocks of the northeastern Transvaal. Ext. Abstr. 16th International Colloquium
leucogranodiorite–granite suites: evidence from the Rio Maria terrane and implica- of African Geology, Mbabane, Swaziland, pp. 54–56.
tions for granite magmatism in the Archean. Lithos 120, 235–257. Brandl, G., Jaeckel, P., Kröner, A., 1996. Single zircon age for the felsic Rubbervale Forma-
Almeida, J.A.C., Dall'Agnol, R., Leite, A.A.S., 2013. Geochemistry and zircon geochronology tion, Murchison greenstone belt, South Africa. South African Journal of Geology 99
of the Archean granite suites of the Rio Maria granite–greenstone terrane, Carajás (3), 229–234.
Province, Brazil. Journal of South American Earth Sciences 42, 103–126. Breaks, F.W., Moore, J.M., 1992. The Ghost Lake batholith, Superior Province of North-
Almeida, J.A.C., Dall'Agnol, R., Oliveira, M.A., Macambira, M.J.B., Pimentel, M.M., Rämö, O.T. western Ontario: a ferile, S-type, peraluminous granite — rare-element pegmatite
, Guimarães, F.V., Leite, A.A.S., 2011. Zircon geochronology, geochemistry and origin system. Canadian Mineralogist 30, 835–875.
of the TTG suites of the Rio Maria granite–greenstone terrane: implications for the Brown, M., 2006. A duality of thermal regimes is the distinctive characteristic of plate tec-
growth of the Archean crust of the Carajás Province, Brazil. Precambrian Research tonics since the Neoarchean. Geology 34, 961–964.
187, 201–221. Cagnard, F., Barbey, P., Gapais, D., 2011. Transition between “Archaean-type” and “mod-
Althoff, F.J., 1996. Étude pétrologique et structurale des granitoïdes de Marajoara (Pará, ern-type” tectonics: insights from the Finnish Lapland Granulite Belt. Precambrian
Brésil): leur rôle dans l'évolution archéenne du craton Amazonien (2.7–3.2 Ga). Research 187, 127–142.
Unpublished Ph.D thesis. Université Henri Poincaré, Nancy I, France, 296 pp. Cassidy, K.F., Barley, M.E., Groves, D.I., Perring, C.S., Hallberg, J.A., 1991. An overview of the
Althoff, F.J., Barbey, P., Boullier, A.M., 2000. 2.8–3.0 Ga plutonism and deformation in the nature, distribution and inferred tectonic setting of granitoids of the late-Archaean
SE Amazonian craton: the Archean granitoids of Marajoara (Carajás Mineral province, Norseman–Wiluna Belt. Precambrian Research 51, 51–83.
Brazil). Precambrian Research 104, 187–206. Cawood, P., Kröner, A., Collins, W.J., Kusky, T.M., Mooney, W.D., Windley, B.F., 2009. Accre-
Armstrong, R.L., 1981. Radiogenic isotopes: the case for crustal recycling on a near- tionary orogens through Earth history. Geological Society of London Special Publica-
steady-state no-continental-growth Earth. Philosophical Transactions of the Royal tions 318, 1–36.
Society of London A 301, 443–472. Champion, D.C., Sheraton, J.W., 1997. Geochemistry and Nd isotope systematics of Ar-
Atherton, M.P., Petford, N., 1993. Generation of sodium-rich magmas from newly chaean granites of the Eastern Goldfields, Yilgarn Craton, Australia: implications for
underplated basaltic crust. Nature 362, 144–146. crustal growth processes. Precambrian Research 83 (1–3), 109–132.
Bagai, Z., Armstrong, R., Kampunzu, A.B., 2002. U– Pb single zircon geochronology of gran- Champion, D.C., Smithies, R.H., 1999. Archaean granites of the Yilgarn and Pilbara cratons,
itoids in the Vumba granite – greenstone terrain (NE Botswana): implication for the Western Australia: secular changes. In: Barbarin, B. (Ed.), The Origin of Granites and
Archaean Zimbabwe craton. Precambrian Research 118, 149–168. Related Rocks—IVth Hutton Symposium Abstracts Doc. BRGM, 290, p. 137.
Barbey, P., Martin, H., 1987. The role of komatiites in plate tectonics: evidence from Champion, D.C., Smithies, R.H., 2003. Archaean granites. In: Blevin, P.L., Chappell, B.W.,
the Archaean and early Proterozoic crust in the eastern Baltic shield. Precambrian Jones, M. (Eds.), Magmas to Mineralisation: the Ishihara Symposium. AGSOGeoscience
Research 35, 1–14. Australia, pp. 19–24 (Record 2003/14).
Barbey, P., Marignac, C., Montel, J.M., Macaudière, J., Gasquet, D., Jabbori, J., 1999. Cordierite Champion, D.C., Smithies, R.H., 2007. Geochemistry of Paleoarchean granites of the East
growth textures and the conditions of genesis and emplacement of crustal granitic Pilbara Terrane, Pilbara Craton, Western Australia: implications for early Archean crustal
magmas: the Velay granite complex (Massif Central, France). Journal of Petrology 40 growth. In: van Kranendonk, M.J., Smithies, R.H., Bennett, V.C. (Eds.), Earth's Oldest
(9), 1425–1441. Rocks, Developments in Precambrian Geology, vol. 15. Elsevier, Amsterdam, pp.
Barley, M.E., Pickard, A.L., 1999. An extensive, crustally-derived, 3325 to 3310 Ma silicic 369–410.
volcanoplutonic suite in the eastern Pilbara Craton: evidence from the Kelly Belt, Chardon, D., Gapais, D., Cagnard, F., 2009. Flow of ultra-hot orogens: a view from Precam-
McPhee Dome and Corunna Downs Batholith. Precambrian Research 96, 41–62. brian, clues for the Phanerozoic. Tectonophysics 477 (3–4), 105–118.
Barker, F., Arth, J.G., 1976. Generation of trondhjemitic–tonalitic liquids and Archaean Chung, S.-L., Liu, D., Ji, J., Chu, M.-F., Lee, H.-Y., Wen, D.-J., Lo, C.-H., Lee, T.-Y., Qian, Q.,
bimodal trondhjémite-basalt suites. Geology 4 (10), 596–600. Zhang, Q., 2003. Adakite from continental collision zones: melting of thickened
Barton Jr., J.M., Klemd, R., Zeh, A., 2006. The Limpopo Belt: a result of Archean to Protero- lower crust beneath southern Tibet. Geology 31, 1021–1024.
zoic, Turkic-type orogenesis? Geological Society of America, Special Paper 405, Choukroune, P., Bouhallier, H., Arndt, N.T., 1995. Soft lithosphere during periods of Arche-
315–332. an crustal growth or crustal reworking. In: Coward, M.P., Ries, A.C. (Eds.), Early
Beard, J.S., Lofgren, G.E., 1991. Dehydration melting and water-saturated melting of basaltic Precambrian Processes. Geological Society Special Publications, 95, pp. 67–86.
and andesitic greenstones and amphibolites at 1, 3 and 6.9 kbar. Journal of Petrology 32 Clemens, J.D., Belcher, R.W., Kisters, A.F.M., 2010. The Heerenveen batholith, Barber-
(2), 365–401. ton Mountain Land, South Africa: Mesoarchaean, potassic, felsic magmas formed
Bédard, L.P., 1996. Archaean high-Mg monzodiorite plutonic suite: a reevaluation of the by melting of an ancient subduction complex. Journal of Petrology 51 (5),
parental magma and differentiation. The Journal of Geology 104, 713–728. 1099–1120.
Bédard, J.H., 2006. A catalytic delamination-driven model for coupled genesis of Archaean Clemens, J.D., Darbyshire, D.P.F., Flinders, J., 2009. Sources of post-orogenic calcalkaline
crust and sub-continental lithospheric mantle. Geochimica Cosmochimica Acta 70, magmas: the Arrochar and Garabal Hill–Glen Fyne complexes, Scotland. Lithos 112,
1188–1214. 524–542.
Bédard, J.H., 2013. How many arcs can dance on the head of a plume? A ‘Comment’ on: a Clift, P., Vannucchi, P., 2004. Controls on tectonic accretion versus erosion in subduction
critical assessment of Neoarchaean ‘plume only’ geodynamics: evidence from the zones: implications for the origin and recycling of the continental crust. Reviews of
Superior Province, by Derek Wyman, Precambrian Research, 2012. Precambrian Geophysics 42, 1–31.
Research 229, 189–197. Clift, P., Schouten, H., Vannucchi, P., 2009. Arc–continent collisions, sediment recycling
Bédard, J.H., Brouillette, P., Madore, L., Berclaz, A., 2003. Archaean cratonization and defor- and the maintenance of the continental crust. In: Cawood, P.A., Kröner, A. (Eds.),
