Eoarchean ultra-depleted mantle domains inferred from ca. 3.81 Ga Anshan trondhjemitic gneisses, North China Craton

Precambrian Research 263 (2015) 88–107

Contents lists available at ScienceDirect

Precambrian Research journal homepage: www.elsevier.com/locate/precamres

Eoarchean ultra-depleted mantle domains inferred from ca. 3.81 Ga Anshan trondhjemitic gneisses, North China Craton Ya-Fei Wang a,b , Xian-Hua Li a,∗ , Wei Jin c , Jia-Hui Zhang c a b c

State Key Laboratory of Lithospheric Evolution, Institute of Geology and Geophysics, Chinese Academy of Sciences, Beijing 100029, China University of Chinese Academy of Sciences, Beijing 100049, China College of Earth Sciences, Jilin University, Changchun 130026, China

a r t i c l e

i n f o

Article history: Received 4 August 2014 Received in revised form 25 February 2015 Accepted 11 March 2015 Available online 21 March 2015 Keywords: Eoarchean Ultra-depleted mantle Zircon U–Pb–Hf–O isotopes Trondhjemitic gneisses Anshan North China Craton

a b s t r a c t Early mantle differentiation is yet to be well constrained because isotopic systems recorded in many ancient rocks are arguably susceptible to later disturbances. Although refractory zircon is of singular importance in deciphering early Earth history, a growing body of work on ancient zircons indicates that interpretations of their U–Pb ages and Hf–O isotopes are neither straightforward nor unique. Owing to these complexities, the question of whether the Anshan Complex of North China Craton (NCC) really preserves 3.8 Ga rocks remains controversial. To better understand these issues, we conducted a systematic in situ zircon U–Pb, Hf and O isotopic study guided by detailed field mapping (1:50 scale) and sampling at Guodishan, northern Anshan Complex. Three rock units were recognized and precisely dated: a ca. 3.13 Ga massive trondhjemitic gneiss that dominates the outcrop, a ca. 3.36–3.30 Ga migmatite complex that is intruded by the trondhjemitic gneiss, and a ca. 3.81 Ga massive to weakly-banded trondhjemitic gneiss that is enclosed by the migmatite complex. U abundances, U–Pb discordances, apparent 207 Pb/206 Pb ages and ␦18 O values of a single generation of magmatic zircons reveal that only low-U (generally U < 500 ppm), concordant zircons can preserve magmatic oxygen isotopes. High-U zircons generally tend towards lower ␦18 O values, possibly due to secondary alterations facilitated by strong metamictization. Concordant ca. 3.81 Ga oscillatory zoned zircons from the oldest unit give εHf (t) = −0.7 to 6.2 and ␦18 O = 5.2–7.0‰. These data combined with whole-rock geochemistry suggest that this Eoarchean trondhjemitic geniss was either formed by magma mixing or probably derived from a heterogeneous basaltic source at ca. 13–14 kbar, 950–1000 ◦ C. The highest initial εHf value of 6.2 at ca. 3.81 Ga exceeds those recorded by contemporaneous zircons worldwide and the modeled depleted mantle value. This provides new isotopic evidence for the existence of an Eoarchean ultra-depleted mantle domain underlying NCC. A literature survey reveals that this Eoarchean ultradepleted signature was not registered in younger zircons worldwide, suggesting that this ultra-depleted mantle signature may be transient, possibly erased by subsequent mantle re-homogenization processes. © 2015 Elsevier B.V. All rights reserved.

1. Introduction There is a general consensus that continental crust is a mantle extract and therefore, ancient crustal rocks can serve as direct archive of early mantle differentiation. Evidence for early mantle differentiation depends largely on isotopic studies (e.g. Sm–Nd and Lu–Hf systematics) of ancient rocks and zircons. Whole-rock Sm–Nd isotope systems of ancient rocks have been widely used to decipher early mantle differentiation, but some rocks were arguably susceptible to later geological disturbances, thus may not preserve primary signatures (e.g., Gruau et al., 1996; Kamber

∗ Corresponding author. Tel.: +86 10 82998512; fax: +86 10 62010846. E-mail address: [email protected] (X.-H. Li). http://dx.doi.org/10.1016/j.precamres.2015.03.005 0301-9268/© 2015 Elsevier B.V. All rights reserved.

et al., 2001; Moorbath et al., 1997; Vervoort and Blichert-Toft, 1999; Vervoort et al., 1996). As an alternative, zircons have been intensively studied to unravel early Earth processes (e.g., Amelin et al., 1999; Harrison et al., 2005). Among Hadean to Eoarchean zircons, most show chondritic to enriched Hf isotopes (e.g., Amelin et al., 2000; Blichert-Toft and Albarède, 2008; Harrison et al., 2008; Iizuka et al., 2009; Kemp et al., 2009, 2010; Nebel-Jacobsen et al., 2010; Vervoort and Blichert-Toft, 1999), which are of limited use for directly tracking mantle depletion history, and only rare zircons exhibit depleted Hf isotope signatures, mainly from West Greenland (e.g., Guitreau et al., 2012), Antarctica (Choi et al., 2006) and North China (e.g., Liu et al., 2008; Wu et al., 2008). It is noteworthy that the extremely positive εHf (t) values for Jack Hills detrital zircons are very rare, and these data have been debated (Bell et al., 2011; Blichert-Toft and Albarède, 2008; Harrison et al.,

Y.-F. Wang et al. / Precambrian Research 263 (2015) 88–107

2008; Kemp et al., 2010). Eoarchean metabasalts in West Greenland yielded highly variable depleted Hf isotopes (Hoffmann et al., 2010), together with zircon εHf (t) values up to 5.6 at 3.85 Ga in Eoarchean gneisses from the Napier Complex, Antarctica (Choi et al., 2006), indicating the possible existence of a highly depleted Archean mantle. These zircon analyses, however, were conducted by the solution MC-ICPMS technique on highly reversely-discordant single-grain zircons with complex age patterns and overgrowth rims, making their interpretations ambiguous. The major problem is whether the isotopic signatures obtained from ancient zircons truly represent primary magmatic signatures or result from secondary alteration, since the interpretation of isotopic compositions of ancient zircons is frequently clouded by: (1) the strong U–Pb discordant nature of zircons, which are susceptible to later disturbances, potentially obliterating their pristine isotopic compositions; (2) the complex age patterns even within a single zircon grain; (3) the difficulty of reliably associating Hf isotope values with the crystallization ages for structurally complex zircons, particularly when obtaining Hf isotopic data by large beam sizes laser-ablation or solution chemistry techniques (e.g., Guitreau et al., 2012; Kemp et al., 2010; Valley et al., 2006). These complexities have resulted in fundamental controversies about early Earth processes, from the crystallization ages of ancient continental rocks (e.g., Kamber et al., 2001) to Hadean tectonics (e.g., Harrison et al., 2005; Kemp et al., 2010). The Archean Anshan Complex (AAC) in the northeastern segment of North China Craton (NCC) (Fig. 1A), is considered as one of a few documented localities worldwide that may have preserved crustal remnants as old as 3.8 Ga (e.g., Liu et al., 1992; Nutman et al., 2001). Substantial amounts of zircon U–Pb and Hf–O isotopic data of AAC have been accumulated in the past two decades (e.g., Liu et al., 1992, 2008; Song et al., 1996; Wan et al., 2005, 2012, 2013; Wu et al., 1998, 2008; Zhang et al., 2013). Results show that zircons within previously reported oldest rocks have complex structures (e.g., recrystallization and overgrowth) and are characterized by the coexistence of a wide range of ages associated with radiogenic Pb loss to varying degrees even within a single sample. Geological interpretation of such ancient, complicated zircons is difficult, resulting in controversy over whether there are any ca. 3.8 Ga rocks existing in Anshan (Liu et al., 2008; Nutman et al., 2009; Wu et al., 2008, 2009). In light of these issues, we carried out a systematic in situ zircon U–Pb, O and Hf isotopic study and detailed field mapping and sampling on outcrops in Guodishan, northern AAC. The new results demonstrate that: (1) there exists a ca. 3.81 Ga trondhjemitic gneiss enclave overprinted by a ca. 3.36–3.30 Ga migmatization event. The younger zircon 207 Pb/206 Pb ages within this enclave are attributed to subsequent radiogenic Pb loss and to the injection of ca. 3.36 Ga leucosome veins; (2) only low-U, concordant zircons can best preserve primary ␦18 O isotope signatures, while high-U, strongly discordant zircons generally record lower ␦18 O values; and (3) pristine ca. 3.81 Ga zircons within this Eoarchean trondhjemitic gneiss recorded highly depleted εHf values, arguing for a significant mantle differentiation in the early Eoarchean–Hadean time. 2. Regional geology The AAC consists of several Archean rock suites, with a total exposed area over 80 km2 in the northeastern segment of NCC (Fig. 1B). These rock units include: (1) Eoarchean Baijiafen, Dongshan and Shengousi gneisses previously dated at ca. 3.78–3.81 Ga (Liu et al., 1992, 2008; Song et al., 1996; Wan et al., 2005; Zhang et al., 2013); (2) Paleoarchean gneissic trondhjemites and porphyritic to fine-grained granites dated at ca. 3.45–3.30 Ga with associated supracrustal rocks (Song et al., 1996; Wan et al., 2013); (3) Mesoarchean Lishan trondhjemitic gneiss and Tiejiashan granite dated at ca. 3.1 and 3.0 Ga, respectively (Wu et al., 1998); (4) ca.

89

2.45 Ga Qidashan alkali-feldspar granite that intrudes Neoarchean Anshan Group metasedimentary rocks (Wu et al., 1998). It is, however, noteworthy that crystallization ages for some of these rock suites, particularly of the Eoarchean and Paleoarchean rocks, are still a subject of considerable debate, because of the coexistence of several generations of zircons within a single sample (e.g., Liu et al., 2008; Nutman et al., 2009; Wu et al., 2008, 2009). In this paper, we document and report results for the newly-discovered North Guodishan Complex, another occurrence of very ancient rocks in the Anshan area (Fig. 1). 3. North Guodishan Complex 3.1. Field geology The Guodishan outcrop (41◦ 08 14 N, 123◦ 03 16 E-WGS-84 datum) in northern AAC has a boomerange shape and three major lithologic units are identified based on detailed field mapping (1:50 scale) (Fig. 1C). (1) A fine-grained massive trondhjemitic gneiss (Unit 1) dominates the outcrop (Fig. 2A). It is locally crosscut by granitic, pegmatitic and quartz veins several centimeters in width. (2) A migmatite complex (Unit 2), consisting of layered migmatite, with paired leucosome and melansome-biotite schist (Fig. 2B and C). This unit is highly deformed and shows phlebitic, stromatic and ptygmatic structures in the field. Leucosomes occur as thin straight and ptygmatic veins, or accumulate to form clots, which cut across the layered migmatite. Biotite-rich layers, however, are spatially closely related to leucosome, occurring either as dark bands of the layered migmatite and thus maybe restitic, or as small rounded blocks within the migmatites. (3) Massive to weakly-banded trondhjemitic gneiss (Unit 3) occurs as enclaves within Unit 2. This unit is bounded by the migmatite complex (Figs. 1C and 2D). Based on field relationships, the Unit 3 trondhjemitic gneiss is the oldest rock. However, it does contain minor leucosome veins of Unit 2 (Fig. 2E and F). The Unit 1 trondhjemitic gneiss intrudes the Unit 2, and is in turn cut by younger granitic veins. 3.2. Petrography Samples C209-4 and C209-6 from Unit 1 show hypidiomorphic textures, consisting of quartz (∼35%), plagioclase (∼50%), biotite (10%), microcline (<5%) and accessory apatite and zircon (Fig. 3A and B). Plagioclase forms subhedral to anhedral, polysynthetic twinned crystals that are partly or entirely sericitized. Quartz is present as anhedral, medium-grained aggregates. Small tabular biotites occur as interstitial crystals among major rock-forming minerals. Biotite was partly replaced by small prismatic clinozoisites. Unit 2 migmatite complex consists of layered migmatite (sample C209-1), coarse-grained trondhjemitic leucosome (sample C209-3) and biotite schist (sample C209-2). C209-3 is composed of subhedral plagioclase (50%) and anhedral quartz (40%), with minor biotite, apatite and zircon (Fig. 3C). C209-2 consists mainly of biotite (80%), quartz (15%) with accessory apatite, zircon. Biotite is partly replaced by epidote (Fig. 3D). C209-1 consists of alternating medium-grained light layers and fine-grained dark layers (Fig. 3E). The light layer consists of plagioclase (60%), quartz (35%) and minor biotite (5%) with accessory zircon, apatite and secondary muscovite; whereas the dark layer consists mainly of biotite (55%) and quartz (40%), and some secondary muscovite (<5%) and epidote. Unit 3 samples C209-8(1) and C209-8(3) consist of plagioclase (55%), quartz (30%), biotite (5–10%) and accessory minerals apatite, zircon (Fig. 3F). Subhedral polysynthetic twinned plagioclase grains are strongly sericitized. Anhedral quartz grains in size of 0.5 to 1.5 mm occur frequently as aggregates. Biotite occurs as small

90

Y.-F. Wang et al. / Precambrian Research 263 (2015) 88–107

Fig. 1. (A) Simplified geological map of Anshan area, showing the tectonic setting and (B) main rock assemblages of the Archean Anshan Complex (AAC) (modified after Wan et al., 2013; Zhang et al., 2013). (C) Lithologic relationships at Guodishan of AAC. The polygons labeled with cap alphabets correspond to field photographs in Fig. 2, which show field relationships of each units and sample locations. This outcrop is dominated by fine-grained massive trondhjemitic gneiss (Unit 1) and the migmatite complex (Unit 2), with a massive to weakly-banded trondhjemitic gneiss enclave (Unit 3) enclosed by Unit 2.

crystals commonly distributed along the edge of plagioclase or presents interstitially. 4. Analytical procedures

ICP-MS (Agilent 7500a), respectively, at the Institute of Geology and Geophysics, Chinese Academy of Sciences (IGGCAS), Beijing, China. Analyses of rock standards indicate precision and accuracy better than 5% for major elements and 10% for trace and rare earth elements.

