Tải bản đầy đủ - 0trang
370



Chapter 4.3: Geochemistry of Paleoarchean Granites of the East Pilbara Terrane



of this age, seems reasonably well understood, many crucial questions remain unanswered.

Two of the more controversial aspects are the actual sites of TTG magma formation (i.e.,

thickened mafic crust, subducting slab, or both), and the roles of intracrustal differentiation

and evolved crustal contributions. The timing and extent of intracrustal differentiation has

implications for the onset of potassic granite magmatism that typically dominates the late

stages of many Archean crustal evolution cycles, and for understanding the influences on

the more expanded evolutionary sequence from fractionated tholeiitic rocks to TTGs to

potassic magmatism (e.g., Champion and Smithies, 2001).

Perhaps the best exposed and best understood Paleoarchean rocks are those of the East

Pilbara Terrane (EPT), in the Pilbara Craton of north Western Australia (Van Kranendonk

et al., 2006a, 2007a, this volume). The Paleoarchean rock record of the East Pilbara Terrane comprises quasi-continuous greenstone development in conjunction with similar aged,

though more episodic, felsic magmatism (extrusive and intrusive), all in an autochthonous

geological environment. The structural integrity of the terrane, combined with the well

preserved geology of the sequences, provide strong constraints both on tectonic models

and on petrogenetic models of granite formation. Particularly helpful are the concurrent

records of felsic and mafic magmatism and, through these, the links that can be established

between the mantle and the crust during Paleoarchean times (e.g., Smithies et al., 2005b,

2007a).

This paper describes the geology and geochemistry of the pre-3.2 Ga TTGs, TTG-like

granites and more potassic granites of the East Pilbara Terrane and the constraints on their

petrogenesis. A companion paper by Smithies et al. (this volume) looks at the evolution

of contemporaneous felsic and mafic extrusive rocks of the East Pilbara Terrane. Insights

from that study are included here to establish more robust petrogenetic and tectonic models

for the pre-3.2 Ga felsic magmatism, and to provide constraints on crustal growth models

for the early Archean. In particular, these models emphasise the role of compositionally

evolved crust in TTG formation, and may possibly have wider application to early Archean

granite genesis, in general.



4.3-2. REGIONAL GEOLOGICAL SUMMARY

The Pilbara Craton is divided into: the 3.53–3.07 Ga East Pilbara Terrane (EPT), which

includes the Pilbara Supergroup; the 3.27–3.11 Ga West Pilbara Superterrane (WPS); and

the 3.2 Ga Kurrana Terrane (KT). Each of these are distinguished by unique lithostratigraphy, structural map patterns, geochemistry and tectonic histories and they are separated

by late tectonic, dominantly clastic sedimentary rocks of the De Grey Supergroup deposited

in the Mallina and Mosquito Creek basins (see Fig. 4.1-1). A detailed account of the geology of these regions is provided by Van Kranendonk et al. (2002, 2006a, 2007a, this

volume) and will not be repeated here.



4.3-3. Granite Geochemistry and Petrology



371



4.3-3. GRANITE GEOCHEMISTRY AND PETROLOGY

4.3-3.1. Introduction

The exposure of preserved granite totals an area of ∼24,000 km2 . Most of the known

periods of granite intrusion in the EPT either broadly correspond to periods of greenstone

development (i.e., are contemporaneous with felsic volcanism in the Pilbara Supergroup),

or postdate greenstone formation. These magmatic periods include ca. pre-3.5 Ga, 3.50–

3.42 Ga, ca. 3.32 Ga, ca. 3.24 Ga, 2.95–2.93 Ga, and ca. 2.85 Ga, with most magmatism

relating to the 3.50–3.42, ca. 3.32, ca. 3.24 and 2.95–2.93 Ga events (see Fig. 4.1-7: Bickle

et al., 1983, 1989, 1993; Buick et al., 1995, 2002; Nelson, 1996–2002; Barley et al., 1998;