mation in the northern Superior Province, Canada: an evaluation of plate tectonic Accretionary Orogens in Space and Time. Geological Society of London, Special Publica-
versus vertical tectonic models. Precambrian Research 127, 61–87. tions, 318, pp. 75–103.
Condie, K.C., 1981. Archaean greenstone belts. In: Windley, B.F. (Ed.), Developments in Foley, S.F., Tiepolo, M., Vannucci, R., 2002. Growth of early continental crust controlled by
Precambrian Geology, 3 (434 pp.). melting of amphibolite in subduction zones. Nature 417, 637–640.
Condie, K.C., 1989. Geochemical changes in basalts and andesites across the Archean– Friend, C.R.L., Nutman, A.P., 2005. Complex 3670–3500 Ma orogenic episodes
Proterozoic boundary: identification and significance. Lithos 23 (1–2), 1–18. superimposed on juvenile crust accreted between 3850 and 3690 Ma, Itsaq Gneiss
Condie, K.C., O'Neill, C., 2010. The Archean–Proterozoic boundary: 500 my of tectonic Complex, Southern West Greenland. The Journal of Geology 113 (4), 375–397.
transition in Earth history. American Journal of Science 310, 775–790. Frost, B.R., Barnes, C.G., Collins, W.J., Arculus, R.J., Ellis, D.J., Frost, C.D., 2001. A geochemical
Condie, K.C., O'Neill, C., Aster, R.C., 2009. Evidence and implications for a widespread mag- classification for granitic rocks. Journal of Petrology 42, 2033–2048.
matic shutdown for 250 My on Earth. Earth and Planetary Science Letters 282 (1–4), Frost, C.D., Frost, B.R., Chamberlain, K.R., Huselbosch, T.P., 1998. The Late Archaean history
294–298. of the Wyoming province as recorded by granitic magmatism in the Wind River
Corfu, F., Krogh, T.E., Kwok, Y.Y., Marmont, S., Jensen, L.S., 1989. U–Pb zircon geochronol- Range, Wyoming. Precambrian Research 98, 145–173.
ogy in the south-western Abitibi greenstone belt, Superior Province. Canadian Journal Frost, C.D., Frost, B.R., Kirkwood, R., Chamberlain, K.R., 2006. The tonalite–trondhjemite–
of Earth Sciences 26, 1747–1763. granodiorite (TTG) to granodiorite–granite (GG) transition in the late Archean
Couturié, J.P., 1977. Le massif granitique de la Margeride (Massif Central Français). plutonic rocks of the central Wyoming Province. Canadian Journal of Earth Sciences
Annales Scientifiques de l'Université de Clermont-Ferrand 62, 319. 43, 1419–1444.
Couzinié, S., Moyen, J.F., Villaros, A., Paquette, J.L., Marignac, C., 2014. Temporal relation- Galán, G., Corretgé, L.G., Laurent, O., 1997. Low-potassium vaugnerites from Guéret (Mas-
ship between Mg–K magmatism and catastrophic melting of the crust during late- sif Central, France). Mafic magma evolution influenced by contemporaneous granit-
collisional period: LA-ICP-MS dating of zircons from granitoids and coeval mafic oids. Mineralogy and Petrology 59, 165–187.
rocks in the southeastern Velay complex, French Massif Central. Journal of Gao, Y., Yang, Z., Hou, Z., Wei, R., Meng, X., Tian, S., 2010. Eocene potassic and ultrapotassic
Geosciences 59 (1), 69–86. volcanism in south Tibet: new constraints on mantle source characteristics and
Dall'Agnol, R., Oliveira, M.A., Almeida, J.A.C., Althoff, F.J., Leite, A.A.S., Oliveira, D.C., Barros, geodynamic processes. Lithos 117 (1–4), 20–32.
C.E.M., 2006. Archean and Paleoproterozoic granitoids of the Carajás metallogenetic Garde, A.A., 1997. Accretion and evolution of an Archaean high-grade grey gneiss-am-
province, eastern Amazonian craton. In: Dall'Agnol, R., Rosa-Costa, L.T., Klein, E.L. phibolite complex: the Fiskefjord area, southern West Greenland. Geology Greenland
(Eds.), Symposium on Magmatism.: Crustal Evolution, and Metallogenesis of the Am- Survey Bulletin (114 pp.)..
azonian Craton. Abstracts Volume and Field Trips Guide. Belém, PRONEX-UFPA/ Gardien, V., Vanderaeghe, O., Arnaud, N., Cocherie, A., Grange, M., Lécuyer, C., 2011.
SBGNO. PRONEX-UFPA/SBGNO, Belém, pp. 99–150. Thermal maturation and exhumation of a middle orogenic crust in the Livradois
Davis, W.J., Hegner, E., 1992. Neodymium isotopic evidence for the accretionary develop- area (French Massif Central). Bulletin de la Societe Geologique de France 182 (1),
ment of the Late Archean Slave Province. Contributions to Mineralogy and Petrology 5–24.
111, 493–503. Guitreau, M., 2012. Les isotopes de l'hafnium dans les TTG et leurs zircons: témoins de la
Davis, W.J., Bleeker, W., 1999. Timing of plutonism, deformation and metamorphism in croissance des premiers continents. Unpublished Ph.D. thesis, École Normale
the Yellowknife domain, Slave Province, Canada. Canadian Journal of Earth Sciences Supérieure de Lyon, 275 pp.
36, 1169–1187. Guitreau, M., Blichert-Toft, J., Martin, H., Mojzsis, S., Albarède, F., 2012. Hafnium isotope
Debon, F., Lemmet, M., 1999. Evolution of Fe/Mg ratios in Late Variscan plutonic rocks evidence from Archean granitic rocks for deep-mantle origin of continental crust.
from the External Crystalline Massif of the Alps (France, Italy, Switzerland). Journal Earth and Planetary Science Letters 337–338, 211–223.
of Petrology 40 (7), 1151–1185. Guo, Z., Wilson, M., 2012. The Himalayan leucogranites: constraints on the nature
Defant, M.J., Drummond, M.S., 1990. A model for trondhjemite–tonalite–dacite genesis of their crustal source region and geodynamic setting. Gondwana Research 22
and crustal growth via slab melting: Archaean to modern comparisons. Journal of (2), 360–376.
Geophysical Research 95 (B-13), 21503–21521. Halla, J., 2005. Late Archean high-Mg granitoids (sanukitoids) in the southern Karelian
de Wit, M.J., Hart, R.A., 1993. Earth's earliest continental lithosphere, hydrothermal flux domain, eastern Finland: Pb and Nd isotopic constraints on crust–mantle interac-
and crustal recycling. Lithos 30 (3–4), 309–335. tions. Lithos 79 (1–2), 161–178.
de Wit, M.J., Armstrong, R.A., Kamo, S.L., Erlank, A.J., 1993. Gold-bearing sediments in the Halla, J., van Hunen, J., Heilimo, E., Hölttä, P., 2009. Geochemical and numerical constraints
Pietersburg greenstone belt: age equivalents of the Witwatersrand Supergroup sedi- on Neoarchaean plate tectonics. Precambrian Research 174, 155–162.
ments, South Africa. Economic Geology 88, 1242–1252. Hargraves, R.B., 1986. Faster spreading or greater ridge length in the Archean? Geology 14
de Wit, M.J., Roering, C., Hart, R.J., Armstrong, R.A., de Ronde, R.E.J., Green, R.W., (9), 750–752.
Tredoux, M., et al., 1992. Formation of an Archaean continent. Nature 357, Harris, N.B.W., Pearce, J.A., Tindle, A.G., 1986. Geochemical characteristics of collision-
553–562. zone magmatism. Geological Society of London Special Publications 19, 67–81.
Dey, S., 2013. Evolution of Archaean crust in the Dharwar craton: the Nd isotope record. Harrison, T.M., Grove, M., McKeegan, K.D., Coath, C.D., Lovera, O.M., Le Fort, P., 1999.
Precambrian Research 227, 227–246. Origin and episodic emplacement of the Manaslu intrusive complex, central
Dey, S., Nandy, J., Choudhary, A.K., Liu, Y., Zong, K., 2014. Origin and evolution of granit- Himalaya. Journal of Petrology 40, 3–19.
oids associated with the Kadiri greenstone belt, eastern Dharwar craton: a history Harrison, T.M., Lovera, O.M., Grove, M., 1997. New insights into the origin of two contrast-
of orogenic to anorogenic magmatism. Precambrian Research 246, 64–90. ing Himalayan granite belts. Geology 25, 899–902.
Dey, S., Pandey, U.K., Rai, A.K., Chaki, A., 2012. Geochemical and Nd isotope constraints on Hawkesworth, C.J., Dhuime, B., Pietranik, A.B., Cawood, P.A., Kemp, A.I.S., Storey, C.D.,
petrogenesis of granitoids from NW part of the eastern Dharwar craton: possible impli- 2010. The generation and evolution of the continental crust. Journal of the Geological