4.1. Major and trace elements 4.2. Cathodoluminescence (CL) images Prior to analysis, all samples were cleaned from visible alteration rims and possible melt veins using a rock saw. Major and trace elements of Unit 3 samples were determined by XRF and

Zircons extracted from samples C209-1, C209-2, C209-3, C2094, C209-6, C209-8(1) and C209-8(3), together with standards

Y.-F. Wang et al. / Precambrian Research 263 (2015) 88–107

91

Fig. 2. Field photographs illustrating field relationships of each units and sample locations in Guodishan outcrop. (A) Photograph showing the intrusive relationship of Unit 1 into surrounding Unit 2 layered migmatite; (B) Occurrences and relationships between strongly deformed migmatite complex; (C) Field relations between the leucosome and layered migmatite and the location of samples C209-1 and C209-3, note the leucosome seemingly cuts the layered migmatite; (D) Photograph showing the field relationships between Unit 2 and Unit 3 trondhjemitic gneiss and sample locations: Unit 2 enclosed Unit 3 samples C209-8(1) and C209-8(3) (lower left); (E) Details of white dashed box of photo (D), and the location of Unit 3 samples C209-8(1) and C209-8(3), note the leucosome on the right side of C209-8(1); (F) Details of lower right part of photo (E), showing the relationships between leucosome and Unit 3 sample C209-8(1), note the clear intrusive relationship of Unit 2 leucosome into Unit 3 trondhjemitic gneiss. The pencils are ∼15 cm long, the red marker is 14 cm and the hammer is 29 cm long and the coins are ∼2 cm in diameter. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Pleˇsovice and Penglai were cast in epoxy resin discs and then polished. CL images were obtained using a Gatan Mono CL3+ equipped on a Quanta 200F Field-Emission Environmental Scanning Electron Microscope at Peking University, in order to examine internal structures and choose potential domains for analysis. The acceleration voltage and current were 15 kV and ∼1 nA, respectively. Representative CL images and their analytical results were presented in Appendix A (Figs. A.1–A.12). These mounts were cleaned using ultrapure water, then processed using alcohol ultrasonic cleaning to remove possible surface Pb contamination and then vacuum-coated with high-purity gold prior to secondary ion mass spectrometry (SIMS) analyses.

4.3. Zircon U–Pb dating Zircon U–Pb isotopic analyses were conducted using a Cameca IMS-1280 SIMS at IGGCAS, following procedures as described by Li et al. (2009). U–Pb isotopic ratios were determined relative to the Pleˇsovice (Sláma et al., 2008), analyses of which were interspersed with unknown grains. U abundances were determined relative to 91500 (Wiedenbeck et al., 1995). A long-term uncertainty of 1.5% (1SD) for 206 Pb/238 U measurements of the standard zircons was propagated to the unknowns (Li et al., 2010a). Common Pb was corrected using measured 204 Pb and Stacey and Kramers’s (1975) model for crustal common Pb isotopic composition was assumed.

92

Y.-F. Wang et al. / Precambrian Research 263 (2015) 88–107

Fig. 3. Photomicrographs showing texture characteristics of various rock units from Guodishan. (A) Unit 1 trondhjemitic gneiss C209-4, showing the constitute minerals: sericitezed and unaltered plagioclase and quartz and minor biotite; (B) Unit 1 trondhjemitic gneiss C209-6, displaying essential minerals: sericitezed and unaltered plagioclase and anhedral quartz with minor biotite; (C) Unit 2 leucosome C209-3 displaying subhedral to anhedral plagioclase and anhedral granular quartz. Note the relative large grain size and the secondary muscovite as well as the sericitized plagioclase; (D) Unit 2 quartz biotite schist C209-2, showing the flaky metablastic textures and essential minerals including biotite, quartz and the epidotization of biotite; (E) Unit 2 layered migmatite C209-1, showing the boundary between light colored layer and dark colored layer. Note the contrasting grain sizes and constituent minerals of these two layers: the light layer contains mainly of quartz and plagioclase and minor small biotite, while the dark layer contains biotite and quartz as essential minerals without plagioclase, (F) Massive trondhjemitic gneiss C209-8(3), consisting predominantly of subhedral altered plagioclase and anhedral quartz with minor biotite.

Data assessment was performed using the Isoplot 3.70 programs (Ludwig, 2008).

precision of a single analysis was generally 0.20–0.40‰ (2SE) for the 18 O/16 O ratio, while the reproducibility of Penglai zircon standard was better than 0.50‰ (2SD).

4.4. Zircon oxygen isotopes 4.5. Zircon Lu–Hf isotopes Prior to oxygen isotopic analysis, the mounts were grounded and repolished to remove pits of earlier U–Pb analyses. All O-isotope analytical spots overlap or within the same homogeneous domain in CL with U–Pb pits. Analytical procedures and instrument parameters follows the well-established method of Li et al. (2010b). The instrumental mass fractionation factor was corrected using Penglai zircon standard with ␦18 O of 5.31‰ (Li et al., 2010c). The internal

Zircon Lu–Hf isotopes were performed at IGGCAS in two sessions, using a Neptune MC-ICPMS equipped with a 193 nm ArF excimer laser ablation system. The analytical procedures have been described in detail by Wu et al. (2006). Spot size of ∼60 ␮m was applied to most analyses, while ca. 44 ␮m was applied for zircons of C209-2, owing to their smaller size. The ablation time is 26 s

Table 1 Summary of zircon U–Pb, ␦18 O and εHf (t) results of samples from Guodishan outcrop (41◦ 08 14 N, 123◦ 03 16 E). Sample spot

U (ppm)

Th (ppm)

207 Pbb /

1␴ (%)

206 Pbb

C209-1 layered migmatite @34-1 309 @34-2 4923 @43 318

206 Pbb /

1␴ (%)

238 U

207 Pb/206 Pb

1␴

Disc. (%)c

␦18 OVSMOW (‰)

2␴m

Measured 176 Lu/177 Hf

176 Hf/177 Hf

age (Ma)

Measured

2␴m

εHf (t)d

2␴e

0.3767 0.0846 0.3670

0.38 10.13 0.31

0.785 0.047 0.769

1.52 1.54 1.50

3819 1306 3779

6 185 5

2 78 3

6.3 2.4 6.5

0.3 0.3 0.4

0.001018 0.001114 0.000481

0.280469 0.280940 0.280402

0.000019 0.000048 0.000018

2.7 −36.9 0.8

0.7 1.7 0.7

C209-2 biotite schist @3f 463 @4 431 @7 665 @17 375

150 112 213 99

0.2712 0.2685 0.2689 0.2707

0.76 0.46 0.50 0.47

0.645 0.623 0.618 0.639

1.64 1.57 1.51 1.51

3313 3297 3300 3310

12 7 8 7

3 5 6 4

6.1 6.3 5.5 6.2

0.3 0.2 0.3 0.3

0.001041 0.000796 0.000950 0.000722

0.280897 0.280870 0.280876 0.280831

0.000024 0.000020 0.000022 0.000021

6.3 5.5 5.5 4.6

0.8 0.7 0.8 0.7

C209-3 leucosome @1 99 @10 96 @13 203 @14 264 @16 286 @17 259 @21 161 @22 102 @26 347 @27 576 @30 507 @32 417 @34 78

73 54 151 197 168 161 93 107 387 303 311 279 62

0.2808 0.2771 0.2793 0.2778 0.2774 0.2788 0.2800 0.2771 0.2770 0.2787 0.2788 0.2803 0.2773

0.38 0.63 0.44 0.23 0.33 0.23 0.31 0.54 0.47 0.35 0.39 0.37 0.99

0.651 0.687 0.685 0.631 0.646 0.669 0.674 0.685 0.662 0.682 0.643 0.678 0.660

1.50 1.51 1.52 1.77 1.52 1.50 1.51 1.51 1.52 1.51 1.50 1.51 1.52

3367 3347 3359 3350 3348 3356 3363 3347 3346 3356 3356 3365 3348

6 10 7 4 5 4 5 8 7 5 6 6 15

4 −1 0 6 4 2 1 −1 2 0 5 1 2

5.9 5.6 5.8 5.9 5.3 6.2 6.1 5.9 6.3 5.7 6.0 6.2 6.2

0.4 0.3 0.2 0.4 0.3 0.2 0.2 0.3 0.3 0.2 0.3 0.4 0.2

0.000838 0.000919 0.001023 0.002725 0.001037 0.001037 0.000829 0.000835 0.001827 0.001282 0.001059 0.001055 0.000787

0.280626 0.280578 0.280600 0.280723 0.280627 0.280597 0.280591 0.280568 0.280629 0.280590 0.280610 0.280613 0.280625

0.000016 0.000017 0.000015 0.000018 0.000014 0.000015 0.000014 0.000013 0.000015 0.000015 0.000013 0.000013 0.000014

−1.6 −4.0 −3.2 −2.9 −2.5 −3.4 −3.0 −4.2 −4.3 −4.2 −3.0 −2.7 −2.0

0.6 0.6 0.5 0.6 0.5 0.5 0.5 0.5 0.6 0.5 0.5 0.5 0.5

C209-4 massive trondhjemitic gneiss @2 197 101 @8 116 46 @9 127 86 @12 139 78

0.2412 0.2432 0.2436 0.2409

0.24 0.99 0.40 0.38

0.603 0.554 0.620 0.637

1.51 1.50 1.58 1.50

3128 3141 3144 3126

4 16 6 6

3 9 1 -2

6.2 5.6 5.9 6.2

0.3 0.4 0.3 0.4

– – – –

– – – –

– – – –

C209-6 massive trondhjemitic gneiss @1 291 266 @4 161 83 @6-1 560 84 @6-2 878 815 @8 364 190 @17 263 190

0.2447 0.2437 0.2709 0.2155 0.2415 0.2408

0.40 0.78 0.54 0.43 0.37 0.41

0.605 0.614 0.659 0.386 0.595 0.614

1.56 1.50 1.56 1.55 1.53 1.51

3151 3144 3311 2947 3130 3125

6 12 8 7 6 7

3 2 1 29 4 1

6.5 6.1 5.2 4.9 5.8 5.6

0.3 0.3 0.2 0.4 0.3 0.3

0.000254 0.000240 0.001493 0.001444 0.000354 0.000450

0.280771 0.280813 0.280809 0.280808 0.280809 0.280855

0.000025 0.000030 0.000020 0.000021 0.000022 0.000027

−0.2 1.2 2.1 −6.0 0.5 1.8

0.9 1.1 0.7 0.8 0.8 1.0

0.33 3.06 0.41 3.07 0.34 3.06 0.31 0.36 3.07 0.56 3.09 0.63 3.08 0.58 0.47

0.653 0.798 0.801 0.836 0.497 0.599 0.759 0.785 0.802 0.740 0.793 0.469 0.454 0.802 0.571

1.51 1.26 1.53 1.30 1.51 1.25 1.51 1.58 1.28 1.59 1.35 1.50 1.24 1.53 1.54

3706 3817 3794 3829 3368 3681 3677 3811 3780 3802 3829 3431 3540 3801 3530

5 45 6 46 5 46 5 5 46 8 46 10 47 9 7

13 1 0 −2 23 18 1 2 0 6 2 28 32 0 17

5.2 – 6.2 – 5.2 – 6.1 6.5 – 6.5 – 5.3 – 6.6 6.4

0.3 – 0.3 – 0.3 – 0.3 0.3 – 0.3 – 0.2 – 0.2 0.2

0.001086 0.001773 0.000993 0.001207 0.000408 0.000688 0.000761 0.001283 0.001021 0.000667 0.000831 0.001042 0.000891 0.000930 0.000572

0.280566 0.280560 0.280376 0.280459 0.280389 0.280441 0.280434 0.280425 0.280467 0.280543 0.280410 0.280403 0.280439 0.280553 0.280444

0.000027 0.000038 0.000028 0.000028 0.000025 0.000025 0.000026 0.000027 0.000034 0.000027 0.000041 0.000031 0.000036 0.000033 0.000023

3.4 3.9 −1.1 2.1 −9.1 −0.6 −1.2 0.2 1.7 5.9 1.3 −8.6 −4.5 5.5 −3.7

1.0 1.4 1.0 1.0 0.9 0.9 0.9 1.0 1.2 1.0 1.5 1.1 1.3 1.2 0.8

C209-8(1) massive to weakly banded trondhjemitic gneiss @1 718 367 0.3498 @1g 383 193 0.3762 @5 349 145 0.3705 @5 192 69 0.3793 @6 1020 76 0.2809 @6 436 111 0.3440 @7 692 429 0.3433 @10 448 231 0.3748 @10 434 258 0.3673 @12 372 199 0.3726 @12 128 50 0.3794 @20 1000 158 0.2925 @20 525 238 0.3139 @22 340 201 0.3723 @23 516 116 0.3119

– – – –

– – – –

Y.-F. Wang et al. / Precambrian Research 263 (2015) 88–107

174 1187 80

93

94

Table 1 (Continued) Sample spot

U (ppm)

Th (ppm)

207 Pbb /

1␴ (%)

206 Pbb

483 541 233 121 330 282 240 129 227 116 190 236 501 174 117 188 333 232 136 246 529 449 334 291 302 730 180 377 275 380 291 548 354 347 237

50 255 123 36 77 110 96 73 92 47 56 125 230 90 48 69 53 132 53 92 227 238 214 135 130 168 69 255 89 237 173 395 119 184 141

C209-8(3) massive trondhjemitic gneiss @1 324 136 154 @1g 960 914 379 @7 @7 199 336 745 531 @9 @10 1266 63 196 @10 364 802 502 @12 @12 583 262 @16 887 650 @16 652 86 a b c d e f g

1␴ (%)

238 U

207 Pb/206 Pb

1␴

Disc. (%)c

␦18 OVSMOW (‰)

2␴m

Measured 176 Lu/177 Hf

176 Hf/177 Hf

age (Ma)

Measured

2␴m

εHf (t)d

2␴e

0.3080 0.3732 0.3757 0.3687 0.3641 0.3681 0.3466 0.3710 0.3761 0.3714 0.3737 0.3678 0.3395 0.3727 0.3599 0.3758 0.3317 0.3740 0.3774 0.3659 0.3760 0.3704 0.3749 0.3661 0.3753 0.3264 0.3765 0.3764 0.3603 0.3667 0.3685 0.3752 0.3491 0.3365 0.3230

3.08 0.33 0.58 0.26 0.15 0.31 0.26 0.62 0.25 0.47 0.37 2.98 0.28 3.03 0.67 0.28 2.99 0.27 0.33 0.40 0.21 3.01 0.45 2.98 0.24 2.97 0.48 0.19 2.99 0.20 2.98 0.22 0.21 0.22 2.99

0.616 0.798 0.821 0.709 0.717 0.762 0.738 0.761 0.813 0.781 0.810 0.802 0.575 0.812 0.787 0.779 0.735 0.820 0.803 0.612 0.797 0.612 0.804 0.805 0.804 0.561 0.808 0.783 0.778 0.735 0.830 0.822 0.784 0.738 0.755

1.26 1.52 1.58 1.50 1.51 1.50 1.50 1.50 1.51 1.58 1.50 1.24 1.50 1.29 2.50 1.51 1.23 1.55 1.51 1.50 1.53 1.23 1.84 1.25 1.61 1.21 1.56 1.51 1.27 1.53 1.23 1.51 1.50 1.52 1.25