Barley and Pickard, 1999; Van Kranendonk et al., 2002, 2006a, 2007a). The younger, post3.1 Ga events are concentrated within the western part of the EPT (Champion and Smithies,

2000), as well as the adjoining Mallina Basin (Smithies and Champion, 2000; Smithies et

al., 2004a; Van Kranendonk et al., 2006a, 2007a, this volume). Sm-Nd isotope data (see

Smithies et al., 2007b) show that there is a marked change in average model crustal age

(T2DM ) between the dominantly 3.6–3.2 Ga EPT (T2DM ages of >3.6–3.3) and the 3.02–

2.94 Ga Mallina Basin (T2DM ages of 3.2 to 2.96 Ga) (Figs. 4.1-1, 4.1-7). This Nd-isotopic

discontinuity represents largely juvenile crustal growth and accretion to the west of the

EPT, related to the tectonic evolution of the WPS and Mallina Basin (Smith et al., 1998;

Champion and Smithies 2000; Smithies and Champion 2000; Smithies et al., 2004a). The

geological and Sm-Nd data further indicate that post-3.2 Ga crustal growth extended to

the western part of the EPT (Champion and Smithies, 2000), and the post 3.2 Ga granites

in that region are, accordingly, petrogenetically grouped with those found in the Mallina

Basin. Broadly coinciding with the Nd isotopic break in the western EPT is the change

to dominantly younger granites, emplaced at 2.94 and ca. 2.85 Ga (e.g., Champion and

Smithies, 2000). Magmatism of this age occurs elsewhere within the EPT, but is not as

voluminous.

4.3-3.2. Ca. 3.5–3.42 Ga granites

The oldest, widely preserved granites range in age from ca. 3.50–3.42 Ga (Fig. 4.1-7).

These are best documented from the Shaw Granitic Complex (Bickle et al., 1983, 1993;

Van Kranendonk et al., 2002, 2006a), but are also present within the Carlindi, Yule, Warrawagine, Mount Edgar, Corunna Downs and Muccan Granitic Complexes (Fig. 4.1-7:

Buick et al., 1995; Nelson, 1998, 1999; Williams and Collins, 1990; Bagas et al., 2003; Van

Kranendonk et al., 2006a, this volume). Units dominantly comprise hornblende-biotite and

biotite tonalites, trondhjemites and granodiorites (e.g., Jahn et al., 1981; Bickle et al., 1983,

1989). They vary from discrete plutons of variably deformed granite, such as in the northern part of the Shaw Granitic Complex (Bickle et al., 1993), to strongly deformed, locally

migmatitic and variably-banded gneisses. The latter range from large zones of probably

mixed ages (e.g., Shaw Granitic Complex; Bickle et al., 1993; Van Kranendonk, 2003:

Mt Edgar Granitic Complex; Collins, 1993), to marginal facies in granitic complexes,



372



Chapter 4.3: Geochemistry of Paleoarchean Granites of the East Pilbara Terrane



commonly extensively intruded by younger granites and locally forming abundant xenoliths/enclaves within them (e.g., Yule Granitic Complex; Blewett and Champion, 2005).

Van Kranendonk et al. (2006a) subdivided the pre-3.42 Ga granites into the 3.5–3.46 Ga

Callina Supersuite and the 3.45–3.42 Ga Tambina Supersuite (Fig. 4.1-7, Table 4.3-1).

These authors also showed that the intrusive rocks were emplaced contemporaneously with

development of the Pilbara Supergroup and, more specifically, with felsic extrusive units

within the supergroup (e.g., the Duffer Formation/Callina Supersuite and Panorama Formation/Tambina Supersuite extrusive/intrusive pairs).

Although there is an apparent age difference between the two supersuites and also some

evidence for structural differences (i.e., the Tambina Supersuite is locally significantly less

migmatised; Van Kranendonk, 2003), distinguishing between members of each supersuite

is difficult, such that granites of both supersuites are discussed together in this paper (see

Table 4.3-1). Older granites, preserved as remnants in gneiss, are also included within this

group (e.g., 3.655 Ga remnants within >3.42 Ga gneiss of Warrawagine Granitic Complex: Nelson, 1999; Williams, 2001). Similarly, composite units (such as the Nandingarra

Granodiorite in the Corunna Downs Granitic Complex; Bagas et al., 2003), which comprise a mixture of ca. 3.45 and 3.315 Ga ages (Nelson, 2002), are here treated with the ca.