cations for late Archaean crustal accretion. Journal of Asian Earth Sciences 45, 40–56. Society of London 167, 229–248.
Dhuime, B., Hawkesworth, C.J., Cawood, P.A., Storey, C.D., 2011. A change in the Heilimo, E., Halla, J., Hölttä, P., 2010. Discrimination and origin of the sanukitoid series:
geodynamics of continental growth 3 billion years ago. Science 335, 1334–1336. geochemical constraints from the Neoarchean western Karelian Province (Finland).
Didier, J., Barbarin, B., Gagny, C., Leistel, J.M., Kerrien, Y., 1989. Notice explicative, Carte Lithos 115, 27–39.
géologique de la France (1/50000e), feuille Noiretable. Bureau de recherches Heilimo, E., Halla, J., Huhma, H., 2011. Single-grain zircon U–Pb age constraints of the
géologiques et minières, Orléans. western and eastern sanukitoid zones in the Finnish part of the Karelian Province.
Downes, H., Shaw, A., Williamson, B.J., Thirlwall, M.F., 1997. Sr, Nd and Pb isotope geo- Lithos 121, 87–99.
chemistry of the Hercynian granodiorites and monzogranites, Massif Central, Henderson, D.R., Long, L.E., Barton Jr., J.M., 2000. Isotopic ages and chemical and isotopic
France. Chemical Geology 136, 99–122. composition of the Archaean Turfloop Batholith, Pietersburg granite–greenstone ter-
Duretz, T., Gerya, T.V., 2013. Slab detachment during continental collision: Influence of crustal rane, Kaapvaal Craton, South Africa. South African Journal of Geology 103 (1), 38–46.
rheology and interactions with lithospheric delamination. Tectonophysics 602, 124–140. Hill, R.I., Chappell, B.W., Campbell, I.H., 1992. Late Archean granites of the southeastern
Eglington, B.M., Armstrong, R.A., 2004. The Kaapvaal Craton and adjacent orogens, south- Yilgarn Block, Western Australia: age, geochemistry, and origin. Transactions of the
ern Africa: a geochronological database and overview of the geological development Royal Society of Edinburgh: Earth Sciences 83 (1–2), 211–226.
of the craton. South African Journal of Geology 107, 13–32. Iizuka, T., Horie, K., Komiya, T., Maruyama, S., Hirata, T., Hidaka, H., Windley, B.F., 2006. 4.
England, P., Bickle, M., 1984. Continental thermal and tectonic regimes during the Archae- 2 Ga zircon xenocryst in an Acasta gneiss from northwestern Canada: evidence for
an. The Journal of Geology 92 (4), 353–367. early continental crust. Geology 34 (4), 245–248.
Ellam, R.M., Hawkesworth, C.J., 1988. Is average continental crust generated at subduction Jaguin, J., Gapais, D., Poujol, M., Boulvais, P., Moyen, J.-F., 2012. the Murchison Greenstone
zones? Geology 16, 314–317. Belt (South Africa): a General Tectonic Framework. South African Journal of Geology
Eriksson, P.G., Banerjee, S., Nelson, D.R., Rigby, M.J., Catuneanu, O., Sarkar, S., Roberts, R.J., 115 (1), 65–76.
et al., 2009. A Kaapvaal craton debate: nucleus of an early small supercontinent or Jahn, B.M., Glikson, A.Y., Peucat, J.-J., Hickman, A.H., 1981. REE geochemistry and isotopic
affected by an enhanced accretion event? Gondwana Research 15 (3–4), 354–372. data of Archaean silicic volcanics and granitoids from the Pilbara Block, western
Feio, G.R.L., Dall'Agnol, R., Dantas, E.L., Macambira, M.J.B., Gomes, A.C.B., Sardinha, A.S., Australia: implications for the early crustal evolution. Geochimica et Cosmochimica
Oliveira, D.C., Santos, R.D., Santos, P.A., 2012. Geochemistry, geochronology, and Acta 45, 1633–1652.
origin of the Neoarchean Planalto Granite suite, Carajás, Amazonian craton: A-type Jahn, B.M., Auvray, B., Shen, Q.H., Liu, D.Y., Zhang, Z.Q., Dong, Y.J., Ye, X.J., Zhang, Q.Z.,
or hydrated charnockitic granites? Lithos 151, 57–73. Cornichet, J., Macé, J., 1988. Archaean crustal evolution in China: the Taishan complex,
Feng, R., Kerrich, R., 1992. Geochemical evolution of granitoids from the Archean Abitibi and evidence for juvenile crustal addition from long-term depleted mantle. Precam-
Southern Volcanic Zone and the Pontiac subprovince, Superior Province, Canada: brian Research 38, 381–403.
implications for tectonic history and source regions. Chemical Geology 98, 23–70. Jayananda, M., Moyen, J.-F., Martin, H., Peucat, J.-J., Auvray, B., Mahabaleswar, B., 2000.
Ferré, E.C., Caby, R., Peucat, J.J., Capdevila, R., Monié, P., 1998. Pan-African, post-collisional, Late Archaean (2550–2520 Ma) juvenile magmatism in the Eastern Dharwar craton,
ferro-potassic granite and quartz–monzonite plutons of Eastern Nigeria. Lithos 45, southern India: constraints from geochronology, Nd–Sr isotopes and whole rock geo-
255–279. chemistry. Precambrian Research 99, 225–254.
Flament, N., Coltice, N., Rey, P.F., 2008. A case for late-Archaean continental emergence Jayananda, M., Chardon, D., Peucat, J.-J., Capdevila, R., 2006. 2.61 Ga potassic granites and
from thermal evolution models and hypsometry. Earth and Planetary Science Letters crustal reworking in the western Dharwar craton, southern India: Tectonic, geochro-
275, 326–336. nologic and geochemical constraints. Precambrian Research 150 (1–2), 1–26.
Jayananda, M., Martin, H., Peucat, J.-J., Mahabaleswar, B., 1995. Late Archaean crust– Laurent, O., Paquette, J.L., Martin, H., Doucelance, R., Moyen, J.F., 2013b. LA-ICP-MS dating
mantle interactions: geochemistry of LREE-enriched mantle derived magmas. of zircons from Meso- and Neoarchean granitoids of the Pietersburg block (South
Example of the Closepet batholith, Southern India. Contributions to Mineralogy Africa): Crustal evolution at the northern margin of the Kaapvaal craton. Precambrian
and Petrology 199, 314–329. Research 230, 209–226.
Jelsma, H.A., Vinyu, M.L., Valbracht, P.J., Davies, G.R., Wijbrans, J.R., Verdurmen, E.A.T., Laurent, O., Rapopo, M., Stevens, G., Moyen, J.F., Martin, H., Doucelance, R., Bosq, C., 2014.
1996. Constraints on Archaean crustal evolution of the Zimbabwe craton: a U–Pb zir- Contrasting petrogenesis of Mg–K and Fe–K granitoids and implications for post-
con, Sm–Nd and Pb–Pb whole-rock isotope study. Contributions to Mineralogy and collisional magmatism: a case study from the late-Archean Matok pluton (Pietersburg
Petrology 124, 55–70. block, South Africa). Lithos 196–197, 131–149.
Johnson, T., Brown, M., Kaus, B.J.P., van Tongeren, J.A., 2013. Delamination and Ledru, P., Courrioux, G., Dallain, C., Lardeaux, J.M., Montel, J.M., Vanderhaeaghe, O., Vitel,
recycling of Archaean crust caused by gravitational instabilities. Nature Geoscience G., 2001. The Velay dome (French Massif Central): melt generation and granite
7, 47–52. emplacement during orogenic evolution. Tectonophysics 342 (3–4), 207–237.
Kamber, B.S., Ewart, A., Collerson, K.D., Bruce, M.C., McDonald, G.D., 2002. Fluid-mobile Le Fort, P., Cuney, M., Deniel, C., France-Lanord, C., Sheppard, S.M.F., Upreti, B.N., Vidal, P.,
trace element constraints on the role of slab melting and implications for 1987. Crustal generation of the Himalayan leucogranites. Tectonophysics 134, 39–57.
Archaean crustal growth models. Contributions to Mineralogy and Petrology 144, Leite, A.A.S., Dall'Agnol, R., Macambira, M.J.B., Althoff, F.J., 2004. Geologia e geocronologia
38–56. dos granitóides arqueanos da região de Xinguara e suas implicações na evolução do
Kamo, S.L., Davis, D.W., 1994. Reassessment of Archean crustal development in the Terreno Granito–Greenstone de Rio Maria, Cráton Amazônico. Revista Brasileira de
Barberton Mountain Land, South Africa, based on U–Pb dating. Tectonics 13 (1), Geociencias 34, 447–458.
167–192. Liégeois, J.P., Navez, J., Hertogen, J., Black, R., 1998. Contrasting origin of post-collisional
Kampunzu, A.B., Tombale, A.R., Zhai, M., Bagai, Z., Majaule, T., Modisi, M.P., 2003. Major high-K calc-alkaline and shoshonitic versus alkaline and peralkaline granitoids. The