3511 3804 3814 3786 3767 3783 3692 3795 3816 3797 3807 3783 3661 3802 3749 3815 3625 3808 3821 3775 3816 3793 3811 3775 3813 3600 3818 3817 3751 3778 3785 3813 3703 3647 3584

47 5 9 4 2 5 4 9 4 7 6 44 4 45 10 4 45 4 5 6 3 45 7 45 4 45 7 3 45 3 44 3 3 3 45

12 1 −1 9 7 4 4 4 0 2 0 0 20 −1 0 3 2 −1 1 18 1 19 0 −1 0 20 0 2 1 6 −3 −1 −1 2 −1

– 6.3 6.6 5.7 5.3 5.2 6.6 6.2 6.4 7.0 6.1 – 5.4 – 6.2 6.5 – 5.9 – – 6.5 – 6.3 – 6.1 – 6.3 6.2 – 5.9 – 5.3 6.8 5.5 –

– 0.3 0.3 0.3 0.3 0.3 0.2 0.3 0.2 0.3 0.3 – 0.3 – 0.1 0.2 – 0.2 – – 0.3 – 0.4 – 0.3 – 0.2 0.3 – 0.3 – 0.3 0.3 0.3 –

0.000378 0.000896 0.001508 0.000724 0.001032 0.000937 0.000993 0.000910 0.001024 0.000859 0.000978 0.001420 0.000960 0.000969 0.001190 0.001052 0.000422 0.000913 – – 0.000921 0.000901 0.000623 0.000540 0.001090 0.000992 0.000910 0.001275 0.000939 0.001336 0.001150 0.001138 0.000900 0.001157 0.000787

0.280436 0.280375 0.280533 0.280552 0.280404 0.280403 0.280598 0.280407 0.280396 0.280471 0.280522 0.280541 0.280598 0.280433 0.280465 0.280393 0.280379 0.280409 – – 0.280562 0.280541 0.280449 0.280463 0.280403 0.280554 0.280463 0.280451 0.280460 0.280487 0.280517 0.280407 0.280471 0.280504 0.280549

0.000032 0.000030 0.000033 0.000031 0.000027 0.000024 0.000029 0.000024 0.000025 0.000019 0.000021 0.000031 0.000021 0.000028 0.000027 0.000029 0.000034 0.000025 – – 0.000020 0.000046 0.000034 0.000029 0.000025 0.000028 0.000024 0.000023 0.000024 0.000023 0.000026 0.000024 0.000021 0.000025 0.000023

−4.0 −0.7 3.6 5.6 −0.9 −0.3 4.4 0.2 0.0 2.7 4.4 4.0 3.8 1.1 0.5 −0.2 −3.5 0.6 – – 6.2 5.0 2.8 2.7 0.0 0.8 2.7 1.3 1.0 1.6 3.3 0.0 0.4 0.8 0.7

1.2 1.1 1.2 1.1 1.0 0.8 1.0 0.9 0.9 0.7 0.7 1.1 0.8 1.0 1.0 1.0 1.2 0.9 – – 0.7 1.6 1.2 1.1 0.9 1.0 0.9 0.8 0.9 0.8 0.9 0.9 0.8 0.9 0.8

0.3764 0.3058 0.3177 0.3511 0.3658 0.2545 0.3556 0.3326 0.3218 0.3461 0.2646

0.64 3.12 0.33 3.08 0.29 0.38 3.08 0.33 3.08 0.30 3.09

0.808 0.300 0.629 0.682 0.774 0.425 0.655 0.591 0.532 0.627 0.335

1.53 1.26 1.54 1.27 1.51 1.51 1.26 1.50 1.25 1.51 1.24

3817 3500 3559 3712 3774 3213 3731 3629 3579 3690 3274

10 47 5 46 4 6 46 5 47 5 48

0 52 12 10 2 29 13 17 23 15 43

6.4 – 5.8 – 6.4 4.5 – 5.6 – 5.5 –

0.2 – 0.2 – 0.2 0.3 – 0.3 – 0.3 –

0.001449 0.001337 0.000903 0.001091 0.001149 0.001027 0.001357 0.001237 0.001410 0.000853 0.000798

0.280486 0.280753 0.280519 0.280541 0.280540 0.280620 0.280463 0.280548 0.280550 0.280532 0.280555

0.000040 0.000029 0.000024 0.000041 0.000024 0.000024 0.000039 0.000023 0.000042 0.000030 0.000032

2.1 4.7 −1.2 2.6 3.9 −5.8 −0.4 0.6 −0.9 2.4 −6.2

1.4 1.0 0.8 1.5 0.8 0.9 1.4 0.8 1.5 1.1 1.1

f206 (%): fraction of 206 Pb that is unradiogenic. Corrected using measured 204 Pb. Disc.(%) = 100 × (1 − (206 Pb/238 U age)/(207 Pb/206 Pb age)). εHf (t) were calculated using the 207 Pb/206 Pb ages, chondritic parameters of Blichert-Toft and Albarède (1997) and 176 Lu of 1.867 × 10−11 y−1 (Söderlund et al., 2004). Errors in εHf (t) are propagated to include uncertainties in 176 Lu/177 Hf, 176 Hf/177 Hf and age. Analysis used for weighted mean or mean calculations. Italics are concurrent Pb–Hf analyses by laser ablation.

Y.-F. Wang et al. / Precambrian Research 263 (2015) 88–107

@23 @25 @27 @30 @34-1 @34-2 @35 @37-1 @37-2 @39-1 @39-2 @39-2 @41 @41 @46 @49 @49 @50-1 @50-2 @50-3 @52 @52 @53 @53 @55 @55 @60-1 @62 @62 @63 @63 @64 @65 @69 @69

206 Pbb /

Y.-F. Wang et al. / Precambrian Research 263 (2015) 88–107

95

Fig. 4. Concordia diagrams and representative zircon CL images. All data-point error ellipses are 2␴ and the all scale bars are 50 ␮m. (A) Unit 1 sample C209-4: the upper intercept is calculated from 12 grains, among which 4 most concordant analyses (shown as red ellipses) yield weighted average 207 Pb/206 Pb age of 3131 ± 11 Ma. The older xenocrysts are shown in dashed ellipses. (B) Unit 1 sample C209-6: the upper intercept age is calculated from all but two old xenocryst analyses 6-1, 6-2. Four nearlyconcordant analyses (shown in red ellipses) yield a weighted average 207 Pb/206 Pb of 3136 ± 20 Ma. (C) Unit 2 leucosome C209-3: note the cluster of ca. 3.36 Ga zircons and one slightly younger ca. 3.30 Ga zircon (spot 12) shown in the inset. The old xenocrysts are shown as dashed ellipses. (D) Unit 2 biotite schist C209-2: 17 grains yield a weighted average 207 Pb/206 Pb age of 3304 ± 4 Ma; (E) Unit 2 layered migmatite C209-1 and some paired analyses and their CL images; (F) Unit 3 sample C209-8(1): nearly-concordant 3.8 Ga zircons and three 3.31–3.37 Ga zircons are shown in red ellipses along with their representative CL images; (G) Unit 3 sample C209-8(3); (H) all <10% discordant analyses of Unit 3 trondhjemitic gneiss: the red ellipses are 16 concordant analyses from 15 grains yielding a weighted mean 207 Pb/206 Pb age of 3814 ± 2 Ma. Four concordant 3.31–3.37 Ga grains are shown as red ellipses, and the remaining analyses are likely affected by ancient and more recent Pb loss. See text for explanation.

Y.-F. Wang et al. / Precambrian Research 263 (2015) 88–107



εHf (0) = 10, 000 ×

εHf (t) = 10, 000





×

fLu/Hf =





176



176 Hf/177 Hf

s

−1

,

Unit 1

disc.≤5%

disc.>5%

Unit 3

disc.≤5%

disc.>5%

(A)

7

δ18O (‰)

6

5

Mantle zircon δ18O = 5.3 ± 0.6 (2 s.d.)

4

3

2

1 1.5 8

7

2.0

Unit 2

2.5

3.0

C209−1:

disc.≤5%

C209−2:

disc.≤5%

disc.>5%

C209−3:

disc.≤5%

disc.>5%

3.5

(B)

disc.>5%

6

Mantle zircon δ18O = 5.3 ± 0.6 (2 s.d.)

5

4

3

2

1 1.0 10 8

1.5

2.0

2.5

3.0

Unit 1

3.13 Ga zircons

Unit 2

leucosome 3.36 Ga zircons Other leucosome zircons

biotite schist zircons layered migmatite zircons

Unit 3

Unit 3, 3.81 Ga zircons

Other Unit 3 zircons

3.5

4.0

Other Unit 1 zircons

(C) 176

Lu/ 177 Hf =

6

0.04

6

MORB-D M 176 Lu/ 177Hf = 0.039

4 2 0 -2

CHUR reference line

-4

ust ic cr .022 maf =0 177 Hf / 176 L u

-6

CHUR,o

4.0

-8 3.0

3.1

3.2

3.3

3.4

3.5

3.6

3.7

3.8

3.9

4.0

Apparent 207Pb/206Pb age (Ga)

176



Hf/177 Hf



176 Hf/177 Hf

176

Hf/177 Hf





8

δ18O (‰)

for each measurement at a repetition rate of 8 Hz at 10 J cm−2 . The interference of 176 Yb and 176 Lu on 176 Hf could significantly affect the accuracy of obtained 176 Hf/177 Hf ratios. For 176 Yb interference, ␤(Yb) was calculated by normalizing 173 Yb/172 Yb ratio to 0.73925, 176 Yb interference was calculated by applying the average ␤(Yb) value of individual spot and assuming 176 Yb/172 Yb = 0.5887 (Wu et al., 2006). The interference of 176 Lu on 176 Hf was corrected by measuring the intensity of interference-free 175 Lu isotope, using a recommended 176 Lu/175 Lu ratio of 0.02655 (Machado and Simonetti, 2001) and assuming mass basis factor ␤(Lu) equals to the average ␤(Yb) value of individual spot. The isobaric interference of 176 Lu on 176 Hf is minor since the measured 176 Lu/177 Hf ratios for our zircons are normally lower than 0.0018. Measured 176 Hf/177 Hf ratios were normalized to 179 Hf/177 Hf = 0.7325. During these two sessions, the obtained 176 Hf/177 Hf ratios of two standard zircon were 0.282507 ± 0.000024 (2SD, n = 89), 0.282502 ± 0.000031 (2SD, n = 66) for Mud Tank and 0.282021 ± 0.000029 (2SD, n = 55), 0.282017 ± 0.000039 (2SD, n = 46) for GJ-1, respectively. These values are identical to the solution MC-ICPMS values of 0.282507 ± 0.000006 (2SD) for Mud Tank and 0.282000 ± 0.000005 (2SD) for GJ-1, respectively (Woodhead and Hergt, 2005; Morel et al., 2008). We also analyzed our in-house Qinghu standard and TEMORA standard before and after measurement of unknown zircons. A total of 133 analyses on Qinghu zircons yielded variable high 176 Yb/177 Hf ratios of 0.0092–0.1145, and consistent 176 Hf/177 Hf ratios between 0.282966 and 0.283048, with a mean of 0.283001 ± 0.000039, identical within error to its recommended value of 0.283002 ± 0.000004 (Li et al., 2013). Twenty analyses of TEMORA zircon yielded 176 Yb/177 Hf between 0.015379 and 0.061064, and 176 Hf/177 Hf of 0.282665 to 0.282726, with an average of 0.282696 ± 0.000036, consistent within errors with its recommended value of 0.282686 ± 0.000007 (Woodhead et al., 2004). To evaluate the age of the volumetrically larger domain sampled by laser drilling against SIMS age, we conducted 21 simultaneous U–Pb–Lu–Hf isotope analyses on 21 Unit 3 zircons previously analyzed by sequential SIMS and LA-MC-ICPMS approach. These simultaneous isotopic analyses were carried out at Northwest University using an Elan 6100 DRC Q-ICPMS and a Nu Plasma HR MC-ICPMS connected to a single GeoLas 2005 excimer ArF laserablation system, following the analytical procedures described by Yuan et al. (2008). Spot size was ca. 44 ␮m and the repetition rate was 6 Hz. The notations for εHf and fLu/Hf are defined as:

εHf(t)

96

s

CHUR,o

− −

176 

  

176 Lu/177 Hf



CHUR





Lu/177 Hf × et − 1

 −1

× et − 1

,



Lu/177 Hf



176 Lu/177 Hf

s

− 1,

CHUR

where (176 Lu/177 Hf)s and (176 Hf/177 Hf)s are the measured ratios, t is the crystallization age (here we use the apparent 207 Pb/206 Pb age) and  is the decay constant of 176 Lu. 5. Results Full analytical results are presented in Appendixes B–D. and a summary is shown in Table 1. Results of zircon U–Pb analyses are plotted on Fig. 4A–F and described as concordant (disc. ≤2%), nearly

Fig. 5. Plots of zircon ␦18 O values against measured 207 Pb/206 Pb ages for (A) Unit 1 and Unit 3 samples C209-4, C209-6, C209-8(1) and C209-8(3) and (B) Unit 2 migmatite samples (C209-1, C209-2 and C209-3). Note that the disc. > 5% analyses yield rather scattered ␦18 O values. (C) εHf (t) versus zircon 207 Pb/206 Pb ages of <10% discordant zircons in this study along with three putative reservoirs differentiated at 4.5 Ga. Solid symbols are analyses used for weighted mean calculations of each unit. The model depleted mantle evolution line (MORB-DM) was derived by assuming present-day depleted mantle with εHf of 18 (Vervoort and Blichert-Toft, 1999) and taking εHf = 0 at 4.5 Ga.

concordant (2% 10%). Oxygen isotopic compositions of trondhjemitic gneisses (Unit 1 and Unit 3) and migmatite complex (Unit 2) are presented in Fig. 5A and B, respectively. Fig. 5C shows initial Hf isotopic compositions against 207 Pb/206 Pb ages for all <10% discordant zircons in this study. All εHf (t) were calculated using CHUR parameters from Blichert-Toft and Albarède (1997), 176 Lu of 1.867 × 10−11 y−1 from Söderlund et al. (2004) and