3.5–3.42 Ga granites. The gneissic rocks of Collins (1993), which may be composite in

part, are treated similarly. A full listing of granite units included in this group is given in

Table 4.3-1.

4.3-3.3. Ca. 3.32–3.24 Ga granites

The ca. 3.32–3.24 Ga granites are best documented in the Mount Edgar (Collins, 1983,

1993; Williams and Collins, 1990) and Corunna Downs Granitic Complexes (Davy, 1988;

Barley and Pickard, 1999; Bagas et al., 2003), where they comprise the bulk of the complexes (e.g., Van Kranendonk et al., 2006a: Fig. 4.1-7). Recent data suggests that the older

granites (ca. 3.325 to 3.290 Ga) of this group also dominate the Warrawagine Granitic

Complex (Williams, 2002) and are present in the Muccan Granitic Complex (Williams,

1999; Fig. 4.1-7). They appear to be largely absent from the Shaw, Yule and Carlindi

Granitic Complexes (Fig. 4.1-7), although the presence of inherited zircons of this age

may suggest that such granites once formed a component of these complexes as well.

Rock types range from hornblende-biotite to biotite, tonalite, granodiorite, and trondhjemite to monzogranite (e.g., Davy, 1988; Collins, 1993; Barley and Pickard, 1999;

Williams, 1999, 2001; Bagas et al., 2003). Van Kranendonk et al. (2006a) grouped granites

of this age (ca. 3.325 to 3.290 Ga) into the Emu Pool Supersuite. Younger granites of this

group (ca. 3.27-3.24 Ga granites) are widespread and occur as minor components within

the Yule, Mount Edgar, Muccan and Warrawagine Granitic Complexes (Nelson, 1998,

1999; Williams, 1999, 2001; Van Kranendonk et al., 2002, 2006a), as well as forming discrete intrusions, such as the Strelley Monzogranite (Buick et al., 2002) (Fig. 4.1-7). Rock

types comprise hornblende-biotite and biotite tonalite, granodiorite and granite (Collins,

1993; Williams, 2001; Buick et al., 2002). Van Kranendonk et al. (2006a) grouped granites



Groups



Granite complex Classification



Data sources



Previous classifications



ca. 3.33–3.25 Ga group

ca. 3.3 to 3.25 Ga

Munganbrina



Mount Edgar



Muccan



Muccan



high-Al?



Collins (1983); GSWA-GA

unpublished

high-Al, low-Al GSWA-GA unpublished



ca. 3.3 Ga

Boobina



Corunna Downs



low-Al



Boodallana



Mount Edgar



high-Al



Carbana



Corunna Downs



low-Al



Carbana low-Th



Corunna Downs



low-Al



Chimingadgi



Mount Edgar



high-Al



Coppin Gap



Mount Edgar



high-Al



Corunna Unit 2



Corunna Downs



low-Al



Corunna Unit 1



Corunna Downs



low-Al



Mondana



Corunna Downs



low-Al



Davy (1988), Bagas et al. (2003),

GSWA-GA unpublished

Collins (1983); GSWA-GA

unpublished

Davy (1988), Bagas et al. (2003),

GSWA-GA unpublished

Davy (1988), Bagas et al. (2003),

GSWA-GA unpublished

Collins (1983); GSWA-GA

unpublished

Collins (1983); GSWA-GA

unpublished



Emu Supersuite (Van Kranendonk et

al., 2006)

Boobina – Group 2 (Bagas et al., 2003)

Boodallana Suite (Collins, 1983, 1993)

Carbana – Group 2 (Bagas et al., 2003)

Carbana – Group 2 (Bagas et al., 2003)

Chimingadgi Suite (Collins, 1983,

1993)