and trace element geochemistry of plutonic rocks from Francistown, NE Botswana: use of sliding normalization. Lithos 45, 1–28.
evidence for a Neoarchaean continental active margin in the Zimbabwe craton. Lithos Liu, S.W., Pan, Y.M., Xie, Q.L., Zhang, J., Li, Q., 2004. Archean geodynamics in the Central
71, 431–460. Zone, North China Craton: constraints from geochemistry of two constrasting series
Käpyaho, A., Mänttäri, I., Huhma, H., 2006. Growth of Archaean crust in the Kuhmo of granitoids in the Fuping and Wutai complexes. Precambrian Research 130,
district, Eastern Finland: U–Pb and Sm–Nd isotope constraints on plutonic rocks. 229–249.
Precambrian Research 146, 95–119. Lobach-Zhuchenko, S.B., Rollinson, H.R., Chekulaev, V.P., Arestova, N.A., Kovalenko, A.V.,
Keller, C.B., Schoene, B., 2012. Statistical geochemistry reveals disruption in secular Ivanikov, V.V., Guseva, N.S., Sergeev, S.A., Matukov, D.I., Jarvis, K.E., 2005. The Archaean
lithospheric evolution about 2.5 Gyr ago. Nature 485, 490–495. sanukitoid series of the Baltic Shield: geological setting, geochemical characteristics
Kemp, A.I.S., Hawkesworth, C.J., 2003. Granitic perspectives on the generation and secular and implications for their origin. Lithos 79, 107–128.
evolution of the continental crust. In: Rudnick, R.L. (Ed.), The Crust. Treatise on Geo- Lobach-Zhuchenko, S.B., Rollinson, H., Chekulaev, V.P., Savatenkov, V.M., Kovalenko, A.V.,
chemistry. Elsevier-Pergamon, Oxford, pp. 349–410. Martin, H., Guseva, N.S., Arestova, N.A., 2008. Petrology of a Late Archaean, Highly
Kemp, A.I.S., Hawkesworth, C.J., Collins, W.J., Gray, C.M., Blevin, P.L., Edinburgh Ion Potassic, Sanukitoid Pluton from the Baltic Shield: insights into Late Archaean Mantle
Microprobe Facility, 2009. Nd, Hf and O isotope evidence for rapid continental growth Metasomatism. Journal of Petrology 49 (3), 393–420.
during accretionary orogenesis in the Tasmanides, eastern Australia. Earth and McDonough, W.F., Sun, S.S., 1995. The composition of the Earth. Chemical Geology 120
Planetary Science Letters 284, 455–466. (3–4), 223–253.
Kinny, P.D., 2000. U–Pb dating of rare metal (Sn–Ta–Li) mineralized pegmatites in Mahéo, G., Blichert-Toft, J., Pin, C., Guillot, S., Pêcher, A., 2009. Partial melting of mantle
Western Australia by SIMS analysis of Tin and Tantalum-bearing ore minerals. Ab- and crustal sources beneath South Karakorum, Pakistan: implications for the Miocene
stracts and Proceedings, Beyond 2000. New Frontiers in Isotope Geoscience, Lorne, geodynamic evolution of the India–Asia Convergence Zone. Journal of Petrology 20
Victoria, pp. 113–116. (3), 427–449.
King, E.M., Valley, J.M., Davis, D.W., Edwards, G.R., 1998. Oxygen isotope ratios of Archean Martin, H., 1986. Effect of steeper Archaean geothermal gradient on geochemistry of
plutonic rocks from granite-greenstone belts of the Superior Province: indicator of subduction-zone magmas. Geology 14, 753–756.
magmatic source. Precambrian Research 92, 365–387. Martin, H., 1987. Petrogenesis of Archaean trondhjemites, tonalites and granodiorites
Kleinhanns, I.C., Kramers, J.D., Kamber, B.S., 2003. Importance of water for Archaean gran- from Eastern Finland: major and trace element geochemistry. Journal of Petrology
itoid petrology: a comparative study of TTG and potassic granitoids from Barberton 28, 921–953.
Mountain Land, South Africa. Contributions to Mineralogy and Petrology 145, Martin, H., 1993. The mechanisms of petrogenesis of the Archaean continental crust,
377–389. comparison with modern processes. Lithos 30, 373–388.
Kovalenko, A., Clemens, J.D., Savantenkov, V., 2005. Petrogenetic constraints for the Martin, H., 1994. The Archaean grey gneisses and the genesis of continental crust. In:
genesis of Archaean sanukitoid suites: geochemistry and isotopic evidence from Condie, K.C. (Ed.), Archaean Crustal Evolution. Developments in Precambrian
Karelia, Batlic Shield. Lithos 79, 147–160. Geology, 11, pp. 205–260.
Kramers, J.D., Houri, H., 2011. The geochronology of the Limpopo Complex: a controversy Martin, H., 1999. Adakite magmas: modern analogs of Archaean granitoids. Lithos 46,
solved. In: van Reenen, D.D., Kramers, J.D., McCourt, S., Perchuk, L.L. (Eds.), Origin and 411–429.
Evolution of Precambrian High-Grade Gneiss Terranes, with Special Emphasis on the Martin, H., Querré, G., 1984. A 2.5 G.a. reworked sialic crust: Rb–Sr ages and isotopic
Limpopo Complex of Southern Africa. Geological Society of America, Memoir, 207, geochemistry of late archaean volcanic and plutonic rocks from E. Finland. Contribu-
pp. 85–106. tions to Mineralogy and Petrology 85, 292–299.
Kramers, J.D., Tolstikhin, I.N., 1997. Two major terrestrial Pb isotope paradoxes, forward Martin, H., Moyen, J.F., 2002. Secular changes in TTG composition as markers of the
transport modeling, core formation and the history of the continental crust. Chemical progressive cooling of the Earth. Geology 30 (4), 319–322.
Geology 139, 75–110. Martin, H., Moyen, J.F., Rapp, R.P., 2009. The sanukitoid series: magmatism at the Archaean–
Kramers, J.D., McCourt, S., van Reenen, D.D., 2006. The Limpopo Belt. In: Johnson, M.R., Proterozoic transition. Earth and Environmental Science Transactions of the
Anhaeusser, C.R., Thomas, R.J. (Eds.), The Geology of South Africa. Geological Society Royal Society of Edinburgh 100 (1–2), 15–33.
of South Africa, Johannesburg, South Africa, pp. 209–236. Martin, H., Moyen, J.-F., Guitreau, M., Blichert-Toft, J., Le Pennec, J.-L., 2014. Why Archaean
Kreissig, K., Nagler, T.F., Kramers, J.D., van Reenen, D.D., Smit, C.A., 2000. An isotopic and TTG cannot be generated by MORB melting in subduction zones. Lithos 198–199,
geochemical study of the Kaapvaal Craton and the Southern Marginal Zone of the 1–13.
Limpopo Belt: are they juxtaposed terranes? Lithos 50, 1–25. Martin, H., Smithies, R., Rapp, R.P., Moyen, J.F., Champion, D., 2005. An overview of
Krogstad, E.J., Hanson, G.N., Rajamani, V., 1995. Sources of continental magmatism adja- adakite, TTG, and sanukitoid: relationships and some implications for crustal evolu-
cent to the late Archean Kolar Suture Zone, South India: distinct isotopic and elemen- tion. Lithos 79 (1–2), 1–24.
tal signatures of two late Archean magmatic series. Contributions to Mineralogy and Medeiros, H., Dall'Agnol, R., 1988. Petrologia da porção leste do Batólito Granodiorítico Rio
Petrology 122, 159–173. Maria, sudeste do Pará. Congresso Brasileiro de Geologia, 35. Anais de Congresso
Kröner, A., Jaeckel, P., Brandl, G., 2000. Single zircon ages for felsic to intermediate rocks Brasileiro de Geologia. SBG, Belém, pp. 1488–1499 (v 3).
from the Pietersburg and Giyani greenstone belts and bordering granitoid Meyer, F.M., Robb, L.J., Reimold, W.U., de Bruiyn, H., 1994. Contrasting low- and high-Ca
orthogneisses, northern Kaapvaal Craton, South Africa. Journal of African Earth granites in the Archean Barberton Mountain Land, Southern Africa. Lithos 32 (1–2),
Sciences 30 (4), 773–793. 63–76.
Lagabrielle, Y., Goslin, J., Martin, H., Thirot, J.L., Auzende, J.M., 1997. Multiple active Michon, G., 1987. Les vaugnérites de l'Est du Massif Central Français: apport de l'analyse
spreading centres in the hot North Fidji Basin (Southwest Pacific): a possible multivariée à l'étude géochimique des éléments majeurs. Bulletin de la Societe
model for Archaean seefloor dynamics? Earth and Planetary Science Letters 149 Geologique de France 3 (3), 591–600.
(1–4), 1–13. Mikkola, P., Salminen, P., Torppa, A., Huhma, H., 2011a. The 2.74 Ga Likamännikkö com-