Y.-F. Wang et al. / Precambrian Research 263 (2015) 88–107

t is apparent 207 Pb/206 Pb ages. Data are reported at 2 confidence interval (c.l.) unless otherwise stated. 5.1. Unit 1 (fine-grained massive trondhjemitic gneiss) 5.1.1. Sample C209-4 Zircons are euhedral, prismatic, and 30 to 150 ␮m long, with aspect ratios of 1:1–6:1. Cracks, inclusions and turbid domains, as well as clear pristine domains, can be clearly seen in transmitted light microphotographs. Under CL, most zircons show oscillatory zoning (Fig. A1), but some also display irregular dark-patches. Eighteen spots were analyzed on 17 zircons. Twelve analyses define an upper intercept of 3151 ± 23 Ma, consistent with the weighted average 207 Pb/206 Pb age of 3131 ± 6 Ma of four concordant to slightly discordant analyses (Fig. 4A). The remaining six analyses on five grains yield older 207 Pb/206 Pb ages (>3.2 Ga); possibly they are xenocrysts. Oxygen isotopes of eighteen zircons were measured. All zircons yield ␦18 O ranging from 4.1‰ to 6.5‰ (Fig. 5A). ␦18 O values of four concordant to slightly discordant zircons range from 5.6‰ to 6.2‰, averaging 6.0 ± 0.6‰. 5.1.2. Sample C209-6 Zircons of this sample are euhedral, mostly short prismatic to stubby. They range in length from 30 to 150 ␮m, with aspect ratios of 1:1 to 4:1. Under CL, most zircons display well-developed concentric oscillatory zoning (Fig. A2). Twenty-three spots on 22 grains were analyzed, including 2 analyses (6–1 and 6–2) conducted on the same oscillatory zone of a single grain. Excluding analyses 6–1 and 6–2, the remaining 20 analyses yield an upper intercept of 3156 ± 36 Ma (Fig. 4B), indistinguishable within error from an average 207 Pb/206 Pb age of 3136 ± 20 Ma from four concordant to nearly-concordant zircons. Oxygen isotopes were measured for all dated zircons. Values of ␦18 O range from 1.8 to 6.7‰ (Fig. 5A). The lowest ␦18 O of 1.8‰ comes from spot analysis 23, which gives rather high U and Th concentrations. Concordant to slightly discordant zircons of C209-6 have a limited range of εHf (t) value between −1.5 and 1.8 (Table 1; Fig. 5C; Appendix C), while the remaining highly discordant zircons range from −20.2 to 2.4. Spots 6–1 and 6–2 have identical measured 176 Hf/177 Hf ratios. In summary, the average 207 Pb/206 Pb age of 3133 ± 8 Ma of eight least discordant (disc. < 10%) analyses suggest that the Unit 1 finegrained massive trondhjemitic gneiss crystallized at ca. 3.13 Ga. A few older zircons dated at 3.2–3.3 Ga may be xenocrysts from older rocks. 5.2. Unit 2 (Migmatite complex) 5.2.1. Leucosome (sample C209-3) Zircons are euhedral, 50 to 100 ␮m in width with aspect ratios of 2:1 to 4:1. Most zircons contain spotty, turbid domains in transmitted light microphotographs. CL images of most zircons display oscillatory zoning, and some show core-mantle-rim textures. There are some dark domains in zircon CL images, corresponding to translucent to opaque domains in transmitted light microphotographs (Fig. A3-A4). Thirty-eight U–Pb analyses were made on 37 grains, with spots 2–1 and 2–2 on core and rim, respectively, of a single zircon grain. Most analyses are discordant (Fig. 4C). Among them, thirteen concordant to nearly concordant analyses yield weighted mean 207 Pb/206 Pb age of 3356 ± 3 Ma, interpreted as the crystallization age of this leucosome sample. Spot 12 yields a slightly younger, concordant age of 3296 ± 12 Ma. The remaining seven analyses give significantly older 207 Pb/206 Pb ages

97

(3558–3865 Ma) than the main population and therefore could be xenocrysts. Thirteen ca. 3.36 Ga zircons yield ␦18 O varying from 5.3‰ to 6.3‰, averaged at 5.9 ± 0.5‰. Their εHf (t) values vary from −4.3 to −1.6, averaged at −3.2 ± 1.7. Spot 12 (ca. 3296 Ma) give ␦18 O of 5.4‰ and εHf (t) of 0.9. Eight analyses on seven xenocrysts give ␦18 O ranging from 5.4‰ to 6.8‰ (Fig. 5B), and εHf (t) from −4.8 to 3.4. The remaining analyses on highly-discordant zircons yield highly variable ␦18 O and εHf (t) of 1.9–7.2‰ and −20.2 to 1.6, respectively. 5.2.2. Biotite schist (sample C209-2) Zircon grains are well rounded and oval in shape, 20 to 60 ␮m in length, with aspect ratios of 1:1 to 2:1. Under CL, most zircons contain dark to faintly-zoned cores, surrounded by white rims, which are too narrow to analyze (Fig. A5). All analyses, therefore, are conducted on the cores. Nineteen analyses on 19 grains define a discordia line with an upper intercept of 3304 ± 9 Ma, consistent with the weighted mean 207 Pb/206 Pb age of 3304 ± 4 Ma (Fig. 4D), which is interpreted to be the metamorphic age of this sample. Apart from one analysis (spot 10) that gives relatively low ␦18 O of 4.7‰, the others have ␦18 O varying from 5.5‰ to 6.7‰, averaged at 6.1 ± 0.5‰ (Fig. 5B). Due to small grain size, Hf isotopic compositions were analyzed only on 4 zircons, yielding εHf (t) values of 4.6 to 6.3, with an average of 5.4 ± 1.4. 5.2.3. Layered migmatite (sample C209-1) Most zircons are euhedral, short- to long-prismatic in shape, and 100 to 450 ␮m in length, with aspect ratios of 3:2 to 4:1. Many are fractured under transmitted light. Under CL, they show oscillatory zoning dispersed with dark domains occurring as either altered remnants of oscillatory zoning or irregular small patches (Fig. A6). Thirty-three spots on 25 grains were analyzed, including eight grains with each having 2 spots. Zircons from this sample are variably discordant, with apparent 207 Pb/206 Pb ages between ca. 3.8 Ga and ca. 1.7 Ga. Four concordant to nearly concordant analyses give similar 207 Pb/206 Pb ages of ca. 3.8 Ga (Fig. 4E). Within a single zircon, low-U domains generally yield much higher 207 Pb/206 Pb ages than those high-U domains. Twenty-eight oxygen isotope measurements were collected and the analysis spots overlap or within the same CL domain of the U–Pb spots. Four ca. 3.8 Ga zircons have ␦18 O ranging from 6.3‰ to 7.0‰ and εHf (t) from −0.4 to 2.7. The remaining analyses on discordant zircons range in ␦18 O from 1.7‰ to 6.4‰ and εHf (t) from −45.5 to 1.5. It is noted that strongly discordant zircons/zircon domains commonly have relatively lower ␦18 O values (Fig. 5A). 5.3. Unit 3 (massive to weakly-banded trondhjemitic gneiss) 5.3.1. Massive to weakly-banded trondhjemitic gneiss (sample C209-8(1)) Most grains are euhedral, prismatic, 100 to 400 ␮m in length, with aspect ratios of 3:2 to 5:1. In transmitted light, most grains are clear and cracked, and some contain dispersed opaque domains. Under CL, zircons show weak luminescence but have well-developed oscillatory zonings (Figs. A7–A10). Seventy-eight U–Pb analyses were conducted on 69 grains, including two grains on which three analyses were carried out and five grains on which two analyses were carried out. Twenty-seven of 78 analyses are highly discordant yielding 207 Pb/206 Pb ages between 1.87 and 3.77 Ga. The remaining 51 analyses can be divided into two groups: one group contains three concordant analyses (spots 16, 24, 26) yielding 207 Pb/206 Pb ages between 3.31 and 3.37 Ga, the remaining 48 analyses (except for spot 44) form another group yielding 207 Pb/206 Pb ages between 3.65 and 3.83 Ga, with the majority clustering around ca. 3.8 Ga (Fig. 4F).

98

Y.-F. Wang et al. / Precambrian Research 263 (2015) 88–107

Table 2 Whole rock geochemistry of Unit 3 and Nûk trondhjemite no. GGU221121. C209-8(1) Major elements (wt %) 72.38 SiO2 TiO2 0.24 Al2 O3 15.20 2.28 Fe2 O3 * – FeO 0.04 MnO 0.59 MgO CaO 2.51 Na2 O 5.07 K2 O 1.26 0.10 P2 O5 0.84 LOI 100.51 Total b 34 Mg# 1.07 A/CNKc 759 TZr (◦ C)d 928 TAp (◦ C)e Trace elements (g/g) Sc 1 19 V 3 Cr 4.0 Co 16.7 Ga Rb 131 373 Sr Y 6.6 Zr 116 Nb 3.40 Cs 5.61 Ba 242 La 17.9 Ce 21.1 Pr 2.99 10.3 Nd 1.94 Sm Eu 0.51 Gd 1.56 Tb 0.22 1.17 Dy Ho 0.22 Er 0.56 Tm 0.08 Yb 0.46 0.07 Lu Hf 2.86 0.22 Ta 13.0 Pb 5.06 Th 0.50 U 59.1 REE 27.9 (La/Yb)N 0.87 ␦Eu 15.5 Nb/Ta * a b c d e

C209-8(2) – – – – – – – – – – – – –

778

2 19 5 4.2 17.8 115 392 6.7 145 3.71 5.83 247 18.2 21.4 3.08 10.5 1.85 0.53 1.52 0.21 1.16 0.23 0.55 0.08 0.47 0.08 3.46 0.27 14.5 5.52 0.53 59.9 27.6 0.93 13.9

71.89 0.24 15.45 2.29 – 0.04 0.57 2.52 5.30 1.26 0.07 0.68 100.31 33 1.05 761 888

71.60 0.25 15.49 2.94 – 0.04 0.65 2.61 5.15 1.32 0.09 0.76 100.90 30 1.06 774 910

2 19 5 4.1 17.5 140 391 6.8 122 3.59 5.88 245 18.3 21.8 3.05 10.4 1.86 0.53 1.52 0.22 1.15 0.23 0.55 0.08 0.48 0.07 2.83 0.27 13.8 5.70 0.51 60.2 27.2 0.93 13.1

2 20 7 4.3 17.1 59 306 6.4 142 4.12 5.93 200 18.7 17.6 3.22 11.1 1.98 0.49 1.65 0.23 1.20 0.22 0.53 0.08 0.44 0.07 3.46 0.31 13.0 4.45 0.57 57.6 30.2 0.81 13.3

– – – – – – – – – – – – –

778

1 16 4 4.0 17.1 64 312 6.3 149 4.11 6.23 191 17.3 16.7 3.08 10.5 1.95 0.48 1.58 0.23 1.15 0.22 0.48 0.07 0.42 0.06 3.54 0.37 13.4 3.28 0.51 54.3 29.8 0.81 11.2

C209-8(3)

72.73 0.18 15.20 2.13 – 0.03 0.45 2.57 4.96 1.21 0.07 0.90 100.43 30 1.07 770 896

73.12 0.19 15.23 2.13 – 0.03 0.45 2.57 4.97 1.22 0.11 0.72 100.74 30 1.07 775 945

70.63 0.23 15.65 3.30 – 0.04 0.55 2.74 5.08 1.34 0.09 0.84 100.49 25 1.06 763 900

70.83 0.23 15.72 3.33 – 0.04 0.55 2.76 5.10 1.35 0.09 0.82 100.82 25 1.06 767 902

1 14 6 3.2 16.1 41 298 5.1 130 4.80 4.09 173 15.4 19.5 2.69 9.2 1.59 0.43 1.24 0.17 0.92 0.18 0.44 0.07 0.42 0.06 3.13 0.35 12.2 3.85 0.60 52.3 26.3 0.90 13.8

1 9 4 2.9 16.2 40 302 5.0 138 4.86 4.25 164 13.7 18.5 2.52 8.4 1.53 0.42 1.15 0.16 0.91 0.17 0.40 0.06 0.38 0.06 3.19 0.40 12.7 2.62 0.52 48.4 25.6 0.93 12.2

2 18 4 4.3 20.8 89 376 6.1 126 12.6 6.62 228 12.7 19.7 2.30 7.9 1.47 0.45 1.26 0.19 1.06 0.21 0.52 0.08 0.49 0.08 3.06 0.49 16.1 7.35 1.18 48.4 18.6 0.98 25.4

1 13 1 3.9 21.0 91 380 6.0 133 12.7 6.85 222 10.9 18.7 2.10 7.0 1.34 0.44 1.20 0.18 1.02 0.21 0.49 0.08 0.47 0.07 3.15 0.55 17.1 6.77 1.26 44.2 16.5 1.04 23.1

Average

No. GGU221121a

71.88 0.22 15.42 2.63 – 0.04 0.54 2.61 5.09 1.28 0.09 0.79 100.60 29 1.06 769 910

71.09 0.20 16.45 0.33 1.05 0.03 0.60 2.76 4.93 2.33 0.05 0.46 100.28 50 1.05 – 849

Total Fe as Fe2 O3 , except for no. GGU221121. data from Johnston and Wyllie, (1988). Mg# = cation Mg2+ /(Mg2+ + Fe2+ ). A/CNK = molar Al2 O3 /(CaO + Na2 O + K2 O). TZr , zirconium saturation temperature, calculated using equation of Watson and Harrison, (1983). TAp , apatite saturation temperature, calculated using equation of Harrison and Watson, (1984).

A total of 75 oxygen isotope analyses were measured on 68 grains. All <10% discordant analyses yield ␦18 O ranging from 5.2‰ to 7.0‰, with the exception of spot 33 which gives ␦18 O of 3.9‰. Among them, the nearly-concordant and concordant zircons, including twenty-seven 3.76–3.83 Ga zircons and three 3.31–3.37 Ga zircons show limited variation of ␦18 O from 5.5‰ to 6.5‰. The remaining strongly discordant zircons yield variable ␦18 O values ranging from 1.5‰ to 6.4‰ (Fig. 5A).