Coppins Gap Suite (Collins, 1983,

1993)



Unnamed unit – Group 2 (Bagas et al.,

2003)

Unnamed unit – Group 2 (Bagas et al.,

2003)

Mondana – Group 3 (Bagas et al.,

2003)



373



Davy (1988), Bagas et al. (2003),

GSWA-GA unpublished

Davy (1988), Bagas et al. (2003),

GSWA-GA unpublished

Davy (1988), Bagas et al. (2003),

GSWA-GA unpublished



Emu & Cleland Supersuites (Van

Kranendonk et al., 2006)

Munganbrina Suite (Collins, 1983,

1993)



4.3-3. Granite Geochemistry and Petrology



Table 4.3-1. Geochemical subdivisions for granite batholiths, as used in this paper, showing current high-Al low-Al groupings and previous

classifications, and data sources



374



Table 4.3-1. (Continued)

Granite complex Classification Data sources



Previous classifications



Triberton



Corunna Downs low-Al



Warrulinya

Yandicoogina



Mount Edgar

Mount Edgar



Davy (1988), Bagas et al. (2003),

GSWA-GA unpublished

Collins (1983)

Collins (1983); GSWA-GA

unpublished



Triberton – Group 1 (Bagas et al.,

2003)

Warrulinya Suite (Collins, 1983, 1993)

Yandicoogina Suite (Collins, 1983,

1993)



Davy (1988), Bagas et al. (2003),

GSWA-GA unpublished



Emu & Callina & Tambina Supersuites

(Van Kranendonk et al., 2006)

Nandingarra – Group 1 (Bagas et al.,

2003)



high-Al

high-Al



ca. 3.5–3.42 Ga group

ca. 3.5 to 3.3 Ga granites

Nandingarra



Corunna Downs low-Al



ca. 3.5 to 3.42 Ga granites



Tambina & Callina Supersuites (Van

Kranendonk et al., 2006)



Carlindi

Mt Edgar high-Al



Carlindi

Mount Edgar



Mt Edgar low-Al



Mount Edgar



Mt Edgar high-Al, low-Th



Mount Edgar



Muccan

Shaw high-Al

Shaw layered

Yule



Muccan

Shaw

Shaw

Yule



high-Al

high-Al



GSWA-GA unpublished

Collins (1983); GSWA-GA

unpublished

low-Al

Collins (1983); GSWA-GA

unpublished

high-Al

Collins (1983); GSWA-GA

unpublished

high-Al

GSWA-GA unpublished

high-Al

Bickle et al. (1989, 1993)

high-Al

Bickle et al. (1989, 1993)

low-Al mostly GSWA-GA unpublished



layered unit (Bickle et al., 1989, 1993)

Yule (Champion & Smithies, 2000)



Age subdivisions follows Van Kranendonk et al. (2006). Refer to Fig. 4.1-7 for distribution of granite complexes and granite age groups



Chapter 4.3: Geochemistry of Paleoarchean Granites of the East Pilbara Terrane



Groups



4.3-4. Geochemistry



375



of this age into the Cleland Supersuite. A full listing of granite units included in the ca.

3.32–3.24 Ga granite group is given in Table 4.3-1.



4.3-4. GEOCHEMISTRY

4.3-4.1. Analytical Methods

Geochemical data (major elements, trace elements and Sm-Nd isotopes) used in this paper include both previously collected published (Bickle et al., 1983, 1993; Collins, 1993;

McCulloch, 1987; Davy, 1988) and unpublished data (Collins, 1983), and new data for

either newly collected rocks (GA and GSWA unpublished data) or reanalysis of previously collected samples (from Davy, 1988). All new major and trace element geochemical

data were analysed at Geoscience Australia, Canberra. Major elements were determined

by wavelength-dispersive XRF on fused disks using methods similar to those of Norrish

and Hutton (1969). Precision is better than ±1% of the reported values. Loss on Ignition

(LOI) was determined by gravimetry after combustion at 1100 ◦ C. FeO abundances were

determined by digestion and electrochemical titration using a modified method based on