Laurent, O., 2012. Les changements géodynamiques à la transition Archéen- plex in Suomussalmi, East Finland: lost between sanukitoids and truly alkaline
Protérozoïque: étude des granitoïdes de la marge Nord du craton du Kaapvaal rocks? Lithos 125, 716–728.
(Afrique du Sud). Unpublished Ph.D. thesis, University Blaise Pascal, Clermont- Mikkola, P., Huhma, H., Heilimo, E., Whitehouse, M., 2011b. Archean crustal evolution of
Ferrand II, 512 pp (in French). the Suomussalmi district as part of the Kianta Complex, Karelia: constraints from geo-
Laurent, O., Doucelance, R., Martin, H., Moyen, J.F., 2013a. Differentiation of the chemistry and isotopes of granitoids. Lithos 125, 287–307.
late-Archaean sanukitoid series and some implications for crustal growth: insights Miller, C., Schuster, R., Klotzli, U., Frank, W., Grasemann, B., 1999. Post-collisional potassic
from geochemical modelling on the Bulai pluton, Central Limpopo Belt, South and ultrapotassic magmatism in SW Tibet: geochemical and Sr–Nd–Pb–O isotopic
Africa. Precambrian Research 227, 186–203. constraints for mantle source characteristics and petrogenesis. Journal of Petrology
Laurent, O., Martin, H., Doucelance, R., Moyen, J.-F., Paquette, J.-L., 2011. Geochemistry and 40, 1399–1424.
petrogenesis of high-K “sanukitoids” from the Bulai pluton, Central Limpopo Belt, Montel, J.-M., Weisbrod, A., 1986. Characteristics and evolution of “vaugneritic magmas”:
South Africa: implications for geodynamic changes at the Archaean–Proterozoic an analytical and experimental approach, on the example of the Cévennes Médianes
boundary. Lithos 123, 73–91. (French Massif Central). Bulletin of Mineralogy 109, 575–587.
Moyen, J.F., 2011. The composite Archaean grey gneisses: petrological significance, and Reichardt, H., Weinberg, R.F., Andersson, U.B., Fanning, C.M., 2010. Hybridization of gra-
evidence for a non-unique tectonic setting for Archaean crustal growth. Lithos 123 nitic magmas in the source: The origin of the Karakoram Batholith, Ladakh, NW
(1–4), 21–36. India. Lithos 116 (3–4), 249–272.
Moyen, J.F., Martin, H., 2012. Forty years of TTG research. Lithos 148, 312–336. Rey, P.F., Coltice, N., 2008. Neoarchean strengthening of the lithosphere and the coupling
Moyen, J.F., Stevens, G., 2006. Experimental constraints on TTG petrogenesis: implications of the Earth's geochemical reservoirs. Geology 36, 635–638.
for Archean geodynamics. In: Benn, K., Mareschal, J.-C., Condie, K.C. (Eds.), Archean Reymer, A., Schubert, G., 1984. Phanerozoic addition rates to the continental crust and
Geodynamics and Environments. AGU, pp. 149–178. crustal growth. Tectonics 3, 63–77.
Moyen, J.F., van Hunen, J., 2012. Short-term episodicity of Archaean plate tectonics. Robb, L.J., Brandl, G., Anhaeusser, C.R., Poujol, M., 2006. Archean granitoid intrusions. In:
Geology 40 (5), 451–454. Johnson, M.R., Anhaeusser, C.R., Thomas, R.J. (Eds.), The Geology of South Africa.
Moyen, J.F., Martin, H., Jayananda, M., 2001. Multi-element geochemical modelling of Geological Society of South Africa: Johannesburg/Council of Geosciences, pp. 57–94.
crust–mantle interactions during late-Archaean crustal growth: the Closepet granite Roering, C., van Reenen, D.D., Smit, C.A., Barton, J.M., de Beer, J.H., de Wit, M.J., Stettler, E.H.,
(South India). Precambrian Research 112, 87–105. van Schalkwyk, J.F., Stevens, G., Pretorius, S., 1992. Tectonic model for the evolution of
Moyen, J.F., Martin, H., Jayananda, M., Auvray, B., 2003. Late Archaean granites: a typology the Limpopo Belt. Precambrian Research 55, 539–552.
based on the Dharwar Craton (India). Precambrian Research 127, 103–123. Rudnick, R.L., Gao, S., 2003. Composition of the continental crust. In: Rudnick, R.L.
Moyen, J.F., Stevens, G., Kisters, A., 2006. Record of mid-Archaean subduction from meta- (Ed.), The Crust. Treatise on Geochemistry. Elsevier-Pergamon, Oxford, pp.
morphism in the Barberton terrain, South Africa. Nature 442 (3), 559–562. 1–64 (Chap. 3).
Mshiu, E.E., Maboko, M.A.H., 2012. Geochemistry and petrogenesis of the late Archaean Sage, R.P., Lightfoot, P.C., Doherty, W., 1996. Geochemical characteristics of granitoid rocks
high-K granites in the southern Musoma-Mara Greenstone Belt: their influence in from within the Archean Michipicoten Greenstone Belt, Wawa Subprovince, Superior
evolution of Archaean Tanzania Craton. Journal of Africa Earth Sciences 66–67, 1–12. Province, Canada: implications for source regions and tectonic evolution. Precambri-
Nutman, A.P., Bennett, V.C., Friend, C.R.L., Horie, K., Hidaka, H., 2007. Ca. 3, 850 Ma an Research 76, 155–190.
tonalites in the Nuuk region, Greenland: geochemistry and their reworking within Samsonov, A.V., Bogina, M.M., Bibikova, E.V., Petrova, A.Y., Shchipansky, A.A., 2005. The
an Eoarchaean gneiss complex. Contributions to Mineralogy and Petrology 154 (4), relationship between adakitic, calc-alkaline volcanic rocks and TTGs: Implications
385–408. for tectonic setting of the Karelian greenstone belts, Baltic Shield. Lithos 79, 83–106.
Oliveira, M.A., Dall'Agnol, R., Almeida, J.A.C., 2011. Petrology of the Mesoarchean Rio Sanchez-Garrido, C.J.M.G., Stevens, G., Armstrong, R.A., Moyen, J.-F., Martin, H.,
Maria suite and the discrimination of sanukitoid series. Lithos 127, 192–209. Doucelance, R., 2011. Diversity in Earth's early felsic crust: Paleoarchean
Nutman, A.P., Bennett, V.C., Friend, C.R.L., Hidaka, H., Yi, K., Lee, S.R., Kamiichi, T., 2013. The peraluminous granites of the Barberton Greenstone Belt. Geology 39 (10), 963–966.
Itsaq Gneiss Complex of Greenland: episodic 3900 to 3660 Ma juvenile crust forma- Sarvothaman, H., 2001. Archaean high-Mg granitoids of mantle origin in the Eastern
tion and recycling in the 3660 to 3600 Ma Isukasian orogeny. American Journal of Dharwar Craton of Andhra Pradesh. Journal of the Geological Society of India 58,
Science 313 (9), 877–911. 261–268.
Oliveira, M.A., Dall'Agnol, R., Althoff, F.J., Leite, A.A.S., 2009. Mesoarchean sanukitoid rocks Schofield, D.I., Thomas, R.J., Goodenough, K.M., de Waele, B., Pitfield, P.E.J., Key, R.M.,
of the Rio Maria Granite–Greenstone Terrane, Amazonian craton, Brazil. Journal of Bauer, W., Walsg, G.J., Lidke, D.J., Ralison, A.V., Rabarimanana, M., Rafahatelo, J.M.,
South American Earth Sciences 27, 146–160. Randriamananjara, T., 2010. Geological evolution of the Antogil Craton, NE
Oliveira, M.A., Dall'Agnol, R., Scaillet, B., Journal of Petrology, 2010. Petrological constraints Madagascar. Precambrian Research 182, 187–203.
on crystallization conditions of MesoArchean Sanukitoid Rocks, southeastern Scholl, D.W., von Huene, R., 2009. Implications of estimated magmatic additions and
Amazonian craton, Brazil. Journal of Petrology 51, 2121–2148. recycling losses at the subduction zones of accretionary (noncollisional) and colli-
O'Neill, C., Jellinek, A.M., Lenardic, A., 2007a. Conditions for the onset of plate sional (suturing) orogens. In: Cawood, P.A., Kröner, A. (Eds.), Earth Accretionary
tectonics on terrestrial planets and moons. Earth and Planetary Science Letters 261, Systems in Space and Time. Geological Society of London Special Publications,
20–32. 318, pp. 105–125.
O'Neill, C., Lenardic, A., Moresi, L., Torsvik, T.H., Lee, C.A., 2007b. Episodic Precambrian Shirey, S.B., Hanson, G.N., 1984. Mantle-derived Archaean monzodiorites and
subduction. Earth and Planetary Science Letters 262, 552–562. trachyandesites. Nature 310, 222–224.
Patiño-Douce, A.E., Johnston, D., 1991. Phase equilibria and melt productivity in the pelitic Shirey, S.B., Richardson, S.H., 2011. Start of the Wilson cycle at 3 Ga shown by diamonds