5.3.2. Massive trondhjemitic gneiss (Sample C209-8(3)) Zircons are euhedral, 50–100 ␮m in width, and up to 300 ␮m in length, with aspect ratios of 1:1 to 4:1. They are similar in morphology and CL features to zircons of sample C209-8(1) (Figs. A11–A12). Nineteen U–Pb analyses were obtained on 19 grains, and many are discordant. All but two (spots 1 and 6) analyses are variably discordant, yielding 207 Pb/206 Pb ages between 2953 and 3752 Ma. Spots 1 and 6 are concordant analyses,

Y.-F. Wang et al. / Precambrian Research 263 (2015) 88–107

giving 207 Pb/206 Pb age of 3817 ± 19 Ma and 3342 ± 18 Ma, respectively (Fig. 4G). A total of 19 oxygen isotope analyses were performed on 19 grains. Seventeen variably discordant zircons give ␦18 O values ranging from 3.0‰ to 6.5‰ (Fig. 5A), whereas two concordant zircons, spots 1 and 6, yield ␦18 O of 6.4‰ and 6.0‰, respectively. 176 Hf/177 Hf ratios of Unit 3 zircons range from 0.280350 to 0.280837 (Table 1 and Appendix C). Generally, the least discordant (<10%) zircons dated at 3.4–3.8 Ga have εHf (t) between −2.9 and 6.2, with an exception of spot 20 of C209-8(1) (9% disc.) yielding εHf (t) value of −5.8 (Fig. 5C). Four concordant 3.31–3.37 Ga zircons have εHf (t) values between −2.3 and 2.9, among which three 3.34–3.37 Ga zircons of −0.3 to −2.3, and one 3.31 Ga zircon of 2.9 (Fig. 5C). 5.4. Geochemistry of Eoarchean trondhjemitic gneiss Major and trace element compositions of Unit 3 trondhjemitic gneiss, along with the starting composition of relevant experiments are listed in Table 2. This unit belongs to high-Al2 O3 subgroup (Barker et al., 1976) with 15.20–15.72 wt % Al2 O3 , and has high SiO2 (70.6–73.1 wt%), Na2 O (5.0–5.3 wt%), low K2 O (1.2–1.4 wt%) and MgO contents (0.5–0.7 wt%). They are trondhjemites in normative An–Ab–Or classification diagram of Barker (1979) and are peraluminous, with A/CNK (molar Al2 O3 /(CaO + Na2 O + K2 O)) values between 1.05 and 1.07. Three samples have low REE abundances (REE = 44–60 ppm), weak Eu anomalies (␦Eu = 0.8–1.0) and strong REE fractionations, with (La/Yb)N ratios between 17 and 30 (Fig. 6A). On primitive-mantle normalized trace element diagram (Fig. 6B), this unit exhibits overall parallel distribution patterns, including

99

negative Nb, Ta and Ti anomalies and positive Zr, Hf anomalies. It is noticed that this unit has large variation in Nb/Ta ratios between 13 and 25. 6. Discussion 6.1. Geochronological framework at the Guodishan outcrop Zircon U–Pb age, ␦18 O and εHf (t) results of the Guodishan outcrop have been presented in the previous section. Based on these results and field relationships, a geochronological framework can be established. Unit 1 trondhjemitic gneiss, the dominant lithology, crystallized at ca. 3.13 Ga, coeval with the adjacent ca. 3.14 Ga Lishan trondhjemitic gneiss (Song et al., 1996). It is characterized by homogeneous zircon ␦18 O (5.6–6.5‰) and εHf (t) (−0.2 to 1.8) values, indicating that this unit was derived from an basaltic source that had not been affected by significant amount of low temperature alteration. Unit 2 migmatite complex is intruded by Unit 1. The leucosome (sample C209-3) is dated at 3.36 Ga, with ␦18 O of 5.9 ± 0.5‰ and εHf (t) of −3.2 ± 1.7. It was likely generated by partial melting of preexisting crustal rocks. The melansome-biotite schist, however, has a slightly younger metamorphic zircon age of 3.30 Ga and comparable ␦18 O of 6.1 ± 0.8‰ with that of ca. 3.36 Ga leucosome zircons. Unit 3 trondhjemitic gneiss occurs as an enclave within Unit 2. Fifty-eight least-discordant (disc. < 10%) analyses were obtained for two samples, yielding 207 Pb/206 Pb ages between 3.82 and 3.31 Ga (Fig. 4H). Among them, 16 analyses give a concordant age of 3812 ± 8 Ma, identical to their weighted mean 207 Pb/206 Pb age of 3814 ± 2 Ma, which is interpret as the crystallization age of this unit. Zircons with apparent 207 Pb/206 Pb ages of 3.8–3.6 Ga may have been subjected to ancient radiogenic Pb loss (see section 6.4.1). 6.2. Recognition of primary zircon oxygen isotopic signatures

Fig. 6. (A) Chondrite-normalized REE patterns and (B) primitive mantle (PM) normalized trace element patterns for Unit 3 trondhjemitic gneiss. Normalizing values are from Sun and McDonough (1989).

Experimental studies demonstrate that diffusion rates for many elements (e.g., U, Pb, Hf) are very slow within pristine zircons (e.g., Cherniak and Watson, 2003), indicating that pristine zircons may preserve primary elemental signatures (e.g., Valley et al., 1994). However, in U- and Th-rich domains of ancient zircons, metamictization can lead to partial to complete destruction of crystal lattices (e.g., Ewing et al., 2003). These metamict domains are susceptible to alteration and Pb loss, which could obliterate their primary isotopic signatures (e.g., Nasdala et al., 1996; Utsunomiya et al., 2004). Previous studies indicated that zircon Hf isotopes can help to identify whether zircons have experienced recent or ancient radiogenic Pb loss (e.g., Amelin et al., 2000) and that oxygen isotopes are susceptible to modification in metamictization-induced alterations (e.g., Booth et al., 2005; Gao et al., 2014). Therefore, whether primary isotopic signatures have been preserved can be evaluated using the combination of U–Pb, O and Hf isotope systems. For zircons from Unit 1 and Unit 3 trondhjemitic gneisses, U abundance, degree of U–Pb discordance, ␦18 O value, and apparent 207 Pb/206 Pb age are interrelated for a single generation of magmatic zircons (Fig. 7B–E). It is clear in Fig. 7B–E that the higher U abundances, the higher degree of U–Pb discordances, the lower ␦18 O values and the younger apparent 207 Pb/206 Pb ages (Fig. 7A), similar to the relationships described by Booth et al. (2005). For instance, in sample C209-8(1), excluding three concordant 3.31–3.37 Ga zircons, U abundances are correlated inversely with degree of discordance and positively with ␦18 O values (Fig. 7B and C). Meanwhile, the degree of discordances and ␦18 O values also increase and decrease, respectively, with decreasing apparent 207 Pb/206 Pb ages (Fig. 7D and E). Similar patterns also exist for zircons from C209-4, C209-6, and C209-8(3) (figures are not shown).

100

Y.-F. Wang et al. / Precambrian Research 263 (2015) 88–107

Fig. 7. (A) Representative paired analyses of zircons from C209-1, C209-6 and C209-8(1). The CL image in the box shows the notation of analytical results and the scale bars are 50 ␮m in length. (B) Degrees of U–Pb discordance (%) increase with increasing U abundances of Unit 3 trondhjemitic gneiss C209-8(1) zircons. (C) Values of ␦18 O decrease with increasing U abundances of C209-8(1) zircons. (D) Degrees of U–Pb discordance increase with decreasing apparent 207 Pb/206 Pb ages of C209-8(1) zircons. (E) ␦18 O values decrease with decreasing apparent 207 Pb/206 Pb age of C209-8(1). These interrelated parameters of U abundances, degrees of U–Pb discordance, ␦18 O values, apparent 207 Pb/206 Pb ages and measured 176 Hf/177 Hf ratios suggest that only low-U abundances, concordant zircons have the greatest potential to preserve primary O and Hf isotopic compositions. (F) Plot of initial 176 Hf/177 Hf ratios (calculated at 3.81 Ga) versus their measured 207 Pb/206 Pb ages for all <10% discordant zircons within Unit 3. Note younger 207 Pb/206 Pb age zircons yield comparable intial 176 Hf/177 Hf ratios as ca. 3.8 Ga zircons, indicating they were probably derived from ancient radiogenic Pb loss from ca. 3.8 Ga zircons. See text for discussion.

Y.-F. Wang et al. / Precambrian Research 263 (2015) 88–107

Similar relationships also exist in CL-bright and dark bands within single oscillatory zoned crystals, corresponding to U-poor and U-rich domains, respectively (Fig. 7A). Thus radiogenic Pb loss might have occurred in U-rich, CL-dark bands, causing variable discordances and scattered 207 Pb/206 Pb ages. Compositional differences between CL-dark and bright bands can be revealed by examining zircons formed by single magmatic event but with contrasting growth zonings under CL. Analyses 34–1 and 34–2 of C209-1 were conducted on center (CL-bright) and rim (CL-dark) of a single oscillatory zoned crystal (Fig. 7A), respectively. The former is lower in U (309 ppm), Th (174 ppm) concentrations and the degree of discordance (2%), but higher in apparent 207 Pb/206 Pb age (3819 Ma) and ␦18 O (6.3‰), compared with the latter (4923 ppm, 1187 ppm, 78%, 1306 Ma and 2.4‰). Similar patterns also exist in paired analyses 6–1, 6–2 of C209-6 and 25–1, 25–2, 15–1, 15–2 of C209-1 within the same oscillatory zoned zircon (see details in Fig. 7A). These relationships among U abundance, degree of discordance, ␦18 O, apparent 207 Pb/206 Pb age of zircon suggest that only lowU abundance (generally <500 ppm), concordant zircons/domains can potentially preserve primary O isotopes, while U-rich, strongly discordant zircons/domains generally tend towards lower ␦18 O. Therefore, only low-U, concordant to slightly discordant zircons are used for proxy of their primary oxygen isotopic compositions. 6.3. Uncertainties in zircon εHf (t) calculation Factors that could affect εHf (t) of ancient zircons include: (1) the mismatched domains between SIMS U–Pb and LA Hf isotope analyses, particularly for complexly zoned zircons; (2) the erroneous age assignments, especially for zircons experienced ancient radiogenic Pb loss; and (3) the uncertainties of 176 Lu and CHUR parameters. 6.3.1. Comparisons between sequential and simultaneous isotopes analyses The first problem can be circumvented by analyzing simple oscillatory zoned zircons or by simultaneous (U-)Pb–Lu–Hf isotopic analyses (e.g., Yuan et al., 2008; Kemp et al., 2009). Our results demonstrate that concordant to slightly discordant analyses by SIMS within single oscillatory zircon yield comparable 176 Hf/177 Hf as shown by analyses 37–1, 37–2 and 60–1, 60–2 of C209-8(1) (Fig. 7A), indicating primary Hf isotopes could be preserved at low degrees of radiogenic Pb loss. Several paired Hf isotopic analyses on single generation of zircon yield identical 176 Hf/177 Hf ratios, which do not show systematic time-resolved 176 Hf/177 Hf variations (Fig. 8A and B). Therefore, simple-zoned, concordant zircons could potentially preserve primary Hf isotopes. Twenty-one sequential (SIMS U–Pb and LA Lu–Hf) and simultaneous (LA U–Pb–Lu–Hf) isotope analyses on 21 zircons from Unit 3 zircons yield complex results, and broadly two groups were recognized: (1) Group 1 contains 10 grains with each yielding comparable 207 Pb/206 Pb ages and 176 Hf/177 Hf ratios (@1, 10, 23, 39–2, 52, 53, 62, 63, 69 of C209-8(1) and @12 of C209-8(3)) by two approaches. Among them, five grains with concordant U–Pb dates were chosen to calculate the mean of crystallization age of Unit 3; (2) Group 2 includes 11 grains with each yielding contrasting results either in 207 Pb/206 Pb ages or 176 Hf/177 Hf ratios, or both (Table 1). Five concordant and two slightly discordant ∼3.8 Ga zircons in Group 1 yield consistent εHf (t) values of 0.2–6.2 and 1.0–5.0 by sequential and simultaneous approaches, respectively (Table 1). These εHf (t) values, therefore, could best represent the pristine signatures of ∼3.8 Ga zircons of Unit 3. The remaining three analyses yield lower 207 Pb/206 Pb ages but comparable 176 Hf/177 Hf ratios as ∼3.8 Ga zircons, suggesting that they were probably formed at ∼3.8 Ga but experienced ancient radiogenic Pb loss.

101

For Group 2 zircons, both approaches yield complex isotopic results: five grains yield comparable Hf isotopes but contrasting 207 Pb/206 Pb ages with ∼3.8 Ga zircons, indicating that these zircons were probably crystallized at 3.8 Ga and subjected to radiogenic Pb loss. The remaining 6 analyses yield contrasting 176 Hf/177 Hf and/or 207 Pb/206 Pb ages, coupled with their highly discordant nature and CL images, suggesting the ages and Hf isotopes by two approaches are not well consistent, due possibly to Hf isotope heterogeneities, or subsequent modifications. These analyses, therefore, may not preserve pristine Hf isotopes. 6.3.2. Other uncertainties Another uncertainty is erroneous age assignment. Calculations show that for 3.9–3.1 Ga zircons, a 100-Myr error in age shifts εHf (t) by 2.1 to 2.3 units. As shown by analyses 6–1 and 6–2 of C209-6 with comparable 176 Hf/177 Hf but contrasting 207 Pb/206 Pb ages and U–Pb discordances, a ca. 360 Ma age difference results in ca. 8 εHf unit deviation. The uncertainties in CHUR parameters have minor effect, whereas revision of 176 Lu has significant influence on εHf (t) values for ancient samples (e.g., Scherer et al., 2007). Calculations show that using new CHUR parameters (Bouvier et al., 2008) will increase the εHf (t) by 0.4–0.6 units for Eoarchean zircons than by using previous parameters (Blichert-Toft and Albarède, 1997). It holds true that for low Lu/Hf samples, such as zircon, using higher 176 Lu would yield higher εHf (t) values. 6.4. Zircon primary isotopic characters of the 3.81 Ga trondhjemitic gneiss 6.4.1. Interpretation of complex zircon age patterns Complex age spectrums are frequently encountered when performing in situ zircon U–Pb geochronology of Archean orthogneisses. Several explanations have been proposed, including ancient and/or recent Pb loss, metamorphism, multiple generations of magma intrusions, recrystallization and mixed sampling of different age domains (e.g., Compston and Kröner, 1988; Hoskin and Black, 2000; Liu et al., 2008; Nutman et al., 2009; Stern and Bleeker, 1998; Zeh et al., 2008). Therefore, whether zircons with apparent 207 Pb/206 Pb ages between 3.8 and 3.6 Ga within Unit 3 indicate multiple magmatic events or reflect subsequent disturbance of 3.8 Ga zircons needs further investigation. Apparent 207 Pb/206 Pb age variations within Unit 3 zircons are not likely caused by mixed sampling, because analytical spots were targeted on simple oscillatory zones that are generally considered as a single generation of zircon growth. Instead, we attribute the age variations to ancient radiogenic Pb loss of ca. 3.8 Ga zircons, possibly facilitated by contrasting behaviors of high- and low-U domains. This interpretation is supported by Pb loss patterns and Hf isotope studies. If ∼3.8 Ga zircons experienced ancient radiogenic Pb loss, their 207 Pb/206 Pb ages would be younger. Ancient radiogenic Pb loss, therefore, can be identified by examining a group of zircons with identical measured 176 Hf/177 Hf ratios but different apparent 207 Pb/206 Pb ages, especially single oscillatory grain with distinct 207 Pb/206 Pb ages, provided Lu–Hf system remained undisturbed (Amelin et al., 2000). Spots 60–1 and 60–2 of C2098(1) were analyzed on an oscillatory zoned grain, yielding identical 176 Hf/177 Hf ratios, but different 207 Pb/206 Pb ages and U–Pb discordances (Fig. 7A), suggesting that this grain had experienced ancient radiogenic Pb loss. An examination of typical concordant zircons with 207 Pb/206 Pb ages slightly younger than 3.8 Ga is taken for six spots: 7, 34–1, 46, 65, 69 of C209-8(1) and 9 of C209-8(3) (shown as dark green ellipses in Fig. 4H). They give apparent 207 Pb/206 Pb ages between 3647 and 3774 Ma and U–Pb discordance of −1 to 7%, but have measured 176 Hf/177 Hf ratios between 0.280404 and 0.280540, within the range of 0.280375 to 0.280562 for concordant 3.8 Ga