Shapiro and Brannock (1962). The trace elements Ba, Cr, Cu, Ni, Sc, V, Zn and Zr were

determined by wavelength-dispersive XRF on a pressed pellet using methods similar to

those of Norrish and Chappell (1977), while Cs, Ga, Nb, Pb, Rb, Sr, Ta, Th, U, Y and

the REE were analysed by ICP-MS (Perkin Elmer ELAN 6000) using methods similar to

those of Eggins et al. (1997), but on solutions obtained by dissolution of fused glass disks

(Pyke, 2000). Sm-Nd isotopic analyses were determined by isotope dilution at VIEPS Radiogenic Isotope Laboratory, Department of Earth Sciences, La Trobe University, Victoria.

Analytical techniques follow the method reported in Waight et al. (2000).

4.3-4.2. Ca. 3.5–3.42 Ga Granites

The geochemistry of 3.5–3.42 Ga granites was previously discussed by Bickle et al. (1983,

1993) for the Shaw Granitic Complex, by Collins (1993) and Davy and Lewis (1986) for

the Mt Edgar Granitic Complex, by Davy (1988) and Bagas et al. (2003) for the Corunna

Downs Granitic Complex, and by Champion and Smithies (2000) for the Yule Granitic

Complex. Granites of this age have an expanded silica range (62–76% SiO2 ). TiO2 , Al2 O3 ,

FeO* , MgO, CaO and P2 O5 are all negatively correlated with SiO2 , generally forming

broad but well-defined trends (Fig. 4.3-1). In detail, chemical differences between and

within granitic complexes are evident for most of the major elements; for example, Nandingarra granites in the Corunna Downs Granitic Complex have elevated MgO (Fig. 4.3-1)

and low P2 O5 relative to other granites with the same SiO2 contents. The most consistent

differences are for Al2 O3 , FeO* and Na2 O, with low Al2 O3 , Na2 O, high FeO* and high

Al2 O3 , Na2 O, low FeO* subgroups present (Fig. 4.3-1). These will be referred to here as

low- and high-Al subgroups, respectively. The majority of the ca. 3.5–3.42 Ga granites

belong to the high-Al group, with low-Al granites occurring in the Yule, Corunna Downs



376



Chapter 4.3: Geochemistry of Paleoarchean Granites of the East Pilbara Terrane



Fig. 4.3-1.



4.3-4. Geochemistry

377



Fig. 4.3-1. (Continued). Major element Harker diagrams for the ca. 3.5–3.42 Ga granites; SiO2 (anhydrous) versus Mg#, TiO2 , Al2 O3 , FeOtot,

MgO, CaO, Na2 O and K2 O. Mg# = 100 molecular (Mg/(Mg + total Fe). Low-Al units and granite complexes shown as ‘+’, ‘×’, and upright

grey triangles. Data from Davy (1988), Bickle et al. (1989, 1993), Collins (1983) and GA-GSWA unpublished. Unit divisions as given in

Table 4.3-1. Low-, medium- and high-K subdivisions after Gill (1981).



378



Chapter 4.3: Geochemistry of Paleoarchean Granites of the East Pilbara Terrane



and Mount Edgar Granitic Complexes (Table 4.3-1). Mg# (mostly 45-10) are moderate to

low (Fig. 4.3-1), as are Cr (mostly <50 ppm) and Ni (<40 ppm) contents, consistent with

other early Archean TTGs (e.g., Smithies, 2000). Mg# is largely independent of subgroup,

reflecting the variable MgO contents of these rocks.

Despite the variation in Na2 O, all granites are sodic with Na2 O/K2 O varying from

>1 to >4. K2 O shows no strong correlation with Na2 O, or with SiO2 , MgO, Mg#, etc.