system: implications for the origin of peraluminous granitoids and aluminous granu- from subcontinental mantle. Science 333, 434–436.
lites. Contributions to Mineralogy and Petrology 107, 202–218. Singh, J., Johannes, W., 1996. Dehydration melting of tonalites. 2. Compositions of melts
Patiño-Douce, A.E., Beard, J.S., 1995. Dehydration-melting of biotite gneiss and quartz and solids. Contributions to Mineralogy and Petrology 125, 26–44.
amphibolite from 3 to 15 kbar. Journal of Petrology 36, 707–738. Sisson, T., Ratajeski, K., Hankins, W., Glazner, A., 2005. Voluminous granitic magmas from
Patiño-Douce, A.E., Beard, J.S., 1996. Effects on P, f(O2) and Mg/Fe ratio on dehydration common basaltic sources. Contributions to Mineralogy and Petrology 148, 635–661.
melting of model metagreywackes. Journal of Petrology 37 (5), 999–1024. Sizova, E., Gerya, T., Brown, M., Perchuk, L.L., 2010. Subduction styles in the Precambrian:
Percival, J.A., 1994. Archaean high-grade metamorphism. In: Condie, K.C. (Ed.), Archaean insight from numerical experiments. Lithos 116, 209–229.
Crustal Evolution. Elsevier, Amsterdam, pp. 357–410. Skjerlie, K.P., Johnston, A.D., 1996. Vapour-absent melting from 10 to 20 kbar of crustal rocks
Percival, J.A., Sanborn-Barrie, M., Skulski, T., Stott, G.M., Helmstaedt, H., White, D.J., 2006. that contain multiple hydrous phases: implications for anatexis in the deep to very deep
Tectonic evolution of the western Superior Province from NATMAP and lithoprobe continental crust and active continental margins. Journal of Petrology 37, 661–691.
studies. Canadian Journal of Earth Sciences 43, 1085–1117. Smithies, R.H., 2000. The Archaean tonalite–trondhjemite–granodiorite (TTG) series is
Poujol, M., 1997. Étude U–Pb et Pb–Pb de la Murchison greenstone belt et du bassin not an analogue of Cenozoic adakite. Earth and Planetary Science Letters 182,
aurifère d'Evander, Afrique du Sud: implications pour l'évolution du Kaapvaal Craton. 115–125.
Unpublished Ph.D. thesis, Université de Montpellier, 247 pp. Smithies, R.H., Champion, D.C., 1999a. High-Mg diorite from the Archaean Pilbara Craton;
Poujol, M., 2001. U–Pb isotopic evidence for episodic granitoid emplacement in the anorogenic magmas derived from a subduction-modified mantle. Geological Survey
Murchison greenstone belt, South Africa. Journal of African Earth Sciences 33 (1), of Western Australia, Annual Review 45–59.
155–163. Smithies, R.H., Champion, D.C., 1999b. Late Archean felsic alkaline igneous rocks in the
Poujol, M., Robb, L.J., 1999. New U–Pb zircon ages on gneisses and pegmatite from south Eastern Goldfields, Yilgarn Craton, Western Australia: a result of lower crustal delam-
of the Murchison greenstone belt, South Africa. South African Journal of Geology 102 ination? Journal of the Geological Society 156, 561–576.
(2), 93–97. Smithies, R.H., Champion, D.C., 2000. The Archaean high-Mg diorite suite: links to
Poujol, M., Robb, L.J., Anhaeusser, C.R., Gericke, B., 2003. A review of the geochronological tonalite–trondhjemite–granodiorite magmatism and implications for early Archaean
constraints on the evolution of the Kaapvaal Craton, South Africa. Precambrian Re- crustal growth. Journal of Petrology 41 (12), 1653–1671.
search 127 (1–3), 181–213. Smithies, R.H., Champion, D.C., Sun, S.S., 2004. Evidence for Early LREE-enriched mantle
Poujol, M., Robb, L.J., Respaut, J.-P., Anhaeusser, C.R., 1996. 3.07–2.97 Ga greenstone belt source regions: diverse magmas from the c. 3.0 Ga Mallina Basin, Pilbara Craton,
formation in the northeastern Kaapvaal Craton: implications for the origin of NW Australia. Journal of Petrology 45 (8), 1515–1537.
the Witwatersrand Basin. Economic Geology 91, 1455–1461. Smithies, R.H., van Kranendonk, M.J., Champion, D.C., 2007. The Mesoarchean emergence
Prabhakar, B.C., Jayananda, M., Shareef, M., Kano, T., 2009. Petrology and geochemistry of of modern-style subduction. Gondwana Research 11 (1–2), 50–68.
Late Archean granitoids in the northern part of Eastern Dharwar, southern India: im- Solgadi, F., Moyen, J.-F., Vanderhaeghe, O., Sawyer, E.W., Reisberg, L., 2007. The relative
plications for transitional geodynamic setting. Journal of the Geological Society of roles of crustal anatexis and mantle-derived magmas: generation of synorogenic,
India 74, 299–317. Hercynian granites in the Livradois area, French Massif Central. Canadian Mineralogist
Rapp, R.P., Watson, E.B., 1995. Dehydration melting of metabasalt at 8–32 kbar: implica- 45, 581–606.
tions for continental growth and crust–mantle recycling. Journal of Petrology 36 (4), Steenfelt, A., Garde, A.A., Moyen, J.F., 2005. Mantle wedge involvement in the petrogenesis
891–931. of Archaean grey gneisses in West Greenland. Lithos 79, 207–228.
Rapp, R.P., Norman, M.D., Laporte, D., Yaxley, G.M., Martin, H., Foley, S.F., 2010. Continent Stern, R.A., Hanson, G.N., 1991. Archaean High-Mg granodiorites: a derivative of light
formation in the Archean and chemical evolution of the cratonic lithosphere: melt– rare earth enriched monzodiorites of mantle origin. Journal of Petrology 32,
rock reaction experiments at 3–4 GPa and petrogenesis of Archean Mg-diorites 201–238.
(sanukitoids). Journal of Petrology 51 (6), 1237–1266. Stern, R.A., Hanson, G.N., Shirey, S.B., 1989. Petrogenenesis of mantle-derived, LILE-
Rapp, R.P., Shimizu, N., Norman, M.D., 2003. Growth of early continental crust by partial enriched Archean monzodiorites and trachyandesites (sanukitoids) in southwestern
melting of eclogite. Nature 425, 605–609. Superior Province. Canadian Journal of Earth Sciences 26, 1688–1712.
Rapp, R.P., Shimizu, N., Norman, M.D., Applegate, G.S., 1999. Reaction between slab- Stevenson, R., Henry, P., Gariépy, C., 1999. Assimilation-fractional crystallization origin of
derived melts and peridotite in the mantle wedge: experimental constraints at 3. Archaean sanukitoid suites: Western Superior Province, Canada. Precambrian
8 GPa. Chemical Geology 160, 335–356. Research 96, 83–99.
Rapp, R.P., Watson, E.B., Miller, C.F., 1991. Partial melting of amphibolite/eclogite and Sutcliffe, R.H., Smith, A.R., Doherty, W., Barnett, R.L., 1990. Mantle derivation of Archean
the origin of Archaean trondhjemites and tonalites. Precambrian Research 51, amphibole-bearing granitoid and associated mafic rocks: evidence from the southern
1–25. Superior Province, Canada. Contributions to Mineralogy and Petrology 105, 255–274.
Sylvester, P.J., 1994. Archaean granite plutons. In: Condie, K.C. (Ed.), Archaean Crustal Whalen, J.B., Percival, J.A., McNicoll, V.J., Longstaffe, F.J., 2002. A mainly crustal origin for
Evolution. Developments in Precambrian Geology, vol. 11. Elsevier, Amsterdam, pp. tonalitic granitoid rocks, Superior Province, Canada: implications for late Archean
261–314. tectonomagmatic processes. Journal of Petrology 43 (8), 1551–1570.
Taylor, J., Nicoli, G., Stevens, G., Frei, D., Moyen, J.-F., 2014. The processes that control Whalen, J.B., Percival, J.A., McNicoll, V.J., Longstaffe, F.J., 2004. Geochemical and isotopic
leucosome compositions in metasedimentary granulites: perspectives from the (Nd–O) evidence bearing on the origin of late- to post-orogenic high-K granitoid
Southern Marginal Zone migmatites, Limpopo Belt, South Africa. Journal of Meta- rocks in the western Superior Province: implications for late-Archaean tectonomagmatic
morphic Geology. http://dx.doi.org/10.1111/jmg.12087. processes. Precambrian Research 132, 303–326.
Taylor, S.R., 1987. Geochemical and petrological significance of the Archaean–Proterozoic Wilde, S.A., Valley, J.W., Peck, W.H., Graham, C.M., 2001. Evidence from detrital zircons for