102

Y.-F. Wang et al. / Precambrian Research 263 (2015) 88–107 0.2812

(A)

measured 176Hf/177Hf

0.2810 0.2808

@27-1

0.2806 0.2804 0.2802 0.2800

@27

C209-8(1)@27 : 176Hf/177Hf=0.280533±33 (2σm)

0.2798

C209-8(1)@27-1: 176Hf/177Hf=0.280532±26 (2σm) 0.2796 0

20

40

60

80

100

120

140

160

180

200

0.2811

(B)

measured 176Hf/177Hf

0.2810 0.2809 0.2808 0.2807

@52

0.2806 0.2805 0.2804 0.2803 0.2802

C209-8(1)@52: 176Hf/177Hf=0.280562±20 (2σm)

0.2801 0

20

40

60

80

100

120

140

cycle number (increasing depth in zircon

160

180

200

)

Fig. 8. Time-resolved isotope ratio signals of two representative zircon grains 27 and 52 of trondhjemitic sample C209-8(1). (A) An additional Hf analysis 27-1 (shown in green circle) yield identical 176 Hf/177 Hf with analysis 27 (shown in blue circle), and no systematic change of measured 176 Hf/177 Hf ratios during the analyses, therefore, suggesting that only one age domain of zircon was sampled during Hf isotope analysis. (B) No systematic changes in measured 176 Hf/177 Hf ratios during analysis of 52, indicating one generation of zircon was sampled during Hf isotope analysis. The white ellipses (30 × 20 ␮m) represent U–Pb analytical spots and the large circles (ca. 60 ␮m in diameter) represent Lu–Hf spots. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

zircons (Table 1, Appendix C). In addition, all <10% discordant zircons within Unit 3 exhibit a wide-horizontal array on initial 176 Hf/177 Hf calculated at 3.81 Ga versus 207 Pb/206 Pb plot (Fig. 7F). We, therefore, suggest that the scattered 207 Pb/206 Pb ages are attributable to ancient and/or more recent radiogenic Pb loss of a single generation of ca. 3.8 Ga zircons. With respect to the 3 concordant 3.34–3.37 Ga zircons, they are most likely attributed to injection of surrounding Unit 2 leucosome as strongly indicated by field observations (Fig. 2E and F), and by their similar age, ␦18 O and Hf isotopic compositions to the 3.36 Ga zircons from the leucosome. 6.4.2. Isotope characters of ca. 3.81 Ga trondhjemitic gneiss Fifteen concordant to nearly concordant 3.81 Ga zircons yield ␦18 O between 5.3‰ and 6.6‰, averaging at 6.3 ± 0.7‰, which is considered as the pristine ␦18 O for Unit 3 zircons. These data are broadly consistent with O-isotopes for the low-U concentration and/or “grey CL image” zircons from Anshan (Wan et al., 2013), and with most Hadean and Archean igneous zircons worldwide (5–7.5‰) as well (Hiess et al., 2009; Valley et al., 2005). Therefore, the ca. 3.81 Ga Guodishan trondhjemitic magma was probably formed by partial melting of a basaltic source without an appreciable component of rocks altered by low-temperature aqueous fluids. In contrast to the relatively homogeneous ␦18 O values, these ca. 3.81 Ga zircons have highly variable εHf (t) values ranging from −0.7 to 6.2 (Table 1). It is noteworthy that the εHf (t) value of 6.2 for spot 52 of sample C209-8(1)

records the highest documented value for any ca. 3.8 Ga zircons worldwide. 6.5. Petrogenesis of Eoarchean trondhjemitic gneiss It has been proposed that Archean TTGs are high pressure (generally ≥10 kbar) partial melts of basaltic rocks (e.g., Martin et al., 2005; Moyen and Stevens, 2006). Here, we combined rock geochemistry with available experimental results to constrain the formation conditions of Unit 3 trondhjemite. Our approach is to identify residual assemblages by rock geochemistry, and correlate these residuals with liquidus minerals of primary magmas, since near-liquidus assemblages are interpreted as potential residuals from which primary melt was extracted under the same P–T–H2 O conditions (e.g., Wyllie, 1984). It seems reasonable to assume this unit represents possible primary melts (but see the alternative magma mixing scenario in Section 6.5.2) given its limited major element variations and the absence of appropriate parental magmas in Anshan area. 6.5.1. P–T conditions This unit exhibits moderate Al2 O3 , Sr, strong REE fractionations and weak Eu anomalies, indicating that significant amount of garnet and minor plagioclase occur as residuals. The peraluminous nature suggests that amphibole also present in the residual (see below). Therefore, the geochemical characteristics of Unit 3 prescribe a residual assemblages consisting of significant amount of garnet,

Y.-F. Wang et al. / Precambrian Research 263 (2015) 88–107

amphibole and minor plagioclase from which this trondhjemitic melt was extracted. Dehydration melting experiments on meta-basalts provide some general constraints on TTG genesis, including: (1) the garnet-in boundary locates at ∼10 kbar (e.g., Wyllie et al., 1997; Moyen and Stevens, 2006); (2) the amphibole-out phase boundary, which exerts a strong control on the composition of derived melts, is restricted to 1000–1100 ◦ C (Moyen and Stevens, 2006; Rapp and Watson, 1995; Qian and Hermann, 2013). With increasing temperature, the coexisting melts are peraluminous (A/CNK >1.0) up to amphibole-out, neutral at the boundary (A/CNK ∼1.0), and increasingly metaluminous (A/CNK <1.0) beyond it (Rapp, 1995). Several experiments aimed at establishing phase relations of natural TTGs, among which the starting material of Van der Laan and Wyllie (1992) resembles closely to Unit 3 trondhjemite in major elements and mode mineralogy. Compared with their starting material, Unit 3 trondhjemite has comparable SiO2 , TiO2 , MgO, CaO, Na2 O, but slightly lower Al2 O3 and K2 O contents. Johnston and Wyllie (1988) and Van der Laan and Wyllie (1992) determined hydrous phase relations of Nûk trondhjemite (#GGU 221121) over 8–18 kbar and 700–1200 ◦ C. Water-undersaturated liquidus surface (their Fig. 7) shows that Nûk trondhjemitic melt with 7–8 wt % water coexists with garnet, hornblende, clinopyroxene, and plagioclase at ∼14 kbar, ∼900 ◦ C and that garnet crystallized at pressures over 13 kbar. Apatite is an early crystallized mineral in metaluminous to slightly peraluminous melts, its saturation temperature (Harrison and Watson, 1984) is an estimation of magma temperatures. Apatite saturation temperature for this trondhjemite ranges from 888 ◦ C to 945 ◦ C, averaging at 910 ◦ C, while zirconium saturation thermometer (Watson and Harrison, 1983) yields lower temperatures of 759–778 ◦ C, averaging at 769 ◦ C. Therefore, if the melt composition of this Eoarchean trondhjemite is primary, the most suitable formation P–T–H2 O conditions are 950–1000 ◦ C and 13–15 kbar with ∼6–8 wt % H2 O in the melt, considering slightly compositional differences between Unit 3 and Nûk trondhjemite. 6.5.2. Magma mixing or source heterogeneities? High-field-strengthen elements are sensitive indicators of residuals during TTG formation (e.g., Foley et al., 2002; Hoffmann et al., 2011). Nine analyses on 3 samples-C209-8(1), C209-8(2) and C2098(3), yield overall parallel trace elements distribution patterns but varying Nb/Ta ratios (the former two of 11–16, the latter of 23–25). It is well documented that TTGs have large variations of Nb/Ta (from 5 to over 40) (e.g., Moyen and Stevens, 2006; Hoffmann et al., 2011) and several explanations were proposed, including source heterogeneities (Rapp et al., 2003), continuum of melting conditions (e.g., Moyen and Stevens, 2006) or mixing of melts derived from garnetamphibolite and rutile-bearing eclogite residuals (e.g., Hoffmann et al., 2011). Two scenarios can explain these trace element and zircon initial εHf variations in Unit 3. (1) This unit was formed by mixing of melts from basaltic sources with depleted (εHf (t) ≥ 6.2) and enriched isotopic signatures (εHf (t) ≤ −0.4), respectively, with presence and absence of rutile in the residues phases at different pressures. These zircons were crystallized from this trondhjemitic melt, since such melt could dissolve 500–1200 ppm zirconium at ca. 950–1000 ◦ C calculated using zirconium saturation empirical equation of Watson and Harrison (1983), much higher than its abundances and therefore inherited zircons would dissolve in such melt. In this scenario, the variable zircon εHf initials and Nb/Ta ratios of Unit 3 reflect a binary mixing origin. Alternatively, (2) the basaltic protolith of Unit 3 was heterogeneous in certain trace elements and Hf isotopes. Similar explanation was

103

proposed for zircons with ∼6 εHf units variations from Neoarchean monzogranite (Kemp et al., 2010). Variations of Nb/Ta in this scenario reflect either source characteristics or existence of residual rutile (Xiong et al., 2011), which depends on protolith compositions and melting pressures. Considering the high temperatures of trondhjemitic magma and the efficiency of isotope exchange, we favor the explanation that Nb/Ta ratio and zircon initial εHf variations in Unit 3 reflect more probably a binary mixing origin (scenario 1) than a heterogeneous source (scenario 2), though the latter scenario cannot be entirely ruled out. Either the mixing or heterogeneous origin for Unit 3 trondhjemitic gneiss indicates derivation from a source, or source component with εHf value exceeding 6. 6.6. Implications for Eoarchean ultra-depleted mantle 6.6.1. Eoarchean ultra-depleted mantle inferred from zircon Hf isotopes Pristine isotopic compositions of sixteen 3.81 Ga zircons from Unit 3 can provide new isotopic constraints on the nature of Eoarchean crust and on the degree of early mantle depletion of NCC. These zircons show consistent ␦18 O of 6.3 ± 0.7‰, but variable εHf (t) values ranging from −0.7 to 6.2. One salient feature is that several zircons exhibit highly positive εHf (t) values up to 6.2 (Table 1), which are significantly higher than the modeled depleted mantle value of ca. 2.8 (corresponding to the present-day MORBsource depleted mantle with εHf = 18 and 176 Lu/177 Hf = 0.039, e.g., Vervoort and Blichert-Toft, 1999). To characterize Lu–Hf fractionation in the early Earth, we summarize Pb–Hf isotopes for all <10% discordant zircons obtained during this study, along with several putative reservoirs with different present-day 176 Lu/177 Hf ratios differentiated at 4.5 Ga (Fig. 5C). Most zircons yield εHf (t) between CHUR and MORB-source depleted mantle (MORB-DM), with a few being significantly higher than MORB-DM. This requires the existence of an ultra-depleted Eoarchean mantle source. The highest zircon εHf (t) of 6.2 invokes a reservoir evolved from chondritic initial 176 Hf/177 Hf = 0.279862 starting at 4.5 Ga with 176 Lu/177 Hf of 0.046, much greater than the time-integrated 176 Lu/177 Hf ratio of 0.039 for the present-day MORB-DM. As shown in Fig. 5C, two main magmatic events at ca. 3.81 Ga, ca. 3.13 Ga and a ca. 3.36–3.30 Ga crustal anatexis event can be recognized. It is noteworthy that such ultra-depleted mantle source signature was not registered in younger zircons from this outcrop. This may indicate that either this Eoarchean ultra-depleted mantle source was not sampled by these younger rocks, or more likely, it was no longer existed due to subsequent mantle rehomogenization processes. 6.6.2. Global comparisons and implications for Hadean-Eoarchean magmatism To compare our data with other ancient zircons, we compiled a zircon U–Pb, Lu–Hf dataset of Eoarchean rocks and detrital zircons worldwide (Fig. 9) and recalculated εHf (t) using CHUR and 176 Lu parameters described above. To minimize possible ambiguous interpretations, we restricted the compilation to ≤5% discordant zircons and screened the data in terms of three criteria. (1) For detrital zircons, only concurrent (U)–Pb–Lu–Hf isotopes or laser ablation Hf isotope data with in situ U–Pb ages were adopted, whereas solution Hf isotopic data were not included; (2) For zircons from Eoarchean meta-igneous rocks, both laser ablation and solution Hf isotopes were included, because only small amounts of laser ablation data were available. Inherited and recrystallized zircons as recognized by the original authors are rejected and εHf (t) were calculated using the crystallization ages of each rock unit; (3) For structurally complex zircons from West Greenland and Antarctica

104

Y.-F. Wang et al. / Precambrian Research 263 (2015) 88–107

Eoarchean rock zircons 15

Detrital zircons

This study Acasta Gneisses, Canada (1, 2) Napier Complex, Antarctica (3) West Greenland (1, 4, 5) Anshan Gneiss (6, 7)

10 176

MORB-DM Lu/177Hf = 0.039 176

Lu/ 177H

5

εHf(t)

Other zircons

Antarctica (12) Jack Hills (8) Greenland (5, 12) Jack Hills (9, 10) Slave Craton, Canada (11, 12) Mt. Narryer, Australia (13) Limpopo Belt, Africa (14) North China Craton (15)

f = 0.0

46

0

-5

CHUR reference line

-10

-15 3.5

3.6

3.7

3.8

3.9

4.0

4.1

4.2

4.3

4.4

4.5

Age (Ga) Fig. 9. A worldwide comparison of Pb–Hf isotopes of Eoarchean gneiss zircons and detrital zircons worldwide. The selection criteria are summarized in the text. All εHf (t) were recalculated using CHUR parameters from Blichert-Toft and Albarède (1997) and 176Lu of 1.867 × 10−11 y−1 from Söderlund et al. (2004). Data sourced from: (1) Amelin et al. (2000), (2) Iizuka et al. (2009), (3) Choi et al. (2006), (4) Hiess et al. (2009), (5) Kemp et al. (2009), (6) Wu et al. (2008), (7) Liu et al. (2008), (8) Harrison et al. (2005), (9) Harrison et al. (2008), (10) Kemp et al. (2010), (11) Pietranik et al. (2008), (12) Guitreau et al. (2012), (13) Jacobsen et al. (2010), (14) Zeh et al. (2008), (15) Wu et al. (2005).

orthogneisses, we use 207 Pb/206 Pb ages in the calculation (labeled as ‘Other zircons’ in Fig. 9). The εHf (t) value was plotted against crystallization ages for Eoarchean rocks, and against 207 Pb/206 Pb ages for detrital zircons and ‘Other zircons’. As shown in Fig. 9, most Hadean to Eoarchean detrital and magmatic zircons exhibit typical subchondritic to chondritic Hf isotopic compositions, and few exceptions define a significant bulge in the εHf -time array. Early analyses on Jack Hills detrital zircons (Harrison et al., 2005) gave some highly positive εHf (t) values, which, however, were not observed in more recent analyses (Harrison et al., 2008; Kemp et al., 2010). This discrepancy might be attributed to either biased samplings or more probably mismatched analytical domains. Highly positive εHf (t) values analyzed by solution MCICPMS were reported for Gage Ridge gneiss of Napier Complex, Antarctic (Choi et al., 2006), but these analyses were conducted on highly reverse-discordant zircons with complex recrystallized core-rim structures, making their interpretations less straightforward. Guitreau et al. (2012) suggest on the basis of a global compilation of Hf isotope dataset that the mantle source of continents has remained essentially unchanged with a 176 Lu/177 Hf ratio of 0.032–0.038 over the last 4.3 Gyr. However, we demonstrate that zircon εHf (t) values up to 6.2 for the ca. 3.81 Ga trondhjemitic gneiss at Guodishan in Anshan are reliable, pointing to an Eoarchean ultradepleted mantle domain with time-integrated 176 Lu/177 Hf = 0.046. If this ultra-depleted source was the prevalent upper mantle composition, it would have been a transient feature that could be subsequently re-homogenized possibly by mixing with recycled enriched crust. This scenario was proposed by Bennett et al. (1993) in terms of Nd isotopic studies on West Greenland Eoarchean orthogneisses. Alternatively, this ultra-depleted mantle source could have existed regionally, thus the mantle source of continents could have been highly heterogeneous in the Eoarchean.