(Fig. 4.3-1; see also Fig. 15 of Bickle et al., 1993). Granites vary from medium- to highK (for given SiO2 ) for both the low-Al and high-Al subgroups (Fig. 4.3-1). The majority

of the granites straddle the Ca-enriched boundary of the ‘trondhjemite’ field on the CaNa-K plot (Fig. 4.3-2) and there is evidence for limited K-enrichment within all granitic

complexes and both subgroups (Fig. 4.3-1, 4.3-2: see also Bickle et al., 1993). The behaviour of Rb mirrors that of K2 O (Fig. 4.3-3), varying over a considerable range of values

(from 40 to 150 ppm for both subgroups), but with no obvious correlation with SiO2 or

Mg#. Ba appears to behave differently to other LILEs, with a pronounced geographic

distribution in abundance. This is best shown by the almost universally elevated values

(500 to >1500 ppm) in the Shaw and Yule Granitic Complexes relative to the nearby

Mount Edgar and Corunna Downs Granitic Complexes (mostly <500 ppm; Fig. 4.3-3).

Sr broadly correlates with Na2 O and amplifies the subgroups evident in Al2 O3 , FeO* and

Na2 O (Fig. 4.3-3), with low values (mostly <350 ppm) in the low-Al group, and high values (to >800 ppm) in the high-Al group (Fig. 4.3-4(b)). Sr is negatively correlated with

SiO2 for both groups (Fig. 4.3-3).

The ca. 3.5–3.42 Ga granites are slightly to strongly LREE-enriched when normalised

to primitive-mantle (e.g., (La/Sm)N 2 varies from 1 to 10; (La/Yb)N from 7 to >100), but

vary from HREE- and Y-depleted, to Y-undepleted (e.g., (Gd/Yb)N ranges from 1.3–4.5

(Fig. 4.3-5(a–c)), with Yb from 0.2 to 2.5 ppm. HREE and Y decrease with increasing

SiO2 , following two distinct trends: a low HREE and Y (Y < 12 ppm, Yb < 1.2 ppm at

65% SiO2 ) trend is best seen in the granites of the Shaw Granitic Complex (see Bickle

et al., 1993); and an elevated HREE and Y (Y < 25 ppm, Yb < 2.5 ppm at 65% SiO2 )

trend that is most common in the Corunna Downs and Mount Edgar Granitic Complexes (Fig. 4.3-3). Samples with elevated (Gd/Yb)N or (Tb/Yb)N are restricted to low

Yb (<1.2 ppm) rocks (Fig. 4.3-5(a–c)). The Y and HREE trends largely reflect the highand low-Al subgroups (low and high-Y, respectively; Fig. 4.3-4), though a small subset of

samples from a number of high-Al units are characterised by elevated Y (up to 25 ppm;

Fig. 4.3-4(a)).

LREE broadly correlate with the HREE (Fig. 4.3-5(a–c)) and generally lie along either a

high or low LREE/HREE trend, for low- or high-Al subgroups, respectively. LREE/LREE

ratios (e.g., La/Sm) are variable but generally greater for the high-Al subgroup. Both positive and negative Eu anomalies are present. Small to moderate negative Eu anomalies

(Eu/Eu* from ∼1.0 to <0.4) characterise the low-Al subgroup, consistent with the low

Sr contents (<500 ppm). These are typical of low-Al TTGs, which are dominantly Yundepleted and Sr-depleted (e.g., Barker and Arth, 1976; Barker, 1979; Fig. 4.3-5(a–c)).

2 All normalised elements and ratios are normalised to Primitive Mantle using the values of Sun and McDonough (1989).



4.3-4. Geochemistry



Fig. 4.3-2. Ca-Na-K ternary diagrams for (A): ca. 3.5–3.42 Ga; and (B): ca. 3.32–3.25 Ga Pilbara granites. High-Al units and granite complexes

shown in blue, green and black, low-Al units and granite complexes in red and magenta. Data sources as for Figs. 4.3-1 and 4.3-7. Abbreviations:

TDH – trondhjemite, CA – calc-alkaline. Calc-alkaline and trondhjemite trends, and shaded trondhjemite field, from Martin (1994).



379



380



Chapter 4.3: Geochemistry of Paleoarchean Granites of the East Pilbara Terrane



Fig. 4.3-3.