boundary. In: Pharaoh, T.C., Beckinsale, R.D., Rickard, D. (Eds.), Geochemistry and the existence of continental crust and oceans on the Earth 4.4 Gyr ago. Nature 409,
Mineralization of Proterozoic Volcanic Suites. Geological Society Special Publication, 175–178.
33, pp. 3–8. Willbold, M., Hegner, E., Stracke, A., Rocholl, A., 2009. Continental geochemical signatures
Taylor, S.R., McLennan, S.M., 1985. The Continental Crust: Its Composition and Evolution. in dacites from Iceland and implications for models of early Archaean crust forma-
Blackwell Scientific, Boston, Mass (312 pp.). tion. Earth and Planetary Science Letters 279, 44–52.
Taylor, S.R., McLennan, S.M., 1995. The geochemical evolution of the continental Williams, H.M., Turner, S.P., Pearce, J.A., Kelley, S.P., Harris, N.B.W., 2004. Nature of the
crust. Reviews of Geophysics 33 (2), 241–265. source regions for post-collisional, potassic magmatism in southern and northern
Thébaud, N., Rey, P.F., 2013. Archean gravity-driven tectonics on hot and flooded conti- Tibet from geochemical variations and inverse trace element modelling. Journal of
nents: controls on long-lived mineralised hydrothermal systems away from conti- Petrology 45, 555–607.
nental margins. Precambrian Research 229, 83–104. Williamson, B., Downes, H., Thirlwall, M., Beard, A., 1997. Geochemical constraints on
Turner, S., Sandiford, M., Foden, J., 1992. Some geodynamic and compositional constraints restite composition and unmixing in the Velay anatectic granite, French Massif
on ‘postorogenic’ magmatism. Geology 20, 931–934. Central. Lithos 40 (2–4), 295–319.
Turpin, L., Cuney, M., Friedrich, M., Bouchez, J.L., Aubertin, M., 1990. Meta-igneous origin Williamson, B.J., Downes, H., Thirlwall, M.F., 1992. The relationship between crustal mag-
of Hercynian peraluminous granites in N.W. French Massif Central: implications for matic underplating and granite genesis—an example from the Velay granite complex,
crustal history reconstructions. Contributions to Mineralogy and Petrology 104 (2), Massif-Central, France. Transactions of the Royal Society of Edinburgh-Earth Sciences
163–172. 83, 235–245.
Valley, J.W., Lackey, J.S., Cavosie, A.J., Clechenko, C.C., Spicuzza, M.J., Basei, M.A.S., Windley, B.F., 1984. The Archaean–Proterozoic boundary. Tectonophysics 105 (1–4),
Bindeman, I.N., Ferreira, V.P., Sial, A.N., King, E.M., Peck, W.H., Sinha, A.K., Wei, C.S., 43–53.
2005. 4.4 billion years of crustal maturation: oxygen isotope ratios of magmatic Wolf, M.B., Wyllie, P.J., 1994. Dehydration-melting of amphibolite at 10 kbar: the ef-
zircon. Contributions to Mineralogy and Petrology 150, 561–580. fects of temperature and time. Contributions to Mineralogy and Petrology 115,
Vanderhaeghe, O., 2010. The thermo–mechanical evolution of crustal orogenic belts at 369–383.
convergent plate boundaries: a reappraisal of the orogenic cycle. Journal of Xu, Y.G., Menzies, M.A., Thirlwall, M.F., Xie, G.H., 2001. Exotic lithosphere mantle beneath
Geodynamics 56–57, 124–145. the western Yangtze craton: petrogenetic links to Tibet using highly magnesian
van Hunen, J., Allen, M.B., 2011. Continental collision and slab break-off: a comparison of ultrapostassic rocks. Geology 29, 863–866.
3-D numerical models with observations. Earth and Planetary Science Letters 302 Yang, J.H., Wu, F.Y., Wilde, S.A., Zhao, G., 2008. Petrogenesis and geodynamics of Late-
(1–2), 27–37. Archaean magmatism in eastern Hebei, eastern North China Craton: geochrono-
van Hunen, J., van den Berg, A.P., 2008. Plate tectonics on the early Earth: Limitations logical, geochemical and Nd–Hf isotopic evidence. Precambrian Research 167,
imposed by strength and buoyancy of subducted lithosphere. Lithos 103, 217–235. 125–149.
van Kranendonk, M.J., Smithies, H.R., Hickman, A.H., Champion, D.C., 2007. Review: secular Yearron, L.M., 2003. Archaean granite petrogenesis and implications for the evolution of
tectonic evolution of Archean continental crust: interplay between horizontal and verti- the Barberton Mountain Land, South Africa. Unpublished Ph.D thesis, Kingston
cal processes in the formation of the Pilbara Craton, Australia. Terra Nova 19 (1), 1–38. University, London, 315 pp.
van Reenen, D.D., Barton Jr., J.M., Roering, C., Smit, C.A., van Schalkwyk, J.F., 1987. Deep Zegers, T.E., van Keken, P.E., 2001. Middle Archean continent formation by crustal delam-
crustal response to continental collision: the Limpopo belt of southern Africa. Geology ination. Geology 29 (12), 1083–1086.
15, 11–14. Zeh, A., Gerdes, A., 2012. U–Pb and Lu–Hf isotope record of detrital zircons from gold-
van Thienen, P., Vlaar, N.J., van den Berg, A.P., 2004. Plate tectonics on the terrestrial bearing sediments of the Pietersburg Greenstone Belt (South Africa)—is there a com-
planets. Physics of the Earth and Planetary Interiors 142, 61–74. mon provenance with the Witwatersrand Basin? Precambrian Research 204–205,
Veizer, J., McKenzie, F.T., 2003. Evolution of sedimentary rocks. Treatise on Geochemistry, 46–56.
vol. 7. Elsevier-Pergamon, Oxford, pp. 369–407. Zeh, A., Gerdes, A., Barton Jr., J.M., 2009. Archean accretion and crustal evolution of the
Vielzeuf, D., Holloway, J.R., 1988. Experimental determination of fluid absent melting Kalahari Craton—the Zircon Age and Hf isotope record of granitic rocks from
relations in the pelitic system. Contributions to Mineralogy and Petrology 98, 257–276. Barberton/Swaziland to the Francistown Arc. Journal of Petrology 50, 933–966.
Vijaya Kumar, K., Ernst, W.G., Leelanandam, C., Wooden, J.L., Grove, M.J., 2011. Origin Zeh, A., Jaguin, J., Poujol, M., Boulvais, P., Block, S., Paquette, J.-L., 2013. Juvenile crust
of ~ 2.5 Ga potassic granite from the Nellore Schist Belt, SE India: textural, formation in the northeastern Kaapvaal Craton at 2.97 Ga—implications for Archean
cathodoluminescence, and SHRIMP U–Pb data. Contributions to Mineralogy and terrane accretion, and the source of the Pietersburg gold. Precambrian Research
Petrology 162, 867–888. 233, 20–43.
von Raumer, J.F., Finger, F., Veselá, P., Stampfli, G.M., 2013. Durbachites–Vaugnerites — a Zeng, L., Gao, L.E., Xie, K., Liu-Zeng, J., 2011. Mid-Eocene high Sr/Y granites in the Northern
geodynamic marker in the central European Variscan orogen. Terra Nova. http://dx. Himalayan Gneiss Domes: melting thickened lower continental crust. Earth and Plan-
doi.org/10.1111/ter.12071. etary Science Letters 303 (3–4), 251–266.
Wang, Y., Zhang, Y., Zhao, G., Fan, W., Xia, X., Zhang, F., Zhang, A., 2009. Zircon U–Pb Zhai, M., Kampunzu, A.B., Modisi, M.P., Bagai, Z., 2006. Sr and Nd isotope systematics of
geochronological and geochemical constraints on the petrogenesis of Taishan the Francistown plutonic rocks, Botswana: implications for Neoarchean crustal evolu-
sanukitoids (Shandong): implications for Neoarchean subduction in the Eastern tion of the Zimbabwe craton. Geologische Rundschau 95, 355–369.
Block, North China Craton. Precambrian Research 174, 273–286. Zhao, Z., Mo, X., Dilek, Y., Niu, Y., dePaolo, D.J., Robinson, P., Zhu, D., Sun, C., Dong, G., Zhou,
Watkins, J.M., Clemens, J.D., Treloar, P.J., 2007. Archaean TTGs as sources of younger S., Luo, Z., Hou, Z., 2009. Geochemical and Sr–Nd–Pb–O isotopic compositions of the
granitic magmas: melting of sodic metatonalites at 0.6–1.2 GPa. Contributions to post-collisional ultrapotassic magmatism in SW Tibet: petrogenesis and implications
Mineralogy and Petrology 154, 91–110. for India intra-continental subduction beneath southern Tibet. Lithos 113, 190–212.