Several lines of evidence, such as enriched Hf isotopic signatures of Hadean to Eoarchean zircons, positive 142 Nd anomalies in Eoarchean rock, and the depleted radiogenic isotopes (Nd and Hf) signatures of Archean rocks, all point to the existence of early enriched reservoirs. But the nature, longevity of Hadean enriched reservoir and its potential contributions to Archean magmatisms remain ambiguous (e.g., Bell et al., 2011; Kemp et al., 2010; Pietranik et al., 2008). The presence of Hadean xenocryst within Eoarchean granitic rocks (Iizuka et al., 2006) indicates that Hadean enriched crust was sampled by younger Archean magmatism. Nontheless, the scarcity of markedly unradiogenic Hf isotopes in Eoarchean gneisses appears inconsistent with available Hadean detrital zircon records by exhibiting an apparent shift in initial εHf by 5–7 units at ca. 3.9–3.8 Ga (Fig. 9). This discrepancy was strengthened by the lack of negative 142 Nd anomalies in extant early Archean rock records (see the compilation of Rizo et al., 2013), suggesting their ultimate derivation from the depleted part of the mantle. The change in Lu–Hf systematic with age suggests that the inferred early enriched reservoir was no longer substantially reworked and sampled by Eoarchean rocks worldwide. The mechanism behind this change remains highly speculative and possible scenarios include the surviving enriched Hadean reservoir was small, perhaps due to late heavy bombardment that may prevail on the early Earth, or eroded away as increasing more Hadean zircons were discovered in younger rocks (e.g., Duo et al., 2007; Diwu et al., 2010), or this enriched reservoir returned back to the mantle becoming inaccessible to subsequent samplings. In these cases, the extant Eoarchean rocks were derived from a juvenile mafic crust that replaced the Hadean enriched crust earlier than 3.85–3.8 Ga. Alternatively, the Hadean enriched Hf isotope signatures were possibly diluted by increasing contributions from juvenile magmas (Kemp et al., 2010). Clearly, deciphering mechanisms behind this critical transition requires further investigations on early Archean

Y.-F. Wang et al. / Precambrian Research 263 (2015) 88–107

terrains and particularly Eoarchean rocks to trace down what happened on Earth around this transition. 7. Conclusions (1) The newly-discovered ca. 3.81 Ga trondhjemitic gneiss corroborates the existence of 3.8 Ga rock fragments in Anshan Complex. It occurs as enclaves hosted by ca. 3.36–3.30 Ga migmatite complex, which in turn was intruded by ca. 3.13 Ga trondhjemitic gneiss. Variably younger zircon ages of the 3.81 Ga trondhjemitic gneiss are attributed mainly to radiogenic Pb loss and minor injection of 3.36 Ga leucosome veins. (2) Concordant pristine domains of zircons with low U- and Th-concentrations have the greatest potential in preserving primary oxygen isotopic features. In contrast, U- and Th-rich, strongly discordant domains fail to retain primary isotopic signatures, possibly due to metamictization and subsequent hydrothermal alterations. They tend to have lower ␦18 O. (3) The ca. 3.81 Ga concordant zircons preserve primary isotopic signatures of the trondhjemitic gneiss, with εHf (t) of −0.7 to 6.2 and ␦18 O of 5.3–7.0‰, suggesting its derivation from a basaltic protolith characterized by heterogeneous Hf isotopes without appreciable involvement of supracrustal rocks. The highest εHf (t) of 6.2 exceeds that reported for contemporaneous zircons and the modeled depleted mantle source, providing new isotopic evidence for the existence of an Eoarchean ultra-depleted mantle domain underlying NCC. (4) Comparisons of Hf isotopes between our data and Eoarchean magmatic and detrital zircons worldwide suggest that either this Eoarchean ultra-depleted mantle was either a widespread transient feature that was re-homogenized subsequently, or was a regional signature pointing to a highly heterogeneous mantle source in the Eoarchean.

Acknowledgements We thank Yu Liu and Guo-Qiang Tang for their help in SIMS U–Pb and O isotope analyses, Yue-Heng Yang for Hf isotope analyses and Prof. Hong-Lin Yuan for simultaneous U–Pb and Lu–Hf isotope analyses. Professors Fu-Yuan Wu and Qiu-Li Li are thanked for their helpful suggestions and discussions. We thank Prof. Robert P. Wintsch, associate professor Robin Offler and Hui-Qing Huang for proofreading the manuscript. Constructive comments from Allan Nutman, Tony Kemp and editor Randall Parrish are greatly appreciated. This research was supported by the National Natural Science Foundation of China (Grant 41221022), and the State Key Laboratory of Lithospheric Evolution, IGGCAS (Grant Z1005). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.precamres. 2015.03.005. References Amelin, Y., Lee, D.C., Halliday, A.N., Pidgeon, R.T., 1999. Nature of the Earth’s earliest crust from hafnium isotopes in single detrital zircons. Nature 399, 252–255. Amelin, Y., Lee, D.-C., Halliday, A.N., 2000. Early-middle Archaean crustal evolution deduced from Lu–Hf and U–Pb isotopic studies of single zircon grains. Geochim. Cosmochim. Acta 64, 4205–4225. Barker, F., Arth, J.G., Peterman, Z.E., Friedman, I., 1976. The 1.7- to 1.8-b.y.-old trondhjemites of southwestern Colorado and northern New Mexico: geochemistry and depths of genesis. Geol. Soc. Am. Bull. 87, 189–198. Barker, F., 1979. Trondhjemite: definition, environment and hypotheses of origin. In: Barker, F. (Ed.), Trondhjemites, Dacites and Related Rocks. Elsevier, Amsterdam, pp. 1–12.

105

Bell, E.A., Harrison, T.M., McCulloch, M.T., Young, E.D., 2011. Early Archean crustal evolution of the Jack Hills Zircon source terrane inferred from Lu–Hf, 207 Pb/206 Pb, and ␦18 O systematics of Jack Hills zircons. Geochim. Cosmochim. Acta 75, 4816–4829. Bennett, V.C., Nutman, A.P., McCulloch, M.T., 1993. Nd isotopic evidence for transient, highly depleted mantle reservoirs in the early history of the Earth. Earth Planet. Sci. Lett. 119, 299–317. Blichert-Toft, J., Albarède, F., 1997. The Lu–Hf isotope geochemistry of chondrites and the evolution of the mantle-crust system. Earth Planet. Sci. Lett. 148, 243–258. Blichert-Toft, J., Albarède, F., 2008. Hafnium isotopes in Jack Hills zircons and the formation of the Hadean crust. Earth Planet. Sci. Lett. 265, 686–702. Booth, A.L., Kolondny, Y., Chamberlain, C.P., McWilliams, M., Schmitt, A.K., Wooden, J., 2005. Oxygen isotopic composition and U–Pb discordance in zircon. Geochim. Cosmochim. Acta 69, 4895–4905. Bouvier, A., Vervoort, J.D., Patchett, P.J., 2008. The Lu–Hf and Sm–Nd isotopic composition of CHUR: constraints from unequilibrated chondrites and implications for the bulk composition of terrestrial planets. Earth Planet. Sci. Lett. 273, 48–57. Cherniak, D.J., Watson, E.B., 2003. Diffusion in zircon. Rev. Mineral. Geochem. 53, 113–143. Choi, S., Mukasa, S., Andronikov, A., Osanai, Y., Harley, S., Kelly, N., 2006. Lu–Hf systematics of the ultra-high temperature Napier Metamorphic Complex in Antarctica: Evidence for the early Archean differentiation of Earth’s mantle. Earth Planet. Sci. Lett. 246, 305–316. Compston, W., Kröner, A., 1988. Multiple zircon growth within early Archaean tonalitic gneiss from the Ancient Gneiss Complex, Swaziland. Earth Planet. Sci. Lett. 87, 113–128. Diwu, C., Sun, Y., Dong, Z., Wang, H., Chen, D., Chen, L., Zhang, H., 2010. In situ U–Pb geochronology of Hadean zircon xenocryst (4.1–3.9 Ga) from the western of the Northern Qingling Orogenic Belt. Acta Petrol. Sin. 26, 1171–1174. Duo, J., Wen, C., Guo, J., Fan, X., Li, X., 2007. 4.1 Ga old detrital zircon in western Tibet of China. Chin. Sci. Bull. 52, 23–36. Ewing, R., Meldrum, A., Wang, L., Weber, W.J., Corralem, L.R., 2003. Radiation effects in zircon. Rev. Mineral. Geochem. 53, 387–425. Foley, S., Tiepolo, M., Vannucci, R., 2002. Growth of early continental crust controlled by melting of amphibolite in subduction zones. Nature 417, 837–840. Gao, Y.Y., Li, X.H., Griffin, W.L., O’Reilly, S.Y., Wang, Y.F., 2014. Screening criteria for reliable U–Pb geochronology and oxygen isotope analysis in uranium-rich zircons: a case study from the Suzhou A-type granites, SE China. Lithos 192-195, 180–191. Gruau, G., Rosing, M., Bridgwater, D., Gill, R.C.O., 1996. Resetting of Sm–Nd systematics during metamorphism of >3.7-Ga rocks: implications for isotopic models of early Earth differentiation. Chem. Geol. 133, 225–240. Guitreau, M., Blichert-Toft, J., Martin, H., Mojzsis, S.J., Albarède, F., 2012. Hafnium isotope evidence from Archean granitic rocks for deep-mantle origin of continental crust. Earth Planet. Sci. Lett. 337–338, 211–223. Harrison, T.M., Blichert-Toft, J., Muller, W., Albarede, F., Holden, P., Mojzsis, S.J., 2005. Heterogeneous Hadean hafnium: evidence of continental crust at 4.4 to 4.5 Ga. Science 310, 1947–1950. Harrison, T.M., Schmitt, A.K., McCulloch, M.T., Lovera, O.M., 2008. Early (≥4.5 Ga) formation of terrestrial crust: Lu–Hf, ␦18 O, and Ti thermometry results for Hadean zircons. Earth Planet. Sci. Lett. 268, 476–486. Harrison, T.M., Watson, E.B., 1984. The behavior of apatite during crustal anatexis: equilibrium and kinetic considerations. Geochim. Cosmochim. Acta 48, 1467–1477. Hiess, J., Bennett, V.C., Nutman, A.P., Williams, I.S., 2009. In situ U–Pb, O and Hf isotopic compositions of zircon and olivine from Eoarchean rocks, West Greenland: new insights to making old crust. Geochim. Cosmochim. Acta 73, 4489–4516. Hoffmann, J.E., Münker, C., Næraa, T., Rosing, M.T., Herwartz, D., Garbe-Schönberg, D., Svahnverg, H., 2011. Mechanisms of Archean crust formation inferred from high-precision HFSE systematics in TTGs. Geochim. Cosmochim. Acta 75, 4157–4178. Hoffmann, J.E., Münker, C., Polat, A., König, S., Mezger, K., Rosing, M.T., 2010. Highly depleted Hadean mantle reservoirs in the sources of early Archean arc-like rocks, Isua supracrustal belt, southern West Greenland. Geochim. Cosmochim. Acta 74, 7236–7260. Hoskin, P.W.O., Black, L.P., 2000. Metamorphic zircon formation by solidstate recrystallization of protolith igneous zircon. J. Metamorph. Geol. 18, 423–439. Iizuka, T., Horie, K., Komiya, T., Maruyama, S., Hirata, T., Hidaka, H., Windley, B.F., 2006. 4.2 Ga zircon xenocryst in an Acasta gneiss from northwestern Canada: evidence for early continental crust. Geology 34, 245–248. Iizuka, T., Komiya, T., Hohnson, S.P., Kon, Y., Maruyama, S., Hirata, Y., 2009. Reworking of Hadean crust in the Acasta gneisses, northwestern Canada: evidence from in-situ Lu–Hf isotope analysis of zircon. Chem. Geol. 259, 230–239. Jacobsen, Y.N., Münker, C., Nebel, O., Gredes, A., Mezger, K., Nelson, D.R., 2010. Reworking of Earth’s first crust: Constraints from Hf isotopes in Archean zircons from Mt. Narryer, Australia. Precambrian Res. 182, 175–186. Johnston, A.D., Wyllie, P.J., 1988. Constraints on the origin of Archean trondhjemites based on phase relationships of Nûk gneiss with H2 O at 15 kbar. Contrib. Mineral. Petrol. 100, 35–46. Kamber, B.S., Moorbath, S., Whitehouse, M.J., 2001. The oldest rocks on Earth: time constraints and geological controversies. Geol. Soc. London Spec. Publ. 190, 177–203.

106

Y.-F. Wang et al. / Precambrian Research 263 (2015) 88–107

Kemp, A.I.S., Foster, G.L., Scherstén, A., Whitehouse, M.J., Darling, J., Storey, C., 2009. Concurrent Pb–Hf isotope analysis of zircon by laser ablation multi-collector ICPMS with implications for the crustal evolution of Greenland and the Himalayas. Chem. Geol. 261, 244–260. Kemp, A.I.S., Wilde, S.A., Hawkesworth, C.J., Coath, C.D., Nemchin, A., Pidgeon, R.T., Vervoort, J.D., DuFrane, S.A., 2010. Hadean crustal evolution revisited: new constraints from Pb–Hf isotope systematics of the Jack Hills zircons. Earth Planet. Sci. Lett. 296, 45–56. Li, Q.L., Li, X.H., Liu, Y., Tang, G.Q., Yang, J.H., Zhu, W.G., 2010a. Precise U–Pb and Pb–Pb dating of Phanerozoic baddeleyite by SIMS with oxygen flooding technique. J. Anal. At. Spectrom. 25, 1107–1113. Li, X.H., Liu, Y., Li, Q.L., Guo, C.H., Chamberlain, K.R., 2009. Precise determination of Phanerozoic zircon Pb/Pb age by multicollector SIMS without external standardization. Geochem. Geophys. Geosyst. 10, Q04010, http://dx.doi.org/10.1029/2009GC002400. Li, X.H., Li, W.X., Li, Q.L., Wang, X.C., Liu, Y., Yang, Y.H., 2010b. Petrogenesis and tectonic significance of the ∼850 Ma Gangbian alkaline complex in South China: evidence from in situ zircon U–Pb dating, Hf–O isotopes and whole-rock geochemistry. Lithos 114, 1–15. Li, X.H., Long, W.G., Li, Q.L., Liu, Y., Zheng, Y.F., Yang, Y.H., Chamberlain, K.R., Wang, D.F., Guo, C.H., Wang, X.C., 2010c. Penglai zircon megacrysts: a potential new working reference material for microbeam determination of Hf–O isotope and U–Pb age. Geostand. Geoanal. Res. 34, 117–134. Li, X.H., Tang, G.Q., Gong, B., Yang, Y.H., Hou, K.J., Hu, Z.C., Li, Q.L., Li, W.X., 2013. Qinghu zircon: a working reference for microbeam analysis of U–Pb age and Hf and O isotopes. Chin. Sci. Bull. 58, 4647–4654. Liu, D., Wilde, S.A., Wan, Y., Wu, J., Zhou, H., Dong, C., Yin, X., 2008. New U–Pb and Hf isotopic data confirm Anshan as the oldest preserved segment of the North China Craton. Am. J. Sci. 308, 200–231. Liu, D.Y., Nutman, A.P., Compston, W., Wu, J.S., Shen, Q.H., 1992. Remnants of ≥3800 Ma crust in the Chinese part of the Sino-Korean craton. Geology 20, 339–342. Ludwig, K.R., 2008. Users’ manual for Isoplot 3.70: a geochronological toolkit for Mircrosoft Excel. In: Berkeley Geochronology Center Special Publication No. 4. Berkeley, California, pp. 77. Machado, N., Simonetti, A., 2001. U–Pb dating and Hf isotopic composition of zircon by laser-ablation ICPMS. In: Sylvester, P. (Ed.), Laser Ablation-ICPMS in the Earth sciences: Principles and applications. Mineralogical Association of Canada, St. John’s Newfoundland, pp. 121–146. Martin, H., Smithies, R.H., Rapp, R., Moyen, J.-F., Champion, D., 2005. An overview of adakite, tonalite–trondhjemite–granodiorite (TTG) and sanukitoid: relationships and some implications for crustal evolution. Lithos 79, 1–24. Moorbath, S., Whitehouse, M.J., Kamber, B.S., 1997. Extreme Nd-isotope heterogeneity in the early Archaean-fact or fiction? Case histories from northern Canada and West Greenland. Chem. Geol. 135, 213–231. Morel, M.L.A., Nebel, O., Jacobsen, Y.J.N., Miller, J.S., Vroon, P.Z., 2008. Hafnium isotope characterization of the GJ-1 zircon reference material by solution and laser-ablation MC-ICPMS. Chem. Geol. 255, 231–235. Moyen, J.-F., Stevens, G., 2006. Experimental constraints on TTG petrogenesis: implications for Archean geodynamics. In Benn, K., Condie, K.C., Mareschal, J.C. (Eds.), Archaean geodynamics and environments. Geoph. Monog. Ser. 164, 149–175. Nasdala, L., Pidgeon, R.T., Wolf, D., 1996. Heterogeneous metamictization of zircon on a microscale. Geochim. Cosmochim. Acta 60, 1091–1097. Nebel-Jacobsen, Y., Münker, C., Nebel, O., Gerdes, A., Mezger, K., Nelson, D.R., 2010. Reworking of Earth’s first crust: constraints from Hf isotopes in Archean zircons from Mt. Narryer, Australia. Precambrian Res. 182, 175–186. Nutman, A.P., Friend, C.R.L., Bennett, V.C., 2001. Review of the oldest (4400–3600 Ma) geological and mineralogical record: glimpses of the beginning. Episodes 24, 93–101. Nutman, A.P., Wan, Y., Liu, D., 2009. Integrated field geological and zircon morphology evidence for ca. 3.8 Ga rocks at Anshan: Comment on Zircon U–Pb and Hf isotopic constraints on the Early Archean crustal evolution in Anshan of the North China Craton by Wu et al. [Precambrian Res. 167 (2008) 339–362]. Precambrian Res. 172, 357–360. Pietranik, A.B., Hawkesworth, C.J., Storey, C.D., Kemp, A.I.S., Sircombe, K.N., Whitehouse, M.J., Bleeker, W., 2008. Episodic, mafic crust formation from 4.5 to 2.8 Ga: new evidence from detrital zircons, Slave craton, Canada. Geology 36, 875–878. Qian, Q., Hermann, J., 2013. Partial melting of lower crust at 10–15 kbar: constraints on adakite and TTG formation. Contrib. Mineral. Petrol. 165, 1195–1224. Rapp, R.P., 1995. Amphibole-out boundary in partially melted metabasalt, its control over liquid fraction and composition, and source permeability. J Geophys. Res. 100, 15601–15610. Rapp, R.P., Shimizu, N., Norman, M.D., 2003. Growth of early continental crust by partial melting of eclogite. Nature 425, 605–609. Rapp, R.P., Watson, E.B., 1995. Dehydration melting of metabasalt at 8–32 kbar: Implications for continental growth and crust-mantle recycling. J. Petrol. 36, 891–931. Rizo, H., Boyet, M., Blichert-Toft, J., Rosing, M.T., 2013. Early mantle dynamics inferred from 142 Nd variations in Archean rocks from southwest Greenland. Earth Planet. Sci. Lett. 377–378, 324–335. Scherer, E.E., Whitehouse, M.J., Munker, C., 2007. Zircon as a monitor of crustal growth. Elements 3, 19–24. Sláma, J., Koˇsler, J., Condon, D.J., Crowley, J.L., Gerdes, A., Hanchar, J.M., Horstwood, M.S.A., Morris, G.A., Nasdala, L., Norberg, N., Schaltegger, U., Schoene, B., Tubrett,

M.N., Whitehouse, M.J., 2008. Pleˇsovice zircon—a new natural reference material for U–Pb and Hf isotopic microanalysis. Chem. Geol. 249, 1–35. Söderlund, U., Patchett, P.J., Vervoort, J.D., Isachsen, C.E., 2004. The 176 Lu decay constant determined by Lu–Hf and U–Pb isotope systematics of Precambrian mafic intrusions. Earth Planet. Sci. Lett. 219, 311–324. Song, B., Nutman, A.P., Liu, D., Wu, J., 1996. 3800 to 2500 Ma crustal evolution in the Anshan area of Liaoning Province, Northeastern China. Precambrian Res. 78, 79–94. Stacey, J.S., Kramer, J., 1975. Approximation of terrestrial lead isotope evolution by a two-stage model. Earth Planet. Sci. Lett. 26, 207–221. Stern, R.A., Bleeker, W., 1998. Age of the world’s oldest rocks refined using Canada’s SHRIMP: the Acasta Gneiss Complex, Northwest Territories, Canada. Geosci. Can. 25, 27–31. Sun, S.S., McDonough, W.F., 1989. Chemical and isotopic systematics of oceanic basalts: implications for mantle composition and process. In: Saunders, A.D., Norry, M.J. (Eds.), Magmatism in the Ocean Basins, vol. 42. Geol. Soc. Spec. Publ., Bath, UK, pp. 313–345. Utsunomiya, S., Palenik, C.S., Valley, J.W., Cavosie, A.J., Wilde, S.A., Ewing, R.C., 2004. Nanoscale occurrence of Pb in an Archean zircon. Geochim. Cosmochim. Acta 68, 4679–4686. Valley, J.W., Cavosie, A.J., Fu, B., Peck, W.H., Wilde, S.A., 2006. Comment on heterogeneous hadean hafnium: evidence of continental crust at 4.4 to 4.5 Ga. Science 26, 1139–11139. Valley, J.W., Chiarenzelli, J.R., McLelland, J.M., 1994. Oxygen isotope geochemistry of zircon. Earth Planet. Sci. Lett. 126, 187–206. Valley, J.W., Lackey, J.S., Cavosie, A.J., Clechenko, C.C., Spicuzza, M.J., Basei, M.A.S., Bindeman, I.N., Ferreira, V.P., Sial, A.N., King, E.M., Peck, W.H., Sinha, A.K., Wei, C.S., 2005. 4.4 Billion years of crustal maturation: oxygen isotope ratios of magmatic zircon. Contrib. Mineral. Petrol. 150, 561–580. Van der Laan, S.R., Wyllie, P.J., 1992. Constraints on Archean trondhjemite genesis from hydrous crystallization experiments on Nûk gneiss at 10–17 kbar. J. Geol. 100, 57–68. Vervoort, J.D., Blichert-Toft, J., 1999. Evolution of the depleted mantle: Hf isotope evidence from juvenile rocks through time. Geochim. Cosmochim. Acta 63, 533–556. Vervoort, J.D., Patchett, P.J., Gehrels, G.E., Nutman, A.P., 1996. Constraints on early Earth differentiation from hafnium and neodymium isotopes. Nature 379, 624–627. Wan, Y., Liu, D., Nutman, A., Zhou, H., Dong, C., Yin, X., Ma, M., 2012. Multiple 3.8–3.1 Ga tectono-magmatic events in a newly discovered area of ancient rocks (the Shengousi Complex), Anshan, North China Craton. J. Asian Earth. Sci. 54–55, 18–30. Wan, Y., Liu, D., Song, B., Wu, J., Yang, C., Zhang, Z., Geng, Y., 2005. Geochemical and Nd isotopic compositions of 3.8 Ga meta-quartz dioritic and trondhjemitic rocks from the Anshan area and their geological significance. J. Asian Earth Sci. 24, 563–575. Wan, Y., Zhang, Y., Williams, I.S., Liu, D., Dong, C., Fan, R., Shi, Y., Ma, M., 2013. Extreme zircon O isotopic compositions from 3.8 to 2.5 Ga magmatic rocks from the Anshan area, North China Craton. Chem. Geol. 352, 108–124. Watson, E.B., Harrison, T.M., 1983. Zircon saturation revisited: temperature and composition effects in a variety of crustal magma types. Earth Planet. Sci. Lett. 64, 295–304. Wiedenbeck, M., Alle, P., Corfu, F., Griffin, W.L., Meier, M., Oberli, F., Vonquadt, A., Roddick, J.C., Speigel, W., 1995. Three natural zircon standards for U–Th–Pb, Lu–Hf, trace-element and REE analyses. Geostand. Newsl. 19, 1–23. Woodhead, J.D., Hergt, J.M., 2005. A preliminary appraisal of seven natural zircon reference materials for in situ Hf isotope determination. Geostand. Geoanal. Res. 29, 183–195. Woodhead, J., Hergt, J., Shelley, M., Eggins, S., Kemp, R., 2004. Zircon Hf-isotope analysis with an excimer laser, depth profiling, ablation of complex geometries, and concomitant age estimation. Chem. Geol. 209, 121–135. Wu, F.Y., Yang, Y.H., Xie, L.W., Yang, J.H., Xu, P., 2006. Hf isotopic compositions of the standard zircons and baddeleyites used in U–Pb geochronology. Chem. Geol. 234, 105–126. Wu, F.Y., Zhang, Y.B., Yang, J.H., Xie, L.W., Yang, Y.H., 2008. Zircon U–Pb and Hf isotopic constraints on the Early Archean crustal evolution in Anshan of the North China Craton. Precambrian Res. 167, 339–362. Wu, F.Y., Zhang, Y.B., Yang, J.H., Xie, L.W., Yang, Y.H., 2009. Are there any 3.8 Ga rock at Anshan in the North China Craton? Reply to comments on Zircon U–Pb and Hf isotopic constraints on the Early Archean crustal evolution in Anshan of the North China Craton by Nutman et al. Precambrian Res. 172, 361–363. Wu, F., Yang, J., Liu, X., Li, T., Xie, L., Yang, Y., 2005. Hf isotopes of the 3.8 Ga zircons in eastern Hebei Province, China: implications for early crustal evolution of the North China Craton. Chin. Sci. Bull. 50, 2473–2480. Wu, J.S., Geng, Y.S., Shen, Q.H., Liu, D.Y., Song, B., 1998. Archean Geology Characteristics and Tectonic Evolution of Sino-Korean Paleo-Continent. Geological Publishing House, Beijing, pp. 1–36 (in Chinese). Wyllie, P.J., 1984. Constraints imposed by experimental petrology on possible and impossible magma sources and products. Philos. Trans. R. Soc. London, Ser. A 310, 439–456. Wyllie, P.J., Wolf, M.B., van der Laan, S.R., 1997. Conditions for formation of tonalites and trondhjmeites: magmatic sources and products. In: De Wit, M.J., Ashwal, L.D. (Eds.), Greenstone Belts. Clarendon, Oxford, pp. 256–266. Xiong, X., Keppler, H., Audetat, A., Ni, H., Sun, W., Li, Y., 2011. Partitioning of Nb and Ta between rutile and felsic melt and the fractionation of Nb/Ta during partial melting of hydrous metabasatl. Geochim. Cosmochim. Acta 75, 1673–1692.

Y.-F. Wang et al. / Precambrian Research 263 (2015) 88–107 Yuan, H.L., Gao, S., Dai, M.N., Zong, C.L., Günther, D., Fontaine, G.H., Liu, X.M., Diwu, C., 2008. Simulataneous determinations of U–Pb age, Hf isotopes and trace element compositions of zircon by excimer laser-ablation quadrupole and multiplecollector ICP-MS. Chem. Geol. 247, 100–118. Zeh, A., Gerdes, A., Klemd, R., Barton Jr., J.M., 2008. U–Pb and Lu–Hf isotope record of detrital zircon grains from the Limpopo Belt—evidence for crustal

107

recycling at the Hadean to early-Archean transition. Geochim. Cosmochim. Acta 72, 5304–5329. Zhang, J.H., Jin, W., Zheng, P.X., Wang, Y.F., Li, B., Cai, L.B., Wang, Q.L., 2013. Identification and zircon U–Pb geochronology of the Yingchengzi Paleoarchean gneiss complex, Anshan area. Acta Petrol. Sin. 29, 399–413.