Classification of Archean Granitoids

Uploaded by

Copyright:

Available Formats

Classification of Archean Granitoids

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Classification of Archean Granitoids

Uploaded by

Copyright:

Available Formats

Lithos 205 (2014) 208–235

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/lithos

Invited review article

The diversity and evolution of late-Archean granitoids: Evidence for the

2.1.3. Petrogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211

1. Introduction Schematically, an Archean crustal segment consists of three

Proterozoic boundary and intruded by large volumes of 3.1 to 2.6 Ga-

silicic, plagioclase-rich plutonic rocks (tonalites, trondhjemites and

O. Laurent et al. / Lithos 205 (2014) 208–235

TTG Sanukitoids s.l. Biotite- and two-mica granites “Hybrid” granitoids

O. Laurent et al. / Lithos 205 (2014) 208–235

29°E 30°E 31°E

Major shear zone Town

TTG Sanukitoids s.l. Biotite- and two-mica Hybrid granites

Major elements (wt.%)

Trace elements (ppm)

Al2O3 / (CaO + Na2O + K2O) (mol.)

Al2O3 / (FeOt + MgO)

Sanukitoids s.l. Biotite- and two-mica granites Hybrid granites

600 1.0 No Eu anomaly

Sanukitoids s.l. Biotite- and two-mica granites Hybrid granites

Na2O / K2O Na2O / K2O

Na2O / K2O Na2O / K2O

(c) Biotite-, (d) Hybrid

Emplacement ages (1 box = 1 sample) Ages of detrital zircons

Stage 1 (~0.5 Ga)

2.8 Stage 2 (~0.15 Ga)

Superior Amazonian Karelian Kaapvaal Dharwar N.China Pilbara

εNd (t) Whole rock

+4 DM +4 DM TTG Model Ages

εNd (t) Whole rock

amphibole-bearing mantle peridotite (e.g. Gao et al., 2010; Mahéo 500

(1) Low FeO t + MgO (b10 wt.%) peraluminous muscovite- or UC

(a) ~20 km (b) Tectonically and magmatically (c) Thermal relaxation

Crustal melting Crustal melting

Magmatic stage 1 Magmatic stage 2

TTG magmatism Sanukitoids s.l. Biotite- and two-mica ‘‘Hybrid’’ granitoids

Vanderhaeghe, 2010), has been consistently proposed as tectonic

5.1. Comparison with Eo- to Mesoarchean granitoids

A two-stage evolution of granitoid magmatism, from dominant

~ 3.25 Ga (Barley and Pickard, 1999; Bickle et al., 1989; van

Kranendonk et al., 2007). 3.2

lent to late-Archean domains. The situation is rather different in the

(a) 50 6. Summary and concluding remarks

99% in relation to the TTG-dominated magmatic record of N3.0 Ga. These

75 magmas or source rocks from the ﬁrst three groups.

tion of these stages are consistent with a subduction–collision evolu-

tion. In particular, granitoids emplaced in modern collisional settings

You might also like